Comprehensive Energy Systems, vol.2 - Energy Materials [2, 1 ed.] 978-0-12-814925-6

Citation preview

2.1 Ammonia Ibrahim Dincer and Yusuf Bicer, University of Ontario Institute of Technology, Oshawa, ON, Canada r 2018 Elsevier Inc. All rights reserved.

2.1.1 Introduction 2.1.1.1 Ammonia Production and Transport 2.1.1.2 Ammonia Storage 2.1.1.3 Ammonia Utilization 2.1.1.3.1 Ammonia in Heating, Ventilation and Air Conditioning applications 2.1.1.3.2 Ammonia as both fuel and refrigerant 2.1.1.3.2.1 Thermo-catalytic decomposition of ammonia 2.1.1.3.3 Ammonia and urea 2.1.2 Analysis and Assessment 2.1.2.1 Ammonia in Direct Ammonia Solid Oxide Fuel Cell 2.1.2.2 Urea in Direct Urea Solid Oxide Fuel Cell 2.1.2.3 Utilization of Fuel and Oxidant 2.1.2.4 Electrochemical Analysis 2.1.2.4.1 Activation overpotential 2.1.2.4.2 Ohmic overpotential 2.1.2.4.3 Concentration overpotential 2.1.2.4.4 Electrochemical efficiency 2.1.2.5 Illustrative Example 2.1.3 Case Study 1: Ammonia Utilization in Road Vehicles 2.1.3.1 Materials and Methods 2.1.3.2 Systems Description 2.1.3.2.1 Passenger car manufacturing 2.1.3.2.2 Maintenance 2.1.3.2.3 Disposal of the vehicles 2.1.3.2.4 Operation of vehicles 2.1.3.3 Case Study 1 Results and Discussion 2.1.3.4 Case Study 1 Conclusions 2.1.4 Case Study 2: Ammonia Utilization in Maritime Applications 2.1.4.1 Introduction 2.1.4.2 Pathways for Production of Ammonia 2.1.4.3 Material and Methods 2.1.4.4 Life Cycle Assessment Phases 2.1.4.4.1 Transoceanic freight ship manufacture 2.1.4.4.2 Maintenance of transoceanic freight ship 2.1.4.4.3 Port facilities 2.1.4.4.4 Maintenance and operation of the port 2.1.4.4.5 Operation of the transoceanic freight ship and transport for 1 tkm 2.1.4.5 Case Study 2 Results and Discussion 2.1.4.6 Case Study 2 Conclusions 2.1.5 Closing Remarks Acknowledgment References Further Reading Relevant Websites

Nomenclature Bo Deff i;j Deff i;k Uf Uo ∆H

Permeability of the porous electrode Effective binary diffusion Effective Knudsen diffusion Utilization of fuel Utilization of air Enthalpy change (kJ kg 1)

Comprehensive Energy Systems, Volume 2

E F j N P R T

doi:10.1016/B978-0-12-809597-3.00201-7

2 5 6 8 11 12 13 14 14 14 15 15 15 15 16 16 17 17 20 20 22 22 23 23 23 24 28 29 29 29 30 31 31 31 31 31 32 32 35 36 37 37 38 39

Open-circuit voltage (V) Faraday constant Current density Molar flux Pressure (bar) Gas constant Temperature (K)

1

2

Ammonia

Working potential or voltage (V) Molar fraction of chemical species

z

Electron number

LPG MSWI MTPM Ni-YSZ NM NMI OCV OEM PAH PTH PV RCV RSZ SMR SOFC SOFC-H

SUR TKM TPB VOC WTH WTP YSZ

Liquefied petroleum gas Municipal waste incineration plant Mean transport pore model Nickel-yittria stabilized zirconia Non-methane Nautical mile Open cell voltage Original equipment manufacturer Polycyclic aromatic hydrocarbon Pump to haul Photovoltaic Refuse collection vehicles Reduced-speed zones Steam methane reforming Solid oxide fuel cell Solid oxide fuel cell with hydrogen proton (H þ ) conducting electrolyte Solid oxide fuel cell with oxygen ion (O2 ) conducting electrolyte Standard temperature of 298 K and pressure of 101.325 Kpa Steam-to-urea ratio Tonne-kilometer Triple phase boundary Volatile organic compounds Well to haul Wheel to pump Yittria-stabilized zirconia

Greek Letters a Charge transfer coefficient G Dimensionless temperature δ Electrolyte thickness (M) D Net change of quantity e Porosity of electrode e/κ Lennard-Jones temperature parameter (K) r Density (kg m 3) sij Mean characteristic length

t jact,an jact,ca jconc,an jconc,ca jO OD,ij Z m

Tortuosity of electrode Anode activation overpotential (V) Cathode activation overpotential (V) Anode concentration overpotential (V) Cathode concentration overpotential (V) Ohmic overpotential of electrolyte (V) Collision integral Efficiency Viscosity

Subscripts Act An Ca Conc Elc

f i Mix o r rev

Fuel Chemical species Mixture Oxidant Reactants Reversible

V y

Acronyms ADF CFC CML DA DB DeNOx DGM DU DWT EV FFE GEM GHG GREET GWP HEV HFC HHV HVAC ICE IPCC ISO KN LCA LHV

2.1.1

Abiotic depletion factor Chlorofluorocarbon Center of Environmental Science of Leiden University Direct ammonia Dichlorobenzene Destruction of nitrous oxides Dusty-gas model Direct urea Dead weight tonnage Electric vehicle Feed plus fuel energy Gibbs energy minimization Greenhouse gas Greenhouse gases, regulated emissions, and energy use in transportation Global warming potential Hybrid electric vehicle Hydrofluorocarbon Higher heating value Heating, ventilation and air conditioning Internal combustion engines Intergovernmental panel on climate change International organization for standards Knot Life cycle assessment Lower heating value

and superscripts Activation Anode Cathode Concentration Electrochemical

SOFC-O STP

Introduction

Inevitable issues of fossil fuels and their extensive use in many sectors, including transportation domain, have resulted in damaging effects on human health and well being as well as the environment. Thus, there is an urgent need to develop and

Ammonia

3

implement some efficient, effective and environmentally benign potential replacements. The relevant studies suggest switching from the fossil fuels to renewable energy sources and carbon-free fuels, such as hydrogen and ammonia [1]. In this respect, there is a critical need for main investments for the growth of competitive hydrogen manufacture, delivery, and the particularly storing technologies that are not commercially fully viable and well developed. There is a need to further study and develop viable technologies for mass implementations [2]. Therefore, there is a drawback regarding hydrogen that hydrogen delivers very little quantities of energy per unit volume as compared to the conventional fuels used in various transportation fuels. On top of this, the growth of hydrogen supply substructure (essentially the infrastructure required) will require some specific safety measures undertaken particularly on its flammability and explosivity risks since hydrogen is recognized as volatile energy carrier with the invisible flame [3]. Ammonia is expected to overcome the above mentioned hydrogen related challenges by offering a potential solution to reduce these challenges, and projected to be a potential hydrogen carrier with high hydrogen content in the near future. In recent years, expectations are rising for hydrogen and hydrogen carriers as a medium for storage and transportation of energy in the mass introduction and use of renewables. Both storage and transport of hydrogen are considered an important issue since hydrogen is in gas form under normal temperature and pressure. Hydrogen carriers are mediums that convert hydrogen into chemical substances containing large amounts of hydrogen, to simplify storage and transport processes. Hydrogen carriers include ammonia synthesized from nitrogen and hydrogen that can be used for direct combustion, cooling, fuel cells, power generation, etc. Ammonia becomes a primary hydrogen carrier that does not contain any carbon atoms and has a high hydrogen ratio. Therefore, it is evaluated as a clean power-generating fuel. Since ammonia produces mainly water and nitrogen on combustion, replacing a part of conventional fuel with ammonia will have a large effect in reducing carbon dioxide emissions. In this regard, a unique list of the specific features of ammonia (NH3) is as follows:

• • • • • • • • • • •

It consists of one nitrogen atom from air separation and three hydrogen atoms from any conventional or renewable resources. It is the second largest synthesized industrial chemical in the world. It is a significant hydrogen carrier and transportation fuel that does not contain any carbon atoms and has a high hydrogen ratio. It does not emit direct greenhouse gas emission during its deployment. It can be used as solid and/or liquid for many purposes. It can be stored and transported under relatively lower pressures compared to hydrogen. It can be produced from a various type of resources ranging from oil sands to renewables. It is a suitable fuel to be transferred using steel pipelines with minor modifications which are currently used for natural gas and oil. It can be used in all types of combustion engines, gas turbines, burners as a sustainable fuel with only small modifications and directly in fuel cells which is a critical advantage compared to other type of fuels. It brings a non-centralized power generation via fuel cells, stationary generators, furnaces/boilers, and enables smart grid applications. It can be used as a refrigerant for cooling in the vehicles and other applications.

Ammonia can be produced by extracting nitrogen from air and hydrogen from water and combining them with the presence of any energy source, including, especially, the renewables. Presently, more than 90% of the ammonia manufacture in the world is done by the Haber–Bosch synthesis process as it was developed in 1913 [4,5]. The Haber–Bosch process combines hydrogen and nitrogen together with an iron-oxide catalyst under very high pressure and elevated temperatures. On the other hand, new expertise such as solid state synthesis and electrochemical processes are currently being developed to reduce the cost and to improve the efficiency of ammonia production processes. Ammonia is one of the major synthesized industrial chemicals on Earth. Ammonia production consumes nearly 1.2% of total primary energy and subsidizes about 1% of GHGs emissions [6]. About 1.5 t of CO2 is released into the environment during the production of 1 t of ammonia depending on the location of the plant [7]. Regarding conventional sources, naphtha, heavy fuel oil, coal, natural gas coke oven gas, and refinery gas can be used as feedstock for ammonia manufacture. Natural gas is the main feedstock used for making ammonia in Canada, and worldwide. Eleven ammonia plants are operating in Canada, producing a yearly average of 4–5 million metric tons per plant [8]. The United States is accepted as the biggest ammonia importing country which corresponds to about 35–40% of world trade. On the other hand, Europe accounts for approximately 25% of commerce while they produce ammonia with a higher cost. In the coming future, it is expected that the majority of import development will be in Asia because of industrial uses and fertilizer products. Though there are a few practical techniques for ammonia synthesis, commonly there is one ammonia synthesis technique available in the world named as the Haber–Bosch process [9] whereas the electrochemical ammonia synthesis processes are also under investigation. In both approaches, nitrogen is supplied through the air separation process. Cryogenic air separation is presently the most efficient and cost-effective technology for generating a large amount of oxygen, nitrogen, and argon [10]. Using cryogenic technology, nitrogen can also be produced in high purity which can further be utilized as a useful by-product stream at quite small incremental price. Among other air separation processes, cryogenic air separation has most mature and developed technology. Since ammonia is produced in high amounts, required nitrogen should be generated in a low cost and high efficient. The required electricity could be supplied either from conventional or alternative sources. The Haber–Bosch process is an

4

Ammonia

Table 1

Thermo-physical properties of ammonia

Property

Liquid

Color Density (01C, 101.3 kPa) Density ( 33.431C, 101.3 kPa) Boiling point (101.3 kPa) Melting point Critical temperature Critical pressure Critical viscosity Lower heating value, LHV (MJ kg 1) Higher heating value, HHV (MJ kg 1)

Colorless 638.6 kg m 3 682 kg m 3 33.431C 77.711C 132.41C 11.28 MPa. 23.90  10 3 mPa s 18.57 22.54

Source: Data from Kaudy L, Rounsaville JF, Schulz G. Ullmann’s encyclopedia of industrial chemistry. Weinheim: VCH; 1987.

Table 2

Some properties of ammonia, hydrogen and other conventional fuels

Feature

Gasoline

Diesel

Natural gas

H2

NH3

Flammability limit, volumes (% in air) Auto-ignition temperature (1C) Peak flame temperature (1C)

1.4–7.6 300 1977

0.6–5.5 230 2053

5–15 450 1884

4–75 571 2000

16–25 651 1850

Source: Data from Zamfirescu C, Dincer I. Ammonia as a green fuel and hydrogen source for vehicular applications. Fuel Process Technol 2009;90:729–37.

exothermic process that combines hydrogen and nitrogen in 3:1 ratio to produce ammonia. The reaction is facilitated by catalyst and the optimal temperature range is 450–6001C [11,12]. Ammonia (NH3), which has a great content of three atoms of hydrogen per unit of capacity, has occasionally been used in the history as a fuel for internal combustion engines (ICEs) and fuel cells [13]. One mole of ammonia comprises 1.5 mol of hydrogen which is 17.8% by weight or 108 kg H2 m 3 entrenched in liquid ammonia at 201C and 8.6 bar. It is remarkable to know that this density is four times greater than that of the most advanced storing approaches in metal hydrides which extend about 25 kg H2 m 3 [14]. The physical properties of ammonia are listed in Table 1. NH3 is one of the most attractive fuels due to the following features:

• • • •

It has a high octane rate of 110–130 [15] and therefore is a decent fuel for ICEs. It can be thermally decomposed into hydrogen and nitrogen using low energy, i.e., approximately 12% from the higher heating value (HHV) [2] to generate hydrogen for fuel cells. The delivery substructure already exists for ammonia [16] to deliver it in quantities superior than 100 million tons annually [17] or more. It is still safer than many fuels and hydrogen due to the following possessions: ✓ If escapes into the air, it disperses fast because its density is lighter than that of air. ✓ It is self-alarming: nose can notice any leak in concentrations as low as 5 ppm [9]. ✓ It has a slight flammability range and so, it is typically evaluated nonflammable and giving no explosion danger when correctly conveyed; this detail is evident from the data listed in Table 2 which were compiled from various sources [18–20].

Although there are some flammability and toxicity concerns linked to NH3 and its use, these challenges are easy to overcome with the appropriate safety measures and control. That is why this has been the most widely used refrigerant and working fluid since the industrial revolution. Especially, there are some side effects with every refrigerant, working fluid and fuel. Nonetheless, ammonia appears to be one of the most environmentally friendly refrigerant and working fluid and hence fuel while considering the ammonia produced by renewable hydrogen and nitrogen (from the air). If we further elaborate, such difficulties have mainly been addressed, and these apparent disadvantages are compensated by the well-established knowledge in NH3 treatment, storing and use in numerous forms (i.e., gaseous, liquid as well as solid), particularly in agriculture, chemical, and cooling applications. Some quantities of NOx will consequence due to the extra nitrogen presence in the burning cavity if NH3 is used as fuel for ICEs. It is also stimulating to note that NH3 is a reduction agent for the NOx, typically in combustion releases. The reaction of NOx with ammonia over catalysts produces only steam and nitrogen. An average car needs approximately 30 mL of NH3 per 100 km to neutralize any NOx emissions [21]. If the vehicles run on NH3 as fuel, this amount is unimportant with respect to the fuel tank volume.

Ammonia

5

Comparison of ammonia with other fuels

Table 3 Fuel/storage

P (bar)

r Density (kg m 3)

HHV (MJ kg 1)

HHV per volume (GJ m 3)

Energy per volume (GJ m 3)

Cost per mass ($ kg 1)

Cost per volume ($ m 3)

Cost per energy ($ GJ 1)

Gasoline, C8H18/liquid CNG, CH4/integrated storage LPG, C3H8/pressurized tank Methanol, CH3OH/liquid Hydrogen, H2/metal hydrides Ammonia, NH3/pressurized tank Ammonia, NH3/metal amines

1 250 14 1 14 10 1

736 188 388 786 25 603 610

46.7 42.5 48.9 14.3 142 22.5 17.1

34.4 10.4 19 11.2 3.6 13.6 10.4

34.4 7.8 11.7 9.6 3 11.9 8.5

1.03 0.91 1.06 0.41 3.02 0.23 0.23

754.86 170.60 413.66 317.80 75.49 136.63 138.14

21.97 21.29 21.74 28.31 21.29 10.04 13.21

Source: Reproduced from Zamfirescu C, Dincer I. Ammonia as a green fuel and hydrogen source for vehicular applications. Fuel Process Technol 2009;90:729–37.

Ammonia is considered as a possible working fluid for thermodynamic cycles, refrigeration, heating, and power [22,23]. It is used for exhaust gases to drive automotive absorption refrigeration system [24]. More lately, Zamfirescu and Dincer [25–27] have examined the option of using the on-board ammonia as refrigerant while it is spent as fuel for vehicle propulsion. Ammonia also seems to be a medium for thermochemical storing of energy as confirmed by numerous works [28,29]. Consequently, ammonia can play a critical part in vehicular applications, as fuel, a hydrogen source, working fluid/refrigerant, NOx reduction agent, or perhaps as an energy storage intermediate. In Table 3, it is listed the fuel and the type of storing option, the fuel pressure in the storing tank, and the fuel density in the full tank. The exact fuel energy per mass is given regarding fuel’s HHV. The volumetric energy of the fuel is found by multiplying the HHV with the density value listed in the third column. Ammonia’s HHV is around half of the one of gasoline, and its density is also inferior. Therefore, liquid NH3 stores 2.5 fewer energy per unit capacity than gasoline. If the NH3 is stored in the form of hexaamminemagnesium chloride [30] to remove the hazard related to its toxicity, the energetic cost to pay for discharging ammonia reduces its HHV. The manufacture of ammonia from fossil fuels is in numerous features similar to hydrogen manufacture since it includes gasification to produce syngas, gas cleaning and CO2 elimination, compression of the reactants and catalytic conversion, and ammonia separation through condensation. An extremely energy consuming constituent of the ammonia manufacture is considered as the makeup gas compression which is needed to ease the synthesis. This apparent disadvantage is compensated by an efficient synthesis procedure that is likely at high pressure. The procedure is eased even more by the facile separation of NH3 product that is made at no extra costs by ammonia condensation and heat recovery. Thus ammonia is continuously drained from the synthesis loop such that the chemical equilibrium in the reactor is shifted toward ammonia formation. Furthermore, ammonia synthesis is an exothermic process, and contemporary technologies use heat recovery to reduce the production prices [31]. Strickland [32] pointed even since 1980 that amongst other fuels, counting liquid hydrogen and methanol, ammonia (NH3) has some excellent advantages which make it the neighboring one to gasoline. Numerous other new studies have proposed NH3 as a possible hydrogen storing medium [9,16,33]. A significant advance in ammonia synthesis was noticed in 1931 when Shell Chemicals produced NH3 from natural gas for the first time [27]. Starting with the natural gas crisis in 1970 ammonia manufacturers became worried to advance and develop methods to efficiently produce ammonia from other fossil fuels, counting coal [34]. In order to decompose ammonia and separate the hydrogen from nitrogen at the same time, Skodras et al. [35] industrialized a catalytic membrane decomposition unit: after the decomposition procedure which take place typically at the catalytic membrane surface, only hydrogen permits through the membrane; thus small traces of unreacted ammonia and the produced nitrogen are separated from the pure H2 stream. As an alternative option, hydrogen can also be obtained via ammonia electrolysis [36]. This will definitely make it more pure, more cost effective, more commercially viable and reliable, and more environmentally benign than the hydrogen produced through some conventional techniques such as hydrogen production from natural gas.

2.1.1.1

Ammonia Production and Transport

The annual increase in ammonia production covering the period from 2002 to 2015 is shown in Fig. 1. Regarding conventional resources, naphtha, heavy fuel oil, coal, natural gas coke oven gas and refinery gas can be used as feedstock for ammonia production. Natural gas is the primary feedstock used for producing ammonia in worldwide as shown in Fig. 2. Ammonia production plants in Canada are ranked internationally as having the highest feed plus fuel energy (FFE) plant efficiency which consumes a typical 33.8 GJ natural gas per ton of produced ammonia [37]. In comparison, the world FFE average is of 38.6 GJ t 1 NH3. FFE is related to the CO2 generated by the ammonia plant, whereas net energy efficiencies contain electrical consumption and modifications for other energy debits and credits, which may have connected offsite CO2 emissions not directly from the ammonia plant [37]. In China, coal is intensively used and is generally characterized by high energy requirements [38]. In 2010, ammonia plants in China had a projected average energy intensity of 49.1 GJ t 1 NH3. About 75% of the ammonia in China was produced with the coal-based method. The average energy intensity of the Chinese coal-based ammonia-producing plants is 54 GJ t 1 NH3 [38]. Natural gas costs constitute 70–90% of the production cost of ammonia. Since ammonia production is based on natural gas in the steam methane reforming (SMR) method, if natural gas prices rise, production costs for ammonia increase in parallel. For the Haber–Bosch process, production of ammonia is based on various hydrogen production techniques as shown in Fig. 3.

6

Ammonia

Global ammonia production (million tonnes)

200 180 160 140 120 100 80 60 40 20 0 2002

2004

2006

2008

2010

2012

2014

2016

Year

Fig. 1 World ammonia production growth. Data from Yara Fertilizer Industry Handbook. Available From: http://www.yara.com; 2014.

Fuel oil 4%

Naphta 1%

Others 1%

Coal 22%

Natural gas 72%

Fig. 2 Sources of global ammonia production based on feedstock use. Data from International Energy Agency. Energy Technology Perspectives 2012; Pathways to a Clean Energy System, France. ISBN: 978-92-64-17488-7.

The storage and distribution structure of ammonia are comparable to the liquefied petroleum gas (LPG) process. Under medium pressures (in the range of 5–15 bar), both of the materials are in liquid state. This qualifies the significant advantage because of storing opportunities. Nowadays, vehicles running on propane are commonly recognized and used by the public since their on-board storage is possible and it is a good case for ammonia-fueled vehicle opportunities since the storage and risk characteristics of both substances are similar to each other. An ammonia pipeline from the Gulf of Mexico to Minnesota and with divisions to Ohio and Texas has served the ammonia industry for many years. It indicates that there is a working ammonia pipeline transportation which can be spread overall the world. The potential of using ammonia in many applications will be dependent on the accessibility of ammonia in the cities. Ammonia is a suitable material to be transported using steel pipelines with slight adjustments which are presently used for natural gas and oil. In this way, the problem of accessibility of ammonia will be abolished. A pipeline can deliver almost 50% more energy when conveying liquid ammonia than carrying compressed natural gas because of the volumetric energy densities [39].

2.1.1.2

Ammonia Storage

Ammonia can be stored in two different ways, pressurized or at low temperature. Pressurized storage keeps ammonia in a liquid phase having a pressure above 8.6 bar at ambient temperature (201C), but ammonia is usually stored at 17 bar to keep ammonia in liquid phase if ambient temperature increases. The energy density of the liquid ammonia stored pressurized is 13.77 MJ L 1. A rule of thumb is that 2.8 t of ammonia may be retained per ton steel. This storage does not require energy to maintain the pressurized state. Low-temperature storage is usually employed for large-scale applications. This type of storage requires energy to maintain its low temperature and thereby avoid boil-off due to ambient temperature. Lower capital cost is the reason why low temperature is preferred for large scale storage. The energy density of the liquid ammonia

Ammonia

Renewable

Solar

Wind

Tidal and wave

Ocean thermal

Conventional

Hydro

Biomass

Geothermal

Coal gasification

UCG

SMR

Cryogenic

Nuclear

7

Non-cryogenic

Heavy oil

H2

N2

NH3 synthesis (Haber–Bosch)

Internal combustion engine

Fuel cell systems

Fig. 3 Ammonia production and usage routes.

10,000 10,000

Cost ($)

8000

Electricity: batteries

Hydrogen (70 bar) ICEHV

Hydrogen (70 bar) FCHEV

CNG

Ammonia (20 bar)

Gasoline, diesel

6000 4000 4000

3000

2000 300

300

100

0 Fig. 4 Estimated OEM costs of on-board ammonia storage tanks for one personal vehicle with 482 km range. Data from Leighty Bill. Energy storage with anhydrous ammonia: comparison with other energy storage. In: 2008 annual NH3 fuel conference NH3 fuel association, Minneapolis; 2008.

stored in this way is 15.37 MJ L 1 compared to 13.77 MJ L 1 for pressurized storage. If storage time is assumed 182 days representing a period between winter and summer, will give a storage cost of 4.03 $ GJ 1 for ammonia. It can be mentioned that this cost is much lower compared to hydrogen storage that costs 98.74 $ GJ 1. As illustrated in Fig. 4 the estimated cost of a storage tank on a personal vehicle is lowest for ammonia after standard and conventional gasoline/diesel tanks for a 482 km range. The hydrogen density is highest for ammonia among other fuels as illustrated in Fig. 5. Because the main application of ammonia is fertilizer, large-capacity seasonal storing tanks are already developed. Ammonia demands peak throughout the summer when it must be spread on farming fields. Ammonia is produced throughout the year, and the winter’s manufacture is stored for the summer period. Tanks with a volume of 15,000–60,000 m3 were constructed before the 1970s [40]. Ammonia is kept in the chilled state at ambient pressure and at its standard boiling point, which is 331C. The tanks are cylindrical with a 38–52 m inner diameter and 18–32 m of height. In order to compensate for the heat diffusions, the whole building is well insulated (a double-wall technology is used) and compressors are employed to eliminate the heat by simulating a cooling plant for which the tank plays the role of an evaporator. Essentially, ammonia vapors existing above the liquid are aspired

8

Ammonia

140

136 126

Hydrogen density

Hydrogen density (kg H2 m–3) Lower heating value (MJ kg–1)

LHV

120.1

116

120

110

106

102

98

100 80

70

60 42.8

50.1 42.5

45.8

40

27 20.1

18.6

20 0

Ammonia

Diesel

Methane

Gasoline

Propane

Ethanol

Methanol

Hydrogen

Hydrogen density

136

126

116

110

106

102

98

70

LHV

18.6

42.8

50.1

42.5

45.8

27

20.1

120.1

Fig. 5 Comparison of hydrogen density and lower heating value (LHV) values of various fuels.

by the compressors and delivered at high pressure where the vapors are condensed, and the liquid is returned to the tank. In this way, the temperature and the pressure in the tank are kept constant [40,41]. Ammonia can be stored onboard of a car in pressurized cylinders in an anhydrous form or in some chemical forms such as metal amines or ammonia boranes, which are produced by means of lately developed physical–chemical reversible methods [30,41]. In this technology, ammonia is adsorbed on a porous metal–amine complex, for example, hexaaminemagnesium chloride, Mg(NH3)6Cl2; to do this, NH3 is passed over an anhydrous magnesium chloride (MgCl2) powder at room temperature. The absorption and desorption of ammonia in and from MgCl2 are entirely reversible. The metal amine can be formed in the desired form and can stock 0.09 kgH2 kg 1 and 100 kgH2 m 3.

2.1.1.3

Ammonia Utilization

The ability to use one fuel in all types of combustion engines, gas turbines, burners, and directly in fuel cells is a tremendous advantage. Storage and delivery infrastructure would be significantly reduced if ammonia is employed rather than hydrogen. NH3 is one of a very short list of fuels that can be used in nearly every type of engine and gas burner with only minor modifications. Gas burners can be equipped with in-line partial reformers to split approximately 5% of the NH3 into hydrogen. This mixture produces a robust, unpolluted burning open flame. One pipeline to a home could provide NH3 to furnaces/boilers, fuel cells, stationary generators and even vehicles. Due to the very minor enthalpy of reforming exhibited by NH3, it can easily be reformed to hydrogen for any application that would require hydrogen. Relatively minor modifications allow efficient use of ammonia as a fuel in diesel engines; high compression ratio spark ignition engines can produce astounding efficiencies of over 50% using NH3 fuel; direct ammonia fuel cells promise to be low-cost, robust, and very efficient; NH3 is also a very suitable fuel for use in solid oxide fuel cell (SOFC) and gas turbines. These medium-temperature (approximately 4001C) fuel cells promise to be low-cost, highly efficient, and very robust [39]. The global ammonia demand is forecasted to grow at an average annual rate of approximately 3% over the next 5 years. The historical growth rate was 1%. Therefore, currently, it is 2% above. The global ammonia consumption amounts are shown in Fig. 6. Solid agricultural materials are expected to drive this growth as fertilizer uses account for approximately 80% of global ammonia demand [42]. The global ammonia exports and imports based on the selected years are illustrated in Fig. 7. The United States and Asia have quite close import rates whereas Western Europe’s imports are almost half of Asia. In recent years, the export of ammonia from Africa and Middle East increases gradually. The United States is recognized as the largest ammonia importer and typically accounts for approximately 35–40% of world trade. Europe, a higher-cost producer, accounts for roughly 25% of commerce. The majority of growth in imports is expected in Asian countries, for industrial uses and the production of fertilizer products. The physical properties of ammonia require high-pressure containers, making it a little costly and difficult to transport. Most of the ammonia is consumed close to where it is produced as illustated in Fig. 8. The domestic sales represent approximately 88% of world ammonia trade. Asia is the main ammonia trading country more than sum of other continents. Almost 53% of ammonia is currently used as fertilizer in the United States as shown in Fig. 9. The direct applications constitute only quarter of whole usage.

Ammonia

Million tonnes 200

Urea

Other fertilizer

DAP/MAP

9

Industrial

180 160 140 120 100 80 60 40 20 0 2001

2003

2005

2007

2009

2011

2013

2015

Fig. 6 World ammonia consumption and distribution. Data from PotashCorp Integrated Annual Report. Annual integrated report. Available From: http://www.potashcorp.com/irc/nitrogen; 2015 [accessed 07.01.17].

Imports 30

W. Europe

Asia

Other

20

30 25

Million tonnes

Million tonnes

25

US

Exports

15 10

Trinidad

FSU

Africa

Others

Middle east

20 15 10 5

5

0

0 2005 2007 2009 2011 2013 2015

2005 2007 2009 2011 2013 2015

Fig. 7 World ammonia trade shares. Data from PotashCorp Integrated Annual Report. Annual integrated report. Available From: http://www. potashcorp.com/irc/nitrogen; 2015 [accessed 07.01.17].

However, when the world ammonia usage is considered, direct applications represent only 4% of overall ammonia consumption as illustrated in Fig. 10. The lack of ammonia using devices and equipment leads indirect applications. Utilization of ammonia in household applications is also possible in various ways. In an ammonia economy, the readiness of a pipeline to the residential area could source ammonia to fuel cells, stationary generators, furnaces/boilers, and even vehicles which will bring a non-centralized power production and allow smart grid applications [39]. Decentralized power generation and utilization is one of the solutions for transmission lines. Ammonia can play a crucial role in this process since it has multiple usage options [43]. It is emphasized that the physical characteristics of ammonia are close to propane. The capability to convert a liquid at adequate pressure permits ammonia to store more hydrogen per unit volume than compressed hydrogen/cryogenic liquid hydrogen. Besides having significant advantages in storing and transporting hydrogen, ammonia may also be burned directly in ICE. Compared to gasoline vehicles, ammonia-fueled vehicles do not produce direct CO2 emission during operation. However, it is important to determine not only direct emissions associated with vehicle operation but also total energy cycle emissions related to fueling the vehicles. Furthermore, ammonia can be produced at locations where oil and natural gas extraction wells are located. In this way, generated CO2 can be reinjected into the ground for sequestration or can be reacted with ammonia for urea production. Ammonia can then be easily transferred through pipelines, railway cars, and ships by delivering to consumption area where it may be utilized as a source of hydrogen, chemical substance, and fertilizer for agriculture, fuel for transportation and power generation sector, working fluid or refrigerant. Ammonia can be utilized in many transportation applications as shown in Fig. 11.

10

Ammonia

Production - Million tonnes

2010

2015F

Asia FSU Exports Middle East

12%

Europe North America Latin America

88%

Africa Domestic sales Oceania 0

10

20

30

40

50

60

70

80

90

Fig. 8 Domestic- and export-based world ammonia production profile. Data from PotashCorp Integrated Annual Report. Annual integrated report. Available From: http://www.potashcorp.com/irc/nitrogen; 2015 [accessed 07.01.17].

Non-fertilizer 23%

53%

Upgraded fertilizers

24% Direct application Fig. 9 Ammonia consumption in the United States for industrial and fertilizer purposes. Data from PotashCorp Integrated Annual Report. Annual integrated report. Available From: http://www.potashcorp.com/irc/nitrogen; 2015 [accessed 07.01.17].

Non fertilizer 19%

Direct application 4% Urea 48% DAP/MAP 7%

Other fertilizer 14%

Ammonium nitrate 8% Fig. 10 World ammonia usage, average of 2010–2013. Data from PotashCorp Integrated Annual Report. Annual integrated report. Available From: http://www.potashcorp.com/irc/nitrogen; 2015 [accessed 07.01.17].

Ammonia

11

Ammonia synthesis

Ammonia storage

Compression ignition engines

Spark ignition engines

Fuel cells

Ammonia

Ammonia dual fuel

Ammonia and gasoline

Ammonia and hydrogen

Ammonia

Ammonia and diesel

Ammonia duel fuel

Ammonia and DME

Ammonia and biodiesel

Fig. 11 Direct ammonia utilization pathways for transportation sector.

The following list summarizes some vehicle powering options and potential applications of ammonia:

• • • • • •

Spark ignited ICE, Diesel ICE with H2 or diesel “spike”, Combustion turbines, Gasoline or ethanol mixture ICEs, Transformed biogas generators, Direct ammonia fuel cells.

For power generation systems, where the storing space is readily accessible, the energy density is not the responsible aspect for the fuel choice, as the cost per MJ and emission stages are characteristically the critical factors. With the new energy efficient systems of making ammonia on the cost per MJ basis, ammonia manufactured via renewable energy resources would be competitive with the fossil-based fuels. The toxicity issue is not also as dangerous for power generation methods since the fuel will be controlled by professionals following well-established handling processes.

2.1.1.3.1

Ammonia in Heating, Ventilation and Air Conditioning applications

Ammonia has been recognized and employed as a leading refrigerant in the industrialized regions due to its outstanding thermal features, zero ozone depletion and zero global warming potential (GWP). Ammonia has the maximum refrigerating outcome per unit mass compared to all the refrigerants being used counting the halocarbons. The notable benefits of ammonia over R-134a could be: inferior overall operational costs of ammonia systems, the flexibility in meeting complex and several refrigeration needs, and inferior initial costs for several applications [44]. Ammonia is obtainable almost everywhere and is the lowest cost of all the regularly used refrigerants. Ammonia has superior heat transfer features than most of the chemical refrigerants and consequently allow for the use of equipment with a smaller heat transfer area. Thus plant building costs will be lower. Furthermore, as these features also benefit the thermodynamic efficiency in the system, it also diminishes the operational costs of the system. In many countries, the cost of ammonia per mass is significantly inferior to the cost of HFCs. This kind of advantage is even increased by the fact that ammonia has a lower density in the liquid phase. Contemporary ammonia systems are entirely closed-loop systems with completely integrated controls, which adjust the pressures all over the system. Additionally, every refrigeration system is regulated by codes, which are effective, mature, and continuously updated and revised, to have safety relief valves to protect the system and its pressure vessels from over-pressurization and possible failure. For a refrigerant to be considered a long-term option, it is advised to meet three criteria:

• • •

Safe, Environmentally friendly, Good thermodynamic performance.

Numerous non-halogen materials, containing ammonia, carbon dioxide, and hydrocarbons, work as refrigerants. All of these materials can be refrigerants for the right use if the system can be planned to meet the main choice criteria. Component and equipment manufacturers continue to research how these refrigerants perform in systems. Ammonia (NH3) has constantly been a

12

Ammonia

Heating

Condenser

Electricity

Compressor

Ammonia power generator

Expansion valve Ammonia refrigerant Pump Evaporator Liquid separator

Cooling Fig. 12 Ammonia-based heating, ventilation and air conditioning (HVAC) system schematic.

leading refrigerant in the industrial segment. It is classified as a B2 refrigerant by ASHRAE 34-2013 (Designation and Safety Classification of Refrigerants) for toxicity and flammability, and therefore governed by strict regulations and codes. Ammonia is used as refrigerant commonly in the refrigeration structures of food industry like dairies, ice creams plants, frozen food production plants, cold storage warehouses, processors of fish, poultry and meat, and a number of other uses. Though the specific volume of ammonia is great, the compressor displacement essential per ton of refrigeration is fairly minor, because small compressor is desired per ton of the cooling capacity. This saves lots of power in the long run. For the typical conditions around 151C in the evaporator, the condenser and the evaporator pressures are about 2.37 and 11.67 bar, respectively. Since the pressures are not very high, lightweight substances can be used for the building of the equipment. The pressure in the evaporator is quite high, so it is not necessary to expand the gas to very low pressure. This also empowers high suction pressure for the compressor and lower compression ratio. The release temperature of the ammonia refrigerant from the compressor is high, hence water cooling of the cylinder heads and the cylinders of the compressor is vital. If high discharge pressure is necessary, it is desirable to use the multi-cylinder compressors instead of the single cylinder compressor to evade overheating of the compressor. Recently, alternative ammonia chilled water systems are also developed. One of these examples is the elimination of compressor. Although there are some chilled water systems in residential applications, they are mostly employed in commercial air conditioning systems. A basic schematic of heating, ventilation and air conditioning (HVAC) systems is illustrated in Fig. 12. Ammonia can be used as a refrigerant in the cycle of HVAC systems. Additionally, for stand-alone applications, the power required for a compressor of HVAC system can be produced by ammonia-based power generation units so called ammonia generators as shown in Fig. 12. The compressor sucks the dry gas (from the evaporator and flash gas) from the separator at evaporating temperature, compresses it to condensing temperature and feeds the superheated discharge gas to the condenser. The condenser liquefies the refrigerant while dissipating the heat from the refrigerant gas to the cooling media. From the condenser, the liquid refrigerant is fed to the expansion device at condensing pressure and close to the condensing temperature. In the expansion device, the ammonia is expanded to evaporating temperature and then fed to the separator. In the separator, liquid and flash gasses are separated. The liquid refrigerant, at evaporating temperature and pressure, is sucked by the pump and delivered to the evaporator. In the evaporator, the heat exchange takes place. A mix of gas and liquid is fed back to the separator, where the liquid is separated from the gas, and the compressor can suck dry gas.

2.1.1.3.2

Ammonia as both fuel and refrigerant

Ammonia has outstanding potentials as a refrigerant and as a fuel. It is also worth to examine the option to cool the engine with ammonia that can act as a refrigerant while it is heated to the temperature at which it is fed to the power producer (ICE or fuel cell). Optionally, the cooling outcome of ammonia, i.e., its high latent heat of evaporation, may be used to harvest some air

Ammonia

13

5

Gasoline LPG

CNG

10

Bateries

15

Ammonia

20

Liquefied H2

GJ m−3

25

Compressed H2 Chemical hydrides

30

Methanol Metal amines

35

Proposed DOE goal 2020

40

0 0

10

20

30

40

50

60

GJ t−1 Fig. 13 Comparison of volumetric energy densities and specific energy densities of various fuels and ammonia. Modified from Zamfirescu C, Dincer I. Ammonia as a green fuel and hydrogen source for vehicular applications. Fuel Process Technol 2009;90:729–37.

Table 4 NH3 fuel

Conversion properties of hydrogen-fueled ICE Ford Focus to run on

Property

Unit

H2

NH3

Volume of storage tank Pressure of storage On-board energy Cost of full tank Range of drive Cost of drive Compactness of tank

Liter Bar MJ $ km $ 100 km L 100 km

217 345 710 18.87 298 6.34 73

76 10 1025 10.57 430 2.42 18

1 1

Source: Reproduced from Zamfirescu C, Dincer I. Ammonia as a green fuel and hydrogen source for vehicular applications. Fuel Process Technol 2009;90:729–37.

conditioning onboard. The comparison of volumetric energy densities and specific energy densities of numerous fuels is illustrated in Fig. 13. Numerous automakers have industrialized the prototypes of hydrogen-fueled cars in recent years. Here, for examination purposes, a Ford Focus H2ICE prototype is selected [3]. In Table 4, it is listed the performance parameters of the real prototype and some calculation results for the similar prototype as converted to NH3 fuel. In calculation it has been assumed that the price of ammonia is $ 0.23 kg 1 and the power-train performance is characterized by 1.19 MJ km 1 shaft power where it is founded on specified 50% efficiency, 710 MJ stored in the full tank and 298 km driving range [3]. The effectiveness of the ammonia engine has been taken the similar as the hydrogen engine. Actually, ammonia can be dissociated onboard at no extra cost (only using the heat rejected by the ICE) and the engine fueled with pure hydrogen [3]. 2.1.1.3.2.1 Thermo-catalytic decomposition of ammonia Ammonia can be decomposed thermo-catalytically to generate hydrogen according to the following endothermic reaction [40]: 2 NH3 þ 30:1 kJ mol 3

1

1 H2 -H2 þ N2 3

ð1Þ

Here, the required enthalpy signifies 10.6% of HHV or 12.5% of the lower heating value (LHV) of the generated hydrogen. The ammonia decomposition reaction does not need catalysis to be performed at high temperatures for example over 1000K; though, at inferior temperatures, the reaction rate is too slow for practical applications such as hydrogen generation for energy conversion. At 4001C, the equilibrium conversion of NH3 is very high at 99.1% [45] and at about 4301C, almost all ammonia is converted to hydrogen at equilibrium, below atmospheric pressure circumstances [11]. There is a big array of catalysts appropriate to ammonia decomposition (e.g., Fe, Ni, Pt, Ir, Pd, and Rh), nonetheless ruthenium (Ru) seems to be the finest one when reinforced with carbon nanotubes, making hydrogen at additional than 60 kW equal power per kilogram of catalyst [45]. Over ruthenium catalysts, at temperatures lower than about 3001C, recombination of nitrogen atoms is rate limiting, while at temperatures higher than 5501C, the cleavage of ammonia’s N–H bond is rate limiting. Though, the activation energy is greater at low temperature (180 kJ mol 1) and inferior at higher temperatures (21 kJ mol 1). The finest temperature range for ammonia decomposition

14

Ammonia

over ruthenium catalysts may be 350–5251C, which proposes that flue gases from hydrogen ICEs, other hot exhausts from burning equipment, or electrochemical power conversion in high-temperature fuel cells can be used to drive ammonia decomposition [40].

2.1.1.3.3

Ammonia and urea

The mission of finding the optimal hydrogen carrier is not easy as it includes multi-criteria decision making and attention of numerous practical and financial characteristics of safety, energy density and cost of processing or recycling. This has led to the inspection of a miscellaneous spectrum of storing resources such as metal hydrides, metal-organic materials, and amide systems [46]. In spite of wide research and improvement determinations, these equipment and composites have main disadvantages revolving around the rate of hydrogen desorption, cyclability, and high cost [47,48]. With this respect, ammonia has been regarded as an excellent hydrogen carrier for its several favorable attributes as shown in Table 5. Large quantities of ammonia are used worldwide for agricultural purposes. The infrastructure and technology of ammonia production are also well established with existing industrial plants around the world to support the increasing demand for fertilizers [9]. Natural gas is the main feedstock for the synthesis of ammonia which uses the steam reforming method. So from a life-cycle perspective, the production of 1 t of ammonia emits about 1.5 metric tons of carbon dioxide most of which can be easily recovered for use in downstream processes such as the manufacture of urea or other derivatives [49]. This figure excludes the potential amount of carbon dioxide emitted if carbon-based fuel is used to provide the energy required to drive the process of ammonia production. On the other hand, ammonia is corrosive, toxic, and life-threatening when released at high concentrations [50]. To lessen these risks, some attention has been focused toward steadying the ammonia by merging it in metal ammine complexes or ammoniaborane systems. This permits for the transportation and long-term storage of fuel in solid state or liquid form and hydrogen can be released on demand [30]. However, such systems are also burdened with disadvantages like to those discussed earlier. Alternatively, urea is a nontoxic chemical which can be found in natural systems as well as human and animal waste (urine). On average, the concentration of urea in human urine is 9.3–23.3 g L 1 [51]. Pure urea is formed as white, odorless prills, or granules when artificially synthesized. Owing to its stable nature, it can be easily and safely handled, transported and stored at room temperature. Also, urea is the most widely used solid fertilizer worldwide. In 2009, the global production of urea reached 146 million tons, and it is anticipated to increase to 210 million tons by 2013 due to increasing global demand. This major increase is, due to the growth of the nonagricultural use of urea in emission control (DeNOx) systems for industrial and automotive applications [52]. As stated earlier, the process of ammonia production normally supplies the feedstock of ammonia and carbon dioxide for the synthesis of urea. Therefore, greenhouse gas is released only when fossil fuel is utilized to provide the required energy for this process.

2.1.2

Analysis and Assessment

In this section, ammonia and urea usage in fuel cell applications is discussed by electrochemical analyses presented before in Ref. [52]. Ammonia and urea are directly used in SOFCs for power generation.

2.1.2.1

Ammonia in Direct Ammonia Solid Oxide Fuel Cell

When heated, ammonia is decomposed to hydrogen and nitrogen according to the following reaction: 3 1 NH3 2 H2 þ N2 2 2

ð2Þ

The above reaction is endothermic and depends on the temperature and pressure of the system. The reaction proceeds to the right as the temperature increases or to the left as the pressure increases thus producing less hydrogen. For the purpose of this analysis, it is assumed that the ammonia is directly fed to the fuel cell where it will be consumed and the only species involved in the decomposition process are NH3, H2, and N2. Catalysts are used to lower the activation energy and speed up the rate of ammonia decomposition reaction within the fuel cell. Much effort has been focused on finding a suitable catalyst to promote the decomposition of ammonia at low temperature and Table 5

Energy density of different energy carriers (based on LHV value)

Energy carrier

Density (kg m 3)

Gravimetric density (%H2)

Volumetric density (kg H2 L 1)

Energy density (MJ L 1)

Gaseous H2 (298K, 10 MPa) Liquid H2 (30K, 10 MPa) Liquid NH3 (298K, 1 MPa) Aqueous urea (76.92%wt – STP)

7.68 72.58 603 1200

100 100 17.76 7.74

0.0077 0.0726 0.1071 0.0930

0.92 8.71 12.85 11.16

Sources: Reproduced from Ma Q, Ma J, Zhou S, et al. A high-performance ammonia-fueled SOFC based on a YSZ thin-film electrolyte. J Power Sources 2007;164 86–9 and Egan EP, Luff BB. Heat of solution, heat capacity, and density of aqueous urea solutions at 251C. J Chem Eng Data 1966;11:192–4.

Ammonia

15

high kinetic rate. Although ruthenium-based catalysts are expensive, they can achieve near complete decomposition of ammonia at temperatures around 600K. More attractive alternatives include nickel-based catalysts which are relatively cheaper and can achieve the same task at about 700–800K [11]. Since the operating temperature of an average SOFC ranges between 900 and 1400K [53], it is safe to neglect the kinetic rate of reaction and assume that full decomposition of ammonia is attained within the porous anode layer.

2.1.2.2

Urea in Direct Urea Solid Oxide Fuel Cell

An aqueous solution composed of equimolar amounts of urea and water (76.92% urea by weight) can be heated and injected into the system resulting in a transitory evaporation of water from the spray droplets such that [52]: ðNH2 Þ2 COðaqÞ -ðNH2 Þ2 COðsÞ þH2 OðgÞ

ð3Þ

The rapid thermolysis of solid urea can be initiated at a relatively low temperature of 406K to yield ammonia and isocyanic acid as ðNH2 Þ2 COðsÞ -NH3ðgÞ þHNCOðgÞ

ð4Þ

while the simultaneous hydrolysis of isocyanic acid over a catalyst like titanium oxide anatase gives HNCOðgÞ þ H2 OðgÞ -NH3ðgÞ þ CO2ðgÞ

ð5Þ

ðNH2 Þ2 COðsÞ þ H2 OðgÞ -2NH3ðgÞ þ CO2ðgÞ

ð6Þ

resulting in the following reaction:

Further heating of the products over a suitable catalyst like nickel will initiate the decomposition of ammonia to hydrogen and nitrogen to yield the final reaction ðNH2 Þ2 COðsÞ þ H2 OðgÞ -3H2ðgÞ þ N2ðgÞ þ CO2ðgÞ

ð7Þ

Furthermore, other chemical species like carbon (graphite) and carbon monoxide formed during intermediate or side reactions are considered for determining the thermodynamic equilibrium of the thermohydrolysis of urea.

2.1.2.3

Utilization of Fuel and Oxidant

The utilization of ammonia or urea can be defined in terms of the actual supply and consumption of the fuel or its hydrogen equivalent such that Uf ¼

Fuelconsumed Fuelsupplied

ð8Þ

In the same way, the utilization of air as an oxidant can be witten as a function of the air supply and usage or its oxygen equivalent because air can be presumed to be composed of 79% nitrogen and 21% oxygen by volume Uf ¼

Air consumed Air supplied

ð9Þ

This applies to ion-conducting and proton-conducting cells.

2.1.2.4

Electrochemical Analysis

The examination of SOFCs at closed-circuit conditions requires the characterization of the overpotentials affecting the cell operation. The working potential or voltage of SOFC can be determined as V ¼E

jactan

jactca

jO

jconcan

jconcca

ð10Þ

The models described below are applicable to all types of SOFC under investigation in this study.

2.1.2.4.1

Activation overpotential

This kind of irreversible loss is linked to the electrode kinetics and signifies the difference between the real potential (closed-circuit conditions) and the reversible potential (open-circuit conditions). The implies connection between the current density and the activation overpotential which can be defined using the Butler–Volmer expression [54,55]: 

 azFjact ð1 aÞzFjact exp ð11Þ J ¼ J0 exp RT RT The exchange current density (J0) is a measure of the magnitude of electron activity at the equilibrium potential of the electrode. The value is greatly inclined by the electrode material, structure and other issues such as the temperature of the reaction and the length of triple phase boundary (TPB) [56]. The charge transfer coefficient (a) designates the effect of the electrical potential on the ratio of the forward to the reverse activation blockade. For most electrochemical reactions in fuel cells, this value is assumed as 0.5 [57,58]. In addition, the parameter (z) refers to the number of electrons transferred per mole of fuel. Therefore, the

16

Ammonia

explicit relationship of the activation overpotential is jacty

RT ¼ sinh F

1

J z J0y

!

ð12Þ

Here, y is anode and cathode.

2.1.2.4.2

Ohmic overpotential

The electronic resistance of the fuel cell electrode and connection is typically subtle and is deliberated negligible when compared to the ionic resistance of the electrolyte which can be described by Ohm’s law [58,59]. jO ¼ JδRO

ð13Þ

The electrolyte resistance (RO) is a function of the electrolyte features and is highly reliant on the operational temperature.

2.1.2.4.3

Concentration overpotential

The concentration overpotential accounts for losses experienced by the resistance of the porous electrode to the carriage of gaseous species between the gas channel and the reaction sites at TPB. The mass transport in the electrodes is driven by the diffusion of reacting species because of concentration gradient as well as the infusion triggered by pressure gradient [60,61]. Numerous mass transport models with changing accuracies have been used to capture the effects of concentration overpotential on the performance of fuel cells. It has been demonstrated that the highest accuracy can be achieved using the dusty-gas model (DGM) [60] and the mean transport pore model (MTPM) [62]. The one-dimensional multi-component mass transport model using DGM can be written as ( !) X yi Ni yi Nj Ni 1 dyi dP B0 P þ yi þ ¼ 1þ ð14Þ P j ¼ 1; ja i dx dx RT Deff Deff mmix Deff ik i;j i;k where the effective Knudsen diffusion accounts for the porosity and tortuosity of the electrode such that [61] sffiffiffiffiffiffiffiffiffiffi 2 ϵ 8 RT eff rp Di;k ¼ 3 t p Mi

The effective binary diffusion of chemical species can be obtained using the Chapman–Enskog equation [63] sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 ϵ 1 1 1 eff T3 þ Dij ¼ 0:0018583 t Mi Mj P s2ij OD;ij

ð15Þ

ð16Þ

and the mean characteristic length (sij) of the molecular collision diameter of species i and j is given by si þ sj 2 The collision integral (OD,ij) is a function of temperature and the Lennard-Jones parameter (k/eij) sij ¼

OD;ij ¼

1:06036 0:193 1:03587 1:76474 þ þ þ G0:15610 e0:47635 G e1:52996 G e3:89411 G

ð17Þ

ð18Þ

where kT eij

ð19Þ

pffiffiffiffiffiffiffi ei ej

ð20Þ

G¼ eij ¼

The values of the parameter in Eqs. (16)–(20) have been obtained from Refs. [63–65]. The permeability of the porous electrode is estimated using the Kozeny–Carman relationship [66] B0 ¼

4 ϵ3 rp2 72t ð1

ð21Þ

ϵÞ2

Furthermore, the viscosity of gas mixture can be calculated using the following semi-empirical equation [67]: pffiffiffiffiffiffi P yi mi Mi ffiffiffiffiffiffi mmix ¼ P p yi Mi

ð22Þ

The gas viscosities are obtained from Ref. [63]. As presented by Zhu and Kee [68], the evaluation of the pressure gradient (dP/ dx) across the electrode can be written as
 P Ni i ¼ 1 Deff dP i;k
 ð23Þ ¼    P dx yi B0 P 1 þ i ¼ 1 Deff mRT RT i;k

Ammonia

17

At TPB, the molar flux of the gaseous reactants involved in the electrochemical reaction, namely hydrogen and oxygen, can be related to the current density as [69] Ni ¼

J zF

ð24Þ

where z is equal to two for hydrogen and four for oxygen. Depending on the type of electrolyte (SOFC-O or SOFC-H), the molar flux of water vapor can be calculated using Graham’s law of diffusion which governs the diffusion in gas mixtures [60]: X pffiffiffiffiffiffi Ni Mi ¼ 0 ð25Þ i

The molar fluxes of all other non-reacting species are equal to zero [69]. The set of simultaneous differential equations can be solved numerically using the software to obtain the molar fractions and partial pressures of the reacting species at TPB. Finally, the relationship between the concentration overpotential and the partial pressures can be written as [57,70] For SOFC-O: ! pTPB RT H2 p H2 O ð26Þ ln jconcan ¼ TPB 2F pH2 PH 2O

jconcca

RT p O2 ln TPB ¼ 4F P O2

!

ð27Þ

jconcan

RT p H2 ln TPB ¼ 2F P H2

!

ð28Þ

For SOFC-H:

jconcca

2.1.2.4.4

1 TPB pffiffiffiffiffiffiffi p P RT B H2 O O2 C qffiffiffiffiffiffiffiffiffiA ¼ ln@ 2F TPB PH2 O PO 2 0

ð29Þ

Electrochemical efficiency

The maximum electrochemical efficiency of SOFC can be defined as [71] Zelc ¼

z F E Uf DH0

ð30Þ

Since only hydrogen is oxidized in the fuel cell, its LHV is used in this particular calculation.

2.1.2.5

Illustrative Example

Using the Gibbs energy minimization (GEM) method described earlier, the thermodynamic equilibrium of ammonia is obtained as depicted in Fig. 14. It can be seen that the decomposition of ammonia is favored at higher temperatures and low pressures. Nonetheless, it is well known that the reaction kinetics is rather slow and often requires the use of a catalyst to promote faster conversion of ammonia. This example is derived from Ref. [52]. Complete transformation of ammonia may be realized over expensive metal catalysts such as ruthenium and platinum at temperatures around 600–650K [72]. However, more economical alternatives include iron, nickel or nickel-based catalysts. Fig. 15 shows the open cell voltage (OCV) of the DA-SOFC as a function of fuel utilization. It is permanently desired to have a high fuel utilization to maximize the system efficiency and evade fuel combustion and its emissions in combustion chambers and afterburners. However, higher OCV value is a challenging goal which tends to decay as the fuel utilization is improved. As a general trend, the OCV of SOFC-O is lower than that of SOFC-H and the difference becomes more pronounced at higher fuel utilizations due to the high rate of the oxidation reaction and water vapor formation. The OCV value of DU-SOFC is usually lower than that of DA-SOFC mainly due to the presence of additional chemical species like carbon dioxide and carbon monoxide which reduce the partial pressure of hydrogen. This is shown in Fig. 16. Figs. 17 and 18 indicate that the OCV of the corresponding types of SOFC can be improved by 5–6% when the operating pressure is increased by five times the atmospheric pressure. This modest enhancement is of particular benefit when SOFC is used in an integrated pressurized system. Figs. 19 and 20 show the maximum electrochemical efficiency of the respective fuel cells for a given fuel utilization. In general, the maximum efficiency of SOFC-H is higher than that of SOFC-O due to higher OCV over the range of fuel utilizations. Moreover, DA-SOFC is more efficient than DU-SOFC by 3–8% at high fuel utilization due to a higher average molar fraction of hydrogen under most conditions. From the figures, it can also be seen that operating SOFC at higher pressure improves the OCV which in turn offers a slight enhancement to the maximum electrochemical efficiency. The vertical dashed lines represent the maximum fuel utilization at which the maximum electrochemical efficiency can be achieved.

18

Ammonia

1.0 0.9 NH3 H2

0.8

Molar fraction (–)

0.7 0.6

101 kPa

0.5

505 kPa 1010 kPa

0.4 N2

0.3 0.2 0.1 0.0 300

400

500

600

700

800

900

1000

1100

1200

1300

Temperature (k) Fig. 14 Thermodynamic equilibrium of ammonia. Data from Ishak F. Thermodynamic analysis of ammonia and urea fed solid oxide fuel cells (SOFCs); 2011.

1.1

SOFC-H

OCV (V)

1.0

0.9 SOFC-O 0.8 P = 101 kPa Air utilization = 25%

0.7

1073K 1273K

0.6 0

20

40

60

80

100

Ammonia utilization (%) Fig. 15 Open cell voltage (OCV) of DA-SOFC at different temperatures. Data from Ishak F. Thermodynamic analysis of ammonia and urea fed solid oxide fuel cells; 2011.

The following concluding remarks can be derived from the illustrative example:

• •

•

The proton-conducting ammonia fed SOFC maintained the highest reversible cell potential under all open-circuit conditions followed by the ion conducting counterpart. The ion and proton-conducting urea fed fuel cells attained lower cell potentials due to higher overpotentials Under closed-circuit conditions, the proton-conducting ammonia fed SOFC demonstrated the best performance due to the higher partial pressure of hydrogen at the anode in comparison to the ion-conducting counterpart. In addition, intermediate and side reactions at the anode of ion and proton conducting urea fed SOFC resulted in a strong decrease of the hydrogen partial pressure hence the lower performance. The reverse water gas shift reaction has a harmful consequence on the partial pressure of hydrogen at the anode of urea fed SOFC. Especially, the effect is most marked in the proton-conducting SOFC primarily due to the nonappearance of sufficient water vapor at the anode of the cell.

Ammonia

1.1 SOFC-H

OCV (V)

1.0

0.9

SOFC-O

0.8

P = 101 kPa SUR=1 Air utilization = 25%

0.7

1073K 1273K

0.6 0

20

40

60

80

100

Urea utilization (%) Fig. 16 Open cell voltage (OCV) of DU-SOFC at different temperatures. Data from Ishak F. Thermodynamic analysis of ammonia and urea fed solid oxide fuel cells (SOFCs); 2011.

1.1

SOFC-H

OCV (V)

1.0

0.9 SOFC-O 0.8 P = 101 kPa P = 505 kPa

T = 1073K Air utilization = 25%

0.7

0.6 0

20

40

60

80

100

Ammonia utilization (%) Fig. 17 Effect of pressure on the open cell voltage (OCV) of DA-SOFC. Data from Ishak F. Thermodynamic analysis of ammonia and urea fed solid oxide fuel cells (SOFCs); 2011.

1.1

SOFC-H

OCV (V)

1.0

0.9 SOFC-O 0.8 101 kPa 505 kPa

T = 1073K SUR=1 Air utilization = 25%

0.7

0.6 0

20

40

60

80

100

Urea utilization (%) Fig. 18 Effect of pressure on the open cell voltage (OCV) of DU-SOFC. Data from Ishak F. Thermodynamic analysis of ammonia and urea fed solid oxide fuel cells (SOFCs); 2011.

19

Ammonia

Maximum electrochemical efficiency (%)

20

100 P = 101 kPa P = 505 kPa

90 80

SOFC-H

70 60 50 40

SOFC-O

30 20 T = 1073K Air utilization = 25%

10 0 0

10

20

30

40

50

60

70

80

90 100

Ammonia utilization (%)

Maximum electrochemical efficiency (%)

Fig. 19 Maximum electrochemical efficiency of DA-SOFC. Data from Ishak F. Thermodynamic analysis of ammonia and urea fed solid oxide fuel cells (SOFCs); 2011.

100 P = 101 kPa P = 505 kPa

90 80

SOFC-H

70 60 50 40

SOFC-O

30 20

T = 1073K SUR = 1 Air utilization = 25%

10 0 0

10

20

30

40

50

60

70

80

90 100

Urea utilization (%) Fig. 20 Maximum electrochemical efficiency of DU-SOFC. Data from Ishak F. Thermodynamic analysis of ammonia and urea fed solid oxide fuel cells (SOFCs); 2011.

2.1.3

Case Study 1: Ammonia Utilization in Road Vehicles

In this case study, a comparative life cycle assessment (LCA) of ICE-based vehicles fueled by various fuels, ranging from hydrogen to gasoline, is conducted in addition to electric and hybrid electric vehicles. Three types of vehicles are considered, such as ICE vehicles using gasoline, diesel, LPG, hydrogen, and ammonia; hybrid electric vehicles using gasoline and electricity; and electric only vehicles for comprehensive comparison and environmental impact assessment. The processes are analyzed from raw material extraction to vehicle disposal using LCA methodology. In order to reflect the sustainability of the vehicles, seven different environmental impact categories are considered: abiotic depletion, acidification, eutrophication, global warming, human toxicity, ozone layer depletion, and terrestrial ecotoxicity. The energy resources are chosen mainly conventional and currently utilized options to indicate the actual performances of the vehicles. The results show that electric and hybrid electric vehicles result in higher human toxicity, terrestrial ecotoxicity, and acidification values because of manufacturing and maintenance phases. In contrast, hydrogen vehicles yield the most environmentally benign option because of high energy density and low energy consumption during operation.

2.1.3.1

Materials and Methods

A typical life cycle of a vehicle technology can be categorized into two main steps, namely fuel cycle and vehicle cycle. In the fuel cycle, the processes beginning from the feedstock production to fuel utilization in the vehicle are considered. The extraction of crude petroleum is accounted for diesel. Transformation of crude oil feedstock into useful fuels is a too energy intensive stage of

Ammonia

21

Vehicle maintenance

Vehicle disposal

Vehicle operation

Vehicle manufacturing Transport services

Transport services Energy production Infrastructure

Energy production Raw material

Energy production

Energy transmission

Energy consumption

Production/assembly/ infrastructure and auxiliaries

Fig. 21 Boundaries of conducted LCA analyses including fuel and vehicle cycle.

the fuel cycle, producing substantial amounts of GHG. Nevertheless, purification of natural gas results in considerably fewer energy usage and GHG. The fuel needs to be transported to be available for vehicle usage. Hence, emissions and energy usages associated with fueling trucks/pipelines are thus accounted for in the fuel delivery step [73]. The boundaries of the current analyses are illustrated in Fig. 21 including fuel and vehicle cycles. A remarkable part of any life cycle analysis involves gathering of reliable data. The excellence of data has a deep influence on the quality of the results predicted or estimated by an LCA tool. Argonne National Laboratory has developed a full life cycle model called GREET (Greenhouse gasses, Regulated Emissions, and Energy use in Transportation), sponsored by the U.S. Department of Energy, Office of Energy Efficiency and Renewable Energy, which allows evaluating various vehicle and fuel combinations on a full fuel cycle or vehicle cycle basis. It is a powerful software with a substantial amount of alternative vehicles. It takes into account the well to pump (WTP) and well to wheel (WTW) processes. WTW phases for each fuel starts with the extraction of the primary energy and ends with the consumption in vehicles. WTP consists of the feedstock and fuel phases, counting fuel feedstock extraction, transmission, distribution, and storage. Pump to the wheel (PTW) represents the energy use and emissions during vehicle operation. The functional unit is the distance traveled by the vehicle which is taken as 1 km for each type of vehicle. The results of the primary energy demand and the GHG emissions for each vehicle type are expressed as MJ km 1 and g CO2 eq. km 1, respectively [67]. In this case study, the GREET 2015 software is utilized to simulate the life cycle GHG emissions for the selected vehicles only for the operation process. After the required operation data is obtained from the GREET software, they are used in the SimaPro LCA tool for complete LCA analyses. The scope of the analyses represents a complete LCA since they include the WTW stages as well as the equipment life cycle. The equipment life cycle includes production, manufacturing, maintenance, and end-of-life of vehicle infrastructure. the GREET software can calculate energy use with high accuracy. In this case study, the vehicle operation emissions belonging to different types of vehicles and transportation fuels are based on GREET 2015 calculations [74]. A part of GREET model characterizes the life cycle of vehicles, including production, maintenance, operation, and disposal [75]. In particular, electric and hybrid electric vehicle data are utilized from production until disposal and then adapted in the SimaPro LCA software. The LCA database ecoinvent, v2.2 was used as source of background LCI data. Life cycle impact assessment (LCIA), quantification of life-cycle environmental weights and potential impacts was carried out using the LCA software SimaPro [76]. Impact assessment stage in LCA study is the part in which collected inputs and outputs of fundamental streams are interpreted into impact pointer results typically linked to human health, environment, and resource depletion. It is essential to remind that LCA and impact assessment analyses indicate the possible environmental impacts formed by exchanges that cross the border between technosphere and ecosphere, and act on the natural environment and humans. The results of LCIA are understood as environmentally suitable indicators of potential impact, rather than estimates of real environmental effects. A few standards from the International Organization for Standards (ISO) such as ISO 14040 and ISO 14044 govern the exact requirements necessary to manage LCA studies [77,78]. A guide for the operation of ISO standards was published by the Center of Environmental Science of Leiden University (CML) in 2001 [79]. This guide defines the process to be realized for studying LCA project agreeing to the ISO standards. In the current study, a CML impact assessment methodology is applied. In the impact assessment phase of LCA, a group of impact categories and the characterization methods and factors for a wide list of materials are suggested. For applying the structures in the ecoinvent life cycle inventory database, it is essential to allocate the characterization factors to the elementary source streams and pollutant streams described in the database. The categories considered in this case study are shortly described here. Human toxicity: Toxic substances on the human environment are the main concerns for this category. 1,4-Dichlorobenzene equivalents kg 1 emissions are used to express each toxic substance. Reliant on the material, the geographical scale varies between local and global indicator. Global warming: The greenhouse gasses to air are related to the climate change. Adverse effects upon ecosystem health, human health, and material welfare can result from climate change. A kg carbon dioxide per kg emission is used to express the GWP for time horizon 500 years (GWP500). It has a global scale. Acidification potential: Acidifying substances causes a broad range of impacts on soil, groundwater, surface water, organisms, ecosystems, and materials. SO2 equivalents kg 1 emission is used to expresses the acidification potential.

22

Ammonia

Eutrophication: This category considers the impacts of to extreme intensities of macro-nutrients in the environment initiated by emissions of nutrients to air, water, and soil. It is stated as kg PO4 equivalents per kg emission and the terrestrial measure differs between local and continental scale, the time span is infinity. Depletion of abiotic resources: This impact group is related to protection of human well-being, human health, and ecology health. The indicator is connected to the extraction of minerals and fossil fuels because of involvements in the structure. It is recognized for every extraction of minerals and fossil fuels depending on the concentration of resources and rate of deaccumulation. The terrestrial scope of this indicator is provided at a global scale. It is expressed in kg antimony equivalents/kg extraction unit. Stratospheric ozone depletion: A superior portion of UV-B radiation spreads the earth surface because of increasing chlorofluorocarbons (CFCs). This may yield damaging impacts on human and animal health, terrestrial and aquatic ecosystems, and biochemical cycles and on substances. Ozone depletion potential of several gasses is specified in kg CFC-11 equivalent per kg emission where the time span is infinity. Terrestrial ecotoxicity: This category denotes to influences of toxic substances on terrestrial ecosystems. Ecotoxicity potential is considered with describing fate, exposure and effects of toxic materials. The time horizon is infinite. The results are stated as 1,4dichlorobenzene equivalents per kg emission. The indicator relates at global/continental/regional and local scale. Marine aquatic eco-toxicity: Marine eco-toxicity refers to impacts of toxic substances on marine aquatic ecosystems. It considers each substance emitted to the air, water or/and soil. The unit of this factor is kg of 1,4-dichlorobenzene equivalents (1,4-DB eq.) per kg of emission. Marine sediment ecotoxicity: Marine sediment ecotoxicity refers to impacts of toxic substances on marine sediment ecosystems. The unit of this indicator is kg of 1,4-dichlorobenzene equivalents (1,4-DB eq.) per kg of emission.

2.1.3.2

Systems Description

On average, lifetime performance of a passenger car is assumed to be 239,000 person km. The average utilization factor is expected to be 1.59–1.6 passengers per car. Henceforth, the lifetime of the selected vehicles is approximately 150,000 km. Here, the assessment comprises the following life cycle phases in which the functional unit is 1 km distance traveled.

• • • •

Manufacturing of the vehicle, Operation of the vehicle, Maintenance of the vehicle, Disposal of the vehicle.

2.1.3.2.1

Passenger car manufacturing

The inventory contains processes of energy, water, and material usage in passenger car manufacturing. Rail and road transport of materials are accounted for. The groundwork of plant is involved together with the issues such as land use, building, road, and parking structure. The material consumption reflects a modern vehicle. The data for vehicle production are representative for manufacturing sites with an environmental management system. Thus, the resulting data may be an underestimation of environmental impacts of an average vehicle fleet. The electricity comes from a mixture of Union for the Co‐ordination of Transmission of Electricity (UCTE) countries. The UCTE is the association of transmission system operators in continental Europe in which about 450 million people are supplied electricity. In the electricity usage process, electricity production in UCTE, the transmission network and direct SF6-emissions to air are included. Electricity losses during medium-voltage transmission and transformation from high-voltage are also accounted for. The conversion of high–medium voltage as well as the transmission of electricity at medium voltage are taken into account. In particular, the required high temperature is assumed to be from methane burned in industrial furnace greater than 100 kW which in turn contains fuel feed from the high-pressure gas network, infrastructure (boiler), emissions, and required electricity of operation. In an electric vehicle, the motion is attained from an electric motor and whole energy used for traction is kept in a battery system. The car can trip if sufficient energy is available in the battery. When the battery energy is consumed, the battery needs recharging by electricity grid or replacement. The operation of EV varies from the conventional vehicles in some characteristics as the first difference is the energy source for operation where electricity is utilized despite petrol or diesel. Henceforth, there are no tailpipe emissions. It is therefore assumed that emissions are restricted to tire and brake wear and abrasion from the surface of the road. Conceptually, a HEV is very parallel to EV with the exception that it involves a fuel tank and ICE. Whenever the battery energy is consumed, the ICE can be used to recharge the battery or for traction power. In the current study, HEV is assumed to be 50% electric and 50% gasoline. For electric and hybrid electric vehicles, the required amount of steel is lower compared to conventional cars since there is no ICE in the car. However, for EVs and HEVs, production of electric motor and lithium-ion batteries are included. The battery is the on-board energy storage system of the car. It contains an array of connected cells, the packaging, and the battery management system. If a car is intended to have a range of 100 km with a regular consumption of about 0.2 kWh km 1, a minimum of 20 kWh of energy needs to be stored on-board. When energy density of 0.1 kWh kg 1 is assumed for the battery, the total battery weight would be about 250 kg. Currently, commercially available electric car batteries range between 100 and 400 kg, depending on automobile size and preferred range [80]. The average masses of an electric motor for and

Ammonia

23

lithium-ion battery are assumed to be 104 kg and 312 kg, respectively, for this case study [81]. The inventory data for battery production and disposal is utilized from GREET model [74] and [82].

2.1.3.2.2

Maintenance

The inventory of maintenance of vehicles contains resources used for alternation parts and energy consumption of garages. Rail and road transportation of supplies are accounted for. For EVs, during the lifetime of the car, one battery change is assumed. Henceforth, lithium-ion battery replacement and disposal processes are also taken into account in the maintenance phase.

2.1.3.2.3

Disposal of the vehicles

The inventory of vehicle disposal contains disposal processes for bulk materials. For the disposal of tires, a cut off allocation is applied. In addition, the transportation of tires to the cement works is taken into account. For the disposal of steel, aluminum, copper and tires, a cut off allocation is applied. Waste specific water together with air emissions from incineration and supplementary supply depletion for flue gas scrubbing are accounted for. Short term releases to river water and long-term emissions to groundwater from slag section and remaining material landfill are considered with process energy loads for municipal waste incineration plant (MSWI). The following processes are applied for the disposal of a vehicle scenario:

•

• • •

Disposal of plastics in mixture with 15.3% water to municipal incineration (65 kg). Energy production net output. The waste yields 0.01693 kg of slag and 0.006594 kg of remains per kg of waste. They are landfilled. Supplementary solidifying with 0.002638 kg of cement is applied. MSWI: 3.48MJ kg 1 waste energy in electric form and 7.03MJ kg 1 waste energy in thermal form. Disposal of glass to municipal incineration (30.1 kg). Energy production in MSWI net output: 3.67MJ kg 1 waste electric energy and 7.39MJ kg 1 waste thermal energy. The waste yields 0.01704 kg of slag and 0.01217 kg of residues per kg of waste. They are landfilled. Supplementary solidifying with 0.004869 kg of cement is applied. Disposal of emulsion paint leftovers to HWI (100 kg). Energy production in hazardous waste incineration (HWI) plant, net output: 17.11 MJ kg 1 electric energy and 1.27MJ kg 1 thermal energy. The waste yields 0.707 kg of remains per kg of waste. They are landfilled. Supplementary solidifying with 0.2828 kg of cement is applied. Disposal of zinc in car shredder remains to MSWI (5.89 kg). One kg of this waste produces 0.6244 kg of slag and 0.6202 kg of residues. They are landfilled. Supplementary solidifying with 0.2481 kg of cement is applied.

Note that lithium-ion batteries are recycled for many purposes. The most noticeable one is the retrieval of valued materials and to follow to ecological laws. Numerous methods are present for recycling lithium-ion batteries with diverse environmental consequences. Usually, battery recycling procedures can be expressed in three main categories: mechanical, pyrometallurgical, and hydrometallurgical processes. Hydrometallurgical processes are evaluated to require considerably lesser energy desires compared to pyrometallurgical processes. In this case study, the hydrometallurgical process for disposal of batteries is selected with an average efficiency of 57.5% and energy use of 140 kWh t 1 [80]. The inventory data for the disposal of batteries are taken from [80]. For ammonia and hydrogen fueled vehicles, required amount of steel and electrical energy is a little higher than other cars because of storage tank infrastructure.

2.1.3.2.4

Operation of vehicles

The operation process of the vehicles is one of the key sections of life cycle analyses. In this phase, fuel consumption is involved. Direct airborne emissions of gaseous materials, particulate matters, and heavy metals are accounted for. Particulate emissions cover exhaust- and abrasions emissions. Hydrocarbon emissions contain evaporation. Heavy metal emissions to soil and water produced by tire abrasion are accounted for. The values are based on the operation of an average passenger car. The specific conditions for the selected vehicles are presented herein:

•

• •

Gasoline: All processes on the refinery site excluding the emissions from combustion facilities, including waste water treatment, process emissions and direct discharges to rivers are accounted for. The inventory data also includes the distribution of petroleum product to the final consumer including all necessary transports. Transportation of product from the refinery to the end user is considered together with the operation of storage tanks and petrol stations. Emissions from evaporation and treatment of effluents are accounted for. Particulate emissions cover exhaust- and abrasions emissions. Diesel: Diesel is evaluated as low-sulfur at local storage with an estimation for the total conversion of refinery production to low-Sulfur diesel. An additional energy use (6% of energy consumption for diesel production in the refinery) has been estimated. The other processes are similar to gasoline. Particulate emissions cover exhaust- and abrasions emissions. Hydrogen: Hydrogen is produced during cracking of hydrocarbons. It includes combined data for all processes from raw material extraction until delivery at the plant. The output fractions from an oil refinery are composite combinations of mainly unreactive saturated hydrocarbons. The first processing step in converting such elements into feedstock suitable for the petrochemical industries is cracking. Essentially a cracker achieves two tasks in (i) raising the complexity of the feed mixture into a smaller number of low molecular mass hydrocarbons and (ii) presenting unsaturation into the hydrocarbons to enable more reactivity. The raw hydrocarbon input from the refinery is fed to the heater unit where the temperature is increased. The forming reaction products vary based on the composition of the contribution, the temperature of the heater and the residence time. The cracker operator selects temperature and residence time to enhance product mix from a supplied input. Cracker feeds can be

24

Ammonia

Table 6

Energy consumptions per km for the selected vehicles

Fuel

Fuel/energy consumption

Unit

Gasoline Diesel Hydrogen Ammonia Electric vehicle

0.0649108 0.0551536 0.0195508 0.0926600 0.2167432

kg km 1 kg km 1 kg km 1 kg km 1 kWh km

Hybrid electric vehicle Electric Gasoline Liquefied petroleum gas

0.1083716 0.0324554 0.0576296

kWh km kg km 1 kg km 1

1

1

Source: Data from GREET 2015. Argonne, IL: Argonne National Laboratory; 2015.

•

• • •

naphtha from oil refining or natural gas or a mixture of both. After exiting the heater, the hydrocarbon gas is cooled to prevent extra reactions. After that, it is sent to the separation phase where the individual hydrocarbons are separated from one another by fractional distillation. Particulate emissions cover exhaust- and abrasions emissions. Ammonia: The ammonia synthesis process is Haber–Bosch which is the most common method in the world. Ammonia production requires nitrogen and hydrogen. In this case study, hydrogen is assumed to be from hydrocarbon cracking as explained in the previous paragraph. Cryogenic air separation is mostly used method for the massive amount of nitrogen production. In the LCA of nitrogen production, electricity for the process, cooling water, waste heat and infrastructure for the air separation plant are included. The Haber–Bosch process is employed in this case study. The reaction is facilitated by catalyst (ironoxide based), and the optimal temperature range is 450–6001C. Particulate emissions cover exhaust- and abrasions emissions. EV: Electricity consumption is included. Particulate emissions comprise exhaust and abrasions emissions. Heavy metal emissions to soil and water caused by tire abrasion are accounted for. In the electricity usage process, electricity production mix, the transmission network and direct SF6-emissions to air are included. HEV: Hybrid car is assumed to be 50% electric and 50% gasoline with ICE. Electricity and gasoline consumptions are included. Particulate emissions comprise exhaust and abrasions emissions. Heavy metal emissions to soil and water caused by tire abrasion are accounted for. LPG: All processes on the refinery site excluding the emissions from combustion facilities, including waste water treatment, process emissions and direct discharges to rivers are considered. All flows of materials and energy due to the throughput of 1 kg crude oil in the refinery is accounted for. Refinery data include desalting, distillation (vacuum and atmospheric), and hydrotreating operations. Particulate emissions cover exhaust- and abrasions emissions. The following fuel consumption rates are considered in the analyses as tabulated in Table 6.

2.1.3.3

Case Study 1 Results and Discussion

The selected vehicle types are environmentally assessed in the SimaPro LCA software based on the energy consumption, and GHG emissions of ICEV and EVs obtained from the wheel to wheel simulations using the GREET 2015 model. The results presented here are given on per traveled km basis. The analyzed impact categories are human toxicity, ozone layer depletion, terrestrial ecotoxicity, eutrophication, acidification, and global warming. The energy use is based on units of MJ. The GWP is presented in kg per CO2 equivalent. Fig. 22(A) illustrates the total human toxicity values of all vehicles regarding kg 1,4-DB eq. per travel km. EVs and HEVs have highest human toxicity values corresponding to 0.26 and 0.14 kg 1,4-DB eq. km 1, respectively. Compared to other vehicles, they yield quite higher values because of mainly production and disposal of batteries as shown in Fig. 22(B). The battery production and assembly requires high amounts of copper and aluminum. Henceforth, top two processes contributing issues related human toxicity are copper and aluminum production at plants. The operation of EVs causes only 1% of total human toxicity values as shown in Fig. 22(B). HEVs are quite similar to EVs regarding battery manufacturing which yields second highest value. The depletion of ozone layer is one of the main reasons for climate change which is actually caused by carbon emissions to the atmosphere. Since diesel, gasoline and LPG fuels are fossil based and have a huge amount of carbon substance, they have higher ozone layer depletion values. The highest is equal to 3.3  10 8 kg CFC-11 eq. km 1 for diesel vehicle as Fig. 22(C) represents. The lowest contributions are from ammonia and hydrogen vehicle corresponding to 7.19  10 9 and 1.48  10 9 kg CFC-11 eq. km 1, respectively. Because, there is no direct CO2 emission during operation of ammonia and hydrogen vehicles. In parallel to human toxicity values, EVs and HEVs have a higher environmental impact regarding terrestrial ecotoxicity with 0.00026 and 0.00016 kg 1,4-DB eq. km 1, respectively, because of copper and steel used in the production process as presented in Fig. 23(A). For ammonia vehicle, the highest contributor is steel production process together with lignite and heavy oil burned in power plant.

Human toxicity 500a (kg 1,4-DB eq. km–1)

Ammonia

25

0.3 0.25 0.2 0.15 0.1 0.05 0 Hydrogen Diesel vehicle vehicle

(A)

2% 3% 4%

1% 1%

LPG Gasoline Ammonia Hybrid vehicle vehicle vehicle electric vehicle

5%

Electric vehicle

Copper, primary, at refinery Aluminium, primary, liquid, at plant Anode, aluminium electrolysis

5%

Ferrochromium, high-carbon, 68% Cr, at plant 6%

Disposal, uranium tailings, non-radioactive emissions 73%

Copper, primary, at refinery Disposal, sulfidic tailings, off-site Electric vehicle-operation

Diesel vehicle

Gasoline vehicle

LPG vehicle

Hybrid electric vehicle

Electric vehicle

Ammonia vehicle

(C)

3.50E−08 3.00E−08 2.50E−08 2.00E−08 1.50E−08 1.00E−08 5.00E−09 0.00E+00 Hydrogen vehicle

Ozone layer depletion steady state (kg CFC-11 eq. km–1)

(B)

Fig. 22 (A) Life cycle comparison of human toxicity results for various vehicles, (B) contribution of different processes to human toxicity values of electric vehicles, and (C) life cycle comparison of ozone layer depletion results for various vehicles.

The eutrophication is the impact of excessive levels of macro-nutrients in the environment which is mainly caused by disposal processes. For EVs, HEVs and ammonia vehicles, the main reason of eutrophication is the disposal of spoil from lignite mining in a surface landfill in which it corresponds to about 66%, 49% and 47% for ammonia, EVs and HEVs, respectively. The lowest eutrophication value is observed in hydrogen vehicle with an amount of 7.29  10 5 kg PO4 eq. km 1 as shown in Fig. 23(B). The acidification values of EVs and HEVs are mainly caused by SO2 emission which corresponds to 70% of overall acidification value. The source of SO2 emission is predominantly the lignite and bituminous coal at mine that is used for electricity production mix and eventually consumed in the EVs and HEVs. Afterward, ammonia vehicle has an acidification value of 0.001 kg SO2 eq. km 1 as Fig. 24(A) specifies. 37% of this value is originated from ammonia vehicle operation, and 9% comes from hydrogen production for ammonia synthesis as shown in Fig. 24(B). The GWPs of assessed vehicles are comparatively shown in Fig. 25. The lowest GHG emissions are observed in hydrogen, electric and ammonia vehicles corresponding to 0.049, 0.15 and 0.17 kg CO2 eq. km 1, respectively. Hydrogen consumption is quite lower than ammonia consumption in the passenger car because of higher energy density. It is an expectable result that EVs also yield lower GWP. However production pathway of electricity has a vital role in GHG emissions. If renewable sources can realize electricity production, total emissions would decrease for both EVs and HEVs. In addition, the hydrocarbon route for ammonia vehicle rises the environmental emissions. However, it is noted that renewable-based ammonia is even better than EVs for most of the impact categories.

Ammonia

Terrestrial ecotoxicity 500a (kg 1,4-DB eq. km–1)

26

0.0003

0.00025

0.0002

0.00015

0.0001

0.00005

0 Electric vehicle

Ammonia vehicle

Hybrid electric vehicle

(A)

Gasoline vehicle

LPG vehicle

Diesel vehicle

Hydrogen vehicle

Electric vehicle Ammonia vehicle Hybrid electric vehicle Gasoline vehicle LPG vehicle Diesel vehicle

E− 0

4

04 7.

8.

00

00

E−

E− 04

E−

00 6.

4.

5.

00

E− 00

00 3.

2.

(B)

04

04

04 E−

E− 04 00

00 1.

0.

00

E+

00

E− 04

Hydrogen vehicle

Eutrophication (kg PO4 eq. km)–1

Gasoline vehicle Ammonia vehicle Hybrid electric vehicle LPG vehicle Diesel vehicle Electric vehicle

(C)

00 2 0.

00 18 0.

00 16 0.

00 14 0.

00 12 0.

00 1 0.

00 08 0.

00 06 0.

00 04 0.

00 02 0.

0

Hydrogen vehicle

Abiotic depletion (kg Sb eq. km–1)

Fig. 23 (A) Life cycle comparison of terrestrial ecotoxicity results for various vehicles, (B) life cycle comparison of eutrophication results for various vehicles, and (C) life cycle comparison of abiotic depletion results for various vehicles.

Ammonia

27

Hybrid electric vehicle Electric vehicle Ammonia vehicle Gasoline vehicle LPG vehicle Diesel vehicle

0. 00 16

0. 00 14

0. 00 12

0. 00 1

0. 00 08

0. 00 06

0. 00 04

0

0. 00 02

Hydrogen vehicle

Acidification (kg SO2 eq. km–1)

(A)

7%

Ammonia vehicle−operationhydrocarbon cracking

4%

Hard coal, burned in power plant

9% 37%

Lignite, burned in power plant Hydrogen, cracking, APME, at plant

20%

Heavy fuel oil, burned in power plant 23%

Operation, transoceanic freight ship

(B)

Fig. 24 (A) Life cycle comparison of acidification results for various vehicles and (B) contribution of different processes to acidification values of ammonia vehicles.

Hydrogen vehicle Electric vehicle Ammonia vehicle Diesel vehicle LPG vehicle Hybrid electric vehicle Gasoline vehicle 0

0.05

0.1

0.15

0.2

Global warming 500a (kg CO2 eq.

0.25

0.3

km–1)

Fig. 25 Life cycle comparison of global warming results for various vehicles.

As Fig. 26 shows, the abiotic depletion is mainly caused by the operation processes of vehicles. Manufacturing, maintenance, and disposal of EVs have higher shares during life cycle primarily initiated by production and disposal of lithium-ion batteries.

28

Ammonia

0.0018

Operation

Manufacturing, maintenance, disposal

Abiotic depletion (kg Sb eq. km–1)

0.0016 0.0014 0.0012 0.001 0.0008 0.0006 0.0004 0.0002 0 Ammonia vehicle

Diesel vehicle

Electric vehicle

Gasoline vehicle

Hybrid EV

Hydrogen vehicle

LPG vehicle

Global warming 500a (kg CO2 eq. km–1)

Fig. 26 Contribution of operation, manufacturing, maintenance, and disposal processes of the vehicles to abiotic depletion.

0.25

Operation Manufacturing, maintenance, disposal

0.2

0.15

0.1

0.05

0 Ammonia vehicle

Diesel vehicle

Electric vehicle

Gasoline vehicle

Hybrid electric vehicle

Hydrogen vehicle

LPG vehicle

Fig. 27 Contribution of operation, manufacturing, maintenance, and disposal processes to overall global warming potential.

On-board storage of hydrogen requires high resistant and strength tanks which lead to higher steel and process requirement. Henceforth, non-operation part of hydrogen vehicle constitutes about 22 and 44% of overall hydrogen vehicle life cycle for abiotic depletion and GWP, respectively as illustrated in Figs. 26 and 27. Overall, the operation of the vehicles is dominant contributors to complete life cycle.

2.1.3.4

Case Study 1 Conclusions

Researchers in recent years intensively investigate alternative fuels for transportation sector. The main criteria for a sustainable fuel are being environmentally friendly and profitable. In this case study, a comparative environmental impact assessment of alternative and conventional fueled vehicles is conducted using the cradle to grave approach via life cycle analyses under different environmental impact categories. Conventional vehicles considered in this case study include diesel, gasoline, LPG. Alternative vehicles comprise hydrogen, ammonia, EV and HEV. The analyses are conducted from manufacturing of passenger cars to disposal including the operation of the vehicles. The results show that hydrogen vehicle is the most environmentally benign one in all environmental impact categories. Ammonia as a sustainable and clean fuel has lowest GWP after EVs and yield lower ozone layer depletion values than EVs. Although EVs do not emit direct CO2 during operation, the production and disposal processes of batteries bring some consequences which harm the environment regarding acidification, eutrophication, and human toxicity. It is concluded that to have sustainable and clean transportation, production pathway of vehicles, batteries, and alternative fuels need to be environmentally friendly.

Ammonia

2.1.4

29

Case Study 2: Ammonia Utilization in Maritime Applications

Sea transportation constitutes a large share of global transportation. It is principally used for the transportation of goods, liquid fuels, all types of products and humans. Transoceanic freight ships need a great amount of energy for an operation which is commonly provided by diesel or heavy fuel oils. In order to reduce the total greenhouse gas emissions caused by maritime transportation, alternative fuels, such as ammonia are a potential replacement and/or supplements for conventional fuels. In this case study, zero carbon fuel – ammonia – is proposed to replace heavy fuel oils in the engines of maritime transportation vehicles. Furthermore, it is also proposed to use ammonia as dual fuels to quantify the total reduction of greenhouse gas emissions. An environmental impact assessment of transoceanic freight ship is implemented to explore the impacts of fuel substituting on the environment. In the life cycle analyses, the complete transport life cycle is taken into account from the manufacture of transoceanic freight ship to production, transportation and utilization of ammonia in the maritime vehicles. Several ammonia production routes ranging from municipal waste to geothermal options are considered to evaluate environmentally benign methods comparatively. Besides GWP, environmental impact categories of marine sediment ecotoxicity and marine aquatic ecotoxicity are also selected to examine the diverse effects on marine environment. Being carbon-neutral fuel, ammonia yields significantly minor global warming impacts during operation. The ecotoxicity impacts on maritime environment vary based on the production route of ammonia. The results imply that even ammonia is utilized as dual fuel in the engines, the GWP is quite lower in comparison with heavy fuel oil driven transoceanic ships.

2.1.4.1

Introduction

Decreasing the GWP caused by current transportation technologies and fuels can be reduced significantly by replacing alternative clean fuels. Sea transportation vehicles mostly use heavy fuel oil or diesel fuel for power generation. Ocean freight ships require a massive amount of energy for operation. Ammonia is considered as alternative fuel for power generation especially in transportation sector such as maritime. The usage of ammonia in the maritime applications eventually depends on the capability of producing clean, low-cost energy. One of the hydrogen carriers is ammonia which is synthesized from nitrogen and hydrogen that can be used for direct combustion in maritime vehicles. Besides having significant advantages in storing and transporting compared to hydrogen, ammonia may also be burned directly in diesel engines. Ammonia can be easily transferred through pipelines, railway, and ships by delivering to consumption area where it may be utilized as a source of hydrogen, chemical substance, and fertilizer for agriculture, fuel for transportation such as maritime applications. Since ammonia produces mainly water and nitrogen on combustion, replacing a part of conventional fuel with ammonia will have a large effect in reducing carbon dioxide emissions. Ammonia (NH3) is colorless, pungent gas composed of nitrogen and hydrogen. It is the simplest stable compound of these elements and serves as a starting material for the production of many commercially important nitrogen compounds. Since the improvement potentials of renewable technologies are most likely greater than fossil fuels, it is significant to implement renewable-based alternative fuel production options from the environmental point of view. In this case study, maritime vehicle, freight ship, is driven with ammonia instead of heavy fuel oils in the power engines. Additionally, dual-fuel options, heavy fuel oil, and ammonia are investigated. A comparative LCA of transoceanic freight ship is performed to examine the effects of clean fuel driven maritime vehicles on the environment. The complete transport life cycle is evaluated in the life cycle analyses comprising manufacture of freight ship; operation of freight ship; construction and land use of seaport; operation, maintenance, and disposal of a seaport; and production and transportation of ammonia. Ammonia is produced using renewable resources, namely, biomass, hydropower, municipal waste and geothermal.

2.1.4.2

Pathways for Production of Ammonia

Currently, SMR is a more common method for ammonia production; developing electrolyzers can be used for water electrolysis driven by renewable energy resources. Renewable resources have lower environmental impacts. Therefore, municipal waste, biomass, hydropower and geothermal sources are utilized for ammonia production. For hydrogen production, electrolysis route is employed where an electrolyzer is used with an energy requirement of 53 kWh electricity to generate 1 kg of hydrogen. The source of electricity is taken from municipal waste, geothermal, biomass and hydropower individually for all cases. The chief commercial method of producing ammonia is by the Haber–Bosch process, which involves the direct reaction of elemental hydrogen and elemental nitrogen. N2 þ 3H2 -2NH3

ð31Þ

This reaction requires the use of a catalyst, high pressure about 200 atm and elevated temperature about 4501C. Generally, the catalyst is iron containing iron oxide. The Haber–Bosch process is utilized for ammonia synthesis in this case study. Hydrogen is taken from electrolysis unit and nitrogen is supplied through the air separation process. Cryogenic air separation is typically chosen technique for a huge quantity of nitrogen manufacture. For the LCA of nitrogen manufacture, electrical work for the procedure, cooling water, surplus heat, and groundwork for the air separation facility are taken into account. The distribution elements were attained from the heat of vaporization and the specific heat capacity multiplied with the temperature difference from 201C to the boiling point [76]. Cryogenic air separation process becomes more cost effective compared to non-cryogenic methods at the level of about 200–300 t per day nitrogen. Since gas phase nitrogen is required in the ammonia production

30

Ammonia

reaction, the energy requirement is lower compared to liquid nitrogen because liquefaction is not required. The major input to the air separation plant is electricity required to compress the air. Here, US mix grid electricity is used for air separation plant. Air is not taken as an input because of inexhaustibility. The separated CO2 and water vapor are not evaluated as emissions in the process. The transportation of nitrogen is not accounted for the analyses since it is considered that the ammonia synthesis plant is located near air separation plant. The transportation of ammonia to the port are also considered in the analyses. 0.6 tonne-kilometer (tkm) via diesel driven rail transport and 0.1 tkm lorry transport (higher than 16 t) per kg of ammonia are considered. Producing liquid products from air separation plant requires about two times higher energy than gaseous products. Commercial cryogenic air separation plants require electricity in the range 0.6–1 kWh kg 1 of liquid nitrogen product. However, as mentioned earlier, gaseous product necessitates lower power input. Hence, in this case study, 0.42 kWh electricity is assumed for nitrogen gas production as taken from the GREET 2016 model [83]. The utilized electricity for nitrogen production is US mix grid. For ammonia generation, municipal waste, geothermal, hydropower, and biomass power plant electricity is used from US power plants. For the transportation of the produced ammonia, an average distance is assumed where 100 km is by lorry with a capacity of higher than 16 t and 600 km by rail transport. Dual fuel operation of vessels is also considered in the study as 50% ammonia and 50% heavy fuel oil.

2.1.4.3

Material and Methods

LCA is a powerful method to inspect environmental impacts of a system or process or product. LCA represents a methodical set of processes for assembling and examining the inputs and outputs of materials and energy, and the related environmental impacts, directly assignable to the product or service during the course of its life cycle. A life cycle is the set of phases of a product or service system, from the extraction of natural resources to last removal. Entire life cycle steps from resource extraction to disposal during the lifetime of a product or process are considered in this case study. LCA is a four-step process, namely, goal and scope definition, inventory analysis, impact assessment, improvement potential. The goal of this case study is to explore the environmental effects of ammonia fueled marine transportation ships in comparison with conventional heavy fuel oil from cradle to grave using GWP, abiotic depletion, acidification, stratospheric ozone layer depletion, marine eco-toxicity, and marine sediment ecotoxicity. The motivation behind this case study is to decrease the environmental impacts caused by current hydrocarbon dependent marine transportation systems. The results of this case study will mainly attract marine transportation sector and academicians working in the area of clean fuel production and utilization technologies. The function of this case study assesses environmental impacts per tonne-kilometer cruise travel where the functional unit is 1 tkm. It is assumed that the processes for ammonia production contain production of hydrogen and nitrogen separately and the mass balance is used to identify the amount of hydrogen and nitrogen required for 1 kg of ammonia production. For this analysis, we account for all the stages in the life cycle of maritime transportation, including feedstock recovery and transportation, fuel production and transportation, and fuel consumption in the ocean freight ships. The exploration and recovery activities from the well to fuel production and the subsequent transportation to the pump constitute the WTP stage. The combustion of fuel during ocean vehicle operation constitutes the pump-to-haul (PTH) stage. These two stages combined comprise the well-to-haul (WTH) cycle. There are numerous assessment procedures established over the time to categorize and depict the environmental effects of the system. The method used for the present analysis is CML 2001 which is a method proposed by a set of scientists under the principal of CML (Center of Environmental Science of Leiden University). The system boundaries show the limitations of unit processes which are needed to be included in an LCA study. The system boundaries for the LCA analyses are defined as shown in Fig. 28, namely: mining of the raw materials and extraction of the nutrients from these materials, transportation of raw materials and pre-products, generation, and supply of required energy, manufacturing of the ammonia and the related field operations. Various environmental impact categories including global warming, marine sediment ecotoxicity, marine aquatic ecotoxicity, acidification, ozone layer depletion, and abiotic depletion are selected in order to study the diverse effects of changing to ammonia fuels in sea transportation. The description of the categories is already explained in case study 1.

Ship maintenance

Fuel transport

Port construction Port maintenance

NH3

Ship operation

Heavy fuel oil

Ship manufacturing

Fig. 28 Life cycle steps of maritime transportation.

Port operation Fuel production

Port disposal

Ammonia

Table 7

Cruise RSZ1 RSZ2 Hotel1 Hotel2

31

Distance, speed, and duration of trip for energy requirement of transoceanic freight ship Distance (nmi)

Speed (kn)

Time (h)

Load factor

2306.041 100.827 25 – –

18.006 18.006 18.006 – –

128.073 5.6 1.388 22 22

0.6 0.6 0.6 0.19 0.19

Source: Data from GREET 2016. Argonne, IL: Argonne National Laboratory; 2016.

2.1.4.4

Life Cycle Assessment Phases

Analysis of WTH of ammonia fuels is performed in this case study where each step is briefly explained in this section. The functional unit in the LCA study is 1 tkm marine transportation. A ton kilometer by shipping is defined as unit of measure of goods transport which represent the transport of 1 t by a vessel over 1 km. Marine diesel engines are generally further categorized into two different groups as slow speed (15 knots in average) and medium speed (25–30 knots). Transoceanic freight ships are slow speed vehicles which include two-stroke cycle with crosshead engines of 4–12 cylinders. In the marine industry, these engines are used for main propulsion and constitute a larger portion of installed power on the ship. Each vessel type is characterized by the power rating of its two engines – main and auxiliary. For each engine, fuel consumption and emission factors are defined. The GREET 2016 software is utilized to find the power ratings and total energy consumptions of the selected ships [83]. In order to determine the average power consumptions, the trip is assumed to be from Pacific to international ports. After determining the total trip distances as listed in Table 7, each trip was divided into segments, including transit through reducedspeed zones (RSZs). Each trip segment may have distinct fuel consumption and emission factors owing to different speeds and load factors and engine/fuel switching. At the origin and destination ports, the vessel will leave hotel and burn fuel dockside using mainly auxiliary engines. After a ship leaves port, it travels in an RSZ, during which it uses a lower load factor and consumes less fuel, thereby emitting fewer pollutants, than when traveling at cruising speed. It will pass through an RSZ before hoteling at the port of destination. The GREET marine module aims to model these trip segments, representatively, for different vessel types leaving or arriving at U.S. ports in the various regions [83].

2.1.4.4.1

Transoceanic freight ship manufacture

Transoceanic freight ship manufacture includes the processes of material production, representing the material composition of an average water vehicle used for solid goods transportation. For manufacturing, electricity and heavy oil burned in the industrial furnace are included. For the transportation of materials, standard distances are applied. Also, waste treatment processes for non-metal components of a water vehicle are accounted for. The exchanges of the ship manufacturing are derived from an assessment of a ship, with a load capacity of 51,500 t. The energy consumption in manufacturing is estimated as 50% of the cumulative energy of the used materials. The split of energies is 10% electricity and 90% heavy fuel oil [76].

2.1.4.4.2

Maintenance of transoceanic freight ship

Maintenance of transoceanic freight ship includes the use paint and emissions of the solvent of the paint as NMVOC. Consumption of lubricates is excluded. It is assumed that the ship is painted six times in its entire lifespan of 25 years. The consumption of lubricates is included in the fuel consumption for vessel operation [76].

2.1.4.4.3

Port facilities

Port facilities comprise the construction and disposal of one of the world’s biggest port in Rotterdam, Netherlands. The inventory contains the processes of material production, representing the material used in the construction phase of the harbor. The building activities and electricity consumption are accounted for building and disposal phases. Emission of NMVOC is included. Also, lorry transport of materials to the constructions site are taken into account. The expenditures due to construction and disposal are addressed. The data represent one seaport for both, sea and inland shipping. The material composition of the sealed industrial area is derived from the construction and disposal expenditures of a European highway. The built-up is modeled as a steel building. The lifetime of port is assumed to be 100 years [76].

2.1.4.4.4

Maintenance and operation of the port

Maintenance and operation of the port include emissions to water due to non-removed oil spills. In this step, the land occupation and transformation due to seaport are taken into account. The energy consumption at the port is based on assumptions for the specific electricity consumption at the harbor of Hamburg, Germany. The land use is further distinguished in a built-up area (0.6%), road area (45.6%) and water bodies (54.29%). Emission to waters includes emission from production sites on the port site, which are not directly connected to the transport activities [76].

32

Ammonia

2.1.4.4.5

Operation of the transoceanic freight ship and transport for 1 tkm

In this phase, the full cycle is represented by the supply of the fuel, the operation of the ship and transport of the goods as one tkm. Direct airborne emissions of gaseous substances, particulate matters, dioxins, PAHs, halogens and heavy metals are accounted for. These emissions are caused by heavy oil burning in the engines. Hence they are mostly eliminated in ammonia driven ships. Also, the disposal of bilge oil and emissions of tributyltin compounds are included. Individual hydrocarbons are estimated based on the share of diesel engines of road vehicles. Heavy metals are estimated from trace elements in fuel. A distinction between distilled (28%) and residual fuel (72%) is applied. The amount of disposed bilge oil is estimated as 0.6% of the consumed fuel. The average data for the steam turbine (5%) and diesel engine (95%) propulsion are considered in the study. The fuel used for conventional ships is heavy fuel oil and is representative of slow speed engine types about 15 knots. The data represents solid bulk transport of about 40,000 dwt (deadweight tonnage) where the ship is driven by the steam turbine and diesel engines [76]. The power ratings of the main engine and auxiliary engines are about 37.5 and 8.3 MW, respectively. The average energy consumption for the freight ship is calculated to be 0.214 MJ per mile ton [83]. The entire transport life cycle namely; the operation of the vessel; production of the vessel; construction and land use of port; operation, maintenance, and disposal of the port; production and transportation of fuel to the port are considered. Port infrastructure expenditures and environmental interventions are allocated based the yearly throughput (0.37). The vessel manufacturing is allocated based on the total kilometric performance corresponding to about 2,000,000 km and its transport performance. Since transport activity requires loading and unloading, for each transport activity two ports are needed [76].

2.1.4.5

Case Study 2 Results and Discussion

The life cycle analyses are performed in the SimaPro and GREET 2016 software for the assessment of environmental impacts. Power consumption and emission values for transoceanic freight ships are derived from GREET 2016 based on the trip scenario listed in Table 7. The obtained values are used in the SimaPro LCA software by employing the impact assessment method of CML 2001. The toxic substances on the marine sediment and aquatic environment are the main concerns of marine sediment ecotoxicity and marine aquatic ecotoxicity categories. 1,4-Dichlorobenzene equivalents/tkm is used to express each toxic substance. Among the selected fuels, the conventional heavy fuel oil for freight ship have the greatest damage on marine sediment and aquatic environment as shown in Figs. 29–34. Ammonia driven vehicles where ammonia is produced from geothermal and municipal waste yielded the lowest marine sediment ecotoxicity impact with a value of about 0.0047 kg 1,4-DB eq. tkm 1 for transoceanic freight ship. Using ammonia where it is produced from the municipal waste plant, in the transoceanic ship as dual fuel with heavy fuel oil lowers the marine sediment ecotoxicity level about 49%. The contributions of different processes to ecotoxicity of marine sediment and marine aqua are shown in Figs. 30–32. The operation of the freight ship is responsible for 45% of marine aquatic ecotoxicity as seen in Fig. 32 for sole ammonia fueled ship whereas exploration and offshore production of heavy oil represents 6% and natural gas extraction represents 15% of total. This is due to natural gas, and oil-fired power plants which are then used for nitrogen production plant. For biomass-based ammonia driven freight ship, barium, tributyltin compounds, and vanadium are top three substances causing marine aquatic ecotoxicity as where barium has an impact corresponding to 0.0017 kg 1,4-DB eq. tkm 1. Some of them are related to tributyltin compounds emitted to water because of bottom paintings of the ships. Fig. 30 comparatively shows marine sediment ecotoxicity values of subprocesses for ship powered by sole ammonia from municipal waste whereas Fig. 31 illustrates for dual fuel option (from hydropower). The operation of freight ship, in this case,

Conventional heavy fuel oil Ammonia (hydropower)/heavy fuel oil Ammonia (hydropower) Ammonia (municipal waste/heavy fuel oil dual fuel) Ammonia (municipal waste) Ammonia (geothermal/heavy fuel oil dual fuel) Ammonia (geothermal) Ammonia (biomass/heavy fuel oil dual fuel)

02 0.

25 0. 01 5 0. 01 75

01 0.

01

0.

0. 00 5 0. 00 75

0 0. 00 25

Ammonia (biomass)

Marine sediment ecotox. 500a (kg 1,4-DB eq. tkm–1)

Fig. 29 Marine sediment ecotoxicity values of transoceanic freight ship per ton kilometer (tkm) for ammonia and conventional heavy fuel oil.

Ammonia

Marine sediment ecotoxicity (kg 1,4-DB eq. tkm–1)

0.003

33

Operation, transoceanic freight ship, 0.0028

0.0025

0.002

0.0015 Remaining processes, 0.0012

0.001 Natural gas, unprocessed, at extraction, 0.0005

0.0005

Heavy fuel oil, burned in power plant, 0.0003

0 Fig. 30 Process contributions to marine sediment ecotoxicity values of transoceanic freight ship fueled by sole ammonia from municipal waste.

Operation, transoceanic freight ship Well for exploration and production, offshore 10% 3% 5% 2% 1% 6%

Natural gas, unprocessed, at extraction 4% Heavy fuel oil, burned in power plant Discharge, produced water, onshore 1%

73% 1%

Discharge, produced water, offshore Heavy fuel oil, burned in refinery furnace Heavy fuel oil, burned in industrial furnace 1MW Remaining processes

Fig. 31 Process contributions to marine sediment ecotoxicity values of transoceanic freight ship driven by dual fuel (50% ammonia from hydropower and 50% heavy fuel oil).

causes more than 80% of total ecotoxicity. The remaining contributor is mainly operation/maintenance of the port corresponding to about 14% because of heavy fuel oil and natural gas-fired power plants for the electricity requirement of the port. The operation of the freight ship is responsible for 73% of marine sediment ecotoxicity for ammonia/heavy fuel oil driven ship whereas exploration, and offshore production of heavy oil represents the 10%. This is due to crude oil production which is then used in a heavy fuel oil refinery and transported to heavy fuel oil regional storage to be combusted in the ship. Ecotoxicity level of marine aqua is significantly lower when ammonia from municipal waste, hydropower and geothermal are utilized in the engines as seen in Fig. 32. This validates that using ammonia as a supplementary fuel to heavy oil decreases the total environmental impact significantly. The process contributions to marine aquatic ecotoxicity values of transoceanic freight ship fueled by sole ammonia from geothermal energy are shown in Fig. 33. Regarding GWP, ammonia (from geothermal energy) fueled freight ship yield the lowest greenhouse gas emissions in the entire life cycle corresponding to 0.0043 kg CO2 eq. tkm 1 for freight ship as shown in Fig. 34. However, it is very high for the conventional heavy fuel oil. The highest values after sole heavy fuel oil are 0.0094 kg CO2 eq. tkm 1 for ammonia from biomass and heavy fuel oil combination, respectively, for freight ship. The acidification values of heavy fuel oil driven transoceanic freight ship are mainly caused by SO2 and NOx emissions which correspond to more than 90% of overall acidification value where comparative results are shown in Fig. 35. The source of SO2

34

Ammonia

Conventional heavy fuel oil Ammonia (hydropower)/heavy fuel oil Ammonia (hydropower) Ammonia (municipal waste/heavy fuel oil dual fuel) Ammonia (municipal waste) Ammonia (geothermal/heavy fuel oil dual fuel) Ammonia (geothermal) Ammonia (biomass/heavy fuel oil dual fuel)

8

6

01 0.

0.

01

4

2

01 0.

01 0.

8

01 0.

6

00 0.

4

00 0.

00

2 0.

00 0.

0

Ammonia (biomass)

Marine aquatic ecotox. 500a (kg 1,4-DB eq. t km–1) Fig. 32 Marine aquatic ecotoxicity values of transoceanic freight ship per ton kilometer (tkm) for ammonia and conventional heavy fuel oil.

Operation, transoceanic freight ship 45%

Remaining processes 27%

Well for exploration and production, offshore 6%

Heavy fuel oil, burned in power plant 7%

Natural gas, unprocessed, at extraction 15%

Fig. 33 Process contributions to marine aquatic ecotoxicity values of transoceanic freight ship fueled by sole ammonia from geothermal energy.

Conventional heavy fuel oil Ammonia (hydropower)/heavy fuel oil Ammonia (hydropower) Ammonia (municipal waste/heavy fuel oil dual fuel) Ammonia (municipal waste) Ammonia (geothermal/heavy fuel oil dual fuel) Ammonia (geothermal) Ammonia (biomass/heavy fuel oil dual fuel) Ammonia (biomass) 0

0.002

0.004

0.006

0.008

0.01

0.012

–1

Global warming 500a (kg CO2 eq. tkm )

Fig. 34 Global warming potential (GWP) of transoceanic freight ship per ton kilometer (tkm) for ammonia and conventional heavy fuel oil.

Ammonia

35

Conventional heavy fuel oil Ammonia (hydropower)/heavy fuel oil Ammonia (hydropower) Ammonia (municipal waste/heavy fuel oil dual fuel) Ammonia (municipal waste) Ammonia (geothermal/heavy fuel oil dual fuel) Ammonia (geothermal) Ammonia (biomass/heavy fuel oil dual fuel) Ammonia (biomass) 0

0.00008

0.00016

0.00024

–1

Acidification (kg SO2 eq. tkm )

Fig. 35 Acidification values of transoceanic freight ship per ton kilometer (tkm) for ammonia and conventional heavy fuel oil.

Conventional heavy fuel oil Ammonia (hydropower)/heavy fuel oil Ammonia (hydropower) Ammonia (municipal waste/heavy fuel oil dual fuel) Ammonia (municipal waste) Ammonia (geothermal/heavy fuel oil dual fuel) Ammonia (geothermal) Ammonia (biomass/heavy fuel oil dual fuel) Ammonia (biomass) 0.00E+00

3.00E−05

6.00E−05

9.00E−05 –1

Abiotic depletion (kg Sb eq. tkm )

Fig. 36 Abiotic depletion values of transoceanic freight ship per ton kilometer (tkm) for ammonia and conventional heavy fuel oil.

emission is predominantly the operation of freight ship (96.8%) for conventional heavy fuel oil ships. This is caused by the sulfur content of the heavy fuel oil hence it is mostly eliminated in clean fuels as seen in Fig. 35. The combustion of diesel and heavy fuel oil have the high impact hence, particularly transportation processes with railway and trucks create high emissions leading to higher acidification values. The abiotic sources are natural sources counting energy sources, such as hard coal and crude oil, which are evaluated as nonliving. This is because of fossil fuels are major basis of energy and feed resource, it shows the huge intake of hard coal and lignite for tonne-kilometer travel of ammonia (from biomass) fueled transoceanic freight ship as shown in Fig. 36. The reason of hard coal and lignite consumption is the electricity US mix usage in air separation plant for nitrogen production of ammonia synthesis process. Heavy fuel oil utilization constitutes about 53% of overall abiotic depletion whereas ammonia production from biomass power plant electricity constitutes about 12.3% of total abiotic depletion for ammonia (from biomass)/ heavy fuel oil combination. Also, operation and maintenance of the port have also a high share corresponding to 31.5% of total where it is similarly originated from electricity mix production. Fig. 37 presents the life cycle kg CFC-11 eq. emissions of the freight ship with ammonia and conventional fuel oil per tonnekilometer traveled. It is quite high for heavy fuel oil and dual-fuel options while it is considerably less for ammonia fuel. Particularly, hydropower options have the lowest environmental impact in terms of ozone layer depletion. For ammonia and heavy fuel oil driven transoceanic freight ship, crude oil production has the highest share corresponding to about 81.2% as shown in Fig. 38. The operation and maintenance of the port are responsible for about 16.6% of overall ozone layer depletion whereas the manufacturing of the ship constitutes only 2.2%. Two main substances causing ozone layer depletion are bromotrifluoromethane (Halon 1301) and bromochlorodifluoromethane (Halon 1211) corresponding to about 1.98E 10 and 3.31E 11 kg CFC-11 eq. tkm 1, respectively.

2.1.4.6

Case Study 2 Conclusions

Reducing the total greenhouse gas emissions from marine transportation is possible using ammonia which is carbon-free fuel. They can be utilized for maritime ship engines directly as supplementary fuels or individual fuels. Ammonia fueled ships yield considerably lower global warming impact during operation. The highest GWPs are calculated to be 0.010 kg CO2 eq. tkm 1 for conventional heavy fuel oil, respectively, for transoceanic freight ship. Using ammonia as a dual fuel in the marine engines can

36

Ammonia

Conventional heavy fuel oil Ammonia (hydropower)/heavy fuel oil Ammonia (hydropower) Ammonia (municipal waste/heavy fuel oil dual fuel) Ammonia (municipal waste) Ammonia (geothermal/heavy fuel oil dual fuel) Ammonia (geothermal) Ammonia (biomass/heavy fuel oil dual fuel) Ammonia (biomass) 0.00E+00

4.00E−10

8.00E−10

1.20E−09

1.60E−09

Ozone layer depletion 40a (kg CFC-11 eq. tkm–1) Fig. 37 Stratospheric ozone layer depletion values of transoceanic freight ship per ton kilometer (tkm) for ammonia and conventional heavy fuel oil.

Ozone layer depletion (kg CFC-11 eq. tkm–1)

4.0E−10

Crude oil, at production onshore

3.5E−10 3.0E−10 2.5E−10 2.0E−10 1.5E−10 1.0E−10 5.0E−11

Transport, natural gas, pipeline, long distance

Crude oil, at production

Remaining processes

0.0E+00 Fig. 38 Process contributions to stratospheric ozone layer depletion of transoceanic freight ship by dual fuel (50% ammonia from municipal waste and 50% heavy fuel oil).

decrease total greenhouse gas emissions up to 33.5% (for geothermal-based ammonia) whereas this number increases to 69% if only ammonia (from geothermal) is used in the engines. The LCA study proves that switching to ammonia in maritime transportation reduces the total GHG emissions and other environmental impacts considerably.

2.1.5

Closing Remarks

Ammonia becomes a primary hydrogen carrier that does not contain any carbon atoms and has a high hydrogen ratio. It consists of one nitrogen atom from air separation and three hydrogen atoms from any conventional or renewable resources. Ammonia as a sustainable fuel can be used in all types of combustion engines, gas turbines, burners with only small modifications and directly in fuel cells which is a very significant advantage compared to another type of fuels. It is also an option for cooling the engine with ammonia that can act as a refrigerant while it is heated to the temperature at which it is fed to the power producing unit in the vehicle. Reducing the total greenhouse gas emissions from marine transportation is possible using ammonia which is carbon-free fuel. They can be utilized for maritime ship engines directly as supplementary fuels or individual fuels. Ammonia fueled ships yield considerably lower global warming impact during operation. Ammonia as a sustainable and clean fuel in road vehicles yield also the lowest GWP after electric and hydrogen vehicles. Moreover, the overall thermal and exergy efficiencies of the ion and proton-conducting direct ammonia SOFC range of 70–85% depending on the operating conditions indicating the suitable arrangement of components of the systems.

Ammonia

37

Acknowledgment The authors acknowledge the support provided by the Natural Sciences and Engineering Research Council of Canada.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51]

Veziroglu TN, S¸ahin S. 21st Century’s energy: Hydrogen energy system. Energy Convers Manag 2008;49:1820–31. Jensen JO, Vestbø AP, Li Q, Bjerrum NJ. The energy efficiency of onboard hydrogen storage. J Alloys Compd 2007;446:723–8. Zamfirescu C, Dincer I. Ammonia as a green fuel and hydrogen source for vehicular applications. Fuel Process Technol 2009;90:729–37. Gálvez ME, Halmann M, Steinfeld A. Ammonia production via a two-step Al2O3/AlN thermochemical cycle. 1. Thermodynamic, environmental, and economic analyses. Ind Eng Chem Res 2007;46:2042–6. Kool A, Marinussen M, Blonk H. LCI data for the calculation tool Feedprint for greenhouse gas emissions of feed production and utilization. GHG Emissions of N, P and K Fertilizer Production Blank Consultants Gravin Beatrixstraat 2012;34. Ganley JC, Holbrook JH, McKinley DE. 2007. Solid state ammonia synthesis. In: Ammonia fuel network conference; 2007. Gilbert P, Thornley P. Energy and carbon balance of ammonia production from biomass gasification. In: Bio-ten; 21 Sep 2010-23 Sep 2010; Birmingham: Holiday Inn; 2010. Anderson K, Bows A, Mander S. From long-term targets to cumulative emission pathways: reframing UK climate policy. Energy Policy 2008;36:3714–22. Thomas G, Parks G. Potential roles of ammonia in a hydrogen economy. Washington, DC: US Department of Energy; 2006. Lan R, Tao S. Ammonia as a suitable fuel for fuel cells. Front Energy Res 2014;2:35. Hacker V, Kordesch K. Ammonia crackers. In: Vielstich W, Gasteiger HA, Lamm A, editors. Handbook of fuel cells. Chichester: John Wiley & Sons, Ltd.; 2010. doi:10.1002/9780470974001.f302011. Home – Industry Overview – PotashCorp. Available From: http://www.potashcorp.com/overview/introduction/company/potashcorp-profile; 2017 [accessed 07.01.17]. Bomelburg HJ. Use of ammonia in energy-related applications. Plant/Operations Prog 1982;1:175–80. Schlapbach L, Züttel A. Hydrogen-storage materials for mobile applications. Nature 2001;414:353–8. Feibelman PJ, Stumpf R. Comments on potential roles of ammonia in a hydrogen economy – a study of issues related to the use of ammonia for on-board vehicular hydrogen storage. Sandia National Laboratories; 2006. Christensen CH, Johannessen T, Sørensen RZ, Nørskov JK. Towards an ammonia-mediated hydrogen economy? Catal Today 2006;111:140–4. Blarigan P Van. Advance internal combustion engine research. In: Proc. 2000 DOE Hydrogen Program Review; 2000. p. 1–19. McFarlan A. Development of direct ammonia fuel cells for efficient stationary CHP applications. Proceedings of the ammonia–sustainable, emission free fuel conference, October. San francisco, CA; 2007. Olson N. BioAmmonia–a comparison with other biofuels. In: Proc. 2007 annual NH3 fuel conference. Iowa Energy Center, San Francisco; 2007. Lide DR. CRC handbook of chemistry and physics: a ready-reference book of chemical and phyical data. Boca Raton, FL: CRC Press; 1990. Elmøe TD, Sørensen RZ, Quaade U, et al. A high-density ammonia storage/delivery system based on Mg(NH3)6Cl2 for SCR–DeNOx in vehicles. Chem Eng Sci 2006;61:2618–25. Zhu L, Wang S, Gu J. Performance investigation of a thermal‐driven refrigeration system. Int J Energy Res 2008;32:939–49. Park YM, Sonntag RE. A preliminary study of the kalina power cycle in connection with a combined cycle system. Int J Energy Res 1990;14:153–62. Hilali ˙I, Söylemez MS. On the optimum sizing of exhaust gas-driven automotive absorption cooling systems. Int. J. Energy Res 2008;32:655–60. Zamfirescu C, Dincer I. Environmentally-benign hydrogen production from ammonia for vehicles. In: Proc. global conference on global warming; 2008. p. 6–10. Zamfirescu C, Dincer I. Using ammonia as a sustainable fuel. J Power Sources 2008;185:459–65. Zamfirescu C, Dincer I. Ammonia as a green fuel for transportation. In: ASME 2008 2nd Int. conference on energy sustainability, vol. 1, ASME; 2008. p. 507–15. doi:10.1115/ES2008-54328. Williams OM, Carden PO. Energy storage efficiency for the ammonia/hydrogen-nitrogen thermochemical energy transfer system. Int J Energy Res 1979;3:29–40. Williams OM, Carden PO. Ammonia dissociation for solar thermochemical absorbers. Int J Energy Res 1979;3:129–42. Christensen CH, Sorensen RZ, Johannessen T, et al. Metal ammine complexes for hydrogen storage. J Mater Chem 2005;15:4106–8. Appl M. Ammonia: principles and industrial practice. Weinheim: Vch Verlagsgesellschaft Mbh; 1999. Strickland GG. 1980. Ammonia as a hydrogen energy-storage medium. Present. 5th annual thermal storage meeting, McLean, VA; 1980. Metkemeijer R, Achard P. Ammonia as a feedstock for a hydrogen fuel cell; reformer and fuel cell behaviour. J. Power Sources 1994;49:271–82. Waitzman DA. Ammonia from coal: a technical/economic review. Process technology and flow sheets. In: Cavaseno V, editor. Chemical engineering. New York: McGrawHill Publ. Co; 1979. p. 63–5. Skodras G, Kaldis S, Topis S, et al. NH3 decomposition and simultaneous H2 separation with a commercial Pd–Cu–Ag/V membrane. In: Proc. Second Int. Green Energy Conference; 2006. p. 25–9. Vitse F, Cooper M, Botte GG. On the use of ammonia electrolysis for hydrogen production. J Power Sources 2005;142:18–26. Canadian Industry Program for Energy Conservation. Benchmarking energy efficiency and carbon dioxide emissions, Canada; 2008. Ammonia|Industrial Efficiency Technology Database; Measures. Available From: http://ietd.iipnetwork.org/content/ammonia#key-data; 2017 [accessed 7.01.17]. Olson NK, Holbrook J. NH3 – the other hydrogen. Richland, WA: Ammonia Fuel Network; 2008. Dincer I, Zamfirescu C., editors. 2011. Ammonia as a potential substance. In: Sustainable energy systems and applications. Boston, MA: Springer US; p. 203–32. doi:10.1007/978-0-387-95861-3_7. Heldebrant DJ, Karkamkar A, Linehan JC, Autrey T. Synthesis of ammonia borane for hydrogen storage applications. Energy Environ Sci 2008;1:156–60. PotashCorp Integrated Annual Report. Annual Integrated Report. Available From: http://www.potashcorp.com/irc/nitrogen; 2015 [accessed 07.01.17]. International Energy Agency. Energy Technology Perspectives 2012; Pathways to a Clean Energy System, France. ISBN: 978-92-64-17488-7. Dossat RJ, Horan TJ. Principles of refrigeration. Upper Saddle River, NJ: Prentice Hall; 2002. Yin SF, Xu BQ, Zhou XP, Au CT. A mini-review on ammonia decomposition catalysts for on-site generation of hydrogen for fuel cell applications. Appl Catal A Gen 2004;277:1–9. Klerke A, Christensen CH, Norskov JK, Vegge T. Ammonia for hydrogen storage: challenges and opportunities. J Mater Chem 2008;18:2304–10. Sakintuna B, Lamari-Darkrim F, Hirscher M. Metal hydride materials for solid hydrogen storage: a review. Int J Hydrogen Energy 2007;32:1121–40. Marder TB. Will we soon be fueling our automobiles with ammonia–borane? Angew Chemie Int Ed 2007;46:8116–8. Best Available Techniques for Pollution Prevention and Control in the European Fertilizer Industry PRODUCTION OF AMMONIA. Brussels, Belgium. 2000. Ammonia: OSH Answers. Can. Cent. Occup. Heal. Saf. Available From: http://www.ccohs.ca/oshanswers/chemicals/chem_profiles/ammonia.html; 2017 [accessed 07.01.17]. Putnam DF. Composition and concentrative properties of human urine Report NASA CR-1802. Washington, DC: National aeronautics and space administration; 1971.

38

Ammonia

[52] Ishak F. Thermodynamic analysis of ammonia and urea fed solid oxide fuel cells; Master of applied science (MASc) dissertation, University of Ontario Institute of Technology; 2011. [53] EG&G Technical Services I. Fuel cell handbook. Morgantown, West Virginia: US Department of Energy; 2004. [54] Barbir F. PEM fuel cells: theory and practice. London: Academic Press; 2013. [55] Mench MM. Fuel cell engines. Hoboken, NJ: John Wiley & Sons; 2008. [56] Ni M, Leung DYC, Leung MKH. Electrochemical modeling and parametric study of methane fed solid oxide fuel cells. Energy Convers Manag 2009;50:268–78. [57] Arpornwichanop A, Patcharavorachot Y, Assabumrungrat S. Analysis of a proton-conducting SOFC with direct internal reforming. Chem Eng Sci 2010;65:581–9. [58] Chan S, Low C, Ding O. Energy and exergy analysis of simple solid-oxide fuel-cell power systems. J Power Sources 2002;103:188–200. [59] Singhal SC, Kendall K. High-temperature solid oxide fuel cells: fundamentals, design, and applicatons. Oxford, UK: Elsevier Advanced Technology; 2003. [60] Suwanwarangkul R, Croiset E, Fowler MW, et al. Performance comparison of Fick’s, dusty-gas and Stefan–Maxwell models to predict the concentration overpotential of a SOFC anode. J Power Sources 2003;122:9–18. [61] Ho CK, Webb SW, Stephen W. Gas transport in porous media. Berlin: Springer; 2006. [62] Janardhanan V. A detailed approach to model transport, heterogeneous chemistry, and electrochemistry in solid-oxide fuel cells. Karlsruhe: Universitätsverlag; 2007. [63] Bird RB, Robert B, Stewart WE, Lightfoot EN. Transport phenomena. New York: J. Wiley; 2007. [64] Wu M, Wang M, Liu J, Huo H. Assessment of potential life-cycle energy and greenhouse gas emission effects from using corn-based butanol as a transportation fuel. Biotechnol Prog 2008;24:1204–14. [65] Poling Bruce E, Prausnitz John M, O’Connell John P. Properties of gases and liquids. New York: McGraw-Hill Education; 2001. [66] Zhu H, Kee RJ, Janardhanan VM, Deutschmann O, Goodwin DG. Modeling elementary heterogeneous chemistry and electrochemistry in solid-oxide fuel cells. J Electrochem Soc 2005;152:A2427. [67] Li M, Zhang X, Li G. A comparative assessment of battery and fuel cell electric vehicles using a well-to-wheel analysis. Energy 2016;94:693–704. [68] Zhu H, Kee RJ. A general mathematical model for analyzing the performance of fuel-cell membrane-electrode assemblies. J Power Sources 2003;117:61–74. [69] Ni M, Leung DYC, Leung MKH. An improved electrochemical model for the NH3 fed proton conducting solid oxide fuel cells at intermediate temperatures. J Power Sources 2008;185:233–40. [70] Patcharavorachot Y, Paengjuntuek W, Assabumrungrat S, Arpornwichanop A. Performance evaluation of combined solid oxide fuel cells with different electrolytes. Int J Hydrogen Energy 2010;35:4301–10. [71] Demin A, Tsiakaras P. Thermodynamic analysis of a hydrogen fed solid oxide fuel cell based on a proton conductor. Int J Hydrogen Energy 2001;26:1103–8. [72] Chein R-Y, Chen Y-C, Chang C-S, Chung JN. Numerical modeling of hydrogen production from ammonia decomposition for fuel cell applications. Int J Hydrogen Energy 2010;35:589–97. [73] Rose L, Hussain M, Ahmed S, et al. A comparative life cycle assessment of diesel and compressed natural gas powered refuse collection vehicles in a Canadian city. Energy Policy 2013;52:453–61. [74] GREET 2015. Argonne, IL: Argonne National Laboratory; 9700 S. Cass Avenue, Building 362, Argonne, IL 60439-4844 USA 2015. [75] Archsmith J, Kendall A, Rapson D. From cradle to junkyard: assessing the life cycle greenhouse gas benefits of electric vehicles. Res Transp Econ 2015;52:72–90. [76] Consultants P. SimaPro life cycle analysis version 7.2 (software). PRé Sustainability, Stationsplein 121, 3818 LE Amersfoort, The Netherlands. [77] ISO 14040: Environmental management – Life cycle assessment – Principles and framework. Available From: http://www.iso.org/iso/catalogue_detail?csnumber=37456; 2006 [accessed 07.01.17]. [78] ISO 14044:2006; Environmental management – Life cycle assessment – Requirements and guidelines. Available From: http://www.iso.org/iso/catalogue_detail? csnumber=38498; 2006 [accessed 07.01.17]. [79] Guinée JB, Heijungs R, Huppes G, et al. Life cycle assessment: past, present, and future. Environ Sci Technol 2011;45:90–6. [80] Duce AD, Egede, P, Öhlschläger G, et al. eLCAr – Guidelines for the LCA of electric vehicles. Report from Project. E-Mobility Life Cycle Assessment Recommendations funded within European Union Seventh Framework Programme. Switzerland: RWTH and EMPA; 2013. [81] Leuenberger M, Frischknecht R. Life cycle assessment of battery electric vehicles and concept cars. Report, Uster, Switzerland: ESU-Services Ltd; 2010. [82] Boyden A. The environmental impacts of recycling portable lithium-ion batteries; [BSc Thesis]. Australian National University; 2014. [83] GREET 2016. Argonne, IL: Argonne National Laboratory; 9700 S. Cass Avenue, Building 362, Argonne, IL 60439-4844 USA2016.

Further Reading Appl M. Ammonia, 3. Production plants. In: Ullmann’s encyclopedia of industrial chemistry. Weinheim, Germany: Wiley-VCH Verlag; 2011. http://dx.doi.org/10.1002/14356007. o02_o12 Appl M. Complete ammonia production plants. Ammonia: principles and industrial practice. Wiley-VCH Verlag GmbH; 2007. p. 177–204. http://dx.doi.org/10.1002/ 9783527613885.ch05 Appl M. Future perspectives. In: Ammonia: principles and industrial practice. Weinheim: Wiley-VCH Verlag GmbH; 2007. p. 245–9. Bartels J.R. A feasibility study of implementing an ammonia economy [Graduate theses and Dissertations]. Iowa State University; 2008. Bicer Y, Dincer I, Vezina G, Raso F. Impact assessment and environmental evaluation of various ammonia production processes. Environ Manage 2017;59(5):842–55. doi:10.1007/s00267-017-0831-6. Bicer Y, Dincer I, Zamfirescu C, Vezina G, Raso F. Comparative life cycle assessment of various ammonia production methods. J Clean Prod 2016;135:1379–95. http://dx.doi. org/10.1016/j.jclepro.2016.07.023 Bicer Y, Dincer I. Performance assessment of electrochemical ammonia synthesis using photoelectrochemically produced hydrogen. Int J Energy Res. 2017; doi:10.1002/ er.3756. Christensen CH, Johannessen T, Sørensen RZ, Nørskov JK. Towards an ammonia-mediated hydrogen economy? Catal Today 2006;111:140–4. Dincer I., Zamfirescu C. Methods and apparatus for using ammonia as sustainable fuel, refrigerant and nox reduction agent. Patent Nr. US20110011354 A1; 2009. Duynslaegher C, Contino F, Vandooren J, Jeanmart H. Modeling of ammonia combustion at low pressure. Combust Flame 2012;159:2799–805. Feng T, Lü L. The characteristics of ammonia storage and the development of model-based control for diesel engine urea-SCR system. J Ind Eng Chem 2015;28:97–109. Garagounis I, Kyriakou V, Skodra A, Vasileiou E, Stoukides M. Electrochemical synthesis of ammonia in solid electrolyte cells. Front Energy Res 2014;2:1. Jennings JR. Catalytic ammonia synthesis: fundamentals and practice. New York. US: Springer; 1991. Kim K, Yoo C-Y, Kim J-N, Yoon HC, Han J-I. Electrochemical synthesis of ammonia from water and nitrogen catalyzed by nano-Fe2O3 and CoFe2O4 suspended in a molten LiCl-KCl-CsCl electrolyte. Korean J Chem Eng 2016;33:1777–80. Ouadha A, El-Gotni Y. Integration of an ammonia-water absorption refrigeration system with a marine diesel engine: a thermodynamic study. Procedia Comput Sci 2013;19:754–61. Vitse F, Cooper M, Botte GG. On the use of ammonia electrolysis for hydrogen production. J Power Sources 2005;142:18–26. Zamfirescu C, Dincer I. Utilization of hydrogen produced from urea on board to improve performance of vehicles. Int J Hydrogen Energy 2011;36:11425–32.

Ammonia

Relevant Websites http://www.ammoniaenergy.org Ammonia Energy. https://ammoniaindustry.com/ Ammonia Industry. https://www.ohio.edu/engineering/ceer/ammonia-electrolysis.cfm Center for Electrochemical Engineering Research (CEER) – Ohio University. https://chemengineering.wikispaces.com/Ammonia+production ChemEngineering. https://www.canada.ca/en/health-canada/services/publications/healthy-living/guidelines-canadian-drinking-water-quality-guideline-technical-document-ammonia.html Government of Canada. http://www.nh3fuel.com/ Hydrofuel Inc. https://www.ipni.net/ International Plant Nutrition Institute (IPNI). http://www.nautilus.org/napsnet/napsnet-policy-forum/monia-as-a-fuel-for-passenger-vehicles-possible-implications-for-greenhouse-gas-reduction-in-korea/ Nautilus Institute. https://www.health.ny.gov/environmental/emergency/chemical_terrorism/ammonia_tech.htm New York State. http://www.nh3car.com/ NH3CAR.com. https://www.nh3fuelassociation.org/ NH3 Fuel Association. https://www.worldfertilizer.com/ Palladian Publications Ltd. http://www.potashcorp.com/ PotashCorp. http://www.spg-corp.com/clean-energy-power-generation.html Space Propulsion Group (SPG). http://www.linde-engineering.com/en/process_plants/hydrogen_and_synthesis_gas_plants/gas_products/ammonia/index.html The Linde Group. https://www.thyssenkrupp-industrial-solutions.com/en/products-and-services/fertilizer-plants/ammonia-plants-by-uhde/ammonia-plants-500mtpd/the-uhde-ammonia-processes/ Thyssenkrupp AG. http://www.titech.ac.jp/english/about/stories/ammonia_synthesis.html Tokyo Institute of Technology. http://www.uni-regensburg.de/chemistry-pharmacy/inorganic-chemistry-korber/index.html Universität Regensburg. https://wcroc.cfans.umn.edu/displacing-diesel-fuel West Central Research and Outreach Center – University of Minnesota. http://www.yara.com/ Yara.

39

2.2 Carbonaceous Materials Yinghuai Zhu, Macau University of Science and Technology, Avenida Wai Long, Taipa, Macau, China Shanmin Gao, Ludong University, Yantai, China Narayan Hosmane, Northern Illinois University, DeKalb, IL, United States r 2018 Elsevier Inc. All rights reserved.

2.2.1 Introduction 2.2.2 Carbonaceous and Boronated Materials for Electronic Applications 2.2.2.1 Zero-Dimensional Carbonaceous Materials for Electronic Applications 2.2.2.2 One-Dimensional Carbonaceous Materials for Electronic Applications 2.2.2.3 Two-Dimensional Carbonaceous Materials for Electronic Applications 2.2.2.4 Three-Dimensional Carbonaceous Materials for Electronic Applications 2.2.2.5 Application of Borocarbonaceous Materials for Supercapacitors 2.2.2.6 Application of Borocarbonaceous Materials in Thermoelectric Energy Conversion 2.2.2.7 Application of Borocarbonaceous Materials in Photovoltaic Devices and Fuel Combustion 2.2.2.8 Application of Borocarbonaceous Materials in Nuclear Reactor 2.2.3 Application of Carboraneous Materials in Oil Adsorption 2.2.3.1 Graphite and Graphene-Based Sorbents 2.2.3.2 Carbon Nanotube-Based Sorbents 2.2.3.3 Carbon Aerogel-Based Sorbents 2.2.3.4 Carbon Nanotube-Based Sorbents 2.2.3.5 Carbon Aerogel-Based Sorbents 2.2.4 Carbonaceous Materials and Boron Hydrides for Hydrogen Storage 2.2.4.1 Graphene-Based Materials for Hydrogen Storage 2.2.4.2 Heteroatoms-Doped Carbon Materials for Hydrogen Storage 2.2.4.3 Transition Metal Nanoparticles-Modified Carbonaceous Materials for Hydrogen Storage 2.2.4.4 Applications of Borohydrides in Hydrogen Generation and Storage 2.2.5 Conclusions and Future Directions Acknowledgment References Further Reading

Nomenclature A/g Barn BET B80 BN BNC C

CVD CNTs DOE

2.2.1

Ampere (electric current unit)/gram Neutron cross section unit Brunauer–Emmett–Teller Boron-80 Boron nitride Boron and nitrogen enriched carbons e0erA/d, where C is the capacitance, in farads; e0 is the electric constant (E 8.854  10 12 F/m); er is the dielectric constant of the material; A is the area of the capacitor, in m2; d is the distance between the capacitor, in meters Chemical vapor deposition Carbon nanotubes US Department of Energy

EDLC F/g GPa GHSC GO K kW/kg PAN PEN PMMA PEMFC PVA S/cm

40 41 41 44 45 49 50 51 52 52 53 54 56 57 57 58 59 59 60 61 62 64 65 65 71

Electric double layer capacitor Farad (capacitance unit)/gram Gigapascal, tensile modulus unit Gravimetric hydrogen storage capacity Graphene oxide Kelvin, temperature unit Power density unit (specific power), kilowatt/ kilogram Polyaniline Polyethylene naphthalene Polymethyl merchacrylate Proton exchange membrane fuel cells Polyvinyl alcohol Siemens per centimeter, conductivity unit

Introduction

Carbon is a well known element with the unique electronic configuration of 1s2 2s2 2p2. It has four valence electrons which can form four identical covalent bonds by four degenerate hybridized sp3 orbitals. The carbon center can bond to many heteroatoms, such as N, S, O, Cl, Br, and P, to form thermodynamically stable compounds. The resulting carbon-hetroatom or carbon–carbon bonds can be single, double, or triple bonds. Carbon is the basis of life on Earth. Most of the carbon-based molecules, around

40

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00202-9

Carbonaceous Materials

41

16 millions, are classified as organic compounds. Carbon also forms various polymers such as polyethylene, polypropylene, and polystyrene, by forming carbon–carbon bonds. In addition to the organic molecules and polymers, carbon exists in major allotropic forms of diamond, graphite, and amorphous carbon. Diamond is one of the world’s hardest, strongest material due to its microscopic structure comprising four covalent C–C single bonds in a giant lattice structure. In contrast, graphite is soft and brittle due to its honeycomb structure and weak interlayer forces. Amorphous carbon, such as soot and carbon black, does not have crystalline structures due to the absence of long-range pattern of carbonaceous atomic positions. Carbonaceous materials have unique diversity of structures such as ball-shaped fullerenes, cylindrical single- (multi-) walled carbon nanotubes, graphenes and mesoporous carbons, and thus they have been widely used in our daily life. For example, in water and air purification machines, activated carbons are used as filters to remove impurities and contaminants by chemical adsorptions to generate pure water and air. While carbon foam is fireproof and used as fire insulators, it is also an excellent thermal insultor. On the other hand, carbon black is used as a main reinforcing filler in rubber products, especially in tires. Advanced polymer composites, reinforced by carbon fibers, have been used in aerospace, automotive and in sporting equipments. The energy storage technology is crucial for efficient utilization of renewable energy from sustainable sources such as wind power. Supercapacitors, also called electric capacitators, posses high power and energy density with reasonable life cycle and these properties make supercapacitors a type of potentially promising energy-storage devices when compared to those of relatively lower power density batteries [1]. Supercapacitor can be classified as electric double layer capacitors (EDLCs) and pseudo-capacitors. Carbonaceous materials have been widely investigated as the excellent electrode composites due to their diversity of structures that generally provide extremely high surface areas and high electrical conductivity. Therefore, it is the material of great interest in the development of supercapacitors. In the energy storage mechanism, supercapacitors are different from commonly used battery. Supercapacitors store and release energy through the interaction of the electrode interface and a electrolye, wheareas the batteries store and release energy via a chemical redox reaction [2,3]. Capacitance (C) can be calculated based on a formula of C ¼ e0erA/d, where C is the capacitance, in faraday; e0 is the electric constant (E 8.854  10 12 F/m); er is the dielectric constant of the material; A is the area of the capacitator, in m2; d is the distance between the capacitators, in meters. Theoretically, the capacitance of the EDLCs is proportional to the electrode surface area, therefore the EDLCs generally possess high capacitance due to their high surface area and great electrical properties of the carbonaceous electrode materials [4–6]. Carbons can be classified as zero-, one- two- and three-dimensional types according to their structure. Zero-dimensional carbonaceous materials include carbon onions and buckyballs such as C60, C70, etc. Nanotubes include single-walled and multiwalled carbon nanotubes representing one-dimensional carbons. While graphene is a typical two-dimensional carbon network, the three-dimensional carbonaceous materials are of activated carbons, carbides and template of carbons as shown in Fig. 1 and Table 1 [7,8]. Recently, various carbonaneous materials, such as activated carbons, graphenes and carbon nanotubes (CNTs), have been investigated as electrode materials to examine their energy storage capability. This section summarizes the latest development of the carbonaceous electrode materials for energy storage applications and thus provides the fundamental insight and offers important guidelines for future trends.

2.2.2 2.2.2.1

Carbonaceous and Boronated Materials for Electronic Applications Zero-Dimensional Carbonaceous Materials for Electronic Applications

Carbon onions and buckyballs, such as fullerene C60, are model compounds of zero-dimensional carbon network. Carbon onions, also called carbon nano-onions or onion-like carbon, are relatively new carbon nanoforms in comparison with fullerene C60, carbon nanotubes and other analogous structures. The high electrical conductivity and larger surface area make them potentially suitable for diverse applications, including energy storage and biomedical imaging [9,10]. The chemical and physical characteristics of the carbon onions are highly dependent on their synthetic methodologies [11–47]. For example, thermal annealing of nanodiamond powders at an extremely high temperature of greater than 17001C has been considered a practical approach to synthesize carbon onions due to high purity of the product, small and relatively narrow size distribution of the resulting particles, as well as precursors of low cost [48–50]. In general, the carbon onions are smaller than 10 nm, while up to 100 nm is also possible. Their shapes are flexible, varying from spherical to polyhedral. The surface area is changeable based on the synthetic conditions. So far, the largest area of 984 m2/g has been achieved by an arc-discharge method from graphite precursors [51]. Similar to other carbon materials, the intrinsic electrical conductivity of the carbon onions play a key role for the electrical conductivity of the carbon onion electrodes. By tuning the synthetic conditions, such as annealing temperature, the intrinsic arrangements of the carbon atoms can be adjusted to change its electrical conductivity. The highest conductivity of around 4 S/cm has been reached, and that is comparable to that of carbon black whose conductivity is about 1–2 S/cm; and much higher than that of activated carbon whose conductivity is less than 0.5 S/cm [10]. As a conductive material, carbon onions have been used for supercapacitator (electrical double layered capacitators) either as the primary material or additives, as well as supports for the pseudo-capacitors. As indicated above, the synthetic conditions determine the electrochemical properties of the resulting carbon onions. It has been found that the conductivity, potential capacitance and electrochemical stability of the onions are highly dependent on the reaction temperature. A higher temperature (415001C) usually generate the material of higher conductivity, potential capacitance and electrochemical stability. The resulting onions can be treated with KOH to activate the material and thus produce extremely large surface area and higher capacitance. A capacitance of 115 F/g has been reported by using the KOH

Carbonaceous Materials

42

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

Fig. 1 Some structures of carbon: (A) diamond, (B) graphite, (C) lonsdaleite, (D)–(F) fullerenes (C60, C540, C70), (G) amorphous carbon, and (H) carbon nanotube. Available from: https://en.wikipedia.org/wiki/Carbon; 2016 [accessed 20.11.16].

activated carbon onions in 2M KNO3 media, and that is comparable to the activated carbon-based capacitator with a capacitance of greater than 100 F/g [52]. Owing to their moderate surface area (o100 m2/g), carbon onions-based supercapacitators generally demonstrated relatively low energy storage capacity when compared to nanoporous carbons which have a typical capacitance of 100–200 F/g [53]. Nevertheless, the electrical conductivity of the carbon onions are much higher than that of the commercially available activated carbon. Therefore, they have been used as an excellent conductive additive to decrease the resistance caused by the activated carbons, to fill the gap between the larger carbon particles and to enhance their interactions [54–56]. It has been reported that by using carbon onion additive, the electrical resistance of activated carbon electrode decreases by 80% [54–56]. Carbon onion has also been used in pseudo-capacitors as a substrate. It is well recognized that redox-active material-based pseudo-capacitors generally offer high specific energy in comparison with carbon material-based supercapacitors which possess a high specific power [57]. To significantly improve the energy storage, hybrid materials comprising carbon onions and redox

Carbonaceous Materials

43

Table 1 Different carbon structures used in electric double layer capacitor (EDLCs) with onion-like carbon (OLC), carbon nanotubes, graphene, activated carbons, and carbide-derived carbons Material

Carbon onions

Carbon nanotubes

Graphene

Activated carbon

Carbide derived carbon

Templated carbon

Dimensionality Conductivity Volumetric capacitance Cost

0D High Low High

1D High Low High

2D High Moderate Moderate

3D Low High Low

3D Moderate High Moderate

3D Low Low High

Structure

Source: Simon P, Gogotsi Y. Capacitive energy storage in nanostructured carbon–electrolyte system. Acc Chem Res 2013;46:1094–103.

5 nm

RuO2/ONCNO

Fig. 2 The transmission electron micrograph (TEM) of the carbon onion-RuO2 material. Reproduced from Borgohain R, Li J, Selegue JP, Cheng Y-T. Electrochemical study of functionalized carbon nano-onions for high-performance supercapacitor electrodes. J Phys Chem C 2012;116:15068–75.

materials have been prepared and used in energy storage. For example, carbon onions have been combined with ruthenium oxide as shown in Fig. 2 [58], nickel hydroxide [59], manganese oxide [60–63], polypyrrole [64], and quinines [65,66]. The carbon onion-ruthenium oxide (around 68 mass%) composite exihibited a high capacitance of 334 F/g in 1M H2SO4 at a scan rate of 20 mV/s. At the scan rate of 8 V/s, the material showed excellent capacitance retention [58]. For nickel hydroxide (oxide)/carbon onion hybrids, a maximum capacitance of 306 F/g was reported for nickel hydroxide and 73 F/g for nickel oxide for the corresponding electrodes made from their hybrid of carbon onions [59]. Nickel oxide-based hybrid materials show either poor power handling ability or low capacitance. Among all carbon onion hybrid materials, manganese oxide (MnO2) has been recognized as a potentially promising candidate for pseudo-capacitator. Manganese oxide materials have high theoretical capacity, and they are also low toxic, economical and environmentally benign when compare to other metal oxide materials [67,68]. The pseudo-capacitative reaction occurs only on the surface of the active material and thus the utilization efficiency of manganese oxide-based electrode material is generally low, only thin surface layer of the oxides participates in the reaction [69]. Makgopa et al. reported a hybrid electrodes made from the material of carbon onion producing from nanodiamond precursor and KMnO4-derived manganese oxide with a loading of around 47 mass%. They yielded a capacitance more than 400 F/g in 1M Na2SO4 at 0.1 A/g [61]. It was observed that the capacitance droped to 250 F/g at higher rate of 5 A/g. In summary, this type of hybrid materials may show moderate to high capacitance but suffering from a low rate handling ability. Fullerenes, such as C60 and C70, are also potentially promising materials for supercapacitator devices because of their high surface area and revisable redox charge storage [70,71]. Based on the cyclic voltammetry study, both C60 and C70 exhibited almost the same first reduction potential in 0.1 M KCF3SO3-NH3 solution at 0.96 and 0.94 V, respectively. It has been concluded that

Carbonaceous Materials

44

these fullerenes are closed conducting structures [70]. Hybrid material of fullerene derivatives such as fullerenol (C60(OH)40) was found to be electrochemically active and thus showed high potential as an electrode material [71]. A specific capitance of 80 F/g was reached for an electric double-layer capacitor at 2 mV/s in 0.5 M H2SO4 solution. Nevertheless, the cost of the zero-dimensional carbonaceous material is generally high in comparison with commercially available activated carbons due to non-maturated technology, and thus further development in the specific synthetic area is highly expected.

2.2.2.2

One-Dimensional Carbonaceous Materials for Electronic Applications

Single- and multiwalled carbon nanotubes (CNTs) are well known one-dimensional carbons with a nanoscaled cylindrical structure. They have found wide applications in materials science and nanotechnology due to their extraordinary electrical, mechanical (exceptionally strong and stiff), optical and thermal conductive properties [72–74]. As shown in Fig. 3, CNTs naturally conduct self-assemblization to form ropes attracted by van der Waals forces. Interestingly, CNTs have been reported to hybrid with fullerenes to generate so-called carbon nanobuds in which the fullerene cages are covalently bonded to the outer surface of CNTs (see Fig. 4(A)) [75]. When fullerene cages are traped inside CNTs, the resulting materials are called “peapod” as shown in Fig. 4(B) [76–79]. Carbon peapods own unique physical properties to exihibit unusual magnetic characteristics when irradiated and heated [76–79]. Integration of CNT-graphenes was also reported, the resulting hybrid, shown in Fig. 4(C) [80–85], posses high surface area and three-dimensional structure and this material was used as supercapacitator to enhance the performance.

Fig. 3 The transmission electron micrograph (TEM) of carbon nanotubes (CNTs).

200 nm

5 nm (A)

(B)

(C)

Fig. 4 Structures of nanobud (A), peapod (B) and graphenated carbon nanotubes (CNTs) (C). Available from: https://en.wikipedia.org/wiki/ carbon_nanotube.

Carbonaceous Materials

45

The CNTs show electrical properties of metallic materials or moderate semiconducting substance instead of semimetallic behavior [86]. Therefore, they are employed as additives to enhance the conductivity of the electronic wires. The CNTs can be functionalized conveniently to attain the desired properties for a variety of applications [87]. Currently, the CNTs are mainly applied in materials industry to produce lightweight and exceptionally strong high-performance materials [88,89]. In energy storage, the CNTs have been used as the electrode substrate materials due to their promising conductivity and chemical activity. To date, various CNTs-based composites have been reported with varying capacitance, power and energy density. Multiwalled CNTs-based electrodes provided a power density of 8 kW/kg and specific capacitance of 102 F/g [90]. In contrast, single-walled CNTs-based counterpart showed a power density of 7 kW/kg and specific capacitance of 180 F/g with an energy density of 20 W/kg [91]. To decline the contact resistance between CNTs-based electrode material and electronic collector, and to improve the supercapaciator performance, CNTs-based multicomponents were widely investigated. For example, a relatively high specific capacitance of 670 F/g was reached by using hybrid material of multiwalled CNTs and polyaniline [92]. Sun et al. reported hybrid fibers made of molybdenum disulfide, reduced graphene oxide and multiwalled CNTs (MoS2-rGO/MWCNTs) and their application as electrodes for asymmetric supercapacitator [93]. It was claimed that the resulting capacitator could be operated in a wide potential window of 1.4 V with high Coulombic performance, enriched energy density and high robustness. It could be cycled (charge–discharge) at least 7000 times with sustained efficiency [93]. Single-walled CNTs-based analog, known as patronitesingle-walled CNTs-reduced graphene oxide (VS4-SWCNTs/rGO) hybrid, was also prepared [94]. The material was used as electrode substrate for supercapacitators and exihibited an extremely high energy density of around 174 Wh/kg with a power density of ca. 13.8 kW/kg. A high specific capacitance of ca. 559 at 1 A/g was claimed with 1000 continuous charge–discharge cyles [94]. The CNTs are also used for flexible electronic devices such as so-called Yarn supercapacitator. The type device can be weared and folded. Transition metal oxide-based pseudo-capacitive materials such as Co3O4 and NiO have been deposited on the CNTs surface by a electrodeposition process [95]. The resulting uniform hybrid materials exhibit promising energy density of 1.10 mWh/cm2 and high capacitance of 52.6 Mf/cm2. In addition, the supercapacitators are mechanically and electrochemically stable, 96% capatance maintains after 1000 charge–discharge cycles [95]. Foldable supercapacitators were made from single-walled CNTs-based multiple composite, macroporous cellulose fibers-single-walled CNTs-polyanine nanoribbons [96]. The foldable electrodes displayed a volumetric of 40.5 F/cm3 and area capacitance of 0.33 F/cm2. It was reported that the electrodes were highly flexiable, they could be folded back and forth more than 1000 times with maintained mechanical property and capacitance. Yang et al. reported a hybrid material of carbon-chitosan-coated CNTs for micro-supercapacitators [97]. The material was prepared by pyrolysis technology and displayed a high energy density of 4.5 mWh/cm3 at a scan rate of 10 mV/s and capacitance of 6.09 mF/cm2. The micro-supercapacitor showed an excellent cyclability with 99.9% capacitance retention after 10,000 cycles [97]. It was also claimed that the optimized photolithographic technique could be a useful supplementaries to the existing arts including water cleaning, spin coating, and pyrolysis. Rangom et al. used single-walled CNTs as conductive binder for supercapacitator electrodes [98]. The resulting device has promising energy storage performance with a maximum capacitance of 100 F/g and area capacitance of 11,500 mF/cm2. A cycle life of 1 million was achieved with less than 2% loss in 0.5 M H2SO4 media as shown in Fig. 5 [98]. The work further demonstrated that electronic transfer between CNTs and metal interface is crucial for the device performance. The CNTs can be produced now in large scale by chemical vapor deposition (CVD), high-pressure carbon monoxide disproportionation, laser ablation or arc discharge. Most of the methods need to be operated in vacuum system. Among the existing technologies, the catalysis and continuous growth processes were claimed to be high priority from an industrial point of view as they are much more commercially feasible [99]. The iron-based catalysts, such as ferrocene derivatives, are commonly used in current technologies. Nevertheless, the advanced development in this area is warranted to significantly reduce the materials cost to enable broader applications of CNTs.

2.2.2.3

Two-Dimensional Carbonaceous Materials for Electronic Applications

Graphene and its derivatives represent the classical two-dimensional carbonaceous materials. The Nobel Prize in Physics in 2010 was awarded to the scientists who made significant contributions in the graphene research [100]. The CNTs can be considered as a 101

1 µm

0.5 0.25 0 –0.25 –0.5 0

(A)

Polished Roughened Gold coated

(B)

Relative capacitance (%)

Thickness: 1.6 µm

Capacitance loss (%)

1 0.75

100.5 100 99.5 99

20

40

60

80

Specific current (A g−1)

100

0

(C)

2

4

6

8

10

Cycle number (×105)

Fig. 5 The SEM image (A) and results of galvanostatic discharge tests (B,C) of single-walled CNTs/NT films. Reproduced from Rangom Y, Tang X, Nazar LF. Carbon nano-tube-based supercapacitators with excellent ac line filtering and rate capability via improved interfacial impedance. ACS Nano 2015;9:7248–55.

46

Carbonaceous Materials

Fig. 6 Structure of grapheme. Graphene properties. Available from: www.graphene-battery.net; 2016 [accessed 27.05.16].

rolling up structure of graphene sheet. As shown in Fig. 6, two-dimensional graphene belongs to the family of flat aromatic hydrocarbons. Graphenes have many attractive physical properties including exceptionally high surface area, strong and high conductivity of both electricity and heat [101–103]. Furthermore, graphene exhibits a remarkable charge transport property and electron mobility [104–108]. Owing to its unique two-dimensional structure, the chemical reagents could easily access and then interact with carbon atoms of graphene from two sides, particularly for the single-layer sheet. Oxidation of a monolayer of graphene flake forms a “graphene oxide paper”, with a tensile modulus of 32 GPa [109]. These oxide flakes have potential applications in many areas including in membrane technology, and as photoconductors and catalyst supports [109–112]. The synthetic technology of graphene has been well developed following its continual growing market, particularly in the electronics, energy storage and semiconductor industry [113]. To date, various methods, such as chemical vapor deposition [114–117] and exfoliation of graphite and its derivatives, have been reported [118–121]. It has been well recognized that the graphene is a useful material for electronic devices used in our daily life. Theoretically, a maximum specific capacitance of around 550 F/g can be reached for a single-layer graphene [122]. Additionally, hybrid materials comprising of graphenes and metal oxides display excellent electrochemical performance due to their synergistic effects. Considerable efforts have been dedicated to investigate graphene-based supercapacitors due to their inherent advantages of fast charge and potentially long life when compared to unsafe batteries used currently. The supercapacitors made from graphene material are potentially alternative devices to replace the battery that we currently use. Following the advanced synthetic technology development, it is viable to produce graphene on large scale with low cost. Therefore, the production of graphene-based supercapacitor is a growing interest in the public sector and attracted all kind of researchers in both academia and industry. Numerous scientific publications have been established in this area of research [123–125]. As discussed above, the graphene films are attractive for next generation supercapacitors, particularly for flexible thin film supercapacitors. However, the graphene films tend to agglomerate and stack to form multi-layered graphene sheets due to strong p–p interaction between aromatic rings and van der Waals forces. The resulting graphene sheets exhibit poor electronic property and thus hinder their wide applications [126–128]. It remained a big challenge to maintain the graphene thin films in the preparation process of the corresponding capacitors. A big breakthrough is highly desired to accelerate the applications of capacitors that could replace the battery that is currently being used. Various methods have been explored to achieve the target. Among the reported studies, an addition of a suitable spacer between the graphene films was found to be an effective approach. The commonly used spacers include metal oxides/hydroxides, metals, polymers and/or carbonaceous materials-beased nanoparticles [129–132]. The hybrid materials of metal oxides/hydroxides, which are also widely used as pseudo-capacitive materials with large capacitances, and graphenes demonstrated an improving performance. For example, a material of nanoscaled Co(OH)2 and graphene have been prepared and used as electrode material that provided a high specific capacitance of 1335 F/g at a current density of 2.8 A/g [133]. The Ni(OH)2-graphene composite was also formed by intercalating Ni(OH)2 nanoplate into graphene layers as shown in Fig. 7 [134]. The electrodes made from Ni(OH)2-graphene materials displayed an impressive volumetric capacitance of 655 F/cm3 with a capacitance of 537 F/g. As shown in Fig. 7, the composite exhibited the flexible property and, thus it is suitable to prepare a foldable capacitor. The electrodes made from MnO2-graphene materials demonstrated a specific capacitance of 389 F/g in 1 M Na2SO4. The capacitor can be charged/discharged 1000 cycles with 95% retension [135]. Also, a bimetallic composite of Fe2O3–MnO2–graphene has been reported. The resulting capacitor showed an energy density of 41.7 Wh/kg, a power density of 13.5 Kw/kg with a working potential window up to 1.8 V [136]. A high cyclability of more than 5000 has been reported at a current density of 16.9 A/g [136]. Such materials can be prepared by a templating assembly technology. As shown in Fig. 8, graphene was first chemically modified by polystyrene colloidal particles to form an organic template [135]. The template was then removed by washing with

Carbonaceous Materials

2 µm

47

1 µm

(A)

(B)

Fig. 7 The SEM images (A, B) of Ni(OH)2–graphene composites. Reproduced from Li M, Tang Z, Leng M, Xue J. Flexible solid-state suercapacitor based on graphene-based hybrid films. Adv Funct Mater 2014;24:7495–502.

Removal of PS

PS-embedded CMG film

Deposition of MnO2

3D macroporous e-CMG film

MnO2/e-CMG composite film

Fig. 8 Synthesis of hybrid material of MnO2/grapheme. Choi BG, Yang M, Hong WH, Choi JW, Huh YS. 3D macroporous graphene frameworks for supercapacitors with high energy and power densities. ACS Nano 2012;6:4020–8.

toluene to produce a macroporous graphene film, on which a thin layer of MnO2 was grown. The resulting hybrid material exhibited a superior electrical conductivity with a large surface area [135]. The reported capacitance reaches 389 F/g with an energy density of 44 Wh/kg, a power density of 25 kW/kg and a working supercapcitor potential window of 2.0 V [135]. The electrodes based on the stainless steel fabricated graphene material were reported by Yu et al [137]. The supercapacitor device made from this material displayed a superior specific capacitance of 180.4 mF/cm2 with 96.8% performance retention after 7500 charge–discharge cycles. It was also reported that the device could be bended and stretched for 800 times with 96.4% capacitance retention [137]. The high performance was due to the unique properties of the employing materials with promising electrical conductivity, stability and high mechanical flexibility. This work demonstrated that stainless steel fabrics could be potentially useful electrical collector for flexible and wearable power devices. Functional polymer-based spacers have also been widely investigated. Graphenes, coated with conductive polyaniline (PANI) by an in situ polymerization process, are used as electrode materials [138–141]. A high specific capacitance of 1046 F/g was achieved with 96% performance retention in a current density range of 10–100 A/g [139]. A nanocomposite of graphenes and PANI nanowire arrays was prepared from the corresponding graphene-polystyrene precursor [140]. The material was claimed to have a dramatically increased active surface area, and thus it provided a high specific capacitance of 740 F/g at a current density of 0.5 A/g. The charge–discharge cycle could be operated for 1000 times with 87% capacitance retention at the current density of 10 A/g [140]. Besides PANI, other polymers, such as polyethylene naphthalene (PEN), polymethyl methacrylate (PMMA) and polyhydroquinone, were also explored to construct graphene-based electrode materials. Wu et al. prepared a flexible micro-supercapacitor made from graphene film and PEN sheet by laser carving method [142]. It was claimed that the method enhanced the accessible surface area for the metal current collector-free and flexible microsupercapacitor. The device exhibited a good specific capacitance of 15.38 mF/cm2 at a current density of 0.1 mA/cm2 [142]. Graphene sheet could be exfoliated by PMMA nanoparticles to form a unique macroporous graphene foam [143]. The polymer nanoparticles were then removed by calcinations at 8001C and the resulting microporous graphene was used to make electrodes. It was claimed that the device had a reasonable capacitance retention of 67.9% with a scanning rate of 1000 mV/s [143]. However, relatively high temperature was used in the procedure to remove the polymer spacer and the resulting graphene was found to aggregate during the course of its annealing. The technology was developed by replacing PMMA with polystyrene, which could be washed away by toluene at low temperature and thus maintaining the graphene microstructure [135].

48

Carbonaceous Materials

6 µm

100 nm

Fig. 9 The SEM images of graphene hydrogels. Reproduced from Cong HP, Ren XC, Wang P, Yu SH. Macroscopic multifunctional graphenebased hydrogels and aerogels by a metal ion induced selfassembly process. ACS Nano 2012;6:2693–703.

The poly(3,4-ethlenedioxythiophene) was found to be highly conductive electrically and robust with a broad potential window. It was also hybridized with graphene by an in situ electropolymerization process on graphene oxide (GO), followed by an electrochemical reduction of the GO [144]. A capacitance of 14 F/cm2 was obtained from the device made of its nanomaterials. When graphene is produced by hydrothermal or chemical reduction methods, graphene aerogels and hydrogels may be formed [145,146]. These materials are ultralight and possess high surface area as shown in Fig. 9. However, they exhibit relatively poor mechanical strength and low flexibility [147]. Therefore, more studies have been focused to overcome the limitations in improving their performance. The electrodes made from hybrid materials of graphene hydrogel and alizarin molecules showed a high specific capacitance of 350 F/g at 1 A/g with a wide voltage window of 1.4 V in 1 M H2SO4 [148]. It was reported that graphene hydrogel coated with electroactive polyhydroquinone produced a highly conductive electrode nanomaterial. A high specific apacitance of 490 F/g was obtained at the current density of 24 A/g [149]. The nanohybrid was prepared conveniently by the reaction between graphene oxide and the hydroquinone reductant. The porous graphene hydrogel was pressed and stacked to form a high density film [150]. A flexible supercapacitor, made from the graphene hydrogel-based electrode material, displayed a volumetric and gravimetric capacitance of 212 F/cm3 and 298 F/g, respectively. This device could charge and discharge 10,000 cycles with a capacitance retention of 87% [150]. Other technologies were also explored to prepare graphene-based supercapacitor. Wang et al. reduced graphene oxides with hydrazine and hydriodic acid to prepare graphene films using the blade-coating technique [151]. It was claimed that the hydrazine-derived graphene films displayed a capacitance of 265 F/g. The device could be charged-discharged thousands of cycles [151]. Graphene-based supercapacitors were fabricated in liquid solution of lithium bis(oxalato)borate in propylene carbonate or gel electrolytes consisting of polyvinyl alcohol and lithium bis(oxalato)borate or solid electrolyte of polyvinyllidene fluoride [152]. A capacitance of 78 mF/cm2 was obtained for gel fabrication method. For the solid electrolyte, a capacitance at 15.5 mF/cm2 was reported with threefold improving of light transmission. The flexible solid showing reasonable chemical and thermal stability was claimed as excellent electrolyte for graphene-based supercapacitors [152]. Xie et al. prepared the N-enriched carbon/graphene composites by a carbonization-activation process [153]. The carbonization was performed by coating polyacrylonitrile nanofiber paper with bulk-doped and/or graphene oxide. The resulting material was activated using KOH at a high temperature of 700 1C to form a sandwich-like composite with a specific surface area of 2631.8 m2/g. Electrodes made of this material exhibited superior energy storage capability [153]. A high specific capacitance of 381.6 F/g was reported at a current density of 0.1 A/g in 6 M KOH aqueous solution. The energy density was 13.2 Wh/kg in a power density range of 4.7–25.0 W/kg with the retention rate of 62.9% [153]. In summary, as the delegate of two-dimensional carbonaceous materials, graphene and derivatives have been widely explored as potentially promising electrode materials for both the supercapacitor and pseudo-capacitor. Following the advanced development in graphene preparation technology, the material can be produced conveniently in relatively large scale by a chemical exfoliation-reduction process. Nevertheless, the product cost of supercapacitor is higher than other energy-storage devices at current stage, particularly for the high-quality graphene. The price of graphene material varies from tens to thousands of dollars per kilogram depending upon the quality [154], and that is significantly higher than the that of activated carbon which is 15 dollars per kilogram [155]. The cost-effective graphene-based supercapacitors with high performance would be favorable to overcome the barrier and to enhance its feasibility on an industrial scale. In terms of large-scale production of graphene, the existing methodologies, such as the exfoliation of graphite or the reduction of graphite oxide, need to be re-evaluated to produce high-quality materials on a large scale in an environmentally friendly approach. As discussed above, the prevention of re-stacking of graphene films remained a big challenge in this area of technologies. The issue has to be addressed before graphene material-based energy storage devices can be developed on a large scale. Furthermore, the performance of the whole device should be consided in addition to graphene-based electrode materials. To materialize the targeted device, more efforts need to be made in terms of improving synthetical technologies, particularly to stabilize the

Carbonaceous Materials

49

graphene films, and material fabrication to improve their performance. In this regard, hybrid nanomaterials of graphene and pseudo-capacitive materials such as graphene/metal oxides and graphene/conductive polymers are the promising graphene products. Nonetheless, the accessibility of the hybrid material to the electrolyte should be of great concern because electron diffusion through the basal plane is limited due to potential graphene wrapping. Considering that the graphene microstructures determine their performance as the electrode material, it should be of high priority to precisely control and functionalize their porosity and pore size at this stage. These materials also show high potential in producing flexible energy storage devices including supercapacitors. Graphene-based energy-storage devices are expected to make a real impact on our daily life.

2.2.2.4

Three-Dimensional Carbonaceous Materials for Electronic Applications

The 3D carbonaceous materials possess complicated porous nanostructures, which include frameworks of lower dimensional carbons such as CNTs, graphene, activated carbon, carbide-derived carbon, and carbon templates. This section discusses the applications of the last three types of 3D carbon materials in energy storage according to the recent literature data. The porous carbon can be classified as hierarchical porous carbon and templated mesoporous carbon. In principle, both the micropores and mesopores (2–8 nm) are crucial for carbonaceous material-based high energy storage. However, micropores are making big contributions for high energy density, whereas mesopores play an important role in speeding up the ion diffusion in the electrodes and that should improve the power density of the capacitors. The 3D hierarchical porous graphitic carbons with different dimensions of micropores (0.7–1 nm), mesoporous walls (100–200 nm) and macroporous cores (1–2 mm) (see Fig. 10) have been reported [156]. The material showed a high power density of around 12,000 W/kg and an energy density of around 80 Wh/kg [156]. More efforts are now dedicated to prepare well-ordered carbonaceous nanomaterial with the 3D interconnected pores [157]. Recently, Xu et al. synthesized the hierarchical hybrids with microporous carbon spheres that are functionalized by graphene frameworks [158]. The material possesses both the porous structure with large surface area and good electrical conductivity. As expected, it demonstrated a superior specific capacitance of 288.77 F/g, and a high electrosorption capacity of 19.8 mg/g [158]. The 3D porous carbon materials can be prepared conveniently by templated procedures. These materials are also called as ordered mesoporous carbons. Li et al. synthesized a nanocomposite with 40 wt% of PANI by in situ polymerization [159]. The electrode made of this material showed a large specific capacitance of 470 F/g and high rate capability of about 87% capacity retention. The electrode could be recycled more than 1000 times with 90.4% stability maintaining [159]. The authors indicated that the MnO2 nanoparticles were also incorporated with the nanocomposite described above [160]. The new conjugate exhibited an improved specific capacitance of 695 F/g with a compatible cycling stability of 88% capacity retention after 1000 cycles. Watermelon rind, as a nitrogen-enriched precursor, was employed to prepare actived carbon with a large surface of around 2277 m2/g and a nitrogen content of 2.48 wt% [161]. The resulting electrode exhibited a high gravimetric specific capacitance of 333.42 F/g in 6 M aqueous KOH at a current density of 1 A/g. It was also reported that this material showed the high rate performance, low impedance and good cycle stability with 96.82% capacitance remaining after 10,000 cycles [161]. Prawn shell was also used as the raw precursor to prepare N-doped activated carbons [162]. This material was obtained by a demineralizationannealing (600–8001C) process. It was reported that the material exhibited a high specific capacitance of 695 F/g in 1M sulfuric acid media at the current density of 50 mA/g. The electrode made of this carbonaceous material delivered a specific energy density of 10 Wh/kg and a specific power density of 1000 W/kg with a sustainable life over 5000 cycles [162]. Marcuzzo et al. prepared microporous activated carbon fiber felt from a 5.0 dtex Brazilian textile PAN fiber by carbonization at a high temperature of 9001C [163]. It was claimed that the resulting material possesses microporous structure with the pores of less than 3.2 nm and a high surface area of 1300 m2/g. This carbonaceous material showed a high specific capacitance of 200 F/g at 1 mA/cm2 and a volumetric capacitance of 104 F/cm3 [163]. Wall Core

Pseudo-active materials

Fig. 10 The 3D hierarchical porous structure. Reproduced from Jiang H, Lee PS, Li C. 3D carbon based nanostructures for advanced supercapacitors. Energy Environ Sci 2013;6:41–53.

50

Carbonaceous Materials

The carbonaceous materials can be intergrated with silicon-based electrochemical capacitors. The composites are of carbidebased carbon films strongly adhering with interdigitated micro-supercapacitors and embedded titanium carbide current collectors [164]. The material has a surface area of about 977 m2/g with a mean pore size of 0.59 nm. It is completely compatible with current microfabrication and silicon-based device technology. The electrode capacitance was stable over 10,000 cycles with a high power capability of 200 F/cm3 and a volumetric capacitance of 350 F/cm3 [164]. Activated carbons, modified with carbon nanodots, were manufactured and used as electrode material for supercapacitors [165]. The carbon dots were introduced into the activated carbon by sonochemical technology. The electrodes, made of these activated carbon–carbon dots, demonstrated a specific capacitance of 0.185 F/cm2 and that is three times higher than the capacitance of pristine activated carbon electrodes. This material also exhibited extremely stable electrochemical behavior with many thousands of cycles with a Coulombic efficiency of around 100% [165]. Comte et al. functionalized porous carbon by directly grafting with 9,10-phenanthrenequinone [166]. The electrochemical capacitance of the tailored carbon–Ni(OH)2 hybrid negative electrode was drastically improved. The capacitance was stable even after 10,000 galvanostatic charge/discharge cycles. Spinel cobalt oxide was used to functionalize activated carbon to enhance its performance in energy storage [167]. A best energy density of 290 W/kg was obtained from a positive/negative weight ratio of 1.25. It was reported that the system exhibited a good electrochemical stability over 3000 cycles with the discaharge capacities of around 62 F/g [167]. Wang et al. prepared Prussian-blue-doped super activated carbons by precipitation method [168]. The Prussian-blue nanoparticles were immobilized on the surfaces of the activated carbon and that was confirmed by transmission electron microscopy (TEM) analysis. The electrodes made of this material demonstrated a specific capacitance up to 263.7 F/g at a current density of 5 A/g. The cyclic stability results indicated that the electrode could retain 94.8% of its initial specific capacitance value over 1500 charge/discharge cycles [168].

2.2.2.5

Application of Borocarbonaceous Materials for Supercapacitors

As discussed above, supercapacitors are of great interest as promising electrochemical energy storage and power output technologies for portable electronics, electric vehicles, and renewable energy systems operated on intermittent sources such as solar and wind mills. Carbonaceous materials are the most widely used electrodes because of their excellent physicochemical properties, including their good electrical conductivity, tunable porosity and ease of processability, and also because of its low cost. Energy storage in carbon-based supercapacitors depends on charge uptake in the carbon/electrolyte interfacial region, which relates to the chemical surface and electronic structure of porous carbons. Surface chemical functional groups significantly increase the interfacial capacitance by introducing pseudocapacitance. It was suggested that the boron doping may improve the specific capacitance per surface area for the multiwalled carbon nanotubes [169]. Boron is a unique element that has been explored for decades as substitution in carbon or diamond materials to promote the properties of oxidation resistance, Li-ion insertion, and electrochemical behavior [170]. Boron enters the carbon lattice by substituting for carbon at the trigonal sites [171] and acts as electron acceptor because it has three valence electrons, causing a shift in the Fermi level to the conducting band and hence modifying the electronic structure of boron-doped carbon [172,173]. The change in electronic structure of carbon electrode materials can affect the electric double layer capacitance. Most importantly, low-level boron doping shows catalytic effect on oxygen chemisorption on carbon surface, rendering the introduction of redox reactions related to oxygen functional groups on carbon surface [174,175]. Therefore, boron doping is able to modify the electrochemical capacitance of carbon materials, involving electrical double layer capacitance and pseudocapacitance [176]. Chen et al. prepared mesoporous carbon with homogeneous boron dopant by co-impregnation and carbonization of sucrose and boric acid confined in mesopores of SBA-15 silica template [176]. Low-level boron doping shows catalytic effect on oxygen chemisorption at edge planes and alters electronic structure of space charge layer of doped mesoporous carbon. These characteristics are responsible for substantial improvement of interfacial capacitance by 1.5–1.6 times higher in boron-doped carbon than that in boron-free carbon with alkaline electrolyte (6 M KOH) and/or acid electrolyte (1 M H2SO4). Such boron-doped mesoporous carbon can be expected to show maximum capacitance if further optimization of the local boron doping environment is done. This finding should be very useful for developing new doped carbon electrode materials for supercapacitors. Guo et al. prepared the boron and nitrogen co-doped porous carbons through a facile procedure using citric acid, boric acid and nitrogen as C, B, and N precursors, respectively [177]. The resulting boron and nitrogen enriched carbon materials showed prominent capacitances. The nitrogen, boron and oxygen incorporated into the carbon matrix enhanced the wettability between the electrolyte and electrode materials, and the introduction of heteroatoms may result in the pseudo-capacitive effect. Two samples of the boron and nitrogen enriched carbons (BNC), BNC-9 and BNC-15 were prepared with high specific surface areas of 894 and 726 m2/g and showed the large specific capacitance up to 268 and 173 F/g, respectively, with the current of 0.1 A/g. When the current was set as 1 A/g, the energy densities were 3.8 and 3.0 Wh/kg and the power densities were 165 and 201 W/kg for BNC-9 and BNC-15, respectively. Thus, BNC-15 is more suitable to apply in high-power-demanded occasion, while BNC-9 tends to store more energy. Wu et al. have demonstrated a simplified prototype device of high-performance all-solid-state supercapacitors, based on three-dimensional nitrogen and boron co-doped monolithic graphene aerogels [178], and this device possesses an electrodeseparator-electrolyte integrated structure, in which the graphenes acted as additive/binder-free electrodes and a polyvinyl alcohol (PVA)/H2SO4 gel as a solid-state electrolyte and thinner separator. The nitrogen and/or boron doping in carbon networks can facilitate charge transfer between neighboring carbon atoms and thus it enhances the electrochemical performance of carbon-based materials [179,180]. The graphene composites show 3D interconnected frameworks with a macroporous architecture, which are favorable for ion diffusion and electron transport in bulk electrode. In addition, the monolithic boron and/or nitrogen-doped

Carbonaceous Materials

51

graphenes can be easily processed into thin electrode plates with a desirable size upon physical pressing. Consequently, the resulting composites exhibit not only minimized device thickness, but also show high specific capacitance (B62 F/g) and enhanced energy density (B8.65 Wh/kg) or power density (B1600 W/kg) with respect to undoped graphenes, or a layerstructured graphene paper [141].

2.2.2.6

Application of Borocarbonaceous Materials in Thermoelectric Energy Conversion

Thermoelectric energy conversion is a very reliable way of generating electrical power than from solar heat or thermal energy from industrial waste. Thermoelectric energy conversion has important technical advantages when compared to other methods of generating electrical energy from heat or to recycle wasted thermal energy. The devices do not contain any movable parts. Therefore, after installation they do not need any service in operation as they are very reliable for a longer period of time. For example, thermoelectic devices installed in satellites in the early sixties have been reliably supplying electrical energy to all electronic equipments ever since [181]. Unfortunately, the efficiency of these devices does not exceeed above 5%. In most cases, this is not sufficient for practical applications, economically and, therefore, thermoelectric energy conversion has been restricted to such cases where extremely high reliability of the devices is more important than economy. The investigations have shown that boron-rich solids are very promising candidates for high efficiency thermoelectric energy conversion. The outstanding high melting points and extraordinary chemical stability allow their use under extreme conditions unlike most other materials. Some boron-rich semiconductors exhibit very favorable transport properties of high Seebeck coefficients, increasing monotonically up to very high temperatures, electrical conductivities with values typical of semiconductors, and very low thermal conductivities [181]. The boron-rich solids are characterized by Bl2 icosahedra, which are common structural elements arranged differently in the various structures. The electronic properties of these solids are essentially determined by these icosahedra and, therefore, the electronic properties are closely related as well [182,183]. The mechanisms of electronic transport in these materials are unique: they differ from those of classical and of amorphous semiconductors. At least some of the structural varieties offer very favoable prerequisites for high effficiencies in thermoelectric energy conversion. In general, the high melting points (e.g., boron carbide, Tm 42600K) are outstanding when compared to all other semiconductors and thus allow operation temperatures to exceed far beyond those that are restricted to other semiconductors. Based on these properties, the Carnot efficiency can reach much higher values than for all other semiconductors known to date. Boron carbide is the most extensively investigated boron-rich solid with respect to properties that are relevant to thermoelectric application. Its homogeneity range extends from B4.3C at the carbon-rich and to about Bl2C at the boron-rich limits. Depending on the composition, the unit cells with 15 atoms in the idealized structure consist of B12 or B11C icosahedra, C–B–C or C–B–B chains. With decreasing carbon content, an increasing number of unit cells without any chains is found [184,185]. There is no chemical composition at which the structure is completely homogeneous. This has an essential influence on modification of the properties as a function of chemical composition. Boron and borides are expected to be candidates for high-temperature thermoelectric materials because some of them exhibit (1) rising Seebeck coefficients and electrical conductivities with increasing temperature, perhaps due to their hopping conduction mechanism, and (2) relatively low values of the high temperature thermal conductivities, due to their complex structures. Nonetheless, they are promising condidates because of their chemical stability at high temperatures. For example, AlB12 [186], B13C2 [187], B–Si–C [188] exhibit p-type conduction and, therefore, n-type boron-based materials are desired to fabricate thermoelectric devices. In order to screen candidates for n-type boride-based thermoelectric materials, the energetics of solid solutions of metallic atoms (Zr, Cr, and V) in b-rhombohedral boron (b-boron) and densities of the physical state of metal borides of the CrB-, FeB-, MoB-, AlB2-, ReB2-, CaB6-, and UB12- types of structures and some tetraborides (YB4, CrB4, WB4, and MgB4) have been calculated by Imai et al., using the first-principles pseudopotential method within the local (spin) density approximation [189]. It was appropriately concluded that Zr occupies the E site of b-boron, while V and Cr occupy the A1 site. As for the metal-monoborides and diborides, the ‘rigid band approach’ seems to be valid. The large negative Seebeck coefficient of FexCo1-xB is hopeful if Mott’s explanation for the trend of the Seebeck coefficient for transition metal elements is valid for these borides as well. The Y- or La- doped Ca (Sr or Ba)B6 can also be expected to be useful for preparing n-type thermoelectric conversion materials. Takeda et al. have synthesized and examined the thermoelectric properties of polycrystalline AlMgB14 and some hexaborides (CaB6, SrB6, YbB6, SmB6, and CeB6) [190]. Single phase of orthorhombic AlMgB14, containing B12 icosahedral clusters as building blocks, was obtained at sintering temperatures between 1573 and 1823K. The Seebeck coefficient (a) and electrical conductivity (s) of the phase were about 500 mV/K and 10 1 1/Om at room temperature, respectively. These values are comparable to those of metal-doped b-rhombohedral boron. On the other hand, metal hexaborides with divalent cation possessed a large negative value, ranging from 100 to 270 mV/K at 1073K. The calculated power factors of CaB6 and SrB6 exceeded 10 3 W/K2m within the entire range of measured temperatures. As a result, they can be considered as promising material for n-type thermoelectric device. As discussed above, the boron-rich semiconductors are very favorable for thermoelectric energy conversion. Efficiencies of 25% or even more seem possible after further systematic scientific and technological development of these materials. Such efficiencies are comparable with smaller electrical power stations, but thermoelectric devices far exceed their reliability. Therefore, using boron-rich solids numerous useful ecological and economical applications of thermoelectric energy conversion have come within our reach.

52 2.2.2.7

Carbonaceous Materials Application of Borocarbonaceous Materials in Photovoltaic Devices and Fuel Combustion

In addition to the application discussed above, the boronated carbonaceous (or simply borocarbonaceous) materials are also used in other fields, such as highly efficient energy transfer in the light harvesting system, photovoltaic films in solar cells, high energy density fuel, and so on. One of the promising technologies for future alternative energy sources is the direct conversion of sunlight into electricity by using photovoltaic (PV) cells [191]. The greatest challenge in the photovoltaic field is to develop new types of advanced materials with the desired electrical and optical properties that will allow the fabrication of robust, high efficiency and inexpensive PV cells. The field of photovoltaic devices has been dominated by solid-state junction devices, typically made of silicon and profiting from the experience and material availability resulting from the semiconductor industry. Currently, crystalline silicon-based PVs offer the best possibilities in terms of large-scale manufacturing of efficient, robust, and relatively low cost devices [192]. Boron-doped p-type carbon films have been used to fabricate photovoltaic cells by chemical vapor deposition (CVD) and ion implantation techniques [193,194]. Ma et al. prepared the boron-doped diamond-like carbon thin film on n-silicon substrate using arc-discharge plasma chemical vapor deposition (PCVD) technique. The resulting film showed a lowresistivity and a high optoelectronic efficiency when the deposition condition and boron concentration were well controlled. It was found that the electronic transportation is strongly dependent on the microstructure of the film [195]. Elemental boron has one of the highest volumetric heats of combustion known to date and, therefore, it is a material of great interest to explore as a high energy density fuel. It is well recognized that the efficient engine performance requires rapid fuel ignition and combustion, but they can be difficult to achieve for a complex hydrocarbon fuel. One approach for speeding up the kinetics is to add a soluble catalyst to the fuel [195–197]. As high energy density materials, boron and boron-rich solids are the only practical materials with both volumetric (135.8 MJ/L) and gravimetric (58.5 MJ/kg) energy densities that are substantially greater than those of hydrocarbons. Boron’s potential as a fuel or fuel additive has not, to date, been realized, partly due to difficulty in igniting and burning it efficiently. The fact that the boron combustion is inherently a heterogeneous process, its rapid and efficient combustion is difficult to achieve. An obvious strategy is to increase the surface area/volume ratio by decreasing the particle size. This approach is limited by the fact that the boron forms B0.5 nm thick native oxide layer, which not only inhibits combustion, but also consumes an increasing fraction of the particle mass as the size is decreased. The heterogeneous chemistry can, in principle, be accelerated by using nanoparticulate boron, leading to large surface-area-to-volume ratios, as observed by a number of researchers [198–201]. The native oxide layer that forms on air-exposed boron surfaces limits this strategy, both by inhibiting combustion and by reducing the energy density for small particles, due to increasing fraction of the particle mass made up by the oxide layer [202]. Various methods have been conducted to mitigate the effects of the oxide layer and enhance the ignition of boron nanoparticles, including the following [203–209]: (1) treating boron particles with TiCl4 and triethylaluminum followed by the addition of ethylene or even coating boron particles with LiF and trimethylolpropane, (2) sodium naphthalenide reduction of BBr3 in 1,2-dimethoxyethene followed by n-octanol, (3) coating of boron particles with metals, such as titanium and Mg, (4) utilization of energetic materials, such as glycidyl azide polymer (GAP) and azide polymer (AP), coating on the boron surface, and (5) capping boron particles with an organic oleic acid layer [210]. Another strategy might be to coat the boron particles with a material (e.g., catalyst) to enhance combustion of either the boron itself or of a hydrocarbon fuel carrier. Such a catalyst would tend to accelerate the combustion of hydrocarbon carrier fuel, and thereby improving the combustion efficiency of the solid particles. Using boron nanoparticles as carrier for a combustion catalyst is, therefore, a possible strategy for simultaneous enhancement of ignition of the hydrocarbon fuel carrier (or the binder in a solid propellant or explosive) and the total energy density. Devener et al. presented a simple, scalable, one-step process for generating air-stable boron nanoparticles that are unoxidized, soluble in hydrocarbons, and coated with a combustion catalyst. Ball milling is used to produce B50 nm particles that were protected against room temperature oxidation by oleic acid functionalization, and optionally coated with catalyst [211]. Recently, Fareghi-Alamdari et al. synthesized a new dicationic ionic liquid (DCIL), based on dicyanamide anoins, and used as a protective ligand for boron nanoparticles [210]. Boron nanoparticles and DCIL-capped boron were produced using a ball mill technique with size distributions of 50 100 nm. The results of X-ray diffraction (XRD), energydispersive X-ray (EDX), Fourier-transform infrared (FT-IR), X-ray photoelectron spectroscopy (XPS), and ζ-potential measurements showed that the surfaces of boron nanoparticle were successfully capped by the DCIL layer, which protects the boron surfaces against air oxidation. Thermogravimetric analysis (TGA) measurements showed that the nanoparticle oxidation occurred after DCIL thermal decomposition and allowed boron ignition at higher temperatures. These results confirmed that ionic liquid binds to boron well enough to protect the boron surfaces from oxidation during air exposure. Despite the desirable properties of boron, the incorporation of boron nanoparticles into the propellants is not widely practiced, because of its prominent oxidation product B2O3. The combustion of boron particles occurs in two successive steps: the first step involves the removal of its oxide layer, while the second step involves the burning of bare boron. The pre-existing oxide layer on the boron surface plays an important role in the ignition and combustion processes; more precisely, it delays the ignition process.

2.2.2.8

Application of Borocarbonaceous Materials in Nuclear Reactor

At present, electricity is mainly generated from fossil fuels and around 33% of the carbon entering the atmosphere annually [212]. Fossil fuels are non-renewable and non-green energy resources. Renewable energy resources of wind and solar power can only provide a small fraction of energy for our daily life. Therefore, the replacement of non-renewable carbon-based fuels by nuclear power is becoming more attractive. Nuclear reactor has been incorporated in many nuke plants and in aircraft carriers and

Carbonaceous Materials

53

Fig. 11 Boron-based control rod assembly for pressurized water reactor (PWR) nuclear reactor. Available from: https://en.wikipedia.org/wiki/ Control_rod; 2016 [accessed 20.11.16].

submarines. In the operation of a nuclear reactor, control of the neutrons, which are produced by fission reactions, is one of the basic requirements. The 10B isotope of boron possesses a very large neutron absorption cross section, about 3840 barns for thermal neutrons [213]. The reaction of thermal neutrons with boron is: 10B þ n - 4He2 þ þ 7Li3 þ þ a (2.31 MeV) þ g (0.48 MeV) [213]. The resulting products (helium and lithium) are stable isotopes. Boron has been widely used as the main composition in control rods (see Fig. 11) in nuclear reactors to absorb neutrons and thus control the rate of fission by quenching the chain reaction that generate them. Boron-based control-rod materials are particularly used for the predominantly pressurized water reactors (PWRs) and boiling water reactors (BWRs). Furthermore, boron is also used as a neutron shield due to its wide adsorption spectrum. However, the elementary forms of boron are unsuitable for nuclear reactor due to their weak mechanical properties, instead alloys or compounds of boron are used. Boron carbide and refractory metal borides are commonly used as control-rod materials in nuclear reactors. They have attractive properties of high melting point, hardness, low density, chemical inertness and excellent thermal and electrical characteristics [214–216]. Boron carbide can be prepared by a carbothermic reduction route at a temperature of greater than 15001C from commercially available boric acid as per the reaction [216]: 4H3BO3 þ 7C - B4C þ 6H2O þ 6CO. However, this route gives a relatively poor yield of boron carbide with 60%–65% in terms of boron content. A new method has been developed to prepare boron carbide by solid-phase reaction between elementary boron and carbon as indicated by the equation [217]: 4B þ C - B4C. In summary, boron-based neutron absorber materials are crucial in nuclear reactors. Processes for the synthesis of high-quality boron carbides and related materials are in the developing stage. Application of advanced nanotechnologies, including three-dimensional printing should improve the process significantly with the desired properties.

2.2.3

Application of Carboraneous Materials in Oil Adsorption

Carbon materials have also found important application in oil field. It is well recognized that development and discovery of renewable energy, such as wind and solar energy, are becoming more important. Nevertheless, there are many challenges that need to be addressed before these sustainable energy can be used widely to replace the oil resource completely. Today, naturally occurring fossil fuels remained as one of the main energy and chemical resources of our daily life. However, oil spilling causes loss of millions of oil barrels routinely. For example, in an incident of tanker collision in Bangladesh in December 2014, the spilled oil covered the shoreline, killed large number of fishes, dolphins, sea birds and other small animals due to poisonous chemicals in the oil. The leaked oil must be collected and cleaned up immediately to reduce the environmental contamination to protect our ecosystem. In addition, following the worldwide ever-increasing energy demand, the depletion of fossil fuels is getting worsening. Therefore, efficient recovery and cleanup of the spilled oil, particularly for the oil spill in the sea, will address both environmental issues and industrial oil-waste. Currently, the oil spills can be cleaned up mainly by mechanical, chemical and biological methods [218–220].

54

Carbonaceous Materials

When oil spills occur, the first and primary step that people take is mechanical cleanup such as skimmers, booms and artificial sorbents. In contrast, chemical cleanup includes treating with solidifiers and dispersants, which may break down the slick oil into droplets, and thus benefits a microorganisms-based biological conversion of oil toward relatively small molecules such as carbonic gas and water. Among above methods, adsorption is recognized as one of the most straightforward, effective and practical approaches. The absorbent materials are variable with low cost, high selectivity and easy modification. The optimized absorbent materials are expected to be superhydrophobic, superoleophilic, highly selective, chemically inert, and recyclable. However, the current commonly used materials, such as wool fiber and zeolite, show some inherent drawbacks of lower absorption capacity and selectivity, as well as limited recyclability [221,222]. Therefore, it is highly demanding to develop new absorbent materials with optimizing properties. In the course of exploring new materials, various candidates have been examined. Among them, carbon materials have demonstrated attractive properties, such as super hydrophobic and superoleophilic with high surface area and, therefore, this material has been recognized as the most applicable candidates due to their extremely high oil uptake capacity as well as sustainability. It is well recognized that carbon is a tough, chemically stable, recyclable, hydrophobic and oleophilic material. Following advanced developments in nanotechnology, various carbonaceous materials, including carbon nanotubes, graphene sponges and activated carbon, with high surface area and low density have been well investigated and documented. They have been used in many areas of oil spill cleanup, gas storage and separation and water treatment [223]. Carbon materials with large pore volume and high surface area are particularly suitable for oil absorption and water/oil separation, due to their excellent oil absorption capacity and reusability. In this section, we summarize the current developments in carbon materials, including graphite and graphene-based sorbents, carbon nanotube-based sorbents and carbon aerogels.

2.2.3.1

Graphite and Graphene-Based Sorbents

Graphite and its derivatives represent classical two-dimensional carbonaceous materials. It is one of the three naturally occurring allotropes of carbon. It is formed by the reduction of sedimentary carbon compounds during metamorphism and can be found in metamorphic rocks globally. Exfoliated graphite absorbs heavy oil quickly and thus it was recognized as one of the promising oil absorbents. Exfoliated graphite can be prepared by various techniques. On a laboratory scale, it can be made by heating graphite with strong sulfuric acid or nitric acid [224]. The resulting carbon material possess an absorption capacity of about 70–80 different oils (absorption capacity ¼ (weight of the material after absorption minus ( ) weight of the material before absorption)/weight of the material before absorption) [224]. Exfoliated graphite can be produced on a large scale conveniently by thermal decomposition of natural graphite at a temperature range of 1000–12001C. During the course of heating, natural graphite generates different types of gases, including CO, and thus improving the exfoliation process to form porous carbon materials [225]. However, the process generally produces a mixture of exfoliated materials with different pore volumes and absorption capacities. This capacity is closely related to the temperature used in the absorption processes. Accordingly, the higher temperature gives increased absorption capacity due to increased oil viscosity. A maximum absortion capacity of 90 was achieved with a high recovery ratio of 80% from an industrially available exfoliated graphite [225]. Wang et al. reported a method to functionalize graphite [226], in which the hydrogen peroxide was used to oxidize the graphite and concentrated sulfuric acid was used as an intercalate. The resulting material showed a worm-like morphology with three different types of pores as detailed in Fig. 12(A)–(C). As claimed, the V-type pores with a dimension of hundreds of mm are the level-I pores, the willow leaf-type pores with dimension from several mm to several dozens of mm, shown in Fig. 12(B), are level-II pores. The pores-III (0.1 mm to several mm) are showing above and below the walls of the level-II pores as shown in Fig. 12(C). This type of materials is particularly flexible and it could absorb different types of industrial oils with an absorption capacity ranging from 43 to 85 for an expanded volume of 100–320 mL/g as shown in Fig. 12(D) [226]. Graphite is a commonly found mineral with many layers of graphene, which is known as one atomic layer of graphite. Graphene shows unique properties with significant chemical, thermal and mechanical stability [227]. It is one of the strongest materials ever found on earth with more than 40 times stronger than diamond and more than 300 times stronger than A36 structural steel [227]. Graphene shows higher electrical conductivity than graphite due to availability of free p-electrons on each carbon atom. Thus, graphene can be prepared by a number of different techniques of mechanical exfoliation of graphite, chemical vapor deposition (CVD) [228], among other mehods. Both mechanical exfoliation and CVD approaches are effective to create single and few layer graphenes. However, there are inherent drawbacks for these methods, including complexity in removing the graphene layers without damaging the graphene’s atomic structure. Therefore, more efforts are needed to develop the most efficient technology to produce the high-quality graphene on a large scale. Among the newly developed techniques, heating and reducing graphene oxide to graphene is gaining significant attention due to its reduced cost (Fig. 13). Bi et al. chemically modified graphene to prepare spongy graphene materials with needle-like microporous structures [229]. It was reported that the contaminated small amount of organic functional groups, during the process, benefited the interaction between polar solvents and graphene surfaces and thus improved its absorption capacity to 20–86 times of its weight [229]. In the process of reducing graphene oxide thermally, the C/O (carbon/oxygen) ratio could be tuned by the conditions used for exfoliation [230]. An increasing C/O ratio increases the surface area and the pore volume, and thus improves absorption capacity of the graphene material. A high absorption capacity of 108–131 has been reached with the C/O ratio of 17:1 [230]. Graphene aerogel is about 7.5 times lighter than air and is the lightest material on Earth [231]. These materials have demonstrated great potential for the oil spill cleanup application due to their extremely high compressibility, ultra-lightness and high surface area. Graphene aerogels, with a low density of around 4.4–7.9 mg/cm3 and a porosity up to 99.6%, showed a high

Carbonaceous Materials

(A)

55

(B)

90 80 70 m2 (g g−1)

60 50 40 30 20 10 0 Kerosene

Crude oil

Gear oil

Different oil

(C)

(D)

100 mL g−1 150 mL g−1 190 mL g−1

250 mL g−1 320 mL g−1

Fig. 12 The FE-SEM images (A–C) of the modified expanded graphite and its absorption capacity for different oils (D). Reproduced from Wang L, Fu X, Chang E, et al. Preparation and its adsorptive property of modified expanded graphite nanomaterials. J Chem 2014;2014:678151.

absorption capacity up to 250 times of its own weight for carbon tetrachloride [232]. This aerogel can be reused by compression and combustion-distillation with a slight drop in its absorption capacity (decrease of 4% after ten cycles). Zhao et al. prepared a threedimensional graphene aerogel (see Fig. 14) with an ultra-light density of 2.170.3 mg/cm3 at room temperature [234]. The material was also claimed to possess excellent mechanical and thermal stability as well as a tremendous absorption capacity of 200–600 times of its own weight and a high rate of 41.7 g/(g
s), which is the highest rate reported so far for any graphene materials [234]. Graphenes have also been used to functionalize porous materials such as fibers, cottons and sponges to form new oil absorbents due to their intrinsically hydrophobic property. Methods have been developed to modify the porous materials with dip coating and grafting polymerization [235,238]. These porous materials usually have good mechanical properties, low-cost and reasonable absorption capacity, and are potentially useful in oil spill cleanup. Even though the commercially available porous materials absorb both oil and water, the functionalized graphene is expected to enhance their hydrophobicity and improve their capability of absorbing oil. Tai et al. reported the coating of melamine sponge with graphene nanosheet [236]. In this case, the coating material displayed superhydrophobic and superoleophilic properties, and a high absorption capacity of 54-165 times of their own weight for different oils and organic compounds. It was reported that a lower graphene loading (o5.1 wt%) could not alter the wettability of the melamine sponge; meanwhile, overloading (47.3 wt%) could not further improve its superhydrophobicity. In contrast, high graphene loading could block the sponge pores and thus significantly decrease its absorption capacity. In addition, the residual oil could not be removed completely from the functional sponge unless the absorption capacity drops to 18 times of their own weight after five cycles [236]. Liu et al. coated polyurethane sponge with graphene oxide by reduction to form the corresponding graphene coating material [237]. This material showed an absorption capacity of 80–160 times

56

Carbonaceous Materials

Fig. 13 Graphene aerogel of 1.8 cm  1.1 cm  1.2 cm balancing on a dandelion. Available from: https://www.extremetech.com/extreme/153063graphene-aerogel-is-seven-times-lighter-than-air-can-balance-on-a-blade-of-grass/3.

160

Contact angle (degree)

140 120 100 80 60 40 20 0 BC

700

1000

1300

Pyrolysis temperature (°C) (A)

(B)

Fig. 14 The SEM image of the carbon aerogels from bacterial cellulose pellicles treated at 13001C (A) and effect of pyrolysis temperature on the water contact angle (B). Reproduced from Wu Z-Y, Li C, Liang H-W, Chen J-F, Yu S-H. Ultralight, flexible, and fire-resistant carbon nanofiber aerogels from bacterial cellulose. Angew Chem Int Ed 2013;52:2925–29.

of its own weight and was better than the corresponding graphene oxide coated polyurethane sponge (70–140 times of its own weight). As claimed by the authors, the results are due to graphene’s higher porosity and less polar groups present on its surface. However, graphene oxide coated sponges are much robust and could be reused at least 50 times with sustainable absorption capacity [236,237]. Liu et al. prepared superhydrophobic foam by grafting amine groups-functionalized graphene oxide with polyureathane sponges containing nitrile groups via in situ amidation reaction [238]. The composite material showed an improved absorption capacity of 26–41 times of its own weight for various oils and organic solvents when compared to pristine sponge.

2.2.3.2

Carbon Nanotube-Based Sorbents

Carbon nanotubes (CNTs) are well known one-dimensional carbons. They also have found applications in oil spilling cleanup, oil-water separation, gas separation and its storage. The CNTs generally have stronger mechanical properties, functionable suface

Carbonaceous Materials

57

and high absorption capacity [239] and they were functionalized by attaching p-phenylenediamine to increase their surface roughness and to decrease their surface free energy [240]. The modified CNTs have a surface area of 285 m2/g according to the Brunauer–Emmett–Teller (BET) analysis. The composite also showed good hydrophobicity and a maximum absorption capacity, that is 3–12 times of its own weight, while repelling water completely [240]. More importantly, the absorbed organic solvents and oil can be recovered conveniently by a heat treatment of solvent washings. This material can be reused a few times with a slight decrease in capacity. The CNTs-based magnetic sponges were prepared by a chemical vapor deposition of ferrocene and dichlorobenzene [241]. There are a few benefits of introducing magnetic property in the absorption of oils and organic solvents, including (1) magnetic sorbents could be collected conveniently by application of a magnetic field; (2) magnetic sorbents could be targeted and tracked to a specific location under the control of an external magnetic field. As claimed, the magnetic sponge has a highly porous, interconnected three-dimensional structure with a broad pore size of a few nanometers to micrometers. In addition, the sponge also showed promising mechanical properties of high volume recovery after strain deformation, structural fatigue resistence in oils, high chemical stability to acid, base, oil and organic solvents [241]. It is hydrophobic with a water contact angle of 1401 and twotimes heavier (15 mg/cm3) than a graphene aerogel. The sponge showed a relatively low absorption capacity of 49–56 times of its own weight for different oils and organic solvents, but that was lower than the corresponding CNT sponges [241]. It was reported that the material could be recovered and used at least 1000 times with sustainable framestructure, hydrophobicity and absorption capacities. Hashim et al. reported a three-dimensional elastic nanotube-based solid material [242], prepared by a chemical vapor deposition and doped with boron element leading to the formation of boron-based junctions. The incorporation of boron species enhanced the interconnections and mechanical properties of the material with an absorption capacity of 22–180 for different oils and organic solvents [242]. Similar with graphite, the CNTs were also used to coat porous materials such as sponges to improve their hydrophobicity and recyclability. This type of CNT coated sponges have been widely used as sorbents for oil spill cleanup. The absorbed oil can be recovered either by mechanical squeezing or vacuum sucking. The porous pristine sponges are three-dimensional polymer materials which absorb water and oil. After coating with CNTs, the nature of the sponge surfaces altered to be more hydrophobic. However, this material demonstrated enhanced thermal and mechanical stability with an absorption capacity of 15–50 times of their own weight [243].

2.2.3.3

Carbon Aerogel-Based Sorbents

Carbon aerogels are often made of carbon nanofibers impregnated with resorcinol–formaldehyde aerogel, and pyrolyzed in an inert atmosphere to form a carbon matrix. Carbon aerogels have found application in energy storage as useful electrodes in creating supercapacitors with a capacitance density of 104 F/g. In addition, these materials are also efficient solar energy collectors as they reflect only 0.3% of the infrared radiation (250 nm–14.3 mm) [244]. Carbon aerogels are commercially available in solid shapes, powders, or as composite papers. Graphenes have also been used to functionalize porous materials such as fibers, cottons and sponges to form new oil absorbents due to their intrinsically hydrophobic property. Methods have been developed to modify the porous materials with dip coating and grafting polymerization [235,238]. These porous materials usually have good mechanical properties, low-cost and reasonable absorption capacity, and are potentially useful in oil spill cleanup. Even though the commercially available porous materials absorb both oil and water, the functionalized graphene is expected to enhance their hydrophobicity and improve their capability of absorbing oil. Tai et al. reported the coating of melamine sponge with graphene nanosheet [236]. In this case, the coating material displayed superhydrophobic and superoleophilic properties, and a high absorption capacity of 54–165 times of their own weight for different oils and organic compounds. It was reported that a lower graphene loading (o5.1 wt%) could not alter the wettability of the melamine sponge; meanwhile, overloading (47.3 wt%) could not further improve its superhydrophobicity. In contrast, high graphene loading could block the sponge pores and thus significantly decrease its absorption capacity. In addition, the residual oil could not be removed completely from the functional sponge unless the absorption capacity drops to 18 times of their own weight after five cycles [236]. Liu et al. coated polyurethane sponge with graphene oxide by reduction to form the corresponding graphene coating material [237]. This material showed an absorption capacity of 80–160 times of its own weight and was better than the corresponding graphene oxide coated polyurethane sponge (70–140 times of its own weight). As claimed by the authors, the results are due to graphene’s higher porosity and less polar groups present on its surface. However, graphene oxide coated sponges are much robust and could be reused at least 50 times with sustainable absorption capacity [236,237]. Liu et al. prepared superhydrophobic foam by grafting amine groups-functionalized graphene oxide with polyureathane sponges containing nitrile groups via in situ amidation reaction [238]. The composite material showed an improved absorption capacity of 26–41 times of its own weight for various oils and organic solvents when compared to pristine sponge.

2.2.3.4

Carbon Nanotube-Based Sorbents

Carbon nanotubes (CNTs) are well known one-dimensional carbons. They also have found applications in oil spilling cleanup, oil-water separation, gas separation and its storage. The CNTs generally have stronger mechanical properties, functionable suface and high absorption capacity [239] and they were functionalized by attaching p-phenylenediamine to increase their surface

58

Carbonaceous Materials

roughness and to decrease their surface free energy [240]. The modified CNTs have a surface area of 285 m2/g according to the BET analysis. The composite also showed good hydrophobicity and a maximum absorption capacity, that is 3–12 times of its own weight, while repelling water completely [240]. More importantly, the absorbed organic solvents and oil can be recovered conveniently by a heat treatment of solvent washings. This material can be reused a few times with a slight decrease in capacity. The CNTs-based magnetic sponges were prepared by a chemical vapor deposition of ferrocene and dichlorobenzene [241]. There are a few benefits of introducing magnetic property in the absorption of oils and organic solvents, including (1) magnetic sorbents could be collected conveniently by application of a magnetic field; (2) magnetic sorbents could be targeted and tracked to a specific location under the control of an external magnetic field. As claimed, the magnetic sponge has a highly porous, interconnected three-dimensional structure with a broad pore size of a few nanometers to micrometers. In addition, the sponge also showed promising mechanical properties of high volume recovery after strain deformation, structural fatigue resistence in oils, high chemical stability to acid, base, oil and organic solvents [241]. It is hydrophobic with a water contact angle of 1401 and twotimes heavier (15 mg/cm3) than a graphene aerogel. The sponge showed a relatively low absorption capacity of 49–56 times of its own weight for different oils and organic solvents, but that was lower than the corresponding CNT sponges [241]. It was reported that the material could be recovered and used at least 1000 times with sustainable framestructure, hydrophobicity and absorption capacities. Hashim et al. reported a three-dimensional elastic nanotube-based solid material [242], prepared by a chemical vapor deposition and doped with boron element leading to the formation of boron-based junctions. The incorporation of boron species enhanced the interconnections and mechanical properties of the material with an absorption capacity of 22–180 for different oils and organic solvents [242]. Similar with graphite, the CNTs were also used to coat porous materials such as sponges to improve their hydrophobicity and recyclability. This type of CNT coated sponges have been widely used as sorbents for oil spill cleanup. The absorbed oil can be recovered either by mechanical squeezing or vacuum sucking. The porous pristine sponges are three-dimensional polymer materials which absorb water and oil. After coating with CNTs, the nature of the sponge surfaces altered to be more hydrophobic. However, this material demonstrated enhanced thermal and mechanical stability with an absorption capacity of 15–50 times of their own weight [243].

2.2.3.5

Carbon Aerogel-Based Sorbents

Carbon aerogels are often made of carbon nanofibers impregnated with resorcinol-formaldehyde aerogel, and pyrolyzed in an inert atmosphere to form a carbon matrix. Carbon aerogels have found application in energy storage as useful electrodes in creating supercapacitors with a capacitance density of 104 F/g. In addition, these materials are also efficient solar energy collectors as they reflect only 0.3% of the infrared radiation (250 nm–14.3 mm) [244]. Carbon aerogels are commercially available in solid shapes, powders, or as composite papers. To significantly improve certain properties of carbon aerogels for specific applications, various additives, such as fiberglass, have been successfully incorporated using various continuous and discontinuous reinforcements. A more sustainable method of pyrolyzing biomass has been reported and recognized as green, economical and feasible process due to their sustainable resources of using biocompatible and biodegradable materials [233,245–247]. Carbon aerogels, produced from biomass by a pyrolysis process, often have porous three-dimensional interconnected frameworks and reasonable mechanical as well as fire-resistant properties. In molecular structures, the hydrophilic functional groups, such as hydroxyl groups, are removed during the pyrolytic process. It has been found that the mechanical and chemical properties of the biomass-derived aerogels are dependent upon the conditions used in the preparative processes, including temperature and biomass resources. Various biomass materials, such as winter melon, Kapok wadding materials and bacterial cellulose pellicles, have been used to prepare carbon aerogels with varying density and absorption capacity. From crystalline bacterial cellulose pellicles, an amorphous carbon aerogel (see Fig. 14(A)) was prepared through a free-drying process, that is followed by pyrolysis. This material shows a high absorption capacity ranging from 106 to 312 times of their own weight [233]. The water contact angles of the resulting aerogels were found to be increasing with the pyrolysis temperature as shown in Fig. 14(B). Wang et al. prepared carbon aerogels from cellulose by a wet chemical process of dissociation, gelation, regeneration, freedrying or carbonization [248]. The product showed fast absorption rate to organic dyes and heavy ions with an absorption capacity of 195–1947 mg/g. The significantly high porous aerogel (98% porosity) also exhibited good fire-resistance and recyclability. This type of aerogels could also be made by a hydrothermal carbonization followed by freeze drying and pyrolysis [249]. The resulting material was found to be high porous (499% porosity), interconnected three-dimensional carbon aerogel with the absorption capacity of 139 times of its own weight. In addition, the aerogel could be reused at least 10 times with sustained absorption capacity [249]. It has been established that graphene and CNTs have good absorption capacity and the corresponding carbon aerogels, prepared from them, have similar properties. Compared to other aerogels, graphene and CNT aerogels have the advantages of low density, good elasticity and excellent mechanical properties. Dong et al. synthesized a three-dimensional graphene-CNTs hybrid foam by the method of chemical vapor deposition [250]. The resulting highly flexible material is superhydrophocic showing a water contact angle of 152.31 with absorption capacity of 80–130 times of its own weight. It was believed that the nano-roughness, generated by CNTs and the trapping air in the macroscopic pores and the nanoscopic voids of CNTs, is responsible for the enhanced hydrophobicity [250]. Sun et al. prepared graphene oxide-CNTs-based aerogels by freeze drying of aqueous solution of CNTs and graphene oxide, followed by a gas reduction of hydrazine vapor [251]. According to SEM analysis, the resulting porous

Carbonaceous Materials

59

three-dimensional materials contributed to the structural frameworks of graphene with an entangled and spaghetti-like CNTs. This ultra-light aerogel showed an extremely low density of 0.16–22.4 mg/cm3 with excellent absorption capacity of 215–913 times of their own weight for different oils and organic solvents, as well as fast absorption rate [251]. In addition, this material could be recovered and reused. In summary, various carbon materials, such as graphite, graphene, carbon nanotubes and carbon aerogels, have been investigated as potential oil-spill cleanup sorbents due to their unique properties of high surface area, high hydrophobility, physical and chemical stability, recyclability, and excellent oil absorption capacity. Such superhydrophobic carbons have also been used to functionalize other materials, including organic polymers, to improve their hydrophobicity, oleophilicity, mechanical toughness, thermal stability and reusability. Composites of different types of carbon materials, such as carbon nanotubes and graphenes to form novel materials with superior properties, has been emerged as a new frontier in this area of study and, therefore, it is highly expected to continue developing carbon materials with ultrahigh absorption capacity and build up continuous absorptionremoval process in the specific area.

2.2.4

Carbonaceous Materials and Boron Hydrides for Hydrogen Storage

Hydrogen storage focuses on storing the lightweight hydrogen for subsequent applications as a compact energy carrier. It is well recognized that hydrogen storage will be playing important roles in grid energy storage, especially for the renewable energy sources such as wind power. Hydrogen is also used directly as the fuel for transportation of airplanes, ships and space shuttles that use liquid or slush hydrogen. Theoretically, hydrogen storage technologies can be classified as physical and chemical storages. In physical storage, the hydrogen molecules are stored by its compression and liquefication, whereas chemical storage use hydrides which reversibly release hydrogen. As of today, various approaches of using high pressures, liquid or slush hydrogen, porous material hydrogen absorption, and chemical compounds, have been explored for hydrogen storage. However, liquid hydrogen is not an economical source for hydrogen storage because both liquefaction and storage of liquid hydrogen require extremely low temperature ( 252.91C), thus a large energy is needed to cool hydrogen down to the required low temperature and maintain it. In addition, liquid hydrogen possesses less energy density by volume than gasoline. Compared to hydrocarbons, compressed hydrogen (B350 bar) has relatively higher energy density by weight rather than by volume. To prepare and store compressed hydrogen, large tanks are required. Further, production of compressed hydrogen also costs more energy. Hydrogen carriers, based on nanostructured carbons, have also been reported. This section discusses current progress in hydrogen storage using carbonaceous materials. Gas sorbents of carbonaceous materials could be produced from various organic sources such as polymers and biomass. These materials also have different morphologies, including carbon nanotubes, fullerenes and porous carbons. Depending on the precursor and synthetic methodologies, the carbonaceous materials can be tuned for hydrogen storage by considering their composition and morphology (e.g., pore size and shape). Mesoporous carbons of nanostructued carbonaceous materials and activated carbons usually have high surface area, low density, reasonable thermal, and chemical stability. These materials have exhibited high hydrogen storage capacities [252,253]. Among the carbonaceous materials, activated carbon is one of the promising sorbents due to its low-cost and readily available starting materials. Nevertheless, more efforts are expected to design and synthesize active carbon materials with controlling pore size [253]. It has been found that the incorporation of heteroatomic and metal species, such as B, N, and alkaline earth metals, into the carbonaceous materials benefits the material performance and improves their hydrogen storage capability [254–257]. Interactions of hydrogen with materials are defined as physisorption, where hydrogen is absorbed by means of van der Waals forces, and chemisorption by formation of hydrides requiring dissociation of the absorbing hydrogens. Physisorption is generally considered as a storage model in porous materials, and lower temperature and higher pressure will benefit the absorption. Under moderate storage conditions (pressure less than 100 bar and temperature of around 300K), carbonaceous materials were found to store hydrogen without excess of 2 wt% and that was far below the US Department of Energy’s (DOE) target of 5.5 wt% [257]. The low capacity is due to the low binding energies of the molecular hydrogen and easy desorption from the carbonaceous materials. Extensive efforts are being made to significantly improve the hydrogen storage capacity. Small amounts of transition metals, alkaline earth metals, boron and nitrogen have been incorporated with the carbonaceous materials, either by physically mixing or chemically doping.

2.2.4.1

Graphene-Based Materials for Hydrogen Storage

Graphenes showed good potential to be an efficient hydrogen storage materials due to their low-density, high surface area and porous structure [258]. The chemical versatility of graphene carbons allow them to interact with hydrogen both by physisorption and chemisorption for sp2 and sp3 hybridized carbons, respectively [259,260]. At room temperature, graphenes provided a best hydrogen storage value of around 2–3 wt% that is not dramatically better than other carbonaceous materials. However, theoretical works showed that the capacity could be higher, around 5.0–6.5 wt%, at a specific interlayer spacing (B7–8 Å ) due to a cooperative effect of van der Waals forces [261]. A single graphene layer is not suitable for hydrogen storage. The hydrogen molecules are proposed to be incorporated into the interlayers of multiple graphene sheets to enhance the hydrogen absorption capacity.

Carbonaceous Materials

60

O2

H2

H2 O

5 nm H2

2 nm−1 5 nm (A)

(B)

Fig. 15 The TEM image (A) of mixed reduced graphene oxide/magnesium nanocrystal hybrids and hydrogen absorption mechanism (B). Reproduced from Cho ES, Ruminski AM, Aloni S, Liu Y-S, Guo J, Urban JJ. Graphene oxide/metal nanocrystal multilaminates as the atomic limit for safe and selective hydrogen storage. Nat Commun 2016;7:10804.

Cho et al. reported that the reduced graphene oxide/magnesium nanocrystal hybrids (as shown in Fig. 15) exhibited high performance for hydrogen storage with an exceptionally dense hydrogen storage of 6.5 wt% and 0.105 kg H2 per liter in total composite [262]. The mixed dimentional hybrids were prepared by a straightforward one-pot co-reduction process. The layered graphene and monodispersed Mg nanocrystals were formed simultaneously by reduction with lithium naphthalenide. Unlike commonly used ball milling technology, which are notoriously polydisperse, the method produced monodisperse Mg nanocrystals with an average size of 3.26 nm and stabilized by reducing graphene oxide. Interestingly, the hybrids with high loading of Mg nanocystals were reported remarkably stable, at least for 3 months of air exposure [262]. Hydrogen absorption was carried out at 2501C and 15 bar H2. The highest absorption capacity of the hybrid was 6.5 wt% and 0.105 kg H2 per liter in the total composite and that was the highest capacity reported as of today under similar conditions. According to the proposed hydrogen absorption mechanism, shown in Fig. 15(B), the in situ formed Mg nanocrystals interact with reducing graphene layers and thus are protected against invasion of oxygen and water molecules, while enabling rapid surface diffusion of hydrogen to form hydrides. The hydrides release hydrogen as being heated. The graphene layers are vital components in the hybrids to prevent oxygen and water molecules from penetrating, while allowing hydrogen molecules to diffuse and further interact with Mg nanocrystalls.

2.2.4.2

Heteroatoms-Doped Carbon Materials for Hydrogen Storage

Boron and other heteroatoms have been investigated as substituents in carbonaceous materials to enhance their stability and electrochemical behaviors. Boron is usually doped into the carbon frameworks in the trigonal coordination form and it behaves as an electron acceptor due to its three valence electrons. Therefore, boron dopping may cause a shift in Fermi level to the conduction band, thus affecting its electrochemical capacity and influencing the oxygen chemisorption process on the carbon surface. Further, it can change both physical and chemical properties of the materials, including their polarizability and solubility. Ariharan et al. prepared boron-substituted carbon materials by pyrolysis of a mixture of resorcinol and triethylborate in an inert atmosphere [263]. The resulting boron-incorporated carbonaceous materials showed sheet like morphology with a boron atomic percent of about 10.5%. Evidently, the boron content, surface area and its hydrogen storage capacity could be controlled by the temperature used in pyrolysis. The material showed 5.9 wt% hydrogen storage capacity at 298K and 100 bar H2 pressure [263]. Boronincorporated carbon materials were also obtained by template and bulk synthesis from sucrose using boric acid as a boron source [263]. Accordingly, the ordered mesoporous silica (SBA-15) was used as a template to prepare boron-doped carbon materials in the presence of boric acid. The boron content in the ordered mesoporous carbons is around 1.5 wt% determined by acid-base titration using mannitol. The material resulting from the template carbonization of sucrose in SBA-15 in the presence of boric acid has hexagonal morphology and surface area of 870 m2/g with a mesopore diameter of ca. 4nm and volume of ca. 1.0 cm3/g [264]. It showed a hydrogen storage capacity of B1.2 wt% and 37.5 mg/cm3 in micropores with the hydrogen pressure of 760 Torr at 77K, and a gravimetric capacity near 270 F/g. The B-doped materials exhibited higher values of gravimetric and interfacial capacity, when compared to those materials without boron. Lin et al. theoretically studied the hydrogen adsorption capacity of boronsubstituted carbon nanostructures, decorated with alkaline earth metals, through ab initio calculations, systematically [265]. It was expected that B-substituted carbon nanostructures with alkaline earth metals, such as Mg2 þ , can enhance the hydrogen storage capacity. The B-doped nano-carbon-Be2 þ can serve as a material of high hydrogen storage capacity reaching up to 13.38 wt%, and that is consistent with the experimental results [265]. Li et al. demonstrated that the hydrogen storage capacity of a B-doped graphene can reach to 15.1 wt% [266].

Carbonaceous Materials

61

Nitrogen-containing carbonaceous materials such as N-doped activated carbons are also potential candidates for hydrogen storage and, as such, various carbon species have been explored as storage materials for hydrogen energy applications. Despite several advantages of high surface area and low density, carbonaceous materials also have inherent drawback that may limit their applictions. One main drawback is their low adsorption heat, which is 15.1 kJ/mol at 298K [267]. Hydrogen molecules are adsorbed on the surface of carbonaceous materials by weak van der Waals force. Therefore, ultra-microporous carbons my offer high hydrogen storage capacity due to their narrower pores which strengthen the interaction between hydrogen and surface [268]. Sethia et al. reported the high nitrogen microporous activated carbons with high nitrogen loading and surface area of 526–2386 m2/g and pore volumes of 0.26–1.16 cm3/g [269]. This type of carbonaceous materials were prepared by heating a nitrogen-rich carbon prescursor with KOH at varying temperature in the range of 550–7001C. Accordingly, these carbons showed high hydrogen storage capacity at 77K and 1 bar hydrogen pressure because of their increasing porosity. One optimizing carbonaceous material with a nitrogen content of 22.3 wt% and a porosity of 0.59 nm in size, exhibited a hydrogen storage capacity of 2.94 wt%. It has been found that the capacity is linearly related to the porosity (0.5–0.7 nm) of the materials rather than their total surface area or the total pore volume [269]. The ultra-micropore volume mainly controls hydrogen adsorption at 77K, whereas the large pores (41 nm) are involved in hydrogen adsorption at higher pressure.

2.2.4.3

Transition Metal Nanoparticles-Modified Carbonaceous Materials for Hydrogen Storage

It has been repeatedly demonstrated that the addition of transition metal nanoparticles to carbonaceous materials is able to enhance the hydrogen storage capacity by dissociating hydrogen molecules and enabling a chemical adsorption process, named as spillover effect. Evidently, hydrogen adsorption to carbonaceous materials is well recognized as a surface-area dependent process in which the hydrogen molecules are adsorbed by van der Waals forces. Nanostructure carbons with specific pore size and shapes show the overlap of attractive van der Waals forces due to multiple surface, which increases the adsorption energy and hydrogen storage capacity. Addition of small amounts of transition metals, including palladium nanoparticles, has been reported to increase the hydrogen storage capacity up to ninefolds [270]. The enhancement mechanism is proposed as hydrogen spillover to carbon nanostructures [271]. Hydrogen spillover mechanism consists of three steps: (1) adsorption and dissociation of hydrogen molecules on a transition metal nanoparticles to form metal-hydrogen bonds; (2) migration of atomic hydrogen from the surfaces of transition metal nanoparticles to carbon surface; and (3) diffusion of the atomic hydrogen along the carbon surface forming a C–H bond, eventually. According to this spillover mechanism, it is understandable that the carbon nanostructure is crucial because the majority of the adsorbing hydrogens interact with carbonaceous materials, rather than with the low loading transition metals. The hydrogen storage capacity and reversibility are closely associated with carbon materials’ morphology (e.g., accessible surface area, pore shape and size), dopants and defects. Shiraz et al. prepared carbon nanotubes decorated with palladium nanoparticles for hydrogen storage [272]. The material was synthezised by electrochemical anodization process through chemical reactions. The carbon nanotubes were grown over a palladium-decorated porous precursor using chemical vapor deposition with methane as a feeding gas at 9801C. The resulting material showed a popcorn-like structure with a hydrogen storage capacity of B2.05 wt% based on the electrochemical charge/ discharge analysis [272]. This material can also be used at least 100 times with slight drop of 0.05%. Nickel-doped graphene oxide was synthesized by wet chemistry process followed by hydrogen reduction at 3501C [273]. The metal nanoparticles with a size distribution of 11–110 nm were inserted into the graphene layers, reductively. The reduced graphene, doped with nickel and palladium nanoparticles, has a surface area of 25.43 m2/g and a hydrogen storage capacity of 0.13 wt% at room temperature (293.15K) with 800 mmHg hydrogen pressure [273]. It was claimed that this type of materials showed a great potential in hydrogen storage under optimum conditions. However, the inherent challenge for the carbon-based material is its capacity to store hydrogen, chemically. The theoretical limit of one hydrogen atom per carbon sets a storage capacity of 7.7 wt% [274,275]. Addition of the transition metal nanoparticles, as the source of atomic hydrogen according to the hydrogen spillover mechanism, could decrease the hydrogen gravimetric storage capacity to a level below the ultimate goal of 7.5 wt% as per US Department of Energy requirement. The total prospects of transition metal-doped carbonaceous materials might not be promising. Nevertheless, there are several possibilities for optimizing transition metal nanoparticle size and loading amount, that are worthy of investigation to improve the hydrogen storage performance. Carbonaceous materials with a high surface area and hexagonally structured carbons appear to be the most promising hydrogen sorbents based on hydrogen spillover mechanism. In summary, nanostructured carbonaceous materials provide considerable opportunities to overcome the drawbacks of traditional bulk hydrogen storage technologies. Carbonaceous materials with increasing surface area and porosity present additional binding sites for atomic hydrogen on the surface and in the pores, thus providing the enhanced hydrogen storage capacity. Besides carbon nanostructures, other materials with high surface area, such as metal-organic-frameworks and organic polymer networks, also demonstrated great potential for hydrogen storage due to their high crystallinity and porosity, and their ultrahigh surface area. The nanostructured hybrids of transion metal-decorated reduced graphenes and boron and/or nitrogen-doped carbons may benefit in terms of both thermodynamics of hydrogen adsorption and the kinetics of the hydrogen uptake and release processes. The nanoporous carbonaceous materials prevent particle growth and agglomeration to form bulky particles, and thus enable the reusability of the hydrogen sorbents with sustainable capacity. Therefore, it is highly desired to develop new materials and structures that strengthen both physisorption and chemisorption capability for hydrogen storage. In addition, the materials should

62

Carbonaceous Materials

be thermally and chemically stable, easily available, low-cost, recyclable along with outstanding hydrogen adsorption and desorption properties.

2.2.4.4

Applications of Borohydrides in Hydrogen Generation and Storage

The sodium borohydride (NaBH4), one of the most common hydrides of boron, has undergone extensive research efforts globaly, as part of sponsored projects of governmental agencies, due to their outstanding ability for hydrogen storage. Borohydrides possess excellent hydrogen densities, release rates and safety characteristics, and they can hold substantial hydrogen by weight and volume [276,277]. Also, they have an excellent potential for high energy density storage at room temperature and at atmospheric pressure [278], and thus they have been extensively studied to evaluate its potential for portable, automobile and stationary applications. It has the advantages of a high potential hydrogen density (maxium 10.9 wt%) together with a safe and ready hydrogen release through a hydrolysis reaction that can be controlled catalytically [279,280]. Metal borohydrides are formally called metal tetrahydroborate, an extensive class having the general composition as M(BH4)n, where M may be Li, Na, K, Mg, Ca, Sc, Ti, V, Cr, Mn, Zn, Zr, Al, U, and so on, and n is normally between 2 and 4. Metal borohydrides can contain substantial amounts of hydrolytically or thermally accessible hydrogen on both the weight and volume basis, depending on the atomic number and valence requirements of the metal. They can appear in either solid or liquid form, and can be heated directly, passed through a catalyst-containing reactor, or combined with water (i.e. hydrolysis) or other reactants to produce hydrogen. It was first noted by Schlesinger et al. that it is possible to form a highly stable aqueous solution of NaBH4 by dissolving it in basic solution. The hydrolysis reaction (NaBH4 þ 2H2O - NaBO2 þ 4H2; DH ¼ 75 kJ mol 1) can then be initiated on demand by bringing the solution into contact with a heterogeneous catalyst, making the release of hydrogen very easy and control [279,280]. Based on the reaction stoichiometry, 1 g of fully hydrolyzed NaBH4 will produce 2.37 L of hydrogen at standard temperature and pressure. On a reactants-only basis, this gives a gravimetric hydrogen storage capacity (GHSC) of 10.8 wt%, which is more than required storage density target set by DOE. Aqueous borohydride is stable under ordinary conditions and liberates hydrogen in a safe and controllable manner and that makes it a potentially promising approach in hydrogen storage, particularly for portable applications. This system is also much safer and more controllable than the hydrolysis of solid hydrides, because the solution effectively acts as a thermal buffer by absorbing the exothermic heats of reaction and preventing thermal loss. The release rate of hydrogen is easily regulated by controlling the amount of solution in contact with the catalyst (or vice versa), allowing the system to meet the dynamic power demands of a fuel cell vehicle. A paper published in 2000 by Millennium Cell, Inc. was the first to demonstrate a portable hydrogen storage system based on aqueous NaBH4 solutions, and thus stimulated further research in this field [280,281]. However, it is arguable about the ability of such systems to meet the DOE targets. The GHSC of real storage systems will invariably be lower than the theoretical 10.8 wt%, due to excess water, required to dissolve the NaBH4 and its by-product, NaBO2, as well as due to added mass of the reaction and storage vessels. These shortcomings were identified in the DOE’s 2007 publication of a review paper on NaBH4 as a hydrogen storage material [282]. Unlike the reversible complex hydrides, borohydrides are considered as “one-way” single-use fuels as the waste materials or byproducts must be removed from the vehicle for off-board regeneration [283]. During the course of catalyst development for the borohydride hydrolysis reaction to release hydrogen, low cost and highly effective transition metal catalysts such as Ru, Pd, Pt, Pd–Pd, Ru–Pd alloys [280,284–288], and Ni–Co–B powders [289] exhibited outstanding activity in the essential hydrogen generation system for practical onboard applications. Amendola et al. [280,284] reported the application of Ru-catalyzed hydrolysis of aqueous BH4 solution as hydrogen generator for proton exchange membrane fuel cells (PEMFC). While the noble metal catalysts show excellent catalytic activity, their use in practical applications is restricted by their high material cost. A highly stable and active nickel boride catalyst (NixB) was prepared and tested for the catalytic hydrolysis of alkaline NaBH4 solution [290]. It was found that after heat treatment at 1501C in vacuum, the NixB catalyst showed greatly enhanced catalytic activity and operational stability. Under suitable experimental conditions, the hydrolysis reaction can produce 6:75 wt% hydrogen at 451C and 44.0 wt% hydrogen even at room temperatures, exhibiting much higher hydrogen storage capacity than currently used alloys for hydrogen storage. Since the NixB catalyst is inexpensive and easy to prepare, it is feasible to use this catalyst in the construction of practical hydrogen generators for portable and in situ applications. Among numerous boron compounds, ammonia borane is indeed a promising chemical hydrogen storage material [291–293]. The appeal for some chemical hydrides originates from their high gravimetric and volumetric capacities and near-ambient operating conditions which are often less than 801C (at 0.1 MPa hydrogen pressure). For example, ammonia borane (NH3BH3) contains over 19 wt% H2 and 150 g H2/L of hydrogen by weight and volume, respectively (materials basis) and practically releases over one equivalent of H2 rapidly at 701C using a transition metal catalyst [294]. Likewise, materials related to ammonia borane, for example ammonia triborane (NH3B3H7) [295], are being actively pursued as potential hydrogen storage materials. The development of ammonia borane in the field of solid-state chemical hydrogen storage is however hindered by issues relating to thermolytic decomposition. Upon heating (e.g., over the temperature range 80–2001C), ammonia borane decomposes more than its dehydrogenation and high amounts of undesired gaseous byproducts (e.g., borazine, diborane and ammonia) are released along with two equivalents of hydrogen [296]. Concomitantly, the formation of a solid residue of complex nature, with the empirical formula [BNHx]n where x o2, is possible (as suggested elsewhere). Therefore, it is most likely to obtain a mixture of polyaminoborane [H2N-BH2]n, polyiminoborane [HN¼ BH]n, o-polyborazylene [B3N3H4]n/3 and graphitic cross-linked polymer [B3N3Hy]n/3 with yo4 in such decomposition processes [297].

Carbonaceous Materials

63

Evidently, the decomposition scheme of ammonia borane is not acceptable from an application point of view. The temperatures are too high and the hydrogen is not pure. The solid residue cannot be totally and properly rehydrogenated by chemical recycling [298]. In this context, strategies of destabilization of ammonia borane have been investigated with the purposes of decreasing the dehydrogenation *temperature below 1001C, releasing pure hydrogen (while avoiding any byproducts), and forming a solid residue of simple composition, ideally that of polyborazylene. To date, five different destabilization strategies have been reported: (1) solubilization in organic solvent or in ionic liquid; (2) solubilization and addition of homogeneous catalyst; (3) doping with solid-state oxidant or acid; (4) nanoconfinement into the porosity of a host material; and (5) chemical modification towards the formation of derivatives of alkali amidoboranes (e.g., LiNH2BH3). By solubilizing ammonia borane in a suitable solvent, the intermolecular dihydrogen N–H


H–B network is disrupted [299]. This results in destabilization of ammonia borane by dehydrocoupling to release hydrogen at temperatures lower than 1001C. An example of this is the triglyme (b.p. 2161C) solution of 6 M ammonia borane, which is able to generate one equivalent of hydrogen in less than 1 h at 701C, and with no induction period [300]. Several metal acetylacetonates, Fe(O2C5H7)3, Co(O2C5H7)2, Ni(O2C5H7)2, Pd(O2C5H7)2, Pt(O2C5H7)2 and Ru(O2C5H7)3, are considered for assisting dehydrocoupling of ammonia borane in diglyme (0.135 M) at 501C [301,302]. The dehydrogenation kinetics can be improved by using metal-based catalysts. For example, ammonia borane (0.4 M) in toluene (b.p. 110.61C) is able to release one equivalent of hydrogen in less than 1 h when catalyzed by homogeneous ruthenium acetylacetonate-based catalyst at 601C [303]. From a mechanistic point of view, cyclic intermediates with singly dehydrogenated boron atoms are supposed to form first, that subsequently polymerizes to yield polyborazylene. Formation of cyclic intermediates, such as B-(cyclodiborazanyl) amine-borane, B-(cyclotriborazanyl)amine-borane and cyclotriborazane, were indeed reported [304]. Among those metal acetylacetonates catalysts, palladium acetylacetonate was found to be very reactive towards ammonia borane, even in the glove box when both solids were put into contact without solvent [305]. Ruthenium acetylacetonate showed to be an efficient precursor for the in situ formation of homogeneous catalyst for ammonia borane dehydrocoupling and the homogeneous character was unequivocally demonstrated with the help of the mercury poisoning test [303]. Perez et al. studied the dehydrocoupling of liquidstate ammonia borane NH3BH3 by the metal acetylacetonate-aided catalytic process [303]. The catalytic effect of the metal acetylacetonates was compared by monitoring the hydrogen evolution with time (over a maximum of 2 h). The main objective was to define the time at which the dehydrocoupling reaction could be stopped after the evolution of less than 0.3 mol of hydrogen per mole of ammonia borane (conversion rates lower than 30%) and to analyze the reaction intermediates at the early stages of the reaction. The hydrogen evolution curves are shown in Fig. 16. The most efficient metal acetylacetonate is Ru (O2C5H7)3 with the release of one equivalent of hydrogen in less than two hours. This is in good agreement with the results reported by Duman and Özkar [303] where one equivalent of hydrogen was liberated in about 70 min at 601C (mol ratio NH3BH3/Ru of 160). The mechanisms of ammonia borane dehydrocoupling are roughly independent on the metal nature of the acetylacetonate salts. Metal acetylacetonates catalyze ammonia borane’s dehydrocoupling by accelerating the reaction; they mainly have effect on the kinetics. Except for the borohydrides and the ammonia borane compounds, boron-based nanostructures can also be used as the hydrogen storage materials [306,307]. Nanostructured form of boron materials has been the main focus in recent research on hydrogen storage [278,313]. Materials at nanoscales can have advantages over their bulk counterparts with respect to molecular adsorptions on large specific surface area with potentially high binding energy. For storage materials operating at near room temperature, the binding energy of hydrogen should be in the range of B0.2–0.5 eV, which was addressed by the van’t Hoff equation. It demonstrated the possibility of designing nanometer materials with appropriate hydrogen binding properties.

Fe Co Ni Pt Ru

mol H2 per mol NH3BH3

1.0 0.8 0.6 0.4

30% conversion

0.2 0.0 0

20

40

60

80

100

120

Time (min) Fig. 16 Time evolution of hydrogenrelease by dehydrocoupling of ammonia borane (0.135 M) solubilized in diglyme at 501C and in the presence of metal acetylacetonate (mol ratio NH3BH3/M of 100; with M as Fe, Co, Ni, Pt or Ru for the metal of the acetylacetonate salts). Reproduced from Manon P, Philippe M, Umit BD. Mechanistic insights of metal acetylacetonate-aided dehydrocoupling of liquid-state ammonia borane NH3BH3, Adv Energy Res 2016;4:177–87.

64

Carbonaceous Materials

Recently, the tubular form of boron nitride has been shown to store hydrogen at elevated temperatures [308,309]. It has been reported that boron nitride (BN) nanotubes can uptake 1.8–2.6 wt% hydrogen under B10 MPa at room temperature [310] and collapsed BN nanotubes exhibit an even higher hydrogen adsorption capacity (4.2 wt%) than any multiwalled carbon nanotubes [311]. Jhi et al. have shown through computational simulations that BN can be a good hydrogen storage medium. Their study shows that deviations from sp2 bonding tends to increase the binding energy of hydrogen in BN. It is possible that layered materials of ionic character, more ionic than boron nitrides, with a moderate substitutional doping would have a substantially large binding energy, enough for storing hydrogen at room temperatures [308]. Chen et al. synthesized BN nanotubes through chemical vapor deposition over a wafer made by a LaNi5/B mixture and nickel powder at 1473K. The results verified that the BN nanotubes could store hydrogens by means of an electrochemical process. It was tentatively concluded that the improvement of the electrocatalytic activity by the surface modification with metal or alloy would enhance the electrochemical hydrogen storage capacity of BN nanotubes [312]. Jhi et al. theoretically studied the activated forms of BN nanotubes for potential applications to hydrogen storage with the use of pseudopotential density functional method [314]. The binding and diffusion energies of adsorbed hydrogen were particularly calculated. The calculated binding energy of hydrogen on activated boron nitride nanotubes was found to be in the right range for room-temperature storage. It was further shown that the diffusion through the active sites enables hydrogen to access the inner surface of the nanotubes, which leads to increase in the storage capacity [314]. This study provides a tangible solution to increase the operating temperature and capacity of hydrogen storage based on heteropolar nanomaterials such as boron nitride nanotubes. Li et al. investigated hydrogen adsorption and storage in Ca-coated boron fullerenes and nanotubes by means of density functional computations [315]. Their study shows that Ca can bind strongly to the surface of boron-80 (B80) fullerene and boron nanotubes, thus avoiding the notorious clustering problem. Fullerene B80 coated with 12 Ca atoms can store up to 60 H2 molecules with an average binding energy of 0.12–0.40 eV, corresponding to a gravimetric density of hydrogen storage of 8.2 wt%. The hydrogen storage capacity of a Ca-covered boron nanotube is 7.6 wt% with a binding energy of 0.10–0.30 eV [315]. The strong interaction between Ca and boron fullerenes and nanotubes is attributed to the charge transfer. The optimal molecular hydrogen adsorption energies make reversible hydrogen adsorption and desorption feasible at ambient conditions. Nonetheless, the hydrogen storage capacity of such boron structures will significantly decrease in their corresponding macroscopic boron materials. It remained a big challenge to assemble the suitable macroscopic materials for practical hydrogen storage. Porous structures with Ca-coated boron nanostructures, as building blocks, might be useful for high gravimetric and volumetric hydrogen storage capacity. In summary, for the hydrogen storage, new materials with improved performance, or new approaches to the synthesis and/or processing of existing materials, are highly desirable. Given the limitations identified in the conventional systems, a novel approach to design new materials is required to achieve the targets set by DOE [316].

2.2.5

Conclusions and Future Directions

In recent years, energy materials are receiving tremendous attention and research interests due to the increasing concern on the sustainable development of energy, economy, and society, which is closely related to the high efficiency storage and consumption of energy. To fulfill the newly emerging applications, advanced energy materials with superior integrated performance that enables high energy and power density and environmentally benign, convenient, and flexible storage of energy are highly demanding [317]. This chapter summarizes latest advanced development of the carbonaceous and related boraneous materials and their applications in energy and hydrogen storage as well as in oil spilling cleanup. Nanostructural carbonaceous materials are extremely important for energy storage, particularly for advanced energy devices, such as capacitors. As of today, different types of carbon nanoframeworks have been investigated as main composites of the pseudo-capacitors and supercapacitors. It can be concluded that the morphology of the carbon constituents are essential for the power capacitance. Based on the recently published results, it is expected that more efforts should be made to material development to construct high performance carbonaceous hybrids. The newly designed materials may show excellent efficiency in ions charge, transmission, adsorption, discharge, and also good balance in surface area and volumetric capacitance. Boron is an electron deficient element and able to enter the carbon lattice by substituting carbon atoms at the trigonal sites and acts as electron acceptor. The boron doping tunes the electronic structure of the carbonaceous material and thus affect the electric double layer capacitance. In addition, boron doping shows catalytic effects on oxygen chemisorption on carbon surface. Therefore, optimizing boron and/or other heteroatom doping is expected to modify the electrochemical capacitance of carbon materials. As above discussed, boron-rich semiconductors show good benefits for thermoelectric energy conversion. The type materials can potentially increase the efficiencies of 25% or even more after further optimization, and that level of efficiency is comparable with smaller electrical power stations. Owing to their inherient unrealibity of the existing thermoelectric devices, it has been highly expected to use boron-rich materials as ecological and economical candidates in thermoelectric energy conversion. At present, studies on boron particle and boron-based fuel-rich propellants are comprehensive. However, studies focused directly on primary combustion products and their secondary combustion are lacking. The composition of primary combustion products is complicated, and different ingredients may react and influence one another. Therefore, a study on primary combustion products requires various methods. Following the worldwide ever-increasing energy demand, the depletion of fossil fuels is getting worsening. Rapid and efficient recovering of the spilled oil will significantly reduce the industrial oil waste and environmental pollution. Carbonaceous materials

Carbonaceous Materials

65

also found wide applications in the unique area of highly efficient adsorbents because of their inherent advantages, such as high surface area, hydrophobility, physical and chemical stability, recyclability, more importantly, their promising oil absorption capacity. Nanostructural carbons have also been used to modify other porous materials to improve their hydrophobicity, oleophilicity, mechanical toughness, thermal stability and reusability. In this area, careful integration of different types of carbon nanomaterials, including carbon nanotubes and graphenes, to prepare nanocomposites with superior properties and high performance has been emerged as a new frontier. In addition, build up of the continuous absorption-removal processes is also warranted in this specific area. For hydrogen storage, carbonaceous materials have demonstrated great advantages to surmount the drawbacks of traditional technologies. They offer higher surface area, more porosity and thus improve their storage performance. The nanostructural hybrids of nickel nanoparticles-decorated with reducing graphenes and boron-doped carbons benefit both the thermodynamics and kinetics of hydrogen adsorption and release processes. It is expected to discover new materials with novel structures that strengthen both physisorption and chemisorption capability for hydrogen storage. For boron materials-based hydrogen adsorbents, it is essential to discover new methodologies to produce macroscopic boron materials with sustainable gravimetric and volumetric hydrogen storage high capacity, similar to that observed for their nanostructure counterparts. The Ca-coated boron nanostructures should prove to be the suitable building blocks to prepare porous boron materials with high capacity.

Acknowledgment The authors thank financial support from the School of Pharmacy, Macau University of Science and Technology and Northern Illinois University.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33]

Miller JR, Burke AF. Electrochemical capacitors: challenges and opportunities for real-world applications. In: Spring 2008. The Electrochemical Society Interface, p. 53–7. Stoller MD, Park S, Zhu Y, An J, Ruoff RS. Graphene-based ultracapacitors. Nano Lett. 2008;8:3498–502. Chmiola J, Yushin G, Gogotsi Y, Portet C, Simon P, Taberna PL. Anomalous increase in carbon capacitance at pore sizes less than 1 nm. Science 2006;313:1760–3. Bureau of Naval Personnel, Fundamentals of Electronics, Vol 1b: basic Electricity, Alternating Current, NAVPERS 93400A-1b; 1965. 197. Pandolfo AG, Hollenkamp AF. Carbon properties and their role in supercapacitors. J Power Sources 2006;157:11–27. Lota G, Centeno TA, Frackowiak E, Stoeckli F. Improvement of the structural and chemical properties of a commercial activated carbon for its application in electrochemical capacitors. Electrochim. Acta 2008;53:2210–6. Available from: https://en.wikipedia.org/wiki/Carbon; 2016 [accessed 20.11.16]. Simon P, Gogotsi Y. Capacitive energy storage in nanostructured carbon-electrolyte system. Acc Chem Res 2013;46:1094–103. Bartelmess J, Giordani S. Carbon nano-onions (multi-layer fullerenes): chemistry and applications. Beilstein J Nanotechnol 2014;5:1980–98. Portet C, Yushin G, Gogotsi Y. Electrochemical performance of carbon onions, nanodiamonds, carbon black and multiwalled nanotubes in electrical double layer capacitors. Carbon 2007;45:2511–8. Serin V, Brydson R, Scott A, et al. Evidence for the solubility of boron in graphite by electron energy loss spectroscopy. Carbon 2000;38:547–54. Tsai S, Lee C, Chao C, Shih H. A novel technique for the formation of carbon-encapsulated metal nanoparticles on silicon. Carbon 2000;38:781–5. Maquin B, Derré A, Labrugère C, Trinquecoste M, Chadeyron P, Delhaès P. Submicronic powders containing carbon, boron and nitrogen: their preparation by chemical vapour deposition and their characterization. Carbon 2000;38:145–56. He C, Zhao N, Shi C, Du X, Li J. Carbon nanotubes and onions from methane decomposition using Ni/Al catalysts. Mater. Chem. Phys. 2006;97:109–15. Hou S-S, Chung D-H, Lin T-H. High-yield synthesis of carbon nano-onions in counterflow diffusion flames. Carbon 2009;47:938–47. Gao Y, Zhou Y, Park J, et al. Resonant excitation of precursor molecules in improving the particle crystallinity, growth rate and optical limiting performance of carbon nano-onions. Nanotechnology 2011;22:165604. Choucair M, Stride JA. The gram-scale synthesis of carbon onions. Carbon 2012;50:1109–15. Zhao M, Song H, Chen X, Lian W. Large-scale synthesis of onion-like carbon nanoparticles by carbonization of phenolic resin. Acta Mater 2007;55:6144–50. Chung D-H, Lin T-H, Hou S-S. Flame synthesis of carbon nano-onions enhanced by acoustic modulation. Nanotechnology 2010;21:435604. Bystrzejewski M, Rummeli MH, Gemming T, Lange H, Huczko A. Catalyst-free synthesis of onion-like carbon nanoparticles. New Carbon Mater 2010;25:1–8. Choi M, Altman IS, Kim YJ, et al. Formation of shell-shaped carbon nanoparticles above a critical laser power in irradiated acetylene. Adv Mater 2004;16:1721–5. Huang J, Yasuda H, Mori H. Highly curved carbon nanostructures produced by ball-milling. Chem Phys Lett 1999;303:130–4. Cabioc'h T, Jaouen M, Thune E, Guerin P, Fayoux C, Denanot M. Carbon onions formation by high-dose carbon ion implantation into copper and silver. Surf Coat Technol 2000;128:43–50. Chen XH, Yang HS, Wu GT, et al. Generation of curved or closed-shell carbon nanostructures by ball-milling of graphite. J Cryst Growth 2000;218:57–61. Thune E, Cabioc'h T, Guérin P, Denanot M-F, Jaouen M. Nucleation and growth of carbon onions synthesized by ion-implantation: a transmission electron microscopy study. Mater Lett 2002;54:222–8. Lange H, Sioda M, Huczko A, Zhu Y, Kroto H, Walton D. Nanocarbon production by arc discharge in water. Carbon 2003;41:1617–23. Tomita S, Burian A, Dore JC, LeBolloch D, Fujii M, Hayashi S. Diamond nanoparticles to carbon onions transformation: x-ray diffraction studies. Carbon 2002;40:1469–74. Ishigami M, Cumings J, Zettl A, Chen S. A simple method for the continuous production of carbon nanotubes. Chem Phys Lett 2000;319:457–9. Sano N, Wang H, Chhowalla M, Alexandrou I, Amaratunga G. Nanotechnology: synthesis of carbon ‘onions’ in water. Nature 2001;414:506–7. Lian W, Song H, Chen X, et al. The transformation of acetylene black into onion-like hollow carbon nanoparticles at 10001C using an iron catalyst. Carbon 2008;46:525–30. Fan J-C, Sung H-H, Lin C-R, Lai M-H. The production of onion-like carbon nanoparticles by heating carbon in a liquid alcohol. J Mater Chem 2012;22:9794. Xu N, Chen J, Deng S. Effect of heat treatment on the properties of nano-diamond under oxygen and argon ambient. Diamond Relat Mater 2002;11:249–56. Mykhaylyk OO, Solonin YM, Batchelder DN, Brydson R. Transformation of nanodiamond into carbon onions: a comparative study by high-resolution transmission electron microscopy, electron energy-loss spectroscopy, X-ray diffraction, small-angle X-ray scattering, and ultraviolet Raman spectroscopy. J Appl Phys 2005;97:074302–16.

66

Carbonaceous Materials

[34] Butenko YV, Kuznetsov VL, Paukshtis EA, et al. The thermal stability of nanodiamond surface groups and onset of nanodiamond graphitization. Fullerenes Nanotubes Carbon Nanostruct 2006;14:557–64. [35] Xie FY, Xie WG, Gong L, et al. Surface characterization on graphitization of nanodiamond powder annealed in nitrogen ambient. Surf Interface Anal 2010;42:1514–8. [36] Zou Q, Li YG, Lv B, Wang MZ, Zou LH, Zhao YC. Transformation of onion-like carbon from nanodiamond by annealing. Inorg Mater 2010;46:127–31. [37] Baidakova M, Siklitsky V, Vul AY. Ultradisperse-diamond nanoclusters. Fractal structure and diamond–graphite phase transition. Chaos Solitons Fractals 1999;10:2153–63. [38] Weingarth D, Zeiger M, Jäckel N, Aslan M, Feng G, et al. Graphitization as a universal tool to tailor the potential-dependent capacitance of carbon supercapacitors. Adv Energy Mater 2014;4:1400316. [39] Aleksenskii A, Baidakova M, Vul AY, Dideikin A, Siklitskii V. Effect of hydrogen on the structure of ultradisperse diamond. Phys Solid State 2000;42:1575–8. [40] Mykhailiv O, Lapinski A, Molina-Ontoria A, Regulska E, Echegoyen L, et al. influence of the synthetic conditions on the structural and electrochemical properties of carbon nano-onions. ChemPhysChem 2015;16:2182–91. [41] Plonska-Brzezinska ME, Lapinski A, Wilczewska AZ, et al. The synthesis and characterization of carbon nano-onions produced by solution ozonolysis. Carbon 2011;49:5079–89. [42] Plonska-Brzezinska ME, Lewandowski M, Błaszyk M, Molina-Ontoria A, Lucin´ski T, Echegoyen L. Preparation and characterization of carbon nano-onion/PEDOT: pSS composites. ChemPhysChem 2012;13:4134–41. [43] Plonska-Brzezinska ME, Mazurczyk J, Palys B, et al. Preparation and characterization of composites that contain small carbon nano-onions and conducting polyaniline. Chem Eur J 2012;18:2600–8. [44] Gubarevich AV, Kitamura J, Usuba S, Yokoi H, Kakudate Y, et al. Onion-like carbon deposition by plasma spraying of nanodiamonds. Carbon 2003;41:2601–6. [45] Plonska-Brzezinska ME, Molina-Ontoria A, Echegoyen L. Post-modification by low-temperature annealing of carbon nano-onions in the presence of carbohydrates. Carbon 2014;67:304–17. [46] Xiao J, Ouyang G, Liu P, Wang C, Yang G. Reversible nanodiamond-carbon onion phase transformations. Nano Lett 2014;14:3645–52. [47] Du AB, Liu XG, Fu DJ, Han PD, Xu BS. Onion-like fullerenes synthesis from coal. Fuel 2007;86:294–8. [48] McDonough JK, Gogotsi Y. Carbon onions: synthesis and electrochemical applications. Electrochem Soc Interface 2013;22:61–5. [49] Kuznetsov VL, Chuvilin AL, Butenko YV, Mal'kov IY, Titov VM. Onion-like carbon from ultra-disperse diamond. Chem Phys Lett. 1994;222:343–8. [50] Zeiger M, Jäckel N, Aslan M, Weingarth D, Presser V. Understanding structure and porosity of nanodiamond-derived carbon onions. Carbon 2015;84:584–98. [51] Sano N, Wang H, Alexandrou I, Chhowalla M, Teo KBK, et al. Properties of carbon onions produced by an arc discharge in water. J Appl Phys. 2002;92:2783. [52] Gao Y, Zhou YS, Qian M, et al. Chemical activation of carbon nano-onions for high-rate supercapacitor electrodes. Carbon 2013;51:52–8. [53] Béguin F, Presser V, Balducci A, Frackowiak E. Carbons and Electrolytes for Advanced Supercapacitors. Adv Mater 2014;26:2219–51. [54] Jäckel N, Weingarth D, Zeiger M, Aslan M, Grobelsek I, Presser V. Comparison of carbon onions and carbon blacks as conductive additives for carbon supercapacitors in organic electrolytes. J Power Sources 2014;272:1122–33. [55] Fulvio PF, Mayes RT, Wang X, et al. Brick-and-mortar self-assembly approach to graphitic mesoporous carbon nanocomposites. Adv Funct Mater 2011;21:2208–15. [56] Kovalenko I, Bucknall DG, Yushin G. Detonation nanodiamond and onion-like-carbon-embedded polyaniline for supercapacitors. Adv Funct Mater 2010;20:3979–86. [57] Makgopa K, Ejikeme PM, Jafta CJ, Raju K, Zeiger M, et al. A high-rate aqueous symmetric pseudocapacitor based on highly graphitized onion-like carbon/birnessite-type manganese oxide nanohybrids. J Mater Chem A 2015;3:3480–90. [58] Borgohain R, Li J, Selegue JP, Cheng Y-T. Electrochemical study of functionalized carbon nano-onions for high-performance supercapacitor electrodes. J Phys Chem C 2012;116:15068–75. [59] Plonska-Brzezinska ME, Brus DM, Molina-Ontoria A, Echegoyen L. Synthesis of carbon nano-onion and nickel hydroxide/oxide composites as supercapacitor electrodes. RSC Adv 2013;3:25891–901. [60] Azhagan MVK, Vaishampayan MV, Shelke MV. Synthesis and electrochemistry of pseudocapacitive multilayer fullerenes and MnO2 nanocomposites. J Mater Chem A 2014;2:2152–9. [61] Borgohain R, Selegue JP, Cheng Y-T. Ternary composites of delaminated-MnO2/PDDA/functionalized-CNOs for high-capacity supercapacitor electrodes. J Mater Chem A 2014;2:20367–73. [62] Wang Y, Yu SF, Sun CY, Zhu TJ, Yang HY. MnO2/onion-like carbon nanocomposites for pseudocapacitors. J Mater Chem 2012;22:17584–8. [63] Mykhailiv O, Imierska M, Petelczyc M, Echegoye L, Plonska-Brzezinska ME. Chemical versus electrochemical synthesis of carbon nano-onion/polypyrrole composites for supercapacitor electrodes. Chem Eur J 2015;21:5783–93. [64] Anjos DM, McDonough JK, Perre E, et al. Pseudocapacitance and performance stability of quinone-coated carbon onions. Nano Energy 2013;2:702–12. [65] Jost K, Perez CR, McDonough JK, et al. Carbon coated textiles for flexible energy storage. Energy Environ Sci 2011;4:5060–7. [66] Liu Y, Kim DY. Enhancement of capacitance by electrochemical oxidation of nanodiamond derived carbon nano-onions. Electrochim Acta 2014;139:82–7. [67] An G, Yu P, Xiao M, Liu Z, Miao Z, Ding K. Low-temperature synthesis of Mn3O4 nanoparticles loaded on multi-walled carbon nanotubes and their application in electrochemical capacitor. Nanotechnology 2008;19:275709. [68] He L, Zhang G, Dong YZ, et al. Polyetheramide template synthesis of monodisperse Mn3O4 nanoparticles with controlled size and study of the electrochemical properties. Nano–Micro Lett 2014;6:38–45. [69] Ma SB, Ahn KY, Lee ES, Oh KH, Kim KB. Synthesis and characterization of manganese dioxide spontaneously coated on carbon nanotubes. Carbon 2007;45:375–82. [70] Tran NE, Lambrakos SG, Lagowski JJ. Analysis of capacitance characteristics of C60, C70, and La@C82. J Mater Eng Perform 2009;18:95–101. [71] Doi Y, Takai A, Makino S, et al. Synthesis of mesoporous carbon using a fullerene-based precursor solution via nanocasting with SBA-15. Chem Lett 2010;39:777–9. [72] Hummers WS, Offeman RE. Preparation of graphitic oxide. J Am Chem Soc 1958;80:1339. [73] He H, Klinowski J, Forster M, Lerf A. A new structural model for graphite oxide. Chem Phys Lett 1998;287:53. [74] Cui H, Zhou O, Stoner BR. Deposition of aligned bamboo-like carbon nanotubes via microwave plasma enhanced chemical vapor deposition. J Appl Phys 2000;88:6072–4. [75] Available from: https://en.wikipedia.org/wiki/carbon_nanotube. [76] Smith BW, Marc M, Luzzi DE. Encapsulated C-60 in carbon nanotubes. Nature 1998;396:323–4. [77] Smith BW, Luzzi DE. Formation mechanism of fullerene peapods and coaxial tubes: a path to large scale synthesis. Chem Phys Lett 2000;321:169–74. [78] Su H, Goddard WA, Zhao Y. Dynamic friction force in a carbon peapod oscillator. Nanotechnology 2006;17:5691–5. [79] Wang M, Li CM. An oscillator in a carbon peapod controllable by an external electric field: a molecular dynamics study. Nanotechnology 2010;21:035704. [80] Stoner BR, Akshay SR, Brown B, Parker CB, Glass JT. Graphenated carbon nanotubes for enhanced electrochemical double layer capacitor performance. Appl Phys Lett 2011;18(99):183104. [81] Yu K, Lu G, Zheng B, Mao S, Chen J. Carbon nanotube with chemically bonded graphene leaves for electronic and optoelectronic applications. J Phys Chem Lett 2011;13(2):1556–62. [82] Parker CB, Raut AS, Brown B, Stoner BR, Glass JT. Three-dimensional arrays of graphenated carbon nanotubes. J. Mater. Res. 2012;7(27):1046–53. [83] Hsu H-C, Wang C-H, Nataraj SK, et al. Stand-up structure of graphene-like carbon nanowalls on CNT directly grown on polyacrylonitrile-based carbon fiber paper as supercapacitor. Diamond Rel Mater 2012;25:176–9. [84] Pham K-C, Chua DHC, McPhail DS, Wee ATS. The direct growth of graphene-carbon nanotube hybrids as catalyst support for high-performance PEM fuel cells. ECS Electrochem Lett 2014;3:F37–40.

Carbonaceous Materials

67

[85] Pham K-C, McPhail DS, Cecilia M, Wee ATS, Chua DHC. Graphene–carbon nanotube hybrids as robust catalyst supports in proton exchange membrane fuel cells. J Electrochem Soc 2016;16:F255–63. [86] Lu X, Chen Z. Curved Pi-Conjugation, aromaticity, and the related chemistry of small fullerenes (C60) and single-walled carbon nanotubes. Chem Rev 2005;105:3643–96. [87] Zhu Y, Ang TP, Carpenter K, et al. Substituted-carborane-appended water soluble single-wall carbon nanotubes: new approach to BNCT drug delivery. J Am Chem Soc 2005;127:9875–80. [88] Pirahna USV built using nano-enhanced carbon prepreg, ReinforcedPlastics.com. 19 February 2009. [89] John P. Amroy aims to become nano-leader, European Plastics News. 5 March 2010. [90] Niu C, Sichel EK, Hoch R, Moy D, Howard TC. High power electrochemical capacitators based on carbon nanotube electrodes. Appl Phys Lett 1997;70:1480–2. [91] An KH, Kim WS, Park YS, et al. Electrochemical properties of high-power supercapacitators using single-walled carbon nanotube electrodes. Adv Funct Mater 2001;11:387–92. [92] Wang G, Zhang L, Zhang J. A review of electrode materials for electrochemical supercapacitators. Chem Soc Rev 2012;41:797–828. [93] Sun G, Zhang X, Lin R, Yang J, Zhang H, Chen P. Hybride fibers made of molybdenum disulfide, reduced graphene oxide, and multi-walled carbon nanotubes for solid-state, flexible, asymmetric supercapacitors. Angew Chem Int Ed 2015;54:4651–6. [94] Ratha S, Marri SR, Behera JN, Rout CS. High-energy-density supercapacitators based on patronite/single-walled carbon nanotubes/reduced graphene oxide hybrides. Eur J Inorg Chem 2016;259–65. [95] Su F, Lv X, Miao M. High-performance two-ply yarn supercapacitators based on carbon nanotube yarns dotted with Co3O4 and NiO nanoparticles. Small 2015;11:854–61. [96] Ge D, Yang L, Fan L, et al. Foldable supercapacitators from triple networks of macroporous cellulose fibers, single-walled carbon nanotubes and polyaniline nanoribbons. Nano Energy 2015;11:568–78. [97] Yang Y, He L, Tang C, et al. Improvedconductivity and capacitance of interdigital carbon microelectrodes through integration with carbon nanotubes for microsupercapacitors. Nano Res. 2016;9:2510–9. [98] Rangom Y, Tang X, Nazar LF. Carbon nano-tube-based supercapacitators with excellent ac line filtering and rate capability via improved interfacial impedance. ACS Nano 2015;9:7248–55. [99] Takeuchi K, Hayashi T, Kim YA, Fujisawa K, Endo M. The state-of-the-art science and applications of carbon nanotubes. Nanosyst: Phys Chem Math 2014;5:15. [100] The Nobel Prize in Physics 2010, The Nobel Foundation. Retrieved 3 December 2013. [101] Graphene properties. Available from: www.graphene-battery.net; 2016 [accessed 27.05.16]. [102] Shinn E, Hubler A, Lyon D, et al. Nuclear energy conversion with stacks of graphene nano-capacitors. Complexity. 2012; doi:10.1002/cplx.21427. [103] Li Z, Chen L, Meng S, et al. Field and temperature dependence of intrinsic diamagnetism in graphene: theory and experiment. Phys. Rev. B 2015;91:094429. [104] Morozov SV, Novoselov K, Katsnelson M, et al. Giant intrinsic carrier mobilities in graphene and its bilayer. Phys Rev Lett 2008;100:016602. [105] Kusmartsev FV, Wu WM, Pierpoint MP, Yung KC. Application of graphene within optoelectronic devices and transistors; 2014. arXiv:1406.0809. [106] Sagade AA, Neumaier D, Schall D, et al. Highly air stable passivation of graphene based field effect devices. Nanoscale 2015;7:3558–64. [107] Graphene devices stand the test of time. Available from: https://spectrum.ieee.org/nanoclast/semiconductors/nanotechnology/graphene-devices-stand-the-test-of-time; 2016 [accessed 03.12.16]. [108] Researchers create superconducting graphene. Available from: https://www.ubyssey.ca/science/ubc-researchers-create-superconducting-graphene/; 2016 [accessed 22.09.16]. [109] Ali E, Bahareh Y. Initiating electropolymerization on graphene sheets in graphite oxide structure. J Polym Sci A: Polym Chem 2010;48:2204–13. [110] Venkatram N, Polavarapu L, Manga KK, et al. Transient photoconductivity and femtosecond nonlinear optical properties of a conjugated polymer–graphene oxide composite. Nanotechnology 2010;21:415203. [111] Nair RR, Wu HA, Jayaram PN, Grigorieva IV, Geim AK. Unimpeded permeation of water through helium-leak-tight graphene-based membranes. Science 2012;335:442–4. [112] Algin Oh B, Vangala VR, Chen CS, et al. Cross-coupling reaction between arylboronic acids and carboranyl iodides catalyzed by graphene oxide (GO)-supported Pd(0) recyclable nanoparticles for the synthesis of carboranylaryl ketones. Dalton Trans 2014;43:5014–20. [113] Global demand for graphene after commercial Production to be enormous, says report. Available from: Azonano.com; 2014 [acccessed 24.07.16]. [114] Sutter P. Epitaxial graphene: how silicon leaves the scene. Nature Mater 2009;8:171–2. [115] Samsung's graphene breakthrough could finally put the wonder material into real-world devices, ExtremeTech. 7 April 2014; Retrieved 13 April 2014. [116] Lee J-H, Lee EK, Joo W-J, et al. Wafer-scale growth of single-crystal monolayer graphene on reusable hydrogen-terminated germanium. Science 2014;344:286–9. [117] Tanesh B, Durcan CA, Nikhil J, et al. Synthesis of few-to-monolayer graphene on rutile titanium dioxide. Carbon 2013;55:168–75. [118] Eigler S, Enzelberger-Heim M, Grimm S, et al. Wet chemical synthesis of graphene. Adv Mater 2013;25:3583–7. [119] Available from: http://www.kurzweilai.net/a-new-method-of-producing-large-volumes-of-high-quality-graphene; 2016 [accessed 10.11.16]. [120] Paton KR. Scalable production of large quantities of defect-free few-layer graphene by shear exfoliation in liquids. Nature Mater 2014;13:624–30. [121] Jennifer M. Researchers develop graphene supercapacitor holding promise for portable electronics/UCLA Newsroom. Available from: Newsroom.ucla.edu; 2012. [122] Xia JL, Chen F, Li JH, Tao NJ. Measurement of the quantum capacitance of graphene. Nat. Nanotechnol 2009;4:505–9. [123] Ei-Kady MF, Shao Y, Kaner RB. Graphene for batteries, supercapacitators and beyond. Nature Rev Mater 2016; doi:10.1038/natrevmats.2016.33. [124] Chee WK, Lim HN, Zainal Z, Huang NM, Harrison I, Andou Y. Flexible graphene-based supercapacitators: a review. J Phys Chem C. 2016;120:4153–72. [125] Ke Q, Wang J. Graphene-based materials for supercapacitator electrodes – a review. J. Materiomics 2016;2:37–54. [126] Yang X, Cheng C, Wang Y, Qiu L, Li D. Liquid-mediated dense integration of graphene materials for compact capacitive energy storage. Science 2013;341:534–7. [127] Cai Y, Zhang G, Zhang YW. Layer-dependent band alignment and work function of few-layer phosphorene. Sci Rep 2014;4:6677. [128] Cai Y, Ke Q, Zhang G, et al. Giant phononic anisotropy and unusual anharmonicity of phosphorene: interlayer coupling and strain engineering. Adv Funct Mater 2015;25:2230–6. [129] Si Y, Samulski ET. Exfoliated graphene separated by platinum nanoparticles. Chem Mater 2008;20:6792–7. [130] Qiu L, Yang X, Gou X, et al. Dispersing carbon nanotubes with graphene oxide in water and synergistic effects between graphene derivatives. Chem Eur J 2010;16:10653–8. [131] Li S, Luo Y, Lv W, et al. Vertically aligned carbon nanotubes grown on graphene paper as electrodes in lithium-ion batteries and dye-sensitized solar cells. Adv Energy Mater 2011;1:486–90. [132] Wang GK, Sun X, Lu FY, et al. Flexible pillared graphene-paper electrodes for high-performance electrochemical supercapacitators. Small 2012;8:452–9. [133] Wang H-W, Hu Z-A, Chang Y-Q, et al. Preparation of reduced graphene oxide/cobaltoxide compositesand their enhanced capacitive behaviors by homogeneous incorporation of graphene oxide sheets in cobalt oxide matrix. Mater Chem Phys 2011;130:672–9. [134] Li M, Tang Z, Leng M, Xue J. Flexible solid-state suercapacitor based on graphene-based hybrid films. Adv Funct Mater 2014;24:7495–502. [135] Choi BG, Yang M, Hong WH, Choi JW, Huh YS. 3D macroporous graphene frameworks for supercapacitors with high energy and power densities. ACS Nano 2012;6:4020–8. [136] Yang M, Lee KG, Lee SJ, et al. Three-dimensional expanded graphene-metal oxide film via solid-state microwave irradiation for aqueous asymmetric supercapacitors. ACS Appl Mater Interfaces 2015;7:22364–71.

68

Carbonaceous Materials

[137] Yu J, Wu J, Wang H, et al. Metallic fabrics as the current collector for high-performance graphene-based flexible solid-state supercapacitor. ACS Appl Mater Interfaces 2016;8:4724–9. [138] Meng Y, Wang K, Zhang Y, Wei Z. Hierarchical porous graphene/polyaniline composite film with superior rate performance for flexible supercapacitors. Adv Mater 2013;25:6985–90. [139] Yan J, Wei T, Shao B, et al. Preparation of a graphene nanosheet/polyaniline composite with high specific capacitance. Carbon 2010;48:487–93. [140] Wang S, Ma L, Gan M, et al. Free-standing 3D graphene/polyaniline composite film electrodes for high-performance supercapacitors. J Power Sources 2015;299:347–55. [141] Wang DW, Li F, Zhao JP, et al. Fabrication of graphene/polyaniline composite paper via in situ anodic electropolymerization for high-performance flexible electrode. ACS Nano 2009;3:1745–52. [142] Wu M, Li Y, Yao B, et al. A high-performance current collector-free flexible in-plane micro-supercapacitor based on a highly conductive reduced graphene oxide film. J. Mater Chem A 2016;4. 16313-16218. [143] Chen CM, Zhang Q, Huang CH, et al. Macroporous ‘bubble’ graphene film via template-directed ordered-assembly for high rate supercapacitors. Chem Commun 2012;48:7149–51. [144] Lehtimaki S, Suominen M, Damlin P, et al. Preparation of supercapacitors on flexible substrates with electrodeposited PEDOT/graphene composites. ACS Appl Mater Interfaces 2015;7:22137–47. [145] Cong HP, Ren XC, Wang P, Yu SH. Macroscopic multifunctional graphene-based hydrogels and aerogels by a metal ion induced selfassembly process. ACS Nano 2012;6:2693–703. [146] Xu Y, Sheng K, Li C, Shi G. Self-assembled graphene hydrogel via a one-step hydrothermal process. ACS Nano 2010;4:4324–30. [147] Liu L, Yu Y, Yan C, Li K, Zheng Z. Wearable energy-dense and powerdense supercapacitor yarns enabled by scalable graphene-metallic textile composite electrodes. Nature Commun 2015;6:7260. [148] An N, An Y, Hu Z, et al. Graphene hydrogels noncovalently functionalized with alizarin: an ideal electrode material for symmetric supercapacitors. J Mater Chem A 2015;3:22239–46. [149] Chen L, Wu J, Zhang A, et al. One-step synthesis of polyhydroquinone-graphene hydrogel composites for high performance supercapacitors. J Mater Chem A 2015;3:16033–9. [150] Xu Y, Lin Z, Huang X, et al. Functionalized graphene hydrogel-based high-performance supercapacitors. Adv Mater 2013;25:5779–84. [151] Wang B, Liu J, Mirri F, et al. High-performance graphene-based supercapacitors made by a scalable blade-coating approach. Nanotechnology 2016;27:165402. [152] Kovalska E, Kocabas C. Organic electrolytes for graphene-based supercapacitor: liquid, gel, or solid. Mater. Today. 2016;7:155–60. [153] Xie Q, Zhou S, Zheng A, et al. Sandwich-like nitrogen-enriched porous carbon/graphene composites as electrodes for aqueous symmetric supercapacitors with high energy density. Electrochim Acta 2016;189:22–31. [154] Chaffarzadeh K. Graphene and 2D materials: markets, Technologies and Opportunities, 2015-2025 (IDtECHeX); 2015. [155] Weinstein I, Dash R. Supercapacitor carbons. Mater Today 2013;10:356–7. [156] Jiang H, Lee PS, Li C. 3D carbon based nanostructures for advanced supercapacitors. Energy Environ Sci 2013;6:41–53. [157] Liu R, Duay J, Lee SB. Heterogeneous nanostructured electrodematerials for electrochemical energy storage. Chem Commun 2011;47:1384–404. [158] Xu X, Liu Y, Wang M, et al. Hierarchical hybrids with microporous carbon spheres decorated three-dimentional graphene frameworks for capacitive applications in supercapacitor and deionization. Electochim. Acta 2016;193:88–95. [159] Yan YF, Cheng QL, Wang GC, Li CZ. Growth of polyaniline nanowhiskers on mesoporous carbon for supercapacitor application. J Power Sources 2011;196:7835–40. [160] Yan YF, Cheng QL, Pavlinek V, Saha P, Li CZ. Fabrication of polyaniline/mesoporous carbon/MnO2 ternary nanocomposites and their enhanced electrochemical performance for supercapacitors. Electrochim Acta 2012;71:27–32. [161] Mo R, Zhao Y, Wu M, et al. Activated carbon from nitrogen rich watermelon rind for high-performance supercapacitors. RSC Adv 2016;6:59333–42. [162] Gao F, Qu J, Zhao Z, Wang Z, Qiu J. Nitrogen-doped activated carbon derived from prawn shells for high-performance supercapacitors. Electrochim Acta 2016;190:1134–41. [163] Marcuzzo JS, Cuna A, Tancredi N, Mendez E, Bernardi HH. Microporus activated carbon fiber felt from Brazilian textile PAN fiber: preparation, characterization and application as super capacitor electrode. Rev Bras Appl Vac, Campinas 2016;35:58–63. [164] Huang P, Lethien C, Pinaud S, et al. On-chip and freestanding elastic carbon films for micro-supercapacitors. Science 2016;351. 691–685. [165] Kumar VB, Borenstein A, Markovsky B, et al. Activated carbon modified with carbon nanodots as novel electrode material for supercapacitors. J Phys Chem C 2016;120:13406–13. [166] Le Comte A, Brousse T, Belanger D. New generation of hybrid carbon/Ni(OH)2 electrochemical capacitor using functionalized carbon electrode. J Power Sources 2016;326:702–10. [167] Godillot G, Tabema P-L, Daffos B, et al. High power density aqueous hybride supercapacitor combing activated carbon and highly conductive spinel cobalt oxide. J Power Sources 2016;331:277–84. [168] Wang Q, Zou Y, Xiang C, et al. Prussian-blue-doped super-activated carbon as a high performance supercapacitor electrode material. Int J Electrochem Sci 2016;11:5679–90. [169] Shiraishi S, Kibe M, Yokoyama T, et al. Electric double layer capacitance of multi-walled carbon nanotubes and B-doping effect. Appl. Phys. A: Mater. Sci. Process 2006;82:585–91. [170] Wu XX, Radovic LR. Ab initio molecular orbital study on the electronic structures and reactivity of boron-substituted carbon. J Phys Chem A 2004;108:9180–7. [171] Lowell CE. Solid solution of boron in graphite. J Am Ceram Soc 1967;50:142–4. [172] Morita T, Takami N. Characterization of oxidized boron-doped carbon fiber anodes for Li-ion batteries by analysis of heat of immersion. Electrochim Acta 2004;49:2591–9. [173] Vishwakarma PN, Subramanyam SV. Hopping conduction in boron doped amorphous carbon films. J Appl Phys 2006;100:113702. [174] Radovic LR, Karra M, Skokova K, Thrower PA. The role of substitutional boron in carbon oxidation. Carbon 1998;36:1841–54. [175] Zhong DH, Sano H, Uchiyama Y, Kobayashi K. Effect of low-level boron doping on oxidation behavior of polyimide-derived carbon films. Carbon 2000;38:1199–206. [176] Wang D–W, Li F, Chen Z–G, Lu GQ, Cheng H–M. Synthesis and electrochemical property of boron-doped mesoporous carbon in supercapacitor. Chem Mater 2008;20:7195–200. [177] Guo HL, Gao QM. Boron and nitrogen co-doped porous carbon and its enhanced properties as supercapacitor. J Power Sources 2009;186:551–6. [178] Wu ZS, Winter A, Chen L, Sun Y, Turchanin A, et al. Three-dimensional nitrogen and boron co-doped graphene for high-performance all-solid-state supercapacitors. Adv Mater 2012;24:5130–5. [179] Jeong HM, Lee JW, Shin WH, et al. Nitrogen-doped graphene for high-performance ultracapacitors and the importance of nitrogen-doped sites at basal planes. Nano Lett 2011;11:2472–7. [180] Wu ZS, Ren WC, Xu L, Li F, Cheng HM. Doped graphene sheets as anode materials with superhigh rate and large capacity for lithium ion batteries. ACS Nano 2011;5:5463–71. [181] Werheit H. Boron-rich solids: a chance for high-efficiency high-temperature thermoelectric energy conversion. Mater Sci Eng B 1995;29:228–32. [182] Werheit H. On the electronic transport properties of boron carbide. In: Freer R, editor. The physics and chemistry of carbides, nitrides and borides. Dordrecht: Kluwer; 1990, p. 677.

Carbonaceous Materials

69

[183] Werheit H. On the electronic properties of icosahedral boron-rich solids. In: Gogotsi YG, Andrievski RA, editors. material science of carbides, nitrides and borides, NATO Science Series ;8. ; 1999. p. 65–86. [184] Kuhlmann U, Werheit H. On the microstructure of boron carbide. Solid State Commun 1992;83:849–52. [185] Kuhlmann U, Werheit H, Sehwetz KA. Distribution of carbon atoms on the boron cabide structure elements. J. Alloys Compounds 1992;189:249–58. [186] Bairamashvili IA, Kekelidze LI, Golikova OA, Orlov VM. J. Less Common Metals 1979;67:461–4. [187] Werheit H H, Groot K. Metal-insulator transition in boron carbide. Phys. Status Solid (b) 1980;97:229–38. [188] Li J, Goto T, Hiral T. Microstructure and thermoelectric properties of B4C-SiBn-Si c omposites prepared by arc melting. J Ceram Soc Japan 1998;106:194–7. [189] Imai Y, Mukaida M, Ueda M, Watanabe A. Screening of the possible boron-based n-type thermoelectric conversion materials on the basis of the calculated densities of states of metal borides and doped b-boron. Intermetallics 2001;9:721–34. [190] Takeda M, Fukuda T, Domingo F, Miura T. Thermoelectric properties of some metal borides. J Solid State Chem 2004;177:471–5. [191] Nayak PK, Garcia-Belmonte G, Kahn A, Bisquert J, Cahen D. Photovoltaic efficiency limits and material disorder. Energy Environ Sci 2012;5:6022–39. [192] Hod I, Zaban A. Materials and interfaces in quantum dot sensitized solar cells: challenges, advances and prospects. Langmuir 2014;30:7264–73. [193] Tian XM, Rusop M, Hayashi Y, et al. A photovoltaic cell from p-type boron-doped amorphous carbon film. Solar Energy Materials & Solar Cells 2003;77:105–12. [194] Ma ZQ, Liu BX. Boron-doped diamond-like amorphous carbon as photovoltaic films in solar cell. Solar Energy Mater Solar Cells 2001;69:339–44. [195] Wang JW, Liu Y, Zhang XW, Mi ZT, Wang L. Facile preparation of hydrocarbon fuel-soluble nano-catalyst and its novel application in catalytic combustion of JP-10. Catal Commun 2009;10:1518–22. [196] Holcomb RR, Holcomb AR. Composite inorganic and organic sols and colloids as combustion improvers for liquid fuels. Orlando, FL: Demeter Systems LLC; 2006. p. 36. [197] Valentine JM, Sprague BN, Peter-Hoblyn JD. Fuel-soluble platinum-cerium-iron catalysts for combustion of diesel fuel and jet fuel, Patent Application No. US 200538371, Jan 19; 2005. p. 12. [198] Slutsky VG, Tsyganov SA, Severin ES, Polenov LA. Synthesis of small-scale boron-rich nano-size particles. Propellants Explos Pyrotech 2005;30:303–9. [199] Risha GA, Boyer E, Evans B, Kuo KK, Malek R. Characterization of nano-sized particles for propulsion applications. Mater Res Soc Symp Proc 2003;800:243. [200] Kuo KK, Risha GA, Evans BJ, Boyer E. Potential usage of energetic nano-sized powders for combustion and rocket propulsion. Mater Res Soc Symp Proc 2003;800 (2003):3–14. [201] Young G, Sullivan K, Zachariah MR, Yu K. Combustion characteristics of boron nanoparticles. Combust Flame 2009;156:322–33. [202] Yeh CL, Kuo KK. Ignition and combustion of boron particles. Prog Energy Combust Sci 1996;22:511–41. [203] Dubois C, Brousseau P, Roy C, Lafleur P. In: Proceedings of the AIChE annual meeting; Austin, TX, November 7–12; 2004. p. 281E/1 281E/9. [204] Anderson JL, Ding R, Ellern A, Armstrong DW. Structure and properties of high stability geminal dicationic ionic liquids. J Am Chem Soc 2005;127:593–604. [205] Pickering AL, Mitterbauer C, Browning ND, Kauzlarich SM, Power PP. Room temperature synthesis of surface-functionalised boron nanoparticles. Chem Commun 2007;6:580–2. [206] Devener BV, Perez JPL, Anderson SL. Air-stable, unoxidized, hydrocarbon-dispersible boron nanoparticles. J Mater Res 2009;24:3462–4. [207] Perez JPL, McMahon BW, Anderson SL. Functionalization and passivation of boron nanoparticles with a hypergolic ionic liquid. J Propul Power 2013;29:489–95. [208] Perez JPL, McMahon BW, Schneider S, et al. Anderson, Exploring the structure of nitrogen-rich ionic liquids and their binding to the surface of oxide-free boron nanoparticles. J Phys Chem C 2013;117:5693–707. [209] Perez JPL, Yu J, Sheppard AJ, et al. Binding of alkenes and ionic liquids to B–H-functionalized boron nanoparticles: creation of particles with controlled dispersibility and minimal surface oxidation. ACS Appl Mater Interfaces 2015;7:9991–10003. [210] Fareghi-Alamdari R, Ghorbani-Zamani F, Shekarriz M. Surface passivation of bare boron nanoparticles using new dicyanamide-based dicationic ionic liquid. Energy Fuels 2016;30:551–9. [211] Devener BV, Perez JPL, Jankovich J, Anderson SL. Oxide-free, catalyst-coated, fuel-soluble, air-stable boron nanopowder as combined combustion catalyst and high energy density fuel. Energy Fuels 2009;23:6111–20. [212] Andrew J, Jelley N. Energy science. ISBN 978-0-19-928112-1 Oxford: Oxford University Press; 2007. [213] Available from: https://en.wikipedia.org/wiki/Control_rod; 2016 [accessed 20.11.16]. [214] Thevenot F. Boron carbide – a comprehensive review. J Eur Ceram Soc 1990;6:205–25. [215] Matkovich VI. Boron and refractory borides. Heidelberg/New York: Singler-Verlag Berlin; 1977. [216] Bose DK, Nair KU, Gupta CK. Production of high purity boron carbide. High Temp Mater. Proc. 1986;7:133–40. [217] Subramanian C, Suri AK. Development of boron based neutron absorber materials. Met Mater Process 2004;16:39–52. [218] Broje V, Keller AA. Improved mechanical oil spill recovery using an optimized geometry for the skimmer surface. Environ Sci Technol 2006;40:7914–8. [219] Zahed MA, Aziz HA, Isa MH, Mohajeri L, Mohajeri S. Optimal conditions for bioremediation of oily seawater. Bioresour Technol 2010;101:9455–60. [220] Gupta S, Tai N-H. Carbon materials as oil sorbents: a review on the synthesis and performance. J Mater Chem A 2016;4:1550–65. [221] Bayat A, Aghamiri SF, Moheb A, Vakili-Nezhaad GR. Oil spill cleanup from sea water by sorbent materials. Chem Eng Technol 2005;28:1525–8. [222] Xie J, Meng W, Wu D, Zhang Z, Kong H. Removal of organic pollutants by surfactant modified zeolite: comparison between ionizable phenolic compounds and nonionizable organic compounds. J Hazard Mater 2012;231-232:57–63. [223] Xue Z, Cao Y, Liu N, Feng L, Jiang L. Special wettable materials for oil/water separation. J Mater Chem A 2014;2:2445–60. [224] Hristea G, Budrugeac P. Characterization of exfoliated graphite for heavy oil sorption. J Therm Anal Calorim 2008;91:817–23. [225] Toyoda M, Inagaki M. Heavy oil sorption using exfoliated graphite: new application of exfoliated graphite to protect heavy oil pollution. Carbon 2000;38:199–210. [226] Wang L, Fu X, Chang E, et al. Preparation and its adsorptive property of modified expanded graphite nanomaterials. J Chem 2014;2014:678151. [227] Cong H-P, Chen J-F, Yu S-H. Graphene-based macroscopic assemblies and architectures: an emerging material system. Chem Soc Rev 2014;43:7295–325. [228] Available from: https://www.theguardian.com/science/2013/nov/26/graphene-molecule-potential-wonder-material; 2017 [accessed 20.05.17]. [229] Bi H, Xie X, Yin K, et al. Spongy graphene as a highly efficient and recyclable sorbent for oils and organic solvents. Adv Funct Mater 2012;22:4421–5. [230] Iqbal M, Abdala A. Oil spill cleanup using grapheme. Environ Sci Pollut Res 2013;20:3271–9. [231] Available from: https://www.extremetech.com/extreme/153063-graphene-aerogel-is-seven-times-lighter-than-air-can-balance-on-a-blade-of-grass/3. [232] Li J, Li J, Meng H, et al. Ultra-light, compressible and fire-resistant graphene aerogel as a highly efficient and recyclable absorbent for organic liquids. J Mater Chem A 2014;2:2934–41. [233] Wu Z-Y, Li C, Liang H-W, Chen J-F, Yu S-H. Ultralight, flexible, and fire-resistant carbon nanofiber aerogels from bacterial cellulose. Angew Chem Int Ed 2013;52:2925–9. [234] Zhao Y, Hu C, Hu Y, et al. A versatile, ultralight, nitrogen-doped graphene framework. Angew Chem, Int Ed 2012;51:11371–5. [235] Ge B, Zhang Z, Zhu X, et al. A graphene coated cotton for oil/water separation. Composites Sci Technol. 2014;102:100–5. [236] Nguyen DD, Tai N-H, Lee S-B, Kuo W-S. Superhydrophobic and superoleophilic properties of graphene-based sponges fabricated using a facile dip coating method. Energy Environ Sci 2012;5:7908–12. [237] Liu Y, Ma J, Wu T, et al. Cost-effective reduced graphene oxide-coated polyurethane sponge as a highly efficient and reusable oil-absorbent. ACS Appl Mater Interfaces 2013;5:10018–26. [238] Liu H-D, Liu Z-Y, Yang M-B, He Q. Surperhydrophobic polyurethane foam modified by graphene oxide. J Appl Polym Sci 2013;130:3530–6. [239] Maphutha S, Moothi K, Meyyappan M, Iyuke SE. A carbon nanotube-infused polysulfone membrane with polyvinyl alcohol layer for treating oil-containing waste water. Sci Rep 2013;3:1509.

70 [240] [241] [242] [243] [244] [245] [246] [247] [248] [249] [250] [251] [252] [253] [254] [255] [256] [257] [258] [259] [260] [261] [262] [263] [264] [265] [266] [267] [268] [269] [270] [271] [272] [273] [274] [275] [276] [277] [278] [279] [280] [281] [282] [283] [284] [285] [286] [287] [288] [289]

Carbonaceous Materials Sun H, La P, Zhu Z, et al. Hydrophobic carbon nanotubes for removal of oils and organics from water. J Mater Sci 2014;49:6855–61. Gui X, Wei J, Wang K, et al. Carbon nanotube sponges. Adv Mater 2010;22:617–21. Hashim DP, Narayanan NT, Romo-Herrera JM, et al. Covalently bonded three-dimensional carbon nanotube solids via boron induced nanojunctions. Sci Rep 2012;2:363. Wang H, Wang E, Liu Z, et al. A novel carbon nanotubes reinforced superhydrophobic and superoleophilic polyurethane sponge for selective oil–water separation through a chemical fabrication. J Mater Chem A 2015;3:266–73. Available from: https://en.wikipedia.org/wiki/Aerogel; 2017 [accessed 20.09.17]. Li Y-Q, Samad YA, Polychronopoulou K, Alhassan SM, Liao K. Carbon aerogel from winter melon for highly efficient and recyclable oils and organic solvents absorption. ACS Sustainable Chem Eng 2014;2:1492–7. Huang S, Shi J. Sulfated mesoporous niobium oxide catalyzed 5-hydroxymethylfurfural formation from sugars. Ind Eng Chem Res 2014;53:4888–93. Bi H, Huang X, Wu X, et al. Carbon microbelt aerogel prepared by waste paper: an efficient and recyclable sorbent for oils and organic solvents. Small 2014;10:3544–50. Wang H, Gong Y, Wang Y. Cellulose-based hydrophobic carbon aerogels as versatile and superior adsorbents for sewage treatment. RSC Adv 2014;4:45753–9. Wu Z-Y, Li C, Liang H-W, et al. Carbon nanofiber aerogels for emergent cleanup of oil spillage and chemical leakage under harsh conditions. Sci Rep 2014;4:4079. Dong X, Chen J, Ma Y, et al. Superhydrophobic and superoleophilic hybrid foam of graphene and carbon nanotube for selective removal of oils or organic solvents from the surface of water. Chem Commun 2012;48:10660–2. Sun H, Xu Z, Gao C. Multifunctional, ultra-flyweight, synergistically assembled carbon aerogels. Adv Mater 2013;25:2554–60. Titirici M-M, White RJ, Brun N, et al. Sustanable carbon materials. Chem Soc Rev 2015;44:250–90. Sevilla M, Mokaya R. Energy storage applications of activated carbons: supercapacitors and hydrogen storage. Energy Environ Sci 2014;7:1250–80. Aijaz A, Akita T, Yang H, Xu Q. From ionic-liquid@ metal–organic framework composites to heteroatom-decorated large-surface area carbons: superior CO2 and H2 uptake. Chem Commun 2015;50:6498–501. Zhu ZH, Lu GQ, Hatori H. New insights into the interaction of hydrogen atoms with boron-substituted carbon. J Phys Chem B 2006;110:1249–55. Wang L, Yang RT. New sorbents for hydrogen storage by hydrogen spillover – a review. Energy Environ Sci 2008;1:268–79. Available from: http://www.energy.gov/sites/prod/files/2015/01/f19/fcto_myrdd_table_onboard_H2_storage_systems_doe_targets_ldv.pdf. Tozzini V, Pellegrini V. Prospects for hydrogen storage in graphene. Phys Chem Chem Phys 2013;15:80. Available from: http://pubs.rsc.org/en/Content/ArticleLanding/ 2013/CP/C2CP42538F#!divAbstract. Goler S, Coletti C, Tozzini V, et al. The influence of graphene curvature on hydrogen adsorption: towards hydrogen storage devices. J Phys Chem C 2013;117:11506–13. Mashoff T, Takamura M, Tanabe S, Hibinio H, Beltram F, Heun S. Hydrogen storage with titanium-functionalized graphene. Appl Phys Lett 2013;103:013903. Tozzini V, Pellegrini V. Reversible hydrogen storage by controlled buckling of graphene layers. J Phys Chem C 2011;115:25523–8. Cho ES, Ruminski AM, Aloni S, Liu Y-S, Guo J, Urban JJ. Graphene oxide/metal nanocrystal multilaminates as the atomic limit for safe and selective hydrogen storage. Nature Commun 2016;7:10804. Ariharan A, Viswanathan B, Nandhakumar V. Hydrogen storage on boron substituted carbon materials. Int J Hydrogen Energy 2016;41:3527–36. Shcherban N, Filonenko S, Yaremov P, Dyadyun V, Bezverkhyy I. Boron-doped nanoporous carbons as promising materials for supercapacitors and hydrogen storage. J Mater Sci 2017;52:1523–33. Lin I-H, Tong Y-J, Hsieh H-J, Huang H-W, Chen H-T. Hydrogen adsorption and storage in boron-substituted and nitrogen-substituted nano-carbon materials decorated with alkaline earth metals. Int J Energy Res 2016;40:230–40. Li D, Ouyang Y, Sun Y, Chen L. Hydrogen storage of beryllium adsorbed on graphene doping with boron: first-principles calculations. Solid State Commun 2012;152:422–5. Bhatia SK, Myers AL. Optimum conditions for adsorptive storage. Langmuir 2006;22:1688–700. Sevilla M, Fuertes AB, Mokaya R. Preparation and hydrogenation storage capacity of highly porous activated carbon materials derived from polythiophene. Int J Hydrogen Energy 2011;36:15658–63. Sethia G, Sayati A. Activated carbon with optimum pore size distribution for hydrogen storage. Carbon 2016;99:289–94. Psofogiannakis GM, Steriotis TA, Bourlinos AB, et al. Enhanced hydrogen storage by spillover on metal-doped carbon foam: an experimental and computational study. Nanoscale 2011;3:933–6. Conner Jr WC, Falconer JL. Spillover in heterogeneous catalysts. Chem Rev 1995;95:759–88. Shiraz HG, Shiraz MG. Palladium nanoparticle and decorated carbon nanotube for electrochemical hydrogen storage. Int J Hydrogen Energy 2017;42:11528–33. Wei L, Mao Y. Enhanced hydrogen storage performance of reduced graphene oxide hybrids with nickel or its metallic mixtures based on spillover mechanism. Int J Hydrogen Energy 2016;41:11692–9. Parambhath VB, Nagar R, Sethupathi K, Ramaprabhu S. Investigation of spillover mechanism in palladium decorated hydrogen exfoliated functionalize graphene. J Phys Chem C 2011;115:15679–85. Pyle DS, Gray EM, Webb CJ. Hydrogen storage in carbon nanostructures via spillover. Int J Hydrogen Energy 2016;41:19098–113. Wang P, Kang X–D. Hydrogen-rich boron-containing materials for hydrogen storage. Dalton Trans 2008;40:5400–13. Orimo S, Nakamori Y, Eliseo JR, Zuttel A, Jensen CM. Complex hydrides for hydrogen storage. Chem Rev 2007;107:4111–32. Schlapbach L, Zuttel A. Hydrogen-storage materials for mobile applications. Nature 2001;414:353–8. Schlesinger HI, Brown HC, Finholt AE, et al. Sodium borohydride, its hydrolysis and its use as a reducing agent and in the generation of hydrogen. J Am Chem Soc 1953;75:215–9. Amendola SC, Sharp-Goldman SL, Saleem MJ, et al. An ultrasafe hydrogen generator: aqueous, alkaline borohydride solutions and Ru catalyst. J Power Sources 2000;85:186–9. Muir SM, Yao XD. Progress in sodium borohydride as a hydrogen storage material: development of hydrolysis catalysts and reaction systems. Int J Hydrogen Energy 2011;36:5983–97. U. S. Department of Energy Hydrogen Program. Go/no-go recommendation for sodium borohydride for on-board vehicular hydrogen storage. Available from: http://www1. eere.energy.gov/hydrogenandfuelcells/pdfs/42220.pdf; 2007 NREL/MP-150–42220. Zhang Q, Smith GM, Wu Y. Catalytic hyrolysis of sodium borohydride in an integrated reactor for hydrogen generation. Int J Hydrogen Energy 2007;32:4731–5. Amendola SC, Sharp-Goldman SL, Janjua MS, et al. A safe, portable, hydrogen gas generator using aqueous borohydride solution and Ru catalyst. Int J Hydrogen Energy 2000;25:969–75. Guella G, Zanchetta C, Patton B, Miotello A. New insights on the mechanism of palladium-catalyzed hydrolysis of sodium borohydride from 11B NMR measurements. J Phys Chem B 2006;110:17024–33. Patel N, Patton B, Zanchetta C, et al. Pd–C powder and thin film catalysts for hydrogen production by hydrolysis of sodium borohydride. Int J Hydrogen Energy 2008;33:287–92. Bai Y, Wu C, Wu F, Yi B. Carbon-supported platinum catalysts for on-site hydrogen generation from NaBH4 solution. Mater Lett 2006;60:2236–9. Kojima Y, Suzuki K, Fukumoto K, et al. Hydrogen generation using sodium borohydride solution and metal catalyst coated on metal oxide. Int J Hydrogen Energy 2002;27:1029–34. Ingersoll JC, Mani N, Thenmozhiyal JC, Muthaiah A. Catalytic hydrolysis of sodium borohydride by a novel nickel–cobalt–boride catalyst. J. Power Sources 2007;173:450–7.

Carbonaceous Materials

71

[290] Dong H, Yang H, Ai X, Cha C. Hydrogen production from catalytic hydrolysis of sodium borohydride solution using nickel boride catalyst. Int J Hydrogen Energy 2003;28:1095–100. [291] Moussa G, Moury R, Demirci UB, S¸ener T, Miele P. Boron-based hydrides for chemical hydrogen storage. Int J Energy Res 2013;37:825–42. [292] Li H, Yang Q, Chen X, Shore SG. Ammonia borane, past as prolog. J Organomet Chem 2014;751:60–6. [293] Stephens FH, Pons V, Baker RT. Ammonia-borane: the hydrogen source par excellence? Dalton Trans 2007;25:2613–26. [294] Shrestha RP, Diyabalanage HVK, Semelsberger TA, Ott KC, Burrell AK. Catalytic dehydrogenation of ammonia borane in non-aqueous medium. Int J Hydrogen Energy 2009;34:2616–21. [295] Yoon CW, Carroll PJ, Sneddon LG. Ammonia triborane: a new synthesis, structural determinations, and hydrolytic hydrogen-release properties. J Am Chem Soc 2009;131:855–64. [296] Hu MG, Geanangel RA, Wendlandt WW. The thermal decomposition of ammonia borane. Thermochim Acta 1978;23:249–55. [297] Summerscales OT, Gordon JC. Regeneration of ammonia borane from spent fuel. Dalton Trans 2013;42:10075–84. [298] Wang P. Solid-state thermolysis of ammonia borane and related materials for high-capacity hydrogen storage. Dalton Trans 2012;41:4296–302. [299] Shaw WJ, Linehan JC, Szymczak N, Heldebrant NJ, Yonker C, et al. In situ multinuclear NMR spectroscopic studies of the thermal decomposition of ammonia borane in solution. Angew Chem Int Ed 2008;47:7493–6. [300] Kostka JF, Schellenberg R, Baitalow F, Smolinka T, Mertens F. Concentration-dependent dehydrogenation of ammonia-borane/triglyme mixtures. Eur J Inorg Chem 2012;2012:49–54. [301] Glüer A, Förster M, Celinski VR, der Günne JS, Holthausen MC, et al. Highly active iron catalyst for ammonia borane dehydrocoupling at room temperature. ACS Catal 2015;5:7214–7. [302] Manon P, Philippe M, Umit BD. Mechanistic insights of metal acetylacetonate-aided dehydrocoupling of liquid-state ammonia borane NH3BH3. Adv Energy Res 2016;4:177–87. [303] Duman S, Özkar S. Hydrogen generation from the dehydrogenation of ammonia-borane in the presence of ruthenium (III) acetylacetonate forming a homogeneous catalyst. Int J Hydrogen Energy 2013;38:180–7. [304] Kalviri HA, Gärtner F, Ye G, Korobkov I, Baker RT. Probing the second dehydrogenation step in ammonia-borane dehydrocoupling: characterization and reactivity of the key intermediate, B-(cyclotriborazanyl)amine-borane. Chem Sci 2015;6:618–24. [305] Toche F, Chiriac R, Demirci UB, Miele P. Ammonia borane thermolytic decomposition in the presence of metal (II) chlorides. Int J Hydrogen Energy 2012;37:6749–55. [306] Lian G, Zhang X, Zhang SJ, Liu D, Cui DL, et al. Controlled fabrication of ultrathin-shell BN hollow spheres with excellent performance in hydrogen storage and wastewater treatment. Energy Environ Sci 2012;5:7072–80. [307] Lei WW, Zhang H, Wu Y, Zhang B, Liu D, et al. Oxygen-doped boron nitride nanosheets with excellent performance in hydrogen storage. Nano Energy 2014;6:219–24. [308] Jhi S-H, Kwon Y-K. Hydrogen adsorption on boron nitride nanotubes: a path to room-temperarue hydrogen storage. Phys Rev B 2004;69:245407. [309] Ma RZ, Bando Y, Sato T, Golberg D, Zhu HW, et al. Synthesis of boron nitride nanofibers and measurement of their hydrogen uptake catacity. Appl Phys Lett 2002;81:5225. [310] Ma RZ, Bando Y, Zhu HW, Sato T, Xu CL, et al. Hydrogen uptake in boron nitride nanotubes at room temperature. J Am Chem Soc 2002;124:7672–3. [311] Tang CC, Bando Y, Ding XX, Qi SR, Golberg D. Catalyzed collapse and enhanced hydrogen storage of BN nanotubes. J Am Chem Soc 2002;124:14550–1. [312] Chen X, Gao XP, Zhang H, Zhou Z, Hu WK, et al. Preparation and electrochemical hydrogen storage of boron nitride nanotubes. J Phys Chem B 2005;109:11525–9. [313] Zuttel A. Materials for hydrogen storage. Mater. Today 2003;6:24–33. [314] Jhi S–H. Activated boron nitride nanotubes: a potential material for room-temperature hydrogen storage. Phys Rev B 2006;74:155424. [315] Li M, Li YF, Zhou Z, Shen PW, Chen ZF. Ca-coated boron fullerenes and nanotubes as superior hydrogen storage materials. Nano Lett 2009;9:1944–8. [316] Yang J, Sudik A, Wolverton C, Siegel DJ. High capacity hydrogen storage materials: attributes for automotive applications and techniques for materials discovery. Chem Soc Rev 2010;39:656–75. [317] Liu C, Li F, Ma LP, Cheng HM. Advanced materials for energy storage. Adv Mater 2010;22:E28–62.

Further Reading Devener BV, Perez JPL, Jankovich J, Anderson SL. Oxide-free, catalyst-coated, fuel-soluble, air-stable boron nanopowder as combined combustion catalyst and high energy density fuel. Energy Fuels 2009;23:6111–20. Dreyer DR, Park S, Bielawski CW, Ruoff RS. The chemistry of graphene oxide. Chem Soc Rev 2010;39:228–40. Gupta S, Tai N-H. Carbon materials as oil sorbents: a review on the synthesis and performance. J Mater Chem A 2016;4:1550–65. Ke Q, Wang J. Graphene-based materials for supercapacitator electrodes – a review. J Materiomics 2016;2:37–54. Liang D, Liu J, Xiao J, et al. Energy release properties of amorphous boron and boron-based propellant primary combustion products. Acta Astronautica 2015;112:182–91. Liu BH, Li ZP. A review: hydrogen generation from borohydride hydrolysis reaction. J Power Sources 2009;187:527–34. Miller JR, Burke AF. Electrochemical capacitors: challenges and opportunities for real-world applications. In: Spring 2008. The Electrochemical Society Interface; 2008. p. 53–57. Muir Sean S, Yao Xiangdong. Progress in sodium borohydride as a hydrogen storage material: development of hydrolysis catalysts and reaction systems. Int J Hydrogen Energy 2011;36:5983–97. Narayanan KL, Goetzgberger O, Khan A, Kojima N, Yamaguchi M. Boron ion implantation effect in C60 films. Solar Energy Mater Sol Cells 2001;65:29–35. Pandolfo AG, Hollenkamp AF. Carbon properties and their role in supercapacitors. J Power Sources 2006;157:11–27. Retnamma R, Novais AQ, Rangel CM. Kinetics of hydrolysis of sodium borohydride for hydrogen production in fuel cell applications: a review. Int J Hydrogen Energy 2011;36:9772–90. Santos DMF, Sequeira CAC. Sodium borohydride as a fuel for the future. Renew Sustain Energy Rev 2011;15:3980–4001. Stoller MD, Park S, Zhu Y, An J, Ruoff RS. Graphene-based ultracapacitors. Nano Lett 2008;8:3498–502. Tozzini V, Pellegrini V. Prospects for hydrogen storage in graphene. Phys Chem Chem Phys. 2013;15:80. Available from: http://pubs.rsc.org/en/Content/ArticleLanding/2013/CP/ C2CP42538F#!divAbstract.

2.3 Boron Prabhuodeyara M Gurubasavaraj, Rani Channamma University, Belagavi, India and Northern Illinois University, DeKalb, IL, United States Shanmin Gao, Ludong University, Yantai, China Narayan Hosmane, Northern Illinois University, DeKalb, IL, United States r 2018 Elsevier Inc. All rights reserved.

2.3.1 Introduction 2.3.2 Applications in Hydrogen Generation and Storage 2.3.3 Applications in Li ion Batteries 2.3.4 Application to Supercapacitors 2.3.5 Application in Thermoelectric Energy Conversion 2.3.6 Applications of Boron in Nuclear Reactor 2.3.7 Applications in Other Fields 2.3.8 Conclusion and Future Perspective Acknowledgment References Further Reading Relevant Websites

2.3.1

72 72 75 78 79 80 81 82 83 83 86 87

Introduction

Energy is essential to the worth of human life. At present, human beings are reliant totally on plenty and continuous supply of energy for livelihood and working. It is a vital component in all sectors of modern economies. Primary energy sources include nuclear energy, fossil energy like coal, natural gas, and oil. At present, fossil fuels are the leading components which contribute mainly to the ever rising world’s energy demand. The population explosion and subsequent economic expansion has led to the dramatic increase in the consumption of fossil fuels. Unfortunately, the combustion of fossil fuels release the undesired products such as greenhouse gas which causes serious environmental problems. To overcome these issues there is urgent need for the growth of green, inexhaustible, and extremely efficient processes of energy conversion and storage technologies. There are some promising alternatives like solar cells, fuel cells, supercapacitors (SCs), lithium-ion batteries, etc. are currently being investigated [1,2]. The materials used to design these devices are crucial to fundamental improvements in conversion and storage of energy. Therefore, creating rationally designed, economically viable and high-performance materials with specific properties for energy is need of an hour [3]. Recent years have seen remarkable research development in material science, including the composition of material, regeneration chemistry and catalysis, which leads to the up to date use of these systems and technologies. Among the numerous elements being investigated for the application in energy, boron and its compounds have attracted immense research attention as nuclear fusion as well as hydrogen storage media. In addition, some boron compounds have found the applications in SCs, lithium-ion batteries, thermochemical storage, etc. [4–7]. Boron is inexpensive, safe, and abundant. The usage of only 10% of boron produced currently is enough to generate all of the world’s electric power. The discovery of boron nanoparticles and boron nitride nanotubes, with variable functionalization patterns, has made the areas associated to boron materials predominantly significant [8]. Another important application in the energy sector of boron is its use as an engine fuel. Boron easily reacts with water and has an elevated combustion feature. Combustion of boron is an exothermic reaction and releases no gas. Boron fusion generates nonradioactive and inert helium as a product. The research on boron as energy fuel has been under way since several decades in countries like Germany, Russia, United States, France, and Canada. Boron can be considered as a potential engine fuel in the future. This chapter will present the most recent progress with regards to boron and its compounds in applications as energy materials concerning energy conversion (e.g., thermoelectric energy conversion, high energy density fuel cells, solar cell) and energy storage (hydrogen storage, SC, and lithium-ion battery). The prospects and future developments of boron materials for energy applications are also discussed.

2.3.2

Applications in Hydrogen Generation and Storage

There is an increasing research interest and worldwide attention for the clean, versatile and efficient use, and storage of energy. One of the most effective energy carriers is hydrogen (B39 kWh kg 1) [9]. The key challenge in utilizing hydrogen energy is its limitation in the large scale storage and transport. “On-board” hydrogen storage attracts the most extensive research in order to realize hydrogen-powered fuel-cell automobiles into the market for commercialization with hydrogen as a carrier. A vital solution

72

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00208-X

Boron

73

is expected to store and transport efficient and safe on-board hydrogen like material-based storage. This is due to the high cost of safety concerns, well-developed pressurized tank, problematic energy efficiency low volumetric hydrogen density, and/or cryogenic liquid hydrogen techniques which makes the primary requirements of vehicular applications unsuccessful [10]. In general, expected materials should have some desirable features like cost-effective, high storage capacity, excellent reversibility, fast kinetics, moderate operating temperature and low toxicity to be an ideal on-board hydrogen storage material (HSM) [11]. The US Department of Energy (DOE) has recommended a minimum storage capacity of 6.5 wt% and 62 Kg m 3 for a hydrogen storage system to be able to use in a model fuel-cell vehicle with a standard driving range of 560 km [12]. During the preceding decade, various methodologies have been tried to store and transport hydrogen, which include liquefaction processes, cryogenic and pressured cylinders, zeolites, hydrogen adsorption materials carbon nanotubes, and metal hydrides [13]. None of these methods were efficient and meet the estimated criteria on hydrogen storage or transport. Although, recent efforts for hydrogen adsorption capabilities were productive in bringing up the values near to 6–8 wt%, the storage method needs low temperature and high pressure [10]. For advanced hydrogen storage, governmental agencies around the world have supported substantial research on boron-based HSMs. When compared to pure liquid hydrogen various boron systems can efficiently store additional hydrogen by weight and volume by making use of exceptional safety features, hydrogen densities, and release rates. Recently, Wang and Kang, Umegaki, Hamilton and Orimo and their coworkers reviewed the several characteristics of HSMs that contain boron [14–17]. Metal hydrides have been excellent precursors and have exceptional ability for increased energy density storage at normal conditions [10]. Particularly, hydroboron-based storage, such as sodium borohydride (NaBH4), has been studied extensively to understand its potential for automobile, transportable, and stationary applications. Interestingly, it offers a high potential hydrogen density (max. 10.9 wt%) with catalytically controlled hydrolysis reaction and releases safe and easy hydrogen [18,19]. The use of NaBH4 to generate hydrogen dates back to 1940s, when the research group of Hermann I. Schlesinger at the University of Chicago, United States reported the work on alkali metal borohydrides [18]. Later on, Schlesinger’s PhD student and 1979 Nobel Prize winner Herbert C. Brown and his coworkers explored the importance of sodium and related boron hydride compounds as vital reagents for organic synthesis. However, recently a lot of attention is gained by the creative research in the field of boron hydrides which is used to store hydrogen. Metal tetrahydroborates are the metal salts and complexes containing BH4 anion which is formally referred as tetrahydroborate. However, these compounds are often informally referred as metal borohydrides following the tradition of former workers in the field [20]. These materials have the general formula, M(BH4)n, where M can be K, Li, Na, Mg, Ca, Zn, Al, I, or transition metals like Sc, Ti, V, Cr, Mn, Zr, U, etc., and n varies from 2 to 4. On the basis of weight and volume metal borohydrides can accommodate significant quantity of hydrolytically or thermally attainable hydrogen which depends on the atomic nature and valence of the metal. Stock [21], Schlesinger [22,23], and Wiberg [24] were the first to isolate metal borohydrides independently. Currently, there are numerous reports and literature available associated with these compounds. Sodium tetrahydroaluminate (NaAlH4) is reported to release B7 wt% hydrogen upon catalytic dehydrogenation at moderate temperature of 701C [25]. Nevertheless, due to their extreme sensitivity to air or moisture, these metal borohydrides have limited practical applications and also raise concern over safety issues. Metal borohydrides, similar to their metal hydride analogous accommodate large amount of hydrogen by weight and volume. They usually occur in solids or liquids and produce hydrogen upon heating directly, or hydrolysis (i.e., combining with water)/ other reactants or when passed through a reactor containing-catalyst. It was Schlesinger et al. initially accounted for the formation of aqueous solution of NaBH4 which is highly stable in basic solution. Further, the controlled hydrolysis reaction of NaBH4 solution with a heterogeneous catalyst can be achieved with the meticulous release of hydrogen [18]. This reaction is shown in Eq. (1) below [26]. NaBH4 þ 2H2 O-NaBO2 þ 4H2

DH ¼

75 kJ mol

1

ð1Þ

According to the reaction stoichiometry, at standard temperature and pressure (STP), 2.37 L of hydrogen is released by the hydrolysis of 1 g of NaBH4. This implies the gravimetric hydrogen storage capacity (GHSC) of 10.8 wt% on a reactants-only basis, which is well over the DOE designated storage density target. The ideal source for hydrogen seems to be aqueous borohydride, because of its stability under ambient conditions and its ability to release hydrogen in a safe and controllable way. On the contrary, numerous metal hydride storage materials, storing hydrogen in the form of aqueous NaBH4 solutions has quite a few benefits, such as the discharge of hydrogen at normal conditions, hence making it model for transportation. This system is also beneficial in regards to safety and control over the hydrolysis when compared to the solid hydrides. The control of hydrolysis may be due to its ability to act as a thermal shield which absorbs the heat released during the exothermic heat of reaction and averting the thermal delinquency. By controlling the amount of solution that is in contact with the catalyst, the release of hydrogen can be regulated and the system is allowed to meet the dynamic power demands of a fuel cell vehicle. The first report which demonstrated the use of aqueous NaBH4 solutions as portable hydrogen storage system was published in 2000 by Millennium Cell Inc., inspiring further research in this field [26,27]. In the mid-1980s, Kaufman and Sen [28] quantitatively studied the BH4 hydrolysis reaction kinetics in an acid environment. Meanwhile, there are uncertainties about the potentiality of these hydride systems to meet the entire DOE target. The GHSC of real storage systems will consistently be lesser in comparison with the theoretical 10.8 wt% because of surplus use of water required to dissolve the NaBH4 and its by-product NaBO2, along storage vessels with the added mass of the reaction. These inadequacies were addressed in the DOE’s report of a review paper on NaBH4 as a best compound to store hydrogen [29].

74

Boron

Furthermore, chemical hydrides are projected as “one-way” single use fuels unlike their reversible complex hydrides. Off-board regeneration discards the remaining by-product (spent material) from the vehicle [30]. With the deepening of the research in boride-based hydrogen storage systems, much of the research focuses on the catalyst. There is a need for developing economically viable highly efficient transition metal catalysts for hydrogen generation and storage applications. There are vast number of reports regarding the catalytic generation of hydrogen by borohyrdides [31–33]. There are number of compounds like Ru, Pt–Ru, Pd, Pt–Pd, Pt alloys [26,34–38], fluorinated Mg-based alloys [39], CoxB [40], NixB [41], and Ni–Co–B powders [31], which catalyse the hydrolysis reaction of NaBH4. Amendola et al. [26,34] reported aqueous BH4 solution hydrolysis to produce hydrogen for proton exchange membrane fuel cells (PFMFS) using Ru-catalysts. Even though noble metal catalysts have tremendous catalytic efficiency, but their expensive material cost makes the usage limited in practical applications. The synthesis of highly stable active nickel boride (NixB) and its application as a catalyst in the hydrolysis of alkaline NaBH4 solution was reported [41]. Moreover, the catalytic activity and operational stability of NixB catalyst at higher temperature of 1501C in vacuum was also reported. The experimental conditions of hydrolysis reaction can produce 44.0 wt% hydrogen at room temperature and 6:75 wt% hydrogen at 451C, exhibiting much higher hydrogen storage capacity than currently used materials for hydrogen storage. Since the NixB catalyst is inexpensive and easy to prepare, it is viable to use this catalyst in the fabrication of practical hydrogen generators for portable and in situ applications. Among numerous boron compounds, ammonia borane is a promising chemical hydrogen storage system [42–44]. The high gravimetric and volumetric capacities of some metal hydrides at near-ambient working conditions (often less than 801C at 0.1 MPa hydrogen pressure) make them as better candidates for storage materials. For instance, in presence of transition metal catalyst at 701C one equivalent of H2 is rapidly released from ammonia borane (NH3BH3) which contains over 19 wt% H2 and 150 g H2 L 1 of hydrogen by weight and volume, respectively (materials basis) [45]. Ammonia triborane (NH3B3H7) which is analogous to ammonia borane [46], is used abundantly as candidate to store hydrogen. The development of ammonia borane in the field of solid-state chemical hydrogen storage is however hindered by issues relating to thermolytic decomposition. On heating (e.g., over the temperature range 80–2001C), ammonia borane decomposes more than it dehydrogenates; large quantities of unwanted gaseous by-products (e.g., borazine, diborane, and ammonia) are released along with two equivalents of hydrogen [47]. Concomitantly, a solid residue of complex nature, with the empirical formula [BNHx]n where xo2, forms; as suggested elsewhere, it is aptly a mixture of polyaminoborane [H2N–BH2]n, polyiminoborane [HN¼ BH]n, o-polyborazylene [B3N3H4]n/3 and graphitic cross-linked polymer [B3N3Hy]n/3 with yo4 [48]. The decomposition scheme of ammonia borane is not adequate from usage point of view. The temperatures are too high and the hydrogen is not pure. The solid residue cannot be totally and properly rehydrogenated by chemical recycling [49]. In this context, strategies of destabilization of ammonia borane have been investigated with the purposes of decreasing the dehydrogenation temperature below 1001C, releasing pure hydrogen (while avoiding any by-products), and forming a solid residue of simple composition, ideally that of polyborazylene. To date, five different destabilization tactics have been reported: i.e., solubilization in an organic solvent or in ionic liquid; solubilization and addition of homogeneous catalyst; doping with solid-state oxidant or acid; nanoconfinement into the porosity of a host material; chemical modification toward the formation of derivatives like alkali amidoboranes (e.g., LiNH2BH3). By solubilising ammonia borane in a suitable solvent, the intermolecular dihydrogen N–H


H–B network is disrupted [50]. This results in the destabilization of ammonia borane allowing then dehydrocoupling and hydrogen evolution at temperatures lower than 1001C. For example, a triglyme (b.p. 2161C) solution of 6 M ammonia borane is able to generate one equivalent of hydrogen in less than 1 h at 701C, and with no induction period [51]. Metal acetylacetonates like Fe(O2C5H7)3, Co(O2C5H7)2, Ni(O2C5H7)2, Pd(O2C5H7)2, Pt(O2C5H7)2, and Ru(O2C5H7)3 are considered for assisting dehydrocoupling of ammonia borane in diglyme (0.135 M) at 501C [52,53]. The dehydrogenation kinetics can also be improved by using metal-based catalyst. For example, ammonia borane (0.4 M) in toluene (b.p. 110.61C) is able to release one equivalent of hydrogen in less than 1 h when catalyzed by a homogeneous ruthenium acetylacetonate-based catalyst at 601C [54]. From a mechanistic point of view, cyclic intermediates with singly dehydrogenated boron atoms are believed to form and to subsequently polymerize with the formation of polyborazylene. Cyclic intermediates like B-(cyclodiborazanyl) amineborane, B-(cyclotriborazanyl) amine-borane, and cyclotriborazane have been reported [55]. Among these metal acetylacetonate catalysts, palladium acetylacetonate was found to be very reactive towards ammonia borane, even in the glove box when both solids were put into contact without solvent [56]. Ruthenium acetylacetonate was shown to be an efficient precursor as an in situ homogeneous catalyst for ammonia borane dehydrocoupling. The homogeneous character was unequivocally demonstrated with the help of a mercury poisoning test [54]. Perez et al. studied the dehydrocoupling of liquid-state ammonia borane, NH3BH3 with assistance from an aided metal acetylacetonate [54]. The catalytic efficiencies of the metal acetylacetonates were compared by monitoring the hydrogen evolution with time (over a maximum of 2 h). The main objective was to determine the time at which the dehydrocoupling reaction could be stopped upon the evolution of less than 0.3 moles of hydrogen per mole of ammonia borane (conversion rates lower than 30%) to get and analyze the reaction intermediates at the early stages of the reaction. The hydrogen evolution curves are shown in Fig. 1. The most efficient metal acetylacetonate is Ru(O2C5H7)3 with the release of one equivalent of hydrogen in less than 2 h. This is in accord with the results reported by Duman and Özkar [54] where one equivalent of hydrogen was liberated in about 70 min at 601C (mol ratio NH3BH3/Ru of 160). The mechanisms of ammonia borane dehydrocoupling is roughly independent on the metal present in the acetylacetonate salts. Metal acetylacetonates catalyze ammonia borane dehydrocoupling by accelerating the reaction; they mainly have effect on the kinetics. In addition to the hydroboron and the ammonia borane compound, boron-based nanostructures also can be used as the HSM [57,58]. Recently, nanostructured materials have been attracting lot of research interest in hydrogen storage systems [10,59].

Boron

Fe Co Ni Pt Ru

1.0 mol H2 per mol NH3BH3

75

0.8 0.6 0.4

30% conversion 0.2 0.0 0

20

40

60 80 Time (min)

100

120

Fig. 1 Time evolution of hydrogen release by dehydrocoupling of ammonia borane (0.135 M) solubilized in diglyme, at 501C and in the presence of metal acetylacetonate (mol ratio NH3BH3/M of 100; with M as Fe, Co, Ni, Pt, or Ru for the metal of the acetylacetonate salts). Reproduced from Manon P. Philippe M. Umit BD. Mechanistic insights of metal acetylacetonate-aided dehydrocoupling of liquid-state ammonia borane NH3BH3. Adv Energy Res 2016;4:177–87.

Because of the larger surface area and high binding energy, nanomaterials offer better molecular adsorptions over their bulk analogous materials. For storage materials operating at ambient conditions, hydrogen must have the binding energy in the range of 0.2–0.5 eV. It demonstrates the ability to construct nanometer materials with proper characteristics of hydrogen binding. Recent studies show that boron nitride nanotubes can store hydrogen at elevated temperatures [60,61]. It has been reported that nanotubes of boron nitride (BN) can accommodate 1.8–2.6 wt% of hydrogen under B10 MPa at normal temperature [62]. Furthermore, higher hydrogen adsorption capacity (4.2 wt%) is exhibited by collapsed BN nanotubes than any multiwalled carbon nanotubes [63]. Computational simulations reported by Jhi et al. have shown that BN can be a best HSM. The studies also reveal that the binding energy to hydrogen in boron nitride increases with increase in deviations from sp2 bonding. It is likely that layered ionic materials, which are more ionic than boron nitrides have substantially high binding energy and can store enough hydrogen with a moderate substitutional doping at normal temperatures [60]. Chen et al. used LaNi5/B mixture and nickel powder to synthesize BN nanotubes through chemical vapor deposition over a wafer at 1473K. Their studies confirm that the BN nanotubes can be used for the hydrogen storage by means of an electrochemical process. It is uncertainly concluded that, the advancement in the electrocatalytic activity by modifying the surface with metal or alloy on the boron nitride tubes can increase the ability of BN nanotubes to electrochemically store hydrogen [64]. In addition, Jhi also studied the activated forms of BN nanotubes for possible applications to store hydrogen with the use of the pseudopotential density functional method [65]. The binding and diffusion energies of hydrogen adsorbed on activated boron nitride nanotubes were calculated and are in the ideal range of ambient temperature storage. It is also noted that the increase in storage capacity is attributed to the hydrogen diffusion through the active sites which allows the access of the inner surface in the nanotubes. These studies offer a tangible solution to enhance the capacity to store hydrogen by increasing operating temperature and based on heteropolar nanomaterials like boron nitride nanotubes. Li et al. reported the density functional computations on Ca-coated boron fullerenes and nanotubes for the adsorption and storage of hydrogen [8]. Their study shows that notorious clustering problem was not observed in these compounds due to the strong binding of Ca to the surface of B80 fullerene and boron nanotubes. 12 Ca atoms are coated on B80 fullerene which can accumulate up to 60 H2 molecules with an average binding energy of 0.12–0.40 eV, analogous to hydrogen storage gravimetric density of 8.2 wt%. Ca-covered (9, 0) boron nanotube has binding energy of 0.10–0.30 eV with the hydrogen storage capacity of 7.6 wt%. The charge transfer between Ca and boron fullerenes and nanotubes is because of the strong binding with each other. At atmospheric conditions, hydrogen adsorption and desorption are reversed feasibly due to the optimal molecular hydrogen adsorption energies. But the hydrogen storage media proposed in the nanoscale; decreases the hydrogen capacity significantly in macroscopic materials. However Ca-coated B nanostructures as building blocks may be prospective for high gravimetric and volumetric hydrogen storage capacity. There is still a huge challenge for the research to gather the ideal media into suitable macroscopic materials for practical hydrogen storage. At present, there is a need to focus research on the new approaches to synthesize and/or process new materials with improved performance as HSMs. It is familiar to a novel design and efficient reaction system would be vital to realise these targets with the limitations identified in conventional systems [66].

2.3.3

Applications in Li ion Batteries

Lithium-ion (Li-ion) batteries (LIBs), have found many applications in portable devices like smart phones, digital cameras, laptops, electric and plug-in hybrid electric vehicles (PHEVs) and hybrid electric vehicles (EVs and HEVs) due to their elevated

76

Boron

Fig. 2 Structure of LiFeBO3 (green: iron ions; orange: b ions; red: li ions). (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of the chapter.) Reproduced from Legagneur V, An Y, Mosbah A, et al. LiMBO3 (M¼ Mn, Fe, Co): synthesis, crystal structure and lithium deinsertion/insertion properties. Solid State Ion 2001;139:37–46.

power storage capability, reduced weight, and low self-discharge capacity [67,68]. The trend in the consumer electronics market is the use of lithium ion secondary batteries as the excellent transferable energy storage devices. Enhancing the ability of the electrodes to release or by enhancing the working ability of the cathode materials, batteries can achieve greater energy densities. LIB has anode, cathode, and electrolyte as three major components of which the electrode that has got immense research focus is cathode as it is most expensive and has the highest weight in the battery. Indeed, the predominant role is played by cathode materials with the determination of life cycle, energy density, and safety of Li-ion batteries. Hence, one of the rising topics at international meetings on Li-ion batteries is the research and development of cathode materials. Most of the research has been focused on the borates LiMBO3 due to the presence of its light BO3, polyanion group, which certainly has higher theoretical energy density when compared to that of other polyanion cathode materials. The reports on the electrochemical properties of (M ¼ Mn, Fe, Co), were originally renowned by Legagneur et al. and suggested that LiMBO3 can only (de)intercalate 0.04Li per formula, i.e., 9 milli ampere hour per gram (mAh g 1), at a rate of C/250 (the theoretical capacity is 220 mAh g 1) [69]. The LiFeBO3 structure is shown in Fig. 2. The FeO5 bipyramids and BO3 trigonal planar are used to construct the framework of three dimensional FeBO3. The FeO5 bipyramids share single chains along the edges [001] and BO3 shares three chains towards the corners. The two tetrahedral sites are occupied by lithium and share an edge to form chains along the [001] direction, in this three dimensional framework. Dong et al. [70] later reported that at the initial discharge, capacity of 91.8 mAh g 1 can be obtained. Carbon has an important application in manufacturing the LiFeBO3/C composite in addition to coating the materials and to obtain higher discharge capacity values (158.3 mAh g 1 at 5 mA g 1 and 122.9 mAh g 1 at 50 mA g 1) [71]. A thorough investigation of this material to understand its potential was not reported till 2010, when Yamada et al. [72], approached a capacity of 200 mAh g 1 a claim sustained by both experimental and theoretical results. The inherent activity of LixFeBO3 is supported by ab initio calculations which indicate that without thermodynamic limitation potential is considered as an electrode material that approaches a theoretical capacity of 220 mAh g 1. Further, they revealed that, the sources of the contamination that occurred in their previous studies are due to the moisture present in the air that poisons the surface. With appropriate measures in sample and electrode handling, they could attain the theoretical capacity at C/20 rate, and at 2C rate, they achieved more than 75% of the theoretical capacity. In addition to LiFeBO3, manganese-based borates have been investigated in the very recent years [69,73,74]. LiMnBO3 has two polymorphs, hexagonal [69,73], and monoclinic [74]. The initial discharge capacity of 75.5 mAh g 1 of hexagonal phase was reported with an extensive voltage window range of 1.0–4.8 V [73]. No reports were found on the electrochemical data for the monoclinic phase until early of this decade [74] and a better confinement was reported in LiMnBO3 coated with carbon with a second discharge of 100 mAh g 1 over multiple cycles. The new generation Li intercalated borate materials are relatively poor performers when correlated to the other cathode materials. Because of the small size and large charge of lithium, they are highly moisture sensitive and the kinetic polarization was also observed. There is still lot of work is needed to understand and emulate the limiting factors of these materials and explore the optimal synthetic and operational conditions [75]. In addition, the recent progress of the secondary batteries like lithium ion batteries is renowned by their applications as innovative structures in the use as anode. To enhance the energy density and to generate economic Li-ion batteries, many elements have been used in creative fabrication to support these anodes [76]. The anode material microstructure made of carbon and graphite decides the working of lithium ion secondary batteries which also include the cyclic stability, charge/discharge capacity, and voltage profile. The ability of a representative lithium ion battery has been significantly upgraded (1.7 times) because of the use of carbon components as anode. However, there is still active research is going on to understand the major parameters of the carbon that offers the enhanced anode characteristics, as carbon and graphite systems have high variations in the crystallinity, morphology, texture and microstructure, which depends on precursor materials and also their synthesizing methods for existence of many forms like powder, fibre, and spherule [77]. As regards the anodic properties of graphite are affected by the structural

Boron

77

factors, especially lesser capacity than the theoretical value of 372 mAh g 1, the effects of a–b axis crystallite size [78], stacking fidelity [79], and basal plane defects have been put forward as being reported. Interestingly, graphite systems doped with hetero atoms like B, N, and P within graphite materials will improve its capacity [80]. Heteroatom-doped carbon, such as BCx, BCxN, and CxN are recommended for probable demands as anode materials in Li ion batteries, because these compounds contain layered structure [81–83]. Particularly, boron-doped carbon materials have been experimentally and theoretically explored for various features, from electronic properties to potential applications, like the protection toward oxidation at elevated temperature for carbon/carbon (C/C) composites and as an anode material for Li ion batteries. Because boron-doping is inducing the formation of an electron acceptor level [81–83], the improved capacity has been predicted. Several researchers [81–83] have reported and proposed synthetic procedures for boron-doped carbons by co-deposition of chemical vapor deposition (CVD) pyrolysis of boron containing organic compounds and the substitutional boron doping processes into carbon structures. It can be noted that, the boron-doped samples have higher voltage profiles of 40 mV relative to those of the undoped samples and this can lead to potential practical cell applications. On inducing an electron acceptor level, the boron-doped samples show characteristic shoulder plateau at about 1.3 V on the discharging cycle, so that lithium insertion yields a higher voltage compared to undoped samples [82]. The average ratio of the discharge and charge process capacities gives rise to irreversible capacity. It is fascinating to note that for boron-doped samples the loss of irreversible capacity is inferior in comparison with the analogous undoped samples. These results may indicate that the introduction of an electron acceptor in the lattice of the boron-doped samples, the Fermi levels are reconstructed and dropped by boron-doping [82]. The existence of boron in some boron-doped graphite like boron nitride and boron carbide degrades lithium insertion capacity. Also the unpredicted adverse effects of boron doping can be linked to the heterogeneous growth of the crystallites dimension, L, due to the borons acting as a graphitization catalyst. In order to better understand the impacts of boron doping, depending on the carbon materials gained from a different precursor with a wide variety of shapes and microstructure, the doping conditions, including the atmosphere and heating rate should be cautiously selected. This process could offer new types of materials for carbon or graphite electrodes. In addition to the use of anode and cathode materials in lithium ion batteries, the boron compounds have also been used as electrolyte materials. Lithium-ion batteries consists of specifically designed electrolyte and highly energetic electrode materials. A number of additives are studied to understand the stabilization carbonate-based electrolytes on 5 V-class cathode surfaces at normal temperature. Among the numerous additives, lithium bis (oxalato) borate (LiBOB) has been identified as a powerful additive that produces a stable solid electrolyte inter phase (SEI) on both graphite and silicon electrodes [84]. LiBOB exhibits several advantages over conventional lithium salt which includes being environmentally benign, low cost and shows good thermal stability. Particularly, graphite structure is stabilized effectively by LiBOB even in pure propylene carbonate (PC) solvent [85]. Moreover, usage of LiBOB as an additive forms an SEI film, thereby protecting the graphite anode from exfoliation by the PC-rich electrolyte. However, the negative impact on the bulk properties like ion transport makes the effect insignificant [86]. Very recently, the consequence of LiBOB as an additive on the cycling performance of high-voltage cathodes, such as LiNi0.5Mn1.5O4 and LiCoPO4, has been explored at normal temperature [87,88]. The report by Ha et al. has indicated that the electrolytes with LiBOB additive can form a stable cathode SEI which prevents electrolyte degradation and improves discharge capacity with retention of LiNi0.5Mn1.5O4 at 601C. This SEI layer may hinder undeviated interaction of the electrolyte compounds with the high-voltage cathode which on continuous cycling diminishes the electrolyte solution significantly by oxidative decomposition. A correlated study on the cycling performance of LiNi0.5Mn1.5O4 and LiBOB-doped electrolytes with the reference standard is shown in Fig. 3. 100

90

80

80

60

70

40

20

Coulombic efficiency (%)

Discharge capacity retention (%)

100

60

Ref 1% LiBOB

50

0 0

20

40 Cycle number

60

80

Fig. 3 Discharge capacity retention (filled symbol) and coulombic efficiency (blank symbol) of LiNi0.5Mn1.5O4/Li cells at 601C. Reproduced from Ha SY, Han JG, Song YM, et al. Using a lithium bis(oxalato) borate additive to improve electrochemical performance of high-voltage spinel LiNi0.5Mn1.5O4 cathodes at 601C. Electrochim Acta 2013;104:170–7.

78

Boron

While capacity diminishing still occurs in the cells containing LiBOB additive, the discharge capacity retention of LiNi0.5Mn1.5O4 was considerably enhanced from 66.9 to 78.7% after 80 cycles at 601C. Moreover, in the absence of LiBOB additive, Li/LiNi0.5Mn1.5O4 half cells showed a low coulombic efficiency of 95%, indicating the loss of 5% of the Li on each cycle mainly due to the electrolyte degradation during cycling at 601C. The ex situ X-ray photoelectron spectroscopy (XPS) of the separators recovered from cells charged up to 5.0 V clearly showed the considerable electrolyte decomposition to light brown color product at the cathode. The ex situ ATR-Fourier transform infrared (FTIR) and XPS results for the oxidative decomposition of electrolyte compounds can be efficiently improved with LiBOB-derived SEI and is confirmed from delithiated cathodes, and measuring the leakage current at a constant voltage of 5.0 V [89]. Pieczonka et al. [90] systematically explored the influence of a LiBOB electrolyte additive on the working of full lithium-ion cells coupling with high-voltage spinel cathode and graphite anode. Significant improvement of cycle life and Columbic efficacy of the full-cells at 30 and 451C were found just by adding 1 wt% of LiBOB to the electrolyte. When compared with LiBOB-free electrolyte, the LiBOB was preferably oxidized and reduced during cycling, their relative contributions to the enhanced capacity retention in full-cells was assessed by pairing fresh and LiBOB-treated electrodes with various combinations. The results indicated that on reduction of LiBOB additive forms powerful and stable SEI film on graphite during cycling at 451C against Mn dissolution problem compared with the SEI formed by the reduction of the base (LiBOB-free) electrolyte. In addition, the fully charged Li1 xNi0.42Fe0.08Mn1.5O4 (LNFMO) electrodes were stored at 601C for 1 month and a 3 wt% LiBOB electrolyte was added, Mn dissolution reduced when compared with the base electrolyte. It was suggested that LiBOB traps PF5 and stabilizes the electrolyte, i.e., sequestering the radical which oxidizes EC and DEC electrolyte solvents. Thus, the positive electrode results in the suppression of oxidation on the carbon black particles. Therefore, Mn dissolution from the spinel cathode reduces due to suppressed HF generation. Although, there are many advantages of using LiBOB as an electrolyte additive, it has limitations of weaker conductivity and low solubility in many solvents (e.g., linear alkyl carbonate). Besides in a cell, it produces the SEI layer with high impedance anode material having carbon on its surface, which severely damages the cryogenic property and discharge capacity of the cell [91]. In order to overcome these shortcomings, other boron compounds, such as lithium difluoro(oxalato)borate (LiODFB) and tris (trimethylsilyl)borate (TMSB) were studied as alternative salts for lithium-ion batteries. As a prospecting salt for lithium-ion battery, LiODFB has following benefits [91]: (1) it has superior solubility over LiBOB in linear alkyl carbonate solvents which lowers viscosity and increases wettability of the electrolyte, in turn enhances the rate ability and low-temperature cycle performance of the lithium-ion battery. (2) Although similar to LiBOB, LiODFB facilitates the use in pure PC as carbonaceous anode material due to the formation of SEI film on its surface which leads to reduced concentration of oxalate group in its molecule and the reduced irreversible capacity for the initial cycle of a lithium-ion battery results at above 1.5 V, and (3) unlike LiBOB-based electrolyte, the SEI layer formed on the surface of carbonaceous anode material in case of LiODFB-based electrolyte has lower impedance, which helps in enhancing the power capability and low-temperature cycle performance of a lithium-ion battery. These features strongly indicate the necessity of exploring new and novel synthetic methods and finding appropriate solvents which can be applied to lithium ion batteries. Li et al. reported the simpler, cheaper, environmentally friendly method for synthesizing LiODFB (99.5%, by weight) along with LiBF4 (99.5%, by weight) as a by-product in high yield. The electrochemical study of sulfolane is carried out (SL) with LiODFB using dimethyl sulfite (DMS) as a solvent. The study revealed that LiODFB-SL/DMS (0.9 mol L 1, 1:1, by volume) electrolyte has significant stability against oxidative decomposition (45.5 V) and shows reasonable conductivity. Moreover, the LiODFB-SL/DMS electrolyte in Li/MCMB cells exhibits exceptional film-forming properties and stable cyclic performance. In addition, 0.9 mol L 1 LiODFB-SL/DMS electrolyte when used in LiFePO4/Li cells, show many advantages over other electrolytes such as 1.0 mol L 1 LiPF6-EC/DMC or 0.9 mol L 1 LiODFB/DMS in the LiFePO4/Li cell. For instance, it exhibits highly stable electrochemical performances at either high or comparatively low temperatures. It has been proposed that a LiODFB-SL/DMS electrolyte must be taken into account as a candidate electrolyte material for lithium-ion batteries [92]. To enhance the cyclability of distinct cathodes including Li [Li0.2Ni0.13Mn0.54Co0.13]O2 [93], LiNi0.5Mn1.5O4 [94] and LiMn2O4 [95], TMSB has been applied successfully. It has been known that electrochemical oxidation of TMSB occurred easily than the electrolyte and produces SEI film consisting of silicon and boron-containing compounds as its oxidation decomposition products. Contrast to the electrolyte, TMSB degradation does not release gaseous products and the protection for cathodes is furnished by the resulting SEI film [96]. The study by Zuo et al. validates that a TMSB addition to the electrolyte can radically increases the cycling performance of the LiNi0.5Co0.2Mn0.3O2/graphite cell at larger voltage operation. TMSB shows potential prospect at higher voltage, in the voltage range of 3.0–4.4 V, LiNi0.5Co0.2Mn0.3O2/graphite cell holds about 92% of its basic capacity with TMSB in the electrolyte when compared to the cell that retains only 28.5% of its initial capacity without additive in the electrolyte after 150 cycles. The improved cycling performance is related to the thinner film produced from TMSB on the LiNi0.5Co0.2Mn0.3O2 and the combination of TMSB with PF6 and F in the electrolyte, which prevents the unwanted decomposition of EC solvents and results in lower interfacial impedance [97].

2.3.4

Application to Supercapacitors

SCs are among the most favourable devices to store electrochemical energy [98]. Based on their mechanism of storing energy, SCs are categorised into two types. One is electrochemical double layer capacitors (EDLC), which consists of conductive porous

Boron

79

electrodes and its capacitance depends on the electrostatic adsorption of electrolyte ions on its surface area. Second category is pseudo-capacitors, in which the reversible redox reactions at the electrode/electrolyte interface are understood to be the basis for energy storage. Both mechanisms depend on the nature of the electrode materials, such as the availability of pores, the wettability of the electrodes the conductivity of the materials, could be known by the nature of surface functionalities. Carbon-based systems due to their remarkable ease of processability, physiochemical properties, low production cost, tunable porosity, excellent electrical conductivity and can be mould into variable forms make them as the most commonly used electrodes in the materials applied in SCs. SCs are the excellent energy storage and power output devices and these are meant for their incredible rising demand for clean energy technologies and are used in electric vehicles, portable electronics and various renewable energy systems which work on intermittent sources like solar and wind energy. Electronic structure and chemical structure of porous carbons decides the storage of energy in SCs that are made of carbon which also depends on intake of charge in the carbon/ electrolyte interfacial region [99–101]. Pseudocapacitance is introduced by the significant increase in the interfacial capacitance due to surface chemical functional groups. It is reported that for multiwalled carbon nanotubes, specific capacitance per surface area increases with boron doping [102]. Boron is an element studied with keen interest and most investigated for decades in terms of improving the features like Li-ion insertion, increased resistance to oxidation and electrochemical behavior by substituting carbon or diamond functionalization [103]. Boron has three valence electrons and acts as an electron acceptor. It enters into the carbon lattice and occupies the trigonal sites [104] by substituting carbon. Thus, the electronic structure of boron doped carbon [105,106] is modified as the electrons in the Fermi level shift to conduction band. The capacitance of electric double layer is affected because of this change in electronic structure of carbon electrode materials. Most essentially, catalytic effect is exhibited by doping boron to a lower extent and leads to oxygen chemisorption on the surface of carbon, and redox reactions take place with the introduction of functional groups of oxygen on carbon surface [103,107,108]. Consequently, boron doping is capable in modifying the electrochemical capacitance of carbon systems, involving pseudocapacitance and electric double layer capacitance [109]. Chen et al. reported the homogeneously boron doped mesoporous carbon of an SBA-15 silica template by co-impregnation and carbonization of sucrose and boric acid [109]. The electronic structure of doped mesoporous carbon alters on low-level boron doping exhibits catalytic effect on oxygen chemisorption at edge planes. These features are important for considerable increase in the interfacial capacitance. Alkaline electrolyte (6 M KOH) and/or an acid electrolyte (1 M H2SO4) increase the capacitance to 1.5–1.6 times larger in boron-doped carbon than that in boron-free carbon. To discover new doped carbon electrode materials for SCs greatest capacitance boron doped mesoporous carbon can be expected for additional optimization of the local boron doping environment. Guo et al. carried out an apparent method for synthesizing boron and nitrogen co-doped porous carbons (BNCs) using precurcors like citric acid, boric acid and nitrogen as C, B, and N, respectively [6]. The promising capacitances were achieved from boron and nitrogen enriched carbon materials achieve (denoted as BNC). The addition of heteroatoms like nitrogen, boron, and oxygen into the carbon matrix leads to pseudocapacitive effect and improves the wettability between the electrolyte and electrode substances. Among BNCs, a high specific capacitance of 268 and 173 F g 1 was seen in BNC-9 and BNC-15 models with corresponding large specific surface areas of 894 and 726 m2 g 1 with the current at 0.1 A g 1. For BNC-9 and BNC-15 the corresponding energy densities were found to be 3.8 and 3.0 W h kg 1 and the power densities were 165 and 201 W kg 1 when the current of 1 Ag 1 was passed through it. Hence, BNC-15 is extremely appropriate to apply in high-power-demanded occasion, whereas BNC-9 is liable to store extra energy. Wu et al. have established a basic model on the basis of three-dimensional (3D) nitrogen and boron co-doped monolithic graphene aerogels (BN-GAs) of all-solid-state supercapacitors (ASSSs) that has exceptional working ability [110]. The system has electrode-separator-electrolyte incorporated into its structure, and the electrodes are GAs which acts as additives, the solid-state electrolyte is H2SO4 gel/polyvinyl alcohol (PVA) with a thin separator. The electrochemical performance of carbon-based materials increases on boron and/or nitrogen doping in carbon systems because of the charge transfer between the neighbouring carbon atoms [111–114]. GAs is formed as macroporous architecture with 3D interrelated scaffold which supports the electron transfer and diffusion of ions in the bulk electrode. Furthermore, gigantic BN-Gas on physical pressing is made into thin electrode plates of preferrable size. As a result, the BN-GAs based all-solid-state-supercapacitors (ASSSs) display merely curtailed system thickness, however exhibits a good rate capability, large specific capacitance (B62 F g 1), and improved energy density (B8.65 W h kg 1) or power density (B1600 W kg 1) with respect to layer-structured graphene paper, boron doped (B-GAs) GAs, nitrogen doped (N-GAs), or undoped (U-GAs) [115].

2.3.5

Application in Thermoelectric Energy Conversion

At present, 11% of world’s electric power is generated by nuclear energy which is recognized to pose danger. There is a need to find an alternative energy sources to produce electric energy. The most reliable method to generate electrical power is by thermoelectric energy conversion, for instance from solar heat or from waste industrial thermal energy. One significant technical advantage of thermoelectric energy compared with other methods is it contains immovable device parts. Therefore, once installed, they do not require further assistance in operation and are dependable for long time. For instance, satellites which contain thermoelectic devices ensure to supply electric energy to all the electronic equipment [116]. Unfortunately, these devices are economically not viable due to their limited efficiency of about 5%. Therefore, there is a need for designing and production of efficient and feasible devices for thermoelectric energy conversion.

80

Boron

Recent investigations have shown that high efficiency thermoelectric energy conversion can be achieved using boron-rich solids. The remarkable stability combined with high melting points of these compounds allows their usage under severe circumstances which are not possible by generally by other substances. Several boron-rich semiconductors display enthusiastic transport features, which include monotonic increase in Seebeck coefficients up to elevated temperatures, very low thermal conductivities and also show electrical conductivities with values similar to semiconductors [116]. The characteristic property of boron-rich solids is the most common structural feature of Bl2 icosahedra with structural constituents that are arranged differently in the distinct structures. These icosahedra structural arrangement governs the electronic properties of these boron-rich solids and therefore there is a close relationship between the electronic properties and structural arrangement around the icosahedra [117,118]. These materials exhibit unique electronic transport mechanism and are different from the classical and amorphous semiconductors. Noticeably, several structural features provide favourable conditions for large efficiencies in the conversion of thermoelectric energy. Another important property is the elevated melting points (e.g., boron carbide, Tm4 2600K) of these solids which are exceptional when compared to former semiconductors and permit superior functioning temperatures that are not common to the other semiconductors. These remarkable properties lead to better Carnot efficiency which has larger values than all other semiconductors known. For the application in thermoelectric property, boron carbide is the most comprehensively studied boron-rich solid. It has a wide uniformity range which extends from B4.3C at the carbon-rich limit and to about Bl2C at the boron-rich limit. Depending on the stoichiometry, the unit cells with 15 atoms in the idealized structure comprise of B12 or B11C icosahedra, C-B-C or C-B-B chains. It has been observed that, reduction of carbon content increases the number of unit cells without any chain [119,120]. Completely homogeneous structure has not been found for any chemical composition of the solid. This indicates the critical influence of chemical composition on the properties of the boron-rich solids. Boron and borides actively participate as high-temperature thermoelectric materials due to their (1) hopping conduction mechanism leads to increased Seebeck coefficients and electrical conductivities with rising temperature and (2) their complex structures will relatively give low values at elevated temperature thermal conductivities. They are extremely potent and have better chemical stability at high temperature atmosphere. As most of those studied earlier (e.g., AlB12 [121], B13C2 [122], B–Si–C [123] exhibit p-type conduction and to design thermoelectric devices, n-type boron-based substances are solemnly required. Imai et al. suggested pseudopotential method within the local (spin) density approximation to determine the densities of states of metal-borides of the CrB-, FeB-, MoB-, AlB2-, ReB2-, CaB6-, and UB12- types of structures and some tetraborides (YB4, CrB4, WB4, and MgB4) and the energetics of solid solutions of metallic atoms (Zr, Cr, and V) in b-rhombohedral boron (b-boron) to screen the candidates for n-type boride-based thermoelectric materials [124]. It was concluded that the E site of b-boron is occupied by Zr whereas the A1 site is occupied by V and Cr. The “rigid band approach” is applicable for metal-monoborides and diborides. These borides depend on the Mott’s explanation for the trend of the Seebeck coefficient for transition metal elements and are also helpful in knowing the large negative Seebeck coefficient of FexCo1 xB. N-type thermoelectric conversion substances are synthesized by using Y- or La- doped Ca(Sr, and Ba)B6. Takeda et al. have prepared and studied the thermoelectric features of polycrystalline AlMgB14 and some hexaborides (CaB6, SrB6, YbB6, SmB6, and CeB6) [125]. A single phase of orthorhombic AlMgB14, which holds B12 icosahedral clusters as building blocks, was obtained at sintering temperatures between 1573 and 1823K. At room temperature the Seebeck coefficient (a) and the electrical conductivity (s) of the phase were found to be about 500 mV K 1 and 10 1 1/Om, respectively. These values are in comparison with metal-doped b-rhombohedral boron. Conversely, at 1073K a large negative value ranging from 100 to 270 mV K 1 was observed in metal hexaborides with divalent cation. Within the complete variety of temperatures considered, it was noticed that 10 3 W K 2 m is the calculated power factors of CaB6 and SrB6. Consequently, they can be characterized as capable candidates for n-type thermoelectric systems. As shown above, boron-rich semiconductors are best candidates for thermoelectric energy conversion. Advanced organized scientific and technical progress of the devices has increased the efficiencies to 25% or even more. Although, these efficiencies are analogous with those of smaller electrical power stations the reliability of thermoelectric devices far exceed their reliability. Hence, wide range of valuable ecological and economical applications of thermoelectric energy conversion has come within reach by the boron-rich solids.

2.3.6

Applications of Boron in Nuclear Reactor

Fossil fuels are the core sources for electricity generation around the globe. However, burning carbon fuels, such as coal, oil or gas, produces huge amounts carbon dioxide, and, sulphur dioxide which affect the environment [126]. Fossil fuels are nonrenewable and hazardous, but alternative energy sources like wind and solar are renewable. They all have limitations and are not able to provide all of the required quantities of electricity for our daily life. On the other hand, nuclear power generation is attracting renewed global interest as regards replacing nonrenewable fossil fuels. In a nuclear reactor operation, it is important to control the neutrons, which are generated by a fission reaction. To control the fission rate, control rods are used in nuclear reactors. Usually, control rods are composed of boron, silver, indium, and cadmium which absorb the neutrons without undergoing fission reactions. Boron has a wide absorption spectrum which makes it suitable as a neutron shield. The 10 B, an isotope of boron consists of a substantial neutron absorption cross-section, for thermal neutrons [127]. A typical nuclear reaction involving thermal neutron absorption is shown in the following equation in which stable isotopes of helium and

Boron

81

Fig. 4 Boron-based control rod assembly for pressurized water reactors (PWR) nuclear reactor.

lithium are produced [127]. 10

B þ 1 n-4 He2þ þ 7 Li3þ þ að2:31 MeV Þ þ gð0:48 MeV Þ

The boron-based control-rod materials are particularly useful for pressurized water reactors (PWRs) (Fig. 4) and boiling water reactors (BWRs). It is important to note that boron in its elemental form is not suitable for use in the nuclear reactor due to its weak mechanical properties and, therefore, alloys or compounds of boron are used. Boron carbide and refractory metal borides are also commonly used as control-rod materials in nuclear reactors. They possess attractive properties of hardness, low density, chemical inertness, high melting point and magnificent thermal and electrical characteristics [128–130]. Boron carbide may possibly be prepared by a carbothermic reduction process at 415001C from commercially available boric acid as per the following reaction [130]: 4H3BO3 þ 7C-B4C þ 6H2O þ 6CO. However, this reaction pathway only gives a yield of about 65% based on boron content. Therefore, an advanced method has been developed to prepare boron carbide by solid–solid reaction between elementary boron and carbon as described by the following equation [131]: 4B þ C-B4C. In summary, boron-based neutron absorber materials are of great importance in nuclear reactors. Research for the synthesis of high quality boron carbides and related materials are currently being investigated. Application of advanced nanotechnologies such as 3D printing may significantly improve the process and the product properties.

2.3.7

Applications in Other Fields

In addition to the above application, the boron compounds are also used in other fields, such as highly productive energy transfer in the light harvesting system, photovoltaic (PV) films in solar cell, high energy density fuel, and so on. Glasses containing PbO and B2O3 have become predominantly a novel category of optical materials for optoelectronics and they have been seeking a great attention in laser research. Lead borate glasses belong to amorphous materials, which are interesting from the structural point of view [132]. Their structural properties can be easily modified by laser irradiation [133], heat treatment process [134], or gamma ray interactions [135]. Borate glasses of rare earth ions with different chemical compositions exhibit optical properties and are renowned in the literature [136–140]. Spectroscopic and laser properties of Nd3 þ , Tb3 þ , and Eu3 þ ions in multicomponent PbO–B2O3 glass systems have been studied [141,142]. The significant increase in measured lifetime, peak stimulated emission cross-quantum sections, efficiency and radiative transition rate is due to the addition of heavy metal oxide like lead oxide, to the borate matrix and is essential for the construction of near-infrared solid-state lasers [135]. Particularly, laser operation at 1061 nm is carried out for the comparatively greater emission cross-sections of Nd-doped lead borate-based glass. One of the budding technologies is the future use of PV cells as alternative energy sources for the direct conversion of sunlight into electricity [143]. The production of robust, high efficiency, and economical PVs seems to be one of the greatest challenges in

82

Boron

the field of PV’s and new varieties of materials with superior features like electrical and optical properties have to be developed. Although semiconductor industries are making a huge profit with the production of solid state junction devices that are usually made of silicon which is available easily and therefore these devices are dominating the field of PV devices. Currently, crystalline silicon-based PVs offer excellent possibilities for the production of robust, efficient, and relatively low cost devices in large-scales [144]. Chemical vapour deposition (CVD) and ion implantation techniques are used for doping boron with p-type carbon films to manufacture PV cells [145,146]. Ma et al. used an arc-discharge plasma chemical vapour deposition (PCVD) technique to synthesize boron-doped diamond-like carbon thin film on n-silicon substrate. On controlling the concentration of boron and deposition conditions the film showed low-resistivity and high optoelectronic efficiency. It is found that the microstructure of the film is totally and highly reliant on the transfer of electron [147]. Boron is found to be of great interest as a high energy density fuel with highest volumetric heats of combustion. An engine to work efficiently in many appliances, rapid fuel ignition, and the combustion are the main requirements and it can be very hard for an intricate hydrocarbon fuel to attain these conditions. The speed of the process can be improved by adding fuel-soluble catalyst to the fuel [148,149]. The considerably superior components are boron and boron-rich solids that have both gravimetric (58.5 MJ kg 1) and volumetric (135.8 MJ L 1) energy densities in comparison with that of hydrocarbons. The trouble involved in igniting and burning boron efficiently has lead to delayed recognition and the lower potential of it as a fuel or fuel additive. The fact that rapid efficient combustion seems to be difficult as combustion of boron is naturally a heterogeneous process. An apparent strategy is to increase the surface area/volume ratio by reducing the size of the particles. This approach is restricted to the fact that boron forms a B0.5 nm thick native oxide layer due to the enhanced portion of the particle mass with decrease in size resulting into further reduction in combustion properties [150]. As suggested by several researchers, nanoparticulate boron can be used to speed up the heterogeneous chemistry which leads to huge surface-area-to-volume ratios [151–155]. However the native oxide layer that forms on boron surfaces on exposure to air limits this strategy as discussed above [150]. A variety of processes have been performed to mitigate the effects of the oxide layer and enhance the ignition of boron nanoparticles, including the following [156–162]: (1) treating boron particles with TiCl4 and triethylaluminum followed by the addition of ethylene or even coating boron particles with LiF and trimethylolpropane, (2) sodium naphtalenid reduction of BBr3 in 1,2-dimethoxyethane followed by n-octanol, (3) metals, such as titanium and Mg coated with boron particles, (4) consumption of active components like glycidylazide polymer (GAP) and azide polymer (AP), coating the surface of boron, and (5) capping particles of boron with an organic oleic acid layer [163]. The combustion of hydrocarbon carrier fuel or boron can be increased by a different approach by coating the boron particles with a component like catalyst. Further development in the combustion efficiency of the solid particles with such a catalyst would result in a faster combustion of the hydrocarbon carrier fuel. A feasible scheme for increasing the ignition of the hydrocarbon carrier fuel (or the binder in a solid propellant or explosive), and the total energy density is the use of a carrier for a combustion catalyst with boron nanoparticles. Devener et al. synthesized air-stable in a single step with an easy and scalable process. Boron nanoparticles obtained are coated with a combustion catalyst, unoxidized, and soluble in hydrocarbons. Particles with sizes of B50 nm are produced using Ball milling, oleic acid functionalization is used to protect the particles against room temperature oxidation, and are optionally coated with catalyst [164]. Recently, Fareghi-Alamdari et al. synthesized a new dicationic ionic liquid (DCIL) based on dicyanamide anions and used as a protective ligand for boron nanoparticles [163]. Boron nanoparticles and DCIL-capped boron were produced using a ball mill technique with a size distribution of 50 100 nm. The results of FTIR, XPS, X-ray diffraction (XRD), energy-dispersive X-ray (EDX), and ζ-potential measurements indicated that boron nanoparticle surfaces were successfully capped by the DCIL layer, which protects the boron surfaces against air oxidation. Thermogravimetric analysis (TGA) measurements showed that the nanoparticle oxidation occurred after DCIL thermal decomposition and allows for boron ignition at higher temperatures. These results suggest that this ionic liquid binds to boron and shields the boron surfaces from oxidation when exposed to air. Despite the desirable properties of boron, the incorporation of boron nanoparticles into the propellants is not widely practiced, because of its oxidation product B2O3. The boron particles combustion occurs in two consecutive steps: oxide layer is removed in the first step, whereas the bare boron is burnt in the second step. The preexisting oxide layer (BO)n on the boron surface plays a significant role in the ignition and combustion processes; more precisely, it delays the ignition process.

2.3.8

Conclusion and Future Perspective

In recent years, new materials for energy are attracting immense research interest and attention due to an increasing economic and social progress and concern regarding the sustainable development of energy. These are nearly depicted to a large efficiency storage and use of power. To accomplish these challenges, there is high demand for developing novel energy systems which have highly developed energy and power density with better-quality integrated working and must also be environmentally friendly, capable of storing large amount of energy and convenient to use [11]. At present, the desire for advanced systems with best working, or innovative methods of production and/or processing of current emerging compounds for the hydrogen storage is increasing. There is a requisite to design a novel reaction system to realise these objectives with the limitations observed in conventional systems [66]. Catalytic activation of low reactivity materials is a useful and productive manner which lowers the activation energy to improve the hydrogen adsorption rate (which is related

Boron

83

equally to solid–gas and solid–liquid interactions). It is challenging to design and develop a reaction system capable of achieving the DOE’s hydrogen storage targets. There is a prerequisite of further research on Li-ion battery and SCs to improve performance efficiency and practical application. Moreover, it is very important to have the compatibility of electrolytes, separators, and anode and cathode materials. To produce better quality energy storage systems various advances like nanostructuring, configuration design, pore-structure control, nano-/microcombination, hybridization, surface modification, composition optimization, and new material plans have been developed. These power storage systems with good safety, long cycle life, good reliability, and high capacity will certainly enhances the performance of devices created based on these strategies and facilitate their wide application. Electrode systems with best working for synthesizing SCs is achieved by designing and constructing new substances with inclined structures and compositions on the basis of the interaction between electrodes and electrolyte ions. Porous structure is associated with ion transfer and charge accommodation, and it has been established that electrochemical performance of electrodes increases with hierarchical pores and considerable high surface areas. Due to their high energy density, symmetric capacitors comprised of a SC electrode and a batterytype electrode is getting increasing research interest. Applications of nanostructured materials in Faradaic battery-type electrode due to their increased ionic transport capability are attributed to their minute dimensions. Optimization of the composition and structure of nanomaterials can additionally enhance the efficiency of the asymmetric capacitors. The pseudocapacitive effect of doped nitrogen as well as boron can improve the specific capacitance. Further, materials with a broader member of mesopores have an improved rate response, whereas an extra charge is accumulated by extremely microporous models [6]. The consequences are beneficial in understanding the impact of capacitive behaviors on carbon matrix doped with heteroatoms and also in the design of various sorts of SCs for application in elevated-energy or elevated-energy demanded circumstances in prospect. In addition, boron-rich semiconductors are found to be the best components for converting thermoelectric energy. The systematic scientific and technical advancement of the components give increased efficiencies of 25%, or even more. Small scale electrical power stations also have similar efficiencies but thermoelectric devices as described are considerably more reliable. Therefore, wide range ecological and economical functions of converting thermoelectric energy using boron-rich solids are made available. Currently, boron particle and boron-based fuel-rich propellants are well studied and reports are widespread. However, there are limited reports on the studies of primary and secondary combustion products. It is complicated to understand and study the composition of primary combustion products due to the dissimilar constituents which may react and influence one another. Hence, there is a need to design various strategies to study and understand the primary combustion products [165].

Acknowledgment On Sabbatical as UGC-Raman Fellow sponsored by University Grants Commission (UGC), Govt. of India.

References [1] Titirici M-M, Antonietti M. Chemistry and materials options of sustainable carbon materials made by hydrothermal carbonization. Chem Soc Rev 2010;39:103–16. [2] Lai XY, Halpert JE, Wang D. Recent advances in micro/nano-structured hollow spheres for energy application: from simple to complex systems. Energy Environ Sci 2012;5:5604–18. [3] Deng J, Li MM, Wang Y. Biomass-derived carbon: synthesis and applications in energy storage and conversion. Green Chem 2016; http://dx.doi.org/10.1039/c6gc01172a. [4] Retnamma R, Novais AQ, Rangel CM. Kinetics of hydrolysis of sodium borohydride for hydrogen production in fuel cell applications: a review. Int Hydrogen Energy 2011;36:9772–90. [5] Abu-Hamed T, Karni J, Epstein M. The use of boron for thermochemical storage and distribution of solar energy. Sol Energy 2007;81:93–101. [6] Guo HL, Gao QM. Boron and nitrogen co-doped porous carbon and its enhanced properties as supercapacitor. J Power Sources 2009;186:551–6. [7] Gong ZL, Yang Y. Recent advances in the research of polyanion-type cathode materials for Li-ion batteries. Energy Environ Sci 2011;4:3223–42. [8] Li M, Li YF, Zhou Z, Shen PW, Chen ZF. Ca-coated boron fullerenes and nanotubes as superior hydrogen storage materials. Nano Lett 2009;9:1944–8. [9] Züttel A. Hydrogen storage methods. Naturwissenschaften 2004;91:157–72. [10] Schlapbach L, Zuttel A. Hydrogen-storage materials for mobile applications. Nature 2001;414:353–8. [11] Liu C, Li F, Ma LP, Cheng HM. Advanced materials for energy storage. Adv Mater 2010;22:E28–62. [12] Partnership plan. Freedomcar and fuel partnership. Available from: http://www1.eere.energy.gov/vehiclesandfuels/pdfs/program/fc_fuel_partnership_plan.pdf; 2006. [13] Aardahl CL, Rassat SD. Overview of systems considerations for on-board chemical hydrogen storage. Int J Hydrogen Energy 2009;34:6676–83. [14] Wang P, Kang XD. Hydrogen-rich boron-containing materials for hydrogen storage. Dalton Trans 2008;40:5400–13. [15] Umegaki T, Yan JM, Xhang XB, et al. Boron- and nitrogen-based hydrogen storage materials. Int J Hydrogen Energy 2009;34:2303–11. [16] Hamilton CW, Baker RT, Staubitz A, Manners I. B–N compounds for chemical hydrogen storage. Chem Soc Rev 2009;38:279–93. [17] Orimo S, Nakamori Y, Eliseo JR, Zuttel A, Jensen CM. Complex hydrides for hydrogen storage. Chem Rev 2007;107:4111–32. [18] Schlesinger HI, Brown HC, Finholt AE, et al. Sodium borohydride, its hydrolysis and its use as a reducing agent and in the generation of hydrogen. J Am Chem Soc 1953;75:215–9. [19] Anthony G, Dariusz S, Simona B, Aline A. New insights into the mechanism of H2 generation through NaBH4 hydrolysis on Co-based nanocatalysts studied by differential reaction calorimetry. Int J Hydrogen Energy 2009;34:1185–99. [20] Schlesinger HI, Brown HC, Abraham B, et al. New developments in the chemistry of diborane and the borohydrides. General summary. J Am Chem Soc 1953;75:186–90. [21] Stock A, Stterlin W, Kurzen F. Boron hydrides. XX. Potassium diborane K2(B2H6). Z Anorg Allgem Chem 1935;225:225–42. [22] Schlesinger HI, Sanderson RT, Burg AB. Metallo borohydrides. I. Aluminum borohydride. J Am Chem Soc 1940;62:3421–5. [23] Schlesinger HI, Brown HC. Metallo borohydrides III. Lithium borohydride. J Am Chem Soc 1940;62:3429–35.

84

Boron

[24] Wiberg E, Bauer R. Magnesium borohydride, Mg(BH4)2. Zeit Naturforsch 1950;5b:397–8. [25] Bogdanovic B, Schwickardi M. Ti-doped alkali metal aluminium hydrides as potential novel reversible hydrogen storage materials. J Alloys Compd 1997;253:1–9. [26] (a) Mohring, R. Luzader, R., A sodium borohydride on-board hydrogen generator for powering fuel cell and internal combustion engine vehicles, SAE Technical Paper 2001-01-2529; 2001; (b) Amendola SC, Sharp-Goldman SL, Saleem MJ, et al. An ultrasafe hydrogen generator: aqueous, alkaline borohydride solutions and Ru catalyst. J Power Sources 2000;85:186–9. [27] Muir SM, Yao XD. Progress in sodium borohydride as a hydrogen storage material: development of hydrolysis catalysts and reaction systems. Int J Hydrogen Energy 2011;36:5983–97. [28] Kaufman CM, Sen B. Hydrogen generation by hydrolysis of sodium tetrahydroborate: effects of acids and transition metals and their salts. J Chem Soc Dalton Trans 1985;2:307–13. [29] US Department of Energy Hydrogen Program. Go/no-go recommendation for sodium borohydride for on-board vehicular hydrogen storage. NREL/MP-150-42220. Available from: http://www1.eere.energy.gov/hydrogenandfuelcells/pdfs/42220.pdf; 2007. [30] Zhang Q, Smith GM, Wu Y. Catalytic hyrolysis of sodium borohydride in an integrated reactor for hydrogen generation. Int J Hydrogen Energy 2007;32:4731–5. [31] Ingersoll JC, Mani N, Thenmozhiyal JC, Muthaiah A. Catalytic hydrolysis of sodium borohydride by a novel nickel–cobalt–boride catalyst. J Power Sources 2007;173:450–7. [32] Xu D, Dai P, Liu X, Cao C, Guo Q. Carbon supported cobalt catalyst for hydrogen generation from alkaline sodium borohydride solution. J Power Sources 2008;182:616–20. [33] Minkina VG, Shabunya SI, Kalinin VI, Martynenko VV, Smirnova AL. Long-term stability of sodium borohydrides for hydrogen generation. Int J Hydrogen Energy 2008;33:5629–35. [34] Amendola SC, Sharp-Goldman SL, Janjua MS, et al. A safe, portable, hydrogen gas generator using aqueous borohydride solution and Ru catalyst. Int J Hydrogen Energy 2000;25:969–75. [35] Guella G, Zanchetta C, Patton B, Miotello A. New insights on the mechanism of palladium-catalyzed hydrolysis of sodium borohydride from 11B NMR measurements. J Phys Chem B 2006;110:17024–33. [36] Patel N, Patton B, Zanchetta C, et al. Pd–C powder and thin film catalysts for hydrogen production by hydrolysis of sodium borohydride. Int J Hydrogen Energy 2008;33:287–92. [37] Bai Y, Wu C, Wu F, Yi B. Carbon-supported platinum catalysts for on-site hydrogen generation from NaBH4 solution. Mater Lett 2006;60:2236–9. [38] Kojima Y, Suzuki K, Fukumoto K, et al. Hydrogen generation using sodium borohydride solution and metal catalyst coated on metal oxide. Int J Hydrogen Energy 2002;27:1029–34. [39] Suda S, Sun YM, Liu BH, et al. Catalytic generation of hydrogen by applying fluorinatedmetal hydrides as catalysts. Appl Phys A: Mater Sci Process 2007;72:209–12. [40] Jeong SU, Kim RK, Cho EA, et al. A study on hydrogen generation from NaBH4 solution using the high-performance Co–B catalyst. J Power Sources 2005;144:129–34. [41] Dong H, Yang H, Ai X, Cha C. Hydrogen production from catalytic hydrolysis of sodium borohydride solution using nickel boride catalyst. Int J Hydrogen Energy 2003;28:1095–100. [42] Moussa G, Moury R, Demirci UB, S¸ener T, Miele P. Boron-based hydrides for chemical hydrogen storage. Int J Energy Res 2013;37:825–42. [43] Li H, Yang Q, Chen X, Shore SG. Ammonia borane, past as prolog. J Organomet Chem 2014;751:60–6. [44] Stephens FH, Pons V, Baker RT. Ammonia-borane: the hydrogen source par excellence? Dalton Trans 2007;25:2613–26. [45] Shrestha RP, Diyabalanage HVK, Semelsberger TA, Ott KC, Burrell AK. Catalytic dehydrogenation of ammonia borane in non-aqueous medium. Int J Hydrogen Energy 2009;34:2616–21. [46] Yoon CW, Carroll PJ, Sneddon LG. Ammonia triborane: a new synthesis, structural determinations, and hydrolytic hydrogen-release properties. J Am Chem Soc 2009;131:855–64. [47] Hu MG, Geanangel RA, Wendlandt WW. The thermal decomposition of ammonia borane. Thermochim Acta 1978;23:249–55. [48] Summerscales OT, Gordon JC. Regeneration of ammonia borane from spent fuel. Dalton Trans 2013;42:10075–84. [49] Wang P. Solid-state thermolysis of ammonia borane and related materials for high-capacity hydrogen storage. Dalton Trans 2012;41:4296–302. [50] Shaw WJ, Linehan JC, Szymczak N, et al. In situ multinuclear NMR spectroscopic studies of the thermal decomposition of ammonia borane in solution. Angew Chem Int Ed 2008;47:7493–6. [51] Kostka JF, Schellenberg R, Baitalow F, Smolinka T, Mertens F. Concentration-dependent dehydrogenation of ammonia-borane/triglyme mixtures. Eur J Inorg Chem 2012;2012:49–54. [52] Glüer A, Förster M, Celinski VR, et al. Highly active iron catalyst for ammonia borane dehydrocoupling at room temperature. ACS Catal 2015;5:7214–7. [53] Manon P, Philippe M, Umit BD. Mechanistic insights of metal acetylacetonate-aided dehydrocoupling of liquid-state ammonia borane NH3BH3. Adv Energy Res 2016;4:177–87. [54] Duman S, Özkar S. Hydrogen generation from the dehydrogenation of ammonia-borane in the presence of ruthenium (III) acetylacetonate forming a homogeneous catalyst. Int J Hydrogen Energy 2013;38:180–7. [55] Kalviri HA, Gärtner F, Ye G, Korobkov I, Baker RT. Probing the second dehydrogenation step in ammonia-borane dehydrocoupling: characterization and reactivity of the key intermediate, B-(cyclotriborazanyl)amine-borane. Chem Sci 2015;6:618–24. [56] Toche F, Chiriac R, Demirci UB, Miele P. Ammonia borane thermolytic decomposition in the presence of metal (II) chlorides. Int J Hydrogen Energy 2012;37:6749–55. [57] Lian G, Zhang X, Zhang SJ, et al. Controlled fabrication of ultrathin-shell BN hollow spheres with excellent performance in hydrogen storage and wastewater treatment. Energy Environ Sci 2012;5:7072–80. [58] Lei W, Zhang H, Wu Y, et al. Oxygen-doped boron nitride nanosheets with excellent performance in hydrogen storage. Nano Energy 2014;6:219–24. [59] Zuttel A. Materials for hydrogen storage. Mater Today 2003;6:24–33. [60] Jhi SH, Kwon YK. Hydrogen adsorption on boron nitride nanotubes: a path to room-temperarue hydrogen storage. Phys Rev B 2004;69:245407. [61] Ma RZ, Bando Y, Sato T, et al. Synthesis of boron nitride nanofibers and measurement of their hydrogen uptake capacity. Appl Phys Lett 2002;81:5225. [62] Ma RZ, Bando Y, Zhu H, et al. Hydrogen uptake in boron nitride nanotubes at room temperature. J Am Chem Soc 2002;124:7672–3. [63] Tang CC, Bando Y, Ding X, Qi S, Golberg D. Catalyzed collapse and enhanced hydrogen storage of BN nanotubes. J Am Chem Soc 2002;124:14550–1. [64] Chen X, Gao XP, Zhang H, et al. Preparation and electrochemical hydrogen storage of boron nitride nanotubes. J Phys Chem B 2005;109:11525–9. [65] Jhi SH. Activated boron nitride nanotubes: a potential material for room-temperature hydrogen storage. Phys Rev B 2006;74:155424. [66] Yang J, Sudik A, Wolverton C, Siegel DJ. High capacity hydrogen storage materials: attributes for automotive applications and techniques for materials discovery. Chem Soc Rev 2010;39:656–75. [67] Larcher D, Tarascon JM. Towards greener and more sustainable batteries for electrical energy storage. Nat Chem 2015;7:19–29. [68] Choi NS, Chen Z, Freunberger SA, et al. Challenges facing lithium batteries and electrical double-layer capacitors. Angew Chem Int Ed 2012;51:9994–10024. [69] Legagneur V, An Y, Mosbah A, et al. LiMBO3 (M ¼ Mn, Fe, Co): synthesis, crystal structure and lithium deinsertion/insertion properties. Solid State Ion 2001;139:37–46. [70] Dong YZ, Zhao YM, Fu P, Zhou H, Hou XM. Phase relations of Li2O–FeO–B2O3 ternary system and electrochemical properties of LiFeBO3 compound. J Alloys Compd 2008;461:585–90. [71] Dong YZ, Zhao YM, Shi ZD, et al. The structure and electrochemical performance of LiFeBO3 as a novel Li-battery cathode material. Electrochim Acta 2008;53:2339–45. [72] Yamada A, Iwane N, Harada Y, et al. Lithium iron borates as high-capacity battery electrodes. Adv Mater 2010;22:3583–7. [73] Chen L, Zhao Y, An X, et al. Sturcture and electrochemical properties of LiMnBO3 as a new cathode materials for lithium-ion batteries. J Alloys Compd 2010;494:415–9.

Boron [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123]

85

Kim JC, Moore CJ, Kang B, et al. Synthesis and electrochemical properties of monoclinic LiMnBO3 as a Li intercalation material. J Electrochem Soc 2011;158:A309–15. Xu B, Qian DN, Wang ZY, Meng YS. Recent progress in cathode materials research for advanced lithium ion batteries. Mater Sci Eng R 2012;73:51–65. Nitta N, Yushin G. High-capacity anode materials for lithium-ion batteries: choice of elements and structures for active particles. Part Part Syst Charact 2013;31:317–36. Endo M, Kim C, Nishimura K, Fujino T, Miyashita K. Recent development of carbon materials for Li ion batteries. Carbon 2000;38:183–97. Fujimoto H, Mabuchi A, Tokumitsu K, Kasuh T, Akuzawa N. Effect of crystallite size on the chemical compositions of the stage 1 alkali metal-graphite interaction compound. Carbon 1994;32:193–8. Tasumi K, Iwashita N, Sakaebe H, et al. The influence of the graphitic structure on the electrochemical characteristics for the anode of secondary lithium batteries. J Electrochem Soc 1995;142:716–20. Dahn JR, Sleigh AK, Shi H, editors. Lithium batteries, new materials and perspectives Amsterdam: Elsevier; 1993. Tatsumi K, Zaghib K, Sawada Y, Abe H, Ohsaki T. Anode performance of vapor-grown arbonfibers in secondary lithium-ion batteries. J Electrochem Soc 1995;142:1090–6. Way BM, Dahn JR. The effect of boron substitution in carbon on the intercalation of lithium in Lix(BzC1-z)6. J Electrochem Soc 1994;141:907–12. Nakajima T, Koh K, Takashima M. Electrochemical behavior of carbon alloy CxN prepared by CVD using a nickel catalyst. Electrochim Acta 1998;43:883–91. Choi NS, Yew KH, Kim H, Kim SS, Choi WU. Surface layer formed on silicon thin-film electrode in lithium bis(oxalato) borate-based electrolyte. J Power Sources 2007;172:404–9. Xu K, Zhang SS, Jow TR. Electrochemical impedance study of graphite/electrolyte interface formed in LiBOB/PC electrolyte. J Power Sources 2005;143:197–202. Xu K, Zhang SS, Lee U, Allen JL, Jow TR. LiBOB: is it an alternative salt for lithium ion chemistry? J Power Sources 2005;146:79–85. Dalavi S, Xu M, Knight B, Lucht BL. Effect of added LiBOB on high voltage (LiNi0.5Mn1.5O4) spinel cathodes. Electrochem Solid-State Lett 2012;15:A28–31. Aravindan V, Cheah YL, Ling WC, Madhavi S. Effect of LiBOB additive on the electrochemical performance of LiCoPO4. J Electrochem Soc 2012;159:A1435–9. Ha SY, Han JG, Song YM, et al. Using a lithium bis(oxalato) borate additive to improve electrochemical performance of high-voltage spinel LiNi0.5Mn1.5O4 cathodes at 601C. Electrochim Acta 2013;104:170–7. Pieczonka NPW, Yang L, Balogh MP, et al. Impact of lithium bis(oxalate)borate electrolyte additive on the performance of high-voltage spinel/graphite Li-ion batteries. J Phys Chem C 2013;117:22603–12. Zhang SS. Electrochemical study of the formation of a solid electrolyte interface on graphite in a LiBC2O4F2-based electrolyte. J Power Sources 2007;163:713–8. Li S, Zhao W, Cui X, et al. An improved method for synthesis of lithium difluoro(oxalato)borate and effects of sulfolane on the electrochemical performances of lithiumion batteries. Electrochim Acta 2013;91:282–92. Li J, Xing L, Zhang R, et al. Tris (trimethylsilyl) borate as an electrolyte additive for improving interfacial stability of high voltage layered lithium-rich oxide cathode/ carbonate-based electrolyte. J Power Sources 2015;285:360–6. Rong HB, Xu M, Xie B, et al. Tris (trimethylsilyl) borate (TMSB) as a cathode surface film forming additive for 5 V Li/LiNi0.5Mn1.5O4 Li-ion cells. Electrochim Acta 2014;147:31–9. Liu Y, Tan L, Li L. Tris(trimethylsilyl) borate as an electrolyte additive to improve the cyclability of LiMn2O4 cathode for lithium-ion battery. J Power Sources 2013;221:90–6. Liao XL, Sun PY, Xu MQ, et al. Application of tris(trimethylsilyl)borate to suppress self-discharge of layered nickel cobalt manganese oxide for high energy battery. Appl Energy 2016;175:505–11. Zuo X, Fan C, Liu J, et al. Effect of tris(trimethylsilyl)borate on the high voltage capacity retention of LiNi0.5Co0.2Mn0.3O2/graphite cells. J Power Sources 2013;229:308–12. Wang HL, Li Z, Mitlin D. Tailoring biomass-derived carbon nanoarchitrrctures for high-performance supercapacitors. ChemElectroChem 2014;1:332–7. Huang Y, Peng L, Liu Y, et al. Biobasednano porous active carbon fibers for high-performance supercapacitors. ACS Appl Mater Interfaces 2016;8:15205–15. Li J, Liu K, Gao X, et al. Oxygen- and nitrogen-enriched 3D porous carbon for supercapacitors of high volumetric capacity. ACS Appl Mater Interfaces 2015;7:24622–8. Zhu C, Liu T, Qian F, et al. Supercapacitors based on three-dimensional hierarchical graphene aerogels with periodic macropores. Nano Lett 2016;16:3448–56. Shiraishi S, Kibe M, Yokoyama T, et al. Electric double layer capacitance of multi-walled carbon nanotubes and B-doping effect. Appl Phys A: Mater Sci Process 2006;82:585–91. Wu X, Radovic LR. Ab initio molecular orbital study on the electronic structures and reactivity of boron-substituted carbon. J Phys Chem A 2004;108:9180–7. Lowell CE. Solid solution of boron in graphite. J Am Ceram Soc 1967;50:142–4. Morita T, Takami N. Characterization of oxidized boron-doped carbon fiber anodes for Li-ion batteries by analysis of heat of immersion. Electrochim Acta 2004;49:2591–9. Vishwakarma PN, Subramanyam SV. Hopping conduction in boron doped amorphous carbon films. J Appl Phys 2006;100:113702. Radovic LR, Karra M, Skokova K, Thrower PA. The role of substitutional boron in carbon oxidation. Carbon 1998;36:1841–54. Zhong DH, Sano H, Uchiyama Y, Kobayashi K. Effect of low-level boron doping on oxidation behavior of polyimide-derived carbon films. Carbon 2000;38:1199–206. Wang D–W, Li F, Chen Z–G, Lu GQ, Cheng H–M. Synthesis and electrochemical property of boron-doped mesoporous carbon in supercapacitor. Chem Mater 2008;20:7195–200. Wu ZS, Winter A, Chen L, et al. Three-dimensional nitrogen and boron co-doped graphene for high-performance all-solid-state supercapacitors. Adv Mater 2012;24:5130–5. Jeong HM, Lee JW, Shin WH, et al. Nitrogen-doped graphene for high-performance ultracapacitors and the importance of nitrogen-doped sites at basal planes. Nano Lett 2011;11:2472–7. Wu ZS, Ren WC, Xu L, Li F, Cheng HM. Doped graphene sheets as anode materials with superhigh rate and large capacity for lithium ion batteries. ACS Nano 2011;5:5463–71. Wang S, Iyyamperumal E, Roy A, et al. Vertically aligned BCN nanotubes as efficient metal-free electrocatalysts for the oxygen reduction reaction: a synergetic effect by co-doping with boron and nitrogen. Angew Chem Int Ed 2011;50:11756–60. Qiu YC, Zhang XF, Yang SH. High performance supercapacitors based on highly conductive nitrogen-doped graphene sheets. Phys Chem Chem Phys 2011;13:12554–8. Wang DW, Li F, Zhao J, et al. Fabrication of graphene/polyaniline composite paper via in situ anodic electropolymerization for high-performance flexible electrode. ACS Nano 2009;3:1745–52. Werheit H. Boron-rich solids: a chance for high-efficiency high-temperature thermoelectric energy conversion. Mater Sci Eng 1995;B29:228–32. Werheit H. On the electronic transport properties of boron carbide. In: Freer R, editor. The physics and chemistry of carbides, nitrides and borides. Dordrecht: Kluwer; 1990, p. 677. Werheit H. On the electronic properties of icosahedral boron-rich solids. In: Gogotsi YG, Andrievski RA, editors. Materials science of carbides, nitrides and borides. vol. 68. Dordrecht: Springer; 1999. p. 65–86. Kuhlmann U, Werheit H. On the microstructure of boron carbide. Solid State Commun 1992;83:849–52. Kuhlmann U, Werheit H, Sehwetz KA. Distribution of carbon atoms on the boron cabide structure elements. J Alloys Compd 1992;189:249–58. Bairamashvili IA, Kekelidze LI, Golikova OA, Orlov VM. The preparation of a-AlB12 and AlMgB14 samples and an investigation of their electrothermal properties. J. Less Common Met 1979;67:461–4. Werheit H, Groot K. Metal-insulator transition in boron carbide. Phys Status Solidi (b) 1980;97:229–38. Li J, Goto T, Hiral T. Microstructure and thermoelectric properties of B4C–SiBn–Si c omposites prepared by arc melting. J Ceram Soc Jpn 1998;106:194–7.

86

Boron

[124] Imai Y, Mukaida M, Ueda M, Watanabe A. Screening of the possible boron-based n-type thermoelectric conversion materials on the basis of the calculated densities of states of metal borides and doped b-boron. Intermetallics 2001;9:721–34. [125] Takeda M, Fukuda T, Domingo F, Miura T. Thermoelectric properties of some metal borides. J Solid State Chem 2004;177:471–5. [126] Andrew J, Jelley N.. Energy science 2007 Oxford: Oxford University Press; 2007. ISBN 978-0-19-928112-1. [127] Wikipedia. Available from: https://en.wikipedia.org/wiki/Control_rod. [accessed 20.11.16]. [128] Thevenot F. Boron carbide – a comprehensive review. J Eur Ceram Soc 1990;6:205–25. [129] Matkovich VI. Boron and refractory borides. Heidelberg/New York, NY: Springer-Verlag, Berlin; 1977. [130] Bose DK, Nair KU, Gupta CK. Production of high purity boron carbide. High Temp Mater Proc 1986;7:133–40. [131] Subramanian C, Suri AK. Development of boron based neutron absorber materials. Met Mater Process 2004;16:39–52. [132] Martin V, Wood B, Werner-Zwanziger U, Zwanziger JW. Structural aspects of the photoelastic response in lead borate glasses. J Non-Cryst Solids 2011;357:2120–5. [133] Rada M, Rada S, Culea E. Structural properties of the tungsten lead borate glasses before and after laser irradiation. J Non-Cryst Solids 2011;357:2024–8. [134] Pisarska J, Lisiecki R, Ryba-Romanowski W, et al. Glass preparation and temperature-induced crystallization in multicomponent B2O3–PbX2–PbO–Al2O3–WO3–Dy2O3 (X ¼F, Cl, Br) system. J Non-Cryst Solids 2011;357:1228–31. [135] ElBatal HA, Abdelghany AM, ElBatal FH, EzzElDin FM. Gamma rays interactions with WO3-doped lead borate glasses. Mater Chem Phys 2012;134:542–8. [136] Tanabe S, Kang J, Hanada T, Soga N. Yellow/blue luminescences of Dy3 þ -doped borate glasses and their anomalous temperature variations. J Non-Cryst Solids 1998;239:170–5. [137] Jayasankar CK, Devi AR. Optical properties of Tm3 þ ions in lithium borate glasses. Opt Mater 1996;6:185–201. [138] Babu P, Jayasankar CK. Spectroscopic properties of Dy3 þ ions in lithium borate and lithium fluoroborate glasses. Opt Mater 2000;15:65–79. [139] Ratnakaram YC, Naidu DT, Kumar AV, Rao JL. Characterization of Tm3 þ doped mixed alkali borate glasses-spectroscopic invertigations. J Phys Chem Solids 2003;64:2487–95. [140] Ratnakaram YC, Naidu DT, Kumar AV, Rao JL. The effect of mixed alkalis on the absorption and fluorescence properties of Ho3 þ ions in borate glasses. J Phys Condens Matter 2004;16:3779. [141] Pisarska J, Pisarski WA, Dominiak-Dzik G, Ryba-Romanowski W. Spectroscopic investigations of Nd3 þ ions in B2O3–PbO–Al2O3–WO3 glasses. J Mol Struct 2006;792–793:201–6. [142] Pisarska J, Kos A, Sołtys M, Z˙ur L, Pisarski WA. Energy transfer from Tb3 þ to Eu3 þ in lead borate glass. J Non-Cryst Solids 2014;388:1–5. [143] Nayak PK, Garcia-Belmonte G, Kahn A, Bisquert J, Cahen D. Photovoltaic efficiency limits and material disorder. Energy Environ Sci 2012;5:6022–39. [144] Hod I, Zaban A. Materials and interfaces in quantum dot sensitized solar cells: challenges, advances and prospects. Langmuir 2014;30:7264–73. [145] Narayanan KL, Goetzgberger O, Khan A, Kojima N, Yamaguchi M. Boron ion implantation effect in C60 films. Sol Energy Mater Sol Cells 2001;65:29–35. [146] Tian XM, Rusop M, Hayashi Y, et al. A photovoltaic cell from p-type boron-doped amorphous carbon film. Sol Energy Mater Sol Cells 2003;77:105–12. [147] Ma ZQ, Liu BX. Boron-doped diamond-like amorphous carbon as photovoltaic films in solar cell. Sol Energy Mater Sol Cells 2001;69:339–44. [148] Wang JW, Liu Y, Zhang XW, Mi ZT, Wang L. Facile preparation of hydrocarbon fuel-soluble nano-catalyst and its novel application in catalytic combustion of JP-10. Catal Commun 2009;10:1518–22. [149] Holcomb RR, Holcomb AR. Composite inorganic and organic sols and colloids as combustion improvers for liquid fuels. Nashville: Demeter Systems LLC; 2006. p. 36. [150] Valentine JM, Sprague BN, Peter-Hoblyn JD. Fuel-soluble platinum-cerium-iron catalysts for combustion of diesel fuel and jet fuel. Patent Application No. US 200538371; 2005. p 12. [151] Slutsky VG, Tsyganov SA, Severin ES, Polenov LA. Synthesis of small-scale boron-rich nano-size particles. Propellants, Explos, Pyrotech 2005;30:303–9. [152] Risha GA, Boyer E, Evans B, Kuo KK, Malek R. Characterization of nano-sized particles for propulsion applications. Mater Res Soc Symp Proc 2003;800:243. [153] Kuo KK, Risha GA, Evans BJ, Boyer E. Potential usage of energetic nano-sized powders for combustion and rocket propulsion. Mater Res Soc Symp Proc 2003;800 (2003):3–14. [154] Young G, Sullivan K, Zachariah MR, Yu K. Combustion characteristics of boron nanoparticles. Combust Flame 2009;156:322–33. [155] Yeh CL, Kuo KK. Ignition and combustion of boron particles. Prog Energy Combust Sci 1996;22:511–41. [156] Dubois C, Brousseau P, Roy C, Lafleur P. Proceedings of the AIChE annual meeting. Austin, TX; 2004. p. 281E/1 281E/9. [157] Anderson JL, Ding R, Ellern A, Armstrong DW. Structure and properties of high stability geminaldicationic ionic liquids. J Am Chem Soc 2005;127:593–604. [158] Pickering AL, Mitterbauer C, Browning ND, Kauzlarich SM, Power PP. Room temperature synthesis of surface-functionalised boron nanoparticles. Chem Commun 2007;6:580–2. [159] Devener BV, Perez JPL, Anderson SL. Air-stable, unoxidized, hydrocarbon-dispersible boron nanoparticles. J Mater Res 2009;24:3462–4. [160] Perez JPL, McMahon BW, Anderson SL. Functionalization and passivation of boron nanoparticles with a hypergolic ionic liquid. J Propul Power 2013;29:489–95. [161] Perez JPL, McMahon BW, et al. Exploring the structure of nitrogen-rich ionic liquids and their binding to the surface of oxide-free boron nanoparticles. J Phys Chem C 2013;117:5693–707. [162] Perez JPL, Yu J, Sheppard AJ, et al. Binding of alkenes and ionic liquids to B–H-functionalized boron nanoparticles: creation of particles with controlled dispersibility and minimal surface oxidation. ACS Appl Mater Interfaces 2015;7:9991–10003. [163] Fareghi-Alamdari R, Ghorbani-Zamani F, Shekarriz M. Surface passivation of bare boron nanoparticles using new dicyanamide-based dicationic ionic liquid. Energy Fuels 2016;30:551–9. [164] Devener BV, Perez JPL, Jankovich J, Anderson SL. Oxide-free, catalyst-coated, fuel-soluble, air-stable boron nanopowder as combined combustion catalyst and high energy density fuel. Energy Fuels 2009;23:6111–20. [165] Liang D, Liu J, Xiao J, et al. Energy release properties of amorphous boron and boron-based propellant primary combustion products. Acta Astronaut 2015;112:182–91.

Further Reading Gong Z, Yang Y. Recent advances in the research of polyanion-type cathode materials for Li-ion batteries. Energy Environ Sci 2011;4:3223–42. Guo H, Gao Q. Boron and nitrogen co-doped porous carbon and its enhanced properties as supercapacitor. J Power Sources 2009;186:551–6. Jhi S-H. Activated boron nitride nanotubes: a potential material for room-temperature hydrogen storage. Phys Rev B 2006;74(4):155424. Liang D, Liu J, Xiao J, et al. Energy release properties of amorphous boron and boron-based propellant primary combustion products. Acta Astronaut 2015;112:182–91. Liao X, Sun P, Xu M, et al. Application of tris(trimethylsilyl)borate to suppress self-discharge of layered nickel cobalt manganese oxide for high energy battery. Appl Energy 2016;175:505–11. Liu BH, Li ZP. A review: hydrogen generation from borohydride hydrolysis reaction. J Power Sources 2009;187:527–34. Liu BH, Li ZP. Current status and progress of direct borohydride fuel cell technology development. J Power Sources 2009;187:291–7. Muir SS, Yao X. Progress in sodium borohydride as a hydrogen storage material: development of hydrolysis catalysts and reaction systems. Int J Hydrogen Energy 2011;36:5983–97. Perez M, Miele P, Demirci UB. Mechanistic insights of metal acetylacetonate-aided dehydrocoupling of liquid-state ammonia borane NH3BH3. Adv Energy Res 2016;4:177–87.

Boron

87

Retnamma R, Novais AQ, Rangel CM. Kinetics of hydrolysis of sodium borohydride for hydrogen production in fuel cell applications: a review. Int J Hydrogen Energy 2011;36:9772–90. Santos DMF, Sequeira CAC. Sodium borohydride as a fuel for the future. Renew Sustain Energy Rev 2011;15:3980–4001. Van Devener B, Perez JPL, Jankovich J, Anderson SL. Oxide-free, catalyst-coated, fuel-soluble, air-stable boron nanopowder as combined combustion catalyst and high energy density fuel. Energy Fuels 2009;23:6111–20. Xu B, Qian D, Wang Z, Meng YS. Recent progress in cathode materials research for advanced lithium ion batteries. Mater Sci Eng R 2012;73:51–65.

Relevant Websites http://www.at0086.com/Ludong-University/ Ludong University, Yantai, China. http://www.niu.edu/hosmane/ Northern Illinois University, DeKalb, United States. http://rcub.ac.in/schools/school-of-science/chemistry Rani Channamma University, Belagavi, India.

2.4 Thin Films Franco Gaspari, University of Ontario Institute of Technology, Oshawa, ON, Canada r 2018 Elsevier Inc. All rights reserved.

2.4.1 Introduction 2.4.2 Background 2.4.3 Background: Thin Film Growth Technologies 2.4.3.1 Chemical Vapor Deposition 2.4.3.2 Plasma Enhanced Chemical Vapor Deposition 2.4.3.2.1 Glow discharge 2.4.3.3 Physical Vapor Deposition 2.4.3.4 Atomic Layer Deposition 2.4.3.5 Non-Vacuum-Based Techniques 2.4.3.5.1 Liquid phase synthesis 2.4.4 Background: Characterization Techniques 2.4.4.1 Methods for Structure and Composition Analysis 2.4.4.2 Thin Layer Optical and Electronic Properties 2.4.4.2.1 Thermal conductivity 2.4.4.2.2 Electrical conductivity 2.4.4.2.3 Optical properties 2.4.5 Applications 2.4.5.1 Hydrogenated Amorphous Silicon 2.4.5.1.1 Structure and density of states of a-Si:H 2.4.5.2 Micro-Crystalline Silicon (mc-Si:H) 2.4.5.3 Gallium Arsenide 2.4.5.3.1 Disadvantages of GaAs 2.4.5.4 Compound Semiconductor Thin Films 2.4.5.5 Organic Thin films 2.4.6 Analysis and Assessment 2.4.7 Examples and Future Directions 2.4.7.1 Applications of Thin Films for Optical Coatings 2.4.7.2 Thin Film Photovoltaic and Solar Cells 2.4.7.3 Organic Light-Emitting Diodes 2.4.7.4 Organic Light-Emitting Transistor (OTFT) 2.4.7.5 Thin Films as Selective Energy Contacts for Solar Cells 2.4.7.6 Thin Films as Fuel Cell Components 2.4.7.7 Other Energy Storage Devices (Ultra/Super-Capacitors) 2.4.8 Conclusions References Further Reading Relevant Websites

Nomenclature Symbol/acronym a-Si:H Hydrogenated amorphous silicon AES Auger electron spectroscopy AFM Atomic force microscopy ALD Atomic layer deposition APCVD Atmospheric pressure chemical vapor deposition ARC Antireflecting coating CBD Chemical bath deposition CdSe Cadmium selenide CdTe Cadmium telluride CIGS Copper–indium–gallium–selenide CIS Copper–indium–selenide

88

CVD CZTS DB DBD DOS DSSC EU EELS EL EM ESC Ga GaAs GaP

89 90 91 92 93 94 96 97 98 98 98 98 99 100 100 100 102 102 103 104 105 106 106 108 109 109 109 110 111 111 112 112 112 113 113 116 116

Chemical vapor deposition Copper–zinc–tin–sulfide Dangling bond Dielectric barrier discharge Density of states Dye sensitized solar cell Urbach energy Electron energy loss spectroscopy Electroluminescence Electron microscopy Energy selective contacts Gallium Gallium arsenide Gallium phosphide

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00214-5

Thin Films

GaN GaSb Ge HRTEM HWE InAs InP InSb IPE ITO LED LEED LEEM LPCVD LPE LPS MBE MOCVD OLED OLET OTFT PECVD PEEM

2.4.1

Gallium nitride Gallium antimonide Germanium High-resolution transmission electron microscopy Hot wall epitaxy Indium arsenide Indium phosphide Indium antimonide Inverse photoemission Indium tin oxide Light-emitting diode Low energy electron diffraction Low energy electron microscopy Low pressure chemical vapor deposition Liquid phase epitaxy Liquid phase synthesis Molecular beam epitaxy Metal-organic chemical vapor deposition Organic light emitting diode Organic light emitting transistor Organic thin film transistor Plasma enhanced chemical vapor deposition Photoemission electron microscopy

PEN PEM PET PL PVD R□ RHEED RBS RF RFID SEM Si SIMS SOFC SPM SNO2 STEM TFT UHV VHF VPE XPS ZnO mc-Si

89

Polyethylene naphtalene Proton exchange membrane Polyethylene terephtalate Photoluminescence Physical vapor deposition Sheet or surface resistance Reflexion high energy electron diffraction Rutherford backscattering Radio-frequency Radio-frequency identification Scanning electron microscopy Silicon Secondary ion mass spectrometry Solid oxide fuel cell Scanning photoelectron (probe) microscopy Tin oxide Scanning transmission electron microscopy Thin film Transistor Ultra-high vacuum Very high frequency Vapor-phase epitaxy X-ray photoelectron spectroscopy Zinc oxide Micro-crystalline silicon

Introduction

The term “thin film” is obviously generic and must be defined within the context of materials research. The range of thicknesses of thin film layers may vary from the nanometer range to a few micrometers, which implies that thin film materials must be deposited as a thin layer on top of a substrate. The main characteristic of thin film materials is the dependence of their properties on the growth technologies. This category of materials has particular relevance when considering energy applications, which include, but are not limited to, electronic semiconductor devices, light emitting diodes (LEDs), optical coatings, thin-film solar cells, and thin-film batteries. Thin films are better understood within the context of materials science, which represents a broad multidisciplinary area of applied and theoretical research. Indeed, materials science aims at achieving not only an understanding of the properties of matter, but also at predicting the correlation between the preparation techniques and the design of materials with particular properties; this aspect is especially topical for the study of thin films. Indeed, a core premise in materials science is that properties and performance are the consequence of structure, and that structure is the consequence of the processes that brought it about, all broadly understood. Thin film structures are influenced dramatically not only by the type of process employed for their preparation, but also by the growth parameters within a particular process. Due to the nano- to microscale range of their thickness, small variations in growth conditions may lead to quite large differences in film properties. In this chapter, a background description of the state of the art is presented in Section 2.4.2, where the differences between thin films and other materials of greater thickness will be introduced. Thin films are grouped into two main categories, inorganic and organic thin films. Section 2.4.2 will examine the differences between the two categories and the implications on modern device applications. In particular, we will discuss the current trend of exploring inorganic thin films for applications in microelectronics and photovoltaics (PV), and how alternative organic materials can offer other advantages and flexibilities. Section 2.4.3 will continue the background overview by addressing the connection between growth process and structure and by discussing the most common and effective preparation techniques for thin films, their main characterization techniques, and their properties. In particular, most thin films require vacuum technology for proper growth, which has led to the development of several vacuum coating processes designed to better control growth dynamics and, ultimately, the film structure. Frey [1] listed the main characteristics of these vacuum processes as follows: they must be able to accommodate different coating materials, film properties must be reproducible, adjustment of the film properties can be accomplished by changing the coating parameters, and the coatings must have little or no contamination (high purity). Indeed, the adjustment of the film properties implies a good understanding of vacuum processes, as each technique can be uniquely manipulated to achieve the desired final product. On the other hand, non-vacuum-based techniques are also used for the preparation of thin films, and they will also be discussed in this chapter. It should be underlined that among the non-vacuum techniques, many alternative processes are used to grow not only inorganic thin films, but also especially organic materials. The relatively novel interest in these materials is

90

Thin Films

motivated by their superior application flexibility (e.g., they can be deposited on flexible substrates), the lower costs (most processes are at low or room temperature, materials are inexpensive, and, as mentioned, vacuum technology is not needed), and their opto-electronic properties, which can be easily adjusted through processing. Section 2.4.4 is related to the final portion of background information. Here, some of the major analytical techniques used to characterize thin films will be described, with a focus on their relevance in providing proper feedback for the growth process. In order to correlate the thin-film properties with the growth conditions, it is important that the appropriate characterization techniques be employed to uncover the various properties of thin film materials. The choice of technique is obviously dependent on the information required for the material; however, the size, ultimate use, and nature of the material require an in-depth understanding of the analytical data provided by the characterization. As mentioned above, the correlation of process parameters with film properties is crucial. It should be noted that by film property one usually refers to the response of the material to a particular external stimulus. The main properties are divided into mechanical (elastic modulus, strength, etc.), electrical (conductivity, carrier mobilities, dielectric constant, etc.), thermal (heat capacity, thermal conductivity, etc.), magnetic (magnetic resonance, etc.), optical (index of refraction, optical absorption, photoconductivity, etc.), and deteriorative, that is any property related to the chemical reactivity of the material [2]. As indicated above, one should always consider the implications of film thickness. If the film is too thin, surface processes will have a dominant effect on the results, while films that are too thick do not have the same opto-electronic responses. Furthermore, plenty of techniques can be used to obtain the overall picture of the material properties. Some of these techniques have been modified to be more accurate in the analysis of thin films and some novel techniques are being developed to continue improving the understanding of this class of materials. Ref. [3] provides a good exhaustive review of characterization techniques for thin film solar cells. Other discussions of analytical techniques can also be found in the suggested literature in this chapter. Section 2.4.5 will discuss the energy applications of thin film materials within the two main energy oriented areas, i.e., PV [4–6] and electronics [7–9]. It should be mentioned that these two categories do not represent the whole of thin-film-based energy devices. For instance, simple coatings for optical uses (such as antireflective coatings) [10,11] and structural uses (like hard coatings) [12,13] have a role in enhancing the energy output and performance of many devices. Furthermore, energy storage has also become a relevant area of thin film research. In this section, we will provide a list and description of the most relevant thin film materials for energy use. Sections 2.4.6 and 2.4.7 will describe how to assess the “energy role” of thin films and provide relevant examples of recent thinfilm energy applications. We will examine how more “traditional” thin films, like amorphous silicon, are still being examined for novel applications, and we will look at the “new” thin film materials, in particular organic, and analyze how novel devices and applications are linked to their opto-electronic characteristics. Future directions of thin film applications will be also discussed in Section 2.4.7. Finally, the conclusions of this chapter will be presented in Section 2.4.8.

2.4.2

Background

The first common application of thin films was in the fabrication of mirrors. A thin metallic film was deposited on the rear of a glass substrate to form a reflecting surface. The original material was silver (Ag). The method was invented in the 16th century in Venice and remained an industrial secret for a century. Nowadays, thin films are mostly associated with semiconductor technology. However, one needs to mention that they have had considerable impact also outside this category. In particular, they have been used in chemistry application to provide diffusion barriers, for sensors design, and for protection against corrosion. Indeed, simple structure-related coating applications are quite common, including “hard” layers on drill bits, adhesion and friction improvements, and hard disk coatings. Thin films are also known for their electrical properties, that is, they can behave as insulators or conductors according to material and/or material processing and are used in a variety of electronic devices. They have also provided the means for numerous applications within the field of optics, including antireflection coating on lenses or solar cells, reflection coatings for mirrors, coatings for decorations, Interference filters, waveguides, flat panel displays, CDs and DVDs, etc. Some of these optical applications will be discussed in more detail, as they relate also to energy uses. Finally, thin films have become a major component for a variety of energy storage devices, including solid oxide fuel cells (SOFCs) and ultra-capacitors. However, it is undeniable that the major impact of thin films for energy applications is found mostly in the PV field. The next major area of energy applications is that of microelectronics. It should be mentioned that, although thin film micro-electronics has been associated originally with inorganic semiconductors, current research is focusing on organic thin film transistors (OTFTs) and related applications. Indeed, it can be stated that thin film energy applications encompass all the relevant aspects of the area, including energy production (PV), energy distribution and handling (micro-electronics), and energy storage (fuel cells, ultra-capacitors). Some relevant examples will be given at the end of the chapter. Crystalline silicon has been the dominant material in microelectronics and PV applications. However, aside from conventional crystalline silicon based electronics, scientific research has been focusing on nanostructures for flexible electronics (preferably rollto-roll processable) and low-cost PV. Indeed, c-Si technology, both PV and microelectronic, has reached maturity, with c-Si-based PV cells approaching the theoretical limit of efficiency. The main focus at the moment is the reduction of the cost-efficiency ratio, and novel materials are needed to achieve this goal. Amorphous silicon (a-Si) and nanocrystalline silicon (nc-Si) have been

Thin Films

Table 1

91

Amorphous silicon commercially available products in 1989

Device

Product

Photovoltaic cell Photoreceptor Photoconductors, image sensors Heat control layer Thin-film FETs High-voltage thin-film transistors

Calculators, watches, battery chargers, etc. Electrophotography, LED printers, etc. Color sensors, light sensors, etc. Heat reflecting float glass, etc. Displays, televisions, logic circuits, etc. Printers, etc.

Source: From LeComber PG. J Non-Cryst Solids 1989;115:1.

extensively investigated and applied as channel materials for thin film transistors (TFTs) to be integrated on flexible, low temperature substrates such as polyethylene terephtalate (PET), polyethylene naphtalene (PEN), polyimide (e.g., Kapton), and polycarbonate (e.g., Lexan). New transparent conductive oxide materials based on ZnO and SnO2 and mixed ternary and quaternary oxides have been developed and deposited as thin film (with deposition techniques spanning from spray pyrolysis to inkjet printing) for ITO replacement in touch screen displays. Also carbon nanotubes (CNTs) and graphene have been used for the same purpose and as gate electrode and channel materials for TFTs. Conductive inks made of dispersed metallic, carbon or conductive polymer nanoparticles have been formulated and deposited on low-cost substrate (i.e. paper) for radio-frequency identification (RFID) capacitive coupled antennas. The exploration of other forms of silicon, possibly less expensive to obtain and more flexible to process, has led to the development of thin-film silicon technology, which in turn can be divided into amorphous silicon and microcrystalline silicon [4]. Already in the 1980s amorphous silicon had been employed in a number of applications (see Table 1). Both hydrogenated amorphous silicon (a-Si:H) and microcrystalline silicon (mc-Si) are deposited using physical vapor deposition (PVD), chemical vapor deposition (CVD) and plasma enhanced chemical vapor deposition (PECVD), all vacuum based technologies. Other thin-film materials, including inorganic materials, have emerged as possible alternatives to silicon. Thin film solar cells based on gallium arsenide (GaAs), copper indium selenide (CIS), and indium phosphide (InP), to mention some of the more relevant, are currently an active area of investigation. These materials are also deposited with the same vacuum-based technologies listed above. On the other hand, solar cells based on cadmium telluride (CdTe), copper–indium–gallium–selenide (CIGS), dye sensitized solar cells (DSSC), and organic solar cells are also being considered as a possible alternative. These materials are usually obtained by simple techniques such as electrodeposition, chemical bath deposition (CBD), and screen printing. The low temperatures needed for the process (o5001C) and the fact that high vacuum is not required, makes these new technology very desirable in terms of processing costs [14]. In order to have a proper perspective of thin films and their applications, it is necessary to underline the importance of film growth technology and characterization techniques. The background overview will continue in the next two sections. In particular, Section 2.4.3 will illustrate in more detail the most common techniques used for the preparation of thin film materials, while Section 2.4.4 will look at the most relevant characterization techniques and their role in thin film optimization.

2.4.3

Background: Thin Film Growth Technologies

There are several publications that describe in great detail the technologies employed for the growth of thin films [3,4,15–17]. Indeed, there exist a large variety of options when considering growth techniques for thin films. However, one can group these techniques in two general categories: the bottom-up approach and the top-down approach. The former includes the miniaturization of materials’ components (up to atomic level) with further self-assembly process leading to the formation of nanoparticles; typical examples are quantum dot formation during epitaxial growth and formation of nanoparticles from colloidal dispersion. Top-down approaches use larger (macroscopic) initial structures, which can be externally controlled in the processing of nanostructures. For instance, milling is a typical top-down method for making colloidal dispersions of nanoparticles. Sometimes, the two synthetic approaches can be used in conjunction and it is hard to distinguish between the two methodologies. Lithography in semiconductor industry, for example, may be considered a hybrid approach, since the growth of thin films is bottom-up whereas etching is top-down. Within these two approaches, one must distinguish among three general categories for the synthesis of thin film and nano-structured materials: The first category is represented by gas phase processes, which in turn include vapor deposition (physical and chemical), flame pyrolysis, high temperature evaporation and plasma synthesis; the second category includes liquid phase methods, where chemical reactions processes in solvents are used in order to achieve the formation of colloids or aerosols. Electrodeposition, sol–gel and hydrothermal methods belong to this class. The final category is that of solid phase processing, which includes solid state reactions, grinding, milling, and alloying. The first category is the most relevant to the growth and design of thin films, with both single layer and multilayer structure, nanotubes, nano-filaments, or nanoparticles. Indeed, to best understand the physics of thin films, one must apply nano-physics

Thin Films

92

and nano-chemistry concepts in order to describe the behavior of many of the relevant thin-film materials. Furthermore, within the vapor deposition category, one should distinguish between PVD or CVD. CVD involves the formation of a thin solid films on a substrate material by a chemical reaction of vapor-phase precursors. On the other hand, processes such as evaporation and reactive sputtering belong to the PVD category, which involves the adsorption of atomic or molecular species on the substrate [15]. In the next sections, a summary of the most relevant gas phase vacuum techniques is presented. It should be noted that other vacuum-based options can be found in the references cited above; however, the general principles are comparable to the ones described below. A description of non-vacuum-based techniques will also be provided in Section 2.4.3.4.

2.4.3.1

Chemical Vapor Deposition

Fig. 1 depict the basic physicochemical steps in an overall CVD reaction. There are several key steps that describe the overall process (see [15] and references therein): 1. The first step involves the process of evaporation of precursors (i.e. the reagents that will provide the components for film growth). These precursor are introduced from the bulk gas flow region into the reactor; it should be noted that mixing two or more precursor gases can be achieved by either controlling the flow rates of the respective gases, or by premixing the desired partial volumes and flowing the gas mixture into the chamber. 2. The second step occurs in the reaction zone, where gas precursors react to produce intermediates and gaseous by-products. 3. In the third step, the reactants are transported to the substrate surface. 4. The reactants are then absorbed on the substrate surface; this phase is highly dependent on the substrate temperature. 5. During absorption, the phenomenon of surface diffusion can occur, that is, the reactants will diffuse toward growth sites. Afterwards, nucleation processes and surface chemical reactions will lead to film formation; again, substrate temperature has a large influence on this process. 6. The final step is represented by the removal of the remaining fragments from the reaction zone via desorption and mass transport. Jones and Hitchman [15] describe most CVD reactions as thermodynamically endothermic. A kinetic energy of activation has to be supplied to the reacting system. Traditionally, a thermal energy input is required to initiate all CVD processes. This can be achieved with a variety of methods, for example: 1. 2. 3. 4.

Direct resistance heating of the substrate or substrate holder; Radio-frequency (RF) induction of the substrate holder or susceptor; Thermal radiation heating; Photoradiation heating.

These choices have been implemented with a variety of standard types, namely, atmospheric pressure chemical vapor deposition (APCVD), which operates, as the name implies, at atmospheric pressure, and can operate with wafers on a horizontal conveyor belt [18]; in low pressure chemical vapor deposition (LPCVD) it is possible to load the wafers in either horizontal or Main gas flow

Gas phase reaction Desorption of volatile surface reaction products Transport to surface

Adsorption of film precursor

Desorption of precursor

Surface diffusion

Step growth

Nucleation and island growth Fig. 1 Precursor transport and reaction processes in chemical vapor deposition (CVD). Adapted from Jones AC, Hitchman ML. Chemical vapour deposition: precursors, processes and applications., London: RSC Publishing; 2009.

Thin Films

93

Horizontal LPCVD reactor Pressure gauge

Water-cooled end cap assembly 5-zone resistive heater

Vacuum break valve

Wafer load/unload end cap Batch of wafers

Silica reactor tube

Dry N2

Particulate filter

Wafer boat

Mechanical booster pump

Dry N2 ballast valve Mass flow controller system Gas cabinet N2 O2 NH3 SiH2 Cl2 PH3 SiH4

Rotary piston mechanical pump

Pump oil purifier

Exhaust Fig. 2 Schematic of low pressure chemical vapor deposition (LPCVD) system used for deposition of different thin films. This technique permits either horizontal (as in this diagram) or vertical loading of the wafers into the furnace and accommodates large numbers of wafers for processing. Downloaded from: https://images.search.yahoo.com/yhs/search?p=LPCVD þ system þ figure&fr=yhs-mozilla-001&hspart=mozilla&hsimp=yhs001&imgurl=http%3A%2F%2Fwww.hindawi.com%2Fjournals%2Fjma%2F2014%2F954618.fig.002.jpg#id=0&iurl=http%3A%2F%2Fwww.hindawi. com%2Fjournals%2Fjma%2F2014%2F954618.fig.002.jpg&action=click

vertical position into the furnace; this allows for the accommodation of large numbers of wafers for processing. The films have excellent purity and uniformity (typically better than APCVD and PECVD). There is less dependence of the resulting layer on gas flow. On the other hand, among the less desirable characteristics of this technique, the process requires higher temperatures and the deposition rate is low. Pressures required can be from 0.1 to 2 Torr. Fig. 2 shows the schematics of a LPCVD system.

2.4.3.2

Plasma Enhanced Chemical Vapor Deposition

The use of thermal CVD can be disadvantageous. For example, heat input can result in damage to temperature-sensitive substrates and so alternative forms of energy input have been developed which allow deposition at lower temperatures [15]. In order to lower substrate temperature, ionized gases (plasma) are added to the process. Schematics of the principles of PECVD are shown in Fig. 3. PECVD is characterized by the electrode configuration and type of electrical power for the formation of dielectric films at relatively low temperatures (up to 3001C). Other types of industrial PECVD equipment and processes have been developed with most of them being based on the use of nonequilibrium, nonthermal plasmas sustained by RF or microwave (MW) sources (also referred to as very high frequency (VHF)). Other plasma source are simple dc, and saddle field [19]. See also Refs. [20,21]. Fig. 3 represents a standard RF-PECVD deposition system. There are two main processes that must occur in PECVD. First, precursor molecules collide inelastically with energetic particles (mainly electrons) formed in the plasma; this leads to the formation of chemically active species which will contribute to film growth. For instance, during the growth of amorphous silicon, silane gas is used as a precursor (SiH4). Silane is similar to methane (CH4) in its molecular structure, with a silicon atom (instead of carbon) bonded to four hydrogen atoms. The inelastic collisions will break these bonds and form several ionized species, Si Hþ n , where n¼ 1,2,3. These species, and their relative ratios in the plasma, will play a fundamental role in the determination of the film structure when reaching the substrate. This phase in fact is the second crucial process in PECVD, that is, the delivery of energy to the substrate surface for enhancement of surface processes. The surface processes include nucleation, which represents the initial self-assembly of the material, particle migration, and heterogeneous kinetics, usually associated to the substrate

94

Thin Films

Plasma enhanced CVD system Inert gas

Process gas

RF power

Plasma Wafer Heated plate By-products Fig. 3 A plasma enhanced chemical vapor deposition (PECVD) system. In this process, radio frequency (RF) is used to induce plasma in the deposition gas. This results in a higher deposition rate at relatively low temperatures. Step coverage is good. Downloaded from: http://www. dowcorning.com/images/body/etronics_cvdtut_plasma_enhanced.jpg.

temperature. The plasma parameters determine the composition, structure, and the properties of the deposited films [22]. The main steps of a deposition process can be summarized as follows:

• • • •

Dissociation of the gas precursors Physical and chemical reaction in the plasma to determine the flux and nature of the species reaching the substrate Plasma–surface interaction Reactions in the growth-zone

The RF-PECVD system depicted in Fig. 3 represents one of the industry standard technique; however, it should be noted that control of the ionized species can be better achieved by using other types of excitation. In order to choose the appropriate method for achieving the desired film properties, an understanding of the physics of the plasma (or glow discharge – GD) is necessary.

2.4.3.2.1

Glow discharge

Glow discharge is the name associated with the ongoing plasma process, which is an electrically driven process. A crucial feature of a GD process is represented by the electrode configuration, which can be described by the cathode–anode relationship in a medium vacuum of approximately 0.4–13 mbar (0.03–10 Torr) [23]. The main process of GD is represented by “gas ionization”, which can be summarized as follows: first, an electrical (or electromagnetic) field is established between two electrodes; the field will then ionize the precursor gas by promoting collisions among the neutral particle and the charged particles present in the gas, thus further ionizing the gas. A continuous plasma will be established (very much like the bright plasma in a neon tube used for illumination, although of different color). This is usually referred to as a “cold plasma”, in which the energy of electrons is much higher than that of ions, as opposed to thermal plasmas where the energies are comparable. It is logical that the nature of the electrical source (DC, RF, VHF) has great implications in the plasma process. Furthermore, gas pressure also plays a crucial role in the plasma dynamics. This is better understood by considering the mean free path, one of the major parameters to be considered in vacuum technology. The mean free path is defined as the average distance traveled by a residual gas atom or molecule between successive collisions. If the atom or molecule is an ionized species (i.e. having a net negative or positive charge), then it will be accelerated by the existing electric field. Obviously, longer mean free paths (corresponding to lower pressures) allow for greater kinetic energies of the ionized species. The mean free path depends on pressure (higher pressures imply a more packed gas and a shorter distance between molecules, leading to shorter mean free paths), on gas species (due to different molecule sizes, which impact on the collision cross section) and on temperature (responsible for the thermal energy, and therefore the speed of the molecules). It is possible to calculate the mean free path of gases at a fixed temperature. For instance, using the standard 3 (cm), where l is the mean free temperature of 201C, the mean free path of air is, with good approximation [21,23,24], l ¼ 6:6510 p path (in cm) and p is the residual gas pressure (in mbar). As mentioned above, there are several forms of discharge that can be implemented. For instance, in a DC GD, there is a fixed cathode and a fixed anode. Electrons will be accelerated from cathode to anode and acquire more energy. Electrons begin to collide with gas molecules, and the collisions can be either elastic or inelastic, where: 1. Elastic collisions, by definition, do not diminish the electron’s energy, just direction and momentum. Indeed, the larger molecules are not significantly influenced due to the great mass difference between electrons and molecules (more than 10 orders of magnitude). 2. Inelastic collisions, on the other hand, are responsible for the energy transfer between the colliding species and the target molecule, which goes either into an excited state or becomes completely ionized by the removal of an electron.

Thin Films

–

95

+ DC

Anode

Cathode

A+

A+ e–

e–

e–

A

e–

A

A

e– A+ e– e–

A: Neutral particles

A+: Neutral particles

Accelerating force

e–: Neutral particles Reaction path

Fig. 4 Schematics of a plasma reaction. Downloaded from: http://piescientific.com/Resource_pages/Resource_DC_glow_discharge.

(The excitation - relaxation processes are responsible for the glow which we can observe with the naked eye. The color of the glow depends on the energy transitions created by the process.) A process similar to avalanche breakdown will happen in the gas chamber. Electron and ion densities will be multiplied and eventually the whole space will be filled with positive ions, negative ions, and electrons. There is an optimum voltage range between cathode and anode to generate a plasma. If the voltage is too low, free electrons do not have enough energy to ionize the neutrals. If the voltage is too high, electrons will move too fast thus avoiding any collisions with neutral atoms. The optimum voltage range is a function of gas species and pressure [24]. Fig. 4 shows the schematics of the process, with the electron on the left ionizing a neutral atom (A) into a positive ion A þ plus two electrons (one being the original one and the other resulting from the ionization of the atom). The process then continues in a cascading sequence. The most important parameters of the GD process are the ionization and plasma currents. However, one of the drawbacks of DC plasma is that it is very difficult to use insulating substrates. In fact, insulating materials will tend to “charge-up” as the plasma generated ions imping the surface. Eventually, the ion current will charge the insulator positively, thus opposing the other incoming positive ions) and ultimately extinguishes the plasma which cannot be sustained. RF discharge is used to allow for deposition on insulating materials, as it employs an alternating (AC) signal. In particular, electrons will be able to respond to the AC signal at frequencies 4100 kHz, while ions (due to their larger mass) do not. The typical RF frequency used in thin film growth is 13.56 MHz (designated by the FCC). Furthermore, the RF signal will make electrons oscillate as the polarity changes, thus increasing their mean free path and the probability of collisions. It should also be noted that the use of a RF source implies the use of a matching network between the RF generator and the GD. This is done to increase the power dissipation in the discharge and protect the generator. Therefore, although the RF PECVD is more flexible and efficient than DC, it is limited in the geometry of the reactor by the need of achieving good coupling with the generator. Alternative solutions have been explored, including the saddle field GD [19]. Fig. 5 shows the schematics of a saddle field system. The saddle-field technique is a remote plasma CVD method. It allows independent control of plasma and deposition parameters. It is scalable and more economic to implement than conventional RF growth techniques. As shown in Fig. 5, three semitransparent electrodes are employed, with a DC source connected to the central anode, while the two cathodes are grounded. This allows for electrons to oscillate about the central anode, thus reproducing the RF effect of longer mean free paths. Furthermore, the substrate is placed outside the three electrodes, allowing for independent biasing and control of the impinging ion energies. The main deposition parameters which influence the properties of thin films in any GD system are gas flow rate, gas pressure, discharge power (i.e. anode voltage, substrate current), substrate bias, substrate temperature, and gas composition (i.e. hydrogen dilution, gas mixture, dopant concentration). Another vacuum technique is the electron cyclotron resonance reactor. In this type of reactor, the discharge energy is provided by MW power and a magnetic field. One can also use RF power by establishing inductive coupling to the plasma, thus achieving what is known as an inductively coupled discharge. These techniques are referred to as “high-density sources” discharges, which are associated with a relatively low gas pressure while maintaining high plasma densities [25]. This technology is very popular in the semiconductor industry. MW plasmas can be induced by a variety of methods, as described in Ref. [25], which include resonance cavity plasmas, free expanding plasma torches, and surface wave discharges. A dielectric barrier discharge (DBD), where the electrodes are typically covered by a dielectric, is a type of AC GD which operates at atmospheric pressures.

96

Thin Films

Electrical feedtrough Anode

Cathode

Substrate holder

To gas mass spectrometer

Cathode

Gas inlet

Turbo pump

Motorized butterfly valve Turbo pump

Discharge chamber

M

Mechanical pump

Mixing bottles

Supply bottles

Fig. 5 Schematics of a saddle field glow discharge system. Reproduced from Kosteski T, Kherani NP, Gaspari F, Zukotynski S, Shmayda WT. Tritiated amorphous silicon films and devices. J Vac Sci Technol A 1998:16(2), 893–96.

Another variation is the pulsed GD: in this technique, short GDs (of the order of milli- or microseconds) induce what is called an afterglow, which usually lasts for a longer time-period. By using short pulses, the average power employed will be low, while high peak electrical powers can still be reached, thus improving the efficiency of the process. Among the high power density techniques it is important to mention the magnetron. This method adds a magnetic field to the GD. Since moving electrons respond both to electric fields and magnetic fields, the resulting electron path will resemble helices around the magnetic field lines, which will increase the mean free path and give rise to more ionization. Therefore, lower gas pressures can be used while achieving higher currents than conventional GDs [25]. Fig. 6 shows the schematics of a magnetron reactor. As previously mentioned, the magnetic field generated near the anode allows for a longer mean free path and greater ionization resulting in greater number of ions (Ar þ in this case) being directed toward the substrate (cathode).

2.4.3.3

Physical Vapor Deposition

PVD describes another group of vacuum deposition methods which can be used to produce thin films. The main differences of PVD from gas-based GD techniques such as CVD are represented by the source materials, which are obviously in gaseous form for CVD, but are in solid form in the case of PVD.

Thin Films

97

Magnetron Power

Ar+

Anode ring

Gas in

Ar+

Ar+

Target

Pumps

Table bias Fig. 6 A Magnetron reactor. Downloaded from: https://www.oxford-instruments.com/OxfordInstruments/media/plasma-technology/Process% 20Techniques/Content%20Small/PVD.jpg

The main characteristic of PVD is represented by the techniques used to vaporize the solid target, and include heating of the target, sputtering, and pulse laser. The vapor produced by these processes is then deposited on the object which requires coating. PVD can also be described by four sequential phases for the overall process, which are [26] evaporation, transportation, reaction, and deposition. A beam of high energy electrons or ions are used to bombard a solid target of the desired material. This will “vaporize” the atoms at the surface (evaporation), which at this point will behave in a manner similar to the species formed by collision in CVD, and will be transported from the target to the substrate to be coated. This constitutes the transport phase. In the reaction phase, for coatings made of metal oxides, nitrides, carbides, and other such materials, the target will consist of the metal. As the metal is vaporized, its atoms will react with the gas used during the transport stage. Representative reactive gases are oxygen, nitrogen, and methane. However, when the coating consists of the target material alone, this step would not be part of the process. Finally, the deposition process (i.e., the film growth, or coating, on the substrate surface) will occur. Depending on the actual process, some reactions between target materials and the reactive gases may also take place at the substrate surface simultaneously with the deposition process.

2.4.3.4

Atomic Layer Deposition

As indicated in the introduction, the term “thin film” is referred to layers thicknesses ranging from tens of nanometers to a few micrometers. Atomic layer deposition (ALD) is another technique capable of depositing a variety of thin film materials from the vapor phase. As for PVD and CVD, the deposition occurs on the surface of an appropriate substrate material. The source materials (targets) can be of metallic nature (Al, Au, Cu, Ar, Cr, Ti, etc.) or an alloy (Al/Si, Al/Cu, Fe/Ni, Fe/Mn, etc.), and could have insulating, magnetic, or semiconducting characteristics (i.e., Si, Ge, Se, Te, etc). ALD has demonstrated potential advantages over CVD and PVD techniques, as it allows for accurate control over materials thickness, conformity, and composition. Conformity refers to the ability of the process to evenly coat the substrate. ALD has proven to be particularly useful for the growth of materials for energy storage, as in SOFCs (see Section 2.4.7.3 in this chapter). Furthermore, ALD grown films exhibit very low number of defects, including pin holes, which are detrimental for the performance of the film. In ALD, two or more chemical gas precursors react sequentially on the substrate surface, producing a solid thin film. In most cases, an inert carrier gas flows through the system and precursors are injected as very short pulses into this carrier flow (this process is referred to as a flow-through traveling wave setup). The carrier gas flow takes the precursor pulses as sequential “waves” through the reaction chamber, followed by a pumping line, filtering systems and, eventually, a vacuum pump. Although the techniques described above cover a large portion of standard methods for thin film growth, several other techniques and processes are being employed for the growth of some novel materials. When appropriate, a brief description of these processes will be given within the context of the thin-film material being discussed; however, it is important to mention some of the most popular non-vacuum-based techniques, as these methods have the clear advantage of being low-cost since they do not require vacuum technology.

98

Thin Films

2.4.3.5

Non-Vacuum-Based Techniques

In the previous sections, a review of vacuum-based techniques was presented. It is important to underline how these techniques use vacuum in order to achieve a much better control of the morphology and growth process of the desired film. Vacuum allows for better control of contaminants and of the dynamics of the growth process. However, generally less expensive techniques can be employed when these issues are not relevant. In particular, one can list a number of processes under the general definition of liquid phase synthesis (LPS).

2.4.3.5.1

Liquid phase synthesis

As previously indicated, the second category for materials synthesis is represented by liquid phase methods. Colloids (which are solutions with particles of intermediate size) and aerosols (finer particles dispersed in a gas) can be formed by chemical reactions in solvents. Electrodeposition, sol–gel and hydrothermal methods belong to this class. Essentially, the main process in LPS is the precipitation of nanoparticles after a solution of chemical compounds has been prepared. However, several methods can be used to achieve this type of process, and they are usually classified into six categories: 1. Colloidal methods, based on liquid chemical reactions of the reactants. In some cases, the growth of nanostructures requires high temperatures, complex reactions, and even advanced equipment. The role of the starting reactants, their concentrations, the reaction temperature, and time of the presence on dispersants is also an important aspect. 2. Sol–gel processing, which is a well-known technology within the field of colloidal chemistry. Sol–gel methods can produce a variety of materials while at the same time being able to easily determine their properties. It is a simple process and has a relatively low production cost. In general terms, the sol–gel process induces a chemical transformation of a liquid (the sol) into a gel state. Indeed, the term “sol” refers to the solid particles dispersed in a liquid. These particles are of the order of a few hundreds nanometers in diameter. The “gel” phase represents the assembly of these particles into a solid macromolecule. The resulting gel can also be subject to posttreatment in order to obtain solid oxide material. 3. The water–oil micro-emulsions method. In general, micro-emulsions can be considered as small-scale versions of emulsions, i.e., droplet type dispersions either of oil-in-water (o/w) or of water-in-oil (w/o), with a size range in the order of 5–50 nm in drop radius. 4. Hydrothermal (or solvo-thermal) synthesis. From the materials science stand point hydrothermal syntheses are defined as reactions occurring under the conditions of high-temperature-high-pressure (41001C, 41 atm) in aqueous solutions in a closed system. More generally, solvo-thermal reaction means any chemical in the presence of a solvent in supercritical or near supercritical conditions. The purpose behind using any solvent other than water in the chemical reactions is essentially to bring down the pressure–temperature conditions. Hydrothermal (and solvo-thermal) synthesis are performed in autoclaves which are pressure chambers used to carry out processes requiring elevated temperature, and pressure different from ambient air pressure. 5. The polyol method. This method can be best described using the definition from the Royal Chemistry Society, that is, the synthesis of metal-containing compounds in poly-(ethylene glycols). The ethylene glycol acts as both the solvent and reducing agent. 6. Electrodeposition. As the name implies, this method used electrical power to produce a thin film. An anode and a cathode are usually employed within a liquid solution, where the cathode is made of the metal to be plated, and the anode is the metal to be plated on the cathode. The liquid solution is an electrolyte, which favours ion currents, and a DC power source is used to supply a direct current to the anode thus oxidizing the metal atoms that are then dispersed in the solution.

2.4.4

Background: Characterization Techniques

Standard characterization of materials’ properties in the field of materials science is usually independent of the size of the materials. However, thin films, by their very definition, have a finite resolution of depth. It is not easy to distinguish between surface and bulk properties, and particular attention must be given to the interpretation of the analytical data. In particular, the composition of the surface is averaged over the depth of the film. Therefore, it is desirable to use characterization methods capable of resolution at the monolayer level, in order to investigate adsorption and desorption processes. The same methods can also be used for the investigation of film growth. An excellent review of thin film characterization techniques is given by Frey and Helmut [27]. The following sections present a selection of some of the most relevant techniques. Ref. [27] can be consulted for more in-depth descriptions.

2.4.4.1

Methods for Structure and Composition Analysis

There exist many techniques used to get information on the structural and compositional properties of a material. It should be kept in mind that these properties are different when the analysis considers only the surface of the material as opposed to the bulk. Furthermore, in order to distinguish between surface and bulk properties, one needs to define what “surface” means, that is, how many layers we can probe before the observed characteristic become “bulk-like”. Therefore, when choosing an analytical method, one must keep in mind the effectiveness of the method in describing surface and/or bulk properties. Usually a surface is delimited by the top 2–10 layers below the top layer (0.5–3 nm). This is because they influence the nature of the top layer, so they can be considered the “real” surface of the material. Rutherford backscattering (RBS) [28] is a classic

Thin Films

99

technique, and is able to probe depths of the order of 10 nm. It consists of analyzing ions after they collide with the target material and are backscattered by the internal collisions. The analysis provides information on the structure and composition of the material. RBS can be operated in the high vacuum range (10 6 mbar), and the contribution of the top layer to the total signal is sufficiently small. One of the disadvantages associated with these methods is that, in order to probe the depth profile of the compound, layers of the sample are removed by sputtering, thus effectively destroying the sample. Therefore, one has to be confident that samples grown during the same deposition process possess identical properties. In order to achieve the sputtering of layers, ions (usually Ar þ ) are used to bombard the film, with energies in the range of 1–2 keV. Differential sputtering must also be considered in alloys, that is, different sputtering rates exist for different elements of the alloy. Indeed, as the different elements are sputtered with different rates, the elemental composition is gradually modified, and can lead to erroneous conclusions. Furthermore, the erosion rate can also change with the depth [27]. Understanding the effectiveness of these techniques at different depths is crucial to achieve a complete understanding of the structure and composition. Secondary ion mass spectroscopy (SIMS) [22] is a very powerful technique for thin-film analysis. SIMS can be used in three main modes: statics SIMS, used for sub-monolayer elemental analysis; dynamic SIMS, used for obtaining compositional information as a function of depth below the surface; and imaging SIMS, used for spatially resolved elemental analysis All of these variations on the technique are based on the same basic physical process. Ion bombardment is used to dislodge surface particles to be analyzed. SIMS ions originate from the top one or two atomic layers. The primary information from the emitted ion signal can provide information on quantitative chemical composition (mass spectrum of ions), and structural information (cracking pattern as conditions varied). The main components of a SIMS apparatus are the primary beam, defined by the mechanism of beam production, i.e., electron bombardment, plasma, surface ionization, field ionization. The choice of beam depends on the needs of the analysis: i.e., spatial resolution, speed, sensitivity, coping with insulating materials, magnitude of damage [22]. There are three main categories of mass analyzers used in SIMS: magnetic sector, quadrupole, and time of flight (TOF). Once again, the choice of mass analyzer is also linked to the desired information. As indicated previously, one of the drawbacks of SIMS is that in most cases (except for the static one), it is a destructive technique, i.e., the sample is being etched during measurement and cannot be recycled. On the other hand, with compounds whose stoichiometry goes beyond three or four atoms, SIMS is a powerful tool for structural characterization, capacity of surface studies of catalysts, adsorbates, and surface transformation. When used with other techniques, like Auger electron spectroscopy (AES), and X-ray photo-electron spectroscopy (XPS), sputtering can sequentially remove layers of material and build-up depth distribution of elements. AES and XPS can be easily calibrated for sputtering rate (Å
min 1) for polycrystalline elements; however, complex materials require the use of “standards” in order to achieve an accurate picture of the depth distribution [29]. A major issue in mass spectroscopy based techniques is the sputtering rate. As mentioned above, in compound materials the sputtering rate is different for the various elements of the compound. Therefore, prolonged sputtering will gradually modify the elemental composition of alloys and compounds. XPS and AES will initially show the true concentrations for the first layers, but will change to a new equilibrium value for deeper layers. On the other hand, after short sputtering period, SIMS is unaffected by differential sputtering yields [17,29]. The techniques described above represent the basic mass spectroscopy based techniques; however, many variations and improvements have been developed over the years, as described in detail in Ref. [17]. Inverse photoemission (IPE), for instance, is used as a complement to photoelectron spectroscopy methods; electron energy loss spectroscopy (EELS) analyzes the scattering of electrons to obtain information from the scattering process energy and momentum conservation; diffraction of slow and fast electrons (LEED and RHEED) uses the diffraction process for the determination of the chemical composition of surfaces and thin films and obtain information about the arrangement of surface atoms [17]. Other techniques include image spectroscopy methods, which give information on the surface of structures up to atomic dimensions. Some of these techniques allow for dynamic scanning with a time resolution in the range from 1 ms to several seconds; in photoemission microscopy the sample surface is irradiated with photons and the variations in the photoelectron yield (due to different electron affinity ef) are observed and used to analyze the surface structure [17]. Scanning photoemission microscopy (SPM)) uses a light beam (hnE6 eV) focused on the sample and moved laterally with stepping motors. Other techniques include photoemission electron microscopy (PEEM) and low energy electron microscopy (LEEM). Both obtain images created by photoelectrons and reflected electrons, respectively. A thin film can be characterized by many other structural related properties, including permeation, mechanical stress, hardness, adhesion, roughness, film thickness, and pinhole density. A detailed description of the standard techniques listed above can be found in Frey [17], Chapter 10.

2.4.4.2

Thin Layer Optical and Electronic Properties

After deposition, it is important to determine the macroscopic properties of the thin film that have the most relevance in commercial and energy applications. In particular, thermal conductivity, electrical conductivity, optical absorption, and luminescence provide crucial information of the quality of the thin film. Once again, these techniques are described in detail in Frey and Khan [17]. A brief summary of the most popular ones is given below.

100

Thin Films

2.4.4.2.1

Thermal conductivity

The thermal conductivity of thin films is important for the nucleation process and the film growth. We can define the heat _ ¼ lð∂δ=∂sÞdA, where dQ _ is the heat flow per unit time through the area dA. This conductivity l of a material by the formula: dQ process is also dependent on the surface-normal existing temperature gradient (∂δ/∂s). However, it should be noted that thin films might exhibit anisotropic behavior in samples of varying layer configuration and materials with different cross-plane and in-plane thermal components.

2.4.4.2.2

Electrical conductivity

A four-point probe is a simple apparatus for measuring the surface resistivity of semiconductor samples. The four probes can be aligned or placed at the corners of a square. The experiment consist of passing a current through two outer probes (or two opposite corners probes) and of measuring the voltage through the inner probes (or the other opposite site corner probes). This allows the measurement of the substrate resistivity. The resistivity of the samples can be calculated using the formula(s): R□ ¼ 4:532 VI for a linear contact configuration, or R□ ¼ 9:06

V for a square contact configuration I

where R□ is the sheet or surface resistance, in O/cm2; V and I are the voltage and the current between probes, respectively. The doping concentration can then be derived from the resistivity. Fig. 7 shows a standard experimental apparatus for a four-point probe electrical resistivity measurement with linear configuration.

2.4.4.2.3

Optical properties

The investigation of optical properties of thin films include very well developed methods like reflectivity (R) and transmission (T) measurement. Simple optics (Fresnel’s law) teaches that the reflectivity of a thin-film coated surface depends on the wavelength of the impinging photons. The transmitted light will also be frequency dependent. Therefore, the wavelength range must be chosen according to the information required. R and T measurements in the UV–visible range lead to the determination of the optical absorption. Indeed, light going through a material can either be reflected (R), transmitted (T), or absorbed (A). The absorption is obviously what has not been reflected or transmitted. In a standard UV–vis experiment, the intensity of incident light at a particular wavelength is compared with the intensity of the reflected light (R) and Current source and measurement I

V

S

S

S

N Depletion region acts as an insulator Keeping current flow in the emitter

t

P

Fig. 7 Schematic diagram of the four-point probe method. Downloaded from: http://www.pveducation.org/sites/default/files/PVCDROM/ Characterisation/Images/4pp_clip_image001.jpg

Thin Films

101

that of transmitted light (T), which are easy to measure, and the absorbed light is then derived from the difference between input signal and R þ T signal. Ellipsometry is another optical technique commonly used in the characterization of thin films. This technique uses the properties of polarized light. A change in polarization is measured after reflecting light from the surface of a thin film. The changes in polarization depend on the thickness, the index of refraction, and the absorption coefficient of the thin film and the substrate. Thin film thickness (t) and optical constants (n, k) are derived from the measurement. It should be pointed out that the optical constants n (refractive index) and k (extinction coefficient) for a material will vary for different wavelengths and must be described at all wavelengths probed with the ellipsometer. A table of optical constants can be used to predict the material’s response at each wavelength. Optical spectrometry also belong to the group of standard optical characterization techniques. All spectrometers, regardless of their setup, possess a light source, a monochromator, which is an instrument capable of selecting wavelength components, and a light detector. The different wavelengths of a spectrum can be separated into a frequency spectrum by using systems of grating and prism spectrometers. An alternative analysis uses the Fourier transforms of the signal to obtain a time-dependent signal (Fourier transform spectrometers). Fabry–Perot spectrometers belong to the class of narrow-band optical filters spectrometers. The most common technologies in thin film characterization are mainly grating spectrometers and Fourier transform spectrometers (see Ref. [27] and references therein, for the evolution of spectrometry and detailed description). The use of grating or Fourier spectrometers is particularly useful to measure the photoluminescence (PL) or electroluminescence (EL) of thin films. PL is the optical emission obtained by photon excitation (usually a laser) and is commonly observed with III–V semiconductor materials, although a-Si:H also show a large PL signal. This type of analysis allows nondestructive characterization of semiconductors (material composition, qualitative investigations, etc). Electroluminescence [30,31] can be explained using the same concepts that describe a LED and is very useful for diode, or solar cell, characterization. Current is fed into a semiconducting diode, carriers are excited from the semiconductor valence band to the conduction band, and then recombine radiatively (that is, with no energy losses) and induce light emission of energy equal to the energy gap of the semiconductor. Radiative recombination is very effective in direct band-gap semiconductors; however, indirect band-gap semiconductors, like crystalline silicon, still possess a small amount of band-to-band recombination producing radiative emission, even though most of the recombination occurs via defects or Auger recombination. Such a small amount of radiative recombination can be sensed using an external optical detector. However, since an electrical impulse must be applied, electrical contacts are needed for the sample and so characterization can be performed only once the metallization has been applied. Optical microscopy [27] is also a useful characterization technique; however, present research relies more heavily on more sophisticated microscopic techniques. Indeed, images of materials where the atoms comprising the materials within the image are individually distinguishable can be obtained via scanning probe microscopy (SPM) and electron microscopy (EM). Atomic scale imaging helps characterize molecular and other nanoscale structures by allowing people to view the exact atomic structure present within the image. These images in particular help improve the process of self-assembly by allowing a view of the structure before and after self-assembly. Currently, the two most popular methods of the two camps are atomic force microscopy (AFM) and highresolution transmission electron microscopy (HRTEM or TEM). Scanning electron microscopy (SEM) is also used widely but does not have the same resolution of TEM. Both SEM and TEM use electron beams. In SEM, the electron beam is used to excite atoms to emit Xrays. The X-ray spectra can be used to identify the surface distribution and the bulk material distribution of the elements with very high resolution (of the order of nanometers). HRTEM analyzes the spectrum of transmitted electrons, providing resolution of individual atoms. By transmitting the electrons through the sample, researchers are able to achieve higher resolutions and magnifications than other forms of electron microscopy. For instance, the crystallite structure of the material can be probed analyzing the electron beam diffraction spectrum. Most implementations of TEMs contain several common components. These include condensing and imaging lenses, in order to focus the electrons used to capture sample information; an electron gun, in order to fire the electrons at the specimen; a specimen stage to hold the samples; and an aperture in order to record the intensities of the captured electrons. Several forms of TEM have been developed, including scanning transmission electron microscopy (STEM) and cryo-TEM. In a conventional TEM, a beam of electrons passes through several lenses in order to focus the beam to illuminate a sample. The electrons which transmit through the sample are projected onto a viewing screen or captured by a camera sensor. The electron beam, or primary electrons, functions similarly to the light beam of an optical microscope. Electrons are diffracted whenever an object is encountered or may pass through the sample if they do not interact with an atom. Diffracted electrons, also known as secondary electrons, can also be captured or viewed. The electron beam can also produce X-rays when interacting with the sample. These three sources of information, the primary electrons, secondary electrons and X-rays can all be used to extract details from the sample and the primary and secondary electrons can both be used to create images. Sample preparation is also crucial for these techniques. For instance, TEM requires very thin films (about 100 nm) to achieve atomic resolution. STEM uses a very fine, highly focused beam of electrons, more so than the conventional TEM, to scan over a sample. In STEM, one uses the beam of electrons to scan the sample slowly, so that an image is constructed, while capturing the transmitted electrons at each point in the scanning process. STEM, being a variety of TEM, also produces the same secondary electrons and X-rays that conventional TEMs produce, albeit on a much finer level. This confers to STEM the same analytical abilities that a TEM has but on a larger resolution. Due to the nature of STEM, one can achieve higher resolutions than with conventional TEMs while also being able to image thicker samples, up to 2-mm thick at 200 keV. On the other hand, STEM is much slower than conventional TEMs as TEMs image the entire sample at once. This means that TEM can provide more information of an ongoing process than STEM can provide.

102

Thin Films

Cryo-TEM is commonly applied to biological samples, as in protein structures, where the specimen must be maintained at very low temperatures in order to prevent a phase transition, which would result in formation of crystalline ice and damage to the specimen. Also, it should be mentioned that the need for high resolution implies very expensive equipment and set-ups for all these techniques. In atomic force microscopy (AFM), images of materials and structures, with nanoscale resolution, are obtained by capturing the amount of force applied near the surface of those materials to keep a cantilever probe at a particular predetermined height as the surface of the material is scanned, hence the name of scanning probe microscopy. AFMs have in fact several methods of operation: the user can choose to either maintain a constant level of force applied to the tip or a constant level of height from the surface. The precursor to the AFM was the scanning tunneling microscope (STM). The STM is another SPM that uses the tip of the probe to send a current through the sample to scan the surface of the material. Due to the nature of the capture process, the STM can only capture information from conductive materials. By using force to measure the surface, instead of current, AFMs are able to measure insulators as well as conductors and can also perform surface measurement in many different mediums, which include liquid and gas mediums. Since AFMs and STMs are both SPM techniques, and are similar in construction and operation, occasionally AFMs are equipped with the components necessary to perform the quantum tunneling necessary for STM captures.

2.4.5

Applications

A great variety of thin film can be prepared with the techniques described in Section 2.4.3, but some materials have been prominent in terms of their properties, flexibility and ease of deposition. When examining thin film applications, it is necessary to single out specific materials, or groups of materials, that can be used for different uses. The physical, chemical, and structural properties of thin film can be quite varied. Furthermore, the same material can provide diverse performances according to preparation conditions. For instance, one of the materials that has been prepared by mostly all of the above-mentioned vacuumbased processes is hydrogenated amorphous silicon (a-Si:H), which has been the most successful and flexible thin film material to date. On the other hand, CVD and other related techniques have also been used extensively for the growth of group III–V compound semiconductor material systems, consisting of Group III elements, such as aluminum (Al), gallium (Ga) and indium (In), and Group V elements, such as nitrogen (N), phosphorus (P), arsenic (As), and antimony (Sb). Compounds consisting of alloys of Group III and Group V elements are classified as semiconductors, similar to the closely related Group IV materials such as silicon (Si) and germanium (Ge). Both column IV elemental and III–V compound semiconductors have covalent bonding as a result of sharing electrons in the outer hybridized sp3 orbitals of the tetrahedrally coordinated nearest neighbor atoms; however, in the case of III–V semiconductors, a certain degree of charge transfer occurs, since the outer shells of the Group III and Group V atoms have a different number of valence electrons available, which creates an ionic character to the covalent bonding of III–V semiconductors. It should be mentioned that the semiconducting properties of many III–V compounds and the study of their physical properties, e.g., InSb, AlSb, GaSb, GaAs and InP began in the early 1950s, although some crucial applications, like solar cells, are much more recent [31]. In the following sections, we will focus on the growth processes and physical characteristics of the most relevant thin films and their uses in energy applications.

2.4.5.1

Hydrogenated Amorphous Silicon

Hydrogenated amorphous silicon (a-Si:H) has been the subject of investigation for almost four decades. The success of this material is due mostly to the discovery that hydrogen atoms could passivate the defects of plain amorphous silicon (a-Si). Indeed, a-Si is a semiconductor but the disordered nature of the material is also conducive to the presence of dangling bonds, that is, unsatisfied chemical bonds that act as traps or recombination centers. In an energy diagram, these defects would fill what would be the energy gap of the semiconductor. However, dangling bonds require a proton to have their molecular level filled, and hydrogen is the natural element that can satisfy this requirement. By introducing hydrogen in the amorphous silicon structure, the semiconductor’s energy gap is “cleaned” of most of the trapping states and optical, direct, transition become possible. One of the main advantages of a-Si:H is the fact that it is a low cost, efficient material, used extensively for electronic devices and PV applications, and can be considered the most studied thin-film material. Indeed, there exists an extensive literature of the physics and applications of amorphous materials and, in particular, hydrogenated amorphous silicon [32–34]. Furthermore, a number of techniques which were designed and optimized for the growth of a-SiH films have been eventually employed for many other thin-film materials. When describing the preparation methods used for a-Si:H films, one must mention the first, original studies on evaporated and sputtered a-Si:H. It was found that these techniques, where a solid silicon piece is usually evaporated or sputtered in a hydrogen plasma in order to create Si–H bonds, produced very poor quality films. Indeed, the density of hydrogen atoms that would bond to silicon has always been too low to obtain an electronically viable material. In order to overcome this issue, RF GD, introduced in Section 2.4.3, was employed and it is now well established that this technique produces the best quality material. Other more recent methods claim similar or better results, but they are all essentially plasma enhanced techniques, with diverse forms of excitation to produce the plasma. Searle [35] and Street [36] provide a thorough review of the different methods employed to grow a-Si:H.

Thin Films

103

As in the case of sputtered or evaporated silicon, the GD-based techniques produce a hydrogen plasma that helps the formation of Si–Hn (n ¼1, 2, 3) ion radicals; they are technically referred to as PECVD techniques, and have become the preferred methodology used for a-Si:H thin film growth. The gas used to create the plasma is usually a mixture of silane (SiH4) and hydrogen, so that SiHn ion species can be formed much more easily from the breakdown of the silane molecule. Other gases, like phosphine (PH3) or diborane (B2H4), are used to introduce dopants. The electric field created by the electrode configuration directs the ions produced in the plasma region toward a substrate, where film growth takes place. These PECVD techniques have in common the possibility of tuning the system, and therefore the film characteristics, by the manipulation of several parameters. These parameters include partial gas pressure, gas mixtures, flow rates, electrode bias, substrate bias, and substrate temperature. Depending on the technique, these parameters can be varied independently from each other within a certain range of values. As mentioned previously, plasma deposition of a-Si:H has been investigated for more than 40 years, and several reviews can be found in the literature, including a comprehensive review by Bruno et al. [37].

2.4.5.1.1

Structure and density of states of a-Si:H

The opto-electronic properties of a-Si:H are better understood when one considers the structure of amorphous silicon in comparison to its crystalline form (c-Si). The crystalline structure of c-Si is the well-known diamond structure, which is characterized by a tetrahedral geometry. The bond length between Si atoms is 23.3 nm and the bond angle of the tetrahedron is 109.5 degrees. The amorphous silicon structure, on the other hand, does not deviate too much from the diamond geometry. Indeed, the variations in bond length is within 710% from that of the crystalline structure, while at the same time the bond angle variations are within 75% from the tetrahedral angle. A major implication of the small changes in these parameters is represented by the fact that a-Si: H possesses good short range order (i.e., the first 2–3 nearest neighbors structure is very close to a crystalline structure); however, the small deviations will compound as we extend the structure and will eventually lead to bond breaking and the formation of dangling bonds, associated with the accumulation of structural stress. As mentioned previously, dangling bonds have a negative effect on the opto-electronic properties of a-Si, since they act as traps or recombination centers. These issues were the prevalent reason for the poor quality of evaporated or sputtered amorphous silicon. The solution to the problem is represented by a proper introduction of the hydrogen atoms within the structure, allowing for effective “passivation” of the dangling bonds; PECVD techniques allow for ionized species to react, in particular Hydrogen atoms with the broken silicon bond. Hydrogen satisfies the electron occupancy requirements, and therefore will “passivate” the trap or recombination center; see, for instance, [32,33,36]. It should be again noted that these traps or recombination centers represent defect states which appear within the energy gap of the amorphous silicon semiconductor. Indeed, the density of these defect states, in the absence of hydrogenation, is so high that the material behavior is closer to a conducting behavior rather than that of a semiconductor. Hydrogen atoms have one electron and are very reactive to chemical bonding. By incorporating Hydrogen into the films we are able to satisfy the covalent bonds at defects and micro-voids and also allow the lattice to relax. This will reduce the density of localized states by several orders of magnitude and re-establish the energy gap. Fig. 8(A) and (B) shows the crystalline structure of c-Si and a 3-d representation of amorphous silicon with dangling bonds passivated by hydrogen atoms, respectively.

(A)

(B)

Fig. 8 A 3-d computer model representation of c-Si (A), and a-Si (B) with dangling bonds passivated by hydrogen atoms (red balls). Reproduced from Gaspari F. Optoelectronic properties of amorphous silicon the role of hydrogen: from experiment to modeling. In: Predeep P, editor. Optoelectronics – materials and techniques. Rijeka: In Tech; 2011.

104

Thin Films

g(E) CB extended states

VB extended states Localized states

Defects (dangling bonds) CB tail states

VB tail states

E Fig. 9 A schematic representation of the electronic density of states (DOS), g(E), of a-Si:H. VB indicates the valence band and CB the conduction band. The dashed, red, vertical lines show the mobility edges, which are defined as the energy level separating extended (non-localized) states from localized states. Reproduced from Gaspari F. Optoelectronic properties of amorphous silicon the role of hydrogen: from experiment to modeling. In: Predeep P, editor. Optoelectronics – materials and techniques. Rijeka: In Tech; 2011.

The electronic density of states (DOS) of a-Si:H is shown in Fig. 9 [38]. Once again, Searle [35] and Street [36] provide an excellent description of these features, which can be summarized as follows: 1. The energy gap is delimited by the dashed, red vertical lines. However, there is no abrupt edge between conduction (or valence) band and the gap region. Indeed, extended band states gradually become localized, and are referred to as tail states. This is primarily a consequence of the long range disorder associated with the amorphous structure. The nature of the tails states can be described by a characteristic energy, EU, or Urbach energy, which is associated with the width of the tails states and is also referred to as the Urbach tail width [36]. Both the conduction band and valence band tail states are characterized by a characteristic width, which has been measured to be about 50 meV for the valence band tail states and about 25 meV for the conduction band tail states; it has been shown that the tail states width are a measure of the “degree” of disorder, and their value must be taken in account when considering the opto-electronic properties of a-Si:H. 2. The localized defect states in the middle of the gap are associated with the formation of DBs. It should be noted that the representation given in Fig. 9 only shows the position and general shape of the DOS in the middle of the gap. There exists a variety of models that try to identify the nature of these dangling bonds. The peak in the middle of the gap, shown in Fig. 9, represents three different DB types: neutral, positive, and negative. In fact, a DB is identified not only by the fact that the bond is unsatisfied, but also by its net charge, which is determined by the number of electrons sharing the dangling bond, i.e., no electrons imply a positive DB (D þ ), a single electron makes the bond a neutral one (D0), while the presence of two electrons lead to negative DBs (D ). 3. Finally, the most crucial feature is represented by the mobility edge. In fact, since there is no sharp delimitation between the bands’ extended states and the gap’s localized states, it is necessary to define a critical boundary, called the mobility edge, where the conduction mechanisms change from hopping between localized states to extended conduction. The energy difference between the two mobility edges then defines the mobility gap of the semiconductor. A more in-depth description of the dependence of the optical and electronic properties of a-Si:H and the impact on the DOS of the hydrogen content can be found in Ref. [38].

2.4.5.2

Micro-Crystalline Silicon (lc-Si:H)

Micro-crystalline silicon (mc-Si:H) is a suitable material for application in thin-film solar cells. It is also amenable to applications in thin-film transistors and other devices [39,40]. mc-Si:H refers to structures between pure crystalline and amorphous. Microcrystalline materials are characterized by the presence of nanoscale crystal structures, in the range of 20–700 nm for silicon, with different orientations (see, for instance, [41] and references therein). The crystals of different orientations grow in columns and are separated by an amorphous phase. Usually mc-Si:H films are grown from SiH4 and H2 gas mixtures, as for a-Si:H, but with different plasmas and temperature parameters to induce micro-crystallinity. Other gas mixtures, including SiF4, H2, and Ar, have also been used. The growth of mc-Si:H films is usually associated with a high H2 dilution, moderated RF power, and high deposition pressure; one can control the crystalline fraction (XC) by optimizing these parameters, and the performance characteristics can be

Thin Films

105

optimized by selecting the appropriate ratios. Common methods for preparation of micro-crystalline silicon are Hot-Wire CVD [40], VHF CVD [42], thermal plasma CVD [43], RF CVD [44], etc. One of the issues with mc-Si:H is its low optical absorption. Higher deposition rates are therefore required to further consider the use of mc-Si:H in industrial applications, since low absorption requires thicker films to collect the desired amount of light [42,45–48]. mc-Si:H IR absorption characteristics, which are larger than that of a-Si:H, make it a suitable material for thin-film solar cell applications. Furthermore, contrary to amorphous silicon, mc-Si:H has also larger stability against sun radiation (light soaking). An interesting device development is linked to the different absorption gaps of a-Si:H and mc-Si:H. In fact, the two materials can be combined to fabricate tandem solar cells, which would employ the same basic element (silicon) for the two cells making up the tandem cell, thus avoiding interface issues. a-Si:H/mc-Si:H tandem solar cells (or micromorph solar cells) with stabilized efficiencies up to 12% have been fabricated to date. The best material quality and large-area deposition have been obtained mostly by parallel plate plasma deposition, either from excitation frequencies in the RF or in the very-high-frequency (vhf) range (see [41] and references therein). Another aspect of mc-Si:H that needs to be taken in account is its own structural composition. The structure, in turn, is highly dependent on deposition parameters, such as silane (SiH4) to hydrogen ratio, discharge power and substrate temperature. For instance, Vetterl et al. [49] observe an “increase of dark conductivity with increasing crystalline volume fraction, suggesting a close relationship between electrical transport and the structural details of the material.” The possibility of tuning the properties of mc-Si:H films also translates to device applications not only for solar cells [5], but also for micro-electronics devices[6–8,50].

2.4.5.3

Gallium Arsenide

GaAs applications can be found in laser diodes, infrared light-emitting diodes, solar cells, integrated circuits, and optical windows [6,8,51]. Gallium arsenide has demonstrated properties that exceed those of the less expensive silicon, including a higher saturated electron velocity and higher electron mobility [6,52,53]. The latter property allows for higher working frequencies of GaAs-based transistors (4250 GHz). Furthermore, GaAs has a wide energy band-gap, which minimizes overheating issues. Another advantage of GaAs transistors, when compared, for instance, to Si-based transistors, is the lower noise level in electronic circuits, in particular at high frequencies. The main reason for this improved performance is due to a higher carrier mobilities and lower parasitic resistive paths. Gallium arsenide is of medium gray appearance, and can be mechanically polished with moderate difficulty; however, chemical etching can produce a bright shiny appearance for most low-order crystal planes (the main properties of GaAs can found in the review by Blakemore [53]). As for a-Si:H and mc-Si:H, many techniques are employed for the preparation of GaAs thin films. The following is a summary of these technique from the general review by Murali et al. [54] and references therein:

• •

•

•

• •

Metal organic chemical vapor deposition (MOCVD): metal halides are used as transport agents for GaAs crystal growth employing a closed tube system. In particular, arsenic trichloride (AsCI), metallic gallium (Ga) were used as sources and H2 as the transporting agent to carry AsCI3 from a bubbler into the Ga source zone. Molecular beam epitaxy (MBE): MBE uses thermal energy beams of atoms or molecules to impact a heated substrate under ultra-high vacuum (UHV) conditions. Epitaxial growth of semiconductors or metals is thus achieved. In order to grow GaAs layers on GaAs substrates, it is necessary to heat the gallium and arsenic components in separate cylindrical cells, along with the dopants. The separated beams are collimated and directed onto the surface of the substrate. Vapor phase epitaxy (VPE): In VPE chemical vapors are passed over a substrate in order to deposit solid epitaxial layers. There are two main methods for VPE growth of III–V compounds. The first is the chloride method, where arsenic in chloride passes over gallium metal to form GaCl3. The other is called the hydride method, which employs, among other gases, ammonia, and hydrogen. Liquid phase epitaxy (LPE): in LPE, a cooling solution induces the precipitation of material onto a substrate. The solution, which is separated from the substrate in the growth apparatus, is saturated with the growth material. When the desired growth temperature is reached, the solution is brought into contact with the substrate surface. The cooling process then occurs, where the cooling rate and the cooling time are determined by the desired layer characteristics. The LPE technique has been employed for the growth of GaAs epitaxial layers by tipping method or sliding boat method. Nearly perfect crystalline layers of GaAs have also been obtained by a temperature difference method. Mobilities of around 105cm2/V s at 77K was reported for GaAs layers grown at 970K. Hot wall epitaxy (HWE): with this technique, epitaxial films are prepared under thermodynamic equilibrium, thus minimizing loss of material. With this technique it is easier to control the stoichiometry of the films. Epitaxial layers of GaAs have been grown on mica, GaAs, Ge substrates by this technique. Sputtering: this technique is among the oldest and more mature methods of film growth. As indicated in Section 2.4.3, the basic principle is that by bombarding a target of a specific material we can create vapor species. The bombardment is done with energetic and nonreactive ions. Sputtering refers to the ejection process, initiated by the momentum transfer between the impinging ions and the surface atoms of the target material. Sputtering has also been employed for the growth of GaAs films.

106

Thin Films

V

Ar inlet

Furnace

Platinum

Nickel /GaAs

Melt containing Ga2O3 B2O3 NaF NaAsO2

Fig. 10 Electrochemical reactor for preparation of GaAs. Reproduced from Murali KR, Jayachandran M, RangaRajan N. Review of techniques on growth of GaAs and related compounds. Bull Electrochem 1987;3(3), 261–5.

• •

Vacuum evaporation: in this technique either pre-synthesized GaAs powder is used as source or two sources containing gallium and arsenic are used. The vapors of the material condense on a substrate in vacuum (10 5 Torr). The two-source evaporation method has been successfully employed for growth of GaAs on glass substrates. Electrodeposition: electrodeposition has been used for the preparation of epitaxial layers. It has the advantages that the growth rate is controlled by electrical parameters and is relatively insensitive to temperature. Electrochemical purification also occurs in addition to normal impurity segregation. In the pioneering work on growth of GaAs by electrodeposition at about 1025K a melt composition containing B203/NaF/Ga203/NaAs02 (6.4:20.3:4.2:8.1 wt%) was used. A cell used for this purpose is shown in Fig. 10.

It should be noted that novel techniques are constantly being developed to improve on cost, quality and speed of deposition. For instance, to reduce costs, “laser liftoff” has been used [55], which allows for recycling of the substrate.

2.4.5.3.1

Disadvantages of GaAs

The main drawbacks of GaAs films are cost (based on material abundance) and safety. The latter is particularly troublesome as it is linked to a variety of issues. Flora and Dwidedi (see [56] and references therein) provide an excellent review of these issues. Indeed, they point out “that workers in semiconductor industries are prone to be exposed to this semiconductor material during various operations. It has been observed that industrial workers exposed to GaAs have significantly elevated urinary arsenic levels. Gallium arsenide has been classified as an immune-toxicant and a Group I carcinogen to humans even though no data on human cancer is available and the conclusions are principally based on few incidences of bronchiolo-alveolar neoplasms observed in female rats. It may, however, be noted that gallium arsenide once reaches inside the body, dissociates into arsenic moiety which is a known human carcinogen. On the other hand, gallium moiety which is generally considered safe is reported to be responsible for pulmonary neoplasm observed in male rats.” In the past few decades, various studies have reported the toxicity of GaAs [56].

2.4.5.4

Compound Semiconductor Thin Films

Copper indium (gallium) selenide (CIS, CIGS) thin films require very expensive deposition techniques for the fabrication of high efficiency CIS and CIGS cells. In some deposition techniques, H2Se andH2S vapors are used as source materials for selenium and sulfur, which are poisonous gases.

Thin Films

107

A large number of thin film growth technologies have been employed for the synthesis of CIS thin film compounds, including RF sputtering, magnetron sputtering, multisource vacuum evaporation, flash evaporation, quasi-flash evaporation, molecular beam epitaxy (MBE), chemical spray, and electrodeposition (see Lei et al. [57] and references therein). CIS and CIGS cell have shown excellent absorption properties and are considered mainly for solar cells. For instance, CIGS are described by the chemical formula CuInxGa(1-x)Se2, and can vary energy gap in the range 1.7–1.0 eV for x ¼ 0–1. Other applications are also being explored, including the fabrication of efficient and stable CIS e-based photo-electrodes to drive the photo-electrolysis of water. Indium phosphide is a semiconducting material similar to GaAs and silicon, but is very much a niche product for several reasons. It can be used for laser diodes and virtually any electronic device that require high speed or high power due to great electron velocity. It is however very expensive and is not amenable to large scale production. The first generation of InP solar cells was produced starting from a wafer, where the junction was formed by thermal diffusion of dopants [58]. To overcome cost related problems, due to the InP substrate, hetero-epitaxial growth was then considered. The main problem for this technology is the presence of an 8% lattice mismatch between InP and Si, leading to increased dislocation density and consequently minor optical and semiconducting properties [59]. Zheng et al. [60] used direct non-epitaxial growth of thin poly-crystalline films of III–Vs on metal substrates by using MOCVD. Electrodeposition has also explored as an alternative to vapor phase deposition [61,62]. Fig. 11 shows the electronic band structure of InP with its direct band gap EG ¼ 1.34 eV. Gallium phosphide (GaP) has a high band gap (2.26 eV) that makes it an excellent material for multi-junction solar cells. GaP is usually grown in crystalline form by using epitaxial techniques, and is employed mostly in optical devices, like light-emitting diodes (LEDs) and photo-cells. One interesting feature of GaP is the fact that it can form ternary and quaternary compound semiconductors, such as InGaP and AlInGaP. These compound materials possess several interesting properties, which are suitable for applications like multi-junction solar cell. The amorphous form of GaP (a-GaP), is characterized by a variety of localized states, contrary to what we observe in a-Si. The electronic structure of GaP is shown in Fig. 12. Note that EG ¼ 2.26 and is an indirect band gap. Indium gallium phosphide (InxGa1-xP) is a pseudo-binary III–V alloy of an indirect band gap semiconductor, gallium phosphide (GaP), and a direct band gap semiconductor, indium phosphide (InP). This alloy possesses a tunable band gap that ranges from 1.35 to 2.26 eV (eV), which depends on the compositional range for x¼ 0–1. The band gap range is particularly attractive for use in solar energy conversion and lighting applications because this range encompasses the majority of the visible light spectrum. Being a combination of two different band gap materials, the optical absorption coefficient represents a good indicator of the effect of composition. Fig. 13 shows the absorption coefficient for x¼ 0 and x¼ 1. The values of the two curves at a ¼104 cm 1 give an approximate measure of the band gap. Copper zinc tin sulfide (Cu2ZnSnS4, or CZTS) has several advantages over the other compound semiconductors described above. The main difference between CZTS and Gallium based compounds is that CZTS is made of abundant, environmentally benign, and inexpensive elements. Another advantage of CZTS is represented by its band gap, B1.5 eV, which lies in the optimal range for maximum conversion of the energy of the solar spectrum into electricity. Finally, CZTS has a high absorption coefficient (4104 cm 1 in the visible region of the electromagnetic spectrum). This is an important property as it allows for thinner layers for maximum absorption.

Energy X- valley Γ- valley

300 K Eg = 1.34 eV EL = 1.93 eV Ex = 2.19 eV Eso = 0.11 eV L- valley

Ex EL

Eg 0

〈100〉

〈111〉 Wave vector Heavy holes

Eso Light holes

Split-off band Fig. 11 Band structure of indium phosphite. Downloaded from: http://www.ioffe.ru/SVA/NSM/Semicond/InP/bandstr.html

108

Thin Films

Energy Γ- valley

X1C 〈100〉

Λ1C

L- valley

Γ1C

X- valley

300 K Eg = 2.26eV EL = 2.6eV

E1

E2

Eso

Eso = 0.08eV

EL

Eo

Eg

Eo = 2.78eV

L6C

〈111〉

Γ15V

Wave vector Λ3V Heavy holes Light holes

X5V

Split-off band

Fig. 12 Band structure of GaP. Downloaded from: http://www.ioffe.ru/SVA/NSM/Semicond/GaP/bandstr.html

Absorption coefficient  (cm−1)

107

106

105

104 1

2

103 0

1

2

3

4

5

6

Photon energy hv (eV) Fig. 13 The absorption coefficient vs. photon energy. 300K. Available from: http://www.ioffe.ru/SVA/NSM/Semicond/GaInP/optic.html

There are two main categories for CZTS thin film deposition techniques: vacuum-based techniques and solution-based techniques. The vacuum-based approaches are favored when the goal is to achieve controlled stoichiometry and high uniformity. Sputtering and evaporation are the most common vacuum based techniques used for CZTS thin films. The film growth involves high temperature sulfidation of stacks of metals, metal sulfides, or a combination of the two. The problem with these processes is the slow deposition time, up to several hours for thin film deposition and annealing. On the other hand, solution-based techniques are desirable for their low-cost and high throughput. These techniques have shown tremendous potential for CZTS processing and include precursor-Ink based approaches like sol–gel synthesis [63] and spincoating [64], nanocrystal (NC) Ink based approaches involving nucleation and growth of NCs by the reaction at an elevated temperature of salts of Cu, Zn, and Sn, spray pyrolysis [65], and electrochemical deposition [66].

2.4.5.5

Organic Thin films

As indicated throughout this chapter, PV represents the main area of energy application for thin films. Future directions in this sector are strictly related to the improvement of the cost-efficiency ratio associated with thin films PV devices. Current materials,

Thin Films

109

such as cadmium telluride (CdTe), copper indium selenide/gallium selenide (CIS / CIGS), and amorphous silicon (a-Si), are still part of a development stage. As the technology reaches maturity, economies of scale will benefit, and costs will be further reduced. A potentially greater advance can be achieved with the development of next-generation thin film PV technologies, such as DSSCs, organic PV, and perovskite PV. In addition, novel multi-junction structures might provide an excellent solution for integrating organic and inorganic materials. OTFTs represent a good example of how organic materials can be exploited to achieve applications that involve flexible substrates. Organic semiconductors can be divided into conducting polymers and small molecules. Polymers have lower mobilities than their counterparts due to their higher molecular weight. For obtaining high mobility, the size of the grains of semiconductors should be larger. Several organic materials are being investigated for these applications, see, for instance, Horowitz [67]. Examples of organic TFTs can be found in Ishikawa et al. [68] who showed that a better performance of p-type bottom contact TFT could be achieved by inserting an additional p þ region near the contacts. Watanabe and Kudo [69] succeeded in improving the Ion/Ioff ratio with the addition of a layer of CuPc organic material between the ITO (source) and the active layer, made of pentacene. The result was an improvement in carrier injection that increased the ratio 540 times. Research in organic semiconductors, both in terms of processing and their applications is quite active, as shown by the numerous publications in the last few years (see, for example, Refs. [70–75]), but several aspects still need improvement. For instance, charge localization and conduction also occurs in the bulk of the organic semiconductor, as opposed to inorganic semiconductor where one can restrict the analysis to the conducting channel of the device. This implies that bulk sheet resistance plays a dominant role and must be taken in consideration when developing a model. The properties of these materials are also very dependent on the thickness, due to the dissimilar paths traversed by the charge carriers between the source and drain of a transistor. The impact of the thickness of both the active layer and the dielectric layers must then be examined for both top and bottom contact structures individually. In addition, although the employment of flexible substrates has been verified as feasible, little work still exists on digital circuits. As different layers of materials are generally used to build an OTFT, a thorough study of the effect of different materials on the performance of inverter circuits must be done, including the synthesis of high-performance novel p- and n-type organic materials.

2.4.6

Analysis and Assessment

The analysis of the impact of thin films on energy applications must perforce look at the major areas outlined in the previous sections, namely, PV and micro-electronics. Other energy related applications are also relevant, like the use of thin films in fuel cell technology. However, a note should be made on how to assess the impact of thin films on energetic and exergetic calculations. As indicated throughout this chapter, thin films properties are directly related to the growth methodology, which in turn should also be subject to this type of analysis as the ultimate impact of thin films must taken into account the cost and efficiency of the growth and processing techniques. A good example is given by Wang et al. [76] where the authors apply exergy analysis on ALD of Al2O3 thin film to analyze the utilization and losses of exergy in ALD system. In this analysis, calculations include parameters such as material flow, heat flow, and work flow. On the other hand, the integration of thin film technology in larger systems allows for great improvements in the exergy efficiency of the systems. For instance, thin films play a big role in improving the energy/exergy efficiencies (i.e., heat transfer) in desalination plants and seawater technology [77], and other complex systems like solar-thermal modules [78], fuel cells [79], PV modules [80], and hybrid renewable energy systems for hydrogen and electricity production, and storage systems [81]. The next section will illustrate examples of energy applications of both organic and inorganic thin films, and include discussion on the main impact of the diverse uses of thin films.

2.4.7

Examples and Future Directions

Thin film applications cover a wide range of areas, including structural reinforcement, surface protection, electronic barriers, etc. However, energy related devices make up a great portion of thin film technology, and cover the three major areas of applications, from energy production to energy distribution to energy storage. The following sections gives an overview of the energy related areas.

2.4.7.1

Applications of Thin Films for Optical Coatings

Thin film technology encompasses a variety of materials, including inorganic semiconductors and organic semiconductors. It should be noted that the optical properties of both types of thin films are also employed for light absorption and light trapping processes, as in antireflecting coating (ARC) for solar cells. These films require the proper refractive index, the right thickness and transparency. ARCs are commonly made of silicon dioxide (SiO2), titanium dioxide (TiO2), zinc sulfide (ZnS) with magnesium fluoride (MgF) or layers of silicon nitride with varying refractive index.

110

Thin Films

Antireflection coatings work by producing two reflections which interfere destructively with each other.

Reflections out of phase

π phase change

¼λ Fig. 14 Antireflection coating principle. Two reflections are produced that cancel each other. Downloaded from: http://hyperphysics.phy-astr.gsu. edu/hbase/phyopt/antiref.html

AR coating types

7.0

Reflectivity (%)

6.0 5.0 V-type 4.0 Single layer MgF2 3.0 Broadband 2.0 1.0 0.0 400

450

500 550 600 Wavelength (nm)

650

700

Fig. 15 Plot of reflection versus wavelength shows the performance of the three most common AR coating types: V-type, single layer magnesium fluoride, and broadband. Downloaded from: https://images.search.yahoo.com/yhs/search;_ylt=A0LEVvNudm5Y6GAA2WcnnIlQ; _ylu=X3oDMTE0anRnZ284BGNvbG8DYmYxBHBvcwMxBHZ0aWQDRkZVSTNDMV8xBHNlYwNzYw–? p=Multiple þ Layer þ Antireflective þ Coatings þ Pictures&fr=yhs-mozilla-001&hspart=mozilla&hsimp=yhs-001#id=65&iurl=http%3A%2F% 2Fspie.org%2FImages%2FGraphics%2FNewsroom%2FImported%2Fnov04%2Filluminatingfig2.jpg&action=click

Thin film antireflection coatings are used to reduce the light loss due to reflection from the main substrate. Since reflectivity is dependent on the index of refraction, one can employ a single quarter-wavelength coating of optimum index to eliminate reflection at one wavelength (see Fig. 14). Multilayer coatings can reduce the loss over the visible spectrum. The idea behind antireflection coatings is that the creation of a double interface by means of a thin film produces two reflected waves. If these waves are out of phase, they partially or totally cancel. If the coating is a quarter-wavelength thickness and the coating has an index of refraction less that the glass it is coating then the two reflections are 180 degrees out of phase. A single layer antireflection coating can be made nonreflective only at one wavelength, usually at the middle of the visible. Multiple layers are more effective over the entire visible spectrum [10,82], and are also called broadband ARC (see, for instance, Fig. 15). Other optical applications of thin films include ARC for glasses, optical filters, mirror coatings, attenuation coatings, and beamsplitters [11].

2.4.7.2

Thin Film Photovoltaic and Solar Cells

PV and solar cells probably represent the most active area of research for energy applications of thin films. Thin-film-based PV is mostly focused on a-Si:H, CIGS, CgTe and Perovskite materials. In particular, the latter has recently opened new avenues of research, which examine the application of perovskites in tandem solar cells, with a particular focus on perovskite/silicon structure. An all-perovskite cell is not deemed efficient enough since it does not absorb the red end of the spectrum. Perovskites are metal halides and represent the only example so far of a solution-processable, large-band-gap material with small energetic losses [83]. The physics of a tandem cell can be found in several textbooks and in Refs. [83] and [87], but a brief summary of the concept will help understanding the properties of this structure. Since a semiconductor is limited by the value of its band gap in the optical

Thin Films

111

absorption process, staking different materials with decreasing gaps (windows) will improve the absorption of the overall solar spectrum, increase the number of carriers generated by sunlight, and improve efficiencies beyond the limit calculated by Schockley and Queissar [84] (B33.7% for a single-junction cell). In brief, a single-junction solar cell, due to its small band gap, generates a low voltage from the available solar spectrum. The large-band-gap cell put on top of the small band-gap cell will help absorbing the high-energy photons, generating a larger voltage from these photons than the small-band-gap cell. Indeed, theoretical efficiencies over 40% have been calculated for tandem cells. The solution processability of metal-halide perovskites provides the potential for a low upgrade cost to an existing manufacturing plant. Current research is examining better processing techniques for Perovskites [85], alternative structures, including two Perovskite layers with different band-gaps [86], and integration with proven technology like CIGS cells [87].

2.4.7.3

Organic Light-Emitting Diodes

It is useful to group optical applications for organic thin-film based research in two main categories: organic photovoltaic cells (OPVs), discussed previously, and organic light emitting diodes (OLEDs). As stated before, organic materials offer a variety of advantages, including the possibility of using polymer substrates, which are lightweight, cheap, transparent, printable and flexible [88]. However, due to their more porous structure, polymers tend to exhibit high gas or vapor permeability, which could impact negatively device performance. Indeed, the main issue with organic devices is that their characteristics should stay constant during prolonged operation and varying environmental conditions. Thin films offer a solution also for this problem, and barrier films can be used to reduce the degradation induced by water and oxygen exposure. Exposure to moisture and oxygen can be avoided by depositing a film of inorganic oxides, however, one must understand the physical mechanism at the basis of the instability to properly choose the right material and design. Encapsulation of the devices offers a viable solution to prolong lifetimes. This can be achieved by attachment of a glass or metal lid to the substrate. The lids can be attached with a low-permeation adhesive. Furthermore, one can add a thin barrier coating at the top and bottom side of the device. Barrier layers require a controllable technology that would deposit films with no pin holes (which create shorting electrical paths) and therefore produce barriers capable of separating effectively the thin film with the outside environment. ALD has proven to be particularly effective in this regard. ALD is capable of depositing very thin barriers, when compared to other techniques, with lower water permeation rates [88]. Indeed, one of the most effective and successful barrier coating on polymer substrates is represented by ALD alumina [88]. However, best performances were achieved using multilayer films. The structure of such Inorganic multilayer barrier films is made up of alternate layers of two ALD metal oxides such as alumina and titania. Furthermore, one can also use alternating layers of ALD metal oxide and a polymer [88]. The basic structure of an OLED is made of thin emissive layers prepared from the organic compounds. This structure exhibits high luminous efficiency, since it does not require any backlight function. The basic physics can be explained as follows: OLEDs generate photons through the emissive layer; following the process of electron–hole pairs generation, recombination occurs and photons, hence light, are emitted. When a current is passed through the multiple layers, it transforms into light.

2.4.7.4

Organic Light-Emitting Transistor (OTFT)

A third major research area of thin film for energy applications is represented by OTFTs. When combined with an array of organic light-emitting diodes (OLEDs) they can be used for large-area display devices with mechanical flexibility and relatively low cost. The above example is one of several uses of organic light emitting transistors (OLETs), which can be considered a device with light emission capability along with the switching characteristics, thus providing a double function. Electroluminescent devices represent one of the current areas of R&D. OLETs have become also potential candidates for futuristic innovations in integrated circuits, processing both electrical and optical signals. An extensive description of the physics of OLETs and some of the most innovative applications can be found in reference [89]. It should be noted that the following examples represent only part of the possible future applications. The transistor function of these devices is analogous to that of standard inorganic semiconductors; however, one should keep in mind that the organic nature of these devices adds flexibility and provides the possibility of making inroads in the fundamental understanding of charge transport, injection, and recombination processes. The basic operation of a unipolar OLET having an organic semiconductor layer of tetracene in a bottom contact/bottom gate (BC/BG) structure (see Ref. [89]). OTFTs have been designed for all the classical transistor functions within digital logic circuits. One example is given by the inverter function, a basic standard application of inorganic semiconductor based transistors. However, contrary to their inorganic counterparts, organic inverters use only p-type transistors, since they exhibit higher mobility and better intrinsic stability in comparison to n-type materials [90–101]. Among the most significant work on OTFTs and OLETs to date, one can list a high-performance pentacene-based stacked TFT comparable in performance to of an a-Si:H transistor [92], the first fully organic active matrix organic light emitting diode (AMOLED) [93], a fully printed multilayer organic LED fabricated through the polymer inking and stamping technique [94], and other variations [95]; however, research and development of these organic devices has still to reach maturity. Other applications of OTFTs include NAND and NOR logic gates using a self-assembled monolayer of dielectric on the plastic substrate [96]; NAND and NOR gates operated at 40 V using poly(triarylamine)-based p-type and acene-diimide-based n-type

112

Thin Films

OTFTs [97]; cascade amplifiers [98]; single-ended differential amplifiers [99]; full printed complementary organic differential amplifiers using low-cost screen printing technology on flexible plastic substrate [100]; and RFID tag circuits [101–103]. It has been noted that OTFTs organic devices, with virtually endless combinations and compounds to embed system-on-chips using low cost, low temperature, and efficient process techniques, can create disruptive technology thanks to such flexibility. Sensors, for instance, provide a major area of application, as organic sensor can be used in the fields of medical diagnosis, industrial safety, household security, food safety, and environmental observations have drawn tremendous attention [104].

2.4.7.5

Thin Films as Selective Energy Contacts for Solar Cells

Thin films play a crucial role not only as the main materials employed in the fabrication of a solar cell, but also as enhancing components of PV and other energy devices. For instance, they can be used to fabricate energy selective contacts (ESC) for hot carrier solar cells. When carriers are excited by light absorption within the main semiconductor absorber, “hot” carriers will be created. The “hot” electrons and holes have energy way above the conduction band edge and below the valence band edge, respectively. These hot carriers will then go through a thermalization process that will bring their energy levels closer to the band edges, and will become subject to possible trapping and recombination processes. In summary, two types of losses can occur, the energy loss due to thermalization and the carrier loss due to trapping or recombination. Quantum well (QW) structures have been proposed to create a resonant energy tunnel that will capture the hot carriers before they have a chance to lose energy and become available for trapping or recombination. The proposed structures have been mostly analyzed theoretically, but experimental trials have been initiated in the past few years. These QWs are usually alternating layers of thin film materials, which include AlGaAS/GaAs layers, InP/PbSe layers, and InXGa1 XN/InN/InXGa1 XN layers. A detailed description of the physics of hot carriers and ESC can be found in Refs. [105–108].

2.4.7.6

Thin Films as Fuel Cell Components

Thin film have become also a major component in another energy device, the fuel cell, which is an electrochemical device that convert chemical energy of a fuel and oxidant into electrical energy. The advantage of fuel cells is that instead of using combustion, energy is produced via the electrochemical reaction, which makes the device more environmentally friendly. There are several types of fuel cells, including proton exchange membrane (PEM) fuel cells, where Hydrogen atoms are made to pass through a membrane and travel through a liquid electrolyte. On the other hand, a SOFC uses an electrolyte made of ceramic materials, and the cells operate at temperatures from 650 to 10001C, although the user only handles the cool surfaces. The research on thin film oxides for SOFC is very active [79,109]. Although research in this field has been progressing for over a decade, the technology is just reaching maturity and novel solutions based on alternative thin film materials are being sought. ALD has shown a good potential as the growth method for thin film oxides for the next-generation fuel cells, particularly SOFCs. A SOFC generally consist of three layers – a porous anode material, an ion-conducting electrolyte, and a cathode material. ALD can be used to grow all three layers. As current research is focused on reducing the operating temperatures from 800 to 10001C to ranges between 300 and 6001C, it is necessary to overcome issues related to the lower temperatures, like lower ionic conductivity and slower reaction kinetics at the electrodes. One solution consists in reducing the resistance of the electrolyte, which can be achieved by either synthesizing a material with higher ionic conductivity or by reducing the electrolyte film thickness. Lower thicknesses are linked also to faster ion transfer between electrodes. ALD has been used to deposit both electrolytes and catalysts (see Ref. [110] for a review of ALD for SOFC, and Refs. [111,112] for examples of SOFC layers grown by ALD). The most common electrolyte material used in SOFCs is in SOFCs is generally yttria stabilized zirconia (YSZ) or gadoliniadoped ceria (GDC), which are dense, ion-conducting ceramics. The conductivity of YSZ is generally considered too low for ITSOFCs, however, it is possible to achieve the desired values, even at lower temperatures, by reducing the electrolyte thickness to the nanometer scale. Several studies have investigated growth of YSZ by ALD, with focus on both improving ionic conductivity and significantly reducing the electrolyte film thickness. For instance, ALD of YSZ can be achieved through the use of alternating ZrO2 and Y2O3 ALD depositions to reach the desired YSZ composition [113,114].

2.4.7.7

Other Energy Storage Devices (Ultra/Super-Capacitors)

Thin film batteries have become a major area of research, going beyond fuel cell applications. One of the main issues in energy storage is to develop energy storage devices with both high energy and power density. Current super/ultra-capacitors have very high power density, but energy densities are one order less than conventional batteries. Ultra-capacitors have the advantage that they can be recharged many more times than batteries, limitation to the amount that can be stored makes these devices good for selected applications, where intense bursts of power are required, but they do not perform well for applications that require steady power over a long period. Current research is focused on developing thin film materials that can overcome such limitations. For instance, thin-film carbon ultra-capacitors have been implemented [115] with storage capacity three times that of conventional ultra-capacitor materials. Several forms and alloys of carbon have been investigated, including graphene, CNTs and nanofibers, silicon-carbides, etc. [116–119]. These carbon films are thin and can be made at temperatures as low as 2001C, which makes it possible to integrate them with flexible electronics and other applications, like RFID chips, chips used in digital watches, or the backside of solar cells in both portable devices and rooftop installations, to store power generated during the day for use after sundown.

Thin Films

2.4.8

113

Conclusions

The investigation of materials for energy applications has focused in the past on their macroscopic physical, chemical and structural characteristics. However, the need to provide low-cost processing with minimal material usage has led to the development of thin-film technology. An important aspect of thin films is the fact that not only they require less material and less expensive processing, but their properties, in particular their opto-electronic characteristics, are greatly dependent on the actual size of the film and on the processing technology. Moreover, low weight and mechanical flexibility have become a major asset for a variety of applications. Energy applications of thin-film, both inorganic and organic, can be found in optoelectronics devices, including transparent conductive films for antireflecting coatings in solar cells, battery electrodes for energy systems, fuel cells, desalination plants, sensors for energy, flat panel display, RFID tags, static random access memory (SRAM), differential amplifiers, ring oscillators, and flexible integrated circuits. This chapter has focused on the major aspects of thin film technology. It was shown that, when dealing with this class of materials, three main areas play a crucial role, namely: 1. Material deposition and processing. This is not a trivial subject, as the choice of deposition technique is dependent on the type of desired material, and on the required opto-electronic properties of the material, that is, different technologies and processes lead to different characteristics for the same material. Furthermore, the deposition and processing phase play a major role in the ultimate cost of the device, so that the choice of the growth technology is also linked to cost-efficiency analysis. The possibility of using inorganic thin film in energy applications has also allowed for a number of alternative, low cost, deposition techniques. 2. Material characterization. It has been demonstrated that thin film can impact current technological advances by their optical, electronic and structural properties. Optical based applications were the first to be investigated, as the nanoscale thickness of thin film is ideal for antireflective coatings. Indeed, it has been pointed out that mirrors were probably the first real application of thin films. Micro-electronics applications take advantage of the unique transport properties of these materials, while the nanostructure of the materials has a direct impact on the hardness, thermal conductivity and other properties. Therefore, it is important to be able to analyze the properties of these films with the appropriate technique, while interpretation of the data must take in account the nanoscale size of the structure. Feedback between characterization and processing is crucial to achieve optimum control of film growth and film properties. 3. Applications. The nanoscale dimensions allow for unique applications of thin film materials. Also, thin film technology opens new avenues with respect to the choice of materials. Both inorganic and organic materials have shown to possess unique properties for diverse uses. Novel low-cost organic electronic materials have been investigated for information storage, integrated circuits, and devices like antennas for RFID, thick-film sensors, or displays. Furthermore, alternative (e.g. flexible) substrates can be used for the development of novel solutions. However, the most impactful energy application of thin-film is definitely in PV. Novel thin-film materials have been developed with major improvements in the cost-efficiency ratio of solar cells, as indicated in this chapter. The study and development of the opto-electronic properties of these materials is crucial for achieving better efficiencies in solar cell technology, while keeping production costs and environmental impact low. Finally, energy storage devices, in the form of fuel cells and super capacitors, have also become a major area of application for thin films. In summary, one can identify several interdependent steps that need to be understood for thin-film related technology applications. The main crucial steps can be listed as follows: • Definition of the final application; • Choice of the substrate; • Definition of the critical properties of the substrate surface; • Development and implementation of the appropriate preparation process for the substrate (cleaning, morphology modification); • Selection of the material to be deposited and identification of the desired structure, thickness and purity of the film; • Choice of process parameters (speed of deposition process, maximum temperatures, adherence, etc.); • Control/monitoring of the processing technique; • Development, implementation of the appropriate characterization(s) technique(s) to validate application area; and • Possible development of re-processing techniques to repair eventual damage/defects.

References [1] Frey H. Relevance of the vacuum technology for thin film coatings. In: Frey H, Khan HR, editors. Handbook of thin-film technology. Berlin, Heidelberg: Springer-Verlag; 2015. p. 5–12. [2] Callister Jr. WD, Rethwisch DG. Materials science and engineering. 9th ed. Hobokem, NJ: Wiley; 2014. [3] Abou-Ras D, Kirchartz T, Rau U, editors. Advanced characterization techniques for thin film solar cells. Weinheim, Germany: Wiley; 2011. [4] Shah A. Thin-film silicon solar cells. Boca Raton, FL: CRC Press; 2010. [5] Lee S, Lee Y, Nam K, et al. Characterization of microcrystalline silicon thin film solar cells prepared by high working pressure plasma-enhanced chemical vapor deposition. J Electroceram 2014;33:149–54.

114

Thin Films

[6] Chu SS. Thin-film gallium arsenide homojunction solar cells. Appl Phys Lett 1981;39(10). [7] Higashi S, Sugakawa K, Kaku H, Okada T, Miyazaki S. Characterization of microcrystalline silicon thin film transistors fabricated by thermal plasma jet crystallization technique. Japan J Appl Phys 2010;49(3S). [8] Holzer K, Walling JS. One watt gallium arsenide class-E power amplifier with a thin-film bulk acoustic resonator filter embedded in the output network. J Eng 2015;1–3. doi:10.1049/joe.2015.0058. [9] Mcculloch I. Thin films: rolling out organic electronics. Nat Mater 2005;4:583–4. [10] Khuram A, Khan SA, Jafri MZM. Structural and optical properties of ITO/TiO2 anti-reflective films for solar cell applications. Nanoscale Res Lett 2014;9.1. [11] Raut HK, Ganesh VA, Nairb AS, Ramakrishna S. Anti-reflective coatings: A critical, in-depth review. Energy Environ Sci 2011;4:3779–804. [12] Mc Donnell SJ, Wallace RM. Atomically-thin layered films for device applications based upon 2D TMDC materials. Thin Solid Films 2016;616:482–501. [13] Saji KJ, Venkata Subbaiah YP, Tian K, Tiwari A. P-type SnO thin films and SnO/ZnO heterostructures for all-oxide electronic and optoelectronic device applications. Thin Solid Films 2016;605:193–201. [14] Dharmadasa IM. Advances in thin-film solar cells. Singapore: Pan Stanford Publishing Pte. Ltd.; 2013. [15] Jones AC, Hitchman ML. Chemical vapour deposition: precursors, processes and applications. London: RSC Publishing; 2009. [16] Morosanu CE, Siddall G. Thin films by chemical vapour deposition. Available from: http://dx.doi.org/10.1016/B978-0-444-98801-0.50001-X Amsterdam: Elsevier; 1990. [17] Frey H, Khan HR, editors. Handbook of thin film technology Berlin: Springer; 2015. [18] http://www.dowcorning.com/content/etronics/etronicschem/ [19] Zukotynski S, Gaspari F, Manage D, Pletnev V, Sagnes E. Hydrogenated amorphous carbon deposition by saddle-field glow discharge. Mat Res Soc Symp Proc 2000;595:239–48. [20] Reif R, Kern W. Plasma-enhanced chemical vapor deposition. In: Vossen JL, Kern W, editors. Thin film process II. London: Academic Press; 1991, p. 525. [21] Lieberman MA, Lichtenberg AJ. Principles of plasma discharges and materials processing. New York: Wiley Interscience; 1994. [22] Alexandrov SG, Hitchman ML. Plasma enhanced chemical vapour deposition processes. Chemical vapour deposition : precursors, processes and applications 2009 London: RSC Publishing 2009 (Chapter 2). [23] Rolinski E. Electrical Discharges in Gases and Principles of Ion Nitriding. Advanced Heat Treat Corp. 2009. Downloaded from: http://r.search.yahoo.com/ _ylt=A0LEVvo_0kZYTg0AS3YnnIlQ;_ylu=X3oDMTE0bXU5OWExBGNvbG8DYmYxBHBvcwM0BHZ0aWQDRkZVSTNDMV8xBHNlYwNzcg–/RV=2/RE=1481065152/RO=10/ RU=http%3a%2f%2f4hv.org%2fe107_files%2fpublic%2f1378344675_396_FT156949_electrical_discharges_in_gases_and_principles_of_ion_nitriding.pdf/RK=0/ RS=r7sUAWLO.kTAKLBdFs2xKJXgT20-. [24] http://piescientific.com/Resource_pages/Resource_DC_glow_discharge. [25] https://www.uantwerpen.be/en/rg/plasmant/research/research-topics/gas-discharge-plasma/. [26] http://www.azom.com/article.aspx?ArticleID=1558. [27] Frey H, Helmut T. Measurements of thin layers after the coating process. In: Frey H, Khan HR, editors. Handbook of thin-film technology. Berlin, Heidelberg: SpringerVerlag; 2015. p. 301–55. [28] Chu WK, Mayer JM, Nicole MA. Backscattering spectrometry. New York: Academic Press; 1978. [29] Vickerman JC, Winograd N. SIMS – a precursor and partner to contemporary mass spectrometry. Int J Mass Spectrom 2014;377:568–79. [30] http://www.pveducation.org/pvcdrom/characterisation/electroluminescence [31] Raguse J. Correlation of electroluminescence with open-circuit voltage from thin-film CdTe solar cells. J Photovolt 2015;4. doi:10.1109/JPHOTOV.2015.2417761. [32] Street RA, editor. Technology and applications of amorphous silicon. New York: Springer Verlag; 2000. ISBN: 3540657142. [33] Kasap S. Principles of electronic materials and devices. New York: McGraw-Hill 2005. Retrieved from http://Materials.Usask.Ca. [34] Baranovski S. Charge transport in disordered solids with applications in electronics. New York: John Wiley & Sons 2006. ISBN: 9780470095041. [35] Searle T, editor. Amorphous silicon and its alloys London: INSPEC; 1998. ISBN: 0852969228. [36] Street RA. Hydrogenated amorphous silicon. New York: Cambridge University Press 1991. ISBN: 0521371562. [37] Bruno G, Capezzuto P, Madan A, editors. Plasma deposition of amorphous silicon-based materials. Burlington, MA: Academic Press; 1995. ISBN: 9780121379407. [38] Gaspari F. Optoelectronic properties of amorphous silicon the role of hydrogen: from experiment to modeling. In: Predeep P, editor. Optoelectronics – materials and techniques. Rijeka: In Tech;2011. [39] Wehrspohn RB, Powell MJ, Deane SC, et al. Appl Phys Lett 2000;77:750. [40] Klein S, Finger F, Carius R, et al. Thin Solid Films 2003;430:202. [41] Moreno M, Ambrosio R, Torres A, et al.. 2016. Amorphous, polymorphous, and microcrystalline silicon thin films deposited by plasma at low temperatures. In: Mandracci P editor. Crystalline and non-crystalline solids. Rijeka: In Tech. doi:10.5772/63522. Available from: http://www.intechopen.com/books/crystalline-and-noncrystalline-solids/amorphous-polymorphous-and-microcrystalline-silicon-thin-films-deposited-by-plasma-at-low-temperature. [42] Smets AHM, Matsui T, Kondo M. High-rate deposition of microcrystalline silicon p-i-n solar cells in the high pressure depletion regime. J Appl Phys 2008;104:034508. [43] van Swaaij RACMM, Jiménez Zambrano R, Smit C, van de Sanden MCM. New insights in microcrystalline silicon deposition with expanding thermal plasma chemical vapor deposition. J Non-Cryst Solids 2006;352(9-20):933–6. [44] Zhang XD, Zhang FR, Amanatides E, Mataras D, Zhao Y. Effect of substrate bias on the plasma enhanced chemical vapor deposition of microcrystalline silicon thin films. Thin Solid Films 2008;516(20):6912–8. [45] Strobel C, Zimmermann T, Albert M, Bartha JW, Kuske J. Productivity potential of an inline deposition system for amorphous and microcrystalline silicon solar cells. Sol Energy Mater Sol Cells 2009;93(9):1598–607. [46] Johnson EV, Pouliquen S, Delattre PA, Booth JP. Hydrogenated microcrystalline silicon thin films deposited by RF-PECVD under low ion bombardment energy using voltage waveform tailoring. J Non-Cryst Solids 2012;358(17):1974–7. [47] Nakanoa Y, Goyaa S, Watanabea T, Yamashitaa N, Yonekurab Y. High-deposition-rate of microcrystalline silicon solar cell by using VHF PECVD. Thin Solid Films 2006;506-507:33–7. [48] Rath JK, Verkerk A, Franken R, et al., Implementation of highly stable microcrystalline silicon by VHF PECVD at high deposition rates in micromorph tandem cells. In: IEEE 4th world conference on photovoltaic energy conference, Waikoloa, Hawaii; 2006, doi:10.1109/WCPEC.2006.279747. [49] Vetterl O, Groß A, Jana T, et al. Changes in electric and optical properties of intrinsic microcrystalline silicon upon variation of the structural composition. J Non-Cryst. Solids 2002;299–302:772–7. [50] Abramov A, Roca i Cabarrocas P, Girotra K, et al. Reliable characterization of microcrystalline silicon films for thin film transistor applications. Japan J Appl Phys 2008;47(9R):7308–10. [51] Lin L. Integration of thin film gallium arsenide MSM photodetector in fully embedded board-level optoelectronic interconnects, ProQuest Dissertations Publishing, the University of Texas at Austin; 2004. 3144558. [52] Chakrabarti A. Transition from metastability to instability in the dynamics of phase separation. Phys Rev B: Condens Matter 1992;45(17):9620–5. [53] Blakemore JS. Semiconducting and other major properties of gallium arsenide. J Appl Phys 1982;53(10). [54] Murali KR, Jayachandran M, RangaRajan N. Review of techniques on growth of GaAs and related compounds. Bull Electrochem 1987;3(3):261–5. [55] Hayes GJ, Clemens BM. Laser liftoff of gallium arsenide thin films. MRS Commun 2015;5(1):1–5. [56] Flora SJS, Dwivedi N. A toxicochemical review of gallium arsenide. Defence Sci J 2012;62(2):95–104. doi:10.14429/dsj.62.1014. [57] Lei Q, Cai KF, Yao X. Preparation and characterization of copper indium selenide powders and films. J Electroceram 2008;21:508–11.

Thin Films

115

[58] Yamamoto A, Yamaguchi M, Uemura C. High conversion efficiency and high radiation resistance InP homojunction solar cells. Appl Phys Lett 1984;44(6):611. [59] Hoffman R, Scheiman DA, Ringel SA, et al. Development of high efficiency p þ /n InP solar cells for hetero-epitaxial applications; 1996. p. 171–4. [60] Zheng M, Yu Z, Joon Seok T, et al. High optical quality polycrystalline indium phosphide grown on metal substrates by metalorganic chemical vapor deposition. J Appl Phys 2012;111(12):123112. [61] Cuomo JJ, Gambino RJ. The synthesis and epitaxial growth of GaP by fused salt electrolysis. J Electochem Soc 1968;115(7):755–9. [62] Elwell D, Feigelson RS, Simkins MM. Electrodeposition of indium phosphide. J Cryst Growth 1981;51(2):171–7. [63] Tanaka K, Moritake N, Uchiki H. Sol Energy Mater Sol Cells 2007;91:1199. [64] Todorov T, Gunawan O, Chey SJ, et al. Thin Solid Films 2011;519:7378. [65] Nakayama N, Ito K. Appl Surf Sci 1996;92:171. [66] Scragg JJ, Dale PJ, Peter LM, Zoppi G, Forbes I. Phys Status Solidi (b) 2008;245:1772. [67] Horowitz G. Organic field-effect transistors. Adv Mater 1998;10(5):365–77. [68] Ishikawa Y, Wada Y, Toyabe T. Origin of characteristics differences between top and bottom contact organic thin film transistors. J Appl Phys 2010;107(5):053709–15. [69] Watanabe Y, Kudo K. Vertical type organic transistor for flexible sheet display. Proc SPIE 2009;7415:741515-1–10. [70] Veres J, Ogier S, Lloyd G. Gate insulators in organic field effect transistors. Chem Mater 2004;16(23):4543–55. [71] Kumar B, Kaushik BK, Negi YS. Organic thin film transistors: structures, models, materials, fabrication and applications – a review. Polym Rev 2014;54(1):33–111. [72] Li Y, Fan P, Zheng Z, et al. The influence of heat treatments on the thermoelectric properties of copper selenide thin films prepared by ion beam sputtering deposition. J Alloy Compd 2016;658:880–4. [73] Dong Y, He J, Li X, et al. Study on the preheating duration of Cu2SnS3 thin films using RF magnetron sputtering technique for photovoltaics. J Alloy Compd 2016;665:69–75. [74] Jayaraman VK, Kuwabara YM, Álvarez AM, Olvera-Amador M, de la L. Importance of substrate rotation speed on the growth of homogeneous ZnO thin films by reactive sputtering. Mater Lett 2016;169:1–4. [75] Kang BJ, Oh JH. Influence of substrate temperature and overlap condition on the evaporation behavior of inkjet-printed semiconductor layers in organic thin film transistors. Thin Solid Films 2016;598:219–25. [76] Wang F, Li T, Zhang H-C, Yuan CY. Exergy analysis of atomic layer deposition for Al2O3 nano-film preparation. In: Re-engineering manufacturing for sustainability; 2013. p. 285–90. doi:10.1007/978-981-4451-48-2_47. [77] Slesarenko VN. Comparison of the efficiency of MSF plants and thin-film desalination. Desalination 2003;158:295–302. [78] Gaur A, Tiwari GN. Analytical expressions for temperature dependent electrical efficiencies of thin film BIOPVT systems. Appl Energy 2015;146:442–52. [79] An J, Shim JH, Kim Y-B, et al. MEMS-based thin-film solid-oxide fuel cells. Low-Temp Solid-Oxide Fuel Cells 2014;39(9):798–804. [80] Gaur A, Tiwari GN. Exergoeconomic and enviroeconomic analysis of photovoltaic modules of different solar cells, J Sol Energy, 2014, Article ID 719424, 8 pages, 2014. doi:10.1155/2014/719424 [81] Caliskan H, Dincer I, Hepbasli A. Energy, exergy and sustainability analyses of hybrid renewable energy based hydrogen and electricity production and storage systems. Appl Therm Eng 2013;61(2):784–98. [82] Chen H, Lu H, Ren Q, et al. Enhanced photovoltaic performance of inverted pyramid-based nanostructured black-silicon solar cells passivated by an atomic-layerdeposited Al2O3 layer. Nanoscale 2015;7:15142–8. doi:10.1039/c5nr03353e. [83] Bailie CD, McGehee MD. High-efficiency tandem perovskite solar cells. MRS Bull, suppl. Perovskite Photovoltaics; Warrendale 2015;40(8):681–6. [84] Shockley W, Queisser HJ. Detailed balance limit of efficiency of p-n junction solar cells. J Appl Phys 1961;32:510–9. [85] Zhang Z, Wei D, Xie B, et al. High reproducibility of perovskite solar cells via a complete spin-coating sequential solution deposition process. Sol Energy 2015;122:97–103. [86] Forgács Dávid, Pérez-del-Rey Daniel, Ávila Jorge, et al. Efficient wide band gap double cation – double halide perovskite solar cells. J Mater Chem A 2017;5(7):3203. [87] Mantilla-Perez Paola, Feurer Thomas, Correa-Baena Juan-Pablo, et al. Monolithic CIGS–perovskite tandem cell for optimal light harvesting without current matching. ACS Photon 2017;4(4):861–7. [88] Jarvis KL, Evans PJ. Growth of thin barrier films on flexible polymer substrates by atomic layer deposition. Thin Solid Films 2017;624:111–35. [89] Kumar Kaushik B, Kumar B, Prajapati S, Mittal P. Organic light-emitting transistors. In: Organic thin-film transistor applications. Boca Raton, FL; 2016. p. 119–52. [90] Raval HN, Tiwari SP, Navan RR, Mhaisalkar SG, Rao VR. Solution processed bootstrapped organic inverters based on P3HT with a high-k gate dielectric material. IEEE Electron Device Lett. 2009;30(5):484–6. [91] Spijkman MJ, Myny K, Smits ECP, et al. Dual gate thin film transistors, integrated circuits and sensors. Adv Mater 2011;23(29):3231–42. [92] Zhou L, Wanga A, Wu SC, et al. All organic active matrix flexible display. Appl Phys Lett 2006;88(3):083502-1–3. [93] Li D, Guo LJ. Organic thin film transistors and polymer light-emitting diodes patterned by polymer inking and stamping. J Phys D: Appl Phys 2008;41(10):105115-1–7. [94] Chu CW, Chen CW, Li SH, Yanga Y. Integration of organic light emitting diode and organic transistor via a tandem structure. Appl Phys Lett 2005;86(25):253503-1–3. [95] Liu PT, Chu LW. Innovative voltage driving pixel circuit using organic thinfilm transistor for AMOLEDs. J Display Tech 2009;5(6):224–8. [96] Kim HS, Won SM, Won M, et al. Self-assembled nanodielectrics and silicon nanomembranes for low voltage, flexible transistors, and logic gates on plastic substrates. Appl Phys Lett 2009;95(18):183504-1–3. [97] Guerin M, Bergeret E, Benevent E, et al. Organic complementary logic circuits and volatile memories integrated on plastic foils. IEEE Trans Electron Devices 2013;60 (6):2045–51. [98] Gay N, Fischer WJ, Halik M, et al. Analog signal processing with organic FETs. Solid-State Circuits Conf. 2006;6:1070–9. [99] Marien H, Steyaert M, van Veenendaal E, Heremans P. Analog techniques for reliable organic circuit design on foil applied to an 18dB single-stage differential amplifier. Organic Electron 2010;11(8):1357–62. [100] Guerin M, Daami A, Jacob S, et al. High-gain fully printed organic complementary circuits on flexible plastic foils. IEEE Trans Electron Devices 2011;58(10):3587–93. [101] Baude PF, Enter DA, Haase MA, et al. Pentacene based radio frequency identification circuitry. Appl Phys Lett 2003;82(22):3964-1–3. [102] Cantatore E, Geuns TCT, Gelinck GH, et al. A 13.56-MHz RFID System based on organic transponders. IEEE J Solid-State Circuits 2007;42(4):84–92. [103] Myny K, Steudel S, Vicca P, et al. Plastic circuits and tags for HF radiofrequency communication. Solid State Electron 2009;53(12):1220–6. [104] Mabeck JT, Malliaras GG. Chemical and biological sensors based on organic thin-film transistors. Anal Bioanal Chem 2006;384(2):343–53. [105] Conibeer GJ, Jiang C-W, Shrestha S, Green MA. Selective energy contacts for hot carrier solar cells. Thin Solid Films 2008;516(20):6968–73. [106] Fenga Y, Aliberti P, Veettil BP, et al. Non-ideal energy selective contacts and their effect on the performance of a hot carrier solar cell with an indium nitride absorber. Appl Phys Lett 2012;100:053502. [107] Emelianov V, Konovalov I. Towards the energy selective contacts electrochemical produced n-InPPbSe heterostructures. Available from: https://www.researchgate.net/ publication/280310950; 2015. [108] Su S, Liao T, Chen X, Su G, Chen J. Hot-carrier solar cells with quantum well and dot energy selective contacts. IEEE J Quantum Electron 2015;51(9):1–8. [109] Beckel D, Bieberle-Hütter A, Harvey A, et al. Thin films for micro solid oxide fuel cells. Journal of Power Sources 2007;173(1):325–45. [110] Johnson RW, Hultquist A, Bent SF. A brief review of atomic layer deposition: from fundamentals to applications. Mater Today 2014;17(5):236–46. [111] Holme TP, Lee C, Prinz FB. Atomic layer deposition of LSM cathodes for solid oxide fuel cells. Solid State Ion 2008;179(27–32):1540–4. [112] Brahim C, Ringuedé A, Cassir M, Putkonen M, Ninisto L. Electrical properties of thin yttria-stabilized zirconia overlayers produced by atomic layer deposition for solid oxide fuel cell applications. Appl Surf Sci 2007;253(8):3962–8.

116

Thin Films

[113] Lee GB, Son KS, Park SW, Shim JH, Choia B-H. Low-temperature atomic layer deposition of Al2O3 on blown polyethylene films with plasma-treated surfaces. J Vac Sci Technol A: Vac Surf Films 2013;31:01A129. [114] Bernay C, Ringuedé A, Colomban P, Lincot D, Cassir M. Yttria-doped zirconia thin films deposited by atomic layer deposition ALD: a structural, morphological and electrical characterisation. J Phys Chem Solids 2003;64(9-10):1761–70. [115] Liu J, Mirri F, Notarianni M, Pasquali M, Motta N. High performance all-carbon thin film supercapacitors. J Power Sources 2015;274:823–30. [116] Sun D, Yan X, Lang J, Xue W. High performance supercapacitor electrode based on graphene paper via flame-induced reduction of graphene oxide paper. J Power Sources 2013;222:52–8. [117] Zhou Z, Wu X-F. Graphene-beaded carbon nanofibers for use in supercapacitor electrodes: synthesis and electrochemical characterization. J Power Sources 2013;222:410–6. [118] Kim SY, Kim B-H, Yang KS, Oshida K. Supercapacitive properties of porous carbon nanofibers via the electrospinning of metal alkoxide-graphene in polyacrylonitrile. Mater Lett 2012;87:157–61. [119] Alper JP, Kim MS, Vincent M, et al. Silicon carbide nanowires as highly robust electrodes for micro-supercapacitors. J Power Sources 2013;230:298–302.

Further Reading Abou-Ras D, Kirchartz T, Rau U, editors. 2011. Advanced characterization techniques for thin film solar cells Weinheim, Germany: Wiley-VCH Verlag GmbH & Co. KgaA; 2011. http://dx.doi.org/10.1002/9783527636280.fmatter Böer KW. Handbook of the physics of thin-film solar cells. New York: Springer; 2013. Brundle CR, Evans CA, Wilson S. Encyclopedia of materials characterization: surfaces, interfaces, thin films. Boston: Butterwirth-Heinemann; 1992. Cao Z, Homma Y, Miller T, Chiang T-C, Voigt A. Thin film growth, physics, materials science and applications. Cambridge: Woodhead Publishing Limited; 2011. Available from: http://dx.doi.org/10.1016/B978-1-84569-736-5.50018-9. Di Bartolo B, Collins J. Principles of photoluminescence, in handbook of luminescent semiconductor materials. CRC Press; 2011. p. 1–20. ISBN: 978-1-4398-3467-1 (Chapter 1). Frey H, Khan HR, editors. 2015. Handbook of thin film technology Berlin: Springer; 2015. Hollas JM. Modern spectroscopy. Chichester, England: John Wiley & Sons; 2004. ISBN-13:978 0 470 84416 8. Krebs FC, Jørgensen M. Polymer and organic solar cells viewed as thin film technologies: What it will take for them to become a success outside academia. Sol Energy Mater Sol Cells 2013;119:73–6. Ley L. Photoemission and optical properties. In: Joannopoulos JD, Lucovski G, editors. The physics of hydrogenated amorphous silicon. vol II. New York: Springer-Verlag; 1983. ISBN:0387128077. Lieberman MA, Lichtenberg AJ. Principles of plasma discharges and materials processing. New York: Wiley Interscience; 1994. Lukaszkowicz K. Review of nanocomposite thin films and coatings deposited by PVD and CVD technology. In: Rahman Mohammed Muzibur, editor. Nanomaterials; 2011. ISBN 978-953-307-913-4. Piegari A, Flory F, editors. 2013. Optical thin films and coatings Cambridge: Woodhead Publishing Limited; 2013. Ryou J, Kanjolia R, Dupuis RD. CVD of III–V compound semiconductors. Chemical vapour deposition: precursors, processes and applications 2009 London: RSC Publishing 2009 (Chapter 6). Shah A, editor. 2010. Thin-film solar cells Boca Raton, FL: CRC Press; 2010. Vickerman JC (editor). Surface analysis – the principal techniques, ISBN-13:978 0471 97272 1.

Relevant Websites https://www.intechopen.com/books/modern-technologies-for-creating-the-thin-film-systems-and-coatings/advance-deposition-techniques-for-thin-film-and-coating An Article on Advanced Deposition Techniques for Thin Film and Coating by Asim Jilani et al. https://www.electrochem.org/dl/interface/fal/fal08/fal08_p44-48.pdf An Article on Thin Film Micro-Batteries by Nancy J. Dudney. http://news.energysage.com/thin-film-solar-panels-make-sense/ An Economic Comparison of Thin-Film Solar Panels vs. Standard Solar Technology. http://irjms.in/sites/irjms/index.php/files/article/view/163/163 A Review Article of Metal Oxide Thin Film Based Supercapacitors, by Abhinandan A. Deshmane and R.B. Bhosale. http://www.sciencedirect.com/science/article/pii/S1369702115000826 A 2015 Review Article on Recent Advances and Remaining Challenges in Thin-Film Silicon Photovoltaic Technology, by F. Meillaud et al. http://pubs.rsc.org/en/content/articlelanding/2015/ee/c4ee03346a#!divAbstract A Review on Light Management in Thin Film Silicon Solar Cells, by F.-J. Haug and C. Ballif. http://icrowdnewswire.com/2017/10/10/organic-thin-film-transistors-otfts-market-global-market-size-analysis-and-business-research-report-2017/ A Website Dedicated to the Organic Thin-Film Transistors Market. http://www.pveducation.org/ A Website on the Physics of Solar Cells, With Several Sections on Thin Film PV, Tandem Cells, etc.. http://thinfilm.no/news-press-releases/ A Website With the Latest Press Releases in the Thin-Film Industry. https://www.nature.com/subjects/surfaces-interfaces-and-thin-films Latest Research and Reviews, Nature.com. http://www.mrsec.harvard.edu/education/ap298r2004/Erli%20chenFabrication%20II%20-%20Deposition-1.pdf Lecture on Thin Film Deposition from Harvard University. https://www.youtube.com/watch?v=qCSIGejNT4M U-Tube Vides on Semiconductor Fabrication Basics – Thin Film Processes, Doping, Photolithography, etc

2.5 Photovoltaic Materials Franco Gaspari and Simone Quaranta, University of Ontario Institute of Technology, Oshawa, ON, Canada r 2018 Elsevier Inc. All rights reserved.

2.5.1 Introduction 2.5.2 Background 2.5.3 Systems, Applications and Materials 2.5.3.1 Silicon-Based Photovoltaics 2.5.3.1.1 Solar grade silicon 2.5.3.1.2 Czochralski, zone floating and multi-crystalline silicon processes 2.5.3.2 Screen Printed Solar Cells 2.5.3.2.1 Rear contact cells 2.5.3.2.2 Passivated emitter, rear locally diffused cell 2.5.3.2.3 Bifacial solar cells 2.5.3.2.4 Heterojunction with intrinsic thin layer cell 2.5.3.2.5 Hydrogenated amorphous silicon (a-Si:H) and microcrystalline silicon (mc-Si:H) 2.5.3.3 Other “Conventional” Photovoltaic Materials 2.5.3.3.1 Gallium arsenide 2.5.3.3.2 Indium phosphide (InP) and gallium indium phosphide (GaInP) 2.5.3.3.3 Copper–indium–(gallium)–selenide (CIS and CIGS) 2.5.3.3.4 Copper zinc tin sulphide 2.5.3.3.5 Cadmium telluride (CdTe) 2.5.3.4 Organic Photovoltaics 2.5.3.4.1 Electron–hole exciton pairs in organic photovoltaics 2.5.3.4.2 Recombination mechanisms in organic photovoltaics 2.5.3.4.3 Inorganic semiconductors versus molecular semiconductors 2.5.3.4.4 Charge transfer and separation states in organic photovoltaic 2.5.3.5 Dye-Sensitized Solar Cells 2.5.3.6 Perovskite Solar Cell 2.5.4 Analysis and Assessment 2.5.5 Examples 2.5.6 Discussion and Future Developments 2.5.7 Closing Remarks Acknowledgment References Further Reading Relevant Websites

Nomenclature Symbol/Acronym Al Aluminum ARC Antireflection coating AM(x) Air mass (x¼0, 1, 1.5, etc.) a-Si:H Hydrogenated amorphous silicon BIPV Building integrated photovoltaics cs,l Concentration of solid (or liquid) phase in solutions CB, VB Conduction band, valence band CdSe Cadmium selenide CdTe Cadmium telluride CIS Copper indium selenide CIGS Copper–indium–gallium–selenide CZ Czochralski Method CZTS Copper–zinc–tin–sulfide DSSC Dye-sensitized solar cell EGS Electronic grade silicon

Comprehensive Energy Systems, Volume 2

Fe FZ Ga GaAs GaP GaN GaSb Ge HIT HOMO IC InGaP InP ISCS kd LID LUMO MCS

doi:10.1016/B978-0-12-809597-3.00215-7

118 118 123 123 126 128 130 132 132 132 134 134 135 135 135 136 137 137 137 138 138 138 139 139 141 143 143 145 146 146 146 149 149

Iron Float-zone method Gallium Gallium arsenide Gallium phosphide Gallium nitride Gallium antimonide Germanium Heterojunction with intrinsic thin layer Highest occupied molecular orbital Integrated circuit Indium gallium phosphite Indium phosphite Inorganic semiconductors Coefficient of solid/liquid partition in solutions Light induced defects Lowest unoccupied molecular orbital Molecular semiconductors

117

118

Photovoltaic Materials

MGS MLCT nc-Si P PECVD S–Q Si

2.5.1

Metallurgical grade silicon Metal-to-ligand charge transfer Nanocrystalline silicon Phosphorus Plasma enhanced chemical vapor deposition Shockley–Queisser Silicon

SiO2 SnO2 So-GS TiO2 V2O5 WO3

Silicon oxide Tin oxide Solar grade silicon Titanium dioxide Vanadium oxide Tungsten oxide

Introduction

The continuing pursuit to achieve sustainable alternative forms of energy production through photovoltaics (PVs) and related technologies that are pollution free, storable, and economical, has produced a variety of solutions. However, each PV technology suffers from either energy conversion efficiency, cost, or stability limitations. For instance, while silicon-based single-junction solar cells (SCs) are still far from their fundamental efficiency limit given by the Shockley–Queisser (S–Q) theorem [1–3], their manufacture and optimization technology seem have reached a point where further energy conversion efficiency’s breakthroughs are improbable. On the other hand, although non-silicon-based solar cells are attractive in terms of energy conversion efficiency capability, plenty of issues connected to their manufacturing and real condition applications remain. For example, the high production cost of multi-junction solar cells relying on III–V semiconductors renders their use for terrestrial applications unfeasible. Conversely, low-cost thin films solar cells are stricken with environmental issues (e.g., CdTe-based devices), lack of materials availability (CIGS, copper indium gallium selenide solar cells), or long-term stability problems (e.g., perovskite thin film solar cells [4–6] and polymer solar cells [7,8]). Finally, hybrid inorganic–organic solar cells (e.g., bulk heterojunctions and dye-sensitized solar cells – DSSCs) have shown little promise of acceptable improvements in the efficiency [9,10]. In recent years, PV research has approached the major issues concerning the development and integration of solar technology by: 1. Improving the current solar cell design and finding new ones in order to increase efficiencies. 2. Employing new materials and processing technologies to lower the cost of the device. Both directions aim at lowering the $/Wpeak ratio (i.e., the cost in US dollars of the overall solar panel fabrication cycle to produce a Watt of electrical power under some predefined conditions, i.e., standard AM1.5 illumination: 1000 W m 2 with $=m2 $ ¼ Z
1000W 37-degree tilt and temperature of 251C). Such a ratio can be calculated through the formula: Wpeak 2 , where Z is the peak =m 2 solar energy conversion efficiency and the area of the panel is expressed in m . On the other hand, increasing the power density, i.e., the efficiency, as a standalone goal is still a major aspect of PV research, since there exist several “niche” applications where the available area is limited (e.g., spacecrafts), and therefore performance becomes a priority over cost. It is also important to point out that solar technology is expanding its scope beyond simple direct energy conversion. In fact, PV cells and modules are also considered for integrated photocatalysis and photoelectrolysis. For this purpose, PV devices can be used as standalone energy providers, or in conjunction with a semiconductor–liquid junction [11,12]. In the latter case, large-band-gap semiconductors belonging to the group of transition metals oxides (TiO2, SnO2, WO3, V2O5, etc.) are usually employed as photoanodes to drive the water oxidation reaction, whereas a solar cell supplies the additional voltage required to carry out the water molecule’s splitting [13]. In this chapter, a review of the most important materials used in PVs (and related technologies) will be presented. A description of the main features of different PV materials and solar cells structures is provided. Special emphasis is placed on the advantages and disadvantages related with each material and solar cell configuration.

2.5.2

Background

In order to discuss the properties and device application of PV materials, it is necessary to keep in mind that “basic” PVs is restricted by the S–Q limit [14]. The authors established a theoretical limit for the performance of solar cells based on the following seven assumptions:

• • • • • • •

A single p–n junction (with a single band gap). One electron–hole pair excited per incoming photon. Incoming photons carrying an amount of energy lower than the semiconductor band-gap are not absorbed. Excess energy compared to the semiconductor band-gap is transferred to the electron–hole pairs but is lost by thermal relaxation. Infinite carriers mobility. Only radiative recombination of electron–hole pairs occurs (no defects-mediated or Auger recombination). Sun and solar cells are considered as black body at 6000 and 300K, respectively. Illumination with un-concentrated sunlight in standard condition (AM 1.5 and 300K).

Photovoltaic Materials

119

Based on these assumptions, it is possible to calculate the maximum theoretical efficiency of a solar cell based on the electronic properties of the material, i.e., the energy gap (EG). An efficiency limit E33.7% has been calculated for an ideal material possessing a 1.4 eV band gap [14,16]. Essentially, choosing a single-junction PV material requires that all the S–Q assumptions are either maximized (one electron–hole pair excited per incoming photon), minimized (thermal relaxation) or optimized (by optical design to optimize light absorption). When determining the potential of a semiconductor for PV applications, it is important to know, among other characteristics, whether we are dealing with a direct or indirect bandgap, and the material’s absorption coefficient to determine the needed thickness of material, and the processing cost. Fig. 1 shows the relation between band-gap and PV cell efficiency limit. It turns out that c-silicon is the favoured PV material even though its indirect bandgap is too narrow. Gallium arsenide is ideal technically but marginal economically. Historically, single-junction mono- or polycrystalline Si (EgE1.1 eV) cells have been dominating the market (see Fig. 2). Although, record efficiencies of 25.6% have been reported [15], silicon solar cells are still far from their theoretical efficiency limit (i.e., 32%). The discrepancy between the S–Q limit predictions and real devices performance stem from the incompleteness of the S–Q approach. In fact, the S–Q assumptions account only for the losses 0.35 0.30 c-Si

Efficiency

0.25 0.20

mc-Si

GaAs

InP AlGaAs

0.15 0.10 0.05

CdTe GaSb a-Si Ge

1.5G efficiency limit

0.00 0.8

1.2 Bandgap (eV)

1.6

2.0

Fig. 1 Solar cell efficiency vs. semiconductor band-gap. Available from http://www.physics.usyd.edu.au/app/solar/research/images/fig1.jpg.

a-Si glass/glass 2%

All others 1%

CIS/CIGS sputter 2%

c-Sip-type multistandard 35%

CdTe (first solar) 4% c-Sin-type 6% c-Sip-type monoadvanced 10%

c-Sip-type multiadvanced 27%

c-Sip-type monostandard 14% Fig. 2 2014 Solar PV Module Production by Technology. Goetzberger, A., Hebling, C., & Schock, H. Photovoltaic materials, history, status and outlook. Note that c-Si based technology made up 92% of the market. (Source: NPD Solarbuzz, PV Equipment Quarterly). Available from: http:// www.renewableenergyworld.com/articles/2016/04/2015-top-ten-pv-cell-manufacturers.html.

120

Photovoltaic Materials

due to the spectral mismatch between the semiconductor band-gap and the solar radiation, solar cell’s radiation re-emission, leakage currents stemming from radiative recombination (open-circuit voltage losses). Hence, the S–Q limit leaves out some practical limitations with huge detrimental effects on solar cells’ efficiency: 1. Non-radiative recombination. For indirect band-gap semiconductors (i.e., silicon), the annihilation of electron–holes pairs either through crystalline lattice impurities (Shockley–Reed–Hall recombination) or by excitation of Auger electron constitute the main recombination paths. 2. Optical losses. Light absorption by PV materials is limited by reflection, shading (i.e., by cells’ contacts) and finite cell’s thickness. 3. Parasitic resistances. Series resistance due to the materials bulk resistivity, contacts and charge collectors mainly hinders the cell’s current output (maximum power point and short circuit current). Similarly, shunt resistance brought about by multiple causes (i.e., improper junction insulation, excessive firing temperature of the metallization screen printing paste, base’s or emitter’s crack and holes, etc.) mainly affect the cell’s open-circuit voltage. Because of large impact of the additional losses on the PV efficiency, most of the research has been devoted to the reduction of bulk and surface carriers recombination (i.e., control of dopants concentration and diffusion passivation of cells front and rear surfaces, back surface field (BSF) layers), to the optimization of antireflecting coatings, to either the enhancement of screen printing metallization pastes` performances or to the development of alternative metallization techniques such electroless deposition and photolithography-free electroplating [17,18]. On the other hand, PVs research has also been focusing on exceeding the S–Q limit by violating one or more of its assumptions. For instance, photons up-conversion [19,20], hot carriers capture [21], tandem or multi-junction configurations [22], or solar concentrators [23] have been implemented. However, many of these solutions are quite expensive, and the $/Wp ratio is still not ideal. A useful way of illustrating the evolution of PV research was proposed by Martin Green [13] in terms of efficiency and cost, as shown in Fig. 3. First generation cells are characterized by high efficiency achieved by using low defect materials (mono-crystalline silicon). Second generation devices are comprised of thin film solar cells. Instead of p–n junction obtained by different crystal growth methods (Czochralski, zone floating, etc.) this technology relies on thin (1–5 mm) films of semiconductors (p-type CdTe, n-type CdS, n-type CIGS, etc.) deposited on a transparent conductive substrate (e.g., ITO). The p–n junction can be formed between two different compounds (i.e., p-type CdTe and n-type CdS) and the cost is much lower than first generation solar cells, although the efficiency is less. Third generation solar cells emcompasses multiple PV technologies: 1. Multi-junction solar cells. Fabrication of multilayer structures represent the most used method to overcome the S–Q limit. By relaxing the single-junction condition and by assuming an infinite number of p–n junction (i.e., layers) the theoretical energy conversion efficiency increases from 32% to 66% under non-concentrated radiation and reaches 86% for concentrated systems. In fact, multiple layers of semiconducting materials possessing different band gaps and absorption coefficients allow for the generation of electron–hole pairs exploiting a wide range of wavelengths in the solar spectrum. Multi-junction devices can be classified according to their number of deposited thin layers, usually 2 or 3, the kind of semiconductor (e.g., III–V semiconductors, heterojunctions comprised of crystalline and amorphous silicon, etc.), or the films’ deposition technique (e.g., metallo-organic chemical vapor deposition, atomic layer deposition, molecular beam epitaxy). The most studied multijunction systems are epitaxial GaAs/InP/Ge structures. 100

$0.10/Wp

$0.50/Wp

$0.20/Wp

Percent efficiency

80

60 $1.00/Wp

III 80 Shockley−Queisser limit 20 I

$3.50/Wp

II 100 200 300 400 Cost in dollars per square meter

500

Fig. 3 The three generation of solar cells. Adapted from Green M. Third generation photovoltaics. In: Springer series in photonics, vol. 12. Berlin/ Heidelberg: Springer; 2013 (Tomkiewicz M, Fay H. Photoelectrolysis of water with semiconductors. Appl Phys 1979;18).

Photovoltaic Materials

121

2. Bulk heterojunction solar cells based on the combination of polymers or hybrid organic–inorganic materials and structures (e.g., copper phtalocyanine, fullerenes, pentacene, polythiophene, and thiazole derivatives, etc.) possessing complementary electronic properties; similarly to conventional p–n junctions electron–hole pairs’ separation (geminate pair separation) and charge collection rely on the electrochemical potential drop at the interface between a donor (low electron affinity) and an acceptor material (high electron affinity). 3. Dye-sensitized solar cells. Such devices do not involve neither the generation of geminate pairs (as in bulk heterojunction solar cells) nor the depletion layer formation (as in p–n junction devices). Indeed, charge generation is predicated on the variable oxidation state of metallo-organic (i.e., Ru-polypyridine complexes) or metal-free (e.g., porphyrine, phenothiazine, quinoxaline, indoline, etc.) “dyes”. Furthermore, charge separation and transport are driven by energy bands position at the different heterojunctions (i.e dye/semiconductor, semiconductor/electrolyte, transparent conductor/semiconductor) and by the different kinetics of the processes involved. 4. Perovskite solar cells. Methylammonium lead halides perovskite (CH3PbNH3X3; X¼ Br, I, Cl) have been originally used instead of ruthenium-based sensitizers to enhance TiO2 light absorption properties in conventional electrolyte-based DSSCs. Then, full solid-state devices have been developed by coupling methylammonium lead chloride and bromide perovskites with p-type semiconductor polymers (e.g., spiro-OMETAD). The ambipolar nature of the electrical conductivity in perovskites absorbers has led to high efficiencies (22.1%, March 2016) through planar p-i-n structures. Such a configuration relies on a 200–300 nm perovskite layer (serving as the intrinsic semiconductor), a n-type contact (usually a 30–50 nm compact, planar layer of TiO2), and a spiro-OMETAD holes conductor. Although there exist numerous challenges concerning material’s chemical stability (i.e., light, temperature and moisture induced degradation due to the polar structure of CH3PbNH3X3) and device’s performance reproducibility (I–V characteristic hysteresis), perovskite-based solar cells remain the most attractive third generation device because of their ability to conjugate relatively low production costs with high energy conversion efficiency. Moreover, tunable optical band-gap and the possibility to improve their chemical stability by replacing protons with alkyl groups, render ammonium lead halide perovskite materials suitable for tandem cell operation (coupled to CIGS or high efficiency polycrystalline solar cells). In Fig. 4 a chart on the efficiency trends of solar cells is presented, updated to 2015. The chart, published annually by the National Renewable Energy Lab (NREL – a research structure affiliated with the US Department of Energy), represents different materials and technologies. Once again, one must associate the efficiency to the production costs in order to properly assess the potential of a particular cell material or design. Table 1 summarizes the best efficiencies according to classification (material and device structure) as of 2016 [14].

Best research-cell efficiencies 50 48 44 40 36 Efficiency (%)

32 28 24 20 16 12 8 4 RCA

0

1975

RCA

1980

1985

1990

1995

2000

2005

2010

2015

Fig. 4 National Renewable Energy Laboratory (NREL) graph relating different solar cells “generations” with their efficiency. Available from: http:// static1.1.sqspcdn.com/static/f/372352/22413458/1365584280200/efficiency_chart.jpg?token¼dCU0zFaQtzhhuodS7TJGxuE2kx0%3D.

122

Photovoltaic Materials

Table 1 Confirmed terrestrial cell and submodule efficiencies measured under the global AM1.5 spectrum (1000 W/m2) at 251C (IEC 60904-3: 2008, ASTM G-173-03 global) Efficiency (%)

Area (cm2)

Voc (V)

Jsc (mA/ cm2)

Fill factor (%)

Test center (date)

Description

25.670.5

143.7 (da)

0.740

41.8a

82.7

AIST (2/14)

21.370.4 21.270.4

242.74 (t) 239.7 (ap)

0.6678 0.687c

39.80 38.50

80.0 80.3

FhG-ISE (11/15) NREL (4/14)

10.570.3

94.0 (ap)

0.492

29.7

72.1

FhG-ISE (8/07)

Panasonic HIT, rear junction Trina Solar Solexel (35mm thick) CSG Solar (o2 mm on glass)

III–V cells GaAs (thin film cell) GaAs (polycrystalline) InP (crystalline cell)

28.870.9 18.470.5 22.170.7

0.9927 (ap) 4.011 (t) 4.02 (t)

1.122 0.994 0.878

29.68 23.2 29.5

86.5 79.7 85.4

NREL (5/12) NREL (11/95) NREL (4/90)

Alta Devices RTI, Ge substrate Spire, epitaxial

Thin film chalcogenide CIGS (cell) CIGS (minimodule)

21.070.6 18.770.6

0.9927 (ap) 15.892 (da)

0.757 0.701

35.70 35.29

77.6 75.6

FhG-ISE (4/14) FhG-ISE (9/13)

CdTe (cell)

21.070.4

1.0623 (ap)

0.8759

30.25

79.4

Newport (8/14)

CZTSSe (cell) CZTS (cell)

9.870.2 7.670.1

1.115 (da) 1.067 (da)

0.5073 0.6585

31.95 20.43

60.2 56.7

Newport (4/16) NREL (4/16)

Solibro, on glass Solibro, 4 serial cells First Solar, on glass IMRA Europe UNSW

Amorphous/microcrystalline Si (amorphous cell) 10.270.3 Si (microcrystalline cell) 11.870.3

1.001 (da) 1.044 (da)

0.896 0.548

16.36 29.39

69.8 73.1

AIST (7/14) AIST (10/14)

AIST AIST

Perovskite Perovskite (cell)

19.770.6

0.9917 (da)

1.104

24.67

72.3

Newport (3/16)

KRICT/UNIST

Dye sensitized Dye (cell) Dye (minimodule)

11.970.4 10.770.4

1.005 (da) 26.55 (da)

0.744 0.754

22.47 20.19

71.2 69.9

AIST (9/12) AIST (2/15)

Dye (submodule)

8.870.3

398.8 (da)

0.697

18.42

68.7

AIST (9/12)

Sharp Sharp, 7 serial cells Sharp, 26 serial cells

Organic (polymer based) Organic (cell) Organic (minimodule)

11.270.3 9.770.3

0.992 (da) 26.14 (da)

0.780 0.806

19.30i 16.47

74.2 73.2

AIST (10/15) AIST (2/15)

Toshiba Toshiba (8 series cells)

38.87 71.2

1.021 (ap)

4.767

9.564

85.2

NREL (7/13)

Spectrolab

37.971.2 34.572.0

1.047 (ap) 27.83 (ap)

3.065 2.66/0.65

14.27 13.1/9.3

86.7 85.6/79.0

AIST (2/13) NREL (4/16)

Sharp UNSW/Azur/Trina

31.671.5 29.871.5

0.999 (ap) 1.006 (da)

2.538 1.46/0.68

14.18 14.1/22.7

87.7 87.9/76.2

NREL (1/16) NREL (10/15)

13.670.4

1.043 (da)

1.901

9.92

72.1

AIST (1/15)

Alta Devices NREL/CSEM, 4terminal AIST

12.770.4

1.000(da)

1.342

13.45d

70.2

AIST (10/14)

AIST

Classification

Silicon Si (mono-crystalline cell) Si (polycrystalline cell) Si (thin transfer submodule) Si (thin film minimodule)

Multijunction Five junction cell (bonded) (2.17/1.68/ 1.40/1.06/0.73 eV) InGaP/GaAs/InGaAs GaInP/GaInAs/Ge; Si (minimodule) GaInP/GaAs (monolithic) GaInP/Si (mech. stack) a-Si/nc-Si/nc-Si (thin film) a-Si/nc-Si (thin film cell)

Abbreviations: CIGS, CuIn1-yGaySe2; a-Si, amorphous silicon/hydrogen alloy; nc-Si, nanocrystalline or microcrystalline silicon; CSTSS, Cu2ZnSnS4-ySey; CZTS, Cu2ZnSnS4; (ap), aperture area; (t), total area; (da), designated illumination area; FhG-ISE, Fraunhofer Institut für Solare Energiesysteme; AIST, Japanese National Institute of Advanced Industrial Science and Technology. Source: Adapted from Green, Hishikawa, Warta et al. [14]. Original references and details have been omitted, and can be found in Shockley W, Queisser HJ. Detailed balance limit of efficiency of p-n junction solar cells. J Appl Phys 1961;32:510.

Photovoltaic Materials

2.5.3

123

Systems, Applications and Materials

Table 1 lists a large assortment of materials, combination of materials, cells configurations and structures. Systems and applications of PV materials are obviously related to the fabrication of solar panels to be integrated with the desired infrastructure or to be used in an independent setting. Solar panels can be employed in a variety of situations and for different purposes. For instance, solar power stations (i.e., solar farms), solar panels for orbiting satellites or other spacecrafts, housing installed PV to supply power to the electrical utility grid, solar cells to activate switches and power up consumer electronic devices are just a few applications of the modern PVs technology. Each specific application requires a proper combination of PV material’s properties, device size and weight, and production’s cost affordability. For example, portable electronics power supply systems’ alternative to batteries are usually fabricated on flexible plastic substrates [22]. Therefore, thin film PV materials that can be deposited at low temperature (e.g., amorphous silicon and nanocrystalline silicon, [23,24]) and capable of sustaining the mechanical stress of the flexible substrate (e.g., light absorbing polymer layers) are mandatory [25,26]. Furthermore, the techniques employed for the fabrication of such devices need to be compatible with the low production cost characteristic of consumer electronics. Thus, since the high-resolution feature specifications (e.g., ultrathin dielectric layers, conductors’ line width on the order of hundreds of nanometers, etc.) dictated by the continuous miniaturization of microelectronic devices do not apply to the PVs market, a plethora of contact (e.g., flexography, gravure, offset and screen printing) and non-contact (i.e., inkjet) printing techniques have been adapted to the production of solar cells and modules [27,28]. Concurrently, also well-established PV technologies such as traditional multi-crystalline and monocrystalline silicon are going through a profound modification of their manufacturing processes to fulfill the expectations of a largescale use of solar modules even in the least developed countries. For instance, most of the actual scientific research concerning crystalline silicon aims at implementing low-cost silicon purification routes and at replacing contacts, grids and busbars precious metals (i.e., silver) with cheaper conductors such as copper, nickel, and tin [29,30]. Two examples which are representative of particular PV materials’ requirements needed for specific applications are concentrated [31] and bifacial PVs [32]. In fact, the former is most suited for power generation in areas with high direct normal irradiance, while the latter is amenable for both vertically installed systems and power generation in areas possessing large ground’s reflectivity (i.e., snow-covered soils). Needless to say, materials optoelectronic properties and encapsulation technologies are significantly different for the two cases. High-efficiency multi-junction solar panels based on binary and ternary III–V semiconductors (i.e., GaAs, InP, AlGaAs, InGaAp, etc.) are the most used devices for concentrated PVs because of their lower temperature coefficient compared to silicon modules. Indeed, III–V semiconductors are less prone to carriers’ lifetime degradation as the temperature increases, and possess larger absorption coefficient stemming from their direct bands structure. Therefore, the possibility of achieving high energy conversion efficiencies under high irradiance conditions (up to 44%) together with small device active areas compensate for the higher production cost and make multi-junction concentrated systems economically feasible. On the other hand, bifacial solar modules are mainly devised either for vertically integrated structures such as windows, sound barriers, and railings or for the collection of diffused light available from the earth, roof tops, clouds, and the atmosphere. Hence, the necessity of maximizing light harvesting from all directions and consequently the need for large light capturing surfaces renders bifacial PV relying on III–V materials financially impracticable. Multi-crystalline or mono-crystalline n-type Czochralski silicon solar cells, by contrast, can be turned into bifacial structures by implementing relatively minor modification (i.e., different BSF materials) to the current PV assembly lines. Finally, the increasing demand for energy production diversification has led to a growing interest in nanostructured materials and specific optoelectronic properties such as photocorrosion resistance and interfacial charge transfer between solid semiconductors and aqueous solutions. In fact, a promising method for the production of hydrogen to be used in fuel cells for automotive, consumer electronics and stationary power generation applications, which is water photoelectrolysis performed through semiconducting electrodes [33]. However, classical “solar cell-oriented” semiconductors (i.e., with band gaps between 1 and 2 eV) are prone to photodegradation when immersed in a water-based environment. Besides, they do not match the band edges position requirements to carry out both the oxygen and the hydrogen evolution reactions [11,12]. Nevertheless, the prospective of integrating solar energy collection and water electrolysis in a single electrode without resorting to separate power generators and electrolyzers remains attractive. Hence, plenty of compositional and structural alteration have been introduced in nanostructured wide-band gap transition metal oxides and calcogenides, stannates, vanadates, tungsten bronzes, titanates, etc. to accomplish water photoelectrochemical splitting. Such modifications span from the materials’ doping for engineering the energy bands position to the sensitization with organic dyes to achieve a band gap suitable for solar light absorption, from the morphology optimization for minimizing the charge transfer at the semiconductor/electrolyte to the deposition of multi-layered heterojunction structures to favor charge separation and transport [12,34]. In the following sections, we will describe the properties and fabrication techniques of single homo-junction (i.e., crystalline silicon) and heterojunctions (e.g., CIGS, CdTe, kesterite, HIT cells and a-Si/mc-Si, DSSCs, polymeric, and perovskite) solar cells.

2.5.3.1

Silicon-Based Photovoltaics

Silicon is one of the most captivating materials for energy production application thanks to its large abundance on the Earth’s crust and its high energy density (see Fig. 5).

124 Photovoltaic Materials

Selected energy densities

90 Aluminum

80 Silicon Anthracite

70

60

MJ/L

50 Magnesium

40

Iron

Polystyrene Lithium borohydride Polyethylene

Fat metabolism Diesel Gasoline Polyester Kerosene Butanol Sugar metabolism LPG butane Gasohol E85 LPG propane Ethanol Glucose Liquid natural gas Lithium Bituminous Hydrazine

Zinc

30

20

Sodium

Methanol Liquid ammonia

10

Liquid hydrogen

Natural gas (250 bar)

Hydrogen gas (700 bar)

Zinc-air battery

0 Lithium ion battery 0 20

Hydrogen gas

Natural gas

40

60

80 MJ/Kg

100

120

140

160

Fig. 5 Energy density of selected materials. Vertical scale is in megajoules per litre, and the horizontal scale is in megajoules per kilogram. Available from: https://commons.wikimedia.org/w/index.php? curid ¼5551431.

Photovoltaic Materials

125

Mono-crystalline and multi-crystalline silicon solar cells and modules (panels) have been dominating the PV market (see Fig. 2) since the early days of the semiconductor industry. The reasons for such a success are mainly related to the widespread use of silicon in the microelectronics industry rather than to its optoelectronic properties (see next section). Indeed, silicon is an indirect semiconductor and requires thicknesses of about 200–300 mm to achieve a useful light absorption coefficient. Films on the order of hundreds of micrometers cannot be fabricated by standard thin film deposition methods (chemical vapor deposition, physical vapor deposition, electroplating, etc.). Hence, silicon solar cells have been produced by cutting wafers of appropriate thickness from Czochralski or float-zone grown silicon ingots. Since the manufacture of mono-crystalline wafers is a fundamental step in the integrated circuits (ICs) fabrication, the PV industry has directly benefitted from purification, synthesis, and assembly techniques originally meant for the microelectronics sector. Nevertheless, in the last three decades crystalline Si PVs have been growing independently from microelectronics as new forms of silicon (e.g., solar grade Si, multi-crystalline wafer, Si ribbons, amorphous Si, and nanocrystalline Si), not suitable for the integrated circuits, have been developed. Besides, materials and methods that have been cornerstones of the silicon PVs for almost 50 years have been recently challenged by the increasing demand for more efficient, less expensive, and more durable solar modules. For instance, boron-doped p-type Czochralski wafers have been chosen since the 1970s as mono-crystalline solar cells’ base (i.e., substrate). The reason for such a choice are the simplicity of the doping process (easy control of the boron diffusion depth) by using diborane gas, and the nature of the charge transport to and from the p–n junction. In fact, solar cells are minority carriers (electrons in the p-type and holes in the n-type material, respectively) based devices. A p-type base/n-type emitter configuration ascribes the longest diffusion path toward the depletion region to the minority electrons into the p-type material. Consequently, the larger electrons diffusion coefficient, related to the lower effective mass of the electrons compared to the holes brings about a higher charge collection efficiency. However, boron interaction with the oxygen incorporated during the Czochralski crystal growth cause a “light-activated” detrimental effect known as LID (light induced degradation). This phenomenon consists of the increase of carriers’ recombination rate because of the trapping effect exerted by the B–O complexes formed by the solar radiation. Hence, prolonged light exposure of p-type solar cells brings about efficiency drops up to 2%. Therefore, the use of Czochralski p-type substrates hinders the development of longer lifespan solar modules. Another example of a well-established technology that is going through deep modifications concerns the assembly of solar cells into a module. Standard solar panels fabrication is based on the series connection of 60 or 72 cells (depending on the power output requirements). The encapsulation procedure is carried out by laminating the cells array between two ethylene-vinyl-acetate (EVA) or PVB (poly-vinyl-butyral). The module front and back sides are then protected with a solar grade layer (i.e., high optical transmittance and coated with a mesoporous silica antireflective coating layer) and with a polymer sheet (back-sheet), respectively. This configuration assures mechanical resistance and prevents moisture from penetrating the panel and degrade the solar cell grids and busbars. Traditional back-sheet polymers are three-layered structures comprised of PVF (polyvinylidene fluoride) and PET (Polyethylene terephthalate) “sub-sheet” that are co-laminated with the solar glass and the EVA. Being non-transparent, the backsheet composite impedes the light passage from the rear side of the module, precluding the possibility to exploit bifacial cells for the generation of power from reflected radiation. Thus, bifacial panels allow for higher power density and smaller areas compared to mono-facial modules. These characteristics are very desirable for “niche” applications where the albedo radiation can easily reach the active part of the solar module and where power output maximization without impacting on the lightness of the system is crucial. Such applications include, among others, PV generation modules’ installation on the wings of unmanned aerial vehicles. Therefore, a large part of academic and industrial research on bifacial silicon solar cell is focused on increasing the optical transparency of the solar grade glass while maintaining the same mechanical robustness (see Fig. 6 for the encapsulation structure of solar panels).

Glass Solar EVA Cell Solar EVA Back sheet

Fig. 6 Silicon solar panel encapsulation structure. EVA, ethylene-vinyl-acetate. Available from: https://www.indiamart.com/virattechnofab/viratsolar-for-b2b.html.

126

Photovoltaic Materials

From this brief introduction, it is clear how even a PV technology that has been proved is reliability several decades such as the crystalline silicon technology is still the object of an intense research aiming at maximizing the energy conversion efficiency while keeping the production costs at an affordable level. Therefore, the next sections will describe the recent developments in the silicon purification, cell design and cost reduction (i.e., silicon wafers’ thinning).

2.5.3.1.1

Solar grade silicon

The progress of PV materials and electronic devices during the last 70 years has been profoundly connected to the methods and technologies used to produce high-purity semiconductors. Over the years, silicon has risen to a prominent role among semiconductor materials because of the large availability of its raw feedstock (SiO2) and for the excellent electrical properties, and thermal stability of the silicon/native oxide interface which is the backbone of MOSFET (metal-oxide-semiconductor field effect transistor) technology [35]. Besides, silicon’s 1.1. eV band-gap guarantees low levels of thermal generation of electron–hole pairs at room temperature allowing for voltage (or current) controllable charge flows in IC’s devices and low recombination current (i.e., dark current) in solar cells [36]. Being silicon the most employed material for both PVs and microelectronics, the fabrication of mono- and polycrystalline solar cells shares a large part (raw materials purification, ingots and wafers preparation, doping, etc.) of the manufacturing process with the integrated circuits (IC) industry [37]. Nevertheless, there are two significant differences between the production of solar cells and other microelectronics devices. First, the purity requirements for the fabrication of bipolar transistors, MOSFETs, DRAMs (dynamic random access memories), CCDs (charge coupled devices), and plenty of other IC components are incommensurably more stringent than the ones needed for solar cells. For example, electronic grade silicon (EGS) possesses impurities concentration on the order of 1 ppb (5  1013 cm 3 9N purity), whereas the maximum impurity level acceptable in solar grade silicon (So-GS) is tens of ppms (4–6N). It should be noted that the purity scale refers to the number of digits (all nines) including non decimals (e.g., 3N corresponds to 99.9% purity, 4N indicates 99.99% purity, and so on). Second, while photolithographic patterning is extensively used in ICs fabrication, only special surface texturization features (i.e., inverted pyramids) obtained by isotropic chemical etching of multi-crystalline silicon require the use of photolithography in solar cell manufacturing [38]. Regardless of the specific application, metallurgical grade silicon MGS constitute the starting material for both EGS and So-GS. MGS is obtained from the carbothermal reduction of quartz (SiO2): SiO2 þ 2C-Si þ 2CO

ð1Þ

Carbon is provided as coke and the reaction is performed at high temperature (e.g., 18001C). MGS stemming from the carbothermal reaction contains up to thousands of metal and other impurities (i.e., Al, P, Fe, etc.) [39]. Although, MGS plays a crucial role in the improvements of steel in the ironwork industry, it does not match neither the electronic nor the PV industry’s purity requirements. Hence, further purification steps are carried out on the metallurgical grade silicon before it could be employed as materials for ICs and solar cells. The overall procedure to achieve high-purity (i.e., EGS) silicon has been developed at the beginning of the 1960s by Siemens corporation and was named after them [40]. Briefly, the MGS is reacted with anhydrous HCl to form trichlorosilane: SiMGS þ 3HCl-SiHCl3 þ H2

ð2Þ

The chlorides by-products stemming from the impurities (e.g., FeCl3, AlCl3, and BCl3) possess boiling points sensitively different from the one of SiHCl3 (321C). Thus, fractional distillation can be used to separate the trichlorosilane from the impurities. Then, SiHCl3 is converted back to silicon and deposited on multi-crystalline silicon rods by a hydrogen reduction reaction performed in a vacuum chamber for long time (e.g., 200–300 h), and at high temperature (E11001C). Such a silicon refining procedure allows for multi-crystalline silicon rods having an impurity concentration lower than 1 ppb (9–11 N). This purified silicon can be used for both microelectronics and PV applications making the trichlorosilane route the most dominant method nowadays for the production of solar grade silicon. However, the high cost of the Siemens process, problems related to the toxicity and corrosiveness of HCl, together with the higher impurity tolerance of PV devices compared to ICs, has led to the development of purification techniques “dedicated” to the So-GS. The methods for refining MGS to produce solar grade silicon can be classified into three categories [41]:

• • •

Metallurgical/chemical routes. Electrolytic route. Mixed methods.

The first class of techniques includes different methodologies for the reduction of various silicon precursors to elemental Si. For instance, the silane process relies on the synthesis of trichlorosilane by reacting MGS with SiCl4 and molecular hydrogen and on the successive disproportion of silicon tetrachloride to silane. Finally, the SiH4 is pyrolysed to solar grade silicon and H2 which is recycled [42]. Among the chemical/metallurgical methods, the fluoride process is particularly intriguing because allows to obtain solar grade silicon by low-cost by-products from the industrial production of phosphoric acids and fertilizers. Indeed, hexafluorosilicic acid (H2SiF6, a by-product of the phosphoric acid synthesis chain starting from apatite and fluoroapatite) acts as a precursor of SiF4 that constitute the primary silicon source for the fluoride purification route [43]. Reduction of silicon tetrafluoride by metallic sodium is performed to obtain a solid mixture of multi-crystalline silicon and NaF. Then, Si is separated from sodium fluoride by hydro-metallugical leaching. Although potentially cost-effective to produce solar grade silicon, the fluoride

Photovoltaic Materials

127

process requires complex apparatus design for the sodium mediated reduction step limiting its applicability to continuous industrial production [44]. A sub-class of the metallurgical/chemical methods is represented by the metallothermic and halidothermic purification techniques. Metallothermic procedures employ Zn, Al, alkali, or alkaline earth metals for the reduction of different silicon halides according to the reaction: SiX4 ðX ¼ F; Cl; Br Þ þ MðM ¼ Zn; Al; K; Mg; NaÞ-Si þ MXy

ð3Þ

where the physical state of the reactants and products (and consequently the apparatus for disposal and/or recycling of the byproducts) depends on the specific reaction used and on the vapor pressure of the precursors [45]. For instance, in the aluminothermic reaction liquid Al is supplied as reducing agent for gaseous SiCl4 giving as reaction products solid Si and AlCl3 which is removed as vapor and electrolyzed to regenerate the aluminum reductant [46]. On the other hand, zincothermic reductions in a fluidized bed reactor produce solid multi-crystalline silicon and ZnCl2 by reacting zinc vapors with gaseous silicon tetrachloride [47]. Despite different by-product and operating conditions, all the metallothermic procedures to refine metallurgical grade silicon need to be coupled to systems designed for recovering the metal-halide salts (e.g., distillation and rectification, leaching, etc.) and to electrolyzers employed for recycling the reducing metal and the halogen gas. Furthermore, most of the metal halide by-products (e.g., KF, NaF, ZnCl2, etc.) condensate on the silicon as solids bringing about contamination of the final multi-crystalline Si. Recently (2011), halidothermic reactions were proven to be a possible alternative to metallothermic reactions. Such a route is based on reacting aluminum sub-chlorides (AlCl2 and AlCl) in their gas form at high temperature (10001C) with gaseous SiCl4. Solid silicon is directly generated together with plenty of by-products (SiCl3, SiCl2, AlCl3) and un-reacted aluminum sub-chlorides [39]. Nonetheless, silicon’s contamination is largely reduced compared to the metallothermic reactions, all the by-products are easily removable gasses and no solid waste is created indeed. Refining of metallurgical grade silicon through electrochemical means has been attempted since the beginning of microelectronics and modern PVs (late 1950s). Two methods have been mainly investigated: three layers electrorefining [48] and direct reduction of SiO2. The former is predicated on the anodic solubilization of impure silicon alloys like in the analogous procedure patented by the Alcoa for Al purification, whereas the latter relies on the electrochemical reduction of high-purity silica performed in a molten salt electrolyte (e.g., CaF2), similarly to the Hall-Herault process for aluminum production starting from Al2O3, where synthetic cryolite is employed as both electrolyte and additive to lower the alumina melting point. A schematic of the three-layer silicon electrorefining is provided in Fig. 7. A liquid phase electrode composed of a MG silicon–copper (or other heavy noble metals) seats on the bottom of a vertically stacked electrolysis cell. The two other layers on top of it are, respectively, a mixture of molten CaF2 and BaF2 (melting point E14101C) serving as ion conductor, and a layer of ultrapure liquid silicon for the collection of material released from the working electrode [49]. In fact, when an electrical current is applied between the top and the bottom layer, the silicon–copper electrode is anodized, resulting in the dissolution into the molten salt electrolyte. Then, Si4 þ ions diffuse toward the less dense top layer where they are deposited as high-purity elemental silicon. Nevertheless, such a method is not effective in removing boron as a contaminant. On the other hand, the direct reduction of silica provides multi-crystalline materials with negligible amounts of boron. In fact, impurities are removed from SiO2 by melting it with glass forming oxides (e.g., P2O5, As2O3, GeO2, etc.). Then, glass fibers are formed by extrusion of the molten mixture. Fibers’ treatment with concentrated, hot HCl is carried out to ensure the removal of all non-silicious oxides. The high-purity SiO2 is then electrolyzed (i.e., reduced) in a rectifier containing molten calcium chloride as electrolyte [50]. Residual metal impurities are finally precipitated from the silicon by directional solidification (see next section). A large variety of metallurgical grade silicon purification methods meant for the achievement of low-cost solar grade silicon have been developed since the beginning of the 1980s. Most of them rely on two steps and are usually named “mixed methods” [51]. The first step is the separation of metallic impurities by directional solidification techniques. The second step involves a combination of different techniques such as liquid or gas extraction, hydrometallurgical refinement, and recrystallization to remove non-metal impurities (i.e., B, P, and C) or metals like Cu, and Ni. Regardless of the different techniques employed (e.g., zone float, Czochralski and Bridgeman-Stockbarger crystallization methods, solidification in a mold cavity, etc.), directional solidification exploits the higher solubility of impurities in molten silicon compared to solid silicon. Therefore, during solidification, impurities will distribute themselves between the solid and the liquid phase according to a material (and impurity)-specific

Si

− DC +

4e− Si

CaF2/BaF2

 = 2.57

Si4+  = 2.8 Si4+

Si/Cu

Si

4e−  = 3.5

Fig. 7 Schematic of the three layers silicon electrorefining process. r is the relative density of each layer. Adapted from Olsen E, Rolseth S. Three-layer electrorefining of silicon. Metall Mater Trans B 2010;41:295.

128

Photovoltaic Materials

coefficient of partition (or distribution coefficient) [35]: kd ¼

cs cL

ð4Þ

The terms on the right represent the concentration of solid (cS) and liquid (cL) phases. Heavy metals possess low partition coefficients at 14141C (Si melting point). Hence, it is possible to achieve silicon multi-crystals containing low amounts of metallic contaminants. Directional solidification has been proved to be particularly effective in separating Ti, Al, and Fe from silicon. Due to the formation of Fe–B light sensitive pairs in both mono- and multi-crystalline solar cells (see next section), iron removal is considered crucial to guarantee high electron’s recombination lifetime and consequently high energy conversion efficiency [52]. The partition coefficient plays a key role also in the hydrometallurgical refining of MGS. Although the purity reached through the hydrometallurgical method (98%–99.9%) is much lower than the one obtained by the trichlorosilane route, such a technique offers great advantages in terms of cost and equipment simplicity [53]. The process relies on the accumulation of metallic impurities on multi-crystalline silicon. In fact, because of their low kd in Si, heavy metals are preferentially distributed along the grain boundaries of polycrystalline silicon. Thus, their removal can be easily carried out by grinding the MGS to induce breakage along the grain boundaries and by leaching with combination of different acids (HCl, HF, H2SO4, and aqua regia) [54]. A main flaw of the hydrometallurgical refinement is its incapability at reducing the content of phosphorous and boron in the metallurgical grade silicon. The slag method, basically a liquid–liquid extraction technique, has been proposed to overcome such a limitation [55]. The slag employed for the extraction of P and B must satisfy the following requirements:

• • • •

phosphorous and boron partition coefficient in the slag need to be sensitively higher than the one in silicon. Silicon should be insoluble in the slag. The slag must not react with Si. Silicon and slag density should be significantly different.

The purification principle is based on the higher oxygen affinity of B and P compared to Si. Thus, the two contaminants can be easily removed if the liquid silicon is treated with molten slags comprised of CaO–SiO2, CaOSiO2–CaF2, CaO–SiO2–Al2O3, CaO–SiO2–Al2O3–MgO, CaO/MgO–SiO2 [56]. Another efficient technique for the elimination of boron and phosphorous from solar grade silicon is alloying Si with a low melting point metal such as Sn or Al to further decrease B and P partition coefficients. For instance, boron partition coefficient passes from 0.8 at the silicon melting point, to 0.038 when an Sn–Si eutectic (melting point E12001C) is used. Therefore, solidification of an Sn–Si or Al–Si alloy can result in the removal of 90% of the boron originally present in the So-GS. However, the use of alloys poses challenges concerning the removal of tin or aluminum which is usually performed by leaching in aqua regia. Finally, more “exotic” methods such as electron beam melting and gas blowing have been also used to obtain solar grade silicon. In the electron beam melting technique [57], a beam of high energy free electrons is focused on the MGS to evaporate impurities possessing higher vapor pressure than silicon. Although highly successful for the elimination of carbon, phosphorus, calcium, and aluminum, the electron beam melting is almost completely ineffective in removing boron because its vapor pressure is much lower than silicon [58]. Pyrometallurgical gas blowing refining uses active gases (oxygen, water vapor and their mixtures) diluted with an inert gas. When bubbling on the molten silicon surface, the oxidizing gases react with the impurities giving volatile oxides and sub-oxides and hydrate species (e.g., in the case of boron BO, B3H3O6, and BHO2). Although most of the current scientific research concerning the production of So-GS is devoted to the optimization of the chemical, electrochemical, and metallurgical techniques for the purification of silicon, at this stage none of the technologies described in this section seems to be really competitive with the Siemens methodology.

2.5.3.1.2

Czochralski, zone floating and multi-crystalline silicon processes

As previously mentioned, one of the most important parameters in determining the efficiency of crystalline silicon solar devices is the crystallographic state of the starting material. Indeed, the presence of grain boundaries in multi-crystalline silicon retards the charge transport and favors the electron–hole recombination through the defects. To obtain Si in a single-crystal state, high-purity silicon (either EGS or So-GS) is melted and then made to reform or solidify very slowly in contact with a single-crystal “seed.” The silicon adapts to the pattern of the single-crystal seed as it cools and gradually solidifies into a rod (also called “boule”). Several different processes can be used to grow a boule of single-crystal silicon. The most established and dependable processes are the Czochralski (Cz) method, the float-zone (FZ) technique, and the “ribbon-growth” technique [58,59]. A schematic of these three crystal growth methodologies is shown in Fig. 8. In the Czochralski process, crystal’s growth and p-type doping are performed simultaneously. Thus, chunks of solar grade silicon and heavily boron-doped silicon are melted together in a quartz-lined graphite crucible. Then, a seed crystal is dipped into the molten silicon and, while rotated, slowly pulled out of the liquid silicon. Rotation is instrumental in preventing temperature gradients that result in compositional inhomogeneities of the final crystal. Specific crystallographic orientations of the Si ingot can be achieved by using appropriate seeds. The Si o1004 surface is the one required in both ICs and solar cells fabrication. In fact, a o1004 crystallographic orientation ensures the formation of high quality (i.e., low defectivity) Si/SiO2 interfaces that prevents charge trapping in the MOSFET and enhance the emitter passivation (i.e., reduced recombination) in the PV devices. For this reason, o1104 oriented seeds are employed. Indeed, crystal’s slicing performed perpendicularly to the o1104 crystal growth

Photovoltaic Materials

Seed holder

Float-zone pulling

Ar

Ar

129

Feed rod holder

Seed Crystal neck

Feed rod

Shoulder (cone) Melt reservoir (SiO)

(SiO)

(Si)

Si sheet

Thermal shield

Melting

Heater

RF heating coil

Crucible susceptor

Molten zone Freezing interface

Crucible Ar+SiO+CO

Ar+SiO+CO

Vg Vs Simelt

Single crystal silicon

Silicon melt

Substrate

Crucible shaft Neck Seed

(A)

(B)

Seed holder

(C)

Fig. 8 The three standard methods for the growth of crystalline silicon wafers. (A) Czochralski process, (B) float-zone process, and (C) ribbongrowth technique. Available from: https://www.google.ca/search?q=Images þ Czochralski þ process þ Float-zone þ process þ Ribbongrowth þ technique&dcr=0&tbm=isch&tbo=u&source=univ&sa=X&ved=0ahUKEwiCjZ-zvYnXAhVl3IMKHV3TAIIQsAQIJQ&biw=1280&bih=588.

direction results into defect free, o1004 oriented wafers. As previously mentioned, one of the main drawbacks of the Czochralski technique is the inclusion of oxygen from the quartz crucible into the silicon ingot [60]. To limit, eliminate or mitigate LID effects stemming from the B–O pair three solutions have been proposed over the years [61]: 1. Magnetic Czochralski process. This technique is predicated on the paramagnetic properties of the oxygen molecule. Indeed, by applying a magnetic field during the crystal growth oxygen interstitial formation inside the silicon ingot and consequently the genesis of B–O pairs upon light exposure is avoided. In fact, a magnetic field applied either horizontally or vertically with respect to the axis of the quartz crucible prevents the bulk of the melted silicon from prolonged contact with the walls of the crucible. Therefore, less oxygen is incorporated into the final ingot. 2. Gallium doping. Ga and other trivalent elements (i.e., In) can be used instead of boron to provide silicon with p-type conductivity. Gallium doped Czochralski wafer based solar cells show no significant carrier lifetime degradation [62]. 3. Thermal recovery. Annealing of silicon solar cells (2001C for 30 min, [63]) has been proved adequate to recover carriers’ bulk recombination-lifetime. However, light induced degradation due to surface defects remains after the thermal regeneration process. Although successful in preventing LID, none of these remedies seem to be cost effective [63]. In the float-zone method, a high-purity multi-silicon rod is set atop a seed crystal and then lowered through an electromagnetic coil. The coil’s magnetic field induces an electric field in the rod, heating and melting the interface between the rod and the seed creating a necking process that leads to dislocation free crystals. Similarly to the CZ method, doping from gas sources (i.e., diborane or phosphine) can be performed concurrently to the single crystal growth. Solar cells manufactured from FZ single crystals are immune to LID because no consumable quartz crucibles are involved in the process. Furthermore, the high purity of FZ silicon reflects in longer electrons (and holes) recombination-lifetimes. Nonetheless, only 10% of mono-crystalline Si solar cells are manufactured by FZ and usually are dedicated only to applications where high energy conversion efficiency is required. The reason for such limited use of zone-float materials is mainly related to the excessive cost of the process. Indeed, the high purity of the multi-silicon rod, and the radiofrequency apparatus needed for the inductive heating are a premium compared to the CZ process. Both CZ and FZ methodologies produce cylindrical ingots. Thus, two problems arise during the wafer production: (1) excessive waste of material; (2) wafers possessing round corners. This second issue poses limitations for the packaging density of solar panels. On the other hand, wafers of multi-crystalline silicon can be obtained in a squared shape (see Fig. 9). In fact, multicrystalline ingots can be simply grown by solidifying the melted silicon into rectangular containers (molds [64]). Hence, the lower efficiency at cell level compared to mono-crystalline silicon is largely compensated by lower production cost and by the higher power density of multi-crystalline solar modules. Due to the high cost of wafer sawing for c-Si within the solar cell processing and module fabrication, research is being conducted on the reduction of these costs and on solutions that do not require wafer slicing. The ribbon technique [65–67] involves forming thin multi-crystalline sheets directly, thus avoiding the slicing step necessary to obtain wafers from bulk ingots.

130

Photovoltaic Materials

Mono-crystalline silicon solar cell

Poly-crystalline silicon solar cell

Fig. 9 Mono- and multi-crystalline silicon solar cells. From Saga T. NPG Asia Mater 2010;2(3):96–102. Available from: https://www.nature.com/ am/journal/v2/n3/full/am201082a.html.

Screen-printed Ag-paste ARC n+ emitter

Screen-printed Al-paste

Back surface filed

Fig. 10 Pictorial representation of a standard screen printed solar cell. Available from: https://www.google.ca/search?hl=en-CA&q=screen þ printed þ solar þ cells&tbm=isch&source=iu&pf=m&ictx=1&tbs=simg:CAESswIJGdeuuB38rxkapwILEKjU2AQaBAgDCAoMCxCwjKcIGmIKY AgDEiiyE5sUpxOeFModtxOfFMkdvQjIHZ098D-SN-4_17T_1wM6g-oz6nPuIzGjCOu7H87ybvLYGsnCtQEmtVqoqOnEIt41PW_1MrDgn8gbW8sHie2KS1EWtd4eOXGFkgBAwLEI6u_1ggaCgoICAESBK8tESAMCxCd7cEJGpIBChoKB2RpYWdyYW3apYj2AwsKCS9tLzAydjBtMgobCghwYXJhbGxlbNqli PYDCwoJL20vMDMwemZuChgKBmRlc2lnbtqliPYDCgoIL20vMDJjd20KGwoIZ3JhcGhpY3PapYj2AwsKCS9tLzAyMXNkZwogCg5ncmFwaGlj IGRlc2lnbtqliPYDCgoIL20vMDNjMzEM&fir=oXVCpw2e4DYPeM%253A%252CikfXNWelO5AARM%252C_&usg=__SU2soqR1LaK3Qh4F3wbXVTIoBo%3D&sa=X&ved=0ahUKEwjgi-S5wInXAhXI8CYKHa-XD0wQ9QEIOzAC#imgrc=oXVCpw2e4DYPeM.

One “ribbon-growth” technique – edge-defined film-fed growth – starts with two crystal seeds that grow and capture a sheet of material between them as they are pulled from a source of molten silicon. A frame entrains a thin sheet of material when drawn from a melt. This technique does not waste much material, but has not been fully commercialized yet.

2.5.3.2

Screen Printed Solar Cells

Screen printed solar cells have been the most widespread typology of silicon solar since the beginning of 1970s. Like the success of Si PV devices was boosted by the development of the ICs fabrication technology, the diffusion of screen printing as a technique to metallize solar cells’ front and rear surfaces is mainly due to the hybrid circuits and automotive industry. In fact, screen printers routinely used for depositing silver conductive traces for the low temperature co-fired ceramic (LTCC) electronic technology and car defrosters can be easily re-adapted to the creation of conductive grids for solar cells. A pictorial representation of a classical screen printed solar cell is reported in Fig. 10. The fabrication processes leading to a screen printed silicon solar cell are conceptually and experimentally simple. Briefly, lightly (i.e., boron concentration in the order of 1016 cm 3) p-type doped Si mono- or multi-crystalline wafer is chemically/ mechanically polished to remove damage from the ingot sawing with a slurry of fine and a concentrated alkaline NaOH solution. Subsequently, n-type diffusion doping from a POCl3 source is carried out in a furnace at high temperature (800–9001C). Such a step guarantees that only the outer layers of the wafer are n-doped. To avoid shunts in the final device, the junction is “isolated” by treating the side edges of the wafers (usually stacked together) with a plasma comprised of CF4 and O2. This procedure etches away the phosphorous containing layer electrically isolating the front of the cell from the rear side. A BSF is then produced by screen printing the back of the cell with an aluminum paste. Firing of the back contact (i.e., 7501C peak temperature) generates heavily p-type doped regions because of the Al diffusion through the silicon. The electric field created by such regions, that is the BSF,

Photovoltaic Materials

131

keeps the holes from recombining with the electrons at the back contact. Finally, a silver paste is screen printed and fired at about 8001C before being encapsulated into the solar panel structure. As previously noted, in order to improve the efficiencies of single junctions, single gap, solar cells one must deal with the SQ assumptions, by minimizing the losses. Hence, some modification have been performed on the basic solar cell preparation procedure, namely [68]:

• • •

Surface texturing and passivation. Antireflection coatings (ARC). Screen printing paste’s composition optimization.

Al/Ag strips for soldering contact tabs

A chemical etching (i.e., with concentrated NaOH solutions) process is applied to perform surface texturization on the n-type silicon layer in order to increase the effective light path [69]. Surface recombination, which affects the cell quantum efficiency at the shorter wavelengths, is usually a minor inconvenience for the actual (200–300 mm thick wafers) silicon solar cells technology [69]. Nevertheless, economic and technical requirements (that is, achieving carrier diffusion length comparable to the thickness of the device) has led to consistent wafers thinning (to 120–140 mm). Such a situation causes the emitter’s surface to be a relevant portion of the complete solar cell. Hence, surface recombination plays an important role in determining the energy conversion efficiency. Therefore, thermally grown SiOx, plasma enhanced chemical vapor deposition (PECVD) synthesized SiNx, and atomic layer deposited Al2O3 emitter’s passivation layers have been employed to reduce charge recombination by surface defects (i.e., silicon dangling bonds [70]). Amorphous SiNx is particularly interesting because acts also as ARC. A thin film (80–120 nm [71]) is then deposited onto the surface of n-type silicon layer to reduce the amount of photons reflected by the surface. The antireflective effect is based on refraction index matching and reflection extinction by interference, therefore the elemental composition of the PECVD film is crucial to maximize the amount of incoming light impinging on the active part of the solar cell. The composition of the amorphous ARC layer can be Si-rich (with a smaller refractive index) or N-rich (with a larger index), and the N/Si ratio is done by controlling the NH3/SiH4 flow ratio during ARC deposition. The last item is connected to the doping procedure of the cell’s top layer. Indeed, when the emitter is formed by phosphorous diffusion from POCl3, a uniformly doped n-type layer throughout the cell’s front side is created. However, high P concentrations underneath the screen printed silver fingers are required to keep the contact resistance low. Needless to say, the high phosphorus levels cause poor cell’s responsivity to the blue wavelengths (“dead layer” effect). This is, for example, one of the reasons why silicon solar cells are not well suited for operation under diffuse light [72]. Therefore, an extensive research work has been carried out in the last 20 years by silver screen printing pastes’ manufacturers to investigate the interaction between the metallization paste components (e.g., Ag flakes, organic vehicles and glass frit) and the silicon substrate in order to reduce the contact resistance and eliminate the dead layer problem [73]. For instance, the glass frit content (a mixture of oxides and carbonates such as PbO, Bi2O3, B2O3, Al2O3, ZnO, and BaCO3) of the metal loaded (70–85 wt% silver) screen printing paste has been tuned to 2–5 wt% to improve the formation of metal silicides that enhance the contact resistance while keeping the overall shunt resistance of the cell high [74]. The glass frit component of the printing paste is also needed to etch through the ARC. This strategy has allowed for lighter emitters’ dopants concentrations (the actual sheet resistance of a standard emitter is 80 O/sq and is expected to increase up to 120 O/sq by 2024) leading to an improvement of silicon screen printed solar cell’s spectral response in the blue wavelengths region [74]. In addition, even the metallization pastes for the cell’s rear contact have been optimized in terms of composition. Aluminum pastes are used to form contacts on the back of the cell and simultaneously for the creation of the BSF layer. The chemical formulation of Al pastes is similar to the Ag pastes used for the front grid. However, additional screen printing of aluminum/silver pastes are needed to preserve back contact solderability during the cells interconnection stage of the solar panel packaging. Thus, Al/Ag busbars are printed on top of the BSF (see Fig. 11).

Fingers

Aluminium full coverage for BSF

Busbars (A)

(B)

Fig. 11 Front (A) and rear (B) metallization pattern of a screen printed silicon solar cell. BSF, back surface field. Adapted from Shih YC, Shi FG. Silicon solar cell metallization pastes, materials for advanced packaging. Basel: Springer International Publishing; 2017.

132

Photovoltaic Materials

Silicon dioxide passivation layer and antireflection coatings

n-type diffusion n-type substrate with high minority carrier lifetime n-type diffusion

p-type diffusion

n-type diffusion

Negative contact Negative contact Positive contact All contacts are on the rear of the cell simplifying interconnection and preventing shading losses Fig. 12 Schematics of a rear contact cell. Availble from: http://pveducation.org/pvcdrom/manufacturing/rear-contact.

Although Si screen printed solar cells are still the most important type of PV device on the market, they suffer from numerous drawbacks hindering future prospective for the improvement of their $/Wpeak ratio. First of all, metallization silver pastes are the second most expensive method (after the solar glass) used in the manufacture of solar modules [75]. Besides, the Ag cost is increasing constantly [75]. Unfortunately, metallization solutions relying on electroless or electroplating of Ni/Cu layers or on the screen printing of copper pastes are not economically satisfactory. Indeed, additional costs stem from the photolithographic patterning for the plating method, and from the annealing in an inert atmosphere for both the copper-based screen printing pastes and electrodeposited nickel or copper films. Furthermore, the relatively low resolution (line width on the order of 120–150 mm) of the screen printed front grid creates shading effects lowering the light absorption and consequently the cell short circuit current. HIT cells, rear-contact cells, PERL and PERC solar cells, and bifacial solar cells represent four successful attempt on achieving a noticeable improvement in cell performance (with cost reduction) compared to conventional screen printed cells, and will be described in more detail in the following sections.

2.5.3.2.1

Rear contact cells

Fig. 12 shows the schematics of a rear contact cell. Rear contact solar cells eliminate shading losses altogether by putting both contacts on the rear of the cell. By using a thin solar cell made from high quality material, electron–hole pairs generated by light that is absorbed at the front surface can still be collected at the rear of the cell [76] where the effect of cell series resistance is greater. An additional benefit is that cells with both contacts on the rear are easier to interconnect and can be placed closer together in the module since there is no need for a space between the cells.

2.5.3.2.2

Passivated emitter, rear locally diffused cell

The PERL cell was first developed by Prof. Martin Green’s group at the University of New South Wales (UNSW) in the late 1980s and early 1990s. The main characteristics and advantages of PERL cells can be summarized as follows [77]. “The optical losses in the front contact area are reduced by implementing a textured inverted pyramid structure coated with an antireflector. The contact area at the front side is made as small as possible in order to enhance the total amount of light coupled into the solar cell by allowing collection of reflected light for the second time with less shadowing losses. The emitter is heavily doped underneath the contacts. In PERL, this is achieved by phosphorous-diffused regions. The rest of the emitter is moderately doped to preserve the blue response. A silicon oxide is passivated on the top of the emitter to suppress the surface recombination velocity. In the rear surface of the solar cell, point contacts are used in combination with the thermal oxide passivation layers to reduce the unwelcome surface recombination at the uncontacted area. Heavily doped boron acts as a local back surface to limit the recombination of the minority electrons at the metal back.” A schematic of a PERL cell is shown in Fig. 13. To date, the maximum efficiency for crystalline silicon-based solar cells is 25.6%, as shown in Table 1. This was achieved with a combination of HIT and rear junction schemes (see next sections) [15].

2.5.3.2.3

Bifacial solar cells

Bifacial solar cells are a relatively mature PV technology originally meant for overcoming one of the conventional silicon solar devices’ limitation [78]. Indeed, mono-facial modules are incapable of generating power by exploiting also the albedo radiation. Therefore, the bifacial approach can be considered one of most renowned strategies to improve the PV conversion efficiency without violating none of the S–Q theorem’s assumptions. Furthermore, bifacial solar panels drastically increase the power density compared to traditional modules, while containing the production costs [79]. Plenty of different bifacial configurations have been developed since the 1960s to collect and convert into electricity light reflected from the ground, from objects, and scattered by clouds and the atmosphere [80]. Notwithstanding such a wealth of designs, bifacial structures are still largely based on the layouts reported in Fig. 14. Regardless of the kind of base (i.e., substrate) used, the main difference compared to mono-facial solar cells is the presence of a screen printed metallization grid located on the rear side of the cell instead of the printed Al BSF layer. However,

Photovoltaic Materials

133

“Inverted” pyramids

Finger

p+ n+

n

Oxide p-silicon

p+

p+

p+

Oxide

Rear contact

Fig. 13 The passivated emitter with rear locally diffused (PERL) cell uses microelectronic techniques to produce cells with efficiencies approaching 25% under the standard AM1.5 spectrum. The passivated emitter refers to the high quality oxide at the front surface that significantly lowers the number of carriers recombining at the surface. The rear is locally diffused only at the metal contacts to minimize recombination at the rear while maintaining good electrical contact. Downloaded from: http://www.pveducation.org/pvcdrom/manufacturing/high-efficiency.

Front contacts ARC p+

n+ emitter

emitter

p-type substrate

n-type substrate n+ BSF

p+ BSF ARC

Back contacts

Fig. 14 Cross-section of basic bifacial solar cells employing standard (Cz) p-type (left) and n-type (right) substrates. BSF, back-surface field; ARC, antireflection coating. Adapted from Guerrero-Lemus R, Vega R, Taehyeon K, Kimm A, Shephard LE. Bifacial solar photovoltaics – a technology review. Renew Sustain Energy Rev 2016;60:1533.

the typology of BSF depends on the kind of silicon substrate (i.e., p þ for the n-type and n þ for the p-type). Moreover, beneficial thermal effects stem from replacing the opaque BSF layer with open metal contacts. In fact, the lack of aluminum back metallization decreases the cell’s temperature (i.e., less infrared absorption). Consequently, the device’s maximum power output and open-circuit voltage are enhanced. Nevertheless, such a beneficial effect is partly attenuated by the higher thermal insulation brought about by the glass needed to encapsulate the rear side of the solar panel (i.e., instead of the standard polymer based backsheet). Moreover, wafer bowing due to the marked difference between Al and Si thermal expansion coefficient is eliminated if aluminum free BSF layers are used [81]. Needless to say, additional costs arise from the fabrication of p–n junctions for bifacial cells. First, both sides of the device need be provided with ARC and surface texturing. Second, a higher thermal budget associated to multiple dopants diffusion processes is required to form the two junctions (see Fig. 14) of a bifacial solar cell. Such an issue has been addressed by the co-diffusion technique, which is a concurrent activation of the dopants used for the emitter and BSF layer formation, respectively [81]. However, traditional gaseous diborane sources are not suitable for co-diffusion and new boron precursors (e.g., BBr3) are being taken into account [82]. Furthermore, transparent phosphorus (for n-type substrates) or boron (for p-type substrates) BSF layers affect also the chemistry underlying the formulation of the Al/Ag screen printing pastes for the rear busbars deposition [83]. Therefore, metallization pastes suppliers have started to introduce formulation optimized for metal particles’ diffusion, thermal stress due the formation of silicides compounds, and contact formation on B and P BSF structures [84]. For instance, complete Al removal from the rear contact pastes has been used to improve the bifacial device’s open circuit voltage [85]. On the other hand, aluminum addition to conventional silver front metallization pastes was proven to be beneficial for reducing the contact resistance in the boron-doped emitters used for bifacial solar cells relying on n-type substrates [85].

134

Photovoltaic Materials

HIT (Heterojunction with intrinsic thin layer) solar cell is composed of thin single crystalline Si wafer sandwiched by ultra-thin a-Si layers n-type Electrode

Anti-reflection layer Top electrode

p

Metal electrode Crystalline Si (p-type)

Conventional c-Si solar cell

p-type amorphous Si Intrinsic amorphous Si

n Bottom electrode

Crystalline Si (n-type) Intrinsic amorphous Si n-type amorphous Si HIT solar cell

Fig. 15 Conventional c-Si solar cell structure and heterojunction intrinsic thin film junction. Available from: https://images.search.yahoo.com/yhs/ search?p=hit þ cells&fr=yhs-mozilla-001&hspart=mozilla&hsimp=yhs-001&imgurl=http%3A%2F%2Fwww.solarchoice.net.au%2Fblog%2Fwpcontent%2Fuploads%2FHIT-cell-technology-amorphous-and-crystalline-silicon-combined1.jpg#id=2&iurl=http%3A%2F%2Fmedia-cache-ec0. pinimg.com%2F736x%2F4e%2F06%2F3d%2F4e063d0b140896642b9617a6c9f26ce2.jpg&action=click.

2.5.3.2.4

Heterojunction with intrinsic thin layer cell

The HIT device is not a proper single-junction cell, since it combines two materials, crystalline silicon and amorphous silicon (see Section 2.5.3.2.5 in this chapter). It is indeed a heterojunction formed by the two different silicon structures. Fig. 15 compares a standard c-Si cell with a HIT cell. The p-doped amorphous layer and the crystalline silicon wafer (slightly n-type) form the p–n junction and BSF. A thin intrinsic layer is introduced and a transparent conducting oxide is added to allow lateral carrier transport to metal contacts. The HIT cell shows superior conversion efficiency performance compared with traditional crystalline silicon cells when operating at high temperatures. A major advantage in the manufacturing process of HIT cells is that the thermal budget is considerably reduced. Temperatures of up to 10001C are needed for dopant diffusion in order to form the p–n junction. The HIT cell doped layers are normally formed by PECVD processes below 3501C. The symmetrical structure of the HIT cell also reduces mechanical and thermal stresses. A key factor in enhancing HIT cell performance is the inclusion of the intrinsic amorphous silicon layer. The intrinsic film provides very good passivation of the silicon surface by reducing interface defect density.

2.5.3.2.5

Hydrogenated amorphous silicon (a-Si:H) and microcrystalline silicon (lc-Si:H)

Hydrogenated amorphous silicon (a-Si:H) and microcrystalline silicon (mc-Si:H) have different structures but can be both included in the “thin film silicon” category. Amorphous silicon (a-Si), and in particular its hydrogenated form (a-Si:H), represents a perfect example of what is referred to as second generation PV. Amorphous solids, like common glass, are materials whose atoms are not arranged in any particular order. They do not form crystalline structures at all, and they contain large numbers of structural and bonding defects. But they have some economic advantages over other materials that make them appealing for use in solar electric, or PV, systems. Amorphous silicon is common in solar-powered consumer devices that have low power requirements, such as wristwatches and calculators. Amorphous silicon absorbs solar radiation 40 times more efficiently than does single-crystal silicon, so a film only about 1 mmthick can absorb 90% of the usable light energy shining on it, with obvious cost-saving implications. Other economic advantages are that it can be produced at lower temperatures and can be deposited on low-cost substrates such as plastic, glass, and metal. Furthermore, a-Si:H has a large optical absorption coefficient; the energy gap can be modulated to allow for near optimum conversion efficiency for sunlight and it can be alloyed with other elements (carbon, germanium) to create multi-junction structures with increased energy conversion efficiency for sunlight; furthermore, the basic material, silicon, is plentiful and can be deposited on a variety of substrates (low temperature, large area, flexible). An important distinction between amorphous and crystalline silicon is in the nature of the energy gap. Crystalline silicon exhibits an indirect band gap of about 1.1 eV, while a-Si:H has a direct band gap in the range 1.5–2.0 eV, depending on growth conditions and hydrogen content [86,87]. However, a-Si:H also has disadvantages due to its structure; in particular, exposure to sunlight leads to an increase in the density of states in the energy gap. This increase in the density of states is linked to the formation of silicon dangling bonds. Annealing at elevated temperatures reduces the density of states back to the original value. This is known as the Staebler-Wronski (SW) effect

Photovoltaic Materials

Cell cross sections

135

Spectral cell sensitivity

a-Si

a-Si

μc-Si

Quantum efficiency (a.u.)

0.9 0.8 0.7 0.6 0.5 0.4 0.3

μc-Si bottom cell

0.2 0.1 0 300

Amorph Single junction

a-Si top cell

Micromorph® Tandem junction 400

500

Micromorph® Tandem junction

600 700 800 900 1000 1100 Wavelength (nm) Near IR Amorph Micromorph tandem

Fig. 16 A-Si:H single junction structure vs. a-Si:H/mc-Si tandem structure (left). The increased absorption range due to the combination is shown in the right. Available from: http://thinfilmsiliconpv.ch/drupal/sites/default/files/fig1.jpg

[88] (also, see Gaspari [89] for a more complete description of a-Si:H and the SW effect). The best efficiencies for a-Si:H solar cell are of the order of 10% [90]. High efficiency microcrystalline (mc)-Si solar cells have shown values in the range 10.5%–10.7% [91,92] although recent data show an improved efficiency of 11.8% (see Table 1). It should be noted that the thickness of mc-Si is kept relatively small (2–5 mm) to allow for the same advantages of a-Si:H (flexible substrates, low temperature, etc.). However, since the micro-crystalline structure does not exhibit an absorption coefficient comparable to a-Si:H, there will be a large portion of incident light which is not used for carrier production. a-Si:H has been combined with mc-Si to form tandem or multi-junction cells, thus going beyond the S–Q assumption of a single p–n junction (see, for instance, [93]). One of the advantages of this cell structures is that both junction can be grown using the same PECVD technique. Fig. 16 shows the structure of a a-Si/mc-Si tandem cell and the relative overlapping absorption spectra.

2.5.3.3

Other “Conventional” Photovoltaic Materials

The following section provides a brief description of non-silicon-based materials and solar cells. Thin film devices whose PV action is predicated on mechanism other than charge separation at a p–n junction interface (and minority carriers’ diffusion) will be treated in the “Organic Photovoltaic” section.

2.5.3.3.1

Gallium arsenide

Gallium arsenide (GaAs) is a compound semiconductor: a mixture of two elements, gallium and arsenic. As indicated in Table 1, thin film GaAs cells have achieved 28.8% efficiencies. These efficiencies are very high although there is still room for improvement, according to the S–Q limit. Although the efficiencies of these cells are the best to date for a single-junction structure, the materials employed have drawbacks: Gallium is a by-product of the smelting of other metals, notably aluminum and zinc, and it is rarer than gold. Arsenic is not rare, but it is poisonous. Gallium arsenide has been developed for use in solar cells at about the same time that it’s been developed for light-emitting diodes (LEDs), lasers, and other electronic devices that use light. A standard Ga-As solar cell structure is shown in Fig. 17. It should be noted that thin film GaAs has produced higher efficiencies than its multi-crystalline structure counterpart. This is a big advantage as the multi-crystalline form requires complex and expensive growth methods [94]. Chemical vapor deposition (CVD) is instead used to induce epitaxial growth of the GaAs thin film [95,96].

2.5.3.3.2

Indium phosphide (InP) and gallium indium phosphide (GaInP)

Indium phosphide (InP) is a binary semiconductor composed of indium and phosphorus and it is similar to GaAs (face-centered cubic “zincblende” crystal structure). The energy gap is EG ¼ 1.35 eV and has efficiencies lower than those of GaAs solar cells as shown in Table 1 (22.1%). Addition of gallium increases the band-gap to 1.81 eV, thus increasing the theoretical S–Q limit. However, to date, record efficiencies have not yet surpassed the InP values [97]. Another variation is obtained by adding aluminum to obtain aluminum gallium indium phosphide (AlGaInP), a semiconductor material used mainly in novel multi-junction PVs and optoelectronic devices, thanks to its direct bandgap in the range 1.8–2.0 eV.

136

Photovoltaic Materials

Front contact (Ti/Pt/Au)

500 nm

p-GaAs

300 nm

Window

p-ln0.5Ga0.5P

200 nm

Emitter

p-GaAs

500 nm

Base

n-GaAs

3500 nm

BSF

n-ln0.5Ga0.5P

Buffer

n-GaAs

200 nm

Substrate

n-GaAs

350 μm

Back contact (AuGe/Ni/Au)

500 nm

Ohmic

50 nm

Fig. 17 Schematic structure of a GaAs single-junction solar cells. Available from: https://www.hindawi.com/journals/ijp/2013/539765.fig.001.jpg

Sun light ZnO

Front contact

CdS

p−n junction

Cu(In,Ga)Se2

Electron hole External load Back contact

Mo

Glass 1 μm Fig. 18 Scanning electron micrograph of a Cu(In,Ga)Se2 solar cell (cross-section) and its mode of operation. Available from: https://www.azom. com/article.aspx?ArticleID¼6466.

2.5.3.3.3

Copper–indium–(gallium)–selenide (CIS and CIGS)

Copper indium diselenide (CuInSe2 or “CIS”) has an extremely high absorptivity, which means that 99% of the light shining on CIS will be absorbed in the first micrometer of the material. Cells made from CIS are usually heterojunction structures – in which the junction is formed between semiconductors having different band gaps. The most common material for the top or window layer in CIS devices is cadmium sulfide (CdS), although zinc is sometimes added to improve transparency. Adding small amounts of gallium to the lower absorbing CIS layer boosts its bandgap from its normal 1.0 electron-volts (eV), which improves the voltage and therefore the efficiency of the device. This particular variation is commonly called a copper indium gallium diselenide or “CIGS” PV cell. CIGS is a I-III-IV2 alloy. Its official chemical formula is CuInxGa(1 x)Se2, where x can take any value between 0 and 1. This allows for different amounts of gallium, changing the chemical and physical properties of the material. The best properties are achieved for an x value of around 0.7 [98]. The exact crystal structure is complex and changes with the ratio of the used materials. It is classified as a chalcopyrite structure [98]. However, it seems to be the case that intrinsic conduction is to large amounts due to built-in defects in the crystal [99]. Although CIGS can be grown in a mono-crystalline fashion, most applications use polycrystalline CIGS. Astonishingly the grain boundaries do not lower the suitability for PV applications, instead poly-crystals show favorable attributes. The reason for this phenomenon is not completely understood [98]. Although it is possible to produce both p- and n-type CIGS, actual CIGS cells feature only n-type crystals as absorption material. The p–n junction is created using CdS, forming a heterojunction. A schematic view of a cell can be found in Fig. 18. The lowest layer is the substrate, usually glass, in many modern applications also flexible films, such as metal foil or polymers [100]. These flexible cells offer a whole new field of applications. The back contact usually consists of a molybdenium layer applied

Photovoltaic Materials

137

to the substrate. The actual CIGS layer is the thickest part of the whole cell, about 2–3 mm thick [99]. The next layer is a very thin, highly p-doped CdS layer. It is thin enough to let photons pass through it and to allow tunneling of charge carriers on their way to the front contact. Cadmium is a toxic substance, therefore possible replacements such as zinc sulfide are being considered. The best efficiency to date for CIGS solar cells is 21.7% [101].

2.5.3.3.4

Copper zinc tin sulphide

As the formula shows, copper zinc tin sulphide (CZTS) is a quaternary compound similar to CIGS cells, but with Indium and gallium replaced by zinc (Zn) and tin (Sn), respectively, thus addressing material availability issues. The CZTS film exhibit a bandgap of approximately 1.5 eV and a large absorption coefficient of 10 4 cm 1 [102], although by varying composition it is possible to tune the band-gap between 1.0 and 1.6 eV. Todoriv, Reuter and Mitzi [103] established a record efficiency for CZTS films of 12.6%, by employing a Cu-poor, Zn-rich stoichiometry with the band gap controlled by the S/Se ratio. Schematics of a CZTS cell are similar to those of a CIGS cell (Fig. 18) with the appropriate changes as described above.

2.5.3.3.5

Cadmium telluride (CdTe)

Up to now CdTe PVs remain the only kind of thin film-based solar technology capable of competing with crystalline silicon solar cells in terms of Wp/$ ratio. The reason of such a “success” of CdTe cells and modules lay in its optoelectronic properties and in the low-cost process employed for the manufacture of CdTe/CdS heterojunctions. With a nearly ideal bandgap of 1.44 eV, CdTe also has a very high absorptivity, requiring about a 2 mm thickness to absorb 99% of the solar spectrum above 1.45 eV [104]. Fig. 19 shows the schematic cross-section of the two most used configurations to fabricate a n-type CdS/ p-type CdTe thin film solar cell: superstrate and substrate. The superstrate configuration allows the light to enter the cell from the substrate side. Therefore transparent conductive materials need to be deposited before the CdS/CdTe heterojunction formation. Hence, transparent conductive oxides (TCOs) (i.e., tin oxide doped with fluorine (FTO), indium oxide doped with tin, zinc oxide doped with aluminum AZO, and cadmium stannate Cd2SnO4 are routinely used as window layers. Like CIS, films of CdTe can be manufactured using low-cost deposition techniques such as electrodeposition, chemical bath deposition, screen printing, close space sublimation and spray deposition [105]. However, junction “activation” with chlorine sources like CdCl2 is necessary during the CdTe/CdS structure annealing (usually performed at 400–5001C) [106]. Efficiencies of 21% (see Table 1) have been achieved. On the other hand, among the drawbacks, tellurium is scarce and massive panel production might be a challenge. Furthermore, due to its high toxicity, handling of cadmium can be extremely dangerous. Indeed, cadmium’s use in manufacturing has already been banned by the European Union. CdTe solar cells violate this ban; however, they have not been labeled as a restricted product, and could be granted exemption. Stricter laws on the use, transport, and recycling of Cd could become an issue. The following sections will describe how third generation PV tries to employ both the nature of the material and alternative solar collection and carrier production schemes to produce inexpensive solar cells. Although some of the following solution also belong to a more general “thin-film” category, the materials and combination thereof described below can be grouped in the category of organic materials.

2.5.3.4

Organic Photovoltaics

Organic PV is a general definition that sometimes includes also DSSCs and perovskites solar cells because of the HTM, methylammonium salts, or light absorbing dyes used for their fabrication. Nevertheless, organic solar cells are primarily intended for low-cost devices on flexible substrates and that limits the discussion to polymer solar cells. DSSCs and perovskite cells will be discussed in the following sections. illumination Metal

TCO

Buffer

n-CdS

+

−

−

+ p-CdTe

p-CdTe

n-CdS

Buffer

TCO Glass or polyimide

Metal Glass or metal

illumination Fig. 19 Pictorial representation of superstrate (left) and substrate (right) configurations for CdTe solar cells. TCO, transparent conductive oxide. Adapted from Buecheler S, Kranz SL, Perrenoud J, Tiwari AN. CdTe solar cells. In: Solar energy. New York: Springer

138

Photovoltaic Materials

The field of organic PVs has been under research for about 50 years, but it has received particular attention due to the realization of the need for inexpensive renewable energy [107]. Compared to inorganic solar cells, organic solar cells, if developed into a common technology, would be beneficial in that they would reduce energy production costs by employing earth-rich materials into its operation [108]. One major drawback of organic solar cells is that they have large Coulombic exciton binding energies, making electron–hole pair production more difficult [108]. Deposition of organics by techniques such as screen-printing, doctor blade, inkjet printing, spin coating, dip coating, spray deposition lends itself to incorporation in high-throughput low-cost roll-to-roll coating systems. These are low temperature deposition techniques that allow the organics to be deposited on plastic substrates so that flexible devices can easily be made. The first solar cells generally consisted of a single layer of polymer semiconductor material sandwiched between two electrodes of different work functions [108]. These devices produced very low photocurrent efficiency measures. In 1986, C. W. Tang (as cited in Clarke & Durrant [109]) remedied this problem using a CuPc/perylene derivative bilayer device. Since CuPc and perylene had two different ionization energies, there existed an energy offset at their common boundary which assisted in the dissociation of exciton electron–hole pairs. There are two major classes of organic semiconductors that organic PV is comprised of [107]. They include molecular semiconductors (MSCs) and p-conjugated polymers. The first component of a bulk heterojunction (BHJ) organic solar cell is a transparent electrode, usually made of glass coated by a conducting oxide [108]. The next layer is a thin interfacial layer normally made of an organic or metal oxide material, followed by a thick absorber layer consisting of a network of donor and acceptor atoms. This is followed by a second thin interfacial layer, (i.e., lithium fluoride [108]), followed by a thick electrode layer composed of a metal with a low work function, such as aluminum. In other words, this contact has low electron affinity, making it easier to remove an electron from its valence band. The smaller the work function, the smaller the difference in the vacuum level and Fermi energy levels, and the easier it is to remove a valence electron from the atom. A schematic view of an organic solar cell is shown in Fig. 20.

2.5.3.4.1

Electron–hole exciton pairs in organic photovoltaics

Typically, the electron–hole binding energy in a solar cell is higher than that in a traditional silicon semiconductor [22,108]. It is typically of the order of 0.3–0.5 eV. Therefore, the absorption of light by the solar cell leads to the generation of bound electron–hole pairs known as exciton pairs. The electrons and holes in these exciton pairs are not free carriers, and therefore do not contribute to the current.

2.5.3.4.2

Recombination mechanisms in organic photovoltaics

There are two major types of recombination in organic solar cells [22]. They include geminate pair recombination and bimolecular recombination. Recall that the active layer is composed of an interwoven network of donor and acceptor atoms. A gemination pair is typically formed at the boundary between the donor and acceptor materials, and remains bound together due to Coulombic attraction. In order for the electrons and holes in these gemination pairs to contribute to the current, they must overcome this Coulombic attraction that binds them at the donor-acceptor interface. However, even if the geminate pair undergoes dissociation, the charge carriers are not completely free. There remains the possibility that the free carriers will recombine with other electrons and vacant holes in the active layer of the semiconductor before reaching the electrodes.

2.5.3.4.3

Inorganic semiconductors versus molecular semiconductors

According to Hains et al. [107], one of the parameters affecting the photo-generation of free carriers is the dielectric constant. This parameter is highly dependent upon the type of material being used as a semiconductor. The higher the dielectric constant, the easier it is to produce photocurrent. In particular, silicon, which is classified as an inorganic material, has a dielectric constant of about 11.9, while MSCs have dielectric constants typically ranging from 3.5 to 5.5 [107]. The valence electrons are more tightly bound to atoms with lower dielectric constants. As a result, the low dielectric constant in an MSC is a photo-generation obstacle. This property also holds true for p-conjugated polymers [109].

Interfacial layer

Transparent electrode Absorber layer

Interfacial layer

+

−

Metal electrode Fig. 20 Schematics of a bulk heterojunction (BHJ) organic photovoltaic (PV). The circle on the right shows an electron–hole exciton pair. Adapted from Servaites JD, Ratner MA, Marks TJ. Organic solar cells: a new look at traditional models. Energy Environ Sci 2011; 4:4410.

Photovoltaic Materials 2.5.3.4.4

139

Charge transfer and separation states in organic photovoltaic

In organic PV, charge transfer can occur at the interface of donor and acceptor materials [110]. Each of these materials holds two special molecular orbitals known as the highest occupied molecular orbital (HOMO) and the lowest unoccupied molecular orbital (LUMO). For simplicity, we will focus on the electron transfer from the donor LUMO to the acceptor LUMO, since the donor material is the primary light absorber in most organic solar cells [108]. In order for the electron transfer to occur, the energy level of the LUMO of the donor atom must be higher than that of the LUMO in the acceptor atom [110]. The difference between these LUMO levels is known as the energy offset, denoted by DELUMO. This parameter is important in facilitating the transfer of an electron from the higher LUMO in the donor atom to the lower LUMO in the acceptor atom. It has also been shown to be important in assisting in geminate pair dissociation and hence the photo-generation of free charge carriers [108]. If we let the parameter DEEx represent the Coulombic binding energy of an exciton pair, it is necessary that DELUMO4DEEx in order for electron–hole dissociation to occur. Fig. 21(A) is a simplified energy band diagram showing the stages of electron transfer. After the electron crosses the interface, it will either obtain a transfer state or undergo geminate recombination if it reaches the lowest level of the LUMO orbital. If it remains in a transfer state, it will eventually achieve a charge separation state [109]. Best organic cells efficiencies are of the order of 11% (see Table 1).

2.5.3.5

Dye-Sensitized Solar Cells

Power conversion efficiency

DSSCs are unusual PV devices because they exploit a metallo-organic molecule for light absorption and carriers’ generation (i.e., a bipyridine Ru dye), a nanostructured wide bang gap semiconductor for the electron transport, and a semiconductor–electrolyte interface for charge separation [111] (see Fig. 22). Although the concept of using an illuminated semiconductor–electrolyte junction to generate a photopotential can be backtracked to the 19th century, stable and efficient photoelectrochemical cells (PECs) have not been developed until materials technology has permitted to synthesize nanostructured wide band gap semiconductors at the beginning of the 1990s. In fact, traditional “low” band gap semiconductors (n-type an p-type Si, n-type GaAs, p-type Inp, n-type CdS, etc.) are subject to Charge transfer states S1

Energy

ΔELUMO Bandgap (Eg)

Charge separated states

Geminate pair recombination S0

(A)

15%

10%

Ideal (complete dissociation) Enhanced PCE

5% Current record efficiencies 0% 0.0

Acceptor

Donor

0.2

1.4

(B)

0.6 0.8 ΔELUMO (eV)

1.0

1.2

Fig. 21 (A) Energy band diagram for donor–acceptor interface in organic solar cell. (B) Power conversion efficiency (PCE) curve of best fit for recorded lowest unoccupied molecular orbital (LUMO) offsets (lower curve), predicted improved PCE curve (middle curve), Ideal PCE limit (upper curve). Adapted from Servaites JD, Ratner MA, Marks TJ. Organic solar cells: a new look at traditional models. Energy Environ Sci 2011; 4:4410.

−

−

−

−

− − −

− −

−

− −

3I−

3I−

I3−

I3−

Light −

TCO glass

TiO2

Dye

Electrolyte

−

Pt

TCO glass

Fig. 22 Structure and principle of operation of a dye-sensitized solar cell (DSSC). TCO, transparent conductive oxide. Available from: https:// www.gamry.com/application-notes/physechem/dssc-dye-sensitized-solar-cells/.

140

Photovoltaic Materials

photocorrosion when in contact with an electrolyte. On the other hand, using wide band gap semiconductor oxides (ZnO, TiO2, Nb2O5, WO3) can result in a “vicious circle”: the large band gap produces a sufficient band bending at the semiconductor–electrolyte interface preventing charge accumulation and then photocorrosion, but it incapacitates carriers photogeneration by the semiconductor. Therefore, an oxide needs to be “sensitized” by light absorbing materials like organic chromophores or quantum dots. The sensitizer absorbs photons in the visible range, promotes an electron to its LUMO and injects the e (if a favorable bands’ alignment exists) into the oxide conduction band (CB). However, the first sensitized electrochemical devices (fabricated in the 1970s) used mono- or polycrystalline materials limiting the amount of sensitizer adsorbable on the photoelectrode surface. Furthermore, the original dyes used to sensitize wide band gap semiconductors (i.e., Rodamine B, fluorescine or chlorophyll) covered only a small portion of the solar spectrum. Thus, two main breakthroughs in the field of sensitized PEC have been:

• •

The synthesis of photoactive metal-polypyridine (based on 2,20 -bipyridine, 2,20 ;60 200 -terpyridine and 1,10-phenanthroline) capable of intense (high molar extinction coefficient, ϵE10000 M 1 cm 1) metal-to-ligand charge transfer (MLCT) optical transitions. High specific surface area oxides capable of linking dyes through carboxylic or phosphonic bonds.

These advances have brought to the fabrication of the first (7% efficient) DSSC by Graetzel and O’Regan in 1991 [112]. The use of ruthenium based bis-bipyridine complexes and nanostructured TiO2 has increased the quantum efficiency to values comparable to conventional solar cells (the IPCE for the most efficient DSSCs is around 80%–85%). Although DSSCs are less efficient than polycrystalline and mono-crystalline silicon solar cells (14.7% vs. E18% of poly- and E25% of mono-crystalline Si), their manufacturing cost is immensely lower. Indeed, DSSCs do not require any high vacuum or materials purification steps (like zone refining for example). Furthermore, material properties tunability (TiO2 morphology, dye light absorption properties, etc.) makes DSSCs more versatile than Si devices (where most of the optoelectronic properties are controlled only by doping). Moreover, DSSCs are capable of operating under low-irradiance and diffuse light exposure (like cloudy days conditions). Energy conversion efficiency of silicon solar cells is strongly reduced in those conditions. Finally, DSSCs’ transparency makes them ideal for building integrated applications [113]. The DSSC generates current from the absorption of a photon by a dye molecule which injects an electron into the conduction band of the semiconducting TiO2 (or other large band-gap materials). The dye is then regenerated by an electrolyte containing an iodide/triiodide (I /I3 ) redox couple which was reduced at the counter electrode. The competition between the transport of the electrons through the TiO2 network and the recombination of electrons with the electrolyte at the photoanode/electrolyte interface, defines the electron collection efficiency of the device and its limitations. The driving force of the whole process is the favorable position of the energy states involved in the charge generation, injection and transport, as shown in Fig. 23 [115]. In fact, no electron injection can be carried out from the dye to the semiconductor unless the dye LUMO is positioned at more negative potentials (higher energy level) than the TiO2 CB. Similarly, no dye regeneration can occur unless the redox shuttle

Conducting glass

Pt counter electrode (cathode) HOMO e−

−0.5

C.B.

Injection

S∗

EF

Maximum voltage

0

Reduced Oxidized Mediator

0.5 e−

(3I− 1.0

e−

S0/S+

V vs. SHE e−

TiO2 (anode)

Dye

I3+)

LUMO e−

Load Fig. 23 Energy diagram and charge transfer processes in a dye-sensitized solar cell (DSSC). The figure reports only the forward processes. Recombination mechanisms and kinetics are not shown. The potential scale is referred to the standard hydrogen electrode (SHE). Adapted from Hardin BE, Snaith HJ, McGehee MD. The renaissance of dye-sensitized solar cells. Nat Photon 2012;6. Available from: www.nature.com/ naturephotonics.

Photovoltaic Materials

141

potential (RE) is situated at more negative potentials than the dye’s HOMO. This simple explanation helps to understand why high efficiency cells can be obtained with the following combination: anatase, Ru-based dyes (with proper ligands) and I3 /I redox shuttle (with proper electrolyte additives). Indeed, anatase is the only TiO2 polymorph with a CB edge position suitable for electron injection from Ru dyes. The disordered network of TiO2 nanoparticles contains numerous grain boundaries which ultimately result in the slow transport of electrons and increased recombination. The DSSC’s performance significantly depends not only on minimizing recombination; but also promoting the forward travel of the electrons through the anode. Developments on the DSSC photoanode include different fabrication methods of the TiO2 layer such as sol–gel, hydrothermal/solvothermal, spray pyrolysis, and atomic layer deposition. Other approaches to improve electron transport have included the incorporation of nanostructured semiconducting/conducting materials such as graphene sheets, metal oxides (e.g., ZnO and SnO2), and carbon nanotubes which provide a direct electron pathway through the anode. Solid-state DSCs (ss-DSCs) use solid hole-conductors instead of a liquid electrolyte, and have shown high potential [114]. The hole conductor is typically made from either wide band-gap small molecules (such as spiro-OMeTAD) or semiconducting polymers such as PEDOT or P3HT). These materials would eliminate the need of a liquid electrolyte and the issue of photocorrosion. A thorough review of DSSCs can be found in Hardin et al. [114]. According to the authors, “Increasing the module efficiencies of DSCs to more than 14% would relax the ultralow-cost constraints, thus providing substantial incentive to create laboratory-scale devices with efficiencies greater than 15%. The relatively slow increase in record values for DSCs over the past 10 years has left the impression of a performance ceiling, which is partially justified given that conventional iodide- and ruthenium-based DSCs have a realistic maximum possible efficiency of little more than 13%”. According to Table 1, best efficiency for a DSSC cell has been reported to be 11.9%, although Hardin et al. [55] report a 12.3% efficiency for new cobalt-based redox couples which are making it possible to obtain higher open-circuit voltages.

2.5.3.6

Perovskite Solar Cell

Perovskite solar cells (PSCs) represent the natural evolution of DSSCs [116], whereas organic solar cells share some characteristics with DSSCs because the charge is generated by organic molecules or polymers and because of the high interpenetration level of the hole and electron conducting media. PSCs are PV devices based on organo-metal halides of general formula CH3MNH3X3 (X¼ Br, I, Cl; M¼ Pb, Sn, Ge) possessing the crystal structure of the perovskite materials (ABX3 where X is either a halogen or oxygen, A the larger cation occupying the cubo-octahedral sites and B the smaller cation occupying the octahedral sites, see Fig. 24). Hybrid organic–inorganic methylammonium lead iodide perovskites were first introduced by Mitzi et al. [117] and have demonstrated excellent semiconducting properties. However, efficiency comparable to standard DSSCs (9.7%, in 2012) have been achieved only by replacing the electrolyte with a solid-state hole spiro-OMeTAD conductor [118]. Hence, the base technology for PSCs is solid-state sensitized solar cells that are based on dye-sensitized Graetzel solar cells. Then, PSCs evolved from the original “sensitized” architecture (see Fig. 25). PSCs configuration evolution has also helped to understand the fundamental principles of their charge generation and separation mechanisms. In fact, the high performances achieved (10.9% energy conversion efficiency in 2012 [118]) by using Al2O3 instead of TiO2 as scaffold material for CH3PbNH3I2Cl have demonstrated that organo-metal halides do not require any nanostructured n-type semiconductor to accept electrons injected from the perovskite conduction band. Since there is no need for a nanostructured semiconductor, PSC could be

A B X

Fig. 24 Crystal structure of CH3NH3PbX3 perovskites (X¼I, Br and/or Cl). The methylammonium cation (CH3NH3þ ) is surrounded by PbX6 octahedra. Available from: https://www.google.ca/search?q=crystal þ structure þ of þ perovskite þ images&dcr=0&tbm=isch&source=iu&pf= m&ictx=1&fir=ANsZs6E1AYu41M%253A%252CRvajOcDhCHIsXM%252C_&usg=__stqwvyeTlpJtJ44qg7r2xzPG7EQ%3D&sa=X&ved=0ahUKEwiveXYxYnXAhUL9YMKHWIZDN4Q9QEIOzAH#imgrc=ANsZs6E1AYu41M:.

142

Photovoltaic Materials

Sensitized perovskite solar cell

Thin-film perovskite solar cell

Cathode

Cathode

p-type contact

p-type contact

mp-TiO2 + perovskite

perovskite

n-type contact

n-type contact

Anode

(A)

Anode

(B)

h p-type contact

h+ + − h

e−

+ h+

perovskite

− e−

n-type contact (C)

n-type contact (D)

Fig. 25 Schematic of a sensitized perovskite solar cell (PSC) and its architecture from the “sensitized” (A) to the thin film structure (B). Sensitized: the active layer consists of a mesoporous TiO2 film coated with the perovskite absorber. The active layer is contacted with an n-type material for electron extraction (usually a compact thin layer of TiO2 deposited by spray pyrolysis) and a p-type material for hole extraction (spiroOMETAD). Thin film: a flat layer of perovskite is sandwiched between to selective contacts and function as both light absorber and carrier transport material. (C) Charge generation and extraction in the sensitized architecture. After light absorption in the perovskite absorber the photogenerated electron is injected into the mesoporous TiO2 through which it is extracted. The concomitantly generated hole is transferred to the HTM. (D) Charge generation and extraction in the thin film architecture. After light absorption both charge generation as well as charge extraction occurs in the perovskite layer. Downloaded from: https://crystallography365.files.wordpress.com/2014/12/figure-2.png?w=584&h=524.

fabricated without involving typical high temperature sintering stages (that are required for the mesoporous oxides used in DSSCs). PSCs seem to rely on a classical p-i-n junction heterojunction for charge separation (like an amorphous silicon thin film solar cell for example). The ambipolar nature of perovskites absorbers [119] has brought to high efficiency (17%–18%) planar p-i-n structures (Fig. 20) that can be produced with standard evaporation methods [120]. The physico-chemical process involved in a p-i-n perovskite-based solar cell can be summarized as follows: electron–hole pairs are created by the light driven promotion of an electron into the perovskite conduction band (the intrinsic layer with a thickness of 200–300 nm). Perovskite high permittivity and the low binding energy are sufficient to separate the geminate pairs at least at room temperature. Then electrons diffuse within the perovskite to ntype contact (a compact, planar layer of TiO2, 30–50 nm thick, that acts as a blocking layer to keep the spiro-OMETAD from touching the FTO) and holes diffuse towards the spiro-OMETAD. The two n-type and p-type contact materials (compact TiO2 and spiro-OMETAD respectively) possess conduction and valence band positions favoring electrons and holes injection at the two heterojunctions (CH3NH3PbX3/compact TiO2 and CH3NH3PbX3/spiro-OMETAD, respectively). Finally, electrons are collected at the FTO contact and holes at an evaporated (Au or Ag) contact. Table 1 reports high efficiency values for perovskite cells (B20%, although efficiencies of 22.1% have recently been reported by NREL in March 2016). However, two major challenges remain for commercialization of perovskites-based solar cells: materials stability and devices’ performance hysteresis. Organo-metal halides perovskites present thermodynamic stability issues. Besides the obvious problem of the decomposition of highly polar structures like the perovskites when exposed to moisture, methylammonium-lead halide perovskites start decomposing to the corresponding solid lead (II) halide and gaseous methylamine and hydrogen halide at moderate temperatures (601C) [121] precluding any long-term operation under solar light. Finally, the I-V response of the PSCs demonstrate anomalous dependence on the voltage scan direction/rate/range, voltage conditioning history, and device configuration [122]. The hysteresis behavior is not fully understood and accumulation of ions at the perovskite/HTM interface, ferroelectric effects and light induced electronic accumulation confined within the Debye length in the vicinity of contacts are usually invoked as explanation.

Photovoltaic Materials

2.5.4

143

Analysis and Assessment

Among the many renewable energy production and interconversion systems, PV has been gathering interest since the 1970s. Indeed, PV possesses plenty of advantages compared to other renewable energy sources such as geothermal, wind mills, hydroelectric and biomass:

• • • • •

PV systems are silent and visually unobtrusive. Small scale PV generation units can be easily installed on unused spaces (e.g., roofs, terraces) in existing buildings. PV cells operating in particular conditions (e.g., space use) can function for long time without any maintenance. Solar energy is a locally available renewable resource. PV systems (i.e., modules) can be adapted to any size for matching special output power requirements. Nevertheless, two main drawbacks are associated to the direct conversion of sunlight into electricity:

• •

The high manufacturing cost of solar cells, in both economical and energy consumption terms. The variability of solar power as energy source. In fact, power shortage may occur if solar modules are not coupled to energy storage facilities (e.g., batteries, hydrogen production for fuel cells) when PV generation is hindered for example by clouds or darkness.

These two issues are dealt with by: (1) optimizing solar cells’ materials parameters and environmental operating conditions to achieve high energy conversion and “exergy” efficiencies; (2) diversifying the production of energy stemming from PV materials. The exergy of a PV material can be defined as “the maximum useful work that could be obtained from the system at a given state in a specified environment (temperature)” [123]. In general, the exergy efficiency of an energy interconversion system can be expressed as Zex ¼

Exout Ex in

ð5Þ

where Exout and Exin are the output and input exergy, respectively. For the special case of solar cells equation (5) can be re-written as [124]   Im Vm 1 TTcella hca AðTcell Ta Þ   Zex ¼ ð6Þ 1 TTas Is A where Im and Vm are the current and the voltage of the solar cell at the maximum power point of operation (i.e., the “knee” of the I–V characteristics), respectively; Tcell, Ta and Ts are the cell, ambient and sun (taken as 5778K) temperatures, respectively. A is the active area of the PV device, and hca is the convective heat transfer which is related to the wind speed ((n)) by the formula: 5.7 þ 3.8 (n). Finally, Is represents the solar radiation’s intensity. Thus, the final exergy balance depends on cells’ manufacturing parameters that directly or indirectly impact on the Im and Vm and on the particular kind of application for which the PV device is used. The following section presents two examples of “unorthodox” application of PV materials proving how the use of solar cells can be “tailor-made” to fulfill special operative conditions. Two kinds of hybrid organic–inorganic PV devices (DSSCs and PSCs) are used as examples to illustrate the potential of semiconductor solar materials for purposes lying outside the traditional employment of PV structures.

2.5.5

Examples

An illustrative example of the different factors affecting the exergy efficiency has been reported by Zouri and Lee [123] for the special case of DSSCs. Such a study systematically investigates the effect of photoanode’s porosity and thickness, operating temperature and light intensity on the exergy balance of DSSCs. For instance, being the porosity the ratio of the volume of the emptiness of the photoanode over its total volume, a maximum Zex of 9.7% can be achieved by keeping the porosity of the nanostructured titanium dioxide at about 0.40–0.41. Indeed, such a porosity value assures large light absorption coefficients without affecting the electron diffusion length. Furthermore, the porous nature of the nanostructured photoanode renders DSSCs ideal for operation under high levels of diffuse radiation such as indoor PV generation or installation in regions characterized by the presence of clouds for most of the year [125]. In fact, photoanodes porosity tuning [9] allows for the collection of diffuse light (especially in the blue wavelengths) and ensures a high power output without using sun-tracking systems. Therefore, commercial DSSCs meant for repowering small portable electronic devices have been produced since 2009 (see Fig. 26) [126]. Two other crucial factors for DSSCs’ exergy efficiency maximization are the operating temperature and the level of sun irradiance. Needless to say, both parameters are affected by weather conditions such as seasonal change, wind speed, and cloudiness. Nevertheless, light concentration and cooling systems can be coupled with DSSCs modules for stationary applications in order to achieve the optimal light exposure and temperature conditions [127]. In fact, DSSCs can reach exergy efficiency close to 10% by operating at a temperature 300K and at 1 sun or less incident radiation intensity. Such conditions are the result of the delicate balance existing between fast charge transport (favoured by the electrolyte’s viscosity reduction caused by temperature

144

Photovoltaic Materials

Fig. 26 Backpack-integrated dye-sensitized solar cell (DSSC) module capable of providing a minimum output power of 0.5 W under 1 sun irradiation by GCell. Downloaded from: http://gcell.com/.

increases), high electron recombination lifetime (enhanced by temperature decreases), large levels of charge generation (improved by high sun irradiances) and low dissipation of solar energy into heat (favoured by low temperature and low light intensities). Hydrogen generation by sunlight constitute a possible methodology for diversifying energy production and supply electrical power through fuel cells when PV devices are incapacitated to operate (i.e., at night). Nonetheless, hydrogen production by photocatalytic or photoelectrochemical water splitting into O2 and H2 still remains in the experimental stages. The main reason for such a technology setback is the lack of semiconductor materials capable of matching all the requirements needed for high solarto-hydrogen energy conversion efficiencies. The principal obstacle to the photoelectrochemical hydrogen production is represented by the cathodic (for H2 evolution) and anodic (for O2 evolution) overpotentials involved in the water electrolysis process. Although thermodynamics predicts that 1.2 eV (the energy difference between the water oxidation and reduction potentials) are sufficient to decompose the H2O molecule, kinetics requirements bring the minimum potential necessary to split water up to 2.4–2.6 V [128]. Thus, wide band-gap semiconductors have been employed to perform water photoelectrochmical splitting. However, the large intra-bands energy gap prevents the absorption of a wide portion of the solar spectrum. Besides, the doping of ZnO, WO3, TiO2 aimed at reducing the semiconductor’s band gap usually results into low carriers’ lifetime and negligible light absorption properties improvements [129,130]. Because of the impossibility to carry out water photosplitting by means of a single semiconductor/electrolyte junction, combined approaches relying on semiconductor photoanode or photocathode coupled to PV devices have been tested for the last 20 years [131,132]. Nevertheless, the efficient use of tandem configurations for water photoelectrolysis is predicated on two conditions:

• •

The paired system needs to be comprised of low-cost semiconductor electrodes and solar cells. The semiconductor photoelectrode and the PV device providing the additional bias to decompose water must have complementary optical absorption properties.

To fulfill both requirements Brillet et al. proposed a tandem PEC for water splitting based on nanostructured tungsten or iron (III) oxides and DSSCs [133]. Recently (2015), also PSCs have been employed in conjunction with hematite (Fe2O3) photoanodes to drive the overall water photosplitting and verify the effectiveness of the dual-absorber tandem cell configuration [134]. Such a combination of materials stems from their particular optoelectronic properties. Indeed, hematite possesses a doping tunable band gap (ranging from 1.4 to 2.1 eV [135]) amenable for solar-to-hydrogen conversion efficiency as high as 16.8% [135]. However, two issues limit the hematite performances when applied to PECs [133,134]: (1) the short holes diffusion length that hinders their transfer to aqueous solution to evolve oxygen; (2) the unfavorable position of the Fe2O3 conduction band edge with respect to the H2/H þ redox potential (i.e., too positive compared to the NHE reference electrode). Thus, water photoelectrochemical splitting through hematite photoanodes can be accomplished only if an auxiliary source of electrical bias is used. The 1.5 eV band-gap of the perovskite material and the high photovoltage output (1.1 V) of the solar cell permits to exploit the wavelengths not absorbed by the hematite electrode and provide the external bias needed to drive the overall water decomposition reaction. However, the

Photovoltaic Materials

145

Fe2O3 O2

CH3NH3Pbl3 H2

Pt

Fig. 27 Schematic of the dual junction perovskite solar cell (PSC)/hematite photoanode tandem cell. Oxygen evolution occurs at the hematite photoanode that is series-connected to the gold contact at the Spiro-OMeTAD pole of a PSC. On the other hand, hydrogen is evolved at a standard platinum cathode connected to the metal-halide perovskite part of the photovoltaic (PV) device. Available from: https://www.researchgate.net/figure/ 283537789_fig1_Figure-1-a-Schematic-of-the-dual-junction-perovskite-solar-cellhematite-photoanode.

short electron diffusion length imposes to use thin (350–350 nm) Fe2O3 photoanodes. Hence, light absorption in the hematite electrode constitute the “bottle-neck” of the system that limits the solar-to-hydrogen conversion efficiency to 2%–3%. Nevertheless, such an approach represents a proof of concept for further improvement of water photoelectrolysis tandem cells based on perovskite devices. For instance, different semiconductors possessing longer carriers’ diffusion length (e.g., stabilized silicon photoelectrodes) and light trapping structures could be successfully used in conjunction with PSCs to generate hydrogen [134]. An example of a perovskite/hematite tandem cell is shown in Fig. 27.

2.5.6

Discussion and Future Developments

Although PV technology has seen its beginning in the 1960s, the S–Q limit for PV energy conversion efficiency, 33.7% for a singlejunction solar cell with band gap of 1.34 eV, has yet to be achieved. Indeed, the PV community has just begun the exploration of new trends leading to further major improvements in efficiency over the next decade. One important consideration is how relevant the record efficiencies are for a future application. Record devices are usually lab scale products, optimized within a controlled environment and designed for the purpose of breaking the efficiency record. These technologies, however, although useful to explore the physical limits of PV, do not necessarily translate to a standard factory environment, where production and management costs become an issue. It should be noted that, for instance, when comparing the efficiency of a module (or panel) with that of the corresponding lab cell, a decrease of the order of 2% is observed, that is a crystalline silicon solar cell, with record efficiencies of the order of 26%, will produce at best panels with conversion efficiencies of 24%. Therefore one of the technological challenges for a better PV output is to bring module efficiencies closer to that of the basic solar cell. Another consideration is on the relevance of the panel efficiency with respect to the specific application. PV technology, as indicated in several points, can find applications in an extraordinary diverse range of environments, each with its own requirements. Therefore an analysis of the energy needs for each application is necessary to assess the proper PV technology. For instance, space applications, or small area applications (like in single house roofs), require a PV panel that can provide the needed power using the allowed area, and therefore high efficiency (and more expensive) panels are needed. On the other hand, applications where the installation area is not a limit might do better with less efficient but cheaper technologies. A final point that needs to be made is that by calculating the cost of production of a solar cells, and obtaining the so-called $/Watt-peak ratio, we are actually examining this technology within a advanced society environment, where energy is provided by hydro, nuclear or gas. Therefore, the competitiveness of PV is related to these other technologies. On the other hand, most places on the planet do not have access to these standard forms of energy production (i.e., islands, deserts, underdeveloped countries, etc.), and the solar solution would already represent the most economic and feasible way to create energy. In this chapter we have introduced a variety of PV materials and device structures, from single crystals like Si and GaAs, which currently represent the high efficiency standards, and are very close to the SQ theoretical limit (meaning that efficiency can improve only a few percent with better technology), to thin films like CdTe, CIGS and Perovskite-based cell, which are getting close to the mono-crystalline efficiencies but present greater room for improvement. Specific applications based on the need of flexible substrates have driven the research towards thin film materials amenable to low temperature growth conditions. Among current solutions, PV paints can be used to apply cells onto a variety of surfaces,

Photovoltaic Materials

146

including paper and plastic sheets. Furthermore, glass covered with ultrathin films can be used as windows providing both functions of letting light in and producing electricity. Indeed, one can foresee in the future solar-powered house paint, road paving, and other advanced applications. An intriguing application of solar energy is that of the solar car. As mentioned above, such an application would require a very high efficiency panel, due to the obvious area limitations. However, current efficiencies do not allow for the possibility of an energy independent solar car. The most likely scenario is that of hybrid configurations, where the solar panels on the car are use to recharge batteries. Alternative applications, like photoelectrolysis, are also part of future developments. However, in order to achieve these goals, improvements are still needed, as indicated by the 2017 International Technology Roadmap for Photovoltaics. The document outlines the major areas of improvement expected in the next 10 years, including: more sophisticated cell structures (like PERC or PERL cells, discussed in this chapter), better current collection schemes, minimizing shadowing losses and/or increasing current collection. The report underlines that two opposite trends are being pursued in this regard, namely: increase the number of busbars (i.e., the metal contacts on the front surface) or remove them from the front surface altogether. Obviously the first approach aims at optimizing current collection in such a way that shadow losses become negligible, while the second aims at increasing the input area for light collection, moving all contacts to the back of the device. Another area of investigation for improving efficiency is related to the so-called cell-to-module power ratio (CTM). That is, cell efficiencies, prepared in labs, usually show at least a couple of percent higher efficiency values than the corresponding modules. Current research aims at bringing the CTM as close to 100% as possible. Finally, alternative device structure designs (e.g., tandem solar cells) can lead to the bypassing of the SQ limit.

2.5.7

Closing Remarks

The present chapter has presented an overview of the most widespread PV materials and technologies spanning from classical single-junction silicon to II–VI and III–V semiconductors, from bulk heterojunction to PSCs. Although mono-crystalline and polycrystalline silicon remain the dominant material on the market, new solutions aimed at reducing the production cost and maximizing the energy conversion efficiency are under constant development. In fact, Si solar cells limitations (i.e., relatively high manufacturing cost, insufficient response to blue and diffuse light, low carrier mobility, etc.) have been dealt with either by modifying the original structure of screen printed solar cells (e.g., PERL, rear contacts or bifacial devices) or by implementing new materials such as organo-lead halide perovskites. It has been shown that one can distinguish between different aspects of PV technology evolution. Firstly, materials properties are crucial to provide a proper conversion of light to electricity. Understanding of the limiting factors in a PV device is also crucial, as numerous solution can be attempted to bypass losses, including the combination of different materials. Device design and engineering is particular important as a way to overcome the SQ limitations, as in multi-junction cells. We have also shown how PV is a technology that can be easily adapted to specific applications, thanks to the diverse solutions indicated above. Therefore, understanding the needs of a particular application and designing the appropriate PV system is also a fundamental aspect of this technology. Finally, one cannot ignore the production cost aspect. Indeed, the most important parameter in PV is in fact the cost incurred in producing a watt of energy at peak conditions. This simple relation has driven most of the current research toward materials and design solutions aimed at lowering that ratio at competitive levels with respect to current standard energy sources.

Acknowledgment The work by the author was supported by the Natural Sciences and Engineering Research Council of Canada (NSERC).

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14]

Razykov TM, Ferekides CS, Morel D, Stefanakos E, Ullal HS, Upadhyaya HM. Solar photovoltaic electricity: current status and future prospects. Sol Energy 2011;85:1580. Terakawa A. Review of thin-film silicon deposition techniques for high-efficiency solar cells developed at Panasonic/Sanyo. Sol Energy Mater Sol Cells 2013;119:204. Wang X, Wang ZM. High-efficiency solar cells: physics, materials, and devices. New York: Springer; 2014. Liu M, Johnston MB, Snaith HJ. Efficient planar heterojunction perovskite solar cells by vapour deposition. Nature 2013;501:395. Hodes G. Perovskite-based solar cells. Science 2013;342:317. Ye S, Sun W, Li , Yan W, et al. CuSCN-based inverted planar perovskite solar cell with an average PCE of 15.6%. Nanoletters 2015;15:3723. Jorgensen M, Norrmann K, Gevorgyan SA, Tromholt T, Andreasen B, Krebs FC. Stability of polymer solar cells. Adv Materials 2012;24:580. Jorgensen M, Norrmann K, Gevorgyan SA, Krebs FC. Stability/degradation of polymer solar cells. Sol Energy Mater Sol Cells 2008;92:686. Cavallo C, Di Pascasio F, Latini A, Dini D. Nanostructured semiconductor materials for dye-sensitized solar cells. J Nanomater 2017;3:1. Wright M, Uddin A. Organic–inorganic hybrid solar cells: a comparative review. Sol Energy Mater Sol Cells 2012;107:1. Butler MA, Ginley DS. Review principles of photoelectrochemical. solar energy conversion. J Mater Sci 1980;15:1. Grätzel M. Photoelectrochemical hydrogen production. In: van de Krol Roel, editor. Electronic materials: science & technology. New York: Springer; 2010. Tomkiewicz M, Fay H. Photoelectrolysis of water with semiconductors. Appl Phys 1979;18. Shockley W, Queisser HJ. Detailed balance limit of efficiency of p–n junction solar cells. J Appl Phys 1961;32:510.

Photovoltaic Materials

147

[15] Masuko K, Shigematsu M, Hashiguchi T, et al. Achievement of more than 25% conversion efficiency with crystalline silicon heterojunction solar cell. IEEE J Photovolt 2014;4:1433. [16] Siebentriet S. Why are kesterite solar cells not 20% efficient? Thin Solid Films 2013;535:1. [17] Bandarenka H, Prischepa SL, Fittipaldii R, et al. Comparative study of initial stages of copper immersion deposition on bulk and porous silicon. Nanoscale Res Lett 2013;8:85. [18] Kholostov K, Serenelli L, Izzi M, Tucci M, Bernardi D, Balucani M. Electroplated nickel/tin solder pads for rear metallization of solar cells. IEEE J Photovolt 2015;6:404. [19] Boriskina SV, Chen G. Exceeding the solar cell Shockley–Queisser limit via thermal up-conversion of low-energy photons. Opt Commun 2014;314:71. [20] Song L, Wei A, Li Z, Liu J, Zhao Y, Xiao Z. Synthesis and up-conversion properties of Ho3 þ –Yb3 þ –F tri-doped TiO2 nanoparticles and their application in dyesensitized solar cells. Mater Res Bull 2017;88:71. [21] König D, Casalenuovo K, Takeda Y, et al. Hot carrier solar cells: Principles, materials and design. Phys E: Low-Dimensional Syst Nanostruct 2017;42:2862. [22] Salleo A. Flexible electronics, materials and applications series. New York: Springer; 2009. [23] Matsui T, Matsuda A, Kondo M. High-rate plasma process for microcrystalline silicon: over 9% Efficiency single junction solar cells. In: Ganguly G, Kondo M, Schiff EA, Carius R, Biswas R, editors. Amorphous and nanocrystalline silicon science and technology. Cambridge: Cambridge Unitversity Press; 2004. [24] Martinu L, Poitras D. Plasmadeposition of optical films and coatings: a review. J Vac Sci Technol 2000;A18:2619. [25] Cheng IC, Kattamis A, Long K, Sturm JC, Wagner S. Stress control for overlay registration in a-Si:HTFTs on flexible organic polymer foil substrates. Appl Phys Lett 2005;74:1177. [26] Gleskova H, Cheng IC, Sturm JC, Wagner S. Mechanics of thin-film transistors and solar cells on flexible substrates. Sol Energy 2006;80:687. [27] Daniel J, Arias AC, Crusor B, Lujan R, Street RA. The road towards large-area electronics without vacuum tools. Electrochem Soc Trans 2006;3:229. [28] Sun K, Xiao Z, Lu S, et al. A molecular nematic liquid crystalline material for high-performance organic photovoltaics. Nat Commun 2015;6:6013. [29] Rudolph D, Olibet S, Hoornstra J, et al. Replacement of silver in silicon solar cell metallization pastes containing a highly reactive glass frit: is it possible? Energy Procedia 2013;43:44. [30] Tutashkonko S, Kaminski Cachopo A, Boulord C, Ares R, Aimez V, Lemiti AM. Light induced silver and copper plating on silver screen-printed contacts of silicon solar cells. Opto-Electron Rev 2011;3:301. [31] Parida B, Iniyan S, Goic R. A review of solar photovoltaic technologies. Renew Sustain Energy Rev 2011;15:1625. [32] Guerrero-Lemus SR, Vega R, Kim Taehyeon, Kimm Amy, Shephard LE. Bifacial solar photovoltaics – a technology review. Renew Sustain Energy Rev 2016;60:1533. [33] Schwerzel RE, Brooman EW, Craig RA, et al. The tandem photoelectrolysis plant” concept: a strategy for fuel production via biomass conversion wastes (2016). (A80-29051 11-44) Solar energy: Chemical conversion and storage 1979;Clifton, NJ: Humana Press, Inc.; 1979. p. 83–115. [34] Fujishima A, Honda K, Kikuchi S. Photochemical reactions of semiconductors. I. Photosensitized electrolytic oxidation on semiconducting n-type TiO2 electrode. Kogyo Kagaku Zasshi 1969;72:108. [35] Streetman B, Sbanerjee B. Solid state electronic devices. 6th ed. Upper Saddle, NJ: Prentice-Hall; 2005. [36] Tze SM, Kwok KG. Physics of semiconductor devices. 3rd ed. New York: Wiley; 2007. [37] Jester TL. Crystalline silicon manufacturing progress. Prog Photovolt. 2002;72:99. [38] Moreno M, Daineka D, Roca P, Cabarrocas I. Plasma texturing for silicon solar cells: from pyramids to inverted pyramids-like structures. Sol Energy Mater Sol Cells 2010;94:733. [39] Li J, Ban B, Li Y, Ba X, Zhang T, Chen Jian. Removal of impurities from metallurgical grade silicon during Ga–Si solvent refining. Silicon 2017;9:77. [40] Safariana J, Tranella G, Tangstada M. Processes for upgrading metallurgical grade silicon to solar grade silicon. Energy Procedia 2012;20:88. [41] Gribov J, Zinovev KV. Preparation of high-purity silicon for solar cells. Inorg Mater 2003;39:653. [42] Bathey BR, Cretella MC. Solar-grade silicon. J Mater Sci 1982;17:3077. [43] Sancier KM. Process and apparatus for obtaining silicon from fluosilicic acid. US Patents 4,529,576, 1985. [44] Sanjurj A. Process and apparatus for obtaining silicon from fluosilicic acid. US Patents 4,753,783, 1988. [45] Wolf RA, Meyer CB. Method of producing silicon. US Patents 4,188,368, 1980. [46] Sakaguchi S. Methods and apparatus for producing silicon PCT International Patent WO2007/119605, 2007. [47] Honda S, Yasuda M, Hayashida S, Yamaguachi M. Process for the production of high-purity polycrstalline silicon, Japanese patent application H19-145663, 2007. [48] Pearson TG, Phillips H. The production and properties of super-purity aluminium. Metall Rev 1957;2:305. [49] Olsen E, Rolseth S. Three-layer electrorefining of silicon. Metall Mater Trans B 2010;41:295. [50] Yasuda K, Nohira T, Hagiwara R, Ogata YH. Direct electrolytic reduction of solid SiO2 in molten CaCl2 for the production of solar grade silicon. Electrochim Acta 2007;53:106. [51] Fishman OS. Solar silicon. Adv Mater Process 2008;33. [52] Lindroos J, Savin H. Sol Energy Mater Sol Cells 2016;147:115. [53] Ma X, Zhang J, Wang T, Li T. Hydrometallurgical purification of metallurgical grade silicon. Rare Met 2009;28:221. [54] Ceccaroli B, Friedsta K. Refining of metallurgical grade silicon US Patents 6,861,040, 2005. [55] Teixeira L, Morita K. Removal of boron from molten silicon using CaO–SiO2 based slags. ISIJ Int 2009;49:783. [56] Johnston MD, Barati M. Distribution of impurity elements in slag-silicon equilibria for oxidative refining of metallurgical silicon for solar cell applications. Sol Energy Mater Sol Cells 2010;94:2084. [57] Pires J, Otubo J, Braga A, Mei P. The purification of metallurgical grade silicon by electron beam melting. J Mater Process Technol 2005;169:16. [58] Yuge N, Abe M, Hanazawa K, et al. Purification of metallurgical grade silicon up to solar grade. Prog Photovolt: Res Appl 2001;9:193. [59] http://meroli.web.cern.ch/meroli/Lecture_silicon_floatzone_czochralski.html [60] Sanchez-Garcia MA, Bokhimi X, Maldonado-Alvarez A, Jimenez-Gonzalez AE. Efficient screen-printed n-type solar cells using a spin-on source and thermal oxide/silicon nitride passivation. IEEE J Photovolt 2011;2:146. [61] Sanchez-Garcia MA, Bokhimi X, Maldonado-Alvarez A, Jimenez-Gonzalez AE. Minority carrier lifetime degradation in boron-doped Czochralski silicon. J Appl Phys 2001;90:2397. [62] Radu R, Fretwurst E, Klanner RT, Lindstroem G, Pintilie I. Radiation damage in n-type silicon diodes after electron irradiation with energies between 1.5 MeV and 15 MeV. Nucl Instrum Methods Phys Res A: Accelerators Spectrometers Detectors Assoc Equipment 2013;730:84. [63] Sopori B, Basnyat P, Devayajanam S, et al. Understanding light-induced degradation of c-Si solar cells. In: 38th IEEE photovoltaic specialists conference (PVSC); 2012. p. 001115–001120). [64] Ferrazza F. Polycrystalline semiconductors IV – physics, chemistry and technology. In: Pizzini S, Strunk HP, Werner JH, editors. Solid state phenomena. Zug, Switzerland: Trans. Tech. Publ.; 1995. [65] Möller HJ. In: Jackson KA, Schröter W, editors. Handbook of semiconductor technology – electronic structure and properties of semiconductors. vol. 1. Weinheim, Germany: Wiley-VCH; 2000, p. 715. [66] Helmreich D. In: Symposium on electronic and optical properties of polycrystrystalline or impure semiconductors and novel silicon growth methods. Princeton: Electrochemical Society; 1977. p. 184. [67] Kawamura R, Sasatani K, Onizuka T. Operation and crystalline properties of electro-magnetic casting. In: Proceedings of 1994 IEEE 1st world conference on photovoltaic energy conversion – WCPEC (a joint conference of PVSC, PVSEC and PSEC); 1994. p. 1652–5.

148

[68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80] [81] [82] [83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105] [106] [107] [108] [109] [110] [111] [112] [113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128]

Photovoltaic Materials

A good review of basic PV and of the optimization techniques described here can be found in the following educational web-site: http://pveducation.org/pvcdrom. Shih YC, Shi FG. Silicon solar cell metallization pastes, materials for advanced packaging. Basel: Springer International Publishing; 2017. Schmidt J, Werner F, Veith B, et al. Advances in the surface passivation of silicon solar cells. Energy Procedia 2015;15:30. Dai BT, Chen HC. Handbook of solar cells and battery technology. The Electronics and Materials Association; 2008. Forbes L. Texturing, reflectivity, diffuse scattering and light trapping in silicon solar cells. Sol Energy 2017;86:319. Brenner P. Printing photovoltaics. Industrial þ Specialty Printing 2010;1(01):26. Green MA. Silicon solar cells: advanced principles and practice. Sydney: Bridge Printery; 1995. Shanmugam V, Wong J. Analysis of fine-line screen and stencil-printed metal contacts for silicon wafer. IEEE J Photovolt 2015;5:525. Swanson. Photovoltaics: the path from niche to mainstream supplier of clean energy. In: Solar power 2006, San Jose, October; 2006. http://solarlove.org/concepts-photovoltaic-technology/ Luque A, Cuevas A, Ruiz JM. Double-sided nþ-p-nþ solar-cell for bifacial concentration. Sol Cells 1980;2:155. Cuevas A, Luque A, Eguren J, DelAlamo J. 50%more output power from an albedo-collecting flat panel using bifacial solar cells. Sol Energy 1982;29:419. Guerrero-Lemus R, Vega R, Taehyeon K, Kimm A, Shephard LE. Bifacial solar photovoltaics – a technology review. Renew Sustain Energy Rev 2016;60:1533. Rothhardt P, Demberger C, Wolf A, Biro D. Co-diffusion from AP-CVD BSG and POCl3 for industrial n-type solar cells. EnergyProcedia 2013;38:305. Kania D, Böscke TS, Braun M, Sadler P, Schöllhorn C, Dupke M. Pilot line production of industrial high-efficient bifacial n-types silicon solar cells with efficiencies exceeding20.6%. In:Proceedings of the 28th Europeanphotovoltaic solar energy conference, Paris, France; 2013. p. 1383–85. Yang L, Ye QH, Ebong A, et al. High efficiency screen printed bifacial solar cells on mono-crystalline CZ silicon. Prog Photovolt: Res Appl 2011;19:275. https://www.pv-tech.org/tags/bifacial Wu W, Roelofs KE, Subramoney S, Lloyd K, Zhang L. Role of aluminum in silver paste contact to boron-doped silicon emitters. AIP Adv 2017;7:015306. Zhu F, Singh J. Approach to study the relation between optical energy gap and hydrogen concentration in hydrogenated amorphous silicon. J Appl Phys 1993;73:4009. Gaspari F, O’Leary SK, Zukotynski S, Perz JM. The dependence of the dark conductivity of hydrogenated amorphous silicon films on the hydride content. J Non-Crystal Solids 1993;155:149. Staebler DL, Wronski CR. Appl Phys Lett 1977;31:292. Gaspari F. Optoelectronic properties of amorphous silicon the role of hydrogen: from experiment to modeling. In: Predeep P, editor. Optoelectronics–materials and techniques. InTech.; 2011. ISBN: 978-953-307-276-0. Matsui T, Sai H, Suezaki T, et al. Development of highly stable and efficient amorphous silicon based solar cells. In: Proceedings of the 28th European photovoltaic solar energy conference and exhibition; 2013. p. 2213–17. doi: 0.4229/28thEUPVSEC2013-3DO.7.2 Hänni S, Bugnon G, Parascandolo G, et al. High‐efficiency microcrystalline silicon single‐junction solar cells. Prog Photovolt: Res Appl 2013;31:821. Sai H, Koida T, Matsui T, Yoshida I, Saito K, Kondo M. Microcrystalline silicon solar cells with 10.5% efficiency realized by improved photon absorption via periodic textures and highly transparent conductive oxide. Appl Phys Express 2013;6:104101. Multone X, Fesquet L, Borrello D, et al. Triple-junction amorphous/microcrystalline silicon solar cells: towards industrially viable thin film solar technology. Sol Energy Mater Sol Cells 2015;140:388. Moss SJ, Ledwith A. The chemistry of the semiconductor industry. New York: Springer; 1987. ISBN 0-216-92005-1. Hayes GJ, Clemens BM. Laser lift off of gallium arsenide thin films. MRS Commun 2015;5:1. Li E, Chaparala P. High-efficiency GaAs thin-film solar cell reliability. In: NREL PV module reliability workshop, February 26–27. Alta Devices, Inc; 2013. Geisz JF, Steiner MA, Garcia I, Kurtz SR, Friedman DJ. Enhanced external radiative efficiency for 20.8% efficient single-junction GaInP solar cells. Appl Phys Lett 2013;103:041118. Hetzer MJ. Chemical and electronic characterization of copper indium gallium diselenide thin film solar cells and correlation of these characteristics to solar cell operation [Dissertation] Ohio State University; 2009. Bube RH. Photovoltaic materials. (chapter 6) London: Imperial College Press; 1998. ISBN: 186094065X. Flisom AG. www.myscience.ch/innovation/spino/isom Jackson P, Hariskos D, Wuerz R, et al. Properties of Cu(In,Ga)Se2 solar cells with new record efficiencies up to 21.7%. Phys Status Solidi RRL 2015;9.28–31. doi:10.1002/pssr.201409520. Katagiri H, Jimbo K, Shwe Maw W, et al. Thin Solid Films 2009;107:2460. Todorov TK, Reuter KB, Mitzi DB. High-efficiency solar cell with Earth-abundant liquid-processed absorber. Adv Mater 2010;22:E156. Morales-Acevedo A. Thin film CdS/CdTe solar cells: research perspectives. Sol Energy 2010;80:675. Buecheler S, Kranz S L, Perrenoud J, Tiwari AN. CdTe solar cells. In: Solar energy. New York: Springer; 2013. McCandless BE, Birkmire RW. Analysis of post deposition processing for CdTe/CdS thin film solar cells. Sol Cells 1991;31:527. Hains AW, Liang Z, Woodhouse MA, Gregg BA. Molecular semiconductors in organic photovoltaic cells. Chem Rev 2010;110:6689. Servaites JD, Ratner MA, Marks TJ. Organic solar cells: a new look at traditional models. Energy Environ Sci 2011;4:4410. Clarke TM, Durrant JR. Charge photogeneration in organic solar cells. Chem Rev 2010;110:6736. Derouichea H, Djarab V. Impact of the energy difference in LUMO and HOMO of the bulk heterojunctions components on the efficiency of organic solar cells. Sol Energy Mater Sol Cells 2007;91:1163. Mathew S, Yella A, Gao P, et al. Dye-sensitized solar cells with 13 efficiency achieved through the molecular engineering of porphyrin sensitizers. Nat Chem 2014;6:242. Graetzel M, O’Regan B. A low-cost, high-efficiency solar cell based on dye-sensitized colloidal TiO2 films. Nature 1991;753:737. http://www.dyesol.com/media/wysiwyg/Documents/dsc-resource-library/What_Physical_Factors_Affect_Current-Voltage_Characteristics_of_Dye_Solar_Cells-Dyesol-HansDesilvestro-2008.pdf. Hardin BE, Snaith HJ, McGehee MD. The renaissance of dye-sensitized solar cells. Nat Photon 2012;6. www.nature.com/naturephotonics.. Woronowicz K, Ahamed S, Biridar A. Near-ir absorbing solar cell sensitized with bacterial photosynthetic membranes. Photochem. Photobiol 2012;88:1467. Park NG. Perovskite solar cells: an emerging photovoltaic technology. Mater Today 2015;18:65. Mitzi DB, Chondroudis K, Kagan CR. Inorg Chem 1999;38:6246. Nazeeruddin MK, Grätzel M. Status and progress in dye and perovskite sensitized solar cells. In: 29th European photovoltaic solar energy conference and exhibition, September 22–26; 2014. Lee MM, Teuscher J, Miyasaka T, Murakami TN. Efficient hybrid solar cells based on meso-superstructured organometal halide perovskites. Science 2012;338:647. Chen Y, Peng J, Su D, Chen X, Liang Z. Efficient and balanced charge transport revealed in planar perovskite solar cells. Appl Mater Interfaces 2015;7:4471. Lee MM, Teuscher J, Miyasaka T, Murakami TN. Efficient planar heterojunction perovskite solar cells by vapour deposition. Nature 2013;338:395. Peumans P, Forrest SR. Separation of geminate charge-pairs at donor/acceptor interfaces in disordered solids. Chem Phys Lett 2004;398:27. Zouri K, Lee SY. Analysis and assessment of dye-sensitized solar cell at different materials parameters and environmental conditions. Int J Energy Res 2016;40:1093. Saidur R, BoroumandJazi G, Mekhlif S, Jameel M. Exergy analysis of solar energy applications. Renew Sustain Energy Rev 2012;16:350. Bourzac K. Wrapping solar cells around an optical fiber. Technology Review 2009 Retrieved 31 October 2009. http://newatlas.com/first-commercial-application-dssc-solar-technology/13100/. Bisquert J, Fabregat-Santiago F, Kalyanasundaram K. Dye-sensitized solar cells. Boca Raton,FL: CRC Press; 2010. Feng X, LaTempa TJ, Basham JI, Mor GK, Varghese OK, Grimes CA. Ta3N5 nanotube arrays for visible light water photoelectrolysis. Nano Lett 2010;10:4774.

Photovoltaic Materials

149

[129] Sakthivel S, Kisch H. Daylight photocatalysis by carbon-modified titanium dioxide. Angew Chem, Int Ed 2003;42:4908. [130] Serpone N. Is the band gap of pristine TiO2 narrowed by anion- and cation-doping of titanium dioxide in second-generation photocatalysts? J Phys Chem B 2010;110:24287. [131] Gao X, Kocha S, Frank A. Photoelectrochemical decomposition of water using modified monolithic tandem cells. Int J Hydrogen Energy 1999;24:319. [132] Kelly N, Gibson T. Design and characterization of a robust photoelectrochemical device to generate hydrogen using solar water splitting. Int. J. Hydrogen Energy 2006;31:1658. [133] Brillet J, Yum JH, Cornuz M, et al. Highly efficient water splitting by a dual-absorber tandem cell. Nat Photon 2012;6:824. [134] Gurudayal G, Sabba D, Kumar MH, et al. Perovskite–Hematite tandem cells for efficient overall solar driven water splitting. Nano Lett 2015;15:3833. [135] Xia C, Jia Y, Tao M, Zhang Q. Tuning the band gap of hematite a-Fe2O3 by sulphur doping. Phys Lett A 2013;377:1943.

Further Reading Conibeer GJ, Willoughby A, editors. Solar cell materials Chichester, West Sussex: Wiley; 2014. Gaspari F. Optoelectronic properties of amorphous silicon the role of hydrogen: from experiment to modeling, optoelectronics – materials and techniques. In: Predeep P, editor; 2011. Green MA. Third generation photovoltaics, springer series in photonics, vol. 12. Berlin/Heidelberg: Springer; 2013. Hardin BE, Snaith HJ, McGehee MD. The renaissance of dye-sensitized solar cells. Nat Photon 2012;6. www.nature.com/naturephotonics. Knobloch J, Noel A, Schaffer E, et al. High-efficiency solar cells from FZ, CZ and MC silicon material. In: Conference record of the 23rd IEEE photovoltaic specialists conference; 1993. Krebs FC, Jørgensen M. Polymer and organic solar cells viewed as thin film technologies: What it will take for them to become a success outside academia. Sol Energy Mater Sol Cells 2013;119:73–6. Lee BG, Eisenlohr J, Feldmann F, et al. High efficiency Si solar cells with photonic crystal rear reflector. In: 31st European photovoltaic solar energy conference and exhibition; 2015. McEvoy A, Castaner L, Markvart T. Solar cells. Materials, manufacture and operation, 2012;2nd ed. Oxford: Elsevier; 2012. Möller HJ. In: Jackson KA, Schröter W, editors. Handbook of semiconductor technology – electronic structure and properties of semiconductors. vol. 1. Weinheim, Germany: Wiley-VCH; 2000. Nakayashiki K, Rounsaville B, Yelundur V, et al. Fabrication and analysis of high-efficiency string ribbon Si solar cells. Solid-State Electron 2006;50:1406. Nelson J. The physics of solar cells. London: Imperial College Press; 2003. Park NG. Perovskite solar cells: an emerging photovoltaic technology. Mater Today 2015;18(2):65–71. Shah A, editor. Thin-film solar cells Boca Raton, FL: CRC Press; 2010. Sopori B. Ed., P. Basnyat and S. Devayajanam, guest editors, Photovoltaic materials and devices. International Journal of Photoenergy, Hindawi. (Note: “Photovoltaic Materials and Devices” is an annual special issue published in “International Journal of Photoenergy.”); 2017. Yang QH, Ye , Ebong A, Song WT, Zhang GJ, Wang JX, Ma Y. High efficiency screen printed bifacial solar cells on mono-crystalline CZ silicon. Prog Photovolt: Res Appl 2011;19:275. Yasuda K, Saegusa K, Okabe TH. Production of solar-grade silicon by halidothermic reduction of silicon tetrachloride. Metall Mater Trans B – Process Metall Mater Process Sci 2011;42:37.

Relevant Websites https://www.ise.fraunhofer.de/content/dam/ise/de/documents/publications/studies/AgoraEnergiewende_Current_and_Future_Cost_of_PV_Feb2015_web.pdf. Agora, Energiewende – Long-term scenarios for market development, system prices and LCOE of utility-scale PV Systems. Fraunhofer ISE (2015). http://www.iea-pvps.org/. An iea Technology Collaboration Programme – Photovoltaic power systems programme of the International Energy Agency (IEA). A wealth of information and IEA reports. https://www.photovoltaic-conference.com/. EU PVSEC –A website with annual conference proceedings on latest developments in PV. https://www.nrel.gov/pv/. NREL – PV research and development at the National Renewable Energy Laboratory (NREL), including annual solar chart on PV efficiencies. http://www.pveducation.org/. PV EDUCATION.ORG – A website on the physics of solar cells, with several sections on semiconductor materials, physics and devices. https://www.pv-magazine.com/news/. pv magazine. http://www.pvresouces.com/. PVTECH – German and European newsletter. http://pvwatts.nrel.gov/. PVWatts Calculator – NREL solar power calculator. http://www.solaraccess.com/. solaraccess – United States site for solar commercial organizations. http://www.solarenergy.org/. Solar Energy International: renewable energy education and training. http://www.top50-solar.de/en/member/id/4836.html. Top50 Solar –A portal devoted entirely to issues of renewable energies and energy saving for the construction industry. http://www.top50-solar.de/en/member/id/73/solarserver.com.html. Top50 Solar – extensive information on solar thermal, photovoltaics, solar buildings and renewables in German and English.

2.6 Dye-Sensitized Materials Yun Hang Hu and Wei Wei, Michigan Technological University, Houghton, MI, United States r 2018 Elsevier Inc. All rights reserved.

2.6.1 Introduction 2.6.2 Structure and Principle of Dye-Sensitized Solar Cells 2.6.3 Photoelectrodes 2.6.3.1 Semiconductor 2.6.3.2 Dyes 2.6.4 Electrolyte 2.6.5 Counter Electrodes 2.6.5.1 Carbon Materials 2.6.5.1.1 Graphene 2.6.5.1.2 Carbon nanotube 2.6.5.2 Transition Metal Compounds 2.6.5.2.1 Metal sulfides and oxides 2.6.5.2.2 Transition metal carbides, nitrides, and phosphides 2.6.5.3 Conductive Polymers 2.6.5.3.1 Porous poly(3,4-propylenedioxythiophene) 2.6.5.3.2 Polypyrrole 2.6.5.3.3 Polyaniline 2.6.5.3.4 Polythiophene 2.6.5.3.5 Poly(3,4-ethylenedioxythiophene) 2.6.6 Commercialization of Dye-Sensitized Solar Cells 2.6.6.1 Photovoltaic Market Overview 2.6.6.2 Applications 2.6.6.2.1 Portable charging 2.6.6.2.2 Flexible dye-sensitized solar cells and wearable electronics 2.6.6.2.3 Decorative elements 2.6.6.2.4 Building integrated photovoltaics 2.6.6.2.5 Automotive 2.6.6.3 Future Direction 2.6.7 Closing Remarks Acknowledgment References

Nomenclature ALD CE CNT CVD DSSC DWCNT EIS ff FTO GW IMPS IMVS

150

Atomic layer deposition Counter electrode nanotube Chemical vapor deposition Dye-sensitized solar cell Double-walled carbon nanotube Electrochemical impedance spectroscopy Fill factor Fluorine-doped tin oxide Gigawatt Intensity-modulated photocurrent spectroscopy Intensity-modulated photovoltage spectroscopy

IPCE Jsc lx MWCNT PV Rct Rw SWCNT TCD TMC TVD TW Voc Z

151 151 152 152 154 156 156 157 157 159 161 161 161 162 163 164 165 165 165 165 165 166 166 167 169 169 172 172 174 174 174

Intensive photon-to-current conversion efficiency Short-circuit current Luminous flux Multi-walled carbon nanotube Photovoltaic Charge-transfer resistance Electron transport resistance Single-walled carbon nanotube Transient-photocurrent decay Transition metal compounds Transient-photovoltage decay Terawatt Open-circuit voltage Power conversion efficiency

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00216-9

Dye-Sensitized Materials

2.6.1

151

Introduction

The current consumption of power is 13 terawatts (TW) worldwide. This amount is likely to increase to 30 TW by the year 2050 due to the increase of the population and the improvement of the living standard [1]. If only fossil fuels are exploited to meet this requirement, the concentration of carbon dioxide in the atmosphere will be more than double, enhancing the global warming. Therefore, it is a great challenge to provide the world with 30 TW power without contribution to environmental issues. Light from the sun is the ideal source of energy. The sun deposits 120,000 TW of radiation on the earth, which is the largest single available source of clean energy. Although technologies have been developed to exploit solar energy efficiently, they are not yet an economically viable alternative to fossil fuels [2]. So far, only about 5 gigawatt (GW) power is generated from solar energy by solar cells [3]. A combination of increased energy prices and fears over global warming are pushing up demand for solar cells [4,5]. The vast majority of solar cells on the market are single-junction silicon wafer devices including single crystal and multicrystalline silicon, which are known as first generation solar cells [6]. Their main drawback is associated with their indirect optical band gap that requires a thick active layer for the solar conversion and thus costly fabrication of large-area materials. Furthermore, those cells suffer many efficiency losses for energy conversion, such as “red losses” that photons with energies below the band gap of the device cannot be absorbed and “blue losses” that photons with energies above the band gap lose their excess energy as heat. The aim of the second generation of photovoltaic (PV) material was to reduce the fabrication cost through the deposition of thin films [6], particularly Si thin films and Cu-based ternary semiconductor films, such as CdS/CuInSe2 and Cu(In, Ga)Se2 [7–11]. Those thin film devices with a single-junction constitute a rapidly growing segment of the solar cell market [6]. However, the second generation devices share the same performance restrictions as the first generation ones. Third generation PVs are designed to exceed the limits of single-junction devices, to lead to a high efficiency, and to reduce production costs. There are various attempts to develop the third generation cells, such as, dye-sensitized solar cells (DSSCs), perovskite solar cells, multi-junction cells, intermediate-band cells, hot carrier cells, and organic cells [1,4,6,12–20]. As a representative of third generation PVs, DSSCs have attracted much attention due to their low cost, easy fabrication, and efficient solar energy conversion [21]. In this chapter, the structure, principle, and materials of DSSCs are discussed. Furthermore, advances in materials for DSSCs are emphasized.

2.6.2

Structure and Principle of Dye-Sensitized Solar Cells

The sensitization of semiconductor materials with dyes started in 1960s, but the real viability of DSSCs was created by O’Regan and Gratzel in 1991 [22]. They pioneered the utilization of high surface area nanoparticle TiO2 as a semiconductor film and a trinuclear Ru dye as a sensitizer for DSSCs, leading to a jump in power conversion efficiency (PCE) from B1% to 7% under AM 1.5 illumination [22]. Furthermore, this breakthrough has established the state-of-the-art DSSCs for two decades: a mesoporousnanoparticle-TiO2 film, a dye, and the I3 /I redox shuttle in a nonaqueous solvent. Therefore, a typical DSSC consists of three basic components: sensitized photoanode, electrolyte, and counter electrode [23–29]. A diagram illustrating the energetics and key kinetic processes occurring in a DSSC is shown in Fig. 1(A). A DSSC presents three important steps to convert sunlight into electrical energy: It relies on the photo-excitation of dyes triggering an electron transfer into the conduction band of the metal oxide semiconductor (generally TiO2), followed by regeneration of the oxidized dye molecules by the electron donation from the redox couple in the electrolyte, and finally migration of electron through the external load to complete the circuit [30,31]. These processes compete with several reactions that inhibit the efficient operation of a DSSC

TiO2

Light LUMO

CB

e−

I−

hv

3I−/I3−

I3−

HOMO Dye TiO2

(A)

20

Pt Current density (mA cm−2)

FTO glass

Dye e−

15 Pmax

Jsc

10 ff =

Pmax Jsc.Voc

5 Voc 0 0.8

FTO glass (B)

0.6

0.4 Voltage (V)

0.2

0.0

Fig. 1 (A) Scheme diagram of dye-sensitized solar cells and (B) simulated photocurrent density vs. voltage curve with Jsc ¼17 mA cm 2, Voc ¼0.82 V, and ff ¼0.82 (values corresponding to an 11% efficient cell in response to 100 mW cm 2 illumination). CB, conduction band; FTO, fluorine-doped tin oxide; HOMO, highest occupied molecular orbit; LUMO, lowest unoccupied molecular orbit. Reproduced from Hamann TW, Ondersma JW. Dye-sensitized solar cell redox shuttles. Energy Environ Sci 2011;4:370–81.

152

Dye-Sensitized Materials

including decay of the dye’s excited state prior to injection as well as back-electron transfer from the TiO2 nanoparticle to the oxidized dye or the oxidized form of the redox shuttle [32–38]. The PCE (Z) of a solar cell is the key parameter to evaluate the solar cells performance. It is the ratio of the maximum electrical power output (Pmax) to the incident solar power (Pin) (Fig. 1(B)). The maximum electrical power output is the product of the photocurrent density and the voltage at the power point, usually expressed as Z¼

Pmax Jsc  Voc  ff ¼ Pin Pin

where Jsc is the short-circuit photocurrent density, Voc the open-circuit photovoltage, and ff the fill factor. The maximum current that can run through the cell at zero applied voltage is called the short-circuit current (Jsc) that occurs at the beginning of the forward-bias sweep. For an ideal cell, Jsc is the total current produced in the solar cell by photon excitation. Similarly, the opencircuit voltage (Voc) is the maximum photovoltage that can be generated in the cell and corresponds to the voltage where current under illumination is zero. Fill factor (ff) is defined as the ratio of the maximum power to the theoretical power that is the product of the open-circuit voltage and short-circuit current. A larger ff is desirable, corresponding to a J–V sweep that is more square shape. The equivalent circuit modeling is widely used to evaluate the device performance for DSSCs. Various techniques of frequencydomain methods have been developed for DSSC characterization, such as, electrochemical impedance spectroscopy (EIS), intensity-modulated photovoltage spectroscopy (IMVS), and intensity-modulated photocurrent spectroscopy (IMPS), and timedomain approaches, such as, charge extraction (CE), transient-photocurrent decay (TCD), and transient-photovoltage decay (TVD) [39–43].

2.6.3

Photoelectrodes

In DSSCs, a dye-sensitized nanocrystalline semiconductor film as a photoanode plays a significant role in converting photons into electrical energy [44–46]. The performance of a photoanode, which would efficiently facilitate light harvesting, electron injection, and electron collection [47], are strongly dependent on the properties of a semiconductor layer and dye molecules.

2.6.3.1

Semiconductor

The semiconductor can strongly influence the photovoltage, the ff, and the photon-to-current conversion efficiency (IPCE), which is determined by the light-harvesting efficiency of the dye, the quantum yield of electron injection, and the efficiency of collecting the injected electrons [48–50]. The ideal semiconductor layer of DSSCs should have a nanostructured mesoscopic morphology to obtain a high specific surface area for dye adsorption. So far, many semiconductors have been explored as photoanodes in DSSCs, and nanostructured TiO2 is the best due to its superior characteristics, such as, higher conduction band edge, surface area, dye loading and electron affinity. The designs of TiO2-based DSSCs were continuously renovated since 1991 to enhance the efficiency by increasing the light-scattering properties of the TiO2 film, suppressing charge recombination, improving the interface, and altering TiO2 particle morphology [51]. TiO2 nanoparticle based cells exhibit an efficiency of 13%, which is the highest among the materials investigated [52]. The main crystalline forms of TiO2 are anatase and rutile. Anatase has been widely used, because it has a high band-gap energy (3.2 eV, and absorbs only below 388 nm) making it invisible to most of the solar spectrum, reducing the recombination rate of photoinjected electrons [53]. In contrast, rutile has a higher dark current (Eg ¼3.0 eV) and thus it is less effective [54]. Furthermore, its photon excitation within the band gap generates holes that act as oxidants making it less chemically stable. Therefore, anatase is perceived as the more active phase of TiO2 [55–59]. However, rutile has a lower cost for its production and the superior lightscattering characteristics, which is a beneficial property for effective light harvesting [60,61]. Similar Voc has been reported for anatase and rutile despite the difference in their conduction bands. The observed Jsc using rutile was about 30% lesser than that of the anatase, which was attributed to its lower surface area, decreasing the amount of dye loading [39]. The film morphology is a major variability factor in DSSCs performance. Nanostructured morphology, such as, nanotubes [62], nanowires [24], and nanorods [63] were explored for TiO2, leading to a PCE of 4.13%, 9.93%, and 7.29%, respectively (Fig. 2). The highlight to electricity conversion was attributed to the network structure of TiO2, which would accelerate the transportation of electrons and thus yield a high conversion efficiency of light-to-electricity. However, further increase in conversion efficiency is difficult due to large energy loss from serious recombination between electrons and either oxidized dye molecules or electronaccepting species in the electrolyte during the charge-transport process [64]. Such a recombination is related to the lack of a depletion layer on TiO2 nanocrystallite surfaces [65]. Therefore, efforts to improve electron transport in TiO2 was made. For example, Hu’s group developed a simple approach to introduce graphene into TiO2 for the increase of electron transport [66]. As a result, the incorporation of graphene could increase the short-circuit current density and PCE by 52.4 and 55.3%, respectively. ZnO, which is another wide band gap semiconductor with high electron mobility, has attracted much attention as a fascinating alternative to TiO2 photoanode in DSSCs [46,67–86]. ZnO and TiO2 exhibit similar lowest conduction band edges and electron injection process from the excited dyes. Furthermore, the lifetime of carriers in ZnO is significantly longer than that in TiO2. Therefore, the rapid electron transferring and collecting in ZnO-based DSSCs are expected to reduce recombination. A maximum efficiency of 7.5% with short-circuit current of 19.8 mA cm 2, open-circuit voltage of 640 mV, and ff of 0.59 was reported for cells based on ZnO hierarchical aggregates [87]. Although the conversion efficiencies obtained for ZnO are much lower than that for

Dye-Sensitized Materials

(A)

153

(B)

(C)

(D)

Fig. 2 (A) Field emission scanning electron microscope (FESEM) micrographs showing the cross-section of TiO2 nanotubes, (B) FESEM micrographs showing the top view of TiO2 nanotubes, (C) transmission electron microscopy (TEM) image of TiO2 nanowire, (D) high-resolution transmission electron microscopy image of TiO2 nanorods. Reproduced (A) and (B) from Zhou R, Guo W, Yu R, Pan C. Highly flexible, conductive and catalytic Pt networks as transparent counter electrodes for wearable dye-sensitized solar cells. J Mater Chem A 2015;3:23028–34; (C) from Adachi M, Murata Y, Takao J, Jiu JT, Sakamoto M, Wang FM. Highly efficient dye-sensitized solar cells with a titania thin-film electrode composed of a network structure of single-crystal-like TiO2 nanowires made by the “oriented attachment” mechanism. J Am Chem Soc 2004;126:14943–9; (D) from Jiu J, Isoda S, Wang F, Adachi M. Dye-sensitized solar cells based on a single-crystalline TiO2 nanorod film. J Phys Chem B 2006;110:2087–92.

TiO2, ZnO is still considered as a distinguished alternative to TiO2 due to its easy crystallization and anisotropic growth. Furthermore, there is huge potential to improve the cell conversion efficiency by tuning ZnO nanostructures. For example, ZnO nanowire–nanoparticle composite DSSCs were developed for higher cell performances by increasing the amount of dye loading and retaining the rapid direct pathways for electron transport in nanowire arrays [86]. The treelike branched ZnO nanowire structures were fabricated using ACG/sop-gel/ACG methods. The formation mechanism of the branched ZnO nanowires is illustrated in Fig. 3(A), and its corresponding Scanning electron microscope (SEM) images in Fig. 3(B)–(D). The branched ZnO nanowire structures with one-dimensional (1D) secondary branches seem directly attached to the main ZnO nanowire backbone. This unique structure could simultaneously afford a direct conduction pathway and achieve the high dye adsorption, increasing the overall conversion efficiency of the DSSC to 1.51% (Fig. 3(E)). The Nyquist plot of the impedance data for bare and branched ZnO nanowire DSSCs was shown in Fig. 3(F). Charge-transfer resistance Rk and electron transport resistance Rw are quite similar for both DSSCs, which indicated the same interfacial recombination and equal crystallinity for either bare ZnO nanowires or branched ZnO nanostructures. On the contrary, the first-order reaction rate constant for the loss of electrons (keff) in the branched ZnO nanowire DSSCs was smaller than the bare nanowire ones, indicating that the electron lifetime could be prolonged by the additional transport distance between branches and conductive electrode [88]. The branched nanostructures were anticipated to be applicable to other semiconductor photoelectrodes in organic–inorganic nanocomposite solar cells [89,90]. Nevertheless, nanostructured ZnO DSSCs still showed relatively low overall conversion efficiencies when compared with TiO2-based systems. The limited performance in ZnO-based DSSCs may be explained by the instability of ZnO in acidic dye. The protons derived from

154

Dye-Sensitized Materials

Fig. 3 (A) The schematic growth procedure from the original ZnO nanowires to the branched ZnO nanowires, (B) before and (C) after recoating a seed layer of the original ZnO nanowires obtained from a solvothermal method, (D) the branched ZnO nanowires after second growth (scale bar ¼1 mm), (E) current density against voltage (J–V) characteristics, and (F) Nyquist plots of the bare ZnO nanowires and the branched ZnO nanowire dye-sensitized solar cells. Reproduced from Cheng HM, Chiu WH, Lee CH, Tsai SY, Hsieh WF. Formation of branched ZnO nanowires from solvothermal method and dye-sensitized solar cells applications. J Phys Chem C 2008;112:16359–64.

Ru complexes such as N3 and N719 make the dye loading solution relatively acidic and dissolve the surface of ZnO, resulting in the formation of excessive Zn2 þ /dye agglomerates [91]. Such aggregates lower the electron-injection kinetics from dye to ZnO. The ZnO aggregates can be modified with TiO2 to form a more efficient system, in which the already-achieved aggregate structure of ZnO can serve as light scattering centers and a TiO2 surface as places for dye adsorption. Many physical- or chemical-deposition methods can be employed for this purpose, such as the recently developed atomic layer deposition (ALD) technique [92], hydrothermal growth [93], emulsion-assisted method [94–97], and electrostatic-spray-deposition [98]. Other metal oxides with more negative conduction band (CB) edge positions versus TiO2 were explored to achieve high opencircuit voltage (Voc). For example, Nb2O5 is a wide band gap semiconductor with band-gap energy of 3.49 eV, which is about 0.29 eV larger than that of anatase TiO2 [99]. High open-circuit potential and good electron injection efficiency were observed for Nb2O5-based cells [100]. DSSCs with various nanostructured Nb2O5 were developed, leading to an impressive efficiencies up to 6% [101,102]. Other nanostructured metal oxides, such as, Al2O3, SnO2, V2O5, ZrO2, CeO2, and Fe2O3, were employed as photoanodes in DSSC [39,103–106]. In those cases, Brunauer–Emmett–Teller surface area and band gap are increased to improve short-circuit photocurrent and open-circuit voltage, respectively.

2.6.3.2

Dyes

Dye molecules (sensitizers) play a critical role in absorbing the incident photons and then generating photoexcited electrons. In other words, the role of a sensitizer in DSSC is as a molecular electron pump. Following light absorption, the excited dye rapidly injects an electron into the semiconductor. Then, the oxidized dye molecule can accept an electron from the redox couple of electrolyte to repeat the cycle. For the process to produce a photocurrent density, the energy of the dye excited state must be higher than the conduction band edge of the semiconductor. A sensitizer should possess certain peculiar characteristics such as (1) strong absorption in the visible range, (2) high stability in the oxidized, ground, and excited states, (3) suitable redox potential, and (4) good efficiency in the charge injection and regeneration processes. Ruthenium-based molecular dyes are very efficient sensitizers for DSSCs. The first high-performance polypyridyl ruthenium complex, which is well-known as N3 dye, is [4,40 -dicarboxylic acid-2,20 -bipyridine ruthenium (II)]. It was reported by Nazzeruddin et al. in 1993 [107]. N3 dye exhibits maximum absorption (extinction coefficient) at 400 and 535 nm of metal-to-ligand charge transfer character, namely, the excitation of the dye involves the transfer of an electron from the metal to the p orbital of the surface anchoring carboxylated bipyridyl ligand [108]. Five years later, N3 performances were surpassed by another ruthenium complex, the black dye [tri(idothiocyanato)-2,20 200 -terpyridyl-4,40 ,400 -tricarboxylate ruthenium (II)], first introduced in 1997. However, N3 returned to the top ranking position by its combination with guanidinium thiocyanate, an additive that increased the cell open-circuit voltage [109]. Currently, ruthenium complexes are the most efficient dyes (Fig. 4), which usually display broad absorption features in the visible region, without the aid of additional energy donors [110]. They are only one type of dyes to achieve over 10% efficiency under standard conditions. Nowadays, more recent research has focused on accomplishing a suitable balance of improved molar absorptivity and stability under thermal and light soaking by extending the p-conjugation of the hydrophobic ligands. C104 dye was reported very recently and presents noteworthy efficiency of 10.5% [111]. However, there are three main drawbacks which limit their applications. First, their preparation normally requires multi-step procedures and time-consuming chromatographic methods [112,113]. Second,

Dye-Sensitized Materials

O

O

OH

TBAO

TBAO

OH

155

O

OH

O

N

N NCS

N

Ru

N

NCS

N

O

N

O

NCS

TBAO Ru

N

NCS

O

Ru

N

NCS

N

NSC

N

OH

NCS

N

O OH

TBAO

N3

O

OH

N749

O

N719

O

TBAO

C4H13 O

C8H17

O

S

C4H13

O

O

S C8H17

NCS

N

NCS

N

Ru

N

N

O

N

NCS

N

N

OH

K19

NCS Ru N

O

NCS

OH

OH O

N

NCS Ru

O

OH 2910

S

N

N N

S

O

OH

C104

O

OH

Fig. 4 Six relevant ruthenium-based dyes used in dye-sensitized solar cells. Reproduced from Tennakone K, Kumara GRRA, Kottegoda IRM, Perera VPS. An efficient dye-sensitized photoelectrochemical solar cell made from oxides of tin and zinc. Chem Commun 1999;15–6. doi:10.1039/ A806801A.

their weak extinction coefficient and feeble absorptivity in the near infrared region limit its application. Other issues in considering Ru-based dyes are their cost as Ru is a rare metal with a high price and its toxic nature. Those have stimulated intensive research for Ru-free dyes. Metal-free organic dyes offer superior molar extinction coefficients, low cost, and a diversity of molecular structures. Recently, novel photosensitizers such as coumarin, merocyanine, cyanine, indoline, hemicyanine, triphenylamin, dialkylaniline, phenothiazine, tetrahydroquinoline, and carbazole-based dyes have achieved solar to electrical PCE up to 5%–9% [114–116]. On the other hand, natural dyes have also been explored for DSSCs due to their large absorption coefficients in visible region, relative abundance, easy preparation, environmental friendliness and low cost [117,118]. However, most of them do not achieve energy conversion efficiencies of over 2% [119]. Metal-complex porphyrin dyes were also investigated. For example, Chlorophyll (Chl) is the pigment responsible for light absorption in photosynthesis [120]. It consists of porphyrin ring structure linked to a hydrocarbon tail. Not surprisingly, Chl-a has been extensively explored for PV applications since 1970s [121]. Numerous porphyrin dyes have been tested as light-harvesting components for DSSCs on account of their intense absorption in the visible region of the solar spectrum and their appropriate redox properties for sensitization of TiO2 films [122]. For example, Zinc (II)-porphyrins (ZnPor) and Zinc (II)-phthalocyanines (ZnPc) are promising candidates as NIR sensitizer. They possess outstanding molecular stability, appropriate frontier orbital energy levels for the efficient electron injection into the conduction band of TiO2 film and the regeneration of oxidized-dyes, and structural versatility for fine-tuning of energy levels. The success of these dyes lies in their tunable physicochemical properties as well as in their high extinction coefficients in the visible part of the solar spectrum [123]. Although various types of dyes have been developed as summarized above, it is practically difficult for a single dye to has highlight-harvesting ability converting from the UV to the NIR region and efficiently inject the photo-generated electron into TiO2. Therefore, the co-sensitization, in which several dyes with different spectral responses are used together, has the theoretical advantage of enhancing photo absorption [124–126]. For example, a cell co-sensitized with black dye and Y1 dye significantly improved both Jsc and Voc, reaching a record efficiency of 11.4% [127]. Other organic dyes such as D131 and NKX-2553 also improved the performance of black dye-based DSSCs to achieve a conversion efficiency of over 11.0% [128]. Furthermore, a cell co-sensitized with two porphyrin dyes (YD2-o-C8 and YDD6) and an organic dye (CD4) showed a high IPCE spectrum in the wavelength region of 400–700 nm and extended the spectrum to the near infrared region [129]. However, some inherent issues, such as energy and electron transfer from one dye to the other, limit the utilization of co-sensitization.

156

2.6.4

Dye-Sensitized Materials

Electrolyte

The electrolyte with redox shuttle is responsible for the inner charge carrier between electrodes in a DSSC. The function of the redox shuttle is to transfer electrons from the counter electrode to the oxidized dyes to complete the electrochemical circuit. There are two kinetic requirements for a successful redox shuttle: it must reduce the dye cation before the dye cation recombines with an electron in the photoanode, but not allow the oxidized form of the shuttle to intercept an electron from the photoanode. The dual criteria of fast dye regeneration and slow interception constitute a very challenging constraint on identifying effective redox shuttles. The iodide/triiodine (I /I3 ) couple is the most common redox shuttle for DSSCs. The fluid I /I3 redox electrolytes can infiltrate deep inside the intertwined semiconductor layers, promoting the mobility of the semiconductor layers and serving as a holetransport materials. The good performance of I3 /I in DSSCs can be attributed to efficient dye regeneration combined with exceedingly slow electron transfer from TiO2 to I3 . However, several disadvantages limit its applications: (1) a large over-potential is needed for efficient dye regeneration, owing to the formation of the intermediate I2 [130]; (2) the I3 and other possible polyiodides formed in the electrolyte (such as I5 , I7 , and I9 ) absorb a considerable part of the visible light, downgrading the efficiencies of DSSCs [131]; (3) the complex redox chemistry in the electrolyte causes great energy loss [131]; and (4) the redox potential of I /I3 limits the photovoltage. The photovoltage is thermodynamically determined by the difference between the electron quasi-Fermi level in the TiO2 film under illumination condition and the redox potential of the redox couple in the electrolyte. Therefore, it is essential to develop a new redox shuttle with lower potential and enough driving force for efficient dye regeneration. Transition metal mediators of copper (I)(II) complexes [132], ferrocene/ferrocenium [133], Ni (III)/Ni (IV) bis(dicarbollide) [134] with good solubility and redox properties have been reported. Ferrocene/ferrocenium improves Voc value up to 0.842V with efficiency of 7.5%. Furthermore, the Co(II)/Co(III) redox potentials of the Co-polypyridyl complexes are more positive than that of I /I3 couple, which can provide a high Voc value of over 1V [135–137]. In 2001, with a new one-electron redox couple of Cocomplex (Co(dbbip)22 þ /3 þ ), DSSCs sensitized by the ruthenium dye (Z316) achieved an overall electrical PCE of 2.2% under AM1.5 irradiations (100 mW cm 2) [138]. However, the slow mass transport in the electrolyte and quick recombination between electrons and Co redox hindered the development of Co-based redox shuttles for DSSCs [135–137]. After 2 years, a series of cobalt complexes used as redox shuttles in electrolytes of DSSCs were reported and DSSCs with those Co complexes achieved the highest efficiency of 4.2% together with a ruthenium-based dye, Z907 [139]. Furthermore, [Co(bpy)3] complex-based redox mediator was reported to achieve a high Voc of 0.92 V and a conversion efficiency of 6.7% by sterically hindered D35 organic sensitizer [140]. To achieve a high efficiency, the following three requirements should be met: First, the blocking hydrophobic groups should be on the sensitizing dyes instead of on the Co complexes, which effectively separates the Co complex from TiO2 surface in order to prevent the charge recombination. Second, thinner TiO2 film less than 10 mm was employed to decrease transport length of Co redox, solving the issues of mass transport problems and recombination losses. Third, use of small Co redox mediator consisting of pristine bipyridine ligand to improve the diffusion in the mesoporous TiO2 electrode, leading to improved photocurrent and photovoltage (Fig. 5). An energy diagram of a DSSC sensitized with D29 and D35 and the chemical structure of the complexes are shown in Fig. 5(A) and (B), respectively. The best efficiencies obtained at 1000 W m 2 AM1.5G illumination are 6.7% for [Co(bpy)3]3 þ /2 þ and 5.5% for iodide/triiodide (Fig. 5(C)). The DSSCs employing cobalt redox mediators are also promising for indoor applications as the efficiency and the open-circuit voltage remain high even at low light intensities. The photovoltage for DSSCs sensitized with D35 employing [Co(bpy)3]3 þ /2 þ at an indoor light illumination of 250 luminous flux (lx) was about 700 mV (Fig. 5(D)). Motivated by this results, the [Co(bpy)3] complex was successfully used in DSSCs sensitized with sterically hindered dyes such as Y123 [141], C213 [142], M19 [143], YD2-0-C8 [144], achieving high Voc values of 0.9–1.0 V and high conversion efficiencies. The other alternative redox couples with more positive potentials than I3 /I , which in principle lead to higher Voc, have been reported. One analogue of I3 /I is Br3 /Br , which showed B500 mV more positive redox potential, was employed as a redox shuttle with a variety of sensitizers [145]. As expected, Voc is somewhat higher with Br3 /Br compared to I3 /I . Furthermore, two pseudohalogen redox couples, (SeCN)2/SeCN and (SCN)2/SCN , with potentials 190 and 430 mV more positive than I3 /I , were also investigated [146]. It was found that the dye regeneration is in the order of I 4SeCN 4SCN . However, no increase in voltage was observed with the pseudohalogen couples. This indicates that selecting a redox couple with a more positive potential does not necessarily translate into an increase of the open-circuit potential, because effective shuttles additionally need to meet the challenging criteria of fast regeneration and slow interception. Recently, there has also been much interest in employing solid organic redox couples (hole conductors) as shuttles in DSSCs, such as 2,20 -7,70 -tetrakis (N,N-di-p-methoxyphenylamine) 9,90 -spriobifluorene (spiro-OMeTAD) [147,148]. Such solid-state electrolytes, however, directly address the challenge of achieving high shuttle concentrations. With the growing ability to tune the semiconductor/dye/electrolyte interface and control recombination reactions, we would efficiently use alternative redox mediators or hole conductors with redox potentials up to 0.5 V more positive than that of iodide/ triiodide. Eventually, DSSCs based on those mediators are expected to surpass I /I3 couple in terms of solar cells efficiency [130].

2.6.5

Counter Electrodes

The counter electrode is the important component of DSSCs. It must efficiently promote the electron transfer from the external circuit back to the electrolyte and catalyzes the reduction of I3 [149,150]. Thus, superior catalytic activity and electrical

Dye-Sensitized Materials

157

Ef E0 ([Co(dmb)3]n+) = 0.43 V

R

E0 ([Co(dtb)3]n+) = 0.43 V E0 0

E

([Co(bpy)3]n+)

R R

= 0.56 V

([Co(phen)3]n+)

N

N

= 0.62 V

R R

Co2+ N

R

(D35) = 1.04 V

N

N N

N

N

E0 (D29) = 0.84 V E0

R

N

R

R

N Co2+ N N

R=H

R R R = H, CH or t-butyl 3

F vs. NHF (A)

(B) 12

20

10

Current density (mA cm−2)

Current density (mA cm−2)

R

8 6 4 [Co(bpy)3]n+

2

15

10

5

lodide/triiodide 0

0

0.2

(C)

0.4 0.6 Voltage (V)

0.8

0

1 (D)

0

0.2

0.4 Voltage (V)

0.6

0.8

Fig. 5 (A) Schematic energy diagram for a nanostructured TiO2 electrode sensitized with D35 or D29 employing [Co(bpy)3]2 þ /3 þ -, [Co (dmb)3]2 þ /3 þ -, [Co(dtb)3]2 þ /3 þ -, and [Co(phen)3]2 þ /3 þ -based electrolytes. Formal potentials indicated are vs. NHE, (B) chemical structure of the different cobalt polypyridyl redox couples, current density vs. applied potential curves under (C) 1000 W m 2 AM1.5G and (D) 250 luminous flux illumination for DSSCs sensitized with D35. Reproduced from Feldt SM, Gibson EA, Gabrielsson E, Sun L, Boschloo G, Hagfeldt A. Design of organic dyes and cobalt polypyridine redox mediators for high-efficiency dye-sensitized solar cells. J Am Chem Soc 2010;132:16714–24.

conductivity is highly desired for CE materials. Pt is a standard catalyst in various fields involving the DSSCs. In conventional DSSCs, Pt-based electrodes electrically catalyze the I3 reduction at the CE/electrolyte interface. To date, a wide range of approaches, such as, thermal decomposition [151], electrodeposition [152], spin coating [153], physical vapor deposition [154], electrochemical deposition [155], and sputtering [156] have been developed to prepare Pt counter electrodes. The performance of Pt electrodes employed by DSSCs can be improved by controlling the structures and morphologies. Pt nanoparticles [157], nanowires [158], nanotubes [159], nanofibers [160], nanoflowers [161], and nanocups [162] have been used in DSSC CEs (Fig. 6). However, Pt is rare and expensive, which limit its application. Therefore Pt-free CEs are necessary for practical DSSCs [163–171]. Carbon materials, transition metal compounds, and conductive polymers, have been explored as alternatives to the Pt CE in DSSCs [172–178].

2.6.5.1

Carbon Materials

The main advantages of carbon materials as CEs in DSSCs are their low cost, large surface area, high electric conductivity, and impressive stability. Early work on carbon CEs for DSSCs was performed to achieve a PCE of 6.67% by Grätzel’s group, including functionalized graphite and high surface area carbon black [172]. So far, various carbon materials have been explored as DSSC CEs. Those are deeply discussed in this section.

2.6.5.1.1

Graphene

Graphene, a single-layer hexagonal lattice of carbon atoms and a novel two-dimensional (2D) materials, is a rising star in the carbon family [179–181]. It possesses a range of unusual properties such as high thermal conductivity, optical transparency, robustness, and stiffness. Individual graphene is also found to have excellent electronic transport properties. Thus, it has been

158

Dye-Sensitized Materials

(A)

(B)

(C)

(E)

(F)

100 nm

(D)

56.3 nm

59.9 nm 47.9 nm 500 nm 63.1 nm

500 nm

Fig. 6 Field emission scanning electron microscope images of Pt: (A) nanoparticles, (B) nanowire, (C) nanotubes, (D) nanofibers, (E) nanoflowers, and (F) nanocups. Reproduced (A) from Song MY, Chaudhari KN, Park J, et al. High efficient pt counter electrode prepared by homogeneous deposition method for dye-sensitized solar cell. Appl Energy 2012;100:132–7, (B) Shakeel Ahmad M, Pandey AK, Abd Rahim N. Advancements in the development of TiO2 photoanodes and its fabrication methods for dye sensitized solar cell (DSSC) applications: a review. Renew Sustain Energy Rev 2017;77:89–108, (C) Wu J, Tang Z, Huang Y, Huang M, Yu H, Lin J. A dye-sensitized solar cell based on platinum nanotube counter electrode with efficiency of 9.05%. J Power Sources 2014;257:84–9, (D) Kim J, Kang J, Jeong U, Kim H, Lee H. Catalytic, conductive, and transparent platinum nanofiber webs for FTO-free dye-sensitized solar cells. ACS Appl Mater Interfaces 2013;5:3176–81. (E) Hsieh T-L, Chen H-W, Kung C-W, Wang C-C, Vittal R, Ho K-C. A highly efficient dye-sensitized solar cell with a platinum nanoflowers counter electrode. J Mater Chem 2012;22:5550–9, and (F) Jeong H, Pak Y, Hwang Y, et al. Enhancing the charge transfer of the counter electrode in dye-sensitized solar cells using periodically aligned platinum nanocups. Small 2012;8:3757–61.

extensively investigated theoretically and experimentally [19,182–184]. Furthermore, graphene has been widely investigated as counter electrodes for DSSCs. The performance of graphene counter electrodes is strongly dependent on the structure and properties of graphene materials [185]. Graphene films for DSSC counter electrodes have been prepared by various techniques, including thermal exfoliation from graphite oxide [186], chemical reduction of graphene oxide [187], and electrophoretic deposition [185]. The components of graphene sheet could modify its catalytic and electrical properties. Functionalizing graphene with oxygen containing groups could create more catalytic centers and thus enhance catalytic activity [188]. Furthermore, defects of graphene can also affect its performance for DSSCs [189]. Those indicates that tuning the number of oxygen functional groups and the C/O ratio could allow ones to improve the redox kinetics and thus achieve optimal catalytic activity in DSSCs. However, the oxygen doping can also lead to the destruction of the conjugated carbon structure, thus reducing the conductivity of the graphene dramatically. Doping graphene with heteroatoms was proved to be an effective strategy to overcome this dilemma. Among all heteroatoms doped in graphene, nitrogen is the most promising since it creates more catalytically active sites with a minimized change of the conjugation length [190,191]. Meanwhile, the lone electron pairs of nitrogen atoms can form a delocalized p-system with the sp2-hybridized carbon framework, facilitating the electron transfer ability of the material [192,193]. In addition, co-doping graphene by N and other elements could further improve the electrochemical properties due to the synergistic effect of heteroatoms [194–196]. S atom is also a promising dopant to adjust the electrocatalytic activity for carbon materials, since the S atoms could increase the charge density asymmetry of carbon atoms, create more catalytically active sites, and decrease the charge-transfer resistance [197–199]. Thus, intercalating S and N heteroatoms into graphene networks is a promising pathway to modulate the electrical, chemical, and electrocatalytic properties [154,192,193,196,200–203]. However, conventional graphene with 2D structure is not very efficient as CEs for DSSCs, because the surface area of 2D graphene is not fully accessible for the electrolyte due to its re-aggregation that inhibits electrolyte ions from coming into contact with its internal surface. For this reason, 3D graphene with meso/macro pores and channels, which is considered as the second generation graphene, is attracting much attention. Moreover, the three-dimensional (3D) structure could also generate defects on graphene sheets, providing more active sites. Very recently, Hu’s group discovered numerous chemical reactions and invented various novel approach to synthesize 3D graphene with controlled shapes, including 3D honeycomb-like structure, 3D cauliflower-fungus like structure, and 3D flower-like structure graphenes (Fig. 7) [204–209]. Those 3D graphene materials were explored as CEs for DSSCs, leading to high power conversion efficiencies up to 10%. Furthermore, the synthesis of 3D graphene

Dye-Sensitized Materials

159

Fig. 7 Field emission scanning electron microscope images of (A) honeycomb-like structured graphene, (B) graphene nanosheet, (C) cauliflowerfungus like structured graphene, and (D) 3D flower-like structured graphene. Reproduced (A) from Wang H, Sun K, Tao F, Stacchiola DJ, Hu YH. 3D honeycomb-like structured graphene and its high efficiency as a counter-electrode catalyst for dye-sensitized solarcells. Angew Chem Int Ed 2013;52:9210–4; (B) Wei W, Hu YH. Synthesis of carbon nanomaterials for dye-sensitized solar cells. Int J Energy Res 2015;39:842–50; (C) Wei W, Sun K, Hu YH. Synthesis of 3d cauliflower-fungus-like graphene from CO2 as a highly efficient counter electrode material for dye-sensitized solar cells. J Mater Chem A 2014;2:16842–6; and (D) Wei W, Sun K, Hu YH. Direct conversion of CO2 to 3D graphene and its excellent performance for dye-sensitized solar cells with 10% efficiency. J Mater Chem A 2016;4:12054–7.

directly from CO2 provides not only a promising electrode material of energy conversion and storage devices, but also offers a novel utilization of a greenhouse gas [210].

2.6.5.1.2

Carbon nanotube

Carbon nanotubes (CNTs) are 1D nanomaterials that possess unique structures and properties. CNTs can be classified into two main categories: single-walled Carbon nanotubes (SWCNTs) and multi-walled Carbon nanotubes (MWCNTs). A SWCNT consists of a single rolled-up graphene sheet, whereas a MWCNT consists of several coaxially arranged graphene sheets. The first successful application of CNTs as a counter electrode material for DSSCs was reported in 2003 by Suzuki et al., who deposited SWCNTs onto a Teflon membrane filter substrate for DSSC fabrication [173]. DSSCs with SWCNT CE exhibited a promising result (PCE of 4.5%), which was even comparable to the Pt sputtered cell (PCE of 5.4%). This early work proved that CNT materials naturally possess excellent electrocatalytic activity and high conductivity, and have the potential to perform a catalytic role in DSSCs. The performance of CNT CEs are strongly dependent on CE fabrication process. For example, CNT CEs prepared using a screen printing technique resulted in a PCE of 8.03%, while CNTs grown directly on a FTO substrate as CEs using a chemical vapor deposition (CVD) technique yielded a PCE of 10.04% [211]. That is because the thermal CVD approach could create excellent contact between the electrolyte and the FTO glass and thus significantly improve electrical conductivity (Fig. 8). CNTs have also been explored for fiber-shaped DSSCs. Thin fiber-shaped cells have advantages such as lightweight and device flexibility and can be integrated into textiles in various forms compared with conventional planar solar cells. The basic structure of a PV fiber consists of two parallel wires twisted together, in which one core wire coated by active layers serves as the primary electrode and a second wire (e.g., Pt) as the counter electrode. Zhang reported a highly flexible DSSCs based on a single wire, which was fabricated by wrapping a CNT film around Ti wire-supported TiO2 tube arrays as the transparent electrode. As shown in Fig. 9(A) and (B), the DSSC wire

160

Dye-Sensitized Materials

Fig. 8 Cross-sectional scanning electron microscope images of (A) screen printing counter electrode carbon nanotubes (CNT) paste and (B) directly formed CNT. Reproduced from Nam JG, Park YJ, Kim BS, Lee JS. Enhancement of the efficiency of dye-sensitized solar cell by utilizing carbon nanotube counter electrode. Scr Mater 2010;62:148–50.

lm

T fi

CN

Cathode

Ti

rray ea

e wir iO e/T Dy

2

tub

CNT network

Anode Ti CNT

TiO2 250 µm

(A)

TiO2 tubes with dye

(B)

e−

e− (D) 3I− Electrolyte

hv

TiO2 layer

e− e−

e− I3−

Dye molecules

100 µm

CNT fiber (C)

(E)

(F)

Fig. 9 (A) Scanning electron microscope (SEM) image of the wire with three parts, including an inner Ti core, electrochemically anodized TiO2 nanotube arrays, and an external carbon nanotubes (CNT) film electrode, (B) illustration of a coaxial single-wire structure dye-sensitized solar cell (DSSC), consisting of a core Ti wire, dye-grafted TiO2 nanotube arrays, and a flexible transparent CNT film wrapping around the wire, (C) SEM images of a twined structure between a CNT/TiO2/N719 composite fiber and a pure CNT fiber, (D) schematic illustration of a wire-shaped DSSC fabricated with two CNT fibers twined into a cell, (E) top view of a cell, and (F) working mechanism. Reproduced (A) and (B) from Zhang S, Ji CY, Bian ZQ, et al. Single-wire dye-sensitized solar cells wrapped by carbon nanotube film electrodes. Nano Lett 2011;11:3383–7 and (C)–(F) from Chen T, Qiu LB, Cai ZB, et al. Intertwined aligned carbon nanotube fiber based dye-sensitized solar cells. Nano Lett 2012;12:2568–72.

Dye-Sensitized Materials

161

has three parts with different functions, including an exposed Ti core wire, a small distance bare TiO2 tube array area, and a CNTwrapped TiO2 tube array active device region. The CNT film ensures full contact with the underlying active layer, as well as uniform illumination along circumference through the entire DSSC. The single-wire DSSC showed a PCE of 1.6% under standard illumination, which was further improved to more than 2.6% assisted by a second conventional metal wire [29]. In another fibershaped DSSC, the work electrode was fabricated with the TiO2 coated CNTs, while a bare CNT fiber served as the CE [26]. The CNT–CNT cell gained 2.60% under standard illumination (100 mW cm 2), which was increased to 2.94% by placing a mirror under the cell (Fig. 9(C)–(F)). For a control study, a conventional Pt wire-based cell was tested under the same conditions and reached a PCE of only 0.16%. Such a low efficiency was caused by the lower flexibility of Pt wire, with the gaps between the work electrode and Pt wire inhibiting the interaction. MWNT counter electrodes were also explored for DSSCs [212,213]. Song et al. reported the bamboo-like-structured multiwall CNT counter electrodes for I3 reduction in DSSCs. Defect-rich edge planes of the bamboo-like-structured CNTs facilitate the electron transfer kinetics at the counter electrode-electrolyte interface, resulting in low charge-transfer resistance and an improved ff. As a result, the device showed a PCE of 7.7% comparable to that of conventional Pt counter electrode DSSCs [212]. Zhang et al. fabricated three types of DSSCs based on SWCNTs, double-walled carbon nanotubes (DWCNTs), and MWCNTs and compared their PV performances under the same condition. The sequence of PV efficiencies was DWCNTs (8.03%)4SWCNTs (7.61%)4MWCNTs (7.06%). The DSSC performance sequence is consistent with the corresponding charge-transfer resistance (Rct) values of the films. This low Rct value of the DWCNTs film, which leads to higher energy conversion efficiency compared to other CNT counter electrode based DSSCs, was attributed to the large inner surface area of activated DWCNTs [214]. The aggregation of CNTs would significantly hamper their charge-transport property, resulting in a lower efficiency than expected [20]. To solve the issue, 1D CNTs were combined with 2D graphene to form composite materials as counter electrodes for DSSCs [215,216]. Choi et al. prepared a MWCNT-graphene composite counter electrode for DSSCs, in which the graphene layers were firstly deposited on a SiO2/Si substrate by drop casting method. Then CVD technique was employed to vertically align MWCNTs on graphene layers. The device fabricated with such a MWCNT-graphene film showed a PCE of 3.0%. This performance was further improved by other researchers [217].

2.6.5.2

Transition Metal Compounds

Various transition metal compounds (TMCs) are explored as counter electrodes for DSSCs, including sulfide, oxides, nitrides, and carbides.

2.6.5.2.1

Metal sulfides and oxides

Sulfides are one class of alternative CE materials that were tested. Normally, ionic compounds are good for adsorption of ions, such as I3 . However, they possess low electrical conductivity, which is unfavorable as counter electrodes for DSSCs. Molybdenum disulfide is an ionic semiconductor with a low conductivity, which shows poor energy conversion efficiency for DSSCs. However, in contrast to 2H phase MoS2 that is a semiconductor with a honeycomb lattice (Fig. 10(A)), 1T phase MoS2, which possesses a hexagonal structure (Fig. 10(B)), is a metallic phase with excellent conductivity. It was demonstrated that the DSSC with the 1T metallic MoS2 CE exhibited an excellent PCE of 7.08%, which is three times larger than that (1.72%) of the DSSC with 2H phase MoS2 (Fig. 10(C)). This happens because the 1T MoS2 sheets are 107 times more conductive than 2H bulk MoS2 [218]. Other metal sulfides were also explored as CEs for DSSCs [219,220]. For example, DSSCs with CEs of CoS, Co8.4S8, Co9S8, Ni3S2, FeS2, MoS2, and WS2 yielded promising PV performances with PCEs of 6.5%, 6.50%, 7.0%, 7.01%, 7.31%, 7.59%, and 7.73%, respectively [221]. Metal sulfides could also be deposited on highly conductive graphene films in FTO/TCO-free DSSCs [222]. As shown in Fig. 10(D)–(I), graphene films were directly grown on an SiO2 substrates by CVD technique, and then decorated with metal sulfide (NiS and CoS) nanoparticles by dip coating Co(C3H5OS2)2 and Ni(C3H4OS2)2 precursors, and then annealed at 4001C for 30 min at Ar gas. As-prepared NiS/graphene and CoS/graphene hybrid CEs in DSSCs showed excellent electrocatalytic activity and higher PCE of 5.25% and 5.04%, respectively. They are comparable to that (5.00%) of Pt electrode with the FTO substrate. Metal oxides also exhibited excellent performance for DSSCs, such as WO2 nanorod [223], NbO2 [224], and sulfur-doped NiO thin film [225]. It was found that the low ratio of oxygen to metal atom of NbO2, WO2, and TaO can improve the catalytic activity of CE for I3 reduction [223,224,226]. However, some oxides (TiO2, SnO2, Nb2O5 and WO3) are not promising CEs for DSSCs, because their energy levels or band gaps may be more preferable for their use as photoanodes, not CEs in DSSCs [169].

2.6.5.2.2

Transition metal carbides, nitrides, and phosphides

Transition metal carbides are important electrode materials due to their interesting physicochemical and catalytic properties. WC is a promising electrocatalyst with low cost, excellent catalytic performance, and good thermal stability under extreme conditions [227,228]. Other carbides, such as MoC, NbC, TiC, VC, and Ta4C3, are also attractive catalytic materials. Among them, Ta4C3 showed the highest electrocatalytic activity for DSSCs, leading to a high PCE of 7.4% [167,229]. Moreover, N-doped carbides, such as TiC(N), VC(N), and NbC(N), where C is substituted by N in the unit cell, showed better performance than the corresponding TiC, VC, and NbC [167]. Various transition metal nitrides showed their feasibility as Pt-free counter electrodes for DSSCs. For example, TiN nanotubes arrays, which were prepared by the anodization of a metallic Ti foil substrate and the subsequent simple nitration in an ammonia

Dye-Sensitized Materials

162

20 18 16 14 12 10 8 6 4 2 0 0.0 (C)

Current density (mA cm−2)

2H-type MoS2 1T-type MoS2

(A)

0.2

0.4 0.6 Voltage (V)

0.8

(B) S O

KS

(i)

M:NI or Co CH3

S O

KS

CH3

DSSC

M Cl (D)

(F)

Cl (E)

SiO2

CVD

Dip coating

Growth

Annealing

MS/graphene

3D graphene (G)

(H)

Fig. 10 High-angle annular dark-field scanning transmission electron microscopy image of (A) 2H phase MoS2 and (B) 1T phase MoS2, (C) J–V curves of dye-sensitized solar cells (DSSCs) with MoS2 counter electrodes. Overall procedure for the fabrication of Mass spectrometry (MS)/ graphene composite charge extractions (CEs) for DSSCs: (D) reaction equation of MS precursor preparation, (E) well dispersed MS precursor solutions in acetone and ethanol solvents, (F) dielectric SiO2 substrate, (G) graphene film directly grown on SiO2 substrate, (H) MS NPs loaded on the graphene surface, and (I) application of MS/graphene composite CEs in DSSCs. Reproduced (A) and (B) from Wei W, Sun K, Hu YH. An efficient counter electrode material for dyesensitized solar cells-flower-structured 1T metallic phase MoS2. J Mater Chem A 2016;4:12398–401 and (C)–(I) Bi H, Zhao W, Sun S, et al. Graphene films decorated with metal sulfide nanoparticles for use as counter electrodes of dye-sensitized solar cells. Carbon 2013;61:116–23

atmosphere, were evaluated as a DSSC CE [152]. Furthermore, Pt-like electrocatalytic activities of MoN, WN, and Fe2N were revealed. From Fig. 11(A)–(C), one can see oval-shaped grains (with a size of about 100 nm) for Fe2N, sphere-like aggregates (with a diameter of about 200 nm) for MoN, and irregular grains (with an average size of 20–30 nm) for WN. Among these transition metal nitrides, MoN exhibits superior electrocatalytic activity and a high PV performance [230]. The PCEs of DSSCs with CEs of W2N, Mo2N, and Ta3N5 could reach 5.81%, 6.38%, and 5.03%, respectively [229]. MoP, Ni5P4, and carbon supported Ni5P4 were also evaluated as cathode catalysts [231]. DSSCs incorporated with MoP and Ni5P4 cathodes yielded a PCE of 4.92% and 5.71%, and the DSSC using Ni5P4/C composite cathode showed a high PCE of 7.54% [231]. Hybrid CEs can offer a synergistic effect on catalytic performance improvements due to a combination of different materials into one material, resulting in more superior performance than individual component CEs. DSSCs efficiencies for transition metal/carbon hybrid are selectively listed in Table 1.

2.6.5.3

Conductive Polymers

To highlight the merits of transparency and flexibility for DSSCs, it is essential to develop transparent and flexible CEs. Recently, extensive studies have been performed using conducting polymers as alternative cathodes in DSSCs due to their excellent catalytic activities, low cost, high transparency, good conductivity, easy preparation, and good environmental stability.

Dye-Sensitized Materials

(A)

(C)

163

(B)

(D)

Current density (mA cm−2)

15 MoN WN Fe2N Pt

10

5

0 0.0

0.2

0.4 0.6 Volatage (V)

0.8

Fig. 11 Transmission electron microscopy images of (A) Fe2N, (B) MoN, and (C) WN samples; (D) characteristic J–V curves of dye-sensitized solar cells using different metal nitrides and Pt counter electrodes, measured under simulated sunlight at 100 mW cm2 (AM1.5). Reproduced from Li GR, Song J, Pan GL, Gao XP. Highly Pt-like electrocatalytic activity of transition metal nitrides for dye-sensitized solar cells. Energy Environ Sci 2011;4:1680–3.

Porous poly(3,4-propylenedioxythiophene)(PProDOT) [141], polypyrrole(PPy) [255], polyaniline (PANI) [256], polythiophene (PTh) [257], and Poly(3,4-ethylenedioxythiophene) (PEDOT) [258] have proven themselves capable of competing with Pt as electrocatalysts in DSSCs.

2.6.5.3.1

Porous poly(3,4-propylenedioxythiophene)

PProDOT can replace Pt electrodes in DSSCs and give better results due to its ultra-high surface area, which helps in forming favorable interactions at the electrode/electrolyte interface. Ho et al. have employed an electropolymerization method to deposit a porous poly(3,3-diethyl-3,4-dihydro-2H-thieno-[3,4-b][1,4]-dioxepine) (PProDOT-Et2) film as the CE material onto FTO glass and reported a PCE of 5.20% for the pertinent cell [259]. PProDOT-Et2 provides an extremely high electrochemical surface area for I3 reduction because of its unique 3D matrix. However, its electrocatalytic activity for I3 reduction is still lower than that of Pt despite its large electrochemical surface area. Furthermore, composite PProDOT-Et2 and Pt (PProDOT-Et2/Pt) films were also used as CEs for DSSCs and were compared with bare polymer (PProDOT-Et2) and bare sputtered-Pt (s-Pt) [260]. To control the microstructure and electrochemical properties of the PProDOT, hydrophobic ionic liquids were employed as a growing medium to prepare PProDOT films with electrical-field-assisted growth technique. The performances of the PProDOT CEs strongly depend on used ionic liquids. I-mediated DSSCs using PProDOT yielded PCE values ranging from 9.12% to 9.25%. Furthermore, when Y123 was used as a dye in Co-mediated DSSCs with PProDOT, the device provided a Voc close 1 V and a high PCE of 10.08% [141]. In addition, different Co complexes were synthesized with various polypyridyl ligands, by which the

164

Dye-Sensitized Materials

Table 1

Photovoltaic parameters of dye-sensitized solar cells with TMC/Carbon composites CEs

CEs

PCE (%)

Voc (mV)

Jsc (mA cm 2)

ff

References

TiO2/C TiN/CNTs WO2/MC WC/MC CoS/MWCNTs TiN/NG VC/MC TiN/MC TiN/carbon black Ni12P5/graphene MoO2/CNTs MoS2/MWCNTs MoS2/graphene WS2/MWCNT TaON/graphene NiS2/RGO Ta3N5/graphene NiS/graphene MoS2/C Fe3C/GC MoS2/graphene MoS2/graphene NiS/MWCNT CoS/MWCNT WS2/MWCNT TaC/MC

5.50 5.41 7.764 8.18 6.96 5.78 7.63 8.41 7.92 5.70 4.34 6.45 5.98 7.36 7.65 8.55 7.85 5.25 7.69 6.04 6.07 5.81 7.9 8.05 6.41 7.93

700 750 808 804 720 728 808 820 791 727 75 730 710 750 829 749 837 724 750 700 750 773 753 751 730 800

12.53 12.74 13.55 14.59 15.96 12.34 13.11 15.30 14.29 12.86 13.67 13.69 12.41 13.63 13.38 16.55 13.53 10.31 15.07 13.77 13.27 12.79 14.18 14.69 13.51 14.51

0.57 0.57 0.71 0.70 0.64 0.64 0.72 0.67 0.7 0.61 0.44 0.65 0.68 0.72 0.69 0.69 0.693 0.70 0.68 0.63 0.61 0.59 0.74 0.73 0.65 0.68

[232] [233] [234] [235] [236] [237] [167] [238] [239] [240] [241] [242] [168] [243] [244] [245] [246] [247] [248] [247] [249] [250] [251] [252] [253] [254]

Abbreviations: CE, counter electrode; CNT, carbon nanotube; PCE, power conversion efficiency; MWCNT, multi-walled carbon nanotube; ff, fill factor.

oxidation potential was tuned from 0.17 to 0.34 V [261]. Those Co complexes act as redox mediators in DSSCs together with polymer-based CEs, resulting in a PCE of 8%–10% and a Voc of B1V at 0.1, 0.5, and 1 Sun. The utilization of PProDOT CEs could make a 20% improvement in PCE (9.9%) and a remarkable drop in Rct (2.5 O cm2) as compared to Pt electrode (8.24%, 50 O cm2). These improvements were attributed to the ultra-high surface area of PProDOT films, which can eliminate passivation of the CE/electrolyte interface.

2.6.5.3.2

Polypyrrole

PPy is one of the most intensively studied conducting polymers as CEs for DSSCs due to its simple and inexpensive preparation, good catalytic activity for I3 reduction, high conductivity, and stability in the presence of liquid electrolytes [255,260,262]. Furthermore, it can be used as flexible CEs [263]. However, electrodes prepared from dispersions of PPy by ex situ coating methods, such as doctor-blade, spin coating or printing techniques, generally reach lower conductivities and have poor adhesion to FTO substrates compared to PPy CEs prepared by electrodeposition or in situ polymerization methods [264]. To reduce the charge-transfer resistance and achieve good electrocatalytic activity, PPy-NP CE with spongy and the rubbery texture can be synthesized by synthetic chemical method [255]. The spongy and rubbery texture provides porosity, which enhances the diffusion and hence electrocatalytic activity. The performance of PPy CE can also be enhanced by the combination of PPy with carbon-based nanomaterials, such as, MWCNTs, which have shown high catalytic activity, high surface area, good stability, and rapid electron transport [265–267]. Since 2002, a variety of methods for producing composites on PPy and CNTs have been reported, of which chemical and electrochemical polymerizations have been used the most. Because CNTs are electron acceptors and PPy is an electron donor, CNT/PPy particles can create charge-transfer complexes. Surface interaction between PPy and CNTs can be improved by surface active substances (such as ionic surfactants), functionalization of the CNT surface with proper chemical groups, or direct covalent binding of PPy onto the CNT surface via intermediate chemical linkages. Most PPy and PPy/MWCNT CEs were prepared by in situ polymerization, electrodeposition, or electropolymerization [268]. For example, Yue et al. [269] prepared a PPy/MWCNT CE by electrochemical polymerization onto FTO glasses with a high DSSC conversion efficiency of 7.4% compared to Z ¼ 6.8% of a DSSC with a standard Pt CE. SWCNT were also used to prepare a composite CE with PPy by in situ polymerization, leading to the enhancement of charge transfer and a high DSSC PCE (Z ¼ 8.3%) compared to a PPy CE (Z ¼ 6.3%) [267]. Only in few research studies, PPy/MWCNT CEs were prepared by ex situ coating techniques. For example, Peng et al. deposited PPy/MWCNT onto glass and polyethylene naphthalene (PEN) substrates by drop casting method. The DSSC with the PPy/MWCNT CE exhibited promising efficiencies: Z ¼ 7% on FTO glass and Z ¼ 4% on PEN foils [263].

Dye-Sensitized Materials 2.6.5.3.3

165

Polyaniline

PANI, which possesses high conductivity, good environmental stability, and interesting redox properties, can be easily synthesized [270]. Li et al. explored PANI as DSSC CEs. They prepared microporous PANI films, which possess lower Rct values and higher electrocatalytic activity for I3 reduction than Pt CE due to its high surface area and microporous structure [256]. The use of PANI CEs in DSSCs yielded a PCE of 7.15%, which is higher than that of Pt CE. To obtain the morphology-controllable CE, an oriented PANI nanowire arrays was grown on FTO substrates using in situ synthesis [271]. The PANI nanowire arrays exhibited much better catalytic properties and chemical stability for the Co2 þ /Co3 þ redox couple, yielding a PCE of 8.24% in DSSCs with a FNE29 dye. The cell performance with Pt CE is better than that with the CE of drop-cast PANI films, but inferior to that of the PANI nanowire arrays. The formation of the PANI nanowire arrays should be responsible for the improvement of Jsc and ff, which are mainly affected by Rct. PSS, TsO, SO42 , ClO4 , Cl and BF4 are often used as the dopants in the design of CEs [178,272–274]. Different doping ions affect the morphologies, electrochemical properties, and doping/dedoping process of polymer films. Among them, PANI-SO4 films have a highly porous morphology, yielding a PCE of 5.6% as a CE in DSSCs, a higher reduction current for I3 , and a lower Rct relative to Pt CEs. Nevertheless, PANI is not an ideal CE material due to its instability, self-oxidation, and carcinogenic properties.

2.6.5.3.4

Polythiophene

Polythiophene (PTh) and its derivatives were used to prepare composite counter electrodes [268,275]. Wang et al. fabricated MWCNT/polythiophene composite film counter electrode by electrophoresis and cyclic voltammetry in sequence. The overall energy conversion efficiency of the DSSC employing the MWCNT/PTh composite film reaches 4.72%, which is close to that of the DSSC with a Pt CE. Compared with the PCE (2.68%) of a standard DSSC with MWCNT CE, the energy conversion efficiency has been increased by 76.12% for the DSSC with MWCNT/PTh CE. These results indicate that the composite film with high conductivity, high active surface area, and good catalytic properties for I3 reduction can be used as the counter electrode for high efficient DSSCs [276]. Ho et al. applied a novel thiophene-based water-soluble conducting polymer, sulfonated-poly(thiophene-3[2-(2-methoxyethoy)ethoxy]-2,5-diyl)(s-PT), for DSSCs. By adding the nano-porous carbon black nanoparticles (CB-NPs) into the s-PT polymer matrix, the composite film can provide a nano-porous morphology and a good adhesion to the flexible titanium foil. Therefore, the nano-porous CB-NPs/s-PT composite film provides large surface area, fast electrolyte penetration, and rapid reaction rate for I3 reduction, leading to a high efficiency of 9.02% [277].

2.6.5.3.5

Poly(3,4-ethylenedioxythiophene)

PEDOT was found to be an efficient catalyst for DSSCs. Saito et al. fabricated PEDOT CE by chemically polymerized PEDOT on a conductive glass [247]. The PEDOT:PSS/carbon-based DSSCs reached PCE of 4.11% under irradiation of a simulated solar light with an intensity of 100 mW cm 2, which is higher 20% than that of the DSSCs with Pt CE [247]. Lee et al. compared PEDOT with two other conductive polymers: poly(3,3-diethyl-3,4-dihydro-2H-thieno-[3,4-b] [1,4] dioxepine) (PproDOT-Et2) and poly(3,4propylenedioxythiophene) (ProDOT) for DSSCs. The cells with a PProDOT-Et2 CE showed a higher PCE of 7.88% compared to the cells fabricated with PEDOT (3.93%), PProDOT (7.08%), and sputtered-Pt (7.77%) counter electrodes. This enhancement was attributed to increases in the effective surface area and good catalytic properties for I3 reduction [272]. To improve catalytic activity, PEDOT were mixed with inorganic compounds as composite electrodes, such as nanostructured TiN [278,279], ZrN [280], CoS [281], ZnO [282], NiO [283], and MgO [284]. For example, Hu’s group revealed that the interaction energy of I3 and ZnO was 30 kcal mol 1, indicating that ZnO can strongly adsorb I3 . In contrast, I3 has a weaker interaction with PEDOT (8.4 kcal mol 1). Therefore, ZnO would have a higher catalytic activity than PEDOT for I3 reduction. As a result, the DSSC with a ZnO/PEDOT:PSS composite counter electrode exhibited excellent PV performance with a maximum PCE of 8.17% [282]. They also demonstrated that the interaction energy is 70.7 kcal mol 1 for I -ZrN and 15.0 kcal mol 1 for I PEDOT. This confirms that ZrN has a very strong interaction with I , leading to very difficult desorption of I from ZrN and thus deactivating its catalytic ability (Fig. 12). Therefore, the promoting effect of ZrN on PEDOT would be due to its electrical conductivity which lead to a remarkable increase in PCE from 3.46 to 6.68% [280].

2.6.6 2.6.6.1

Commercialization of Dye-Sensitized Solar Cells Photovoltaic Market Overview

The PV industry generated $82 billion in global revenues, while the worldwide solar cell production reached 20.5 GW in 2010. The top ten major manufacturers accounted for 45% of total production. Most of them use crystalline silicon, and three companies (JA Solar, First Solar, and Trina Solar) employ Thin Film technologies. It is expected that the solar industry’s production will grow at a compound annual growth rate (CAGR) of 20%B30% for the next 5 years. On the other hand, the other portable electronics, especially for indoor applications and with a multitude of different applications, are expected to be developed in the next few years. They are predicted to reach $5 million in market share by 2017, which is expected to grow sixfold by 2023. Technological innovations have resulted in good price to performance ratio of these solar cells and are anticipated to impact positively industry growth. As an efficient and environmentally benign alternative to Si-based PV cells, DSSCs have attracted increasing interest during

Dye-Sensitized Materials

166

H

C

O

S

(A)

H

Zn

(B)

H

C

N

O

Zr

(C)

S

C

O

Zn

S

I

(D)

I

H O C N S Zr

2.2946 Å

(E)

(F)

2.2220 Å

(G)

Fig. 12 Models of (A) ZnO and (B) Poly(3,4-ethylenedioxythiophene) (PEDOT) for discrete Fourier transform (DFT) calculations. DFT predicted structures for adsorption of I3 on (C) ZnO and (D) PEDOT. DFT predicted structures for adsorption of I on (E) ZrN and (F) PEDOT, (G) DFT predicted structure for the ZrN/PEDOT complex. Reproduced (A)–(D) from Wang H, Wei W, Hu YH. Efficient ZnO-based counter electrodes for dyesensitized solar cells. J Mater Chem A 2013;1:6622–8 and (E)–(G) from Wei W, Wang H, Hu YH. Unusual particle-size-induced promoter-to-poison transition of ZrN in counter electrodes for dye-sensitized solar cells. J Mater Chem A 2013;1:14350–7.

the past years [285]. A new record of photoelectric conversion efficiency of DSSCs up to 11.14% could compete well with a-Si solar cell. Several institutes and companies have started to upscale DSSCs for commercial production. STI in Australia finished its showcase with 200 m2 in 2003 [286], and European group obtained 8.18% in efficiency and up to 8300 h stability under “2.5 Sun” equivalent intensity indoor test (equivalent to10 years outdoor running). The DSSC technology is anticipated to address the three basic energy needs of future including economy growth, environment protection, and energy security. Advantages associated with DSSC, including cost effectiveness, ability to provide electricity at low light conditions, and high-performance ratio, are anticipated to drive demand over the forecast period. DSSCs will take market share from Crystalline Silicon and Thin Film technologies. Global DSSC market value was estimated to be USD 49.6 million in 2014 and it will grow at a CAGR of over 12% from 2015 to 2022. Although initial products are aimed toward indoor and portable applications, DSSCs are expected to be incorporated into much bigger installations. Forecasted growth of DSSCs in different market segments were illustrated in Fig. 13. There are numerous companies, which sell DSSCs, such as 3G solar, CSIRO, Dyesol, Fujikura, G24i Power, Nissha Printing, Oxfor Photovoltaics, S Samsung SDI, SHARP, Solaronix, SolarPrint, SONY Technology Centre, and TiSol. As a brief summary, Table 2 lists industrial sectors that have emerged to develop the PV technology, their inferred core business and key achievements.

2.6.6.2 2.6.6.2.1

Applications Portable charging

Portable charging was the largest application segment and accounted for 33.0% of revenue share in 2014 and is expected to grow at a CAGR of 12.6% from 2015 to 2022. To fulfill the requirement for such applications, some evolutions of the structure of solar cells are mandatory: to fully exploit their capabilities, devices have to be flexible and adaptable to complex shapes. The availability of lightweight flexible dye sensitized cells or modules are attractive for applications in room or outdoor light powered calculators, gadgets, and mobiles. Fig. 14 shows several applications of DSSCs as portable charger. For example, Fig. 14(B) shows the wireless

Dye-Sensitized Materials

167

30.00 25.00 20.00 15.00 10.00 5.00 2012

2013

2014

2015

2016

2017

2018

2019

2020

2021 2022

Portable charging

BIPV/BAPV

Embedded electronics

Outdoor advertising

Automotive (AIPV)

Others

Fig. 13 Forecasted growth of dye-sensitized solar cells in different market segments. AIPV, automotive integrated photovoltaic; BAPV, building applied photovoltaic; BIPV, building integrated photovoltaic. Reproduced from Shakeel Ahmad M, Pandey AK, Abd Rahim N. Advancements in the development of TiO2 photoanodes and its fabrication methods for dye sensitized solar cell (DSSC) applications: a review. Renew Sustain Energy Rev 2017;77:89–108.

sensors based on printable solar cell technology developed by Analog Devices Incorporated and Gas Sensing Solutions Limited and Fig. 14(C) shows a self-powered electronic shelf lable (ESL) module using DSSC developed by MKE Technology Co., Ltd. Fig. 14 shows the images of few commercial DSSC products on the market are indoor electronics by G24 ((A)–(C)), while (D) and (E) are developed by 3G Solar.

2.6.6.2.2

Flexible dye-sensitized solar cells and wearable electronics

The Australian company Dyesol has pioneered the commercialization of DSSCs after obtaining a license from the inventors and has developed the technology in practically every aspect. The company recently introduced a flexible, foldable, light weight and camouflaged solar panel for military applications which has been found to be superior to other PV technologies in maintaining voltage under a very wide range of light conditions, even in the dappled light under trees. The first commercial shipment of low light, ultra thin DSSC technology, created by G24 Innovations, has been sent to Hong Kong-based consumer electronics bag manufacturer, Mascotte Industrial Associates for use in backpacks and bags. Ideal for clothing and portable applications, DSSCs are less than 1 mm thick, inexpensive, without silicon or cadmium and usable indoors, making them ideal for powering cell telephones, cameras, and portable electronics. The DSSCs can also be embedded into tent material to power LED lighting systems for camping. The solar panels built into these backpacks and bags will harvest energy to repower mobile electronic devices. The cells are light and bendable, making them ideal for canvas bags. According to the bag manufacturing company, Mascotte, under sunny conditions, it takes about 4–5 h to fully charge a cell phone using a 15 cm  20 cm solar panel, and indoors, takes as many as 12 h, depending on the lighting. On a larger scale, G24i’s flexible DSSC modules can be integrated into electronic advertising displays that can even work indoors. As well as embedding the film into fabrics, G24i’s advanced solar cells can also be layered onto laptops, mobile telephones, GPS units, and AV devices for supplemental power, significantly extending their up-time. Wearable electronics fabricated on lightweight and flexible substrate are widely believed to have great potential for portable devices [287]. Fu et al. designed a new type of integrated power fiber by incorporating a DSSC and a solar cell for harvesting solar energy and storage to realize a self-powered system for driving a commercial light-emitting diode (LED) [288]. Du et al. also proposed self-powered electronics by integrating of flexible graphene-based supercapacitors into perovskite hybrid solar cells [289]. However, a PV cell works only under sufficient light illumination. Furthermore, solar energy is not always available, strongly depending on the weather, working conditions, etc. The intermittent and unpredictable nature of solar energy is an inevitable challenge for its expansion as a reliable power supply system in wearable electronics. To develop a practical strategy to simultaneously scavenge multiple types of energies from the environment, the concept of a hybridized energy harvester incorporating two kinds of conversion cells for concurrently scavenging solar and mechanical energies was proposed, so that the energy resources could be effectively and complementarily used [290–293]. Recently, Wen et al. proposed a hybridized self-charging power textile system with the aim of simultaneously collecting outdoor sunshine and random body motion energies and then storing them in an energy storage unit. Both of the harvested energies can be easily converted into electricity by using fiber-shaped DSSCs (for solar energy) and fiber-shaped triboelectric nano generatores (for random body motion energy) and then further stored as chemical energy in fiber-shaped supercapacitors. Because of the all-fiber-shaped structure of the entire system, their proposed hybridized self-charging textile system can be easily woven into electronic textiles to fabricate smart clothes to sustainably operate mobile or wearable electronics (Fig. 15) [294].

168

Global commercial companies that develop dye solar cells/modules. The information is inferred from their websites/published reports on websites

Company

Affiliation

Core business

Major achievements in DSSCs

Dyesol G24

Australia United Kingdom

DSSC materials and commercial development Solar power, especially third generation PVs

S Samsung SDI Sharp Konarka Solaronix

Korea Japan United States Switzerland

Electronic devices such as LCDs, mobile phones Electronic products Spin off company of MIT for DSSCs DSSC material and commercial development

Dynamo Dyepower

Sweden Italy

EEPL Yingkou OPV Tech New Energy Co., Ltd. Gunze Ltd. Energy Research Center of the Netherlands

Switzerland China

DSSC materials Development of dye solar modules for BIPV and facades applications Organic PVs, materials and characterization Upscaling DSSCs

Japan Netherlands

Electronic components and garments Energy research institute

Institue of Plasma Physics Fraunhofer ISE

China Germany

J Touch Taiwan

Taiwan

Utilization of fusion energy Environmentally friendly energy harvesting and storage research Touch panel solutions

KIST Panasonic Denko Co. Ltd. Ricoh Merck Exeger Dye Tec Solar Tata Steel Europe

Korean Japan Japan Germany Sweden United States India

Materials for DSSC Electronic appliances Electronic Chemicals and pharmaceuticals Industrial production BIPVs Steel roofing

BIPVs, integration of DSSCs in roof materials (in process) Started DSSC plant in 2007, flexible waterproof bags, commercial applications for indoor electronics, such as key-boards, mouse, ebook covers, solar bags Dye solar panels for BIPVs, smart windows with integrated storage Certified highest module PCE (8.2%) in W-type module Licensed its DSSCs IP to G24 Panels with an active area of 200 m2 with an estimated annual production of 2000 kWh Preparing Co-based electrolytes and porphyrin dyes Running an automated pilot line for the production of A4 size dye solar modules and for large-area panels Held the main patent on mesoporous dye solar cell structure Colorful, artistic and transparent dye solar modules in the form of glass windows and screen Wearable DSSC First EU lab for DSSC development started in 1995. Introduced master plate design installed semi-automated DSSC manufacturing up to 100 cm2 A 500 W DSSC power station was installed in 2004 Scalable development and stability research on DSSC, first large area (30  30 cm2) glass frit-based module Started using DSSCs in indoor electronics such as portable time clocks Flexible DSSCs Stability improvement DSSC for indoor lighting Electrolytes for DSSCs Pilot plant installed for BIPVs and AIPVs Joint venture of Dyesol and Pikington Joint venture of Dyesol and Tate Steel

Abbreviations: AIPV, automotive integrated photovoltaics; BIPVs, building-integrated photovoltaics; DSSC, dye-sensitized solar cells; EEPL, Enrich Energy Pvt. Ltd.; LCD, liquid-crystal display; PCE, power conversion efficiency; PV, photovoltaic.

Dye-Sensitized Materials

Table 2

Dye-Sensitized Materials

169

Mouse

vers k co

o e-bo

(A)

(B)

(D)

(C)

(E)

Fig. 14 Images of few commercial dye-sensitized solar cells products on the market are indoor electronics by G24 ((A)–(C)), while (D) and (E) are developed by 3G Solar.

2.6.6.2.3

Decorative elements

Color tunability is one of the prominent properties of DSSCs, in fact the pigment responsible for light harvesting could be easily changed and with it the color of the device [21], different combinations of pigments could also be used and thus complex coloring achieved. In Fig. 16 are reported a few examples of differently colored semitransparent DSSCs and building integrated cells.

2.6.6.2.4

Building integrated photovoltaics

The EU is committed to reducing carbon dioxide emissions with a target of 20% renewable energy by 2020 and has mandates for zero energy buildings. This initiative is expected to lower and meet energy needs of buildings to a large extent with renewable energies. Such favorable initiatives by governments are anticipated to drive the DSSC market over the forecast period. There are many features that make DSSC a clean, green technology inherently suitable to application in the built environment where the largest part of human activity occurs and where electricity demand is highest. (1) Unique chemistry and materials architecture. DSSC is the technology with the thinnest-possible photoactive absorbing layer: one single molecular layer of a sensitizer or dye spread out over a high surface area, low cost material, such as, titanium dioxide. Thus, DSSC is the ultimate “miser” when it comes to the usage of natural resources. Another important aspect of DSSC, which distinguishes this technology from all other PV systems, is its nanotechnology basis. Nanostructured TiO2, which provides the host matrix for the photoactive dye, offers unique electronic properties, optical properties (such as transparency), and mechanical properties.

170

Dye-Sensitized Materials

(A)

(B)

(C)

(E)

(G)

(J)

(D)

(F)

(H)

(K)

(I)

(L)

Fig. 15 (A) Schematic diagram and (B) photograph (scale bar¼1 cm) of a single F-dye-sensitized solar cells (DSSC), consisting of N719 dye–adsorbed TiO2 nanotube arrays on a Ti wire as a working electrode and a Pt-coated carbon fiber as a charge extraction (CE) in an I /I3 -based electrolyte, (C) low-magnification and (D) high-magnification scanning electron microscope (SEM) images of the TiO2 nanotube arrays on the Ti wire (scale bars¼100 mm (C) and 100 nm (D)], (E) J–V curve of a single F-DSSC (inset shows the Nyquist plot of an F-DSSC, which is measured under VOC with frequencies ranging from 100 kHz to 10 MHz), (F) normalized current density of the single F-DSSC at different bending angles (0–180 degrees) (insets show the photograph of a single F-DSSC at different bending angles). Photograph of the self-charging power textile woven with F-triboelectric nanogenerators (TENGs), F-DSSCs, and F-SCs under outdoor (G), indoor (H), and movement (I) conditions, (J) circuit diagram of the self-charging powered textile for wearable electronics (WE), (K) charging curve of the F-DSSC and the F-TENG, where the light blue–shaded area corresponds to the charging curve of the F-DSSC and the light red–shaded area corresponds to the charging curve of the F-DSSC–F-TENG hybrid. The top left corner inset shows an enlarged curve during the F-DSSC charging period, and the bottom right corner inset shows the rectified ISC of F-TENGs, (L) normalized QSC values of F-TENGs, ISC values of F-DSSCs, and capacitances of F-SCs bent between 0 and 180 degrees for 1000 cycles. Insets show the photographs of the two final bending statuses (both scale bars¼1 cm). a.u., arbitrary units. Reproduced from Wen Z, Yeh M-H, Guo H, et al. Self-powered textile for wearable electronics by hybridizing fiber-shaped nanogenerators, solar cells, and supercapacitors. Sci Adv 2016;2:e1600097.

Dye-Sensitized Materials

171

Fig. 16 Application fields for colorful, highly transparent and decorative dye-sensitized solar cells. Reproduced from Cannavale A, Cossari P, Eperon GE, et al. Forthcoming perspectives of photoelectrochromic devices: a critical review. Energy Environ Sci 2016;9:2682–719.

(2) Low toxicity raw materials. The major DSSC materials, such as carbon and titanium are elements we are in daily contact with through our food, through the air we breathe, the toothpaste or sunscreen we use, the plates, cups and glasses we eat and drink from the cutlery we daily used as “food-to-mouth” interface. (3) Low manufacturing costs for DSSC. Since traditional PV technologies heavily rely on vacuum processing and require extremely high purity of materials and stringent cleanliness for the manufacturing environment, these technologies are generally based on expensive equipment, including the most sophisticated and energy-hungry clean rooms and all factory workers wearing “space-suit”-type work gear. In contrast, manufacture of DSSC relies mainly on printing, baking, and packaging processes. (4) Better performance in diffuse light. Most standard solar panels suffer from significant loss of efficiency at lower light levels because losses due to electron-hole pair recombination within the semiconductor phase becoming relatively more important. With DSSC in contrast, photogenerated positive and negative charges are effectively separated into two different phases: the electron is injected very rapidly into the solid phase of titanium dioxide, while the positive charge is chemically “neutralized” within nanoseconds from the liquid electrolyte phase. DSSC application on glass building facades is one of the most exciting areas. Large expanses of glass in modern office buildings and smaller windows in residential buildings are an ideal application for DSSC clean energy generation right at the point of use in the often cloudy and shaded urban environment. DSSC application in building integrated photovoltaics (BIPV) and building applied photovoltaics (BAPV) is expected to witness significant growth to reach a net worth exceeding USD 30 million by 2022. Growing demand to cater increasing energy requirements for residential areas including water heating and roof-top panels is anticipated to drive the market growth over the forecast period. DSSCs work well in diffused or low light and are less affected by the angle of incidence, with efficiencies that increase with angle of light incidence by B10%–16% [295]. Moreover, BIPVs have an advantage over centralized solar power generation because the electricity generated is consumed in place, avoiding transmission line losses, and infrastructure cost. The Australian

172

Dye-Sensitized Materials

Fig. 17 (A) and (B) Dye-sensitized solar cells modules installation recently completed at the Enrich Energy Pvt. Ltd. (EEPL) campus. Photos are courtesy of David Martineau of Solaronix Ltd.

company Sustainable Technologies International has produced electric-power-producing glass tiles on a large scale for field testing and the first building has been equipped with a wall of this type. In the United States of America, Dyesol Limited’s American subsidiary, Dyesol Inc., continues advancing work in BIPV market with particular focus on DSSC enabled glass building products. In March 2012, Dyesol announced progress on this front, including surmounting specific challenges associated with assembly of large sized glass based DSSC panels and also successful assembly of a large scale glass DSSC panel. This large DSSC glass panel exceeded 1.2 m  60 cm in size and represents the largest continuous substrate, single circuit series connected DSSC device made to date. DSSC with active areas of B200 m2 were installed as windows of a green building in April, 2014 at the École Polytechnique Fédérale De Lausanne (EFPL) campus, Switzerland (Fig. 17(A) and (B)). These panels are estimated to be able to generate about 2000 kW h of an annual solar electricity. The transparent panels are in five different colors (representing a desirable unique feature of DSSCs) and were developed by Solaronix. This massive production is an important milestone for the commercial deployment of DSSCs. Oxford PVs, a spin off company of Oxford University, is working on all solid state perovskite-based modules. However, the lifetime of the solid-state perovskite-based modules under outdoor conditions is yet to be determined. For residential and commercial users, a significant portion (10%–30%) of the peak electricity demand arises from overheating caused by glass windows that allow heat radiation to enter a building [296]. This share can be reduced by (1) controlling the optical transparency of the windows to block heat radiation from entering the buildings and (2) converting the passive windows into smart windows that can harvest the solar light into electricity [297–299]. Toward the second strategy, an energy storage smart window, also called an electrochromic device (ECD), that simultaneously harvests and stores solar energy by integrating two electrochemical devices (DSSC and supercapacitors) has been developed [300,301]. The ECD can be divided into two parts: PV and EC components separated by an ionic conductor with negligible electronic conductivity to avoid short circuits between them. Such devices are self-powered unlike the conventional ECD windows, which need an energy supply for operation. The window changes from a bleached to a colored state due to the reversible reaction in the electrolyte upon light absorption [302]. The main concerns for future large-area applications are the possible loss of the energy generated by the PV device for larger dimensions, a small range of optical modulation and rather low transmittances in the clear state. Dyesol has recently launched projects on the integration of DSSCs into buildings with their various industrial partners (Tata Steel Europe of United Kingdom, Pilkington North America of United States, Timo Technologies of South Korea) and on DSSC-powered combined energy generation and storage devices. Theses projects are expected to further reduce the cost kW 1 h 1 as double-coated glass windows in the buildings meant for UV protection and antireflective coatings are used as substrates for DSSC.

2.6.6.2.5

Automotive

Cost effective DSSC modules are used in the automotive sector to meet energy requirements. Growing demand for smart cars integrated with parking guidance systems is expected to drive dye solar cell demand in automotive integrated photovoltaics (AIPV). Growing AIPV market, particularly in Europe and North America, is anticipated to steer the market expansion in these regions. DSSC demand in AIPV is expected to witness average growth over the forecast period. Growing consumer demand for interior lighting and ambiance in vehicles is expected to drive energy requirements in automobiles.

2.6.6.3

Future Direction

Several major advancements have been made to upscale DSSCs in terms of their various interconnection designs, material components, scalable fabrication processes, and innovative applications such as a hybrid energy harvesting and storage devices.

Dye-Sensitized Materials

173

G24-i FhG-ISE: 2012

DSSC modules Perovskites SCs

ECN, first stability testing (Prog photovol: Res app, 2001;9:425)

DSSC achieved outdoor performance specification standards laid out for silicon solar cells (APL 86, 123508, 2005

BIPVs

2012:1st commercial module (60×1002 cm) reported (FhG-ISE) Oct:2009, First commercial shipment by G24i (0.5 kW) 2009, Highest confirmed PCE (8.2%, A=~25 cm2) by SHARP

t

e ur

To

he

fu

Smart windows

t

14

K-19 dye + robust electrolyte

20

2011−2013 10

0

8 00

Monolithic design (EFPL, 1999)

−2

2

2005−2007

DSSC modules

4

00

2 2−

Dyesol: DSSCs as bathroom interior

0

96

9 −1

29

5.29% (EFPL-1998) First DSSC module series type connection

E rep CN: f rod irst s u of ctible emia DS u SC man toma mo ufac ted du tur les ing

1999−2001 98

20

DSSC participated in rally, 2008, travelled 225 km in 3 days

2013: First commercial shipment, Solaronix

First stability testing on large modules

Fig. 18 Step-by-step historical evaluation of dye-sensitized solar modules. BIPV, building integrated photovoltaics; DSSC, dye-sensitized solar cell; ECN, Energy Research Centre of the Netherlands; SC, solar cell. Reproduced from Fakharuddin A, Jose R, Brown TM, Fabregat-Santiago F, Bisquert J. A perspective on the production of dye-sensitized solar modules. Energy Environ Sci 2014;7:3952–81.

Fig. 18 shows a timeline of DSSC development since their first report in 1996. It highlights the major achievements in DSSC development and also projects their future progress (Fig. 18) [303]. These advancements are paving the way toward the ultimate goals of this technology: successful commercial deployment as a fossil fuel alternative and cost-effective competition with the conventional PVs. However, there are still problems need to be overcome. The present highest efficiency of DSSCs is relatively low in comparison with that of 25% for bulk silicon solar cells. High energy conversion efficiency is one of the most important keys to the commercialization of DSSCs in the huge electricity generation market. A tandem structure is commonly used in solar cells to increase the efficiency. Until now, tandem DSSCs based on black dye and N719 dye had successfully obtained efficiency of over 10% [304–306]. For higher efficiency of over 20%, the large loss-in-potential in present DSSCs, which is estimated to be at least 0.5 V, with 0.2 V over-potential for efficient electron injection and another 0.3 V over-potential for the regeneration of oxidized dyes, should be minimized. This large loss-in-potential is mainly due to the potential dispersion in the materials of TiO2, dyes, and redox shuttles. Hence, the most promising way to significantly enhance the efficiency is to develop new materials with less potential dispersion, such as nanocrystalline TiO2 particles with only one crystallographic plane exposed for dye adsorption by which the potential dispersion of conduction bands of different planes will be reduced, new dyes with rigid structure to decrease the dispersion of vibrational energy of the dyes, and redox shuttles with less derivatives to lower the dispersion in redox potentials. If the loss-in-potential is reduced to less than 0.2 V, we will obtain a much wider IPCE spectrum and higher Voc simultaneously. Therefore, we expect the long-term goal to improve the DSSCs efficiency to over 20%, as high as that of the champion efficiency of crystalline silicon solar cells, can be achieved when all the aforementioned issues are solved. However, this will be very hard to achieve because it requires deeper fundamental researches and application of new concepts. We hope more young scientists with different backgrounds will courageously accept these challenges. Reproducibility of cell fabrication and scale-up while the DSSC can be produced in a relatively simple way in the laboratory without employing a glovebox or high vacuum steps, a rigorous protocol needs to be applied during cell fabrication to achieve high efficiencies in a reproducible manner. By taking the appropriate precautions, relative variations of the efficiency of less than 2%–3% can be readily achieved for laboratory cells. Thus, a detailed procedure providing a guide to realize reproducibly cell efficiency values over 10% has been published recently. Reproducible manufacturing of DSSC modules on a semi-automated baseline has also been reported. Because of significant industrial up-scaling efforts, the conversion efficiency of DSSC modules has

Dye-Sensitized Materials

174

been steadily rising over the past few years, with the certified value measured under AM1.5 standard conditions reaching currently 8.2%. Stability and commercial development of the DSSC long-term stability is a key requirement for all types of solar cells. A vast amount of tests have therefore been carried over the last 15 years to scrutinize the stability of the DSSC by both academic and industrial institutions. Most of the earlier work has been reviewed. Long-term accelerated light-soaking experiments performed over many thousands of hours under full or even concentrated sunlight have confirmed the intrinsic stability of current DSSC embodiments. Stable operation under high-temperature stress at 80–851C as well as under damp heat and temperature cycling has been achieved by judicious molecular engineering of the sensitizers with the use of robust and nonvolatile electrolytes, such as, ionic liquids and adequate sealing materials. In the early development stage of the DSSC technology, the quality of device sealing was sometimes not appropriate in laboratory test cells, causing leakage of the volatile solvents. Most research groups with longer practical experience, including industrial enterprises, have overcome this by improving the sealing methods. Because of the direct relevance to the manufacturing of commercial products, little is published on these processing issues though. Good results on overall system endurance have been reported for several years, demonstrating excellent stability under accelerated laboratory test conditions. These promising results are presently being confirmed under real outdoor conditions. From these extensive studies, confidence has emerged that the DSSCs can match the stability requirements needed to sustain outdoor operation for at least 20 years. This has paved the way for the recent worldwide surge in the industrial development and commercialization of the DSSC.

2.6.7

Closing Remarks

Four main components of DSSCs are photoanode, dye, electrolyte, and counter electrode. Since the breakthrough in 1991, all aspects related to DSSCs have been subjected to improvement for better overall PV performance. Furthermore, it will be the positive combination of efficiency and stability which will determine the commercial success of DSSCs. The start of significant market exploitation is on the way. Companies in Japan, Switzerland, Australia, United Kingdom, and United States are manufacturing and commercializing-or intend to commercialize DSSCs.

Acknowledgment This work was supported by the ACS Petroleum Research Fund (PRF-51799-ND10). WW and YHH thank Charles and Carroll McArthur for their great support.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24]

McGehee MD, Goh C. Organic semiconductors for low-cost solar cells. AIP Conf Proc 2008;1044:322–30. Lewis NS. Powering the planet. MRS Bull 2011;32:808–20. Hamann TW, Jensen RA, Martinson ABF, Van Ryswyk H, Hupp JT. Advancing beyond current generation dye-sensitized solar cells. Energy Environ Sci 2008;1:66–78. Hirsch RL. Mitigation of maximum world oil production: shortage scenarios. Energy Policy 2008;36:881–9. Hu YH, Wang H, Hu B. Thinnest two-dimensional nanomaterial – graphene for solar energy. ChemSusChem 2010;3:782–96. Brown GF, Wu J. Third generation photovoltaics. Laser Photon Rev 2009;3:394–405. Kazmerski LL, Ayyagari MS, White FR, Sanborn GA. Growth and properties of vacuum deposited CuInSe2 thin films. J Vac Sci Technol 1976;13:139–44. Brown BJ, Bates Jr. CW. The effect of substrate on the morphology of CuInSe2 films prepared by chemical spray pyrolysis for CuInSe2/Cds solar cells. J Appl Phys 1990;68:2517–9. Hong KJ, Jeong TS, Youn CJ. Temperature dependence and valence band splitting of the photocurrent response in undoped p-type CuInSe2 layers. J Appl Phys 2006;100:123518. Gloeckler M, Sites JR, Metzger WK. Grain-boundary recombination in Cu(In,Ga)Se2 solar cells. J Appl Phys 2005;98:113704. Malm U, Edoff M. Influence from front contact sheet resistance on extracted diode parameters in cigs solar cells. Prog Photovolt Res Appl 2008;16:113–21. Li G, Shrotriya V, Huang JS, Yao Y, Moriarty T. High-efficiency solution processable polymer photovoltaic cells by self-organization of polymer blends. Nat Mater 2005;4:864–8. Hoppe H, Sariciftci NS. Organic solar cells: an overview. J Mater Res 2004;19:1924–45. Niggemann M, Glatthaar M, Gombert A, Hinsch A, Wittwer V. Diffraction gratings and buried nano-electrodes – architectures for organic solar cells. Thin Solid Films 2004;451–452:619–23. Niggemann M, Zimmermann B, Haschke J, Glatthaar M, Gombert A. Organic solar cell modules for specific applications – from energy autonomous systems to large area photovoltaics. Thin Solid Films 2008;516:7181–7. Oseni SO, Mola GT. Properties of functional layers in inverted thin film organic solar cells. Sol Energy Mater Sol Cells 2017;160:241–56. Peumans P, Yakimov A, Forrest SR. Small molecular weight organic thin-film photodetectors and solar cells. J Appl Phys 2003;93:3693–723. Pivrikas A, Sariciftci NS, Juska G, Osterbacka R. A review of charge transport and recombination in polymer/fullerene organic solar cells. Prog Photovolt 2007;15:677–96. Wang X, Zhi L, Tsao N, Tomović Ž, Li J, Müllen K. Transparent carbon films as electrodes in organic solar cells. Angew Chem Int Ed 2008;47:2990–2. Yang Z, Chen T, He R, et al. Aligned carbon nanotube sheets for the electrodes of organic solar cells. Adv Mater 2011;23:5436–9. Hagfeldt A, Boschloo G, Sun L, Kloo L, Pettersson H. Dye-sensitized solar cells. Chem Rev 2010;110:6595–663. O’Regan B, Gratzel M. A low-cost, high-efficiency solar cell based on dye-sensitized colloidal TiO2 films. Nature 1991;353:737–40. Baxter JB, Aydil ES. Nanowire-based dye-sensitized solar cells. Appl Phys Lett 2005;86:053114. Adachi M, Murata Y, Takao J, Jiu JT, Sakamoto M, Wang FM. Highly efficient dye-sensitized solar cells with a titania thin-film electrode composed of a network structure of single-crystal-like TiO2 nanowires made by the “oriented attachment” mechanism. J Am Chem Soc 2004;126:14943–9.

Dye-Sensitized Materials

[25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71]

175

Baxter JB, Aydil ES. Dye-sensitized solar cells based on semiconductor morphologies with ZnO nanowires. Sol Energy Mater Sol Cells 2006;90:607–22. Chen T, Qiu LB, Cai ZB, et al. Intertwined aligned carbon nanotube fiber based dye-sensitized solar cells. Nano Lett 2012;12:2568–72. Jose R, Thavasi V, Ramakrishna S. Metal oxides for dye-sensitized solar cells. J Am Ceram Soc 2009;92:289–301. Law M, Greene LE, Radenovic A, Kuykendall T, Liphardt J, Yang PD. ZnO–Al2O3 and ZnO–TiO2 core-shell nanowire dye-sensitized solar cells. J Phys Chem B 2006;110:22652–63. Zhang S, Ji CY, Bian ZQ, et al. Single-wire dye-sensitized solar cells wrapped by carbon nanotube film electrodes. Nano Lett 2011;11:3383–7. Ganesan S, Muthuraaman B, Mathew V, Madhavan J, Maruthamuthu R. Performance of a new polymer electrolyte incorporated with diphenylamine in nanocrystalline dye-sensitized solar cell. Sol Energy Mater Sol Cells 2008;92:1718–22. Boschloo G, Haggman L, Hagfeldt A. Quantification of the effect of 4-tert-butylpyridine addition to I /I3 redox electrolytes in dye-sensitized nanostructured TiO2 solar cells. J Phys Chem B 2006;110:13144–50. Adachi M, Sakamoto M, Jiu JT, Ogata Y, Isoda S. Determination of parameters of electron transport in dye-sensitized solar cells using electrochemical impedance spectroscopy. J Phys Chem B 2006;110:13872–80. Benkstein KD, Kopidakis N, van de Lagemaat J, Frank AJ. Influence of the percolation network geometry on electron transport in dye-sensitized titanium dioxide solar cells. J Phys Chem B 2003;107:7759–67. Bisquert J, Fabregat-Santiago F, Mora-Sero I, Garcia-Belmonte G, Gimenez S. Electron lifetime in dye-sensitized solar cells: theory and interpretation of measurements. J Phys Chem C 2009;113:17278–90. Bisquert J, Zaban A, Greenshtein M, Mora-Sero I. Determination of rate constants for charge transfer and the distribution of semiconductor and electrolyte electronic energy levels in dye-sensitized solar cells by open-circuit photovoltage decay method. J Am Chem Soc 2004;126:13550–9. Bisquert J, Zaban A, Salvador P. Analysis of the mechanisms of electron recombination in nanoporous TiO2 dye-sensitized solar cells. Nonequilibrium steady-state statistics and interfacial electron transfer via surface states. J Phys Chem B 2002;106:8774–82. Cameron PJ, Peter LM, Hore S. How important is the back reaction of electrons via the substrate in dye-sensitized nanocrystalline solar cells? J Phys Chem B 2005;109:930–6. Duffy NW, Peter LM, Rajapakse RMG, Wijayantha KGU. A novel charge extraction method for the study of electron transport and interfacial transfer in dye sensitised nanocrystalline solar cells. Electrochem Commun 2000;2:658–62. Park NG, van de Lagemaat J, Frank AJ. Comparison of dye-sensitized rutile- and anatase-based TiO2 solar cells. J Phys Chem B 2000;104:8989–94. Schlichthörl G, Huang SY, Sprague J, Frank AJ. Band edge movement and recombination kinetics in dye-sensitized nanocrystalline TiO2 solar cells: a study by intensity modulated photovoltage spectroscopy. J Phys Chem B 1997;101:8141–55. Wang Q, Ito S, Grätzel M, et al. Characteristics of high efficiency dye-sensitized solar cells. J Phys Chem B 2006;110:25210–21. Bisquert J. Theory of the impedance of electron diffusion and recombination in a thin layer. J Phys Chem B 2002;106:325–33. Li L-L, Chang Y-C, Wu H-P, Diau EW-G. Characterisation of electron transport and charge recombination using temporally resolved and frequency-domain techniques for dye-sensitised solar cells. Int Rev Phys Chem 2012;31:420–67. Park JH, Lee TW, Kang MG. Growth, detachment and transfer of highly-ordered TiO2 nanotube arrays: use in dye-sensitized solar cells. Chem Commun 2008;2867–9. doi:10.1039/b800660a. Hore S, Vetter C, Kern R, Smit H, Hinsch A. Influence of scattering layers on efficiency of dye-sensitized solar cells. Sol Energy Mater Sol Cells 2006;90:1176–88. Wu KJ, Yu Y, Shen K, Xia CR, Wang DL. Effect of ultra-thin ZnO coating layer on the device performance of TiO2 dye sensitized solar cell. Sol Energy 2013;94:195–201. Gratzel M. Conversion of sunlight to electric power by nanocrystalline dye-sensitized solar cells. J Photochem Photobiol A 2004;164:3–14. Adachi M, Murata Y, Okada I, Yoshikawa S. Formation of titania nanotubes and applications for dye-sensitized solar cells. J Electrochem Soc 2003;150:G488–93. Bach U, Lupo D, Comte P, et al. Solid-state dye-sensitized mesoporous TiO2 solar cells with high photon-to-electron conversion efficiencies. Nature 1998;395:583–5. Bisquert J, Vikhrenko VS. Interpretation of the time constants measured by kinetic techniques in nanostructured semiconductor electrodes and dye-sensitized solar cells. J Phys Chem B 2004;108:2313–22. Barbé CJ, Arendse F, Comte P, et al. Nanocrystalline titanium oxide electrodes for photovoltaic applications. J Am Ceram Soc 1997;80:3157–71. Mathew S, Yella A, Gao P, et al. Dye-sensitized solar cells with 13% efficiency achieved through the molecular engineering of porphyrin sensitizers. Nat Chem 2014;6:242–7. Kalyanasundaram K, Grätzel M. Applications of functionalized transition metal complexes in photonic and optoelectronic devices. Coord Chem Rev 1998;177:347–414. Deb SK. Dye-sensitized TiO2 thin-film solar cell research at the national renewable energy laboratory (NREL). Sol Energy Mater Sol Cells 2005;88:1–10. Wold A. Photocatalytic properties of titanium dioxide (TiO2). Chem Mater 1993;5:280–3. Hoffmann MR, Martin ST, Choi W, Bahnemann DW. Environmental applications of semiconductor photocatalysis. Chem Rev 1995;95:69–96. Perekrestov R, Kudrna P, Tichy M, Khalakhan I, Danis S. Crystalline structure and morphology of TiO2 thin films deposited by means of hollow cathode plasma jet with supporting anode. Surf Coat Technol 2016;291:123–9. Das P, Sengupta D, Kasinadhuni U, Mondal B, Mukherjee K. Nano-crystalline thin and nano-particulate thick TiO2 layer: cost effective sequential deposition and study on dye sensitized solar cell characteristics. Mater Res Bull 2015;66:32–8. Yen Y-C, Chen J-Z, Lin K-J. Hydrothermal synthesis of high-surface-area anatase TiO2 nanoparticles for enhancing the photovoltaic performance of solar cells. J Chin Chem Soc 2013;60:705–9. Kim B-M, Rho S-G, Kang C-H. Effects of TiO2 structures in dye-sensitized solar cell. J Nansci Nanotechnol 2011;11:1515–7. Gajjela S, Ananthanarayanan K, Yap C, Graetzel M, Balaya P. Synthesis of mesoporous titanium dioxide by soft template based approach: characterization and application in dye-sensitized solar cells. Energy Environ Sci 2010;3:838–45. Zhou R, Guo W, Yu R, Pan C. Highly flexible, conductive and catalytic Pt networks as transparent counter electrodes for wearable dye-sensitized solar cells. J Mater Chem A 2015;3:23028–34. Jiu J, Isoda S, Wang F, Adachi M. Dye-sensitized solar cells based on a single-crystalline TiO2 nanorod film. J Phys Chem B 2006;110:2087–92. Gratzel M. Solar energy conversion by dye-sensitized photovoltaic cells. Inorg Chem 2005;44:6841–51. Law M, Greene LE, Johnson JC, Saykally R, Yang P. Nanowire dye-sensitized solar cells. Nat Mater 2005;4:455–9. Wang H, Leonard SL, Hu YH. Promoting effect of graphene on dye-sensitized solar cells. Ind Eng Chem Res 2012;51:10613–20. Chandiran AK, Abdi-Jalebi M, Nazeeruddin MK, Grätzel M. Analysis of electron transfer properties of ZnO and TiO2 photoanodes for dye-sensitized solar cells. ACS Nano 2014;8:2261–8. Banik A, Ansari MS, Sahu TK, Qureshi M. Understanding the role of silica nanospheres with their light scattering and energy barrier properties in enhancing the photovoltaic performance of ZnO based solar cells. Phys Chem Chem Phys 2016;18:27818–28. Bavir M, Fattah A. An investigation and simulation of the graphene performance in dye-sensitized solar cell. Opt Quantum Electron 2016;48:559. Habibi MH, Mardani M, Habibi M, Zendehdel M. Enhanced photovoltage (Voc) of nano-structured zinc tin oxide (ZtO) working electrode prepared by a green hydrothermal route for dye-sensitized solar cell (DSSC). J Mater Sci Mater Electron 2017;28:3789–95. Lai FI, Yang JF, Kuo SY. Efficiency enhancement of dye-sensitized solar cells’ performance with ZnO nanorods grown by low-temperature hydrothermal reaction. Materials 2015;8:8860–7.

176

Dye-Sensitized Materials

[72] Lee SY, Kim SH. Deposition of zinc oxide photoelectrode using plasma enhanced chemical vapor deposition for dye-sensitized solar cells. J Nanosci Nanotechnol 2014;14:9272–8. [73] Ma CW, Chang CM, Huang PC, Yang YJ. Sea-urchin-like ZnO nanoparticle film for dye-sensitized solar cells. J Nanomater 2015;2015:679474–9. [74] Mahmoud SA, Fouad OA. Synthesis and application of zinc/tin oxide nanostructures in photocatalysis and dye sensitized solar cells. Sol Energy Mater Sol Cells 2015;136:38–43. [75] Pandey P, Parra MR, Haque FZ, Kurchania R. Effects of annealing temperature optimization on the efficiency of ZnO nanoparticles photoanode based dye sensitized solar cells. J Mater Sci Mater Electron 2017;28:1537–45. [76] Rani AA, Ernest S. Enhancement of efficiency in dye-sensitized solar cell using spray deposited Ga doped ZnO thin films. J Mater Sci Mater Electron 2015;26:762–8. [77] Zarghami Z, Ramezani M, Motevalli K. ZnO nanorods/nanoparticles: novel hydrothermal synthesis, characterization and formation mechanism for increasing the efficiency of dye-sensitized solar cells. J Cluster Sci 2016;27:1451–62. [78] Bahadur L, Kushwaha S. Structural and optical properties of tripod-like ZnO thin film and its application in dye-sensitized solar cell. J Solid State Electrochem 2013;17:2001–8. [79] Habibi MH, Karimi B, Zendehdel M, Habibi M. Fabrication, characterization of two nano-composite CuO–ZnO working electrodes for dye-sensitized solar cell. Spectrochim Acta Part A Mol Biomol Spectrosc 2013;116:374–80. [80] Kaur M, Verma NK. Study on CaCO3-coated ZnO nanoparticles based dye sensitized solar cell. J Mater Sci 2013;24:4980–6. [81] Kaur M, Verma NK. Performance of Eu2O3 coated ZnO nanoparticles-based DSSC. J Mater Sci 2013;24:3617–23. [82] Kim YR, Park JY, Lee DK, Ahn KS, Kim JH. Dye-sensitized solar cells composed of well-aligned ZnO nanorod array grown with chemical bath deposition method as the photo-electrode. Mol Cryst Liq Cryst 2014;597:120–7. [83] Kushwaha S, Bahadur L. Studies of structural and morphological characteristics of flower-like ZnO thin film and its application as photovoltaic material. Optik 2013;124:5696–701. [84] Li H, Bai JF, Feng B, et al. Dye-sensitized solar cells with a tri-layer ZnO photo-electrode. J Alloys Compd 2013;578:507–11. [85] Raj CJ, Prabakar K, Karthick SN, Hemalatha KV, Son MK, Kim HJ. Banyan root structured Mg-doped ZnO photoanode dye-sensitized solar cells. J Phys Chem C 2013;117:2600–7. [86] Wang H, Li HY, Wang JS, Wu JS, Liu M. Influence of applied voltage on anodized TiO2 nanotube arrays and their performance on dye sensitized solar cells. J Nanosci Nanotechnol 2013;13:4183–8. [87] Memarian N, Concina I, Braga A, Rozati SM, Vomiero A, Sberveglieri G. Hierarchically assembled ZnO nanocrystallites for high-efficiency dye-sensitized solar cells. Angew Chem Int Ed 2011;50:12321–5. [88] Cheng HM, Chiu WH, Lee CH, Tsai SY, Hsieh WF. Formation of branched ZnO nanowires from solvothermal method and dye-sensitized solar cells applications. J Phys Chem C 2008;112:16359–64. [89] Li Z, Yang H, Wu F, Fu J, Wang L. Single-crystalline self-branched anatase titania nanowires for dye-sensitized solar cells. Electron Mater Lett 2017;13:174–8. [90] Hu H, Shen J, Cao X, Wang H, Lv H. Photo-assisted deposition of Ag nanoparticles on branched TiO2 nanorod arrays for dye-sensitized solar cells with enhanced efficiency. J Alloys Compd 2017;694:653–61. [91] Keis K, Lindgren J, Lindquist S-E, Hagfeldt A. Studies of the adsorption process of Ru complexes in nanoporous ZnO electrodes. Langmuir 2000;16:4688–94. [92] Strobel A, Schnabel H-D, Reinhold U, Rauer S, Neidhardt A. Room temperature plasma enhanced atomic layer deposition for TiO2 and WO3 films. J Vac Sci Technol A 2016;34:01A118. [93] Li H, Duan X, Liu G, Jia X, Liu X. Morphology controllable synthesis of TiO2 by a facile hydrothermal process. Mater Lett 2008;62:4035–7. [94] Kim S-H, Cho Y-S, Jeon S-J, Eun TH, Yi G-R, Yang S-M. Microspheres with tunable refractive index by controlled assembly of nanoparticles. Adv Mater 2008;20:3268–73. [95] Yi GR, Manoharan VN, Michel E, Elsesser MT, Yang SM, Pine DJ. Colloidal clusters of silica or polymer microspheres. Adv Mater 2004;16:1204–8. [96] Cho Y-S, Yi G-R, Kim S-H, Pine DJ, Yang S-M. Colloidal clusters of microspheres from water-in-oil emulsions. Chem Mater 2005;17:5006–13. [97] Kim S-H, Lee SY, Yi G-R, Pine DJ, Yang S-M. Microwave-assisted self-organization of colloidal particles in confining aqueous droplets. J Am Chem Soc 2006;128:10897–904. [98] Moon JH, Yi GR, Yang SM, Pine DJ, Park SB. Electrospray-assisted fabrication of uniform photonic balls. Adv Mater 2004;16:605–9. [99] Sancho-Parramon J, Janicki V, Zorc H. Compositional dependence of absorption coefficient and band-gap for Nb2O5–SiO2 mixture thin films. Thin Solid Films 2008;516:5478–82. [100] Nowak I, Ziolek M. Niobium compounds: preparation, characterization, and application in heterogeneous catalysis. Chem Rev 1999;99:3603–24. [101] Zaban A, Chen SG, Chappel S, Gregg BA. Bilayer nanoporous electrodes for dye sensitized solar cells. Chem Commun 2000;2231–2. doi:10.1039/B005921H. [102] Ahn K-S, Kang M-S, Lee J-K, Shin B-C, Lee J-W. Enhanced electron diffusion length of mesoporous TiO2 film by using Nb2O5 energy barrier for dye-sensitized solar cells. Appl Phys Lett 2006;89:013103. [103] Pfeifer V, Erhart P, Li S, et al. Energy band alignment between anatase and rutile TiO2. J Phys Chem Lett 2013;4:4182–7. [104] Scanlon DO, Dunnill CW, Buckeridge J, et al. Band alignment of rutile and anatase TiO2. Nat Mater 2013;12:798–801. [105] Kitiyanan A, Yoshikawa S. The use of ZrO2 mixed TiO2 nanostructures as efficient dye-sensitized solar cells’ electrodes. Mater Lett 2005;59:4038–40. [106] Tennakone K, Kumara GRRA, Kottegoda IRM, Perera VPS. An efficient dye-sensitized photoelectrochemical solar cell made from oxides of TiN and zinc. Chem Commun 1999;15–6. doi:10.1039/A806801A. [107] Nazeeruddin MK, Kay A, Rodicio I, et al. Conversion of light to electricity by cis-X2bis(2,2'-bipyridyl-4,4'-dicarboxylate)ruthenium(II) charge-transfer sensitizers (X ¼Cl-, Br-, I-, CN-, and SCN-) on nanocrystalline titanium dioxide electrodes. J Am Chem Soc 1993;115:6382–90. [108] Nazeeruddin MK, Baranoff E, Grätzel M. Dye-sensitized solar cells: a brief overview. Sol Energy 2011;85:1172–8. [109] Grätzel M. Dye-sensitized solar cells. J Photochem Photobiol C: Photochem Rev 2003;4:145–53. [110] Goncalves LM, Bermudez VD, Ribeiro HA, Mendes AM. Dye-sensitized solar cells: a safe bet for the future. Energy Environ Sci 2008;1:655–67. [111] Gao F, Wang Y, Zhang J, et al. A new heteroleptic ruthenium sensitizer enhances the absorptivity of mesoporous titania film for a high efficiency dye-sensitized solar cell. Chem Commun 2008;2635–7. doi:10.1039/B802909A. [112] Al-Ghamdi AA, Gupta RK, Kahol PK, et al. Improved solar efficiency by introducing graphene oxide in purple cabbage dye sensitized TiO2 based solar cell. Solid State Commun 2014;183:56–9. [113] Maldonado-Valdivia AI, Galindo EG, Ariza MJ, García-Salinas MJ. Surfactant influence in the performance of titanium dioxide photoelectrodes for dye-sensitized solar cells. Sol Energy 2013;91:263–72. [114] Mishra A, Fischer MKR, Bauerle P. Metal-free organic dyes for dye-sensitized solar cells: From structure: Property relationships to design rules. Angew Chem Int Ed 2009;48:2474–99. [115] Wang ZS, Cui Y, Dan-oh Y, Kasada C, Shinpo A, Hara K. Molecular design of coumarin dyes for stable and efficient organic dye-sensitized solar cells. J Phys Chem C 2008;112:17011–7. [116] Hara K, Kurashige M, Dan-Oh Y, et al. Design of new coumarin dyes having thiophene moieties for highly efficient organic-dye-sensitized solar cells. New J Chem 2003;27:783–5. [117] Luo P, Niu H, Zheng G, Bai X, Zhang M, Wang W. From salmon pink to blue natural sensitizers for solar cells: Canna indica L., salvia splendens, cowberry and Solanum nigrum L. Spectrochim Acta A Mol Biomol Spectrosc 2009;74:936–42.

Dye-Sensitized Materials

177

[118] Hemalatha KV, Karthick SN, Justin Raj C, Hong NY, Kim SK, Kim HJ. Performance of Kerria japonica and Rosa chinensis flower dyes as sensitizers for dye-sensitized solar cells. Spectrochim Acta A Mol Biomol Spectrosc 2012;96:305–9. [119] Ludin NA, Al-Alwani Mahmoud AM, Bakar Mohamad A, Kadhum AAH, Sopian K, Abdul Karim NS. Review on the development of natural dye photosensitizer for dyesensitized solar cells. Renew Sustain Energy Rev 2014;31:386–96. [120] Fleming IAN. Absolute configuration and the structure of chlorophyll. Nature 1967;216:151–2. [121] Tang CW, Albrecht AC. Photovoltaic effects of metal–chlorophyll‐a–metal sandwich cells. J Chem Phys 1975;62:2139–49. [122] Ragoussi M-E, Ince M, Torres T. Recent advances in phthalocyanine-based sensitizers for dye-sensitized solar cells. Eur J Org Chem 2013;2013:6475–89. [123] Chang Y-C, Wang C-L, Pan T-Y, et al. A strategy to design highly efficient porphyrin sensitizers for dye-sensitized solar cells. Chem Commun 2011;47:8910–2. [124] Sayama K, Tsukagoshi S, Mori T, et al. Efficient sensitization of nanocrystalline TiO2 films with cyanine and merocyanine organic dyes. Sol Energy Mater Sol Cells 2003;80:47–71. [125] Zhao W, Jun Hou Y, Song Wang X, et al. Study on squarylium cyanine dyes for photoelectric conversion. Sol Energy Mater Sol Cells 1999;58:173–83. [126] Clifford JN, Palomares E, Nazeeruddin MK, Thampi R, Grätzel M, Durrant JR. Multistep electron transfer processes on dye co-sensitized nanocrystalline TiO2 films. J Am Chem Soc 2004;126:5670–1. [127] Han L, Islam A, Chen H, et al. High-efficiency dye-sensitized solar cell with a novel co-adsorbent. Energy Environ Sci 2012;5:6057–60. [128] Ozawa H, Shimizu R, Arakawa H. Significant improvement in the conversion efficiency of black-dye-based dye-sensitized solar cells by cosensitization with organic dye. RSC Adv 2012;2:3198–200. [129] Wu H-P, Ou Z-W, Pan T-Y, et al. Molecular engineering of cocktail co-sensitization for efficient panchromatic porphyrin-sensitized solar cells. Energy Environ Sci 2012;5:9843–8. [130] Boschloo G, Hagfeldt A. Characteristics of the iodide/triiodide redox mediator in dye-sensitized solar cells. Acc Chem Res 2009;42:1819–26. [131] Snaith HJ. Estimating the maximum attainable efficiency in dye-sensitized solar cells. Adv Funct Mater 2010;20:13–9. [132] Hattori S, Wada Y, Yanagida S, Fukuzumi S. Blue copper model complexes with distorted tetragonal geometry acting as effective electron-transfer mediators in dyesensitized solar cells. J Am Chem Soc 2005;127:9648–54. [133] Daeneke T, Kwon T-H, Holmes AB, Duffy NW, Bach U, Spiccia L. High-efficiency dye-sensitized solar cells with ferrocene-based electrolytes. Nat Chem 2011;3:211–5. [134] Li TC, Spokoyny AM, She C, et al. Ni(iii)/(iv) bis(dicarbollide) as a fast, noncorrosive redox shuttle for dye-sensitized solar cells. J Am Chem Soc 2010;132:4580–2. [135] Nelson JJ, Amick TJ, Elliott CM. Mass transport of polypyridyl cobalt complexes in dye-sensitized solar cells with mesoporous TiO2 photoanodes. J Phys Chem C 2008;112:18255–63. [136] Klahr BM, Hamann TW. Performance enhancement and limitations of cobalt bipyridyl redox shuttles in dye-sensitized solar cells. J Phys Chem C 2009;113:14040–5. [137] Kashif MK, Axelson JC, Duffy NW, et al. A new direction in dye-sensitized solar cells redox mediator development: in situ fine-tuning of the cobalt(ii)/(iii) redox potential through Lewis base interactions. J Am Chem Soc 2012;134:16646–53. [138] Nusbaumer H, Moser J-E, Zakeeruddin SM, Nazeeruddin MK, Grätzel M. Coii(dbbip)22 þ complex rivals tri-iodide/iodide redox mediator in dye-sensitized photovoltaic cells. J Phys Chem B 2001;105:10461–4. [139] Nusbaumer H, Zakeeruddin SM, Moser J-E, Grätzel M. An alternative efficient redox couple for the dye-sensitized solar cell system. Chemistry 2003;9:3756–63. [140] Feldt SM, Gibson EA, Gabrielsson E, Sun L, Boschloo G, Hagfeldt A. Design of organic dyes and cobalt polypyridine redox mediators for high-efficiency dye-sensitized solar cells. J Am Chem Soc 2010;132:16714–24. [141] Yum JH, Baranoff E, Kessler F, et al. A cobalt complex redox shuttle for dye-sensitized solar cells with high open-circuit potentials. Nat Commun 2012;3:631. [142] Liu J, Zhang J, Xu M, Zhou D, Jing X, Wang P. Mesoscopic titania solar cells with the tris(1,10-phenanthroline)cobalt redox shuttle: uniped versus biped organic dyes. Energy Environ Sci 2011;4:3021–9. [143] Zong X, Liang M, Fan C, et al. Design of truxene-based organic dyes for high-efficiency dye-sensitized solar cells employing cobalt redox shuttle. J Phys Chem C 2012;116:11241–50. [144] Yella A, Lee H-W, Tsao HN, et al. Porphyrin-sensitized solar cells with cobalt (ii/iii)-based redox electrolyte exceed 12 percent efficiency. Science 2011;334:629–34. [145] Wang Z-S, Sayama K, Sugihara H. Efficient eosin Y dye-sensitized solar cell containing Br /Br3 electrolyte. J Phys Chem B 2005;109:22449–55. [146)] Oskam G, Bergeron BV, Meyer GJ, Searson PC. Pseudohalogens for dye-sensitized TiO2 photoelectrochemical cells. J Phys Chem B 2001;105:6867–73. [147] Howie WH, Claeyssens F, Miura H, Peter LM. Characterization of solid-state dye-sensitized solar cells utilizing high absorption coefficient metal-free organic dyes. J Am Chem Soc 2008;130:1367–75. [148] Snaith HJ, Moule AJ, Klein C, Meerholz K, Friend RH, Grätzel M. Efficiency enhancements in solid-state hybrid solar cells via reduced charge recombination and increased light capture. Nano Lett 2007;7:3372–6. [149] Murakami TN, Grätzel M. Counter electrodes for DSC: application of functional materials as catalysts. Inorg Chim Acta 2008;361:572–80. [150] Papageorgiou N. Counter-electrode function in nanocrystalline photoelectrochemical cell configurations. Coord Chem Rev 2004;248:1421–46. [151] Papageorgiou N, Maier WF, Grätzel M. An iodine/triiodide reduction electrocatalyst for aqueous and organic media. J Electrochem Soc 1997;144:876–84. [152] Jiang QW, Li GR, Gao XP. Highly ordered tin nanotube arrays as counter electrodes for dye-sensitized solar cells. Chem Commun 2009;6720–2. doi:10.1039/B912776C. [153] Lee KS, Lee HK, Wang DH, et al. Dye-sensitized solar cells with Pt- and TCO-free counter electrodes. Chem. Commun. 2010;46:4505–7. [154] Wang S-F, Rao KK, Yang TCK, Wang H-P. Investigation of nitrogen doped diamond like carbon films as counter electrodes in dye sensitized solar cells. J Alloys Compd 2011;509:1969–74. [155] Wang MK, Anghel AM, Marsan B, et al. CoS supersedes Pt as efficient electrocatalyst for triiodide reduction in dye-sensitized solar cells. J Am Chem Soc 2009;131:15976. [156] Ikegami M, Miyoshi K, Miyasaka T, et al. Platinum/titanium bilayer deposited on polymer film as efficient counter electrodes for plastic dye-sensitized solar cells. Appl Phys Lett 2007;90:153122. [157] Song MY, Chaudhari KN, Park J, et al. High efficient pt counter electrode prepared by homogeneous deposition method for dye-sensitized solar cell. Appl Energy 2012;100:132–7. [158] Shakeel Ahmad M, Pandey AK, Abd Rahim N. Advancements in the development of TiO2 photoanodes and its fabrication methods for dye sensitized solar cell (DSSC) applications: a review. Renew Sustain Energy Rev 2017;77:89–108. [159] Wu J, Tang Z, Huang Y, Huang M, Yu H, Lin J. A dye-sensitized solar cell based on platinum nanotube counter electrode with efficiency of 9.05%. J Power Sources 2014;257:84–9. [160] Kim J, Kang J, Jeong U, Kim H, Lee H. Catalytic, conductive, and transparent platinum nanofiber webs for FTO-free dye-sensitized solar cells. ACS Appl Mater Interfaces 2013;5:3176–81. [161] Hsieh T-L, Chen H-W, Kung C-W, Wang C-C, Vittal R, Ho K-C. A highly efficient dye-sensitized solar cell with a platinum nanoflowers counter electrode. J Mater Chem 2012;22:5550–9. [162] Jeong H, Pak Y, Hwang Y, et al. Enhancing the charge transfer of the counter electrode in dye-sensitized solar cells using periodically aligned platinum nanocups. Small 2012;8:3757–61. [163] Dong P. Vertically aligned carbon nanotubes/graphene hybrid electrode as a TCO- and Pt-free flexible cathode for application in solar cells. J Mater Chem B 2014;2:20902. [164] Heo SY. Three-dimensional conducting polymer films for Pt-free counter electrodes in quasi-solid-state dye-sensitized solar cells. Electrochim Acta 2014;137:34.

178

Dye-Sensitized Materials

[165] Huang YJ. Mose 2 nanosheet/poly(3,4-ethylenedioxythiophene): poly(styrenesulfonate) composite film as a Pt-free counter electrode for dye-sensitized solar cells. Electrochim Acta 2016;211:794. [166] Park JT. High performance electrocatalyst consisting of CoS nanoparticles on an organized mesoporous SnO2 film: its use as a counter electrode for Pt-free, dye-sensitized solar cells. Nanoscale 2015;7:670. [167] Wu MX, Lin X, Wang YD, et al. Economical Pt-free catalysts for counter electrodes of dye-sensitized solar cells. J Am Chem Soc 2012;134:3419–28. [168] Yue G, Lin J-Y, Tai S-Y, Xiao Y, Wu J. A catalytic composite film of MoS2/graphene flake as a counter electrode for Pt-free dye-sensitized solar cells. Electrochim Acta 2012;85:162–8. [169] Yun S, Hagfeldt A, Ma T. Pt-free counter electrode for dye-sensitized solar cells with high efficiency. Adv Mater 2014;26:6210–37. [170] Zheng M. An in situ polymerized PEDOT/Fe3O4 composite as a Pt-free counter electrode for highly efficient dye sensitized solar cells. RSC Adv 2016;6:1637. [171] Zheng M, Huo J, Tu Y, Wu J, Hu L, Dai S. Flowerlike molybdenum sulfide/multi-walled carbon nanotube hybrid as Pt-free counter electrode used in dye-sensitized solar cells. Electrochim Acta 2015;173:252–9. [172] Kay A, Gratzel M. Low cost photovoltaic modules based on dye sensitized nanocrystalline titanium dioxide and carbon powder. Sol Energy Mater Sol Cells 1996;44:99–117. [173] Suzuki K, Yamaguchi M, Kumagai M, Yanagida S. Application of carbon nanotubes to counter electrodes of dye-sensitized solar cells. Chem Lett 2003;32:28–9. [174] Huang Z, Liu X, Li K, et al. Application of carbon materials as counter electrodes of dye-sensitized solar cells. Electrochem Commun 2007;9:596–8. [175] Saito Y, Kitamura T, Wada Y, Yanagida S. Application of poly(3,4-ethylenedioxythiophene) to counter electrode in dye-sensitized solar cells. Chem Lett 2002;1060–1. [176] Muto T, Ikegami M, Kobayashi K, Miyasaka T. Conductive polymer-based mesoscopic counterelectrodes for plastic dye-sensitized solar cells. Chem Lett 2007;36:804–5. [177] Saito Y, Kubo W, Kitamura T, Wada Y, Yanagida S. I /I3 redox reaction behavior on poly(3,4-ethylenedioxythiophene) counter electrode in dye-sensitized solar cells. J Photochem Photobiol A 2004;164:153–7. [178] Xia J, Masaki N, Jiang K, Yanagida S. The influence of doping ions on poly(3,4-ethylenedioxythiophene) as a counter electrode of a dye-sensitized solar cell. J Mater Chem 2007;17:2845–50. [179] Novoselov KS, Geim AK, Morozov SV, et al. Two-dimensional gas of massless dirac fermions in graphene. Nature 2005;438:197–200. [180] Bonaccorso F, Sun Z, Hasan T, Ferrari AC. Graphene photonics and optoelectronics. Nat Photon 2010;4:611–22. [181] Geim AK. Graphene: status and prospects. Science 2009;324:1530–4. [182] Bae S, Kim H, Lee Y, et al. Roll-to-roll production of 30-inch graphene films for transparent electrodes. Nat Nanotech 2010;5:574–8. [183] Di CA, Wei D, Yu G, Liu Y, Guo Y, Zhu D. Patterned graphene as source/drain electrodes for bottom-contact organic field-effect transistors. Adv Mater 2008;20:3289–93. [184)] Schedin F, Geim AK, Morozov SV, et al. Detection of individual gas molecules adsorbed on graphene. Nat Mater 2007;6:652–5. [185] Wang H, Hu YH. Graphene as a counter electrode material for dye-sensitized solar cells. Energy Environ Sci 2012;5:8182–8. [186] Hsieh C-T, Yang B-H, Lin J-Y. One- and two-dimensional carbon nanomaterials as counter electrodes for dye-sensitized solar cells. Carbon 2011;49:3092–7. [187] Choi H, Kim H, Hwang S, Han Y, Jeon M. Graphene counter electrodes for dye-sensitized solar cells prepared by electrophoretic deposition. J Mater Chem 2011;21:7548–51. [188] Roy-Mayhew JD, Bozym DJ, Punckt C, Aksay IA. Functionalized graphene as a catalytic counter electrode in dye-sensitized solar cells. ACS Nano 2010;4:6203–11. [189] Wang G, Zhuo S, Xing W. Graphene/polyaniline nanocomposite as counter electrode of dye-sensitized solar cells. Mater Lett 2012;69:27–9. [190] Yang X, Qiu L, Cheng C, Wu Y, Ma Z-F, Li D. Ordered gelation of chemically converted graphene for next-generation electroconductive hydrogel films. Angew Chem Int Ed 2011;50:7325–8. [191] Yu D, Nagelli E, Du F, Dai L. Metal-free carbon nanomaterials become more active than metal catalysts and last longer. J Phys Chem Lett 2010;1:2165–73. [192] Wang WW, Dang JS, Zhao X, Nagase S. Formation mechanisms of graphitic-N: oxygen reduction and nitrogen doping of graphene oxides. J Phys Chem C 2016;120:5673–81. [193] Wu J, Ma L, Yadav RM, et al. Nitrogen-doped graphene with pyridinic dominance as a highly active and stable electrocatalyst for oxygen reduction. ACS Appl Mater Interfaces 2015;7:14763–9. [194] Wang S, Iyyamperumal E, Roy A, Xue Y, Yu D, Dai L. Vertically aligned BCN nanotubes as efficient metal-free electrocatalysts for the oxygen reduction reaction: a synergetic effect by co-doping with boron and nitrogen. Angew Chem Int Ed 2011;50:11756–60. [195] Wang S, Zhang L, Xia Z, et al. BCN graphene as efficient metal-free electrocatalyst for the oxygen reduction reaction. Angew Chem 2012;124:4285–8. [196] Kannan AG, Zhao J, Jo SG, Kang YS, Kim DW. Nitrogen and sulfur co-doped graphene counter electrodes with synergistically enhanced performance for dye-sensitized solar cells. J Mater Chem A 2014;2:12232–9. [197] Chen L, Cui X, Wang Y, et al. One-step synthesis of sulfur doped graphene foam for oxygen reduction reactions. Dalton Trans 2014;43:3420–3. [198] Zhang Y, Chu M, Yang L, et al. Synthesis and oxygen reduction properties of three-dimensional sulfur-doped graphene networks. Chem Commun 2014;50:6382–5. [199] Yang W, Ma X, Xu X, et al. Sulfur-doped porous carbon as metal-free counter electrode for high-efficiency dye-sensitized solar cells. J Power Sources 2015;282:228–34. [200] Liang J, Jiao Y, Jaroniec M, Qiao SZ. Sulfur and nitrogen dual-doped mesoporous graphene electrocatalyst for oxygen reduction with synergistically enhanced performance. Angew Chem Int Ed 2012;51:11496–500. [201] Atchudan R. Green synthesis of nitrogen-doped graphitic carbon sheets with use of prunus persica for supercapacitor applications. Appl Surf Sci 2017;393:276. [202] Chen PY. PEDOT-decorated nitrogen-doped graphene as the transparent composite film for the counter electrode of a dye-sensitized solar cell. Nano Energy 2015;12:374. [203] Xue YH, Liu J, Chen H, et al. Nitrogen-doped graphene foams as metal-free counter electrodes in high-performance dye-sensitized solar cells. Angew Chem Int Ed 2012;51:12124–7. [204] Wang H, Sun K, Tao F, Stacchiola DJ, Hu YH. 3D honeycomb-like structured graphene and its high efficiency as a counter-electrode catalyst for dye-sensitized solar cells. Angew Chem Int Ed 2013;52:9210–4. [205] Wang H, Wei W, Hu YH. Tuning shape of three dimensional graphene sheets. Catal Today 2016;274:99–102. [206] Wei W, Chang L, Sun K, et al. The bright future for electrode materials of energy devices: highly conductive porous Na-embedded carbon. Nano Lett 2016;16:8029–33. [207] Wei W, Hu YH. Synthesis of carbon nanomaterials for dye-sensitized solar cells. Int J Energy Res 2015;39:842–50. [208] Wei W, Sun K, Hu YH. Synthesis of 3d cauliflower-fungus-like graphene from CO2 as a highly efficient counter electrode material for dye-sensitized solar cells. J Mater Chem A 2014;2:16842–6. [209] Wei W, Sun K, Hu YH. Direct conversion of CO2 to 3D graphene and its excellent performance for dye-sensitized solar cells with 10% efficiency. J Mater Chem A 2016;4:12054–7. [210] Wei W, Sun K, Hu YH. Synthesis of mesochannel carbon nanowall material from CO2 and its excellent performance for perovskite solar cells. Ind Eng Chem Res 2017;56:1803–9. [211] Nam JG, Park YJ, Kim BS, Lee JS. Enhancement of the efficiency of dye-sensitized solar cell by utilizing carbon nanotube counter electrode. Scr Mater 2010;62:148–50. [212] Lee WJ, Ramasamy E, Lee DY, Song JS. Efficient dye-sensitized cells with catalytic multiwall carbon nanotube counter electrodes. ACS Appl Mater Interfaces 2009;1:1145–9. [213] Ramasamy E, Lee WJ, Lee DY, Song JS. Spray coated multi-wall carbon nanotube counter electrode for tri-iodide (I3-) reduction in dye-sensitized solar cells. Electrochem Commun 2008;10:1087–9. [214] Zhang D, Li X, Chen S, Sun Z, Yin XJ, Huang S. Performance of dye-sensitized solar cells with various carbon nanotube counter electrodes. Microchim Acta 2011;174:73.

Dye-Sensitized Materials

179

[215] Choi H, Kim H, Hwang S, Choi W, Jeon M. Dye-sensitized solar cells using graphene-based carbon nano composite as counter electrode. Sol Energy Mater Sol Cells 2011;95:323–5. [216] Choi H, Kim H, Hwang S, Kang M, Jung D-W, Jeon M. Electrochemical electrodes of graphene-based carbon nanotubes grown by chemical vapor deposition. Scr. Mater 2011;64:601–4. [217] Zhu G, Pan L, Lu T, Xu T, Sun Z. Electrophoretic deposition of reduced graphene-carbon nanotubes composite films as counter electrodes of dye-sensitized solar cells. J Mater Chem 2011;21:14869–75. [218] Wei W, Sun K, Hu YH. An efficient counter electrode material for dyesensitized solar cells-flower-structured 1T metallic phase MoS2. J Mater Chem A 2016;4:12398–401. [219] Olsen E, Hagen G, Lindquist SE. Dissolution of platinum in methoxy propionitrile containing liI/I2. Sol Energy Mater Sol Cells 2000;63:267–73. [220] Mulmudi HK, Batabyal SK, Rao M, et al. Solution processed transition metal sulfides: application as counter electrodes in dye sensitized solar cells (DSSC). Phys Chem Chem Phys 2011;13:19307–9. [221] Wu MX, Wang YD, Lin X, et al. Economical and effective sulfide catalysts for dye-sensitized solar cells as counter electrodes. Phys Chem Chem Phys 2011;13:19298–301. [222] Bi H, Zhao W, Sun S, et al. Graphene films decorated with metal sulfide nanoparticles for use as counter electrodes of dye-sensitized solar cells. Carbon 2013;61:116–23. [223] Wu M, Lin X, Hagfeldt A, Ma T. A novel catalyst of WO2 nanorod for the counter electrode of dye-sensitized solar cells. Chem Commun 2011;47:4535–7. [224] Lin X, Wu M, Wang Y, Hagfeldt A, Ma T. Novel counter electrode catalysts of niobium oxides supersede pt for dye-sensitized solar cells. Chem Commun 2011;47:11489–91. [225] Guai GH, Leiw MY, Ng CM, Li CM. Sulfur-doped nickel oxide thin film as an alternative to Pt for dye-sensitized solar cell counter electrodes. Adv Energy Mater 2012;2:334–8. [226] Yun S, Wang L, Guo W, Ma T. Non-Pt counter electrode catalysts using tantalum oxide for low-cost dye-sensitized solar cells. Electrochem Commun 2012;24:69–73. [227] Ganesan R, Lee JS. Tungsten carbide microspheres as a noble-metal-economic electrocatalyst for methanol oxidation. Angew Chem Int Ed 2005;44:6557–60. [228] Rosenbaum M, Zhao F, Schröder U, Scholz F. Interfacing electrocatalysis and biocatalysis with tungsten carbide: a high-performance, noble-metal-free microbial fuel cell. Angew Chem Int Ed 2006;45:6658–61. [229] Yun S, Wu M, Wang Y, et al. Pt-like behavior of high-performance counter electrodes prepared from binary tantalum compounds showing high electrocatalytic activity for dye-sensitized solar cells. ChemSusChem 2013;6:411–6. [230] Li GR, Song J, Pan GL, Gao XP. Highly Pt-like electrocatalytic activity of transition metal nitrides for dye-sensitized solar cells. Energy Environ Sci 2011;4:1680–3. [231] Wu M, Bai J, Wang Y, et al. High-performance phosphide/carbon counter electrode for both iodide and organic redox couples in dye-sensitized solar cells. J Mater Chem 2012;22:11121–7. [232] Joshi P, Xie Y, Ropp M, Galipeau D, Bailey S, Qiao Q. Dye-sensitized solar cells based on low cost nanoscale carbon/TiO2 composite counter electrode. Energy Environ Sci 2009;2:426–9. [233] Li GR, Wang F, Jiang QW, Gao XP, Shen PW. Carbon nanotubes with titanium nitride as a low-cost counter-electrode material for dye-sensitized solar cells. Angew Chem Int Ed 2010;49:3653–6. [234] Wu MX, Lin X, Wang L, et al. In situ synthesized economical tungsten dioxide imbedded in mesoporous carbon for dye-sensitized solar cells as counter electrode catalyst. J Phys Chem C 2011;115:22598–602. [235] Wu MX, Lin XA, Hagfeldt A, Ma TL. Low-cost molybdenum carbide and tungsten carbide counter electrodes for dye-sensitized solar cells. Angew Chem Int Ed 2011;50:3520–4. [236] Lin JY, Liao JH, Hung TY. A composite counter electrode of CoS/MWCNT with high electrocatalytic activity for dye-sensitized solar cells. Electrochem Commun 2011;13:977–80. [237] Wen ZH, Cui SM, Pu HH, et al. Metal nitride/graphene nanohybrids: general synthesis and multifunctional titanium nitride/graphene electrocatalyst. Adv Mater 2011;23:5445. [238] Ramasamy E, Jo C, Anthonysamy A, Jeong I, Kim JK, Lee J. Soft-template simple synthesis of ordered mesoporous titanium nitride-carbon nanocomposite for high performance dye-sensitized solar cell counter electrodes. Chem Mater 2012;24:1575–82. [239] Li GR, Wang F, Song J, Xiong FY, Gao XP. Tin-conductive carbon black composite as counter electrode for dye-sensitized solar cells. Electrochim Acta 2012;65:216–20. [240] Dou YY, Li GR, Song J, Gao XP. Nickel phosphide-embedded graphene as counter electrode for dye-sensitized solar cells. Phys Chem Chem Phys 2012;14:1339–42. [241] Song J. Synergistic effect of molybdenum nitride and carbon nanotubes on electrocatalysis for dye-sensitized solar cells. J Mater Chem 2012;22:20580–5. [242] Tai S-Y, Liu C-J, Chou S-W, Chien FS-S, Lin J-Y, Lin T-W. Few-layer MoS2 nanosheets coated onto multi-walled carbon nanotubes as a low-cost and highly electrocatalytic counter electrode for dye-sensitized solar cells. J Mater Chem 2012;22:24753–9. [243] Wu J, Yue G, Xiao Y, et al. Glucose aided preparation of tungsten sulfide/multi-wall carbon nanotube hybrid and use as counter electrode in dye-sensitized solar cells. ACS Appl Mater Interfaces 2012;4:6530–6. [244] Li Y, Wang H, Feng Q, Zhou G, Wang Z-S. Reduced graphene oxide–TaON composite as a high-performance counter electrode for Co(bpy)33 þ /2 þ -mediated dyesensitized solar cells. ACS Appl Mater Interfaces 2013;5:8217–24. [245] Li ZQ, Gong F, Zhou G, Wang ZS. NiS2/reduced graphene oxide nanocomposites for efficient dye-sensitized solar cells. J Phys Chem C 2013;117:6561–6. [246] Li Y, Feng QY, Wang H, Zhou G, Wang ZS. Reduced graphene oxide–Ta3N5 composite: a potential cathode for efficient Co(bpy)(3)(3 þ /2 þ ) mediated dye-sensitized solar cells. J Mater Chem A 2013;1:6342–9. [247] Yue G, Wu J, Xiao Y, Lin J, Huang M, Lan Z. Application of poly(3,4-ethylenedioxythiophene):polystyrenesulfonate/polypyrrole counter electrode for dye-sensitized solar cells. J Phys Chem C 2012;116:18057–63. [248] Yue G, Wu J, Xiao Y, Huang M, Lin J, Lin J-Y. High performance platinum-free counter electrode of molybdenum sulfide-carbon used in dye-sensitized solar cells. J Mater Chem A 2013;1:1495–501. [249] Lin JY, Yue GT, Tai SY, et al. Hydrothermal synthesis of graphene flake embedded nanosheet-like molybdenum sulfide hybrids as counter electrode catalysts for dyesensitized solar cells. Mater Chem Phys 2013;143:53–9. [250] Lin JY, Chan CY, Chou SW. Electrophoretic deposition of transparent MoS2-graphene nanosheet composite films as counter electrodes in dye-sensitized solar cells. Chem Commun 2013;49:1440–2. [251] Xiao Y, Wu J, Lin J, et al. A high performance Pt-free counter electrode of nickel sulfide/multi-wall carbon nanotube/titanium used in dye-sensitized solar cells. J Mater Chem A 2013;1:13885–9. [252] Xiao Y, Wu J, Lin J-Y, Tai S-Y, Yue G. Pulse electrodeposition of CoS on MWCNT/Ti as a high performance counter electrode for a Pt-free dye-sensitized solar cell. J Mater Chem A 2013;1:1289–95. [253] Yue G, Wu J, Lin J-Y, et al. A counter electrode of multi-wall carbon nanotubes decorated with tungsten sulfide used in dye-sensitized solar cells. Carbon 2013;55:1–9. [254] Yun SN, Zhang H, Pu HH, Chen JH, Hagfeldt A, Ma TL. Metal oxide/carbide/carbon nanocomposites: in situ synthesis, characterization, calculation, and their application as an efficient counter electrode catalyst for dye-sensitized solar cells. Adv Energy Mater 2013;3:1407–12. [255] Wu JH, Li QH, Fan LQ, et al. High-performance polypyrrole nanoparticles counter electrode for dye-sensitized solar cells. J Power Sources 2008;181:172–6. [256] Li QH, Wu JH, Tang QW, et al. Application of microporous polyaniline counter electrode for dye-sensitized solar cells. Electrochem Commun 2008;10:1299–302. [257] Muto T, Ikegami M, Miyasaka T. Polythiophene-based mesoporous counter electrodes for plastic dye-sensitized solar cells. J Electrochem Soc 2010;157:B1195–200.

180

Dye-Sensitized Materials

[258] Pringle JM, Armel V, MacFarlane DR. Electrodeposited PEDOT-on-plastic cathodes for dye-sensitized solar cells. Chem Commun 2010;46:5367–9. [259] Lee K-M, Hsu C-Y, Chen P-Y, Ikegami M, Miyasaka T, Ho K-C. Highly porous PProDOT-Et2 film as counter electrode for plastic dye-sensitized solar cells. Phys Chem Chem Phys 2009;11:3375–9. [260] Yeh M-H, Lee C-P, Lin L-Y, et al. A composite poly(3,3-diethyl-3,4-dihydro-2h-thieno-[3,4-b][1,4]-dioxepine) and Pt film as a counter electrode catalyst in dyesensitized solar cells. Electrochim Acta 2011;56:6157–64. [261] Ahmad S, Bessho T, Kessler F, et al. A new generation of platinum and iodine free efficient dye-sensitized solar cells. Phys Chem Chem Phys 2012;14:10631–9. [262] Veerender P, Saxena V, Jha P, et al. Free-standing polypyrrole films as substrate-free and Pt-free counter electrodes for quasi-solid dye-sensitized solar cells. Org Electron 2012;13:3032–9. [263] Peng S, Wu Y, Zhu P, Thavasi V, Mhaisalkar SG, Ramakrishna S. Facile fabrication of polypyrrole/functionalized multiwalled carbon nanotubes composite as counter electrodes in low-cost dye-sensitized solar cells. J Photochem Photobiol A 2011;223:97–102. [264] Weng B, Shepherd RL, Crowley K, Killard AJ, Wallace GG. Printing conducting polymers. Analyst 2010;135:2779–89. [265] Wang WY, Ting PN, Luo SH, Lin JY. Pulse-reversal electropolymerization of polypyrrole on functionalized carbon nanotubes as composite counter electrodes in dyesensitized solar cells. Electrochim Acta 2014;137:721–7. [266] Yue G, Zhang X, Wang L, et al. Highly efficient and stable dye-sensitized solar cells based on nanographite/polypyrrole counter electrode. Electrochim Acta 2014;129:229–36. [267] Luo J, Niu HJ, Wen HL, et al. Enhancement of the efficiency of dye-sensitized solar cell with multi-wall carbon nanotubes/polypyrrole composite counter electrodes prepared by electrophoresis/electrochemical polymerization. Mater Res Bull 2013;48:988–94. [268] He B, Tang Q, Luo J, Li Q, Chen X, Cai H. Rapid charge-transfer in polypyrrole-single wall carbon nanotube complex counter electrodes: improved photovoltaic performances of dye-sensitized solar cells. J Power Sources 2014;256:170–7. [269] Yue G, Wang L, Zhang X, et al. Fabrication of high performance multi-walled carbon nanotubes/polypyrrole counter electrode for dye-sensitized solar cells. Energy 2014;67:460–7. [270] Kang ET, Neoh KG, Tan KL. Polyaniline: a polymer with many interesting intrinsic redox states. Prog Polym Sci (Oxford) 1998;23:277–324. [271] Wang H, Feng Q, Gong F, Li Y, Zhou G, Wang Z-S. In situ growth of oriented polyaniline nanowires array for efficient cathode of Co(iii)/Co(ii) mediated dye-sensitized solar cell. J Mater Chem A 2013;1:97–104. [272] Lee K-M, Chen P-Y, Hsu C-Y, et al. A high-performance counter electrode based on poly(3,4-alkylenedioxythiophene) for dye-sensitized solar cells. J Power Sources 2009;188:313–8. [273] Saito Y, Kubo W, Kitamura T, Wada Y, Yanagida S. I /I3 redox reaction behavior on poly(3,4-ethylenedioxythiophene) counter electrode in dye-sensitized solar cells. J Photochem Photobiol A 2004;164:153–7. [274] Li Z, Ye B, Hu X, Ma X, Zhang X, Deng Y. Facile electropolymerized-PANI as counter electrode for low cost dye-sensitized solar cell. Electrochem Commun 2009;11:1768–71. [275] Chen JG, Wei HY, Ho KC. Using modified poly(3,4-ethylene dioxythiophene): poly(styrene sulfonate) film as a counter electrode in dye-sensitized solar cells. Sol Energy Mater Sol Cells 2007;91:1472–7. [276] Luo J, Niu H-j, Wu W-j, Wang C, Bai X-d, Wang W. Enhancement of the efficiency of dye-sensitized solar cell with multi-wall carbon nanotubes/polythiophene composite counter electrodes prepared by electrodeposition. Solid State Sci 2012;14:145–9. [277] Li C-T, Lee C-T, Li S-R, et al. Composite films of carbon black nanoparticles and sulfonated-polythiophene as flexible counter electrodes for dye-sensitized solar cells. J Power Sources 2016;302:155–63. [278] Xu H, Zhang X, Zhang C, et al. Nanostructured titanium nitride/PEDOT:PSS composite films as counter electrodes of dye-sensitized solar cells. ACS Appl Mater Interfaces 2012;4:1087–92. [279] Yeh M-H, Lin L-Y, Lee C-P, et al. A composite catalytic film of PEDOT:PSS/TiN-NPs on a flexible counter-electrode substrate for a dye-sensitized solar cell. J Mater Chem 2011;21:19021–9. [280] Wei W, Wang H, Hu YH. Unusual particle-size-induced promoter-to-poison transition of ZrN in counter electrodes for dye-sensitized solar cells. J Mater Chem A 2013;1:14350–7. [281] Sudhagar P, Nagarajan S, Lee Y-G, et al. Synergistic catalytic effect of a composite (CoS/PEDOT:PSS) counter electrode on triiodide reduction in dye-sensitized solar cells. ACS Appl Mater Interfaces 2011;3:1838–43. [282] Wang H, Wei W, Hu YH. Efficient ZnO-based counter electrodes for dye-sensitized solar cells. J Mater Chem A 2013;1:6622–8. [283] Wang H, Wei W, Hu YH. NiO as an efficient counter electrode catalyst for dye-sensitized solar cells. Top Catal 2014;57:607–11. [284] Wang H, Hu YH. Electro-catalytic role of insulator/conductor interface in MgO/PEDOT composite electrodes for dye-sensitized solar cells. Sci China Chem 2015;58:101. [285] Dai S, Weng J, Sui Y, et al. The design and outdoor application of dye-sensitized solar cells. Inorg Chim Acta 2008;361:786–91. [286] Service RF. Tricks for beating the heat help panels see the light. Science 2003;300:1219. [287] Ha M, Park J, Lee Y, Ko H. Triboelectric generators and sensors for self-powered wearable electronics. ACS Nano 2015;9:3421–7. [288] Fu Y, Wu H, Ye S, et al. Integrated power fiber for energy conversion and storage. Energy Environ Sci 2013;6:805–12. [289] Du P, Hu X, Yi C, et al. Self-powered electronics by integration of flexible solid-state graphene-based supercapacitors with high performance perovskite hybrid solar cells. Adv Funct Mater 2015;25:2420–7. [290] Xu C, Wang ZL. Compact hybrid cell based on a convoluted nanowire structure for harvesting solar and mechanical energy. Adv Mater 2011;23:873–7. [291] Xu C, Wang X, Wang ZL. Nanowire structured hybrid cell for concurrently scavenging solar and mechanical energies. J Am Chem Soc 2009;131:5866–72. [292] Yang Y, Zhang H, Zhu G, Lee S, Lin Z-H, Wang ZL. Flexible hybrid energy cell for simultaneously harvesting thermal, mechanical, and solar energies. ACS Nano 2013;7:785–90. [293] Liang Q, Yan X, Gu Y, et al. Highly transparent triboelectric nanogenerator for harvesting water-related energy reinforced by antireflection coating. Sci Rep 2015;5:9080. [294] Wen Z, Yeh M-H, Guo H, et al. Self-powered textile for wearable electronics by hybridizing fiber-shaped nanogenerators, solar cells, and supercapacitors. Sci Adv 2016;2:e1600097. [295] D’Ercole D, Dominici L, Brown TM, Michelotti F, Reale A, Carlo AD. Angular response of dye solar cells to solar and spectrally resolved light. Appl Phys Lett 2011;99:213301. [296] Huang L-M, Hu C-W, Liu H-C, Hsu C-Y, Chen C-H, Ho K-C. Photovoltaic electrochromic device for solar cell module and self-powered smart glass applications. Sol Energy Mater Sol Cells 2012;99:154–9. [297] Zhou H, Song T-B, Chung C-H, et al. Solution-processed TiO2 nanoparticles as the window layer for CuIn(S,Se)2 devices. Adv Energy Mater 2012;2:1368–74. [298] Lopez-Lopez C, Colodrero S, Calvo ME, Miguez H. Angular response of photonic crystal based dye sensitized solar cells. Energy Environ Sci 2013;6:1260–6. [299] Xie Z, Jin X, Chen G, Xu J, Chen D, Shen G. Integrated smart electrochromic windows for energy saving and storage applications. Chem Commun 2014;50:608–10. [300] Wang K, Wu H, Meng Y, Zhang Y, Wei Z. Integrated energy storage and electrochromic function in one flexible device: an energy storage smart window. Energy Environ Sci 2012;5:8384–9. [301] Wei D, Scherer MRJ, Bower C, Andrew P, Ryhänen T, Steiner U. A nanostructured electrochromic supercapacitor. Nano Lett 2012;12:1857–62. [302] Thakur VK, Ding G, Ma J, Lee PS, Lu X. Hybrid materials and polymer electrolytes for electrochromic device applications. Adv Mater 2012;24:4071–96. [303] Fakharuddin A, Jose R, Brown TM, Fabregat-Santiago F, Bisquert J. A perspective on the production of dye-sensitized solar modules. Energy Environ Sci 2014;7:3952–81.

Dye-Sensitized Materials

181

[304] Yanagida M, Onozawa-Komatsuzaki N, Kurashige M, Sayama K, Sugihara H. Optimization of tandem-structured dye-sensitized solar cell. Sol Energy Mater Sol Cells 2010;94:297–302. [305] Yamaguchi T, Uchida Y, Agatsuma S, Arakawa H. Series-connected tandem dye-sensitized solar cell for improving efficiency to more than 10%. Sol Energy Mater Sol Cells 2009;93:733–6. [306] Dürr M, Bamedi A, Yasuda A, Nelles G. Tandem dye-sensitized solar cell for improved power conversion efficiencies. Appl Phys Lett 2004;84:3397–9.

2.7 Porous Materials Seda Keskin Avci, Koc University, Istanbul, Turkey Ilknur Erucar, Ozyegin University, Istanbul, Turkey r 2018 Elsevier Inc. All rights reserved.

2.7.1 Introduction 2.7.2 Background 2.7.3 Methods to Quantify CH4 Storage in Metal Organic Frameworks 2.7.3.1 Experimental Methods 2.7.3.2 Computational Methods 2.7.4 Assessment of CH4 Storage Performances of Metal Organic Frameworks 2.7.4.1 Experimental Studies 2.7.4.2 Computational Studies 2.7.5 Case Studies 2.7.5.1 Case Study 1: CH4 Storage for Transportation 2.7.5.2 Case Study 2: A Thermodynamic Tank Model for Natural Gas Storage in Metal Organic Frameworks 2.7.5.3 Case Study 3: Experimental Investigation of MIL-101(Cr) for CH4 Storage 2.7.5.4 Case Study 4: High-Throughput Screening of Metal Organic Frameworks 2.7.5.5 Case Study 5: Hydrogen Enriched Methane 2.7.6 Future Directions 2.7.7 Conclusions References Further Reading Relevant Websites

Nomenclature Ev Eg mexc ms mst N NCH4 ðvÞ NH2 ðvÞ NCH4 ðgÞ NH2 ðgÞ p p0 Qst r R SH2 =CH4 T U(r) Uangle-bend Ubonded Ubond-strecth Uelectrostatic ULJ

Total volumetric energy density Total gravimetric energy density Excess adsorption Total mass of sample in the system Total mass of gas stored in an ANG system Number of molecules Volumetric gas uptake of methane Volumetric gas uptake of hydrogen Gravimetric gas uptake of methane Gravimetric gas uptake of hydrogen Pressure Vapor pressure of the adsorbed gas Isosteric heat of adsorption Distance between molecule pairs Ideal gas constant Hydrogen adsorption selectivity over methane Temperature Total potential energy of the system Internal potential energy Intramolecular energy Bond stretching potential Electrostatic energy between polar particles LJ potential between two particles

Abbreviations Å AC

182

Angstrom Activated carbon

183 184 186 186 189 190 190 192 193 193 196 196 197 199 200 201 202 203 203

〈Nads〉 e m s DCH4 D H2

Intermolecular energy Torsional potential Energy caused by the weak van der Waals forces Molar fractions of species in the adsorbed phase Molar fractions of species in the bulk phase Volume Total volume accessible to the gas Volume adsorbed at pressure, p, and at absolute temperature, T Pore volume Volume of the solid Volumetric storage capacity Bulk density Gas density as a function of pressure (p) and temperature (T) Skeletal density Average potential energy of the adsorbed region Average number of particles in the system Energetic potential parameter Chemical potential LJ size parameter Combustion heat of methane Combustion heat of hydrogen

ANG ARPA-E BDC

Adsorbed natural gas Advanced Research Projects Agency-Energy Benzenedicarboxylate

Unonbonded Utorsion Uvan der Waals x y V V0 Vads Vp Vs Vst rbulk r0(p,T) rs 〈Uads〉

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00218-2

Porous Materials

[Bmim][BF4] 1-butyl-3-methylimidazolium tetrafluoroborate [Bmim][Tf2N] 1-butyl-3-methylimidazolium bis (trifluoromethyl-sulfonyl) imide BTB Polycarboxylate C Excess gravimetric CH4 uptake CCDC Cambridge Crystallographic Data Centre Methane CH4 cm Centimeter CNG Compressed natural gas CO Carbon monoxide CO2 Carbon dioxide COF Covalent organic frameworks COM Center of mass CoRE Computation-ready experimental DOE The US Department of Energy DP Dominant pore diameter DTs Decision trees EMD Equilibrium molecular dynamics g Gram GCMC Grand canonical Monte Carlo gge Gallon gasoline equivalent H2 Hydrogen HEM Hydrogen enriched methane HKUST-1 Hong Kong University of Science and Technology-1 hMOFs Hypothetical MOFs

2.7.1

hZIFs IAST IL IRMOFs K kJ L LCD LJ LNG m MJ MLR MOF NPN ns PCN PET PP PPNs QSPR STP SVMs TraPPE UFF UTSA ZIF

183

Predicted ZIFs Ideal adsorbed solution theory Ionic liquid Isoreticular metal organic frameworks Kelvin Kilojoule Liter Largest cavity diameter Lennard-Jones Liquefied natural gas Meter Megajoule Multilinear regression Metal organic framework Nitoso polymer networks Nanosecond Porous coordination network Polyethylene terephthalate Porous polymer Porous polymer networks Quantitative structure–property relationship Standard temperature and pressure Nonlinear support vector machines Transferable potentials for phase equilibria Universal force field University of Texas at San Antonio Zeolite imidazolate framework

Introduction

In recent years, natural gas has gained considerable attention as a clean energy source due to its usage in transportation and electricity generation. For example, modern automobile manufacturing industry aims to replace gasoline and diesel with alternative fuels such as natural gas. Natural gas consists mainly of CH4 and it has many environmental advantages such as low emissions of sulfur oxides, carbon monoxide, and hydrocarbons compared to traditional energy sources. Combusted CH4 produces less amount of carbon dioxide (CO2) (13.5 kg C 10 9 J) compared to gasoline (18.9 kg C 10 9 J), diesel oil (19.7 kg C 10 9 J), and bituminous coal (23.8 kg C 10 9 J) [1]. However, natural gas has very low volumetric energy density (0.04 MJ L 1) at ambient temperature and pressure compared to gasoline (32.4 MJ L 1) [2]. In order to replace gasoline with another type of fuel, the candidate fuel must have high energy density. This is required because if the energy density is high, more energy can be transported considering the same amount of volume. Otherwise, very large, high-pressure cylindrical tanks are required for on-board gas storage [3]. For this reason, four different technologies have been considered to increase the volumetric energy density of natural gas to implement natural gas vehicles: liquefied natural gas (LNG), compressed natural gas (CNG), natural gas hydrate (NGH), and adsorbed natural gas (ANG) [4]. The LNG process requires cooling of CH4 to a liquid at 111K. Cryogenic fuel tanks, which are costly and dangerous, are required for LNG. The CNG process, on the other hand, is based on compression of natural gas to 200 bar at ambient temperature using multistage compressors. Cylindrical or spherical tanks are required for the CNG technique. This system is also highly costly and bulky which hinders its implementation in the vehicles [5]. NGH is also a noneconomic and inefficient way to store natural gas because it requires rigorous hydrate formation conditions such as low temperature, high pressure, sufficient amount of water, and low hydrate formation rate. It also has the difficulty in terms of releasing the stored natural gas using pressure difference as a driving force [4]. Among these four different technologies, ANG is a widely used one to storage CH4 molecules within the porous materials at room temperature and at much lower pressures compared to CNG. Due to the physical interactions such as van der Waals between CH4 and a nanoporous material, gas molecules are adsorbed within the pores of the material. This nanoporous material is generally called as adsorbent whereas the CH4 molecules are referred as adsorbates. The low energy density problem of the natural gas can be resolved using this technique to store CH4 under ambient conditions. ANG technology requires lighter and more comfortable fuel tanks compared to CNG and LNG systems. Therefore, ANG systems can be competitive to gasoline or diesel fuel due to their low capital and operating cost. With the optimal choice of the adsorbent material, natural gas can be directly used for natural gas vehicles. The search for a promising adsorbent material for ANG technology has been an active area of research today. A promising adsorbent which possesses high gas storage capacity can reduce the cost of the storage and refueling systems.

184

Porous Materials

Metal organic frameworks (MOFs) have appeared as fascinating nanoporous materials with a large variety of applications in the energy area in the last decade. MOFs have been widely studied and tested as adsorbent materials for the storage of CH4. Several experimental and computational studies as we will discuss in detail below suggested that MOFs can outperform traditional adsorbent materials such as zeolites, carbons, and silica gels in terms of CH4 storage capacities. This is attributed to the very large surface areas and high porosities of MOFs compared to traditional nanoporous materials. The aim of this chapter is to provide a contemporary look at the current state of the art in the use of MOFs for CH4 storage. First, an informative background on the natural gas storage using nanoporous materials was given. MOFs were then introduced and then the fundamentals which are required to assess the natural gas storage performances of MOFs were addressed. Both experimental and computational techniques to determine CH4 storage capacities of MOFs were explained. Recent studies on assessing CH4 storage performances of MOFs were reviewed and several case studies were discussed in detail to develop further understanding of potential applications of MOFs. Finally, future research directions in using MOFs as potential adsorbents for natural gas storage were provided.

2.7.2

Background

Identification of the porous materials that can efficiently store CH4 is needed for the ANG technology. The US Department of Energy (DOE) reported the Advanced Research Projects Agency-Energy (ARPA-E) target to compare the adsorbed CH4 amounts of different adsorbents. In 1993, the DOE defined the initial target as 150 cm3 (STP: standard temperature and pressure) cm 3, which is the volume of gas adsorbed per volume of the storage vessel, at 298K and 35 bar. In 2000, the target was increased to 180 cm3(STP) cm 3 at 298K and 35 bar to compare the energy density of ANG with that of CNG [6]. Recently, the new target was set to 350 cm3(STP) cm 3 and a 25% loss in the volumetric capacity was considered because of the packing and pelletization of the adsorbent in a fuel tank [2]. Searching for the high performance adsorbents to meet the new DOE target for CH4 storage has attracted wide interest in recent years. At that point, it is important to note that an adsorbent material having high CH4 storage capacity is not enough to commercialize the ANG technology. A promising adsorbent material should have both high storage and high deliverable capacity. The deliverable capacity is also known as working capacity in the literature. It represents the difference in the adsorbed gas amounts between the adsorption and regeneration (desorption) steps. The DOE target for the deliverable CH4 capacity was defined as 315 cm3(STP) cm 3 at a storage (adsorption) pressure of 65 bar and a delivery (desorption) pressure of 5.8 bar and ambient temperature [5]. In order to calculate working capacity of an adsorbent, generally 35 bar or 65 bar is used as the adsorption pressure because these pressures are in the range of operating pressures of single-stage and two-stage compressors [4]. As a desorption pressure 5 bar is used because natural gas powered combustion engines operates at this pressure. Schematic representation of working capacity of a porous adsorbent material is given in Fig. 1 in addition to representing unused CH4 under isothermal conditions. As can be seen from this figure, it is required to maximize CH4 working capacity to minimize the unused CH4 under isothermal conditions. Nanoporous materials that can store and deliver maximum amount of CH4 under ambient conditions are highly desired; therefore, numerous various types of materials have been investigated for CH4 storage and delivery to date. Different adsorbents such as zeolites, silica gels, and activated carbons (ACs) have been widely used to examine CH4 storage [7]. However, the feasibility studies of these materials for ANG technology showed that these materials have some structural drawbacks. 120

Total CH4 uptake (cm3 STP cm–3)

100

80 Working capacity

60

40

20

Unused CH4 Padsorption

Pdesorption 0 0

10

20

30

40

50

60

70

80

90

100

Pressure (bar) Fig. 1 Graphical representation of methane (CH4) uptake, CH4 working capacity and unused CH4 under isothermal conditions.

Porous Materials

1

8

13

2

3

9

4

5

10

14

6

7

11

15

185

12

16

Fig. 2 Examples for Isoreticular metal organic framework – IRMOF-n, n:1–16. Reprinted from Rowsell JLC, Yaghi OM. Metal-organic frameworks: a new class of porous materials. Microporous Mesoporous Mater 2004;73:3–14, with permission of the Elsevier.

Zeolites are crystalline aluminum silicates which have been widely used as adsorbents for gas storage and separation applications. Zeolites possess generally low surface area (o1000 m2 g 1) and they are extremely hydrophilic which reduces their CH4 adsorption capacity because of the adsorbed moisture. In other words, water molecules are generally adsorbed in the pores of zeolites and the number of adsorption sites available for CH4 adsorption decreases. The CH4 uptake capacity of zeolites is generally around 74–87 cm3 (STP) cm 3 at 35 bar and 274K [7]. Zeolite packing is also one of the obstacles which hinders the on-board fuel storage of vehicles [7]. ACs possess high CH4 uptake capacities (50–160 cm3(STP) cm 3) but they have low packing densities which decline their volumetric CH4 storage capacities [4]. Silica gels, on the other hand, have high packing densities, but their surface areas are relatively low (20–715 m2 g 1) which limit their CH4 uptake. A new group of nanoporous materials called metal organic frameworks (MOFs) has appeared as an alternative to traditional materials for CH4 storage. MOFs are crystalline and highly porous materials consisted of metal complexes such as Zn2 þ , Co2 þ , Ni2 þ , Cu2 þ , Cd2 þ , Fe2 þ , Mg2 þ , Al3 þ , and Mn2 þ and organic ligands such as benzenedicarboxylate (BDC), polycarboxylate (BTB), and imidazole. MOFs have been synthesized by systematically changing the metal complexes and organic linkers to obtain various chemical compositions, pore dimensions, and functionalities [8]. The synthesis approach refereed as building block approach which assembles rigid molecular building blocks to achieve prearranged structures was used to make MOFs [9]. Large families of isoreticular MOFs having the same topology with different pore sizes and functionalities have been synthesized. As an example, Fig. 2 represents a series of 16 IRMOFs (isoreticular MOFs) which have the same octahedral Zn–O–C clusters that exhibit a cubic framework. The pore sizes and fractional free volumes of these MOFs were reported up to 29 Å and 0.91, respectively, with the expansion of the network [10]. Changing the combinations of metals and organic linkers increased the probability to synthesize numerous materials, over 50,000 with various chemical compositions and physical properties [11]. In fact, one of the most important advantages of MOFs over other traditional porous materials is that a large variety of materials having different physical and chemical properties can be synthesized. Theoretically an infinite number of MOFs can be synthesized by combining different metals and organic linkers [12]. MOFs commonly have very large surface areas (500–6000 m2 g 1), high pore volumes (1–4 cm3 g 1), wide range of pore sizes (1–98 Å ), reasonable thermal and mechanical stabilities. These fascinating physical properties in addition to the direct control of the structures during synthesis make MOFs promising materials for many different applications such as gas storage, gas separation, catalysis, chemical sensing, nonlinearoptics/ferroelectricity, biomedical imaging, drug storage, and drug delivery [13]. Fig. 3 shows the quick growth of the number of publications on MOFs. As can be seen from Fig. 3, MOFs are currently a very hot research topic with the very high number of publications about 1500 in 2015. Most of these publications have focused on synthesis of new MOF materials having interesting physical, chemical, and structural properties. Among various applications listed above, gas storage received a significant interest. Due to the high porosities, large surface areas, and low densities of MOFs, storage of several gases, especially CH4, H2, and CO2 has been widely studied. In this chapter, we will specifically focus on studies related to CH4 storage with MOFs. Several excellent reviews are available in the literature addressing the potential of MOFs for H2 [14–18], and CO2 [19–21] storage. We will briefly mention about H2 and CO2 storage performances of MOFs before addressing and evaluating CH4 storage potential of MOFs in detail. Adsorption of H2 into nanoporous materials has been examined for on-board gas storage since it has almost triple energy density that of gasoline per mass unit and generates zero emission. Recent studies showed that MOFs have strong potential in H2 storage. For example, H2 saturation uptake of MOF-177 was reported as 11 wt% at about 100 atm and 77K [22]. It is important to note that H2 uptakes of the majority of MOFs measured at room temperature are below the DOE target (5.5 wt% by 2017) because

186

Porous Materials

1600

Number of publications

1400 1200 1000 800 600 400 200 2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

0

Years

Fig. 3 Number of publications featuring the term “metal organic framework (MOF)” in their titles. Available from Web of Science [accessed 23.08.16].

of the weak interactions between H2 molecules and the framework atoms [22]. However, at 77K the H2 storage capacities of MOFs were reported in the range 6–15 wt% which are above the DOE target [14]. Studies such as functionalization of MOFs continue to improve H2 storage capacity of MOFs under ambient temperatures. The capture and storage of CO2 using MOFs are also well studied. MOFs are highly promising adsorbents for CO2 storage because there is generally no chemical bond formation between gas molecules and the host which requires less energy for regeneration comparing to conventional methods such as absorption of CO2 in aqueous solutions of alkanolamines [19]. For example, Sumida et al. [20] revealed that the volumetric CO2 uptake capacity of MOF-177 is 320 cm3 (STP) cm 3 at 35 bar. This value is almost 9 times higher than the quantity stored in an empty tank at the same pressure and also higher than the capacity of the traditional, well-known zeolites such as zeolite 13X [20]. In another study, Schoedel et al. [21] reported that Mg-MOF-74 has the highest CO2 uptake (193 cm3 (STP) cm 3) at 1 bar and room temperature among a total of 125 MOFs including MOFs with open metal sites, without open metal sites and amine functionalized ones. The high CO2 uptake of Mg-MOF-74 was attributed to the strong interactions between CO2 molecules and the open metal sites of the adsorbent. At that point, it is useful to explain what the open metal site is. MOFs are synthesized using solvothermal or hydrothermal techniques. During the synthesis, solvents are used and these solvents may remain in the pores of a MOF. An activation procedure is generally required to remove the solvents from the MOF to make the pores available for guest (gas) molecules. If the solvents are coordinated to the metal atoms, once the solvents are removed, the metal atoms become accessible. These MOFs are referred as MOFs having open metal sites. Several MOFs with open metal sites are known to exhibit higher gas adsorption property compared to the ones without open metal sites. The CH4 uptake performance of different types of MOFs has been recently evaluated and some MOFs have been found as promising candidates. The highest CH4 working capacities were reported to be 153 and 200 cm3(STP) cm 3 at 35 and 80 bar, respectively, for a copper-based MOF, Hong Kong University of Science and Technology-1 (HKUST-1), also known as CuBTC in the literature) [23]. HKUST-1 is one of the most widely studied MOFs in the literature and it is already commercially available. However, the reported values for CH4 working capacities of HKUST-1 are much below the current limit which is required for on-board CH4 storage. Therefore, identification of new MOFs to meet the DOE’s target is a very important research topic. Recently, aluminum-based MOFs were synthesized and reported to exhibit higher CH4 working capacity (230 cm3(STP) cm 3 at 80 bar) compared to HKUST-1. Fig. 4 compares CH4 working capacities of aluminum-based MOFs (MOF-519 and MOF-520) with the porous carbon AX-21. Bulk CH4 data were also presented for comparison. Results show that using a porous material such as a MOF, working capacity can be enhanced dramatically. Newly synthesized aluminum-based MOF, MOF-519 can outperform HKUST-1 and AX-21 in terms of CH4 working capacity [23]. Current literature suggests that it may be difficult to reach the ARPA-E target for CH4 storage. However, MOFs have great potential as promising adsorbents for CH4 storage and both experimental and computational efforts are on the way to design new MOFs with better CH4 storage capacities and to enhance the CH4 storage performances of existing MOFs.

2.7.3

Methods to Quantify CH4 Storage in Metal Organic Frameworks

In this section, both experimental and computational methods used to quantify the CH4 storage performances of MOFs were briefly explained.

2.7.3.1

Experimental Methods

Adsorption is the enrichment of a solid surface with a gas component due to the van der Waals forces. The solid is called adsorbent and the adsorbed gas molecules are called adsorbates. The adsorbed gas amount is measured as a function of pressure at a constant

Porous Materials

250

187

80 bar

Working capacity (cm3 cm–3)

35 bar 200

150

100

Bulk CH4 80 bar

50 Bulk CH4 35 bar

74

0

Fi-M

N

M

O

O

F-

-2

21

1

5 O

M

AX

-1

F-

4

7 N

PC

17

5 M

O

F-

20

0 F-

52 O M

O

F-

ST M

KU H

M

O

F-

51

9

-1

0

Adsorbent

Vads

Fig. 4 Comparison of CH4 working capacities of some metal organic frameworks (MOFs) and the porous carbon AX-21. Working capacities were calculated at two different adsorption pressures: 35 bar (blue) and 80 bar (orange). Bulk CH4 data were also shown as a reference. Reprinted from Gándara F, Furukawa H, Lee S, Yaghi OM. High methane storage capacity in aluminum metal-organic frameworks. J Am Chem Soc 2014;136 (2014):5271–4, with permission of the American Chemical Society.

Type I Type II

Vads



Type III





0

0

Type IV

Type V



0

 P0

P0

Fig. 5 Comparison of five different types of adsorption isotherms. Reprinted from Brunauer S, Deming LS, Deming WE, Teller E. On a theory of the van der Waals adsorption of gases. J Am Chem Soc 1940;62:1723–32, with permission of the American Chemical Society.

temperature. The output of this measurement is a curve which is known as an adsorption isotherm. Adsorption isotherms are commonly used to investigate the maximum adsorbed amount of gas molecules in MOFs and they provide an understanding for the adsorption mechanism. For example, at low pressures the adsorbed gas amount increases linearly with the pressure which is explained by Henry’s law [24]. Five different types of adsorption isotherms are defined in Ref. [25]. These were illustrated in Fig. 5. In these figures, y-axis represents Vads, the volume adsorbed at pressure p and at absolute temperature T and x axis represents pressure where p0 is the vapor pressure of the adsorbed gas. Type I is known as Langmuir adsorption isotherm which assumes monolayer adsorption. Type II isotherm can be observed in nonporous or macroporous adsorbents. At low pressures monolayer adsorption occurs and then at high pressures more layers adsorb on the first layer. Type III isotherm occurs when the gas molecules preferentially bind on the adsorbed molecules due to the weak adsorbent–adsorbate interactions [24]. In type IV and type V isotherms hysteresis loops can be observed due to the capillary condensation in mesoporous materials.

188

Porous Materials

In order to evaluate the CH4 storage performance of MOFs, high-pressure adsorption experiments are required. The amount of adsorbed gas can be measured by using either the Sieverts apparatus based on the volumetric measurement method or microbalance based on the gravimetric method [4]. Fig. 6 has the representation of both volumetric and gravimetric methods. In volumetric approach, the change in pressure within a sealed system composed of the gas molecules and adsorbent is measured to determine the amount of adsorbed gas. The adsorbed gas amount is then estimated by using an equation of state to consider real gas behavior at the specified pressure and temperature. In the gravimetric approach, a microbalance is used to measure the mass of the adsorbed gas [26]. Both volumetric and gravimetric methods can be used to measure adsorption isotherms of CH4 in a MOF material. In the literature, absolute, excess, and total adsorption terms are generally used to describe CH4 adsorption in a MOF adsorbent. CH4 adsorption in MOFs is driven by the interplay of the weak dispersive forces between gas molecules and the adsorbent. The distance between adsorbates and surface of the adsorbent determines the strength of the interaction. As the separation distance between gas molecules and the surface of the adsorbent increases, attractive forces of the surface become unimportant and only free gas molecules are available in the system [2]. At that point, Gibbs dividing surface is described as illustrated in Fig. 7. There are two states localized by this surface: adsorbed gas region (green) or bulk gas region (blue). Experiments commonly use excess adsorption term because the accurate location of this red boundary is not known. Excess adsorption can be defined as the amount of adsorbed gas which is in contact with the adsorbent. Absolute adsorption can be defined as the amount of adsorbed gas which is in contact with both the adsorbent and bulk region. If the gas–solid interactions are negligible, then the presence of bulk gas molecules is assumed. In order to express the gas adsorption in MOFs, using a total adsorption term is more reasonable. As shown in Fig. 7(B), total adsorption is the sum of excess adsorption amount and the gas amount within the pore of the material.

Adsorptive

Vacuum

Vacuum

Microbalance

Pressure transducer

Adsorptive

Vacuum

Adsorptive Valve Calibrated valve train Pressure vessel Magnetic coupling

Adsorption vessel (A)

Adsorption vessel (B)

(C)

Fig. 6 (A) Simple volumetric apparatus, the valve train itself serves as the reference volume, (B) volumetric apparatus with separate pressure vessel to act as a reference volume, and (C) gravimetric instrument. Reprinted from Konstas K, Osl T, Yang Y, et al. Methane storage in metal organic frameworks. J Mater Chem 2012;22:16698–708, with permission of the Royal Society of Chemistry.

Adsorbed Bulk +

(A)

=

Excess

+

Excess

Absolute

Bulk

=

Pore volume × bulk denstiy

Total

(B)

Fig. 7 (A) Representation of the excess and absolute gas adsorption on a two-dimensional rectangular surface. (B) Total adsorption for porous adsorbents. Reprinted from Mason JA, Veenstra M, Long JR. Evaluating metal-organic frameworks for natural gas storage. Chem Sci 2014;5:32–51, with permission of the Royal Society of Chemistry.

Porous Materials

189

A thermodynamic variable, namely isosteric heat of adsorption, is commonly used in order to explain the CH4 adsorption mechanisms in MOFs. Isosteric heat of adsorption describes the average binding energy of an adsorbed molecule at a specific surface coverage and can be calculated by the interpolation of adsorption isotherms with respect to the temperature using the following equation [27]:
 ∂lnP Qst ¼ R ð1Þ ∂ð1=TÞ n In this equation, Qst is the differential enthalpy of adsorption which has a negative quantity due to the exothermic adsorption process, T is the temperature, and P is the pressure. The partial differentiation is performed at constant n, which is the adsorbed amount of gas (moles of gas per kilogram of solid). The absolute value of Qst is referred as isosteric heat of adsorption in Ref. [27]. Three different trends are generally observable related to Qst: (1) Qst can be constant with increasing adsorbate uptake for an energetically homogenous adsorbent, (2) Qst may decrease with an increase in adsorbate uptake due to energetic heterogeneity of the adsorbent and (3) Qst may increase with an increase in adsorbate uptake due to stronger lateral interactions (repulsive or attractive forces) between the adsorbed molecules [28]. To summarize, adsorption isotherm and isosteric heat of adsorption are used to examine CH4 adsorption mechanisms in MOFs.

2.7.3.2

Computational Methods

In addition to experimental methods, computational techniques have been widely used to predict CH4 storage capacities of nanoporous materials. The most well-known method is to use molecular simulations to estimate the amount of adsorbed gas molecules in an adsorbent. In statistical mechanics, a system’s macroscopic properties such as temperature and pressure can be described by the ensemble averages of all possible microstates. An ensemble is defined as an imaginary collection of a large number of microstates. The predictions of molecular simulations are dependent on the choice of the ensemble. In an experimental setup of gas adsorption, adsorbed gas is in equilibrium with the gas in the reservoir. The chemical potential (m) and the temperature (T) of the gas inside and outside of the adsorbent are assumed to be equal at equilibrium. Grand canonical Monte Carlo (GCMC) simulations are widely used to mimic an adsorption experiment and the grand-canonical ensemble (m, V, T) is used to predict the average number of adsorbed particles in the system [29]. In a typical GCMC simulation, MOF is the adsorbent and gas molecules are the adsorbates. The initial step in GCMC simulations is to specify the positions of MOF atoms. The crystal structures of MOFs are commonly obtained from Cambridge Crystallographic Data Center (CCDC) [11]. CCDC is a database which contains experimentally reported atomic coordinates for the organic and inorganic compounds. In this database, each MOF has a unique “refcode” (entry ID) consisting of six letters. The atomic coordinates of each element in the MOF, which are obtained from the X-ray diffraction experiments, are available under the refcodes. Once the crystal structure is ready, the reservoir applies constant chemical potential and temperature by exchanging particles and energy during the simulations. The number of adsorbed molecules in a MOF structure is allowed to fluctuate during the simulation based on trial moves such as particle displacement, particle insertion, and deletion. Each move has a certain probability that is related to its energy. When the energy of the particle is low, the acceptance probability is high whereas when the energy of the particle is high, the acceptance probability is low. The total potential energy of the system is calculated by using the following equation: UðrÞ ¼ Ubonded þ Unonbonded

ð2Þ

where U(r) is the total potential energy of the system which is dependent on the atomic coordinates (r), Ubonded is the intramolecular energy, and Unonbonded is the intermolecular energy. The bonded energy can be estimated as follows: Ubonded ¼ Ubond

strecth

þ Uangle

bend

þ Utorsion

ð3Þ

where Ubond-strecth is the bond stretching potential which depends on the bond length, Uangle-bend is the internal potential energy due to the bond bending, and Utorsion is the torsional potential which is dependent on the torsional angles between four pairs. In most GCMC simulations, it is reasonable to assume rigid framework; therefore, the intramolecular interactions (bonded) can be neglected [30]. The predicted CH4 adsorption isotherms were found to be in a good agreement with experimental measurements in several studies when intramolecular interactions are neglected [6,31,32]. The nonbonded energy can be described as follows: Unonbonded ¼ Uvan der waals þ Uelectrostatic

ð4Þ

Here, Uvan der Waals is the energy caused by the weak van der Waals forces and Uelectrostatic is the electrostatic energy between polar particles. Since CH4 is a nonpolar molecule, electrostatic energy is not considered in the molecular simulations of MOFs for CH4 uptake. The Lennard-Jones (LJ) 12–6 potential is commonly used to calculate the nonbonded interactions between two particles. The functional form of the LJ potential given in Eq. (5) is based on two parameters, s is the size parameter and e the energy parameter: "   6 # sij 12 sij ð5Þ ULJ ¼ 4eij rij rij In this equation, ULJ is the intermolecular potential between two particles (i and j), r is the distance of the center of masses of two particles, eij is the depth of the potential well and it describes how strongly the two particles attract each other and sij defines a

190

Porous Materials

molecular length scale based on the particle diameter and affects the scale of interaction. Parameters of the LJ function are commonly obtained from the generic force fields which describe the same types of atoms in all MOFs with the same parameter [33]. The universal force field (UFF) [34] and Dreiding [35] are the two generic force fields that are used as transferable force fields in molecular simulations of MOFs. The validity of these force fields is shown by several studies where adsorption of various gas molecules has been computed [36,37]. For van der Waals terms, the atom-based summation method is used and the interactions between all pair of atoms which are further beyond the cutoff radius are excluded. Periodic boundary conditions are used in molecular simulations to neglect the surface effects. Lorentz–Berthelot mixing rules are assumed to calculate cross-interactions between two dissimilar LJ sites [38,39]. CH4 molecules are generally represented as single spheres with united atom model in molecular simulations and transferable potentials for phase equilibria (TraPPE) force field [40] is widely used to model CH4 molecules. In order to compute adsorption isotherms using GCMC simulations, chemical potential (m) must be converted to the pressure (P). At low pressures, the interactions between gas molecules can be assumed to be negligible and therefore the fugacity and the pressure can be taken as equal. At high pressures, on the other hand, gases behave nonideally. Herein, the ideal gas law is not capable of predicting the real behavior of gas molecules in the system. Therefore, an equation of state such as the Peng–Robinson [41] or Benedict–Webb–Rubin [42] can be used. In some molecular simulations, “widom test particle insertion method” is used to calculate the chemical potential [43]. In this method, a ghost particle (a particle that has no interaction with the system) is added into the system. Based on the trial insertions of the ghost particle, the energy difference can be computed after each Monte Carlo step. Using the average energy change, the chemical potential of a species can be calculated. After a GCMC simulation, the number of adsorbed gas molecules as a function of pressure is obtained. These data give the adsorption isotherm of a gas. The isosteric heat of adsorption can be also calculated from the ensemble average fluctuations in a GCMC simulation [29]: Qst ¼ RT



〈Uads  Nads 〉 〈Uads 〉  〈Nads 〉 2 〉 〈Nads 〉  〈Nads 〉 〈Nads



ð6Þ

where Qst is the isosteric heat of adsorption, T is the temperature, R is the ideal gas constant, 〈Uads〉 is the average potential energy of the adsorbed region, and 〈Nads〉 is the average number of particles in the system. Both the adsorption isotherm and the isosteric heat of adsorption obtained from the molecular simulations are comparable to the ones measured in experiments. The readers are referred to the literature for more technical details of molecular simulations used to estimate and characterize gas adsorption in nanoporous materials [29].

2.7.4

Assessment of CH4 Storage Performances of Metal Organic Frameworks

In 2002, Eddaoudi et al. [8] synthesized an MOF, IRMOF-6 and its CH4 uptake capacity was reported as 240 cm3(STP) g 1 at 298K and 36 atm. IRMOF-6 showed much higher CH4 storage capacity than other well-known porous materials such as zeolite 5A (87 cm3(STP) cm 3) and other coordination frameworks (up to 213 cm3(STP) g 1) [8]. After this pioneering work, a large family of MOFs was further investigated for CH4 storage [4]. In this chapter, both experimental and computational studies that examined MOFs for CH4 storage were briefly discussed. Empirical equations for predicting the CH4 storage capacities of MOFs and largescale computational studies to identify the CH4 storage performance limits of MOFs were also reviewed. Finally, strategies to improve CH4 storage capacities of MOF were addressed by providing recent examples from current applications.

2.7.4.1

Experimental Studies

Several types of MOFs such as IRMOFs, HKUST-1, porous coordination networks (PCNs), MOF-74, Nottingham University (NOTT) and University of Texas at San Antonio (UTSA ) have been identified as promising storage materials due to their high CH4 uptake capacities [4]. Mason et al. [2] and He et al. [4] reported experimental CH4 uptake data of various MOFs in their review articles. It can be seen from these two reviews that several different CH4 adsorption isotherm data are available for the same MOF material and CH4 uptake amounts of MOFs may alter depending on the activation procedure of the MOF and the measurement technique. For example, Ni-MOF-74 was recorded to have total CH4 uptake changing in a wide range between 208 and 230 cm3(STP) cm 3 at 35 bar and 298K [4]. The different CH4 uptake values recorded by researchers may be due to the different measurement method used in the experiments (gravimetric/volumetric), and/or differences in activation of the MOF samples before adsorbing CH4. In order to accurately assess and compare the CH4 storage performance of different types of MOFs, it is vital to use the same experimental methodology including activation methods and measurement techniques for high-pressure CH4 adsorption. He et al. [4] carried out a detailed literature search and collected 154 experimental CH4 adsorption data for 122 different types of MOFs together with MOFs’ crystal densities, pore volumes, and surface areas. The highest total volumetric CH4 storage capacity (267 cm3(STP) cm 3) was reported for HKUST-1 at 298K and 65 bar [44]. If the packing density is excluded, HKUST-1 can reach the DOE’s new volumetric target at 65 bar (263 cm3(STP) cm 3), which indicates that MOFs have strong potential to reach the desired target in CH4 storage. It is important to discuss that the packing density of a MOF alters CH4 uptake capacity of the adsorbents. If the packing density of a material is high, the total volumetric uptake can decrease with decrease in the porosity.

Porous Materials

191

For this reason, it is vital to consider the exact packing density of a MOF adsorbent for ANG technology. Similar to HKUST-1, NiMOF-74 was also reported to have high CH4 uptake capacity (260 cm3(STP) cm 3 at 298K and 65 bar) [2]. In these two MOFs, Cu2 þ and Ni2 þ are the favorable binding sites for gas adsorption and these sites are known as open metal sites as we discussed before. In order to compare the storage performances of adsorbent materials for practical ANG applications, both storage and delivery capacities of the materials must be considered. For example, Ni-MOF-74 has a comparable CH4 uptake capacity with HKUST-1, but it has lower delivery capacity (142 cm3(STP) cm 3) than HKUST-1 (190 cm3(STP) cm 3) at 298K (storage at 65 bar and delivery at 5 bar). This result can be attributed to the weaker interactions of CH4 molecules with the atoms of HKUST-1 than those of Ni-MOF-74 at the delivery (desorption) pressure. This result can be also directly seen when the values of isosteric heat of adsorption are considered. The isosteric heat of adsorption values of the two MOFs were experimentally reported at 25 1C as a function of the total amount of CH4 adsorbed in both materials. HKUST-1 has lower isosteric heat of adsorption value (17 kJ mol 1) than Ni-MOF-74 (20.6 kJ mol 1) which can be also attributed to the higher delivery capacity of HKUST-1. In order to increase the CH4 uptake of MOFs to meet the DOE target, several strategies such as introducing open metal sites and functionalization of the pores were tried. For example, Li et al. [45] incorporated functional groups including pyridine, pyridazine, and pyrimidine into the pores of NOTT-101 and reported that total volumetric CH4 uptake capacities increased from 237 cm3(STP) cm 3 to B249–257 cm3(STP) cm 3 at 65 bar and room temperature. This increase in CH4 uptake was attributed to the Lewis basic nitrogen sites and dynamic freedom of the functionalized linkers. It is also important to note that the functional groups did not affect the CH4 uptake at 5 bar, but they have significant influence on the CH4 uptake of the material at 65 bar. Therefore, the strategy of incorporation of functional groups can be used to improve the CH4 working capacities of MOFs. Similarly, Song et al. [46] developed a new organic linker and synthesized a NbO-type MOF which shows high CH4 storage and working capacity of 241 and 190 cm3(STP) cm 3, respectively, at 298K and 65 bar. Using the ligand modification, pore structure was optimized and CH4 uptake was enhanced at 65 bar due to the creation of small pore sizes and polarized pi-electron surface which favors CH4 interactions at higher loadings. Song et al. [46] pointed out that at 5 bar open copper sites and windows are favorable locations for the adsorbed gas molecules whereas at 65 bar secondary adsorption sites such as surface area of the adsorbent have significant effect on the CH4 adsorption. Another recent idea to improve gas uptake of MOFs is to incorporate ionic liquids into the pores of MOFs. Ionic liquids are molten salts at room temperature and several gases are known to be highly soluble in ionic liquids. However, ionic liquid incorporation into MOFs did not improve the CH4 storage in MOFs as expected. In a recent study, Ban et al. [47] incorporated an ionic liquid, namely, 1-butyl-3-methylimidazolium bis(trifluoromethyl-sulfonyl) imide ([Bmim][Tf2N] ) imide) into a widely studied, commercial MOF known as ZIF-8. They measured the adsorption isotherm of CH4 at room temperature and their results showed that the adsorbed amount of CH4 was decreased in ionic liquid-incorporated ZIF-8 due to the decrease in the pore volume of the MOF upon incorporation of the ionic liquid. Similarly, Sezginel et al. [48] incorporated an ionic liquid, 1-butyl-3-methylimidazolium tetrafluoroborate ([BMIM][BF4]) into CuBTC (HKUST-1) to tune its gas storage properties. They reported a decreasing trend with increasing pressure in the adsorption isotherm of CH4 in ionic liquidincorporated CuBTC. This was explained by the decrease in the surface area and the number of adsorption sites of CuBTC upon incorporation of an ionic liquid. All these studies reported a decrease in the CH4 adsorption of ionic liquid-functionalized MOFs. However, the gas affinities of ionic liquid-functionalized MOFs can be improved with the tunable physicochemical properties of MOF and ionic liquid pairs. Considering the huge number of available MOF and ionic liquid structures, there is still plenty of room for developments of CH4 uptake capacities of ionic liquid-incorporated MOFs. Another strategy to improve CH4 uptake of MOFs is to design flexible frameworks. Flexible MOFs have gate-opening behavior when they are exposed to an external stimuli such as pressure or temperature. Mason et al. [45] stated that if a flexible MOF exhibits a high CH4 storage capacity at 35–65 bar and collapses (transforms into a nonporous framework) to release all adsorbed gas molecules at almost 5.8 bar, then it can be possible to reach higher CH4 working capacity compared to that of rigid adsorbents. Mason et al. [45] used a flexible MOF, namely Co(bdp), to investigate its ANG potential and reported high CH4 working capacity, 155 cm3(STP) cm 3 at 35 bar and 197 cm3(STP) cm 3 at 65 bar at room temperature. Beside the working capacity of an adsorbent, the heat flowing throughout the adsorbent should be also considered for transportation. A large temperature differences between adsorption and desorption cycles may decrease the working capacity of the adsorbent. Therefore, Mason et al. [45] also investigated the differential adsorption enthalpy of flexible MOFs. They reported that flexible MOFs require a much lower differential adsorption enthalpy ( 8.4 kJ mol 1) than rigid adsorbents (from 15 to 25 kJ mol 1) which have high CH4 working capacities such as HKUST-1 since the expansion of the framework during adsorption is endothermic whereas the transition to the collapsed phase (nonporous framework) is exothermic. This result is very important for on-board gas storage because it suggests that flexible MOFs can provide cheaper and more efficient systems compared to traditional rigid adsorbents for vehicular applications. At that point, it is worth noting that experimental studies commonly have focused on only a limited number of MOFs because high-pressure adsorption measurements are not straightforward, in fact these experiments are challenging. First of all, at high pressures gas molecules behave nonideally and large errors are expected especially in free space calculations after the measurement [2]. Therefore, comparing the adsorption isotherms of even two identical MOFs may not give the same results at high pressures. Second issue is the impurities such as water in the gas adsorption system. Impurities can affect the adsorbed gas amount and mislead the results. Additionally, calibration techniques, sample mass measurements, thermocouple reading and pressure transducer reading may all affect the recorded adsorbed gas amount [2]. Experimental studies can be complemented by the computational approaches that provide adsorption isotherms of different gases under various operating conditions, even at high pressures and low temperatures that are difficult to reach within the experiments. Computational studies can also screen a large number of MOFs having different

192

Porous Materials

chemistry, topology, porosity, etc. rather than focusing on a single MOF. The recent advances in computational methods enabled highthroughput screening of a large number of different MOFs for natural gas storage as we will discuss below.

2.7.4.2

Computational Studies

Considering the high number of synthesized MOFs and challenges in experimental high-pressure CH4 adsorption measurements, it is difficult to measure CH4 uptake capacities of every single existing MOF using purely experimental manners. Investigating the relationship between the physical properties of MOFs such as pore volume, pore size, surface area, and CH4 uptake capacities of these materials is essential to identify the common properties of MOFs that have high CH4 storage performance. These relations are also useful to design and develop new materials with predetermined structural properties to reach extraordinarily high CH4 storage capacities. Therefore, substantial efforts have been devoted to develop empirical models to predict CH4 storage capacities of MOFs in the literature. He et al. [49] synthesized a series of activated MOFs, NOTT-100, NOTT-101, NOTT-102, NOTT-103, and NOTT-109, for which the crystal structures are shown in Fig. 8. These MOFs except NOTT-109 have the same topology (NbO-type) but the lengths of the bridging ligands vary. The delivery amount of CH4 for these MOFs was reported to be in the range 181–196 cm3(STP) cm 3 at 35 bar and 300K. It was shown that the gravimetric CH4 storage capacities (cm3(STP) g 1) of these MOFs are highly correlated with their pore volumes and the following model was proposed: C¼

126:69  Vp2 þ 381:62  Vp

ð7Þ

12:57

In Eq. (7), C is the excess gravimetric CH4 uptake of the MOF at 35 bar and 300K in cm3(STP) g 1, and Vp is the pore volume of the material in cm3 g 1. Predictions obtained from this model were compared with the experimentally reported data. The average deviation between experiments and model predictions was found to be around 8.1% indicating the validity of the suggested empirical model. CH4 storage capacities of porous materials are commonly given in volumetric units in the literature. For this reason, once the gravimetric CH4 uptake is predicted by using Eq. (7), it is possible to estimate the volumetric CH4 storage capacities as multiplication of the framework density and gravimetric CH4 storage capacity. In 2012, Wilmer et al. [50] generated and screened a database of 137,953 hypothetical MOFs. These MOFs are not synthesized, real MOFs but they were built using 102 building blocks by computational methods. They calculated CH4 uptake of these materials at 35 bar and 298K using GCMC simulations and they also estimated the volumetric surface area, the gravimetric surface area, and the void fraction of hypothetical MOFs using computational techniques. The structure–property relationship for volumetric CH4 storage was shown in Fig. 9. Results showed that volumetric CH4 storage density increased linearly with the volumetric surface area whereas decreased with the gravimetric surface area after a certain value. The optimal gravimetric surface area and void fraction to maximize the CH4 storage of MOFs were reported as 2500–3000 m2 g 1 and 0.8, respectively. Using the same hypothetical MOF database, Fernandez et al. [51] reported a large-scale, quantitative structure–property relationship (QSPR) analysis for MOFs. They investigated the effect of geometrical properties of MOFs, such as pore size and void fraction, on the CH4 storage capacities of B130,000 hypothetical MOFs at 1, 35, and 100 bar at 298K. Using data mining techniques such as multilinear regression (MLR) models, decision trees (DTs), and nonlinear support vector machines (SVMs), they offered several models

NOTT-100

NOTT-101

NOTT-103

NOTT-102

NOTT-109

Fig. 8 Schematic view of NOTT-100, NOTT-101, NOTT-103, NOTT-102, and NOTT-109. Two types of polyhedral cages are interconnected. NOTT, Nottingham University. Reprinted from He Y, Zhou W, Yildirim T, Chen B. A series of metal-organic frameworks with high methane uptake and an empirical equation for predicting methane storage capacity. Energy Environ Sci 2013;6:2735–44, with permission of the Royal Society of Chemistry.

250 200

150 100

50

0

250 200

150

100

50 0

0 (A)

Absolute CH4 adsorption (volSTP vol–1)

Absolute CH4 adsorption (volSTP vol–1)

Absolute CH4 adsorption (volSTP vol–1)

Porous Materials

1,000

2,000

3,000

Volumetric surface area (m2 cm–3)

250

200

150 100

50

0 0

(B)

193

2,000

4,000

0

6,000

Gravimetric surface area (m2 g–1)

(C)

0.2 0.4 0.6 0.8 Void fraction (helium probe)

1

Fig. 9 Structure–property relationships for 137,953 hypothetical metal organic frameworks (MOFs). (A) Volumetric methane adsorption shows a clear linear relationship with volumetric surface area. Red dots correspond to MOFs that have enough space to interpenetrate, but are not interpenetrated. (B) Methane adsorption initially increases with gravimetric surface area, but then begins to decrease when it has reached the optimal gravimetric surface area. (C) A void fraction of 0.8 is optimal for volumetric methane uptake at 35 bar. Reprinted from Wilmer CE, Leaf M, Lee CY, et al. Large-scale screening of hypothetical metal-organic frameworks, Nat Chem 2012;4:83–9, with permission of the Springer Nature.

to predict the CH4 storage in MOFs at 35 and 100 bar as a function of dominant pore diameter (DP) (Å ), void fraction, and gravimetric surface area (m2 g 1) [51]. Similarly, Gómez-Gualdrón et al. [52] performed molecular simulations of 122,835 hypothetical MOFs to develop an analytical equation to predict the CH4 storage. They found that the maximum deliverable capacity (between 65 and 5.8 bar) among the hypothetical MOFs was 206 cm3 (STP) cm 3 at 298K. They proposed following materials’ properties to reach specific CH4 deliverable capacity targets: volumetric surface area between 2100 and 2300 m2 cm 3, largest cavity diameter (LCD) between 10 and 12 Å and, isosteric heat of adsorption (Qst) between 10.5 and 13.0 kJ mol 1. Martin et al. [53] constructed the first database of hypothetical porous polymer networks (PPNs), a class of porous materials consisting only of organic materials and screened this database (B18,000 materials) to predict the CH4 storage. They observed that the CH4 uptake in PPNs is highly dependent on the pore size of the materials. Results suggested that the CH4 uptake capacities of these materials can be improved by adjusting the pore sizes of the materials due to the optimization of CH4 adsorption sites for CH4–CH4 interactions. Studies that we reviewed so far investigated hypothetical MOFs which have been computationally generated but not experimentally synthesized. Sezginel et al. [54] performed GCMC simulations to predict CH4 uptake of 45 real MOFs that were already synthesized. Their simulation results showed a good agreement with the experimentally available CH4 uptake data of several MOFs. Based on the QSPR analysis, they suggested the optimal material properties must have the following characteristics to exceed the DOE target of 0.5 g g 1 CH4 uptake at a storage pressure of 35 bar: high gravimetric surface area (6000 m2 g 1), high void fraction (0.9), DP of 30 Å , and Qst of B30 kJ mol 1. The geometrical parameters used in the QSPR model for real MOFs were the same with those for hypothetical MOFs developed by Fernandez et al. [51]. However, considering the very large of number synthesized MOF materials, a larger trial set of real MOFs should be used to have a better statistical analysis to compare the performance results of real and hypothetical materials. To summarize, it can be seen that all these high-throughput computational studies have been very useful to uncover the structure–property relationships and to identify the desired structural properties of MOFs required to reach the CH4 storage target.

2.7.5

Case Studies

In this section, we present a brief discussion of potential applications of CH4 storage using MOFs. First, we discussed the CH4 storage for vehicular applications by comparing the storage performances of different adsorbent materials including MOFs. Then a thermodynamic tank model was discussed for natural gas storage in MOFs. A promising MOF, MIL-101(Cr) was compared with other MOFs in ANG storage using both experimental and computational approaches. We also discussed the materials genome approach to accelerate the development of new generation materials for efficient CH4 storage. Finally, the storage of hydrogen enriched methane (HEM) in porous materials was reviewed and the methods to calculate the volumetric and gravimetric energy densities of HEM adsorption were summarized.

2.7.5.1

Case Study 1: CH4 Storage for Transportation

A recent experimental study by Beckner and Dailly [55] evaluated the CH4 storage performances of different adsorbents in ANG and CNG systems. Beckner and Dailly [55] investigated the adsorption of CH4 in five different adsorbents, one AC, three MOFs,

194

Porous Materials

CuBTC, MIL-88a and MOF-177, and one porous polymer (PP) at 251C up to 250 bar. They calculated the total storage of a 110 L test tank considering the efficacy of the adsorbents. Their study provided insights into how the current benchmark materials can be designed for practical gas storage applications. Two important points given in this study are the difference between the bulk density and crystal density and the definition of storage gain. In order to calculate the volumetric gas uptake in MOFs, crystal densities of MOFs are commonly used in the literature. The crystal density is a microscopic quantity and it is calculated as the ratio of the mass of a unit cell to the volume of the unit cell. On the other hand, bulk density is a macroscopic quantity which is obtained from experiments and it considers the pore blockages and defects. Beckner and Dailly [55] showed that bulk density of the adsorbent must be used to compare the storage performance of the adsorbent-containing tank with other storage technologies because the material packing has a considerable effect on the gas storage performance of an adsorbent. The total gas storage is dependent on the amount of macropores and interparticular voids in the material. Therefore, using bulk density to calculate the storage density of an adsorbent-containing tank is more reasonable. Bulk density was used to calculate the volumetric storage capacity in that study, Vst ¼ mst  rbulk

ð8Þ

where Vst is the volumetric storage capacity (g CH4 L 1 material), mst is the total mass of gas stored in an ANG system (g CH4 g 1 material), rbulk is the bulk density (g material cm 3 material). The total mass of gas stored in an ANG system (mst) is calculated from the excess adsorption using the following formula: mst ¼ mexc þ

r0 ðp; TÞ V0 ms

ð8Þ

where mexc is the excess adsorption (g CH4 g 1 material), r0(p,T) is the gas density (g cm 3) as a function of pressure (p), and temperature (T), ms is the total mass of sample in the system (g material 1), and V0 is the total volume accessible to the gas (cm3). Beckner and Dailly [55] defined the storage gain by subtracting the total amount of gas in the adsorbed system from the amount of gas in an empty tank at the same density and storage volume as follows: Gain ¼ mst

r0 ðp; TÞ ðV0 ms

VS Þ

ð9Þ

In Eq. (9), Vs is the volume of the solid and defined in Eq. (10). Vtank ¼ V0 þ Vs

ð10Þ

By combining Eqs. (8) and (9) they calculated the storage gain using the measured quantities: Gain ¼ mexc

r0 ðp; TÞ rs

ð11Þ

where rs (ms Vs 1) is the skeletal density (g cm 3). The skeletal density of each sample was obtained from the volumetric measurements. Fig. 10 shows both excess adsorption and storage gain for five different adsorbents. AC and CuBTC (referred as Cu-MOF in the label of the figure) have the highest excess adsorption uptakes and they exhibited a positive storage gain at 250 bar. If the storage gain has a positive value, for example, 0.1 g g, 1, this means that an ANG tank which has 10 g adsorbent can store 1 g of CH4 more than a CNG tank with the same pressure and volume. The negative storage gain can be explained by the storage lost due to the dead volume of the adsorbents because of the high pressure. As shown in Fig. 10, 5–11 wt% storage gain can be obtained at around 50 bar using AC and CuBTC adsorbents. This result shows that ANG technology can be used for home refueling options. Beckner and Dailly [55] assumed 0.39 gallon gasoline equivalent (gge) per kilogram of gas to predict the energy content of CH4. The adsorbed CH4 within three MOFs, one AC and one polymer were compared with CNG at 50 and 250 bar. All adsorbents outperformed CNG at 50 bar. However, Zn-MOF, Al-MOF, and porous polymers did not exhibit high performance at 250 bar. AC and CuBTC exhibited similar performance with CNG but indeed their performances cannot compete with that of CNG for practical applications. Results of this study suggested that ANG is a useful technology for low pressure applications (50 bar) and can be used for natural gas home fueling stations. Several industrial projects were carried out for vehicular applications of MOFs. One of the famous projects was EcoFuel Asia Tour in 2007 [56]. In this project, a Wolkswagen Caddy EcoFuel car was designed using MOF-enhanced fuel tanks and a journey was planned for 14 countries. Fig. 11 shows MOF-enhanced fuel tanks with CH4 sponsored by BASF and the journey map for the EcoFuel Asia Tour. A commercial MOF, Basolite C300, (other common names are HKUST-1 and CuBTC) was used as a potential carrier for natural gas. MOF was used as pellets in a traditional tank and the tank was filled with CNG. Analysis of this tour showed that the car consumed almost 7 kg of natural gas per 100 km and produced 1.3 t less CO2 emissions compared to normal Volkswagen Caddy 1.6 L petrol engine [56]. Similar commercial applications of MOFs are still under investigation. However, it is important to note that there are some critical points such as system weight, volume, cost, efficiency, and durability which hinder the potential applications of MOFs for on-board gas storage [56]. In automotive applications, the weight of vehicles’ storage tank is one of the constraints. The complexity of the vehicle system increases its weight and also the required energy. Therefore, reducing the size and weight of the storage tank is a key issue for overall energy efficiency. In order to use MOFs as adsorbents in ANG technology for commercial applications it is also necessary to balance the cost/performance relationship. Since the large volume of

Excess adsorption (g g−1)

Porous Materials

195

Activated carbon Zn MOF AI MOF Cu MOF Porous polymer

0.15

0.10

0.05

Storage gain (g g−1)

0.00 0.10

0.05

0.00

–0.05

–0.10 0

50

100 150 Pressure (bar)

200

250

Fig. 10 Comparison of excess adsorption (top) and storage gain for five adsorbents. Reprinted from Beckner M, Dailly A. Adsorbed methane storage for vehicular applications. Appl Energy 2015;149:69–74, with permission of the Elsevier.

(C) Berlin Orenburg

Shanghai Amman (A)

(B)

Delhi

(D) Bangkok

Fig. 11 (A) Basolite C300 (HKUST-1), (B) MOF-enhanced fuel tanks with CH4 sponsored by BASF, (C) Volkswagen Caddy EcoFuel prototype car, and (D) map for EcoFuel Asia Tour 2007, from Berlin to Bangkok: 32,000 km with Basolite C300 in tank. Reprinted from Silva P, Vilela SMF, Tome JPC, Almeida Paz FA. Multifunctional metal-organic frameworks: from academia to industrial applications. Chem Soc Rev 2015;44:6774–803, with permission of the Royal Society of Chemistry.

the storage tank and the increase in weight are reflected to the cost, the minimal and useful storage space must be designed for feasible applications. Durability of the adsorbent particles is also important to maintain the overall efficiency. All these issues can be addressed in the near future due to the continuously increasing academic and industrial interests on MOFs.

196

Porous Materials

2.7.5.2

Case Study 2: A Thermodynamic Tank Model for Natural Gas Storage in Metal Organic Frameworks

The current literature that we summarized so far considers only CH4 uptake in MOFs for natural gas storage. However, natural gas also contains small amount of higher hydrocarbons such as ethane and propane. In order to understand the effects of presence of other hydrocarbons, ethane and propane, on the overall CH4 storage performance of MOFs in an ANG tank, Zhang et al. [57] developed a theoretical model. This model was developed for an ANG tank which assumes isothermal conditions. It requires single-component adsorption isotherm data as an input and the ideal adsorbed solution theory (IAST) was used to predict ternary (methane–ethane–propane) mixture adsorption isotherms. IAST is a well-known method that estimates adsorption data of gas mixtures using single-component adsorption isotherm data at the same temperature by assuming that the adsorbed species are in an ideal solution [58]. Zhang et al. [57] initially tested the ANG tank model for five promising, widely examined MOFs, HKUST-1, IRMOF-1, NU-111, NU-125, and NU-800 which have high volumetric working capacity for CH4. They estimated deliverable energies considering 65 bar as the storage pressure and 5.8 bar as the deliverable pressure. Fig. 12 shows the comparison of predicted deliverable energy of pure CH4 and natural gas mixtures in five MOFs. The dashed line in this figure illustrates the ARPA-E target (9 MJ L 1) for the deliverable energy. As shown in Fig. 12, none of the MOFs can reach the ARPA-E target. ANG tank system has lower deliverable energy compared to pure CH4. Since ethane and propane have stronger interactions with the framework compared to pure CH4, their desorption process is difficult which causes accumulation over multiple cycles of adsorption and desorption. Although the mole fractions of ethane and propane are low in the natural gas, it can be discussed that they have a negative impact on the overall performance of storage materials. One striking example can be seen for HKUST-1. This MOF was identified as one of the promising adsorbents for CH4 storage in the literature. However, when the ANG tank system was considered, the deliverable energy of HKUST-1 significantly decreased. Zhang et al. [57] demonstrated that due to the small cavities of HKUST-1, ethane, and propane molecules adsorb more strongly than CH4 molecules. Analysis of GCMC simulations showed that propane molecules are available in small cavities of HKUST-1 at 5.8 bar and they block the adsorption of CH4 and ethane molecules. These results suggest that the influence of higher hydrocarbons on the performance of an ANG tank is significant for MOFs which have smaller cavities with diameters of 4–10 Å and the presence of the impurities must be considered in assessing the delivery energy of MOFs, especially the narrow-pored ones. Zhang et al. [57] later computationally investigated 120 MOFs and observed correlations between the deliverable energy and structural properties of these materials. They reported that the optimal MOF adsorbent should possess a helium void fraction of 0.9, a volumetric surface area of 1850 m2 cm 3, a gravimetric surface area of 5500 m2 g 1 and a heat of adsorption of 10 kJ mol 1 at low loading. Among 120 MOFs, MOF-143 was identified as the best material for natural gas storage which has the maximum deliverable energy (5.43 MJ L 1) [57]. Results of this study have been very useful for understanding the realistic performance of MOFs for natural gas storage.

2.7.5.3

Case Study 3: Experimental Investigation of MIL-101(Cr) for CH4 Storage

In order to increase the CH4 working capacity, several novel approaches have been suggested in the literature. Kayal et al. [59] attempted to combine the ANG and LNG technologies in an efficient way as we will briefly summarize here. Practical applications of the ANG technology have several drawbacks. For example, since the enthalpy of CH4 adsorption is high at ambient temperature, additional cooling systems are required for on-board CH4 storage. In addition to this, a thermal equilibrium time must be 10 Deliverable energy (MJ L−1 of tank)

9 CNG (250–5.8 bar): 9.0 MJ L−1 of tank

8

ANG tank system (200th cycle)

7

Pure methane

6 5 4 3 2 1

11 N

U

-1

25 U N

-8 U N

-1

00

1 FM O IR

H

KU

ST

-1

0

Fig. 12 Comparison of predicted deliverable energy of pure methane and natural gas mixtures in five metal organic frameworks (MOFs). Reprinted from Zhang H, Deria P, Farha OK, Hupp JT, Snurr RQ. A thermodynamic tank model for studying the effect of higher hydrocarbons on natural gas storage in metal-organic frameworks. Energy Environ Sci 2015;8:1501–10, with permission of the Royal Society of Chemistry.

Porous Materials

197

NG ANG Regasification

LNG P

(Coupling of ANG and LNG) Charging of NG Mesoscopic diffusion through MOF

MOFs LNG out

LNG in

1 mm 10 cm

Microscopic diffusion through MOF

Macroscopic flow through MOFs packing

Adsorption on MOF surface-pores

Fig. 13 The schematic representation of liquefied natural gas (LNG) regasification and adsorbed natural gas (ANG) tank filling system. The mass transport in metal organic frameworks (MOFs) is demonstrated by four different steps: macroscopic flow through MOFs packing, mesoscopic diffusion, microscopic pore diffusion, and adsorption process on MOF surface. Reprinted from Kayal S, Sun B, Chakraborty A. Study of metalorganic framework MIL-101(Cr) for natural gas (methane) storage and compare with other MOFs (metal-organic frameworks). Energy 2015;91:772–81, with permission of the Elsevier.

considered during the CH4 adsorption process in porous materials [59]. In order to overcome these limitations of the ANG technology, Kayal et al. [59] developed a novel methodology as shown in Fig. 13. They combined LNG regasification and ANG tank filling processes. In this methodology, the latent heat of LNG vaporization is used to precool the ANG adsorption bed. They reported that combining ANG and LNG technologies increased the overall energy efficiency of the storage chamber up to 25% based on the filling conditions such as temperature and pressure. MIL-101(Cr) was used in that study as a porous adsorbent. The CH4 delivery capacity of MIL-101(Cr) was reported to be as 180 cm3(STP) cm 3 at 65 bar at 298K. When the coupling of LNG vaporization and the ANG tank filling process was carried out, the CH4 delivery capacity of MIL-101(Cr) was reported to be as 240 cm3(STP) cm 3 at the adsorption (desorption) temperature of 160K (298K) and pressure of 6 bar (5 bar). This CH4 delivery capacity of MIL-101(Cr) is much higher than those of HKUST-1 (205 cm3(STP) cm 3 at 65 bar at 298K) and Al-MOF-519 (210 cm3(STP) cm 3 at 65 bar at 298K) which have the highest CH4 delivery capacities reported in Ref. [59]. Fig. 14 compares the CH4 delivery capacities in MIL-101(Cr) under different operating conditions. For example, the CH4 delivery capacity of MIL-101(Cr) was reported to be 125 cm3(STP) cm 3 at the adsorption (desorption) temperature and pressure of 160K (298K) and 1.2 bar (5 bar), respectively. It is important to note that CH4 adsorption at 160K and 1.2 bar is equal to the CH4 adsorption at 298K and 35 bar. Therefore, the CH4 delivery capacity estimated at the adsorption (desorption) temperature and pressure of 160K (298K) and 1.2 bar (5 bar) is equal to the one obtained at the adsorption (desorption) temperature and pressure of 298K (298K) and 35 bar (5 bar), respectively. When the adsorption pressure was increased up to 6 bar, the CH4 delivery capacity was improved (240 cm3(STP) cm 3) as shown in Fig. 14. As can be seen from the figure, the CH4 delivery capacity (240 cm3(STP) cm 3) was much higher than the delivery capacity (125 cm3(STP) cm 3) reported at 298K. Results of this study show that coupling of ANG and LNG technologies can be a good strategy to enhance the CH4 working capacities of MOFs.

2.7.5.4

Case Study 4: High-Throughput Screening of Metal Organic Frameworks

In the past few years, we have witnessed the development of novel computational techniques that are used to efficiently screen a large number of materials for CH4 storage. High-throughput screening studies are very useful to provide initial

Porous Materials

0.4 0.35 Gravimetric uptake (kg kg–1)

CH4 charging at 160 K and 6 bar

CH4 charging at 160 K and 1.2 bar

300 CH4

0.3

250 Working capacity

CH4

0.25

200

0.2 150 Desorption

0.15

Working capacity

Adsorption

100

0.1

Volumetric uptake (cm3 cm–3)

198

50

0.05 CH4 discharge at 5 bar (298 K)

0 0

2

4

6

8

CH4

0 10

Pressure (bar) Fig. 14 Methane (CH4) adsorption (desorption) in MIL-101 (Cr) at 160K (298K). Reprinted from Kayal S, Sun B, Chakraborty A. Study of metalorganic framework MIL-101(Cr) for natural gas (methane) storage and compare with other MOFs (metal-organic frameworks). Energy 2015;91:772–81, with permission of the Elsevier.

estimate about the performances of materials for a desired application before extensive calculations and experiments. The idea is to use the results of screening studies to narrow down the number of promising materials from thousands to tens and then direct experimental efforts, time and resources to those promising materials. Efficient computational algorithms are required to screen a large number of materials in a reasonable amount of time. Simon et al. [5] recently screened a very large number of porous materials, 650,000 materials, including predicted zeolite structures, predicted PPNs, hypothetical MOFs and hypothetical ZIFs, and experimentally synthesized MOFs from the computation-ready experimental (CoRE) MOF database. They initially estimated the CH4 delivery capacity of 20,000 hypothetical ZIFs and 10,000 hypothetical MOFs and found that the materials which have large cavity diameter (LCD) in the range 8–14.5 Å and void fraction in the range 0.25-0.7 exhibited the highest CH4 delivery capacity. After obtaining this initial result, they extended their study by adding 37,437 hypothetical ZIF and 34,363 hypothetical MOF structures into their library to further investigate the CH4 delivery capacity of these materials. Fig. 15 shows the CH4 delivery capacity (in terms of v(STP) v 1, v-volume) of different porous materials including hypothetical MOFs, hypothetical ZIFs, zeolites, PPNs, and experimentally synthesized real MOFs. Data for this group were obtained from the GCMC simulations using a random sample of 3701 structures from each group. In addition, MOFs which have experimental data are shown in this figure with different symbols and the delivery capacities of these MOFs were estimated from the experimental adsorption isotherms. Results showed that PPNs and MOFs have higher CH4 delivery capacities than zeolites. The highest CH4 delivery capacity among the 650,000 materials was reported to be 196 v(STP) v 1. Among the experimentally tested MOFs, MOF-5 (185 v(STP) v 1), HKUST-1 (185 v(STP) v 1), UTSA-76 (194 v(STP) v 1), and MOF-519 (208 v(STP) v 1) exhibited the best CH4 delivery capacities. However, none of the materials can exceed the ARPA-E delivery capacity target (315 v(STP) v 1). Simon et al. [5] highlighted that experimentally tested MOFs have already reached the boundary of the thermodynamic and material performance limit. They showed that the materials which have high CH4 delivery capacities exhibit a narrow range of crystal densities. They also reported that the optimal pore diameter is almost 11 Å for CH4 storage. When the pore diameter was increased from 11 to 50 Å as shown in Fig. 15, the CH4 delivery capacity of materials was found to decrease and it approached the delivery capacity of an empty tank. This was explained by the weak van der Waals interactions between CH4 molecules and the walls due to the extreme increase in the pore sizes. In order to tune the delivery capacity of materials, one possible strategy is to change the storage and delivery conditions. Simon et al. [5] demonstrated that a lower discharge pressure or a higher discharge temperature can be used for materials that have strong interactions with CH4 molecules in order to increase the delivery capacity of the materials. On the other hand, a high storage pressure can be used for materials which have weak van der Waals interactions due to their larger pore volumes to improve the delivery capacity. In another study, Fu et al. [60] selected 1200 materials from the hypothetical MOF database considering the top performing materials based on the excess CH4 adsorption per material weight, the excess CH4 adsorption per material volume, the material’s void fraction and the material’s specific surface area. They performed GCMC simulations to determine the volumetric CH4 delivery capacity of these materials under different operating conditions. Their results showed that both volumetric and gravimetric CH4 delivery capacities increase (30%) when the storage pressure increases from 35 to 70 bar at 298K. However, these delivery capacities were still under the ARPA-E target. When the storage pressure increased up to 170 bar, the volumetric delivery capacities were still under the target. However, both total and gravimetric delivery amounts were above the ARPA-E target (0.5 g g 1). If the

Porous Materials

HKUST-1 MOF-5 Mg2(dobdc) Ni2(dobdc) PCN-14 IRMOF-6 PCN-16(300K) UTSA-20 UTSA-76a NU-125 MOF-520 ZIF-8(303K) ZIF-76(303K) PPN-4(295K) PPN-3(295K) PPN-2(295K) PPN-1(295K) DD3R Silicalite Free space tank

200 Deliverable capacity (v STP/v)

199

150

100

50

0 0

500

1000

1500

2000

2500

Crystal density (kg m−3) 200 MOFs Deliverable capacity (v STP/v)

Expt’l MOFs ZIFs

150

Zeolites PPNs 100

50

0

0

10

20 30 Largest included sphere (Å)

40

50

Fig. 15 The relationships between the delivery capacity for methane (CH4) and (A) crystal density and (B) largest included sphere. The straight line shows the delivery capacity of an empty tank (62 v(STP)
v 1). Reprinted from Simon CM, Kim J, Gomez-Gualdron DA, et al. The materials genome in action: identifying the performance limits for methane storage. Energy Environ Sci 2015;8:1190–9, with permission of the Royal Society of Chemistry.

storage pressure was set to 250 bar as in the case of a CNG tank, the total volumetric adsorption amounts could exceed the target. However, volumetric delivery capacities were still under the target. Finally, Fu et al. [60] decreased the storage temperature from 298 to 233K and reported that more than half of these materials can exceed both the volumetric and gravimetric ARPA-E targets. They also examined the chemical properties of a MOF that have high CH4 storage capacity and suggested that a promising MOF adsorbent for CH4 storage should have large conjugated phenyl ring in the organic unit and copper sites as the inorganic unit. The MOF was suggested to have a surface area around 5000 m2 g 1 and a void fraction around 0.9 in order to achieve high CH4 storage [60]. As the current literature review we provided above suggested, none of the MOF materials that has been examined to date could exceed the ARPA-E volumetric CH4 delivery target at room temperature and 35 bar or 65 bar. Increasing the adsorption pressure or decreasing the adsorption temperature can improve the delivery capacities of MOFs as stated in both cases. Although the results are promising, further developments are required in this research area for commercial applications of MOFs for onboard CH4 storage.

2.7.5.5

Case Study 5: Hydrogen Enriched Methane

Recent studies showed that the use of HEM for on-board natural gas storage vehicle can enhance the combustion performance of the engine and decrease the noxious emissions [61]. Additionally, there is a tremendous effort to separate H2 or CH4 from the synthetic gas and separation of this gas mixture is a challenging process due to the weak van der Waals forces. Therefore,

200

Porous Materials

co-adsorption of H2 and CH4 in an ANG system can be useful for the refinery off-gas separation processes due to the time and cost savings. Liao et al. [61] investigated the energy density of combustion of stored HEM by studying 33 different covalent organic frameworks (COFs). COFs can be considered as a sub-family of MOFs. They are composed of light elements such as B, C, N, and O linked together by strong covalent bonds. Due to the existence of strong covalent bonds, COFs are considered to have better chemical and thermal stabilities compared to MOFs and ZIFs. They have also low densities due to the loss of transition metals in their building units. Liao et al. [61] performed GCMC simulations for 33 different COFs at room temperature with the pressure varying from 0.01 to 85 bar. They assumed equimolar mixture of H2 and CH4 in the system. The total volumetric and gravimetric energy density of HEM adsorption in COFs was calculated using the output of molecular simulations by [61] Ev ¼ NCH4 ðvÞ  ð DCH4 Þ þ NH2 ðvÞ  ð DH2 Þ

ð12Þ

Eg ¼ NCH4 ðgÞ  ð DCH4 Þ þ NH2 ðgÞ  ð DH2 Þ

ð13Þ

where Ev and Eg are the total volumetric and gravimetric energy density, NCH4 ðvÞ and NH2 ðvÞ are the volumetric gas uptake of each species (mol unit volume of the COF adsorbent 1), NCH4 ðgÞ and NH2 ðgÞ are the gravimetric gas uptake of each species (mol unit mass of the COF adsorbent 1), DCH4 and DH2 represent the combustion heat of CH4 and H2, respectively. Liao et al. [61] discussed that one has to also consider H2 adsorption selectivity in COFs to assess the adsorbents’ realistic HEM storage performance. Adsorption selectivity of H2 from CH4 (SH2 =CH4 ) can be calculated by using SH2 =CH4 ¼ ðxH2 =xCH4 Þ=ðyH2 =yCH4 Þ

ð14Þ

where x and y represent the molar fractions of CH4 and H2 in the adsorbed and bulk phase, respectively. Liao et al. [61] reported that the total gravimetric energy density of HEM in COF-102, COF-103, COF-105, and COF-108 can reach the 2015 DOE target (6480 kJ kg 1) over 30 bar. They also stated that the total volumetric energy density of HEM in COF102, COF-103, and NPN-1 (nitoso polymer networks, NPN type COF) can reach the 2015 DOE target (4680 kJ L 1) above 40 bar. The H2 selectivity is also important to compare the HEM adsorption performance of porous materials. However, the majority of COFs shows low H2 selectivity. Narrow-pored COFs generally showed low H2 selectivity because in these MOFs CH4 adsorption was favored due to the stronger interaction between CH4 molecules and the host material compared to the interaction between H2 molecules and the material. On the other hand, COFs which have larger pore diameters (15 Å ) such as COF-105, COF-108, and COF-43 can offer relatively high H2 selectivities. Results of this study showed that the HEM adsorption in COFs and H2 selectivity of COFs are affected by the physical properties of the materials such as surface area, porosity, and pore sizes of COFs and the chemical properties such as isosteric heat of adsorption of H2 and CH4 in COFs. High surface area and isosteric heat of adsorption of CH4 increased the energy density of HEM in COFs and COFs which have large pores and high porosities showed high H2 selectivities. This case study suggests that instead of considering only CH4 uptake in MOFs, HEM storage capacity of MOFs can be also examined and the energy density of HEM adsorption in MOFs can be analyzed.

2.7.6

Future Directions

MOFs offer a great potential for CH4 storage due to their permanent porosities, well-defined pore structures, availability of large ranges of pore sizes and shapes, low densities. In addition to these structural properties, the most important advantage of MOFs over well-known nanoporous materials is that chemical and physical properties of MOFs can be tuned during synthesis. As a result, thousands of MOFs have been synthesized to date and several MOFs have been already tested for CH4 storage. As discussed above, most MOFs have higher volumetric CH4 storage capacities compared to traditional porous materials. In order to use a MOF as an efficient medium for natural gas storage in a standard vehicle several issues must be addressed:

•

•

Currently only a few MOFs, namely Basolites C300 (HKUST-1), A100 (MIL-53 (Al)), Z1200 (ZIF-8), F300 (Fe-BTC), A520 (Alfumarate) and Basosiv M050 (Mg-formate) are commercially available which are produced by BASF and distributed by Sigma Aldrich. UiO-66 is also commercially available and distributed by Strem Chemicals [62]. These MOFs are not sold in large amounts, generally not more than 500 g. Since the synthesis of MOFs requires long reaction times (12–48 h) and expensive solvents, the production process is costly. In addition to this, yield is not very high in the MOF synthesis. Therefore, novel methods for large-scale synthesis of MOFs should be developed to decrease the high production costs in order to widen commercialization of a large number of MOFs. Recently, a novel synthesis method was developed to synthesize MIL-101(Cr). Researchers used waste PET (polyethylene terephthalate) bottles as the source of BDC acid (as an organic linker of the MOF) to synthesize Cr-MOFs [63]. Their results were highly promising for the potential applications of waste PET bottles to synthesize MOFs for H2 storage [63]. Similar approaches can be used to synthesize other MOFs which have BDC linkers and these MOFs can be further investigated for CH4 storage. Advances in synthesis of MOFs and obtaining high-yield from these syntheses will accelerate commercialization of a large number and variety of MOFs. Another issue which hinders the wide-spread industrial applications of MOFs is their thermal and mechanical stabilities. In order to use MOFs in vehicular applications, long-term stability of MOFs should be investigated. Both experimental and computational studies are highly required to assess the realistic performance of MOFs for on-board natural gas storage of vehicles. For example, several MOFs are known to be sensitive to atmospheric moisture and these materials lose their crystal

Porous Materials

•

•

•

•

201

structures when exposed to water [64]. This may be a significant problem when MOFs are used as storage devices for natural gas streams since natural gas reserves also contain some amount of water. Structural, mechanical, chemical, and thermal stabilities are issues that are more likely to be examined by experimental studies under realistic operating conditions. Examining the natural gas storage and delivery performances of MOFs under practical operating conditions will be very useful to unlock the potential of MOFs as on-board natural gas storage vehicles. In order to make MOF-based ANG systems competitive to gasoline fuel, the heat transfer characteristics of the adsorbent materials should be also examined. The knowledge of specific heat capacities of promising MOFs for natural gas storage is required to assess the thermal performance of ANG systems. Temperature variations for the whole working temperature range should be considered to reduce the capital and operating costs. Current literature that we reviewed above considers only CH4 uptake in MOFs for natural gas storage. However, as we discussed in the thermodynamic tank model given in the second case study, natural gas also contains small amount of higher hydrocarbons such as ethane and propane. Therefore, the effects of higher hydrocarbons in natural gas on the methane delivery performance of MOFs should be investigated in future experimental and computational studies. Most studies in the literature about CH4 uptake with MOFs assume a steady-state system that is at the thermodynamic equilibrium. However, for realistic applications it is vital to consider the dynamics of the whole system (adsorbent–adsorbate). Therefore, it may be useful to model the complete system considering the unsteady state conditions. Using computational methods, diffusion of CH4 molecules through the pores of materials can be observed. Equilibrium molecular dynamics (EMD) simulations can be performed to observe the motion of the gas molecules in a certain time and to understand the interactions between gas molecules and the framework. Non-EMD can be also performed to observe the changes in the system under different circumstances such a sudden decrease in pressure or temperature. These atomic-level simulations can provide information that will be helpful for design and development of new materials that can achieve better CH4 storage. Synthesis and testing of the computer generated materials that have high CH4 storage capacity and high CH4 delivery are very important. Molecular simulations assume ideal conditions such as perfect, defect-free MOF structures. However, in reality synthesized materials may have defects that can give different CH4 storage/delivery capacities than the simulations. Similarly, most molecular simulations examine pure CH4 storage performance of materials by neglecting impurities in the natural gas pipelines. Therefore, results of molecular simulations to predict the CH4 storage/delivery capacities of MOFs must be envisioned as the best performance results that can be expected from a MOF material under ideal conditions. Another important issue which is required for the adoption of MOFs in vehicles is overall cost-performance analysis of the ANG fuel tanks. None of the studies we reviewed in this chapter has focused on cost analysis of MOF-based CH4 storage systems. This type of analysis should be done in order to foresee if it is possible to replace gasoline-sourced energy systems with the natural gas stored in MOFs. Cost of the synthesis of a MOF in the large scale would have a significant impact on the overall cost of the MOF-based CH4 storage systems. Therefore, future studies on cost-effective, high-yield MOF synthesis will be very valuable as discussed at the beginning of this section.

2.7.7

Conclusions

Developing energy-efficient, sustainable and safe systems for ANG is important due to the economic and environmental reasons. The critical issue for ANG technology is the careful selection of the adsorbent/storage material. A promising adsorbent material should have high natural gas storage capacity, high working capacity, easy regeneration, high chemical, mechanical, structural stabilities at a reasonable cost. Since natural gas is mainly composed of CH4, most of the research has focused on identification of materials that can store CH4 as efficiently as possible. The review of the recent literature given in this chapter suggests that a new porous material family named as MOFs is very promising for CH4 storage applications. Both experimental and computational studies have shown that MOFs exhibit similar or better CH4 storage performances compared to different adsorbent materials such as ACs and zeolites. Although MOFs were not able to reach the recent CH4 storage target of the DOE, there are opportunities to tune the structural properties of MOFs in order to improve their storage performance and utilize these new materials for natural gas home fueling stations in near future. Studies suggested that storing and delivering CH4 at lower temperatures can be useful to attain the new targets using the current technologies although low temperatures result in some additional cost. At that point, it is important to note that focusing on deliverable CH4 amount instead of the stored CH4 amount can be critical for the design of novel materials because deliverable amounts are more closely related to the practical usage of materials. Therefore, future studies on new material development should be focused on improving the volumetric CH4 delivery amount of materials rather than the total volumetric CH4 uptake capacity of the materials. The number of studies on computational investigation of MOFs for CH4 storage is continuously increasing because there are thousands of MOFs with different chemical and physical properties and it is practically impossible to test these materials at the lab scale using purely experimental methods. In addition to identifying the most promising MOF material for CH4 storage, computational studies that utilize atomic-level simulations also provide useful insights into the chemical and geometric characteristics of materials. This information can be then used to design and synthesize new materials with the desired characteristics to reach the high CH4 storage performance. We believe that we will continue to witness the exponential growth of the research on the development of more novel adsorbent materials, in addition to MOFs, for CH4 storage given the concerns over the climate change, continuously increasing energy demands and depletion of the oil resources.

202

Porous Materials

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54]

MacDonald GJ. The future of methane as an energy resource. Annu Rev Energy 1990;15:53–83. Mason JA, Veenstra M, Long JR. Evaluating metal-organic frameworks for natural gas storage. Chem Sci 2014;5:32–51. Beckner M, Dailly A. Hydrogen and methane storage in adsorbent materials for automotive applications. Int J Energy Res 2016;40:91–9. He Y, Zhou W, Qian G, Chen B. Methane storage in metal-organic frameworks. Chem Soc Rev 2014;43:5657–78. Simon CM, Kim J, Gomez-Gualdron DA, et al. The materials genome in action: identifying the performance limits for methane storage. Energy Environ Sci 2015;8:1190–9. Düren T, Sarkisov L, Yaghi OM, Snurr RQ. Design of new materials for methane storage. Langmuir 2004;20:2683–9. Menon V, Komarneni S. Porous adsorbents for vehicular natural gas storage: a review. J Porous Mater 1998;5:43–58. Eddaoudi M, Kim J, Rosi N, et al. Systematic design of pore size and functionality in isoreticular MOFs and their application in methane storage. Science 2002;295:469–72. Yaghi OM, O’Keeffe M, Ockwig NW, et al. Reticular synthesis and the design of new materials. Nature 2003;423:705–14. Rowsell JLC, Yaghi OM. Metal-organic frameworks: a new class of porous materials. Microporous Mesoporous Mater 2004;73:3–14. Allen FH. The Cambridge Structural Database: a quarter of a million crystal structures and rising. Acta Crystallogr, Sect B: Struct Sci 2002;58:380–8. Wilmer CE, Kim KC, Snurr RQ. An extended charge equilibration method. J Phys Chem Lett 2012;3:2506–11. Zhou H-C, Long JR, Yaghi OM. Introduction to metal-organic frameworks. Chem Rev 2012;112:673–4. Langmi HW, Ren J, North B, Mathe M, Bessarabov D. Hydrogen storage in metal-organic frameworks: a review. Electrochim Acta 2014;128:368–92. Sculley J, Yuan D, Zhou H-C. The current status of hydrogen storage in metal-organic frameworks-updated. Energy Environ Sci 2011;4:2721–35. Suh MP, Park HJ, Prasad TK, Lim D-W. Hydrogen storage in metal-organic frameworks. Chem Rev 2011;112:782–835. Murray LJ, Dinca M, Long JR. Hydrogen storage in metal-organic frameworks. Chem Soc Rev 2009;38:1294–314. Hu YH, Zhang L. Hydrogen storage in metal-organic frameworks. Adv Mater 2010;22. Li J-R, Ma Y, McCarthy MC, et al. Carbon dioxide capture-related gas adsorption and separation in metal-organic frameworks. Coord Chem Rev 2011;255:1791–823. Sumida K, Rogow DL, Mason JA, et al. Carbon dioxide capture in metal-organic frameworks. Chem Rev 2012;112:724–81. Schoedel A, Ji Z, Yaghi OM. The role of metal-organic frameworks in a carbon-neutral energy cycle. Nat Energy 2016;1:16034–47. Zou R, Abdel-Fattah AI, Xu H, Zhao Y, Hickmott DD. Storage and separation applications of nanoporous metal-organic frameworks. CrystEngComm 2010;12:1337–53. Gándara F, Furukawa H, Lee S, Yaghi OM. High methane storage capacity in aluminum metal-organic frameworks. J Am Chem Soc 2014;136:5271–4. Butt HJ, Graf K, Kappl M. Physics and chemistry of interfaces. Weinheim: Wiley; 2006. Brunauer S, Deming LS, Deming WE, Teller E. On a theory of the van der Waals adsorption of gases. J Am Chem Soc 1940;62:1723–32. Konstas K, Osl T, Yang Y, et al. Methane storage in metal organic frameworks. J Mater Chem 2012;22:16698–708. Myers AL. Thermodynamics of adsorption. Chem Thermodyn Ind 2004;243–52. Chakraborty A, Saha B, Koyama S, Ng K. On the thermodynamic modeling of the isosteric heat of adsorption and comparison with experiments. Appl Phys Lett 2006;89:171901–3. Frenkel D, Smit B. Understanding molecular simulation: from algorithms to applications. London: Academic press; 2002. Getman RB, Bae Y-S, Wilmer CE, Snurr RQ. Review and analysis of molecular simulations of methane, hydrogen, and acetylene storage in metal-organic frameworks. Chem Rev 2011;112:703–23. Wang S. Comparative molecular simulation study of methane adsorption in metal-organic frameworks. Energy Fuels 2007;21:953–6. Gallo M, Glossman-Mitnik D. Fuel gas storage and separations by metal-organic frameworks: simulated adsorption isotherms for H2 and CH4 and their equimolar mixture. J Phys Chem C 2009;113:6634–42. Coudert F-X, Fuchs AH. Computational characterization and prediction of metal-organic framework properties. Coord Chem Rev 2016;307:211–36. Rappe AK, Casewit CJ, Colwell KS, Goddard WA, Skiff WM. UFF, a full periodic table force field for molecular mechanics and molecular dynamics simulations. J Am Chem Soc 1992;114:10024–35. Mayo SL, Olafson BD, Goddard WA. Dreiding: a generic force field for molecular simulations. J Phys Chem 1990;94:8897–909. Keskin S, Liu J, Johnson JK, Sholl DS. Atomically detailed models of gas mixture diffusion through CuBTC membranes. Microporous Mesoporous Mater 2009;125:101–6. Guo H-C, Shi F, Ma Z-F, Liu X-Q. Molecular simulation for adsorption and separation of CH4/H2 in zeolitic imidazolate frameworks. J Phys Chem C 2010;114:12158–65. Lorentz H. Ueber die anwendung des satzes vom virial in der kinetischen theorie der gase. Ann Phys (Berlin, Germany) 1881;248:127–36. Berthelot D. Sur le melange des gaz. C R Hebd Seanc Acad Sci (Paris) 1898;126:1703–855. Martin MG, Siepmann JI. Transferable potentials for phase equilibria. 1. United-atom description of n-alkanes. J Phys Chem B 1998;102:2569–77. Babarao R, Hu Z, Jiang J, Chempath S, Sandler SI. Storage and separation of CO2 and CH4 in silicalite, C168 schwarzite, and IRMOF-1: a comparative study from Monte Carlo simulation. Langmuir 2007;23:659–66. Macedonia MD, Maginn EJ. A biased grand canonical Monte Carlo method for simulating adsorption using all-atom and branched united atom models. Mol Phys 1999;96:1375–90. Widom B. Some topics in the theory of fluids. J Chem Phys 1963;39:2808–12. Peng Y, Krungleviciute V, Eryazici I, et al. Methane storage in metal-organic frameworks: current records, surprise findings, and challenges. J Am Chem Soc 2013;135:11887–94. Mason JA, Oktawiec J, Taylor MK, et al. Methane storage in flexible metal-organic frameworks with intrinsic thermal management. Nature 2015;527:357–61. Song C, Liu H, Jiao J, et al. High methane storage and working capacities in a NbO-type metal-organic framework. Dalton Trans 2016;45:7559–62. Ban Y, Li Z, Li Y, et al. Confinement of ionic liquids in nanocages: tailoring the molecular sieving properties of ZIF‐8 for membrane based CO2 capture. Angew Chem 2015;127:15703–7. Sezginel KB, Keskin S, Uzun A. Tuning the gas separation performance of CuBTC by ionic liquid incorporation. Langmuir 2016;32:1139–47. He Y, Zhou W, Yildirim T, Chen B. A series of metal-organic frameworks with high methane uptake and an empirical equation for predicting methane storage capacity. Energy Environ Sci 2013;6:2735–44. Wilmer CE, Leaf M, Lee CY, et al. Large-scale screening of hypothetical metal-organic frameworks. Nat Chem 2012;4:83–9. Fernandez M, Woo TK, Wilmer CE, Snurr RQ. Large-scale quantitative structure–property relationship (QSPR) analysis of methane storage in metal-organic frameworks. J Phys Chem C 2013;117:7681–9. Gómez-Gualdrón DA, Wilmer CE, Farha OK, Hupp JT, Snurr RQ. Exploring the limits of methane storage and delivery in nanoporous materials. J Phys Chem C 2014;118:6941–51. Martin RL, Simon CM, Smit B, Haranczyk M. In silico design of porous polymer networks: high-throughput screening for methane storage materials. J Am Chem Soc 2014;136:5006–22. Sezginel KB, Uzun A, Keskin S. Multivariable linear models of structural parameters to predict methane uptake in metal-organic frameworks. Chem Eng Sci 2015;124:125–34.

Porous Materials

203

[55] Beckner M, Dailly A. Adsorbed methane storage for vehicular applications. Appl Energy 2015;149:69–74. [56] Silva P, Vilela SMF, Tome JPC, Almeida Paz FA. Multifunctional metal-organic frameworks: from academia to industrial applications. Chem Soc Rev 2015;44:6774–803. [57] Zhang H, Deria P, Farha OK, Hupp JT, Snurr RQ. A thermodynamic tank model for studying the effect of higher hydrocarbons on natural gas storage in metal-organic frameworks. Energy Environ Sci 2015;8:1501–10. [58] Myers A, Prausnitz JM. Thermodynamics of mixed‐gas adsorption. AIChE J 1965;11:121–7. [59] Kayal S, Sun B, Chakraborty A. Study of metal-organic framework MIL-101(Cr) for natural gas (methane) storage and compare with other MOFs (metal-organic frameworks). Energy 2015;91:772–81. [60] Fu J, Tian Y, Wu J. Seeking metal-organic frameworks for methane storage in natural gas vehicles. Adsorption 2015;21:499–507. [61] Liao J, Yazaydin AO, Yang S, Li F, Ding L. Molecular simulation studies of hydrogen enriched methane (HEM) storage in covalent organic frameworks. Microporous Mesoporous Mater 2016;231:138–46. [62] Taddei M, Steitz DA, Van Bokhoven JA, Ranocchiari M. Continuous‐flow microwave synthesis of metal-organic frameworks: a highly efficient method for large‐scale production. Chem – Eur J 2016;22:3245–9. [63] Ren J, Dyosiba X, Musyoka NM, et al. Green synthesis of chromium-based metal-organic framework (Cr-MOF) from waste polyethylene terephthalate (PET) bottles for hydrogen storage applications. Int J Hydrogen Energy 2016;41:18141–6. [64] Schoenecker PM, Carson CG, Jasuja H, Flemming CJJ, Walton KS. Effect of water adsorption on retention of structure and surface area of metal-organic frameworks. Ind Eng Chem Res 2012;51:6513–9.

Further Reading Faramawy S, Zaki T, Sakr AA-E. 2016. Natural gas origin, composition, and processing: a review. J Nat Gas Sci Eng 2016;34:34–54. Li B, Wen H-M, Zhou W, Xu JQ, Chen B. 2016. Porous metal-organic frameworks: promising materials for methane storage. Chem 2016;1:557–80. PrajwalK BP, Ayappa G. 2014. Evaluating methane storage targets: from powder samples to onboard storage systems. Adsorption 2014;20:769–76. Wang X, Fordham S, Zhou H-C. 2015. Metal-organic frameworks for methane storage. In: Liu JL, Bashir S, editors. Nanomaterials for sustainable energy. Washington, DC: American Chemical Society; 2015. p. 173–91.

Relevant Websites https://materialsproject.org/ The Materials Project, open web-based access for predicted materials. For more information, visit. http://webcsd.ccdc.cam.ac.uk/index.php The online portal to the Cambridge Structural Database, WebCSD. For more information. http://www.energy.gov/natural-gas The U.S Department of Energy. For more information, visit.

2.8 Magnetic Materials RK Kotnala and Jyoti Shah, CSIR-National Physical Laboratory, New Delhi, India r 2018 Elsevier Inc. All rights reserved.

2.8.1 2.8.2 2.8.2.1 2.8.2.1.1 2.8.2.1.2 2.8.2.1.3 2.8.2.1.3.1 2.8.2.1.3.2 2.8.2.1.4 2.8.2.1.5 2.8.2.1.6 2.8.2.1.7 2.8.2.1.8 2.8.2.1.9 2.8.2.1.10 2.8.2.1.11 2.8.3 2.8.3.1 2.8.3.2 2.8.3.3 2.8.3.4 2.8.3.5 2.8.4 2.8.4.1 2.8.4.2 2.8.4.3 2.8.4.4 2.8.4.5 2.8.5 2.8.5.1 2.8.5.1.1 2.8.5.1.2 2.8.5.1.3 2.8.5.1.4 2.8.6 2.8.6.1 2.8.6.2 2.8.6.3 2.8.6.4 2.8.6.5 2.8.6.6 2.8.6.7 2.8.6.8 2.8.6.8.1 2.8.6.8.2 2.8.7 2.8.7.1 2.8.7.2 2.8.7.3 2.8.7.4 2.8.8 Acknowledgment

204

Introduction Background/Fundamentals Energy Conservation Through Magnetic Materials Energy saving in magnetic computing technology Multiferroic materials for low power devices Transformer core electrical steel materials in minimizing power losses Electrical steel production Energy saving in transformers by electrical steel Magnetic materials for high frequency power loss Permanent magnetic materials for energy efficient electric motors Power saving in refrigeration by magnetic material Global energy scenario Solar cell Hydropower Wind energy Fuel cell Systems and Applications Water Splitting for H2 Gas Production Electrolysis Photosynthesis Photocatalyst Photobiosynthesis Analysis and Assessment Water Splitting at Room Temperature on the Surface of Metal Oxides Protonic Conduction in Absorbed Water Layers Protonic Conduction in Magnesium Ferrite Due to Water Molecule Adsorption Magnetic Behavior of Magnesium Ferrite Generation of Electric Current Due to Water Splitting Generates Electrical Power in Lithium Substituted Magnesium Ferrite Illustrative Examples or Case Studies Energy Sources Using Water Precursor/Byproduct Fuel cell Seawater battery Cement/soil battery Energy harvesting by surface energy of condensed water droplets Results and Discussions Porous Lithium Substituted Magnesium Ferrite Water Splitting by Lithium Substituted Magnesium Ferrite at Room Temperature Trapped Hydronium Ion Imaging by Electrostatic Force Microscopy Chemidissociated Water on Li-Magnesium Ferrite by Fourier Transform Infrared Spectroscopy Collection of Hydroxide and Hydronium Ions Produced by Lithium Substituted Magnesium Ferrite V–I Performance of Hydroelectric Cell Ionic Current Flow on Water Splitting in Hydroelectric Cell by Nyquist Plot Hydroelectric Cell Byproducts Zinc hydroxide Hydrogen gas Future Directions Permanent Magnet Energy Clean Energy and Sustainable Environment Environmental Magnetism New Magnetic Material as Green Energy Device – Hydroelectric Cell Invention Conclusions

Comprehensive Energy Systems, Volume 2

205 206 206 206 206 207 207 207 208 208 208 209 210 210 210 210 211 211 211 211 212 212 212 212 213 214 216 217 218 218 218 218 218 219 219 219 220 221 222 222 223 224 225 225 227 227 229 229 230 230 230 230

doi:10.1016/B978-0-12-809597-3.00219-4

Magnetic Materials References Relevant Websites

231 234

Nomenclature AC A/m AMR BET (BH)max BJH CGO CO CO2 COP21 DC DFT DI DMFC DOE US DRI EFM EHV EIS EMF FTIR GE GMR GO Gt/Yr HDD HEC HER HGO H 2O HRTEM Hz IUPAC kbits/in2 kg kHz kOe

2.8.1

205

Alternating current Ampere per meter Anisotropic magnetoresistance Brunauer–Emmett–Teller (Magnetic induction  magnetic field)maximum Barrett–Joyner–Halenda Conventional grain oriented Carbon monoxide Carbon dioxide Conference of Parties in 2015 Direct current Density function theory Deionized Direct methanol fuel cell Department of Energy, United States Direct reduced iron Electrostatic force microscopy Extra high voltage Electrochemical impedance spectroscopy Electromotive force Fourier transformation infrared General Electronics Giant magnetoresistance Grain-oriented Gigaton per year Hard drive disk Hydroelectric cell Hydrogen evolution reaction Highly grain oriented Hydrogen dioxide High resolution transmission Hertz International Union of Pure and Applied Chemistry Kilobits per inch square Kilogram Kilohertz Kilo Orested

kV kVA kW LED LEED LEIS Li-magnesium ferrite mA mA/cm2 mAh MGOe MHz MR MRAM MV mV MVA mW mW/cm2 NADPH NdFeB NEMA NO PME pW RH SAED SCWG SD SEM SHE SOFC SSD STM T TB Tbits/in2 UV

Kilovolts Kilovolts ampere Kilowatt Light emitting diode Low energy electron diffraction Low energy ion scattering Lithium substituted magnesium ferrite Microampere Mill ampere per centimeter square Milliampere hour Mega Gauss Orested Megahertz Magnetoresistance Magnetic random access memory Medium voltage Millivolt Megavolts ampere Milliwatt Microwatt per centimeter squared Nicotinamide adenine dinucleotide Neodium iron boron National Electrical Manufacturers Association Nonoriented Polymer electrolyte membrane Picowatt Relative humidity Selected angle electron diffraction Supercritical water gasification SanDisk Scanning electron microscopy Standard hydrogen electrode Solid oxide electrolyte Solid state drive Scanning tunneling microscopy Tesla Terabits Terabits per inch square Ultraviolet

Introduction

Magnetic materials are being efficiently used for saving electrical energy, which in turn cuts down the harmful gas emissions in the environment. Even the nonpolluting renewable energy sources like windmills, hydropower dams, etc. utilize permanent magnetic materials in generators to convert energy [1]. According to the US Greenhouse Gas Emission Report the highest 30% greenhouse gas emission is produced by the electricity sector using coal combustion followed by 26% from the transportation sector [2]. The increasing demand for clean technologies will lead to huge growth of magnetic materials over the next 25 years. Thus efficiency of renewable energy sources significantly depends on high energy density magnets, which are being partially fulfilled by permanent rare earth magnets. Presently, the highest energy density of 50 MGOe with coercivity 12 kOe at room temperature has been reported for neodium iron boron (NdFeB) magnet [3] and many efforts have been initiated to replace the rare earth component from high energy density magnetic material. Contrary to permanent magnets soft magnetic material reduces the electric losses in

206

Magnetic Materials

power generation and transformation for the electric grid [4]. High frequency 10-kW power converters have been already developed by Hitachi metals using nanocrystalline ferrite and amorphous alloy materials [5]. Energy produced by fossil fuel and thermal power stations has to be distributed by magnetically linked power transformers extra high voltage (EHV) (1000 MVA–250 kVA), distribution transformers medium voltage (MV) (250 kVA–20 kVA) and utilization kV (less than 20 kVA), low voltage (LV) transformers. Current is transformed to magnetic field lines concentration in the transformer core connecting primary and secondary coils. Electrical steel cores are one of the key factors in the performance and efficiency of power transformers. High frequency transformers need precise magnetic loss analysis and insulation compared to low frequency 50 Hz line transformers. Medium frequency high power transformers are key components used in offshore wind plants and future distribution system grids [6]. Some magnetic material produces heat by a magnetization–demagnetization cycle known as the magnetocaloric effect [7]. The magnetocaloric effect is being utilized to establish magnetic refrigeration systems [8]. In urban areas, refrigerators and air conditioning consume highest electrical energy. Such magnetic refrigeration would be able to provide a huge reduction in energy consumption. General electric (GE) has demonstrated 44-degree cooling attained by a magnetic refrigeration device [9]. Magnetic refrigerators have been found 20%–30% more efficient than compressed gas refrigeration. A shift in mass transport systems from gasoline-based engines to electric motor vehicles is getting more popular to protect the environment. It is well known that carbon dioxide, carbon monoxide, nitrogen oxides, and hydrocarbons are released when fuel is burnt in an internal combustion engine and these are emitted through the vehicle exhaust. But in electrically motorized automobiles no such emissions are exhausted. In magnetically levitated trains lot of energy is saved by using magnetic superconductors compared to electrified trains. More usage of electric motor vehicles has initiated a focused research to improve efficiency of electric motors and distribution transformers. In this direction a big scope is to enhance energy density of permanent magnets used in the electric motor of a transport system. Distribution transformer efficiency can be improved further by using a better quality electrical steel magnetic core material with low power losses. A huge share of magnetic materials is used in automobiles, transformers, and the power sector, which demands special attention to a new class of magnetic materials. Instead of utilizing the magnetic properties of magnetic materials for reducing energy consumption, in this chapter we introduce electrical energy generation by a magnetic material without using its magnetic property. This has opened a new field for magnetic materials as a green energy source. The oxygen deficient and nanoporous character of magnesium ferrite has been utilized for water molecule dissociation at room temperature. Further, electrode chemistry has been adopted to collect dissociated hydroxide and hydronium ions to develop current/voltage. The material along with electrodes provides electricity by water, and the invented device has been termed the hydroelectric cell (HEC).

2.8.2

Background/Fundamentals

2.8.2.1

Energy Conservation Through Magnetic Materials

Magnetic materials have always been strategic materials. From traditional navigation, to microwave material, to the frontier field of spintronics, magnetic materials have been explored extensively. Magnetic materials in alloy form are being used in electrical steel, permanent magnets, magnetic memory, and magnetocaloric effect. Oxide magnetic material, soft ferrite, has wide applications in high frequency, low frequency regimes. Hexagonal ferrites are cost-effective permanent magnet alternatives to the rare earth magnets. Magnetic material thin films in few-nanometer thicknesses have opened the emerging area of spintronics. Improvement in efficiency of magnetic material by research has enhanced the efficiency of machines.

2.8.2.1.1

Energy saving in magnetic computing technology

Low energy consumption in magnetic computing is an emerging field. Electrical conductivity changes with applying magnetic field in ferromagnetic metals, half metals, and alloys are known as anisotropic magnetoresistance (AMR). Read-out head directly converts information recorded in a magnetic storage medium into electrical signal by change in magnetoresistance (MR). Giant magnetoresistance (GMR) in Fe/Cr multilayers has been a great breakthrough in magnetic computing [10,11]. The GMR effect has revealed that spin dependent scattering can induce a significantly large 180% MR change at room temperature in a Fe/MgO/Fe magnetic tunnel junction [12]. MR brought the revolution in high performance magnetic random access memory (MRAM). An advantage of GMR over existing ferrite core memory and magnetic bubble is that it consumes low power. GMR devices also operate at fast speed. The magnetic storage hard disk drive technology (HDD) is still very important due to its lower cost compared to solid state drive (SSD). From the mid-1950s the number of bits per unit area has increased from 2 kbits/in2 to 1.4 Tbits/in2 in 2015 using heat-assisted magnetic recording [13]. In the present scenario one can fit 2 TB onto a SD card the size of a postage stamp. The drastically increased areal density in magnetic memory has reduced the size, access time, weight, and price of the disk.

2.8.2.1.2

Multiferroic materials for low power devices

Smart technologies demand low power nanodevices, such as fast switching material, gyrators, current transformers, actuators, etc. To combine various features within one microchip, composite materials like multiferroics has been extensively studied. Multiferroic materials are the composites of ferroelectric, ferromagnetic/antiferromagnetic, and ferroelastic materials in which two or

Magnetic Materials

207

more than two ferroic orders coexist simultaneously [14]. A class of multiferroic materials in which electric polarization is controlled by applied magnetic field and vice versa are magnetoelectrically coupled multiferroics. Electric polarization of magnetoelectric material by applied magnetic field can reduce the elaborative system of electric field generation and ultimately reduces energy consumption of the system. Recently, a multiferroic read–write cell of strain mediated magnetoelectric heterostructure has been investigated [15]. Strain developed on magnetic materials due to magnetic field, known as magnetostriction, when combined with piezoelectric material, provide strain, which in turn induces electric charges. Voltage induced elastic strain is used to generate different magnetization orientations. A large piezoelectric actuator hybrid device of Ni/PbZrTiO3 has been investigated extensively using simultaneous magnetooptical Kerr effect and magnetotransport measurements [16]. Pyroelectric effect of multiferroic composite of ferroelectric and ferromagnetic materials has been explored for efficient thermal energy harvesting. Enhancement of Carnot efficiency of 6.45% has been estimated at room temperature in Co/BTO multilayer composite, which depends upon strong magnetoelectric coupling [17]. Strong atomic level strain mediated magneto-electric coupling provides less power dissipation in gyrator [18,19]. A high power efficiency of 73.9% has been observed in Terfenol-D and PbZrTiO3 composite [20].

2.8.2.1.3

Transformer core electrical steel materials in minimizing power losses

The major application of magnetic material is in electrical power generation and distribution. To achieve high magnetic induction and low loss magnetic materials, research is still going on. By reducing losses in transformer core magnetic material the power handling capacity would be increased without increasing size of the transformer. Initially, iron was the exclusive soft magnetic material for transformer core due to high flux density in comparison to other magnetic material. But due to high conductivity of iron metal, eddy current losses are very high. The small amount (3%) of silicon addition in iron was demonstrated to reduce magnetic losses by a factor of 4 due to increased resistivity. However, silicon content is also critical for the performance of electrical steels since it reduces anisotropy, magnetostriction, permeability and increases brittleness [21,22]. Hysteresis loss is also the dominating loss at power frequency. Hysteresis losses are the function of impurities, grain orientation, grain boundaries, and domain wall spacing, which originate coercivity. Coercivity is closely related to the permeability of the magnetic material. Low coercivity, high resistivity, and high permeability are essential for reducing power loss in electrical steel [23]. Presently in global electricity generation 5% is dissipated as heat/core loss from transformers. Moreover operation of transformers releases 35,000 t of sulfur dioxide and 0.4 Gt of carbon dioxide into the air annually. Thousand tons of electrical grade steel have been produced per year for the electrical industries. Iron–silicon alloy gives high permeability with low electrical and magnetic losses suitable for electrical steel. The no-load loss in distribution transformer is 70%, which can be reduced with better magnetic performance of the transformer core [24,25]. By using energy-efficient transformers emission of CO2 can be significantly reduced [26]. 2.8.2.1.3.1 Electrical steel production Electrical steel is the highest soft magnetic material, produced around 97% by iron ore [27]. Annually 12 million tons of electrical steel are produced. Almost 5% of the global CO2 emission is produced by steel manufacturing. Raw iron is made by reducing iron ore by reacting it with air and coke at high temperature (B12001C). About half of the carbon in the coke oxidizes by taking oxygen from the iron ore to make CO2. Around 1% of the carbon from the coke remains in the raw iron to provide the source of the carbon to produce steel. The rest of the coke provides heat to melt the iron, producing more CO2. Thus, a large amount of CO2 is released by today’s steel industry. Coal acts as a reducing agent providing a source of carbon to be incorporated to make steel as well as a source of energy to drive the process. This process is known as iron reduction and produces molten iron that is converted to steel. Substantial reduction in CO2 emission is not feasible if only conventional technologies are still used. Serious efforts are going on worldwide to find ways of reducing carbon emissions from steel making. Several processes are available that can reduce the carbon dioxide emission from steel production. Major reduction in CO2 emission in steel production can be accomplished by use of electric arc furnaces or by direct ore reduction using hydrogen. Use of alternative sources of energy that are CO2-less resources, like natural gas, renewable electrical power, or renewable gases, for example, synthetic methane or hydrogen, is preferred. This depends on the availability of cheaper electric energy from a renewable source of energy, along with quantity of scrap steels, which can be melted to produce steel. Some of the methods available are described in Fig. 1. These emerging technologies can enable greater independence from coal reducing CO2 emissions produced from the steel industry. 2.8.2.1.3.2 Energy saving in transformers by electrical steel Electrical steel has high demand as a core material for electromagnetic induction in transformers. Transformers convert alternating current (AC) of one voltage into AC of another voltage by means of electromagnetic induction at an invariable frequency [28]. Electrical steel is mainly responsible for the efficiency of the transformer. Electrical steels are being chosen based on their magnetic losses at alternating frequency. Magnetic losses of electrical steel are divided into three parts: (1) quasi-static loss, (2) parasitic loss, and (3) anomalous loss. Silicon electrical steel was initially obtained in 1900 by alloying iron with silicon. Resistivity of alloy substantially increased, thus decreased the loss due to eddy currents and reduced hysteresis upon reversal magnetization [29]. At low frequency generally hysteresis loss dominates; anomalous loss becomes significant at medium frequency, while at high frequency eddy current loss becomes highest. In 1906 commercial production of silicon steel as a magnetically soft material was started. In the same year licensed steel with 3 wt% Si was started in the United States [30]. In 1920 dependence of intensity of magnetization in iron single crystal on the crystallographic directions was discovered later formulated in steel [31]. Goss in 1935 has invented grain-oriented (GO) silicon steel and its production process [32]. GO silicon steel consists of the easiest directions of magnetization along /001S crystal directions. Goss orientation {110}//001S is technologically realized to minimize magnetic

208

Magnetic Materials

Electrical arc furnace Direct reduction of iron ore using hydrogen

Top gas recycling blast furnace Electrolysis of iron ore

Direct reduced iron process

Fig. 1 Various techniques for steel production industries to reduce CO2 emission.

losses in electrical transformers. Around 80% of electrical steel production is nonoriented (NO) grades, 20% being GO. GO steels have been divided into conventional grain oriented (CGO) and highly grain oriented (HGO) grades [33,34]. HGO grade steel produces controlled grain growth during high temperature annealing. By improvement in the sharpness of {110}//001S orientation, hysteresis losses decrease due to increase of magnetic induction. The transformer cores are produced from anisotropic electrical steels nowadays. The level of its properties is evaluated in terms of specific magnetic loss in W/kg at magnetic induction 1.7 T and frequency 50 Hz and electromagnetic induction measured at fixed 800 A/m magnetic field intensity [35]. Attempts on high silicon electrical steel started more than 40 years ago to develop 6.5 wt% silicon electrical steel [36]. The alloy has zero magnetostriction and good high frequency characteristics but a major challenge is its high brittleness. Magnetic loss due to AC magnetization in the transformer core produces 5% as waste heat of total energy production [27]. Now strict regulations have been made to transformers in view of global climate change prevention. Recently, Top Runner transformers introduced in Japan showed a large effect in CO2 reduction. The new Siemens National Electrical Manufacturers Association (NEMA) Premium Efficiency Transformer designation requires 30% fewer losses than existing Department of Energy, United States (DOE US) requirements.

2.8.2.1.4

Magnetic materials for high frequency power loss

In high frequency communication range 103–1011 Hz ferrites are being used due to their high electrical resistivity. At high frequencies, the use of soft ferrites is relatively more systematic with respect to other circuit components. An important factor in the choice of soft ferrites is that they are generally less expensive than magnetic metals and alloys. Soft ferrites are the best option for the core material for operating frequencies from 10 kHz to a few hundred MHz with proper combination of low cost, high inductor quality, high stability, and low volume. Moreover, high permeability and time/temperature stability are some additional important features that have expanded the usage of soft ferrites in high frequency and delay lines, broadband transformers, adjustable inductors, quality filter circuits, and other high-frequency circuit electronics. Ferrites, due to their low magnetic losses, are utilized as load coil band pass filters for long distances. Thus ferrite inductors have a significant role in telecommunication. In wireless power transmission using power line communication systems, ferrites are being used as antennas [37]. Fig. 2 illustrates the applications of magnetic materials.

2.8.2.1.5

Permanent magnetic materials for energy efficient electric motors

The applications of permanent magnets range from electric motors, magnetic bearing, generators, loudspeakers, ore separators, water filters, electric watches, and microwave tubes. The function of a permanent magnet is to generate a magnetic field in an air gap of a magnet system. The quality of a permanent magnet is defined by maximum energy product (BH)max. The energy product is a measure of the magnetic flux that can be produced by the magnet in a given volume. Rare earth based magnet SmCo5 was made of energy product 20 MGOe in 1960. Magnetically hard magnets in a medium energy product of hexagonal ferrite, such as barium ferrite and strontium ferrite, were made due to low cost. Renewable energy source windmills of typical 225 kW power have rotor weight approximately 3.3 t by using 536 kg NdFeB and 10.8 t using 4789 kg ferrite permanent magnet respectively [38]. This shows that the quality and quantity of permanent magnets has a major role in the efficiency of windmills. Table 1 compares the demand for magnets in global market.

2.8.2.1.6

Power saving in refrigeration by magnetic material

Magnetic material emits or absorbs heat in the presence of a magnetic field under adiabatic conditions as a result of variation of its internal energy [39]. Refrigerators and air conditioners, commonly used in houses and stores, consume lot of energy based on the gas compression system. Magnetic refrigeration is an emerging technology exhibiting higher cooling efficiency, and is more environmentally friendly compared to gas-compression refrigeration. This is because adiabatic cooling/heating does not use harmful, ozone-depleting coolant gases [40,41].

Magnetic Materials

209

Gyrators Nonvolatile memory storage

Transformer cores

Electric motors

Magnetic materials

Read/write head

Microwave absorber

Ferrite inductor

Adiabatic cooling

Fig. 2 Various applications of magnetic materials can conserve energy by improving material.

Table 1 Global production of rare earth permanent magnet and ferrite magnet with price for the year 2016 Magnetic material

Production (tons)

Year

$Million

Permanent magnet Soft ferrite

1,00,000 5,65,000

2016 2016

13,220 34

Worldwide CO2 emission in gigatonns

Source: Reproduced from Benecki WT. The global permanent magnet industry. A consultancy confidentially serving the global permanent magnet industry. In: The international forum on magnetic applications, technologies and materials; 2017.

35 34 33 32 31 30 2013

2020 Years

2025

2050

Fig. 3 Worldwide energy related CO2 emissions vs. years.

2.8.2.1.7

Global energy scenario

Energy has joined the rank of food, shelter, and clothing: three essential commodities for mankind, going beyond its basic usage for human comfort. Energy consumption per capita has become the barometer of a nation’s prosperity. Contrary to high energy demands, mankind needs clean air, water, and environment for good health. In the present situation carbon dioxide (CO2) emission has scaled to 32.2 Gt/year in our atmosphere and is expected to be at 43 Gt/year by 2016 [42]. With this rate of CO2 release into the atmosphere, global emissions would reach to 34.6 Gt as represented in Fig. 3. Such huge emission of CO2 has an adverse effect on global climate change from reduction of ice in Arctic to the atmospheric pollution [43]. The Conference of Parties in 2015 (COP21) meeting in December 2015 in Paris on global climate change has agreed to meet the goal of 21C increase in climate globally by energy sector through making deep cuts in CO2 emission [44]. Greenhouse gas emission has increased the public concern about climate change. To make energy efficient devices and use of clean energy will reduce deadly air pollution and provide a clean environment. Therefore, in the current scenario, the energy generation focus has shifted more toward alternative

210

Magnetic Materials

2% 3.7%

Fossil fuels 1.2% Hydropower

16.6%

Wind energy

76.3%

Biomass

Solar photovoltaic

Fig. 4 Worldwide estimated energy share of electricity production in 2015.

natural resources, such as solar, wind, hydroenergy, etc. Existence of sustainable clean renewable solar, wind, and hydro energy resources are much larger than the global energy demand. Globally an estimated 147 GW of renewable power capacity have been added in 2015 for electricity [45]. Use of renewable alternative sources not only saves money but also ensures protection of the environment and induces future generations to look for more efficient energy systems and means. The share of renewables in total energy consumption is not growing as quickly as it should be. Solar photovoltaic (PV) and wind energy were the most dynamic markets in past years. Still hydropower provides the largest share in total power capacity and generation globally. Although, bioenergy is still the principal source for the industries and transport sectors [11]. The share of renewable energy globally for the year 2015 has been depicted in Fig. 4. Harnessing solar energy would yield a never-ending energy supply but the difficulty associated with it is to convert solar energy in an efficient and cost effective way. Thousands of research papers and reports are being published each year on this subject all over the world. However, solar cell cost is still very high, and there are limited hours of sunshine in a day, varying remarkably with geographical location. Opportunities exist for technologies that promise either significantly higher energy conversion efficiencies or significantly lower processing costs. To mitigate the cost, as well as geographical and weather constraints of renewable energy sources, there has been increasing research trend toward alternative clean energy sources like fuel cells, biomass, etc.

2.8.2.1.8

Solar cell

The sun is a huge source of energy; about 60% of the sun’s energy reaches the earth’s surface. Solar energy has been converted into electricity by using PV solar cells. Solar cells made of silicon convert photons into electricity while the sun is shining. The global solar PV power generation reached to 227 GW in 2015 [10], but their high manufacturing cost limits their broad application. Solar radiation at the surface depends on time, latitude, atmosphere, and surface conditions. For a silicon solar cell, it appears 3 years of (non carbon emitting) power generation is required to compensate for the CO2 emitted during fabrication [46]. Moreover solar cell power can be exploited during the day; otherwise storage batteries are used. Hence even today its use is unaffordable for the average person.

2.8.2.1.9

Hydropower

Hydropower is used to generate electricity by releasing stored water in dams to run turbines coupled with generators. Hydropower capacity/potential depends upon the size of the dam, its ability to store volume runoff, as well as the water and its height of fall. Hydropower plants neither consume resources to create electricity nor pollute the air, land, or water. Hydropower has a significant share in global renewable energy production. In the 1920s, hydroelectric plants were 40% efficient but now the efficiency of hydroelectric plant has reached 90%. The global hydropower capacity reached to about 1064 GW in 2015 [10].

2.8.2.1.10

Wind energy

Wind on the Earth’s surface has been utilized through windmills that convert wind power into mechanical energy to run a turbine attached to a dynamo, which in turn generates electricity. Modern wind turbine aerodynamic performance has been linearly improved over past 20 years. To capture the more energetic winds at higher elevated height the size of wind turbines has increased tremendously. However, transportation constraints limit the practical size of wind turbines. Wind turbine subsystems include many components and thus fault detection and isolation presents some complexity. In spite of huge expensive wind turbine subsystems globally, a record 63 GW has been added in 2015 only for a total wind power of about 433 GW [10]. Besides, a major limitation of wind turbines is that installation is possible only near shore or very selective windy locations.

2.8.2.1.11

Fuel cell

Among the emerging clean energy technologies, fuel cells, developed by Sir William Robert Grove [47] (1811–96), are growing rapidly. A fuel cell basically consists of two electrodes on either side of a polymeric electrolyte membrane. Hydrogen and oxygen

Magnetic Materials

211

are purged over each of the electrodes, loaded with catalyst, and through a chemical reaction, electricity, heat, and water are produced. Fuel cells are generally based on H2 purged polymer electrolyte membrane (PME), direct methanol fuel cell (DMFC), and solid oxide electrolyte (SOFC) [48]. Fuel cell technology is a highly interdisciplinary system. Efficiency of H2-PEM fuel cells relies on various parameters from fuel (H2 and O2) pressure, temperature, gas concentration to membrane operating temperature and its blocking by water vapor. The risk involved in storage, distribution, and production of hydrogen directs the use of methanol instead of pure hydrogen. Fuel leaks through membrane and Ohmic losses are the two major contributors to irreversibilities in the DMFC. High temperature ceramics are also being utilized as SOFCs due to less possibility of poisoning of material. However, its operational temperature makes it impractical for general applications. The expensive cost of fuel cells and greater safety concerns compared with other forms of power conversion are still the major disadvantages.

2.8.3 2.8.3.1

Systems and Applications Water Splitting for H2 Gas Production

We have learnt from nature how to convert available natural resources into energy productive fuel and oxygen through existing photosynthesis processes. Researchers have tried to achieve artificial photosynthesis by fabricating p–n junctions, photocatalytic material, and artificial bacteria. Among all methods artificial photosynthesis [49], photocatalytic [50], and photobiological [51] processes using solar energy have been found to be better alternatives to achieve clean sustainable water splitting. Recent literature reports efficient and economical ways of water splitting processes for hydrogen production as depicted in Fig. 5. Water splitting can be accomplished via techniques, such as electrolysis, photosynthesis, thermochemical, and biophotolysis.

2.8.3.2

Electrolysis

Hydrogen gas production from pure water is rarely feasible as an electrolyte. Water is a poor ionic conductor and hence it presents a high Ohmic overpotential. To proceed with the water splitting reaction practically, the conductivity of the water is inevitably increased by the addition of acid/alkali. Most of the electrolyzers use a concentration of 25%–35% KOH in water [52]. The history of water electrolysis is 215 years old, and is done by electrolytic splitting of water molecules on applying direct current (DC) to the electrodes. Nicholson and Carlisle in 1800 were the first to discover the ability of electrolytic water splitting [53]. At that time electrolysis of water was exploited to obtain hydrogen and oxygen. On applying DC the electrochemical reactions in alkaline electrolyte taking place at anode and cathode are given as:
ð1Þ Cathode: 2H2 O þ 2e ¼ H2 þ 2OH E0 ¼ 0:40 V vs: SHE Anode: 2OH ¼

1 O2 þ H2 O þ 2e 2

E0 ¼


0:83 V vs: SHE

ð2Þ

Advanced alkaline electrolyzer cells operate at low voltage (1.6 V) with high current densities (about 2 A/cm2) producing hydrogen gas with a nominal purity of 99.8% [54].

2.8.3.3

Photosynthesis

Using sunlight for water splitting into H2 and O2 has been extensively explored since H2 is being used as a direct fuel for the production of electricity by the fuel cell. Water has positive Gibbs free energy, and thus it is highly stable. Water cannot directly

Oxidation− reduction rection

Thermal decompostion

Electrolysis

Water splitting methods

Photobiological

Fig. 5 Different techniques of water molecule splitting.

Photo electro chemical

212

Magnetic Materials

absorb the photons of the solar spectrum, therefore some photocatalytic material is added to initiate a photochemical reaction. Overall water splitting by a photocatalyst typically refers to a photoactive colloidal suspension, which produces hydrogen and oxygen in close proximity to each other. Since 1977 several sacrificial model systems of H2 production from water have been proposed [55]. To a great extent 130 metal oxides have been identified as possible photocatalysts for water splitting [56]. Initially photocatalytic titanium dioxide was used for the production of hydrogen and oxygen using UV light and a cocatalyst [57,58]. Other metal oxides have also been investigated as possible visible light absorbing photoanodes [59–61]. Recently, metal oxide heterostructures improved the performance of photoanode material development. There are several practical implications to absorb solar energy and splitting of water molecules [62]. Hydrogen and/or oxygen production from water by visible light depends on photon intensity, pH, concentration photosensitizer, proton source, catalysis, and sacrificial electron donor of the system. Such complexity does not allow easy and cost-effective techniques for water splitting.

2.8.3.4

Photocatalyst

Water molecule dissociation by using the oxidation–reduction property of metal oxides at high temperature has been also explored. In thermochemical cycle, catalytic materials have been used due to redox activity and ability to host oxygen vacancies in nonstoichiometric crystal lattices [63]. In this process metal oxide redox pair XOox/XOred is utilized for water splitting in two-step thermal cycle. In the first step, a high valence metal oxide is reduced to a low valence metal oxide by the high temperature heat energy, releasing oxygen. In the second step the oxygen deficient metal oxide then reacts with water at a lower temperature, releasing hydrogen and regaining stoichiometry. The thermochemical cycle reaction steps follow the two reactions 1 O XO ox -X red þ O2 2

ð3Þ

XO red þ H2 Oðg Þ-XOox þ H2 ðg Þ

ð4Þ

The redox property and hosting oxygen vacancies in crystal lattices are the main criteria for the reductive-oxidative thermocycle. Water splitting by iron oxide, zinc oxide, manganese oxide, and ferrites based on the thermocycle process has been explored extensively [64–66]. Spinel ferrites are the most stable structure among oxide materials and can withstand in reducing atmosphere without degradation. At high temperature under reductive atmosphere ferrites have the tendency to form nonstoichiometric compounds, though preventing their basic spinel structure [67]. The oxygen vacancies created in a spinel lattice tend to be filled by oxygen by splitting water molecules. Ferrites are promising redox materials due to the fact that the exchange interactions with the gaseous oxygen proceed reversibly and do not require high temperatures [68]. The highest water splitting conversion (62.5%) to H2 gas was reported by doped zinc ferrite. Splitting of water by spinel ferrite at moderate and workable operative high temperatures is also reported [69].

2.8.3.5

Photobiosynthesis

Photosynthesis by green plants produces oxygen and sugar in the presence of sunlight, water, and carbon dioxide. The hydrogen production by green algae is a method of photobiological water splitting. Photobiological water splitting is done in a closed photobioreactor with algae that produces hydrogen by utilizing solar energy [70]. The direction of photobiological hydrogen production research is focused generally on microalgae, both green algae and blue-green algae, which possess enzymes capable of acting as catalysts for the photoassisted production of hydrogen. Microalgae have the highest photosynthetic capability per unit volume [71]. They grow much faster than higher-level plants, with minimum nutrition requirements. Plants produce hydrogen under certain conditions. Higher green plants and algae absorb light and split water into oxygen and nicotinamide adenine dinucleotide (NADPH). The unique hydrogen production ability of green algae and blue green algae is due to the presence of certain enzymes that are absent in other plant species in which only CO2 reduction takes place. Recently, an artificial bionic leaf has been reported that is a combination of Co–P cobalt phosphorous as cathode (hydrogen evolution reaction, HER) and cobalt phosphate Co–Pi as (OER) utilized to split water molecules to produce H2, which has been converted into bioenergy by photosynthesis [51]. In this system water split to oxygen by cobalt phosphate (Co–Pi) catalyst and H2 is produced by a NiMoZn alloy at applied voltage of Eappl ¼3 V. Hydrogen generation from biomass using techniques, such as supercritical water gasification (SCWG), pyrolysis, and employment of microorganisms, have been also researched [72–74]. Using microorganisms for hydrogen production is sustainable and energy saving; moreover the strains can be cocultured or metabolically engineered to enhance the corresponding enzyme activity [75–77]. Hydrogen production is also being achieved via biophotolysis of water, photodecomposition of organic compounds, and photo and dark fermentation [78].

2.8.4 2.8.4.1

Analysis and Assessment Water Splitting at Room Temperature on the Surface of Metal Oxides

Dissociation of water molecules by the defective oxide surface at room temperature has been widely explored experimentally and theoretically. Surface conductivity of the oxide surface increases with increasing amount of water vapor. Strong chemidissociation of water molecules on the Al2O3(110) surface site consequently resulted in negative surface energy [79]. Certain metal oxide

Magnetic Materials

213

surfaces are very effective in dissociating water at room temperature. An important factor that affects water splitting on the surface is the position of the surface O 2p level relative to the top of the valence band. These oxide surfaces are characterized by having their surface O 2p level lying within the band gap. Adsorption energy per water molecule was found to be completely dissociating water molecules at all coverage on the surface determined by density functional theory [80]. The adsorption energy of a single water molecule was calculated to be 140.0 kJ/mol on Feoct2, atom of clean magnetite surface studied by density function theory (DFT) [81]. Thus a single water molecule spontaneously dissociates and a strongly exothermic reaction occurs on the Feoct2 termination of magnetite. Oxygen vacancies have a significant role in both molecular and dissociative water adsorption in TiO2 and ZnO [82–86]. The role of oxygen vacancies in water dissociation demonstrates by scanning tunneling microscopy (STM) that there is a direct correlation between oxygen vacancies and the water molecule. Water adsorbed gets dissociated in oxygen vacancies, proton transfers to a neighboring bridging oxygen atom by creating two bridging hydroxyl groups. Further using DFT water dissociation has been energetically found feasible only at the defect sites [87]. Scanning tunneling microscope tip parameter has been optimized to selectively adsorb H atoms in order to image the reaction of water molecule with rutile TiO2(110). Image contrast has been analyzed to distinguish between the oxygen vacancies and OH groups. That suggested minority defects played a major role in water molecule dissociation [88]. Role of surface oxygen in the dissociation process of water molecules has been estimated by density function calculations. In-plane oxygen atoms acted as intermediate sites for dissociation [89]. On a stoichiometric TiO2(110) surface, the direct hydrogen transfer process from adsorbate to bridging oxygen results in two hydroxyl groups. Bridging oxygen atoms are easily removed to produce vacancies that act as preferential adsorption sites; these vacancies give rise to two hydroxyl groups at bridging sites [90,91]. Water molecule dissociation at oxygen vacancies competes with water dissociation on defect-free regions such that one vacancy-assisted dissociation event cancels another dissociation event on defectfree regions. As a result, molecular adsorption becomes favored for low coverage on a surface where all vacancies have been hydroxylated [92]. For the perfect surface, dissociation occurs through H2O adsorption on a Ti ion followed by proton donation to a neighboring O ion. This leads to an OH group (on O atom) paired with an OH group terminally adsorbed on a Ti ion [93]. Water adsorption on zinc oxide occurs via oxygen atom bonded to the coordinately unsaturated zinc ion on the ZnO surface [94]. Chemisorption of the H2O molecule occurs on active zinc sites of (1010) surface rather than the oxygen sites. The surface terminated by O atoms interacts very weakly with water. Interaction of H2O with ZnO(0001) and (0001 ) polar surfaces, discarding the dissociation mechanism of H2O into individual OH and H ions [95] Ab initio calculations and hybrid density functional theory calculations on ZnO polar surface (0001) ruled out the adsorption of H2O completely. Deficiency of O atoms is confirmed from the analysis of X-ray photoelectron spectral data, the missing O atoms are the vacancies, responsible for the dissociative adsorption of the water molecule. High reactivity of O–ZnO surface toward water indicated large number of O vacancies. H2O is dissociated upon interaction with the surface on the oxygen vacancies, forming two OH-species per dissociated H2O molecule [96]. Oxygen anions exposed on the ZnO(0001) surface at step edges has been found responsible for water adsorption observed by STM images. H2O molecule binds to zinc cation, form Zn–OH bond, H atom of H–O–H leaving it to bind with undercoordinated oxygen anions. Increasing water exposure changed the surface morphology considerably to a more disordered arrangement [97]. The surface of epitaxial FeO(111) is observed to be chemically inert due to the absence of surface Fe atoms demonstrated by STM and low energy electron diffraction (LEED) techniques. While on epitaxial Fe3O4(111) film surface terminated by Fe atoms, water molecule dissociates resulted in adsorbed OH groups [98]. The water molecule dissociation has not been due to any defects, but is dependent on the Fe atoms exposed on the surface. Heterolytic dissociation has been suggested as an acid–base reaction where the coordinately unsaturated Fe site acts as a Lewis acid that attracts the lone pair of electrons of the water molecule. Room temperature water adsorption at a nondefective, stoichiometric Fe3O4(001) surface was reported as a low coverage of surface hydroxyl (OH ) due to H adsorbed on the oxygen sublattice [99]. Isolated water molecules adsorb in dissociative state at both defect sites and regular terrace by theoretical molecular dynamics simulations [100]. Such isolated molecule dissociation at oxygen vacancies/surface defect site or nondefective surface on Fe3O4 (001) surface has been estimated by ab initio calculations [101]. Higher coverage leads to a hydrogen-bonded networking with alternate molecular and dissociated species. Water dissociation significantly takes place at defects and cation rich faces [102,103]. Cation sites exposed on the surface demonstrate high chemical reactivity in ZnO [104].

2.8.4.2

Protonic Conduction in Absorbed Water Layers

The splitting of water molecules on metal oxide surface is measured directly by conductivity of the material at different relative humidity (RH). Protonic conduction in ionized water molecule layers on metal oxides arises due to their surface reactivity and porous microstructure. As dry oxides are kept in contact with humid air, water molecules chemisorb on the available sites of the oxide surface, mainly at the neck parts of crystal grains, by a dissociative mechanism to form two hydroxyl ions for each water molecule. The hydroxyl ion adsorbs on the metal cation present in the surface layer of the particles, which possess high local charge density and a strong electrostatic field, and the proton reacts with an adjacent surface O2 group to form a second OH group [105]. The chemisorbed layer, once formed, is not further affected by exposure to humidity. When the first layer of water molecules is formed, subsequent layers of water molecules are physically adsorbed on the first hydroxyl layer. The physisorbed water easily dissociates to form H3O þ because of the high electrostatic fields in the chemisorbed layer. Water molecules in the succeeding physisorbed layers are only singly bonded and form a liquid-like network [106]. Porous microstructure allows water vapor to diffuse easily through the open pores at the surface and condensation of water vapor in the capillary-like pores. Thus due

214

Magnetic Materials

H Protonic conduction

H

O

O

H+

H

H

H

Hydrogen bonding

O H

H O

Physisorbed layer

O

H O

H O

M+

M+

H+ H

H O H O

M+

Chemisorbed layer

Surface cations

Vo− O

O

O

Fig. 6 Grotthuss chain reaction: chemisorbed OH layer on metal surface cations and protonic conduction in physisorbed water layer through H-bonding.

to high electrostatic field H þ ions hop from one water molecule to the next neighboring molecule forming a hydronium ion (H3O þ ) as indicated in following reactions [107]: 2H2 O3 H3 Oþ þOH

ð5Þ

H3 Oþ 3 H2 O þ Hþ

ð6Þ

Hence, the proton starts conducting in a chain of physisorbed layers as demonstrated by the schematic diagram in Fig. 6, known as the Grotthuss chain reaction.

2.8.4.3

Protonic Conduction in Magnesium Ferrite Due to Water Molecule Adsorption

The protonic conduction in physisorbed water molecules by porous and defective magnesium ferrite has been represented due to water molecule dissociation at room temperature. Magnesium ferrite belongs to a class of compounds having the general formula AB2O4 (MgFe2O4) and crystallizing in a spinel structure. The structure may be characterized as a slightly distorted cubic closely packed array of oxygen atoms containing metal ions in prescribed interstitial positions. The lattice sites occupied by metal ions are of two types, commonly designated as A and B, and distinguished by the way in which the nearest oxygen neighbors are arranged. The configuration of surrounding oxygen is tetrahedral in the case of A sites and octahedral in the case of B sites. The unit cell contains eight molecules with 24 metal ions distributed among 16 octahedral or B sites and eight tetrahedral or A sites. In the so-called normal or regular spinel the divalent metal ions occupy the A sites and the trivalent metal ions the B sites. Without violating the symmetry requirements it is possible for the divalent ions on A sites to exchange places with trivalent ions on B sites. The extent to which this process of inversion occurs depends in some cases on the method of preparation. The inverted spinel is then one in which the A sites are occupied solely by trivalent and divalent metal ions. A spinel lattice of magnesium ferrite is shown in Fig. 7. Spinel ferrite is a highly stable compound that can withstand reduced conditions without degradation. Tetrahedral cations have low coordination numbers, and thus form strong bonds with oxygen atoms. The octahedral cation is surrounded by greater numbers of oxygen, and the octahedral B–O bonds become weaker thus more polar. Thus octahedral cations provide acidic sites exposed on the surface. Magnesium ferrite is a ternary spinel compound, nonstoichiometry in oxygen inherently present in the spinel structure due to imbalance in oxygen and cation components. Magnesium ferrite is an oxygen deficient porous material, possesses large specific surface area, and has high electrical resistance of the order of 108 Ω. Magnesium ferrite has the lowest density, 4.5 g/cm3, among all ferrites. Its porosity can be increased further depending upon synthesis temperature. Magnesium ferrite’s porous character makes its suitable for gas sensing. The electrical properties of magnesium ferrite change because of protonic conduction in the absorbed water molecules on their surface, which permit their use as humidity sensors. Grotthuss was the first to explain the protonic conduction in physisorbed water a hydrogen-bonded network [108]. As dry magnesium ferrite is kept in contact with humid air, water molecules chemisorb on the available cations and defective sites of the oxide surface. This process mainly occurs at the neck parts of crystal grains, by a dissociative mechanism to form two hydroxyl ions for each water molecule. The hydroxyl group adsorbs on metal cations present on the surface layer of the grains possess high local charge density and a strong electrostatic field, and the proton reacts with an adjacent surface O2 group to form a second OH group [68]. When the first layer of water molecules is formed, subsequent layers of water molecules are physically adsorbed on the first hydroxyl layer. The physisorbed water easily dissociates to form H3O þ because of the high electrostatic fields in the chemisorbed OH layer, as shown in Fig. 8. The first layer of physisorbed water molecules is characterized by double hydrogen bonding of a single water molecule. The physisorption changes from monolayer to multilayer as the water-molecule concentration increases. Water molecules in the succeeding physisorbed layers are only singly bonded and form a liquid-like network. After water molecule adsorption on the

Magnetic Materials

215

Octahedral site (B site)

Tetrahedral site (A site)

M+2 a

O−2

Fe+3

a

Fig. 7 Spinel lattice of magnesium ferrite with tetrahedral and octahedral sublattices.

H

H2O

Mg −

(A)

H−

Mg+

−Mg

(B)

H

H+

−Mg(OH)

Mg−

(C)

Mg(OH)−

H

H

H

H

−Mg(OH) Mg(OH)−

(D)

Fig. 8 Surface conduction mechanism steps. (A) Magnesium ferrite sample surface. (B) Adsorption of water molecule. (C) Dissociation and chemisorption of water molecule. (D) Physisorption of water molecules, conduction of electrons by tunneling on the sample surface, and protonic transportation through hydrogen bonding in physisorbed water molecules.

ferrite surface H þ ions hop from one water molecule to the next neighboring molecule forming hydronium ion (H3O þ ). Drop in impedance of ferrite pellet in humidity compared to dry atmosphere was performed by taking Nyquist plot in dry conditions and at 93% RH, as shown in Fig. 9. Impedance of ferrite pellet with silver comb electrodes on the opposite face of a 4.8-cm2 pellet exhibited 104 order decrease in impedance at 93% RH compared to dry atmosphere. A low frequency tail appeared at 93% RH demonstrating protonic conduction in the pellet. It confirmed the proton hopping in physisorbed water multilayer on the surface and inside the pores of the ferrite pellet at high humidity. The conductivity of porous material increases when exposed to humidity due to increasing water vapor adsorption on the material surface [109–111]. Chemisorption and physisorption of water vapor are associated with heat exchange during the adsorption process. Conduction mechanism due to physisorption of water vapor on magnesium ferrite has been experimentally confirmed by measuring isosteric heat of adsorption. Predominant physisorption of water molecules on magnesium ferrite surface and protonic conduction mechanism was supported by isosteric heat of adsorption, RH hysteresis plots, and impedance spectroscopy [112]. For sensing water vapor cations, oxygen vacancies on ferrite surface and porosity have a key role for the operation of humidity sensors. In order to improve the surface water molecule sensing of magnesium ferrite, doping/substitution of alkali and rare earth atoms were investigated.

216

Magnetic Materials

Dry 93% RH

5×106

f

4×106

Increasing frequency 1500 at 93% RH

3×106 2×106

Z′/Ohm

-Imaginary impedance Z ″(Ohm)

6×106

1000

500

Tail

1×106 0

0 0.0

0

1000

2000 Z′/(Ohm)

3000

4000

3.0×106 6.0×106 9.0×106 Real impedance Z′(Ohm)

1.2×107

Fig. 9 Nyquist plot of magnesium ferrite one-inch-square pellet with silver comb electrodes on both faces, taken in dry environment and at 93% RH (inset). RH, relative humidity.

The specific surface area of the magnesium ferrite was found (72 m2/g) and porosity (38%) synthesized at 10001C by solid state reaction method. Humidity dependent properties of ferrites are enhanced by controlled doping percentages of alkali ions [113]. Addition of cerium oxide to pure MgFe2O4 increased sample resistance and humidity sensitivity at low RH. The 4 wt% cerium oxide addition showed a good linearity of resistance in a wide RH range [114]. Humidity sensitivity of pure magnesium ferrite increases with lithium substitution due to the large surface area and high surface charge density. Dissolution of Li þ ions in spinel lattice facilitates quicker nucleation leading to smaller grain size distribution. Pore size distribution becomes smaller with lithium substitution than the pure magnesium ferrite improving the water molecule sensing. The highest humidity sensitivity was observed for 20% lithium substitution at the magnesium site in ferrite [115]. On rare earth doping of praseodymium in magnesium ferrite spin density was enhanced. Effective spin density influenced the water vapor adsorption at lower RH [116]. A colossal seven order (B107) decrease in resistance was reported in 1 wt% ceria added magnesium ferrite thin film with 10%–90% RH [117] change. Nano to macro open pore formed on the magnesium ferrite surface, as demonstrated by scanning electron microscopy images. Porous systems usually consist of interconnected networks. The intragrain pores are generally of molecular dimensions, and form regular networks. Such large specific surface area nanopores are highly sensitive toward water molecule adsorption. Nanoporous TiO2 showed four times increased proton conductivity at given RH compared with mesoporous TiO2 [118]. Porous materials exhibited B5–6 times faster rate of change of electrical conductivity than dense material at 0.15 atm H2O partial pressure and 7001C temperature due to proton hopping [119]. Humidity sensing performance of porous In2O3 with larger surface area was better obtained [120]. TiO2 porous gas sensor significantly enhanced sensor sensitivity by a significant twofold increase (B33%) compared to that of a nonporous TiO2 sensor at 5001C for oxygen detection [121].

2.8.4.4

Magnetic Behavior of Magnesium Ferrite

Magnesium ferrite is a soft magnetic material and it is an important member of the spinel family. It is an n-type semiconducting material having number of applications in adsorption, sensors, and in magnetic material technologies. In the spinel ferrite family magnesium–manganese ferrite was introduced in 1952. It is useful in microwave devices operating at low to medium power levels. In nanocrystalline magnesium ferrite, many of the useful properties of its crystalline counterpart, such as magnetization, are enhanced. Further, magnesium ferrite belongs to the class of soft magnetic materials that is easy to magnetize and demagnetize, so is used as an electromagnetic material. Owing to its nanocrystalline nature and useful properties, the material shows a good potential for novel applications in humidity, gas sensing, and drug delivery. Apart from its magnetic and electronic applications, MgFe2O4 finds a number of applications in heterogeneous catalysis. Moreover, magnesium ferrite and its allied compounds have widespread applications in microwave devices, such as circulators, insulators, phase shifters, and multifunctional devices, because of their low magnetic and dielectric losses and high resistivity. Most of the technologically important magnetic materials like iron and metallic alloys exhibit low electrical resistivity, making these materials useless for applications in transformer cores, inductor cores, and other applications operating at higher frequencies. Magnesium ferrite and its metal doped derivatives have advantage over other electromagnetic materials because of their high inherent resistivity, high permeability, and stability over wide range of temperature [122]. With these advantages, substituted magnesium ferrites outweigh all other magnetic materials. The trouble with the other electromagnetic materials is their low electrical resistivity, which causes the induced currents (called eddy currents) to flow within the materials, thus producing heat. This generated heat is often a serious problem and becomes a cause for loss of energy. Thus, the materials become inefficient due to wastage of energy, especially at high frequency. However, the performance of soft magnetic materials like magnesium ferrites is much better at high frequencies due to its high electrical resistivity [123]. In addition, no other magnetic material possesses magnetic and mechanical parameters as flexible as those of soft ferrites. Magnetic

Magnetic Materials

217

30

Moment/mass (emu/g)

20 10 0 −10 −20 −30 −6000

−4000

−2000

0

2000

4000

6000

Applied magnetic filed (gauss)

Fig. 10 Room temperature magnetic field vs. magnetic moment plot of magnesium ferrite.

12

Moment/mass (emu/g)

10 8 6 4 2 0 0

100

200

300

400

500

Temperature (°C)

Fig. 11 Temperature vs. magnetic moment plot of magnesium ferrite showing Curie temperature.

behavior of chemically synthesized ferrite pellet sintered at 11001C has been shown in Fig. 10 by taking a magnetization loop. The shape of the magnetic hysteresis loop implies a soft magnetic nature. The saturation magnetization of magnesium ferrite exhibited saturation magnetization of 20.8 emu/g. Coercivity has been found to be 63 Gauss and retentivity 4 emu/g. In spinel structure Mg2 þ and Fe3 þ ions are almost inversely distributed over its tetrahedral and octahedral sites as [MgxFey]tet[Mg1 xFe1 y]octO4. The net magnetic moment of magnesium ferrite arises entirely due to the uncompensated magnetic moments of the iron ions occupying two different lattice sites with antiferromagnetic interaction since Mg ions are nonmagnetic. The ferromagnetic to paramagnetic transition temperature of magnesium ferrite has shown by the temperature dependent magnetic moment curve in Fig. 11. Curie temperature of magnesium ferrite has been obtained B4001C. At Curie temperature exchange interaction between the spins is destroyed by thermal energy. Nanoparticles of magnesium ferrite show random distribution of cations in tetrahedral and octahedral sites and do not follow magnetic ordering. Due to their nanosize the structural coherence of the domain is B50–60 Å thus nanoparticles exhibit spin canting and superparamagnetic behavior [124].

2.8.4.5

Generation of Electric Current Due to Water Splitting Generates Electrical Power in Lithium Substituted Magnesium Ferrite

Recently the water molecule dissociation property of porous magnesium ferrite has been exploited for the direct generation of electric current by inventing a new device, the HEC [125]. The HEC consists of two electrodes, zinc plate as anode and silver as inert cathode, attached on lithium substituted magnesium ferrite (Li-magnesium ferrite) pellet, immersed in ion free water in a glass container. Spontaneous dissociation of water molecules occurs at defective sites of the porous surface and inside the pore

218

Magnetic Materials

wall of ferrite and conducts ions with an instant formation of zinc hydroxide at the zinc electrode. Oxidation reaction of OH ions at the Zn electrode releases two electrons (the site of oxidation, the anode) and these are collected on another silver electrode, which converts hydronium ions into hydrogen gas (the site of reduction, the cathode) and water. So there is a net flow of current in the cell due to a potential difference developed between its electrodes, which can be used to do work. HEC cells used in series and parallel combinations can drive a small electric motor fan and power an light emitting diode (LED) table lamp. The prototype workable HEC is an economical proposition, besides being eco-friendly, as only water is being used for its operation.

2.8.5

Illustrative Examples or Case Studies

2.8.5.1

Energy Sources Using Water Precursor/Byproduct

There are energy generating sources that directly or indirectly use water or whose byproduct is water. These are clean energy sources for the environment.

2.8.5.1.1

Fuel cell

Water dissociation is one of the clean sources of H2 production. The potential role of hydrogen in the energy economy of the future stems from an optimistic view; however, its usage demands an elaborate setup due to its stringent safety protocols, dangerous explosive possibilities, and high input cost of pure hydrogen. Such conditions impose a big constraint on liberal hydrogen usage as an energy source for domestic applications. The most attractive option for hydrogen energy is through fuel cell use [126,127]. A fuel cell directly converts chemical energy into electrical energy and its maximum theoretical efficiency is not bound by the Carnot cycle or device size. In fuel cells, electrochemical reactions occur at a catalyzed anode and cathode through a PEM but output voltage of the cell depends on operating conditions, such as temperature, applied load, and fuel/oxidizer flow rates. Use of carbon paper and PEM, in addition to high input cost of pure hydrogen gas, makes fuel cell use very expensive for the average person. Another serious associated problem is hydrogen storage; safe delivery is a roadblock to mass commercialization of fuel cells for automotive and stationary applications. Hydrogen storage and delivery on a mass scale is a long-term issue with certain reservations even today, 60 years since its first use. While at present it is a universally accepted fact, the future switching to a “hydrogen economy” will be extremely costly and it will take decades to complete. Even though installing local hydrogen generation plants in select locations the government subsidy would be high. Although several initiatives to utilize compressed hydrogen in vehicles have been a reality at an exorbitant cost. At present, for a 600-km drive, a 450-Kg fuel system is required with conventional hydrides, an impractical dead weight for vehicles.

2.8.5.1.2

Seawater battery

Seawater contains various chemical species and living organisms; the major chemical constituents of seawater are sodium and chlorides [128]. The underwater surveillance low power seawater battery was developed in 1959–61. Batteries use seawater not only as electrolyte, but also as oxidant with steel wool inert cathodes and magnesium anode. The battery consumes magnesium alloy and dissolved oxygen from the seawater. Power density of a seawater cell is 2 W/ft2 for vertically oriented electrodes. The polarized cell electromotive force (EMF) varied from 0.4 to 0.5 V [129]. A novel rechargeable battery system using seawater as an electrode material has been demonstrated [130]. The seawater provides both anode (Na) and cathode (O2) materials for the battery. During charging, the Na ions in the seawater were extracted and transported into the negative electrode compartment, followed by reduction to metallic Na at the negative electrode. On the positive electrode side, evolution of gaseous O2 and Cl2 phases occurs at the positive electrode. Based on the discharge voltage (B2.9 V) with participation of O2 and the charge voltage (B4.1 V) with Cl2 evolution during the first cycle, a voltage efficiency of about 73% is obtained.

2.8.5.1.3

Cement/soil battery

Cement is a solution of different salts and clays that acts as an ionic conductor. This facilitates its potential as a good electrolyte for novel cement battery designs. Cement batteries utilize the pore solution in cement-based electrolyte in presence of water [131,132]. Cement battery consists of zinc as the anode, manganese dioxide as the cathode, and water as the electrolyte of a battery that provides power during concrete construction. Despite a decrease in mobility of water during setting, the use of the pore solution in set cement as the electrolyte greatly widens the applicability of the cement-battery technology. Cement sealed in 70 mm  70 mm  40 mm plastic molds with electrode plates pairs Cu/Al of dimensions 60 mm  30 mm  0.5 mm provide current 0.101 mA. For every 1% increase in RH, the current was found to increase the current of sealed cells by 0.02 mA compared with 0.01 mA for unsealed cells [133]. The cement-based battery provides open-circuit voltage up to 0.72 V, current up to 120 mA (current density up to 3.8 mA/cm2), power output up to 1.4 mW/cm2, capacity up to 0.2 mAh, and fraction of zinc consumed up to 5  10 5. The performance is very low compared to commercial alkaline batteries [134]. The soil battery consists of a zinc anode, copper cathode, and electrolytes in the form of soil [135,136]. The soil-based battery of dimensions 100 mm  40 mm  20 mm had current density 0.25 mA/cm2 and open-circuit voltage 0.24 V. The power output was 43 mW/cm2, with a capacity up to 179 mAh [137].

Magnetic Materials 2.8.5.1.4

219

Energy harvesting by surface energy of condensed water droplets

The process of condensation of water droplets on a cooled surface swept away by gravity or by vapor shear transfers heat. Since then there have been various experiments performed to improve and implement the energy evolved for practical application. Autonomous removal of condensed water droplet demonstrated on a super hydrophobic surface without applying any external forces [138]. On a hydrophobic surface the surface energy released upon water drop coalescence thus out-of-plane jumping results from in-plane coalescence naturally [139,140]. Water on a hydrophobic surface gains charge when cooled and jumps with surface energy, when coalescence occurs. Utilizing the surface energy gained by coalescence water droplets, jumping water droplet-based energy harvesting demonstrated with nanoengineered superhydrophobic surfaces [141]. Recent studies have shown that when small water droplets (10–100 mm) merge on super hydrophobic nanostructured surfaces, droplets can spontaneously eject via the release of excess surface energy irrespective of gravity. Power densities of 15 pW/cm2 were generated using jumping water droplet from superhydrophobic surface to hydrophilic surface.

2.8.6

Results and Discussions

The capability of Li-magnesium ferrite to dissociate water molecule into ions at room temperature without any external energy has been characterized and analyzed by various experimental techniques. Li-magnesium ferrite was synthesized by solid state reaction method in such a way to make it porous and defective structure. The role of nanopores and surface lattice defects in spontaneous dissociation of water molecules has been analyzed by microscopic studies.

2.8.6.1

Porous Lithium Substituted Magnesium Ferrite

Porosity in Li-magnesium ferrite was created by optimizing sintering temperature and time for sintering. Direct image of porous microstructure of Li-magnesium ferrite pellet captured by scanning electron microscopic (SEM) is shown in Fig. 11. Small distinctly nucleated grains are distributed over the entire surface. Lesser sintering time impedes the grain growth. Aggregates of small grains and distribution of nanopores are distinctly visible in a SEM image Fig. 12(A). Grain contrast is alike embedded with dark spots as pores. Pores are irregularly distributed among the grain boundaries. A combination of macropores (450 nm) and mesopores (2–50 nm) are visibly connected through the grain neck over the entire surface. Presence of micropores (o2 nm) within the grain is not visible in SEM resolution range. Fractured image of ferrite pellet in Fig. 12(B) shows the highly porous three dimensional interconnected pore meshes in ferrite pellet. Pores are distributed throughout the thickness of the pellet. Closed and open pores are widely distributed through an interconnected pore network. Quantitative analysis of pore size distribution, specific surface area, and total porosity of the ferrite pellet is determined by Brunauer–Emmett–Teller (BET) and Barrett–Joyner–Halenda (BJH) technique. Adsorption/desorption (BJH) isotherms for surface area determination of Li-magnesium ferrite particles was carried out in nitrogen atmosphere at 77K. The N2 desorption branch does not follow the adsorption branch exhibiting a hysteresis plot supporting multilayer N2 adsorption. Ferrite pellet follows type III isotherm according to International Union of Pure and Applied Chemistry (IUPAC) convention shown in Fig. 13 (inset). The typical isotherm possessed a wide range of pore size distribution. The average pore diameter distribution in the pellet was 4.2 nm as shown in Fig. 13. Specific surface area of ferrite particles was 165 m2/g determined by BET. The total pore volume of ferrite pellet was obtained 0.74 cc/g for pores smaller than 455-nm diameter depicted in Table 2. Total porosity of the pellet was about 30%.

2 μm

(A)

10 μm

(B)

Fig. 12 (A) Scanning electron micrograph (SEM) of synthesized cell pellet showing porous microstructure, grain and pore size distribution at micrometer scale. (B) Fractured pellet image showing 3-dimentional porous network of nano- to micron-size pore distribution.

Magnetic Materials

220

Volume adsorbed (cc/g)

0.008

dV/dD (cc/A/g)

0.006

0.004

120

N2 adsorption

100

N2 desorption

80 60 40 20 0

0.002

0.0

0.2 0.4 0.6 0.8 Relative pressure (P/Po)

1.0

0.000 0

100 200 Pore diameter (Angstrom)

300

Fig. 13 BJH average pore size distribution plot of nanoporous lithium substituted magnesium ferrite (Li-magnesium ferrite) N2 adsorption–desorption isotherm (inset). Reproduced from Kotnala RK, Shah J. Green hydroelectrical energy source based on water dissociation by nanoporous ferrite. Int J Energy Res 2016;40(11):1652–1661.

Table 2

Brunauer–Emmett–Teller (BET), Barrett–Joyner–Halenda (BJH), and Dollimore-Heal (DH) measurement data of the pellet

Surface area data Multipoint BET BJH method cumulative desorption surface area DH method cumulative desorption surface area

1.653E þ 02 m²/g 2.834E þ 02 m²/g 2.878E þ 02 m²/g

Pore volume data Total pore volume for pores with diameter less than 4554.8 Å at P/Po¼ 0.99577 BJH method cumulative desorption pore volume BJH interpolated cumulative desorption pore volume for pores in the range of 5000.0–0.0 Å diameter DH method cumulative desorption pore volume

1.748E-01 2.542E-01 2.542E-01 2.467E-01

Pore size data Average pore diameter BJH method desorption pore diameter (mode) DH method desorption pore diameter (mode)

4.230E þ 01 2.517E þ 01 2.517E þ 01

cc/g cc/g cc/g cc/g Å Å Å

Source: Reproduced from Kotnala RK, Shah J. Green hydroelectrical energy source based on water dissociation by nanoporous ferrite. Int J Energy Res 2016;40 (11):1652–1661.

2.8.6.2

Water Splitting by Lithium Substituted Magnesium Ferrite at Room Temperature

High sensitivity of magnesium ferrite for water vapor sensing has been explored for the dissociation of water molecules at room temperature. Water molecule sensitivity of ferrite has been further improved by lithium substitution and controlling the processing temperature conditions. Synthesis of Li-magnesium ferrite at low temperature and time is mainly responsible to create defects and porosity in it. Lithium substitution created high nucleation rate in spinel structure formation thus promoting small particle size distribution ranging from 50 to 900 nm, as displayed in Fig. 14(A). Smaller alkali 20% lithium ion substitution at the magnesium site enhanced the surface reactivity of magnesium ferrite with water molecules. Monovalent lithium substitution at divalent magnesium site creates oxygen vacancies that acted as dangling/unsaturated bonds due to trapped electrons. Lithium substitution also developed strain in the ferrite lattice demonstrated by the Moire fringe pattern in high resolution transmission (HRTEM) images, circled in Fig. 14(B). Moire fringes arise due to the interference of overlapped strained fringes of other crystallite [142,143]. It may be also due to the fact that the electron beam diffracted locally by defects produced by smaller lithium ions than magnesium ions resulted in bright and dark field images. In a spinel lattice the tetrahedral cation is four coordinated, and thus forms a highly stable bond. In an octahedral lattice the cation is six coordinated by oxygen atoms hence the surface atom is highly unstable. Lithium ion has preferably higher enthalpy ELi þ ¼ 41 kJ/mole, compared with Fe3 þ ¼ 22.4 kJ/mole and Mg2 þ ¼ 25.1 kJ/mole to occupy the octahedral site [144]. Thus the octahedral site would be most oxygen deficient. Defects were

Magnetic Materials

800 nm

221

1 nm

0.25 nm

(A)

(C)

5 nm

0.2 (1/A)

(B)

(D)

Fig. 14 High resolution transmission (HRTEM) image of ferrite particles (A) showing ferrite particle size and shape distribution. (B) Particle lattice fringes depicted Moire fringes. (C) High magnification image of lattice fringes encircled area shows the grain boundary of crystal lattice having more defects and irregular lattice pattern. (D) Selected area electron diffraction pattern of particle exhibiting (311) plane. Reproduced from Kotnala RK, Shah J. Green hydroelectrical energy source based on water dissociation by nanoporous ferrite. Int J Energy Res 2016;40(11):1652–1661.

observed in high magnified HRTEM image. Fig. 14(C) depicts the irregular fringe patterns in the lattice. Grain boundaries often exhibit irregular atomic arrangement and more defective sites. The lattice fringe width measured 0.25 nm, corresponding to (311) lattice planes of spinel ferrite. It has been reported by low energy ion scattering (LEIS) that octahedral sites (110) and (111) are preferentially exposed on the surface layer of magnesium ferrite crystallite [145]. Thus the most defective octahedral lattice (311) in Li-MgF is exposed significantly on the surface layer. Selected angle electron diffraction pattern (SAED) of ring diameter 0.25 nm also corresponds to (311) spacing. Oxygen vacancies with trapped electrons produce high electric field capture hydroxyl ions of water molecules to compensate its vacancy [146]. The strong Coulomb attraction between surface metal cations Mg2 þ , Li þ , Fe3/2 þ ions and oxygen of polar water molecule forms hydrogen bonding that weakens the O–H covalent bond in a water molecule exhibiting a dynamical process of water dissociation over surface cations. Thus the first layer of the water molecule chemidissociated by surface cations and oxygen vacancies form an OH layer. The surface OH produces high local surface charge density to further physisorb water molecules via hydrogen bonding. Water molecule dissociation mechanism on (311) plane of Li-MgF is schematically shown in Fig. 15. Initially the chemidissociated and physisorbed water molecule forms a multilayer by hydrogen bonding. Physisorbed multilayer formation inside nanopores traps H3O þ ions inside a pore of, for example, 100-nm length and 50-nm diameter, developing a high electrostatic field. The electrostatic field developed inside the nanopore was calculated by applying the Gauss theorem. The electric field developed inside the nanopore was 2.16  104 V/cm, which is high enough to dissociate the physisorbed water molecule spontaneously. It needs few eV to dissociate physically adsorbed (Eadr0.1 eV) water molecules.

2.8.6.3

Trapped Hydronium Ion Imaging by Electrostatic Force Microscopy

Electrostatic force microscopy (EFM) measures local DC voltage developed on the surface charge. Conducting tip is driven into oscillation at the resonance frequency. The cantilever oscillates freely with a reasonable quality coefficient of resonance, and the phase shifts induced by the force gradients are detected due to surface charge presence. The change in frequency of cantilever due to influence of electric field gradient generated by electrostatic surface charge has been confirmed by EFM measurement. Physisorbed multilayer water molecule traps proton in the H3O þ state on the ferrite surface. An EFM image of wet ferrite pellet for 1-mm2 scanning area captured is shown in Fig. 16(A). On applying 0.6 V bias voltage lower than the water dissociation voltage 1.23 V on the tip, the bright contrast of image demonstrated attractive force due to hydronium ions’ presence in physisorbed water molecules. The similar scanning area exhibited a darker image on applying þ 0.6 V bias, displaying the repulsive force shown in

222

Magnetic Materials

H2O molecule

Magnesium ion

Oxygen atom

Iron ion

Oxygen vacancy

Lithium ion

Mg0.8Li0.2Fe2O4(111) plane Fig. 15 Water molecule adsorption and dissociation on surface unsaturated octahedral Mg, Li ions, and oxygen vacancies in (111) plane of Mg0.8Li0.2Fe2O4 spinel lattice.

17.6

16.2

10.0

10.0

5.0 500 nm

500 nm

5.0

0.0 (A)

(B)

Fig. 16 Electrostatic force microscopic images of wet ferrite pellet on applying 0.6 V DC bias at conducting tip (A) and þ 0.6 V at tip (B). Color bar shows developed voltage on wet surface in volts. Reproduced from Kotnala RK, Shah J. Green hydroelectrical energy source based on water dissociation by nanoporous ferrite. Int J Energy Res 2016;40(11):1652–1661.

Fig. 16(B). The EFM image contrast changes with the polarity of dc bias, demonstrating attractive and repulsive forces between the tip and surface charge [147,148]. Thus EFM results confirm the presence of H3O þ ions on the wet Li-magnesium ferrite surface.

2.8.6.4

Chemidissociated Water on Li-Magnesium Ferrite by Fourier Transform Infrared Spectroscopy

Fourier transformation infrared spectroscopy (FTIR) of Li-magnesium ferrite powder was taken to identify chemical bonding of water molecules. The interaction of the hydroxyl group with the oxide surface has been explicitly observed by IR spectroscopic data [149,150]. Two peaks appearing at 433 and 596 cm 1 correspond to tetrahedral and octahedral bonds in the spinel lattice, as shown in Fig. 17. Substituted lithium in magnesium ferrite occupying the octahedral site in the spinel lattice is observed by a kink at 358 cm 1. The peak at 1632 cm 1 signifies bending frequency of the O–H molecule. Peak presence at 3486 cm 1 is attributed to a three-coordinated hydrogen bonded water molecule by physisorption. Lower valance of lithium ions occupied at the octahedral site, as observed by FTIR, enhances the charge imbalance. The octahedral sites of spinel oxide are mainly exposed on the surface [151], which increased the surface reactivity of the ferrite with that water molecule.

2.8.6.5

Collection of Hydroxide and Hydronium Ions Produced by Lithium Substituted Magnesium Ferrite

Two electrodes zinc and silver have been used onto Li-magnesium ferrite surfaces for collection of dissociated ions of water molecule. Water molecule dissociation occurred spontaneously on ferrite surface of the pellet. The water molecule dissociating at the interface of ferrite pellet and zinc foil immediately oxidizes the zinc into zinc hydroxide by releasing two electrons. Oxidation reaction occurs at Zn electrode as: Zn þ 2OH -ZnðOHÞ2 þ 2e ; Eoxd ¼

0:76 V

ð7Þ

Magnetic Materials

120

0

3486 cm−1 O−H bond

Tetrahedral

Octahedral 400

596 cm−1

20

433 cm−1

60 358 cm−1

% Transmittance

80

1632 cm−1

1402 cm−1

O−H bending

100

40

223

600

800 1000

2000

4000

Log10 (cm−1)

Fig. 17 Chemically adsorbed water molecules’ presence in Li-magnesium ferrite by Fourier transformation infrared spectroscopy (FTIR) plot. Reproduced from Kotnala RK, Jyoti Shah. Green hydroelectrical energy source based on water dissociation by nanoporous ferrite. Int J Energy Res 2016;40(11):1652–1661.

Proton as H3O+ conducts through physisorbed multilayers of water molecules. These protons migrate toward Ag comb electrode on other face of pellet. On Ag electrode proton get reduced to mono atomic hydrogen by accepting electron migrated from Zn electrode via external circuit. Atomic hydrogen combines with other hydrogen atom resulting into H2 gas with a potential as: 2H3 Oþ þ 2e -H2 ↑ þ 2H2 O; Ered ¼ þ 0:22 V

ð8Þ

The combined effect of water dissociation by ferrite and collection of ions by electrodes is termed as a complete cell or HEC. The overall HEC voltage is Ecell ¼0.22 þ 0.76¼0.98 V. The generated voltage additionally helps to transport H3O þ and OH ions toward the respective electrodes. Nanopores in ferrite pellet trapped protons in the form of hydronium ions create a high-enough electric field to further dissociate the physisorbed water molecule bonding spontaneously. The ionic current transported through the cell transformed to electric current by oxidation and reduction reaction at electrode pairs. The working mechanism of a HEC is shown schematically in Fig. 18. In the schematic diagram, the brown disk shows Li-magnesium ferrite attached back to back with single zinc foil electrode sheet as anode. Comb pattern on the opposite phase of the disk is a silver electrode as inert cathode. Both back to back disks are connected with wires at silver electrodes. When water is sprayed by a sprayer, water molecules are adsorbed by ferrite disks. Water molecule chemisorption at surface ions followed by physisorption has been shown on the disks. Protonic conduction via hydrogen bonded physisorbed water molecules as hydronium ions toward silver electrodes and movement of hydroxide ions toward zinc electrode foil have been represented. White dots on zinc foil depict the zinc hydroxide formation with hydroxide ions. As the zinc hydroxide is formed, electrons are released through connecting wires. The proton reaching the silver electrodes forms H2 gas by receiving electrons from zinc electrodes. The zoomed image demonstrates the phenomena of chemisorption and physisorption inside the nanopore. The trapped hydronium ions generated an electric field inside the nanopores that is enough for continuous dissociation of further water molecules. Fabricated cell picture of RKJ92 (2  17 cm2) is shown in Fig. 19. The brown disk is Li-magnesium ferrite of 2-in. diameter. Silver paste on one face has been printed by screen printing. The opposite face of the disk has been attached to zinc foil as anode electrode. Another disk has been attached back to back with zinc foil. Both disks were soldered with wire and two connections have been taken from zinc and silver of one disk. That is the complete HEC.

2.8.6.6

V–I Performance of Hydroelectric Cell

Electric current and voltage response of 34 cm2 area cell RKJ92 was demonstrated by standard performance V–I curve representing losses in the cell voltage with operating current [152]. Due to losses resulting from undesired species crossover from one electrode through the pellet the actual open circuit voltage is below the theoretical value. The drop from the open circuit voltage arises due to activation polarization, Ohmic polarization, and concentration polarization. Activation and concentration polarization occurs at the electrode regions, while the resistive polarization is due to Ohmic losses throughout the cell. Polarization plot of overvoltage versus cell current for RKJ92 cell is shown in Fig. 20. Polarization curve illustrates different types of control regions in overvoltage curves. Overvoltage of RKJ92 cell exhibited 0.9 V. Polarization curve is subdivided into three regions, AB, BC, and CD, according to the losses in the cell. Actual open circuit voltage of the cell was represented by point A. At low current density, activation polarization region AB is the voltage drop to overcome activation energy of the electrochemical reaction on the surface and nanopores of the ferrite pellet. A linear region BC on the I–V curve is the voltage loss by internal Ohmic resistance through the cell, primarily due to high resistivity of ferrite, DI water, and electrode contact resistance. High current density region CD is the voltage

224

Magnetic Materials

2H2O↔H3O++OH−

100 200 mW

0

Ecell = 0.22+0.76 = 0.98 V

Electrostatic field

Power meter 2e−

e−

e− H3O+

H3O+

Water spray

OH−

OH− H+

H H+ H H+

H H+ O

O

H H+ OH OH Mg V− O

O

O

H2 (gas)

H+ H OH OH V−O Mg

Zn (anode)

MgLiF

Ag (cathode)

Nanopores

Zn(OH)2

Fig. 18 Schematic diagram of a hydroelectric cell (HEC) showing water molecule dissociation and conduction of hydronium and hydroxide ions. Hydronium ions hop toward the Ag electrode and hydroxide ions toward the Zn electrode, releasing two electrons through external circuit. Zoomed image shows nanopores on the pellet surface, and further proton hopping inside the nanopore generates enough electric field to dissociate water molecules instantly.

Ag electrode

Zn foil

Mg0.8Li0.2Fe2O4 Fig. 19 Lithium substituted magnesium ferrite (Li-magnesium ferrite), RKJ92 cell pellet with silver comb inert electrode and zinc sheet as anode pasted on the other face (back) of the ferrite pellet.

drop due to high concentration of ions on each electrode surface limiting the mass transport [153,154]. HEC RKJ92 exhibited peak power 28 mW, and maximum cell power generated for the cell is quite high at 145 mW. Maximum 150 mA current was generated by the RKJ92 cell. The capacity of the RKJ92 cell is 235 mAh.

2.8.6.7

Ionic Current Flow on Water Splitting in Hydroelectric Cell by Nyquist Plot

Electrochemical impedance spectroscopy (EIS) is a powerful measurement technique used mainly to characterize and model the dynamic behavior of Li-ion batteries [155,156]. The impedance spectra, Nyquist plot displays different time constants corresponding to conduction of charges. The EIS Nyquist plot is divided into three different frequency bands (high, middle, and low frequency band semicircle). Each semicircle is represented by parallel RC time constant networks connected in series. The

Magnetic Materials

1.0 A

30 25 20

0.6 Ohmic loss

15

0.4

Power (mW)

Voltage (volts)

Voltage Power

Activation loss

B 0.8

225

10 0.2

C 5

Concentration loss 0.0 0

20

40

D

60 80 100 120 140 160 Current (mA)

0

- Imaginary impedance Z″ (Ohm)

1.5×107

- Imaginary impedance Z″ (Ohm)

Fig. 20 Polarization curve and energy generated by RKJ92 cell partially dipped in deionized (DI) water.

HEC in dry HEC in water

1.2×107

9.0×106

Increasing frequency

6.0×106

35

HEC in water

30

High frequency tail

25 20 15 Low frequency tail

10 5 0 0

3.5×106

(B)

20 40 60 Real impedance Z′ (Ohm)

80

0.0 0.0

5.0×106

1.0×107

1.5×107

2.0×107

2.5×107

Real impedance Z′ (Ohm)

(A)

Fig. 21 Nyquist plot of RKJ29 cell in dry (A) and in water (B) showing low frequency tail due to H3O þ ion diffusion and high frequency capacitive tail due to formation of passive film Zn(OH)2 at Zn anode.

impedance of the HEC is measured over a frequency range of 100 MHz–20 Hz. A Nyquist plot of RKJ92 cell taken in dry conditions and with water is shown in Fig. 21(A) and (B). In the dry environment, cell pellets exhibited high reactance of the order 108 Ω. When deionized (DI) water is sprayed on a cell, the semicircle disappeared and reactance of cell pellet decreased to approximately 70 Ω due to high protonic conduction. Moving from left to right in the Nyquist plot, the first curve represents the high frequency capacitive tail, middle frequency semicircle, and low frequency tail. The Nyquist plot shows a high-frequency tail due to electrode polarization effect owing to the formation of passive Zn(OH)2 film formation at the Zn electrode [157]. The lowfrequency tail of the relaxation articulates diffusion of H3O þ due to orientational polarization in physisorbed water multilayer observed in Fig. 21(B) [158,159]. Working mechanism and conditions of the recently invented HEC with well established solar and fuel cells have been presented in Table 3.

2.8.6.8

Hydroelectric Cell Byproducts

Although energy is a basic need of human beings, harmful byproducts of energy sources pollute the environment and have an adverse effect on human health. HECs, as an energy source, do not produce any byproducts harmful to the environment or health.

2.8.6.8.1

Zinc hydroxide

HECs consume the zinc electrode during energy generation and forms byproducts zinc hydroxide and hydrogen gas at silver electrode. Zinc hydroxide is converted into very useful nanoparticles of zinc oxide, which possesses a multifunctional character. It

226

Table 3

Magnetic Materials

Comparisons of fuel cell and solar cell with hydroelectric Cell (HEC)

Fuel cell

Solar cell

HEC

A unique primary electrical source operates at 901C or higher temperature

A solid state device to produce electricity in the presence of light

A future of power generation for specific requirements only, due to elaborate system of hydrogen and oxygen storage, etc Expensive precursors H2, O2, proton exchange membrane (PEM), Pt, and carbon paper are used Potential for a high operating efficiency (up to 50%–70%), about 35% presently Excessive heat is produced during its operation therefore cooling is must Zero or near zero greenhouse emissions, with level of pollution reduction but not pollution free No moving parts, vibration free, and highly reliable operation A highly scalable design, output voltage 0.70 V and 20 mA/cm2 current Byproduct is water

Electrical power output at room temperature for general purposes

A fascinating electrochemical cell operates on new principle based device at room temperature Parallel electrical power generation for all requirements with no elaborative system

Only sunlight is required (freely available)

Precursors used are inexpensive, such as water, zinc, magnesium ferrite, and Ag contacts

Efficiency is limited.

No limitation for its efficiency.

Infrared radiation from sunlight heat solar cell to 801C Zero greenhouse emissions

No excessive heat is produced during its operation No pollution, no greenhouse emissions, no chemicals are used for its operation

Same as in fuel cell

Same as in fuel cell

Same as in fuel cell

Costly and risky system Fuel cell lifetime is large, about 200 days tested (development history of 60 years)

Costly system but risk free Solar cell life time is more than 20 years

Highly scalable simple design output voltage 0.9 V and 4.8 mA/cm2 current Byproducts hydrogen gas and zinc hydroxide are very useful Risk free highly economical system At the moment its lifetime tested for 240 days. (development history of 2 years only)

No byproducts

has diverse applications, such as semiconductor, optoelectronic, gas sensor, diluted magnetic semiconductor, and dielectric material [160–166]. Zinc oxide nanoparticles have remarkable potential in the field of spintronics and photonics [167,168]. The optical and semiconducting band gap of ZnO nanoparticles increases and is highly dependent on size and shape of nanoparticles. Visible exciton emission in ZnO nanoparticles is observed due to defects and vacancies created on the surface [169]. These sizes also play a role in its dielectric constant. Dielectric constant of ZnO is tuned with nanoparticle size [170]. There are various methods reported for the synthesis of ZnO nanoparticles. The electrochemical method is one of the simplest, lowest cost, and ambient temperature reaction processes to synthesize ZnO nanoparticles [171,172]. Yet electrolyte is a must in all electrochemical methods for conduction and reaction on applying external current source. HECs generate their own current by protonic/hydroxyl ion conduction due to chemidissociation and physisorption of water molecules. Hydroxyl ions migrate toward zinc electrode and form Zn(OH)2 by oxidation reaction without adding any electrolyte/external energy source in water. Zinc hydroxide nanoparticles are obtained as a byproduct (anode mud) during electrical power generation by HEC. Zinc hydroxide byproduct during electrical power generation by HEC was thermally decomposed to ZnO nanoparticles. The yield of ZnO nanoparticles by HEC is reported 0.017 g with self generated 60 µA/cm2 current density. without using any electrolyte [173]. Polycrystalline orthorhombic Zn(OH)2 and ZnO phase has been identified by X-ray diffraction peaks as shown in Fig. 22. Morphology of ZnO nanoparticles with very fine agglomerated particles is depicted by scanning electron microscopy image Fig. 23. Yield of ZnO nanoparticles obtained by self generated energy source HEC compared with other reported electrochemical process is given in Table 4. High resolution transverse electron microscopic image of fine nanocrystals with narrow size distribution from 5 to 20 nm have been formed by this process as shown in Fig. 24. Finely distributed nanocrystals are grown under near-equilibrium conditions at room temperature without any external energy source that maintained the reaction rate, thus the crystal growth happens. Optical band gap of ZnO nanoparticles measured by photoluminescence (PL) emission spectra is shown in Fig. 25. It exhibits strong PL emission at 399 nm and satellite peak at 412 nm near UV band edge emission peak. Further it suggested energy transition from zinc vacancies (Vzn) to valence band [175]. Weak violet-blue 436 nm and blue emission peak at 439 nm can be ascribed to transition from zinc interstitial (Zni) to zinc vacancies (Vzn) defect level in ZnO [176,177]. Vacancy defects of Zn produced due to H2 gas evolution at silver electrode provides reducing environment in DI water. ZnO nanoparticles obtained by HEC process also exhibited remarkable dielectric constant values as observed in Fig. 26. Dielectric constant er linearly decreased with increasing frequency. The loss decreased exponentially at lower frequency and slowly at higher frequency. The dielectric constant value obtained 6.5 at 1 KHz and 5 at 1 MHz with a loss 0.2 at 1 KHz and 0.01 at 1 MHz. Generally nanoparticles of ZnO exhibit lower dielectric constant than micron size bulk ZnO [178,179]. However, nanoparticles synthesized by HEC method exhibited higher value of dielectric constant than the reported ZnO nanoparticles.

Magnetic Materials

#

227

# ZnO

Intensity (arb. unit)

∗ Zn(OH)2 # #

∗

#

∗

#

#

# ∗∗ 10

20

30

40 50 Angle (2)

∗

60

# #

# 70

80

Fig. 22 X-ray diffraction pattern of white powder deposited during cell reaction at zinc anode. Reproduced from Shah J, Kotnala RK. Rapid green synthesis of ZnO nanoparticles using a hydroelectric cell without an electrolyte. J Phys Chem Sol 2017;108:15–20.

2 μm Fig. 23 Scanning electron micrograph of ZnO nanoparticles.

2.8.6.8.2

Hydrogen gas

Hydrogen gas is a direct source of clean energy and has applications in food industries, petrochemical refinery, cutting/welding, metallurgy, etc. [180–183]. Hydrogen gas is commercially produced by electrolysis of water, reforming of natural gas, and gasification of coal. Each hydrogen gas production process is encompassed with some complexities, such as high current density input, high temperature, and emission of CO2 gas. Nonpolluting H2 gas generation methods are underway, such as photobiosynthesis, photoelectrocatalysis, and thermochemical, etc. [73,184,185]. Hydrogen gas generation by HEC at room temperature is a novel, challenging process. Li-magnesium ferrite dissociates water molecules in hydronium and hydroxide ions. Two electrodes, silver and zinc, are attached onto the ferrite pellet for collecting ions. At the inert cathode silver electrode H2 gas is evolved as a result of hydronium ion neutralization. Hydrogen gas evolved at the silver electrode was confirmed by hydrogen gas sensing detector MQ-8 with. HECs in continuous operation. H2 gas evolved was also collected by water displacement method during cell operation. For rapid H2 gas collection 2 V external voltage was applied on one-inch-square Li-magnesium ferrite pellet kept for 1 h. Hydrogen gas by volume 10 mL was collected in this process. Collected gas injected into a gas chromatograph showed two major peaks. The first peak eluted at retention time 0.94 min, which corresponds to the hydrogen reference peak at 0.92 min shown in Fig. 27. The purity of hydrogen gas is close to 99.99% generated by HEC. Such amount and purity of hydrogen gas collected is a significant value compared to the existing photocatalytic, thermochemical, and photosynthesis methods [186].

2.8.7

Future Directions

The advent of electrical power in the hands of human beings brought a big comfort in day to day life, initially for the purpose of lighting. Later, this was followed by running motors and machinery using electricity, which accelerated the industrial revolution in

228

Magnetic Materials

Table 4

Comparison of ZnO nanoparticle yield with reported electrochemical methods

Zn metal dimension (cm)

ZnO yield

Electrolyte

Current density by external current source

Temperature/time

References

5  6  0.8 75 33

0.17 g – 0.017 g

30 mmol (NaHCO3) 0.5 mol (NaCl) No electrolyte

5 mA/cm2 – –

Room/1 h 981C/4 h Room/3 h

[174] [129] Hydroelectric cell (HEC)

Source: Reproduced from Shah J, Kotnala RK. Rapid green synthesis of ZnO nanoparticles using a hydroelectric cell without an electrolyte. J Phys Chem Sol 2017;108:15–20.

(103) (110) (102)

(101) (002) (100)

0.29 nm (100)

0.2 (1/Å) 0.29 nm

200 nm

5 nm

Fig. 24 ZnO nanoparticle size and shape distribution images with crystallite lattice spacing. Reproduced from Shah J, Kotnala RK. Rapid green synthesis of ZnO nanoparticles using a hydroelectric cell without an electrolyte. J Phys Chem Sol 2017;108:15–20.

PL spectra Fitted curves

Intensity (arb. unit)

3.01

2.70

3.18 2.84

2.45

3.41

350

400

450 500 Wavelength (nm)

550

600

Fig. 25 Photoluminescence spectra of ZnO nanoparticles with multiple peak fitting. Reproduced from Shah J, Kotnala RK. Rapid green synthesis of ZnO nanoparticles using a hydroelectric cell without an electrolyte. J Phys Chem Sol 2017;108:15–20.

Europe. But use of a motor mass transport system was not possible without use of magnets along with electrical current. Moreover transmission of electricity over long distances became inevitable, which increased dependence on transformers. A transformer core is made up of electrical steel strips, which are a special magnetic material.

Magnetic Materials

8

229

1.6

D

1.2

7 0.8 6

tan  (loss)

Dielectric constant (r)

r

0.4

5 102

0.0 106

104 105 103 Frequency, Hz (log scale)

Fig. 26 Dielectric constant vs. frequency plot of nanoparticles of ZnO pellet. Reproduced from Shah J, Kotnala RK. Rapid green synthesis of ZnO nanoparticles using a hydroelectric cell without an electrolyte. J Phys Chem Sol 2017;108:15–20.

Intensity (arb. unit)

H2

Air

0

2

4 6 Retention time (min)

8

10

Fig. 27 Retention time (minutes) vs. intensity of gas peak by gas chromatography.

2.8.7.1

Permanent Magnet Energy

To manufacture more efficient motors and transformers, magnetic hysteresis losses in their core require special attention. In this direction many efforts have been initiated to develop different magnetic materials with high permeability, low coercivity, and high magnetic energy product (BH) parameters. At the initial stage use of carbon free and iron alloys dominated as core materials and in permanent magnets but energy saving efforts replaced them by electrical steel, amorphous thin foils of Fe-Si composition, rare earth NdFeB materials. At present demand for magnetic materials is very high, and must be low cost with high magnetic energy density. Rare earth magnet costs being high, alternative rare earth free materials are being explored. In this direction efforts are on to develop nanocomposite magnetic materials. Magnetic field based energy storage/conversion is being tried out to conserve surplus electrical energy generated. Superconducting magnet system has got potential to store electrical energy, in present scenario it seems to be more viable way to conserve surplus electrical energy.

2.8.7.2

Clean Energy and Sustainable Environment

Intensive research efforts have been considered to create new magnetic materials that can improve efficiency of special induction motors designed for running a car or other vehicles. It mainly depends on magnetic material used in induction motors. Use of electric vehicles has already made inroads to replace gasoline based vehicles worldwide. For the last three decades the steep rise in crude oil price has forced the energy sector to adopt alternative/renewable energy sources. In light of this situation demand for magnetic materials is continuously increasing.

230

Magnetic Materials

Renewable energy sources are emerging as the future paradigm for sustainable development of human society. Energy and environmental protection have become integral parts of societies’ prosperity worldwide. It is well known that energy generation very negatively affects a clean environment on earth. Energy generation processes are shifting natural physical and chemical equilibriums of environment besides emission of greenhouse gases, toxic gases, and harmful particulate matter PM2.5. This is creating environmental pollution, which has very harmful effects on all living beings. Therefore a quest for clean energy sources has sped up, while magnetic energy utilization is one of the best options.

2.8.7.3

Environmental Magnetism

Worldwide, scientists are in pursuit of greener and cleaner energy generation technologies from the available limited set of sources in nature. However, exploiting energy from cleaner sources is a difficult challenge. At present tariffs from energy generation from renewable energy sources has reached parity with conventional modes of power generation. Recently active innovations have been initiated in environmental magnetism that involve interdisciplinary science taken from lakes, marine sediments, biomagnetism, and pollution. Environmental magnetism will help to curb pollution and to understand climate change in a better way.

2.8.7.4

New Magnetic Material as Green Energy Device – Hydroelectric Cell Invention

Still there is a big scope to synthesize new materials for green energy generation without disturbing the ecological balance of nature. A serious concern is prevailing over the use of primary and secondary sources of electrical power being used in trillion numbers. Commonly used batteries/cells are button cells, dry cells, lithium ion and lead acid batteries, which are toxic and not eco-friendly. Their disposal is harmful to humans and plants and to the environment. From all of the above future directions/ requirements there is a need for a new material to generate electrical power without disturbing the ecological and environmental balance of nature. For last 70 years intensive efforts have been put on different metal oxides to generate hydrogen and oxygen gases in the presence of ultraviolet (UV) light and catalyst, and some serious attempts have been made to produce electricity by expensive and elaborative means. Finally, a big success has been achieved by utilizing a magnetic material for the generation of green electrical energy using simple technology at low cost. Magnesium ferrite partially substituted by lithium has been innovated to dissociate water molecules into hydronium and hydroxide ions at room temperature. The dissociation of water molecules has happened by exploiting the oxygen deficient nature of magnesium ferrite and nanopores created in the ferrite. To separate and collect the two ions two electrodes attached to Li-magnesium ferrite pellet gives birth to the HEC. Invention of the HEC is a boon for masses as a green energy source that produces electricity from water droplets using no acid/ alkali. It has opened a new class of green energy by offering a safe, clean, low cost, environmental friendly green energy source that is safe for human health. In the near future HEC can replace solar cells, fuel cells, and energy storage batteries. Thus, new research directions should be stipulated not only for magnesium ferrite but also for other ferrites, metal oxides based on the HEC working principle.

2.8.8

Conclusions

Use of magnetic materials and their demand have been increasing day by day for their need in energy efficient devices. New material composition and processes have been developed to fabricate energy efficient magnetic materials. In this chapter soft ferrite with lowest magnetization magnesium ferrite has been explored for electric energy generation without utilizing its magnetic properties. Its oxygen defective state and nanoporous character have been utilized for water molecule dissociation at room temperature. The ferrite material along with two electrodes is called a HEC – a revolutionary invention. HECs generate current and voltage by water molecule dissociation on octahedrally coordinated unsaturated surface Mg, Li ions and oxygen vacancy in magnesium ferrite. Hydronium ions trapped in nanopores of ferrite develop enough electric field to dissociate physisorbed water molecules spontaneously. The process of water molecule dissociation is accelerated in a bigger way due to nanopores enhancing current in the cell. Cell pellets of area 2  17 cm2 with Zn and Ag electrodes in DI water generated 150 mA short circuit current and 0.95 V emf. Cell produced 28 mW peak power, although maximum possible power 145 mW. Zinc oxide nanoparticles of good crystallinity are produced by the HEC during its operation as energy source. Blue emission by ZnO nanoparticles has potential application in blue laser and LEDs. A significant value of dielectric constant 5 at 1 MHz is uniquely observed by HEC only. Using DI water without any electrolyte in HECs is a clean and facile process. Hydrogen gas generation using HECs is a novel, challenging method. Hydrogen gas obtained from HECs is also a clean source of energy. A substantial amount, 10 mL, of hydrogen gas is being collected by the HEC method. Thus HECs are a cost effective, completely safe, nonpolluting alternative to portable green energy sources. Moreover the role of nanoporous magnesium ferrite pellets as proton exchange membranes (PEMs) in fuel cells would be also be highly beneficial.

Acknowledgment Authors are thankful to the Director NPL for supporting this work.

Magnetic Materials

231

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54]

Gutfleisch O, Willard MA, Bruck E, et al. Magnetic materials and devices for the 21st century: stronger, lighter, and more energy efficient. Adv Mater 2011;23:821–42. United States Environmental Protection Agency. Sources of greenhouse gas emissions. U.S. Climate Action Report; 2014. Alliance LLC. Magnet guide and tutorial. Available from: http://www.Allianceorg.com. Snelling EC. Soft ferrites; properties and applications. London: Butterworth & Co. (Publishers) Ltd.; 1988. Hitachi Metals Ltd. Available from: http://www.hitachi-metals.co.jp; 2017. Vaisambhayana S, Dincan C, Shuyu C, et al. State of art survey for design of medium frequency high power transformer. In: 2016 Asian conference on energy, power and transportation electrification (ACEPT), Singapore; 2016. p. 1–9. Giauque WF, MacDougall DP. Attainment of temperatures below 11 absolute by demagnetization of Gd2(SO4)3.8H2O. Phys Rev 1933;43:768. Pecharsky VK, Gschneidner Jr. KA. Magnetocaloric effect and magnetic refrigeration. J Magn Magn Mater 1999;200:44–56. Anthony S. GE develops high-tech fridge magnets that could save the world billions of dollars in energy costs. Available from: https://www.extremetech.com; 2014. Binasch G, Grunberg P, Saurenbach F, Zinn W. Enhanced magnetoresistance in layered magnetic structures with antiferromagnetic interlayer exchange. Phys Rev B 1989;39:4828–30. Fert A. The origin, development and future of spintronics. Phys Usp 2008;51:1336. Yuasa S, Nagahama T, Fukushima A, Suzuki Y, Ando K. Giant room-temperature magnetoresistance in single-crystal Fe/MgO/Fe magnetic tunnel junctions. Nat Mater 2004;3:868–71. Ju G, Peng Y, Chang EKC, et al. High density heat-assisted magnetic recording media and advanced characterization – progress and challenges. IEEE Trans Magn 2015;51:1–9. Spaldin NA, Cheong SW, Ramesh R. Multiferroics: past, present, and future. Phys Today 2010;63:38. Klimov A, Tiercelin N, Dusch Y, et al. Magnetoelectric write and read operations in a stress-mediated multiferroic memory cell. Appl Phys Lett 2017;110:222401. Brandlmaier A, Geprägs S, Woltersdorf G, Gross R, Goennenwein STB. Nonvolatile, reversible electric-field controlled switching of remanent magnetization in multifunctional ferromagnetic/ferroelectric hybrids. J Appl Phys 2011;110:043913. Etesami SR, Berakdar J. Efficient thermal energy harvesting using nanoscale magnetoelectric heterostructures. Appl Phys Lett 2016;108:053903. Fiebig M. Revival of the magnetoelectric effect. J Phys D Appl Phys 2005;38:R123. Leung CM, Zhuang X, Xu J, et al. Power conversion efficiency and resistance tunability in coil-magnetoelectric gyrators. Appl Phys Lett 2016;109:202907. Leung CM, Zhuang X, Xu J, et al. Importance of composite parameters in enhanced power conversion efficiency of Terfenol-D/PZT magnetoelectric gyrators. Appl Phys Lett 2017;110:112904. Moses AJ. Electrical steels: past, present and future development. IEE Proc 1990;137:233–45. Armco Steel Corporation. The metallurgy of iron and silicon iron for soft magnetic applications: technical report, Middletown, OH; 1970. Shokrollani H, Janghorban K. Soft magnetic composite materials (SMCs). Journal of Materials Processing Technology 2007;189:1–12. Yurekten S, Kara A, Mardikyan K. International conference on ‘renewable energy research and applications’ (ICRERA), October 20–23, 2013; 2013. Yurekten S, Kara A, Mardikyan K. Energy efficient green transformer manufacturing with amorphous cores. In: International conference on renewable energy research and applications, Madrid, Spain, October, 20–23, 2013; 2013. Georgilakis PS. Environmental cost of distribution transformer losses. Appl Energy 2011;88:3146–55. Moses T. Opportunities for exploitation of magnetic materials in an energy conscious world. Interdiscip Sci Rev 2002;27:100–13. Jeszensky S. History of transformers. IEEE Power Eng Rev 1996;16:9–12. Barret WF, Brown W, Hadfield RA. Conductivity and permeability of iron alloys. R Dublin Soc Trans 1900;7:87–126. Taguchi S. State-of-the-art of electrotechnical steels. Tetsu to Hagane 1976;62:905–15. Honda K, Kaya S. On the magnetisation of single crystals of iron. Sci Rep Tohuki Imp Univ 1926;15:721–54. Goss NP. New development in electrical strip steels characterized by fine grain structure approaching the properties of a single crystal. Trans Am Soc Met 1935;23:511–44. Bitti RR, Candiotti M. Surface energy and primary texture influence on the improvement of goss texture in high-permeability electrical steel. J Magn Magn Mater 1982;26:11–4. Mishra S, Darmann C, Lucke K. On the development of the goss texture in iron-3% silicon. Acta Metall 1984;32:2185–201. Lobanov ML, Rusakov GM, Redikul’tsev AA. Electrotechnical anisotropic steel. Part 1. History of development. Met Sci Heat Treat 2011;53:326–32. Shoji K, Misao N, Tatsuhiko H. Recent progress of high silicon electrical steel in JFE steel. JFE technical report, no. 21; 2016. Florez SAJ. Communication system using ferrite antennas in electrical transmission lines. In: SBMO/IEEE MTT-S international microwave and optoelectronics conference (IMOC); 2007. Eriksson S, Ekergard B, Leijon M. Design for PM generators reflecting the unstable neodymium price. In: Proceedings of the international conference on electrical machines ICEM 2012, Marseille, France; 2012. Bruck E. Developments in magnetocaloric refrigeration. J Phys D Appl Phys 2005;38:R381. Aprea C, Greco A, Maiorino A, Massell C. Magnetic refrigeration: an eco-friendly technology for the refrigeration at room temperature. J Phys Conf Ser 2015;655:012026. Saito AT, Kobayashi T, Kaji S, Li J, Nakagome H. Environmentally friendly magnetic refrigeration technology using ferromagnetic Gd Alloys. Int J Environ Sci Dev 2016;7:316–20. International Energy Agency. World Energy Outlook special report on energy and climate change, France; 2015. Oelke C, Zhang T, Serreze MC. Modeling evidence for recent warming of the arctic soil thermal regime. Geophys Res Lett 2004;31:L07208. International Energy Agency. World Energy Outlook, special report; 2015. ISES. Renewable energy policy network for the 21st century. Renewables 2017 global status report; 2017. Turner JA. A realizable renewable energy future. Science 1999;285:687–9. Appleby AJ. From Sir William Grove to today: fuel cells and the future. J Power Sources 1990;29:3–11. Mench MM, Wang CY, Thynell ST. An introduction to fuel cells and related transport phenomena. Int J Trans Phenom 2001;3:151–76. Bard AJ, Fox MA. Artificial photosynthesis: solar splitting of water to hydrogen and oxygen. Acc Chem Res 1995;28:141–5. Hoffmann MR, Martin ST, Choi W, Bahnemann DW. Environmental applications of semiconductor photocatalysis. Chem Rev 1995;95:69–96. Liu C, Colon BC, Ziesack M, Silver PA, Nocera DG. Water splitting–biosynthetic system with CO2 reduction efficiencies exceeding photosynthesis. Science 2016;352:1210–3. Dutta S. Technology assessment of advanced electrolytic hydrogen production. Int J Hydrogen Energy 1990;15:379–86. Andujar JM, Segura F. Fuel cells: history and updating. A walk along two centuries. Renew Sustain Energy Rev 2009;13:2309–22. Wendt H, Imarisio G. Nine years of research and development on advanced water electrolysis: a review of research program of the commission of the european communities. J Appl Electrochem 1988;18:1–14.

232

Magnetic Materials

[55] Bak T, Nowotny J, Rekas M, Sorrell CC. Photo-electrochemical hydrogen generation from water using solar energy. Materials-related aspects. Int J Hydrogen Energy 2002;27:991–1022. [56] Osterloh FE. Inorganic materials as catalysts for photochemical splitting of water. Chem Mater 2008;20:35–54. [57] Schrauzer G, Guth TD. Photocatalytic rections. 1. Photolysis of water and photoreduction of nitrogen on titanium dioxide. J Am Chem Soc 1977;99:7189–93. [58] Yamaguti K, Sato S. Photolysis of water over metallized powdered titanium dioxide. J Chem Soc Faraday Trans 1985;81:1237–46. [59] Liu X, Wang F, Wang Q. Nanostructure-based WO3 photoanodes for photo electrochemical water splitting. J Phys Chem Chem Phys 2012;14:7894–911. [60] Abdi FF, van de Krol R. Nature and light dependence of bulk recombination in Co-Pi catalyzed BiVO4 photoanodes. J Phys Chem C 2012;116:9398–404. [61] Sivula K, Le Formal F, Gratzel M. Solar water splitting: progress using hematite(a-Fe2O3) photoelectrodes. ChemSusChem 2011;4:432–49. [62] Zoulias E, Varkaraki E, Lymberopoulos N, Christodoulou CN, Karagiorgis GN. A review on water electrolysis. TCJST 2004;4:41–71. [63] Sturzenegger M, Nuesch P. Efficiency analysis for a manganese-oxide-based thermo chemical cycle. Energy 1999;24:959–70. [64] Steinfeld A, Sanders S, Palumbo R. Design aspects of solar thermochemical engineering-A case study: two-step water-splitting cycle using the FeO/FeO redox system. Sol Energy 1999;65:43–53. [65] Kudo A, Miseki Y. Heterogeneous photocatalyst materials for water splitting. Chem Soc Rev 2009;38:253–78. [66] Steinfeld A. Solar hydrogen production via a two-step water-splitting thermochemical cycle based on Zn/ZnO redox reactions. Int J Hydrogen Energy 2002;27:611–9. [67] Kojima M, Sano T, Wada Y, et al. Thermochemical decomposition of H2O to H2 on cation-excess ferrite. J Phys Chem Solids 1996;57:1757. [68] Tamaura Y, Ueda Y, Matsunami J, et al. Solar hydrogen production by using ferrites. Sol Energy 1999;65:55–7. [69] Smestad GP, Steinfeld A. Review: photochemical and thermochemical production of solar fuels from H2O and CO2 using metal oxide catalysts. Ind Eng Chem Res 2012;51:11828–40. [70] Bolton JR. Solar photo production of hydrogen: a review. Sol Energy 1996;57:37–50. [71] Volkov AG. Oxygen evolution in the course of photosynthesis: molecular mechanisms. Bioelectrochem Bioenerget 1989;21:3–24. [72] Kruse A. Supercritical water gasification for biomass based hydrogen production. In: Stolten D, Emonts B, editors. Hydrogen science and engineering: materials, processes, systems and technology. 1st ed. Weinheim: Wiley-VCH Verlag GmbH & Co. KGaA; 2016. [73] Kapdan IK, Kargi F. Bio-hydrogen production from waste materials. Enzyme Microb Technol 2006;38:569–82. [74] Ni M, Leung DYC, Leung MKH, Sumathy K. An overview of hydrogen production from biomass. Fuel Process Technol 2006;87:461–72. [75] Zhang H, Chen G, Zhang Q, et al. Photosynthetic hydrogen production by alginate immobilized bacterial consortium. Bioresou Technol 2017;236:44–8. [76] Cheng J, Liu M, Song W, et al. Enhanced hydrogen production of Enterobacter aerogenes mutated by nuclear irradiation. Bioresou Technol 2016;227:50–5. [77] Khanal SK, Chen WH, Li L, Sung S. Biological hydrogen production: effects of pH and intermediate products. Int J Hydrogen Energy 2004;29:1123–31. [78] Kwon KC, Choi S, Hong K, et al. Wafer-scale transferable molybdenum disulfide thin-film catalysts for photo electrochemical hydrogen production. Energy Environ Sci 2016;9:2240–8. [79] Lodziana Z, Topsoe NY, Norskov JK. Negative surface energy for alumina. Nature 2004;3:289–93. [80] Xu H, Zhang RQ, Ng AMC, et al. Splitting water on metal oxide surfaces. J Phys Chem C 2011;115:19710–5. [81] Li X, Paier J. Adsorption of water on the Fe3O4(111) surface: structures, stabilities, and vibrational properties studied by density functional theory. J Phys Chem C 2016;120:1056–65. [82] Diebold U, Li SC, Schmid M. Oxide suface science. Annu Rev Phys Chem 2010;61:129–48. [83] Sun CH, Liu LM, Selloni A, Lu GQ, Smith SC. Titania water interactions: a review of theoretical studies. J Mater Chem 2010;20:10319–34. [84] Calzolari A, Catellani A. Water adsorption on nonpolar ZnO (1010) surface: a microscopic understanding. J Phys Chem C 2009;113:2896–902. [85] Kunat M, Girol SG, Burghaus U, Woll C. The interaction of water with the oxygen-terminated, polar surface of ZnO. J Phys Chem B 2003;107:14350–6. [86] Onsten A, Stoltz D, Palmgren P, et al. Water adsorption on ZnO (0001): Transition from triangular surface structures to a disordered hydroxyl terminated phase. J Phys Chem C 2010;114:11157. [87] Schaub R, Thostrup P, Lopez N, et al. Oxygen vacancies as active sites for water dissociation on rutile TiO2(110). Phys Rev Lett 2001;87:1–4. [88] Bikondoa O, Pang C, Ithnin R, et al. Direct visualization of defect-mediated dissociation of water on TiO2(110). Nat Mater 2006;5:189–92. [89] Oviedo J, Armas R, Miguel M, Sanz JF. Methanol and water dissociation on TiO2 (110): the role of surface oxygen. J Phys Chem C 2008;112:17737–40. [90] Pang CL, Sasahara A, Onishi H, Chen Q, Thornton G. Noncontact atomic force microscopy imaging of water dissociation products on TiO2(110). Phys Rev B 2006;74:073411. [91] Zhang Z, Bondarchuk O, Kay BD, White JM, Dohnalek Z. Imaging water dissociation on TiO2(110): evidence for inequivalent geminate OH groups. J Phys Chem B 2006;110:21840–5. [92] Walle LE, Ragazzon D, Borg A, Uvdal P, Sandell A. Competing water dissociation channels on rutile TiO2(110). Surf Sci 2014;621:77–81. [93] Pang CL, Lindsay R, Thornton G. Chemical reactions on rutile TiO2(110). Chem Soc Rev 2008;37:2328–53. [94] Zwicker G, Jacobi K. Site-specific interaction of H2O with ZnO single-crystal surfaces studied by thermal desorption and UV photoelectron spectroscopy. Surf Sci 1983;131:179–94. [95] Wander A, Harrison NM. The stability of polar oxide surfaces: the interaction of H2O with ZnO(0001) and ZnO(0001). J Chem Phys 2001;115:2313–6. [96] Kunat St. M, Girol G, Burghaus U, Wo C. The interaction of water with the oxygen-terminated, polar surface of ZnO. J Phys Chem B 2003;107:14350–6. [97] Onsten A, Stoltz D, Palmgren P, et al. Water adsorption on ZnO(0001): transition from triangular surface structures to a disordered hydroxyl terminated phase. J Phys Chem C 2010;114:11157–61. [98] Joseph Y, Kuhrs C, Ranke W, Ritter M, Weiss W. Adsorption of water on FeO(111) and Fe3O4 (111): identification of active sites for dissociation. Chem Phys Lett 1999;314:195–202. [99] Parkinson GS, Novotny Z, Jacobson P, Schmid M, Diebold U. Room temperature water splitting at the surface of magnetite. J Am Chem Soc 2011;133:12650–5. [100] Kundu TK, Rao KH, Parker SC. Atomistic simulation studies of magnetic surface structures and adsorption behavior in the presence of molecular and dissociated water and formic acid. J Colloid Interface Sci 2006;295:364–73. [101] Mulakaluri N, Pentcheva R, Wieland M, Moritz W, Scheffler M. Partial dissociation of water on Fe3O4(001): adsorbate induced charge and orbital order. Phys Rev Lett 2009;103:176102. [102] Salmeron M, Bluhm H, Tatarkhanov M, et al. Water growth on metals and oxides: binding, dissociation and role of hydroxyl groups. Faraday Discuss 2009;141:221–9. [103] Hass KC, Schneider WF, Curioni A, Andreoni W. The chemistry of water on alumina surfaces: reaction dynamics from first principles. Science 1998;282:265–8. [104] Zwicker G, Jacobi K. Site-specific interaction of H2O with ZnO single-crystal surfaces studied by thermal desorption and UV photoelectron spectroscopy. Surf Sci 1983;131:179–94. [105] McCafferty E, Zettlemoyer AC. Adsorption of water vapor on a-Fe2O3. Discuss Faraday Soc 1971;52:239–63. [106] Shimizu Y, Arai H, Seiyama T. Ceramic humidity sensors 1, microstructure and humidity sensitive characteristics of spinel type oxides. Denki Kagaku 1982;50:831–4. [107] Sadaoka Y, Sakai Y. Humidity sensitivity of sintered porous alumina and alumina-bona. Denki Kagaku 1983;51:873–8. [108] De Grotthuss CJT. Sur la décomposition de l'eau et des corps qu'elle tient en dissolution à l'aide de l'electricité galvanique. Ann Chim 1806;58:54–73. [109] Cukierman S. Et tu, Grotthuss! and other unfinished stories. Biochim Biophys Acta 2006;1757:876–85. [110] Ghalamboran M, Saedi Y. TiO2–TiO2 composite resistive humidity sensor: ethanol crosssensitivity. IOP Conf Ser Mater Sci Eng 2016;108:012039. [111] Yamazoe N, Shimizu Y. Humidity sensors: principles and application. Sens Actuators 1986;10:379–98. [112] Shah J, Kotnala RK. Humidity sensing exclusively by physisorption of water vapors on magnesium ferrite. Sens Actuators B 2012;171–172:832–7.

Magnetic Materials

[113] [114] [115] [116] [117] [118] [119] [120] [121] [122] [123] [124] [125] [126] [127] [128] [129] [130] [131] [132] [133] [134] [135] [136] [137] [138] [139] [140] [141] [142] [143] [144] [145] [146] [147] [148] [149] [150] [151] [152] [153] [154] [155] [156] [157] [158] [159] [160] [161] [162] [163] [164] [165] [166] [167] [168] [169] [170]

233

Okamoto T, Suzuki T, Matsui N. EXAFS and XANES study on spinel structure materials and their application to humidity sensor. J Phys Colloq 1986;47:737–40. Shah J, Kotnala RK, Singh B, Kishan H. Microstructure dependent humidity sensitivity of porous MgFe2O4–CeO2 ceramic. Sens Actuators B: Chem 2007;128:306–11. Kotnala RK, Shah J, Singh B, et al. Humidity response of Li-substituted magnesium ferrite. Sens Actuators B: Chem 2008;129:909–14. Shah J, Arora M, Purohit LP, Kotnala RK. Significant increase in humidity sensing characteristics of praseodymium doped magnesium ferrite. Sens Actuators A 2011;167:332–7. Kotnala RK, Shah J, Gupta R. Colossal humido resistance in ceria added magnesium ferrite thin film by pulsed laser deposition. Sens Actuators B: Chem 2013;181:402–9. Tejedor-Tejedor MI, Vichi FM, Anderson MA. Effect of pore structure on proton conductivity and water uptake in nanoporous TiO2. J Porous Mater 2005;12:201–14. Wang W, Virkar AV. A conductimetric humidity sensor based on proton conducting perovskite oxides. Sens Actuators B 2004;98:282–90. Huang C-C, Yeh C-S. Porous cube-like In2O3 nanoparticles and their sensing characteristics toward ethanol. J Mater Sci Technol 2008;24:667–74. Lu C-C, Huang Y-S, Huang J-W, Chang C-K, Wu S-P. A macroporous TiO2 oxygen sensor fabricated using anodic aluminium oxide as an etching mask. Sensors 2010;10:670–83. Dionne GF. A review of ferrites for microwave applications. Proc IEEE 1975;63:777–89. Song J, Wang L, Xu N, Zhang Q. Microwave absorbing properties of magnesium-substituted MnZn ferrites prepared by citrate-EDTA complexing method. J Mater Sci Technol 2010;26:787–92. Gateshki M, Petkov V, Pradhan SK, Vogt T. Structure of nanocrystalline MgFe2O4 from X-ray diffraction, rietveld and atomic pair distribution function analysis. J Appl Cryst 2005;38:772–9. Kotnala RK, Shah J. Green hydroelectrical energy source based on water dissociation by nanoporous ferrite. Int J Energy Res 2016;40:1652–61. Clark JH, Dell RM, Rand DAJ. Hydrogen energy. In: Dell RM, Rand DAJ, editors. Clean energy. London: The Royal Society of Chemistry; 2004. p. 198–241. ISBN: 9780-85404-546-4. Steele BCH, Heinzel A. Materials for fuel-cell technologies. Nature 2001;414:345–52. Brown E. Sea water: its composition, properties and behaviour. Oxford: Butterworth-Heinmann; 1995. p. 29. ISBN: 978-0-7506-3715-2. Hasvold O. Seawater batteries for low power, long term applications. In: Proceedings of the 34th international power sources symposium; 1990. p. 50–3. Kim JK, Lee E, Kim H, et al. Rechargeable seawater battery and its electrochemical mechanism. Chem Electro Chem 2015;2:328–32. Burstein GT, Speckert EIP. Developing a battery using set concrete as electrolyte. ECS Trans Electrochem Soc 2007;3:13–20. Sakai T, Kaneko M, Masuko M. Batteries with fresh concrete for their electrolyte and method for ensuring the filling conditions of concrete on their placing. Jpn Kokai Tokkyo Koho. JP 2007263735; 2007. Byrne A, Holmes N, Norton B. Cement based batteries and their potential for use in low power operations. Mat Sci Eng 2015;96:012073. Meng Q, Chung DDL. Battery in the form of a cement-matrix composite. Cem Concr Comp 2010;32:829–39. Gnomes’land BLLL. Ground soil battery for low-power electric receptors. Rom RO 114053, 6; 1998. Mahowald PH. Soil battery. U.S. Patent Application Pub. US 2006019149; 2006. Meng Q, Kenayeti Y, Chung DDL. Battery in the form of a soil-matrix composite. J Energy Eng 2015;141:1–8. Boreyko JB, Chen CH. Self-propelled dropwise condensate on super hydrophobic surfaces. Phys Rev Lett 2009;103:184501. Chen CH, Cai Q, Tsai C, et al. Dropwise condensation on superhydrophobic surfaces with two-tier roughness. Appl Phys Lett 2007;90:173108. Dorrer C, Ruhe J. Wetting of silicon nanograss: from superhydrophilic to superhydrophobic surfaces. Adv Mater 2008;20:159–63. Miljkovic N, Preston DJ, Enright R, Wang EN. Jumping-droplet electrostatic energy harvesting. Appl Phys Lett 2014;105:013111. Huang JY, Zhu YT, Jiang H, Lowe TC. Microstructures and dislocation configurations in nanostructured Cu processed by repetitive corrugation and straightening. Acta Mater 2001;49:1497–505. Zhou W. Microscopic study of crystal defects enriches our knowledge of materials chemistry. J Mater Chem 2008;18:5321–5. Kovtunenko PV. Defect formation in spinels in oxygen nonstoichiometry. Glass Ceram 1997;54:5–6. Anantharaman MR, Reijne S, Jacobs JP, et al. Preferential exposure of certain crystallographic planes on the surface of polycrystalline spinel ferrites – a study by LEIS. J Mater Sci 1999;34:4279–83. Yamazoe N, Shimizu Y. Humidity sensors: principles and applications. Sens Actuators 1986;10:379–98. Rodriguez BJ, Yang WC, Nemanich RJ, Gruverman A. Scanning probe investigation of surface charge and surface potential of GaN-based heterostructures. Appl Phys Lett 2005;86:112115. Bhatt K, Tan S, Karumuri S, Kalkan AK. Charge-selective raman scattering and fluorescence quenching by “nanometal on semiconductor” substrates. Nano Lett 2010;10:3880–7. Tsyganenko AA, Filimonov VN. Infrared spectra of surface hydroxyl groups and crystalline structure of oxides. J Mol Struct 1973;19:579–89. Gurlo A, Riedel R. Insitu and operando spectroscopy for assessing mechanisms of gas sensing. Angew Chemie Int Ed 2007;46:3826–48. Jacobs JP, Maltha A, Reintjes JGH, et al. The surface of catalytically active spinels. J Catal 1994;147:294–300. Dell RM, Rand DAJ. Energy and production and use. In: Clark JH, editor. Clean energy. Cambridge: Royal Society of Chemistry; 2004. Delahay P. Double layer and electrode kinetics. London; NewYork, NY: Interscience Publishers (Division of John Wiley & Sons); 1965. ISBN-10: 0470206004. Bockris JO'M, Reddy AKN. Modern electrochemistry: an introduction to an interdisciplinary area. New York, NY: Plenum Press; 1970. ISBN: 9781461574675. Barsoukov E, Macdonald JR, editors. Impedance spectroscopy: theory, experiment and applications. 2nd ed. Hoboken, NJ: Wiley; 2005. ISBN: 978-0-471-64749-2. Christophersen JP, Glenn DF, Motloch CG, et al. Electrochemical impedance spectroscopy testing on the advanced technology development program lithium-ion cells. In: Proceedings of vehicular technology conference, IEEE; 2002. Sekine I, Masuko A, Senoo K. Corrosion behavior of AISI 316 stainless steel in formic and acetic acid solutions. Corrosion 1987;43:553–60. Yeh YC, Tseng TY. Analysis of the dc and ac properties of K2O-doped porous Ba0.5Sr0.5TiO3 ceramic humidity sensor. J Mater Sci 1989;24:2739–45. Zhang H, Li Z, Wang W, Wang C, Liu L. Na þ -doped zinc oxide nanofiber membrane for high speed humidity sensor. J Am Ceram Soc 2010;93:142–6. Aoki T, Hatanaka Y, Look DC. ZnO diode fabricated by excimer-laser doping. Appl Phys Lett 2000;76:3257–8. Janotti A, de Walle CGV. Fundamentals of zinc oxide as a semiconductor. Rep Prog Phys 2009;72:126501. Panda D, Tseng TY. One-dimensional ZnO nanostructures: fabrication, optoelectronic properties, and device applications. J Mater Sci 2013;48:6849–77. Xu XY, Xu CX, Lin Y, Li J, Hu JG. Comparison on photoluminescence and magnetism between two kinds of undoped ZnO nanorods. J Phys Chem C 2013;117:24549–53. Chawla S, Jayanthi K, Kotnala RK. Room-temperature ferromagnetism in Li-doped p-type luminescent ZnO nanorods. Phys Rev B 2009;79:125204. Deng W, Jin L, Zhang B, et al. A flexible field-limited ordered ZnO nanorod-based self-powered tactile sensor array for electronic skin. Nanoscale 2016;8:16302–6. Lanje AS, Sharma SJ, Ningthoujam RS, Ahn JS, Pode RB. Low temperature dielectric studies of zinc oxide (ZnO) nanoparticles prepared by precipitation method. Adv Powder Technol 2013;24:331–5. Sato K, Katayama-Yoshida H. Material design for transparent ferromagnets with ZnO-based magnetic semiconductors. Jpn J Appl Phys 2000;39:L555–8. Wang ZL. Novel nanostructures of ZnO for nanoscale photonics, optoelectronics, piezoelectricity, and sensing. Appl Phys A 2007;88:7–15. Mouslmani M, Patra D. Modifying emission of ZnO nanoparticles in ZnO interceded polymer based hierarchical ordered nanocapsules. Mater Lett 2015;143:135–9. Yang Y, Guo W, Wang X, et al. Size dependence of dielectric constant in a single pencil-like ZnO nanowire. Nano Lett 2012;12:1919.

234

Magnetic Materials

[171] Gaevs’kyi VR, Nechyporuk BD, Novoselets’kyi NY, Rudyk BP. Electrolytic fabrication of zinc oxide nanoparticles. Ukr J Phys 2013;58:385–8. [172] Malevu TD, Ocaya RO. Synthesis of ZnO nanoparticles using a zinc-air cell and investigation of the effect of electrolyte concentration. Int J Electrochem Sci 2014;9:8011–23. [173] Shah J, Kotnala RK. Rapid green synthesis of ZnO nanoparticles using a hydroelectric cell without an electrolyte. J Phys Chem Solids 2017;108:15–20. [174] Chandrappa KG, Venkatesha TV, Vathsala K, Shivakumara C. A hybrid electrochemical-thermal method for the preparation of large ZnO nanoparticles. J Nanopart Res 2010;12:2667–78. [175] Fu YS, Du XW, Kulinich SA, et al. Stable aqueous dispersion of ZnO quantum dots with strong blue emission via simple solution route. J Am Chem Soc 2007;129:16029–33. [176] Li JW, Liu XJ, Yang LW, et al. Photoluminescence and photoabsorption blueshift of nanostructured ZnO: skin-depth quantum trapping and electron-phonon coupling. Appl Phys Lett 2009;95:031906. [177] Van Dijkena A, Meulenkamp EA, Vanmaekelbergh D, Meijerink A. Identification of the transition responsible for the visible emission in ZnO using quantum size effects. J Lumin 2000;90:123–8. [178] Djurisic AB, Leung YH, Tam KH, et al. Defect emissions in ZnO nanostructures. Nanotechnology 2007;18:095702. [179] Liu D, Lv Y, Zhang M, et al. Defect-related photoluminescence and photocatalytic properties of porous ZnO nanosheets. J Mater Chem A 2014;2:15377–88. [180] Puri PS. Hydrogenation of oils and fats. J Am Oil Chem Soc 1980;57:A850–4. [181] Ramachandran R, Menon RK. An overview of industrial uses of hydrogen. Int J Hydrog Energy 1998;23:593–8. [182] Durgutlu A. Experimental investigation of the effect of hydrogen in argon as a shielding gas on TIG welding of austenitic stainless steel. Mater Des 2004;25:19–23. [183] Lyu Q, Qie Y, Liu X, et al. Effect of hydrogen addition on reduction behavior of iron oxides in gas-injection blast furnace. Thermochim Acta 2017;648:79–90. [184] Ye M, Gong J, Lai Y, Lin C, Lin Z. High-efficiency photoelectrocatalytic hydrogen generation enabled by palladium quantum dots-sensitized TiO2 nanotube arrays. J Am Chem Soc 2012;134:15720–3. [185] Steinfeld A. Solar thermochemical production of hydrogen – a review. Sol Energy 2005;78:603–15. [186] Shah J, Jain S, Shukla A, Gupta R, Kotnala RK. A facile non-photocatalytic technique for hydrogen gas production by hydrolectric cell. Int J Hydrog Energy 2017;42 (52):30584–90.

Relevant Websites http://corefficientsrl.com/electrical-transformer-core-brief-history/ CorEfficient – A Brief History of the Electrical Transformer Core. https://yearbook.enerdata.net/ Enerdata – Global Energy Statistical Yearbook 2017. https://www.enerdata.net/publications/reports-presentations/peak-energy-demand-co2-emissions-2016-world-energy.html Enerdata – Global Energy Trends 2016 Report. https://energy.gov/eere/fuelcells/fuel-cells Energy.Gov – A Brief Look on Fuel Cells. http://www.fchea.org/ Fuel Cell and Hydrogen Energy Association. http://www.iea-coal.org.uk/documents/82861/8363/CO2-abatement-in-the-iron-and-steel-industry,-CCC/193 IEA Clean Coal Centre. http://www.ipcc.ch/ Intergovernmental Panel on Climate Change. http://www.iahe.org/ International Association for Hydrogen Energy. https://www.hydropower.org/ International Hydro Power Association. http://www.zinc.org/compounds/ International Zinc Association. http://www.elect.mrt.ac.lk/Transformer_history_2000.pdf University of Moratuwa. http://www.ngsa.org/ Natural Gas Supply Association. http://www.uni-koeln.de/Baei53/PhotoHydrogen/ PhotoHydrogen – Light Driven Production of Hydrogen Gas. http://www.renewableenergyworld.com/index/tech.html Renewable Energy World Magazine – About Renewable Energy. http://www.seia.org/ Solar Energy Industries Association, America. www.xs4all.nl/Bwjtbeek/history1.html The Complete History of the Transformers. http://www.electrochem.org/redcat-blog/new-catalyst-efficient-water-splitting/ The Electrochemical Society – New Catalyst for More Efficient Water Splitting. https://www.eia.gov/outlooks/ieo/world.php U.S. Energy Information Administration – International Energy Outlook 2016. https://www.eia.gov/energyexplained/?page=renewable_home U.S. Energy Information Administration – Renewable Energy Sources and Its Role. https://www.worldsteel.org/steel-by-topic/sustainability/environmental-sustainability/climate-change/data-collection.html Worldsteel Association – CO2 Emission Data Collection. http://www.worldsteel.org/steelstory/ Worldsteel Association – The Steel Story. http://www.wwindea.org/ World Wind Energy Association.

2.9 Composite Materials Sergio Oller, UPC Technical University of Catalonia (Barcelona Tech), Barcelona, Spain Sergio A Oller Aramayo and Liz G Nallim, National University of Salta, Salta, Argentina Xavier Martinez, UPC Technical University of Catalonia (Barcelona Tech), Barcelona, Spain r 2018 Elsevier Inc. All rights reserved.

2.9.1 2.9.2 2.9.3 2.9.3.1 2.9.3.2 2.9.3.3 2.9.4 2.9.5 2.9.5.1 2.9.5.2 2.9.5.3 2.9.5.4 2.9.5.4.1 2.9.5.4.2 2.9.5.4.3 2.9.5.5 2.9.5.5.1 2.9.5.5.2 2.9.5.6 2.9.5.7 2.9.6 2.9.7

General Introduction Hydrokinetic Turbines – Introduction Rotor Hydrokinetic Turbine Design (Fluid–Solid Interaction) Hydrofoil Profile and Rotor Simplified Hydrofoil Analytical Pre-Design Brief Comment About the 3D Fluid-Dynamic Numerical Simulation of the Hydrofoil Blade Structural Rotor Design of a Hydrokinetic Turbine: A Composite Material Structure as a Solution Numerical Model for the Analysis of Composite Material Rotor Classical Mixing Theory Serial/Parallel Mixing Theory for One Layer Serial/Parallel Mixing Theory for a Stacking Layers Composites Additional Formulations Used by S/P Mixing Theory to Simulate Reinforcement Composite Materials Anisotropy using a mapping space theory Fiber–matrix debonding Tangent constitutive stiffness tensor Local Plastic Damage Model for a Component Material Definition of the plastic damage variable Definition of the cohesion or uniaxial strength evolution law c dp Local Damage (Elastic Degradation) Model for a Component Material Global Composite Homogenized Laminate Damage Index “Micromodel” Versus “Mixing Theory”: Conceptual Comparative Behavior Example “Micromodel” Versus “Mixing Theory”: Conceptual “Fiber–Matrix Displacement” (Debonding) Behavior Example Numerical Simulation of a Structural Analysis of a “Composite Material Rotor-Hydrofoil” of a WCT Geometry, Boundary Conditions and Finite Element Mesh Action on the Hydrofoil’s Rotor Numerical Simulations of the Rotor Made of Steel and Composite Material Conclusions Future Works

2.9.8 2.9.8.1 2.9.8.2 2.9.8.3 2.9.9 2.9.10 References Further Reading Relevant Websites

Nomenclature A a0 a amax b C c cy cx dx, dy, dz g KL Lavg Lx p

Swept area Design attack angle (degree) Real angle of attack (degree) Maximum aerodynamic profile’s a (degree) Hydrofoil’s camber angle (degree) Airfoil chord (m) Relative flow velocity Lift coefficient Drag coefficient VC2 dimensions Gravity acceleration (m/s2) Ratio between X and Lavg Chord average Airfoil’s chord for each x (m) Pressure (Pa)

Comprehensive Energy Systems, Volume 2

r R SP SPx T t y u v l W o x X nwing F E

doi:10.1016/B978-0-12-809597-3.00220-0

236 237 238 238 239 240 241 241 241 242 244 244 245 245 246 246 248 249 249 250 251 254 259 261 261 262 263 263 264 265 265

Fluid density (Kg/m3) Radius of the rotor Shape factor Shape factor for each x Torque (Nm) Time (s) Hydrofoil’s sustentation angle (degree) Blade linear rotational speed (m/s) Absolute flow velocity (m/s) TSR: Ratio between c and u Power (w) Angular speed (rad/s) Length of each hydrofoil segment (m) Wingspan (m) Number of blades of the rotor Load (N) Young modulus (N/m2)

235

236

Composite Materials

(n) fT e e eP i e eV ki Ci s si ℂSi ; ℂTi es, ep ss, sp NP PP As

2.9.1

Poisson ratio Tension strength (N/m2) Thickness (m) Material strain Plastic material strain i-Component material strain Volumetric deformation Volumetric participation coefficient of the ith component Free energy of the ith component Composite material stress (N/m2) i-Component material stress (N/m2) i-Secant and i-Tangent Constitutive tensors Serial and parallel strains Serial and parallel stresses (N/m2) Fourth-order parallel projector tensor Fourth-order complementary serial projector tensor Fourth-order stress mapping tensor

Ae r e Z Ξm (fR)fib (fN)fib (fN)mat (fN)fib-mat FP(s;qp), Fd(s0;qd) qp, qd GP(s;qp) Gd Gdc dp, dd dL

Fourth-order strain mapping tensor Isotropic fictitious stress (N/m2) Isotropic fictitious strain Kinematic plastic flow orientation Plastic dissipation energy Fiber strength (N/m2) Nominal fiber strength (N/m2) Nominal matrix strength (N/m2) Nominal fiber-matrix strength (N/m2) Plastic and damage thresholds Group of the plastic and damage internal variables Plastic potential Function of the damage strength softening evolution Damage compression energy of the material Plastic and damage internal variables Structural damage index

General Introduction

According to United Nations, 20% of the global population does not have access to electricity, and a further 14% lack reliable access [1]. The use of an axial flow rotor turbine in remote area was claimed to have for pumping irrigation and electrical power generation. Hydrokinetic turbines bring newer, greater possibilities and advantages for hydroelectric power generation. There are applications in water currents of 0.5 m/s or greater [2]. Development of renewal energy production in rivers and channels still preserve a very interesting power production potential, being not subjected to the classical hydraulic power exploitation. This solution avoids the construction of expensive dams and reduces considerably the environmental impact produced by classical hydropower generation [3]. Low speed flux and lack of depth are the main obstacles in hydrokinetic operation. For this reason, achieving a very high lift rotor to take the maximum advantage of the kinetic energy of a slow velocity water flow, which belongs to a lowland river type, is a very important topic. The use of a high lift aerodynamic/hydrodynamic profile and composite material for the blades serve to accomplish the task. The main purpose of this chapter is to describe a general procedure to achieve a very low inertia rotor minimizing the start-stop effect for the axial water flow turbine, in which it is important to take the maximum advantage of the kinetic energy. The composite hydrofoil of the turbine rotor can be designed using reinforced laminate composites, to obtain the maximum strength and lower rotational inertia. The mechanical and geometrical parameters involved in the design of this fiber-reinforced composite material are the fiber orientation, number of layers, stacking sequence, and laminate thickness. For this reason, it will be briefly described the features of hydrokinetic turbines (water current turbine (WCT)) for river use, their basic design requirements and the response by using matrix-reinforced composite structures. Design requirements for these turbines need a numerical process simulation of the fluid-dynamic problem coupled with the behavior of the structure made of composite materials. From the structural viewpoint it is necessary the use of an advanced composite material formulation that allows an appropriate structural design. For this purpose, a “mixing theory” [4–16] and/or “homogenization theory” [16–25] of simple substances are used, with a mapping spaces formulation [26] that allow considering the anisotropy of the constituent and composite materials in the most general possible way, and a fiber–matrix debonding formulation [5,9,11,14]. Moreover, within these general formulations, it is also taken into account the nonlinear mechanical behavior of the component materials (matrix and fiber), which allows to know precisely the limits of participation of each one of them into the composite. The study of composite materials has been one of the major objectives of computational mechanics in the last decade. The numerical simulation of orthotropic composite materials has been done by means of the average properties of their constituents, but this approximation, no model has been found able to work beyond the constituents’ elastic limit state. Thus, these procedures are limited to the numerical computation to elastic cases. Different theories have been proposed to solve this problem, taking into account the internal configuration of the composite to predict its behavior. The two most commonly used are herein remarked.

•

Homogenization theory: This method deals with the global problem of composite material in a two-scale context. The macroscopic scale uses the composite materials to obtain the global response of the structure; composites are considered as homogeneous materials in this scale. The microscopic scale corresponds to an elemental characteristic volume in which the microscopic fields inside the composite are obtained; this scale deals with the component materials. Homogenization theory assumes a periodical configuration of the composite material to relate these two scales [17–25].

Composite Materials

237

S1223 RTL

0.3 0.2 0.1 0 −0.1 −0.2 −0.3

0

0.2

0.1

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Fig. 1 Hydrofoil S1223 profile.

u v

Plane of rotation  

z x



c y

Chord(L)

Fig. 2 Notation for angles and velocities on the blade profile.

•

Mixing theory: The first formulation of the mixing theory corresponds to Trusdell and Toupin [6] and it is based in two main hypothesis: (1) All composite constituents have the same strains. (2) Each constituent collaborates to the composite behavior according to its volumetric participation. The main problem of the mixing theory is the iso-strain condition, which forces a parallel distribution of the constituents in the composite. Some improvements to the original formulation can be found in Refs. [4–6,8–16].

In this chapter a brief introduction to the “Serial/Parallel theory of mixtures,” a more advanced formulation than the classic one, is presented. The election of the mixing theory instead of a homogenization theory is based in the better relation between model accuracy versus computational cost provided by the former one [19]. A homogenization theory requires a micromodel for each point of the structure that becomes nonlinear. Despite the advances made in strategies to reduce the amount of micromodels solved [20,23,25,27,28], the resolution of a real structure with this procedure generates such a big amount of degrees of freedom that the calculation is beyond the computation capabilities of nowadays personal computers. On the other hand, the mixing theory does not increase the degrees of freedom of the problem, as it is only present in the constitutive section of the finite element code.

2.9.2

Hydrokinetic Turbines – Introduction

This section provides an overview of a WCT allowing understand the hydrodynamic basis for their design and its requirements for the structural function [29–35]. Then, Section 2.9.5 presents the basis for the analysis of its structure made-up in a reinforced composite material, and a simple application in the examples section are shown too. Rivers kinetic energy for electric power generation is a very valuable alternative source. This emerging class of renewable energy technology, the hydrokinetic conversion device (HCD), offers ways to capture the energy of flowing water without the impoundment or diversion of the conventional hydroelectric facilities based on dams and penstocks. Hydrokinetic technologies are designed for deployment in natural streams, like rivers, tidal estuaries, ocean currents, and in some constructed waterways [29–35]. As opposed to the rigid, expensive, and environmentally aggressive construction of tidal barrages, the modularity and scalability of hydrokinetic devices are attractive features [30].

238

Composite Materials

X

X

Fig. 3 Length of the segments of the Rotor-hydrofoil.

Outlet

dy

Inlet

dz

dx

Fig. 4 VC2 confined fluid domain control volume.

River streams and other artificial channels have potential for generating electric power through several hydrokinetic energy technologies. This nascent class of renewable energy technology is being strongly considered as an exclusive and unconventional solution falling within the area of both in-land water resource and marine energy [31]. Conventional large or small hydroelectric systems use reservoirs and penstocks to create an artificial water head and extract the potential energy of downwardly falling water through suitable turbo-machinery. In contrast, a river turbine, which could be built as a free-rotor or part of a channel augmented system, may provide an effective alternative mean for generating power. Such systems would potentially require little or no civil work, causing less environmental impact [32,33]. Khan, Iqbal, and Quaicoe [32] showed values that indicate the possibility of higher energy capacity through a river turbine when compared to an equally sized wind energy converter. Wind turbines are usually designed to operate with rated wind speed of 11–13 m/s while, in contrast, river turbines with augmentation channels could be designed for low effective water velocities of 1.75–2.25 m/s or even higher, depending on site resources. Unlike wind energy, the size of these engines is a limitation for this type of energy generation and must be reduced according to the river depth. Another drawback is the low flow velocity, and it requires a set of blades and rotor with a specific design to generate the greater amount of kinetic energy as possible from the water flow [34]. This chapter describes a general procedure for an efficient fluid-mechanical design of the rotor’s blades. The use of high lift airfoils, and composite materials structural design for low rotational inertia, guarantees the hydrodynamics efficiency. Thus, the chapter is structured taking into account the analysis of this axial hydrokinetic river turbine as the fluid-dynamic design of the rotor turbine, and the structural design of the rotor by composite materials. These areas converge in a multidisciplinary methodology depicted in Fig. 5.

2.9.3

Rotor Hydrokinetic Turbine Design (Fluid–Solid Interaction)

Hydrokinetic turbines, unlike conventional hydraulic turbines utilize the kinetic energy of river/channels water for power generation. The performance of these turbines depends of the number of blades, tip speed ratio (TSR), type of airfoil, blade pitch, chord length and twist and its distribution along the blade span [36]. Knowing the inlet and outlet pressure in the microscale volume control (VC1), a procedure for the rotor design of a hydrokinetic turbine for riverbed operation is described in this section. The study is focused on the conditions of a standard largemedium sized lowland river. The structural analysis of this rotor engineered in composite materials with reduced inertia and better functionality for low speed currents fluvial beds is described in Sections 2.9.4 and 2.9.5. The results of the numerical simulation of the composite rotor structure can also be found in the section of examples.

2.9.3.1

Hydrofoil Profile and Rotor

Inside of the microscale control volume (VC2) a composite material rotor turbine is placed. A brief hydrofoil design of its profile is here presented.

Composite Materials

239

Starting the process

Fluid–dynamic design (Section 3) Returns: i, pi

Variable transference: New geometry

Variable transference: Velocities and pressures

Composite material rotor design (Section 5) Returns: ij, ui

End of process Fig. 5 Flow diagram of solid–fluid interaction of numeric simulation.

The supplied turbine power W is directly proportional to the machine’s operating angular speed o and its torque T produced at that specific speed, W ¼T
o

ð1Þ

If more lift is obtained by one blade, more torque and angular velocity will be obtained by the turbine. This commitment is achieved by selecting the S1223 foil [35], which belongs to the high lift low Reynolds profiles class (see Fig. 1). Initially designed as an airfoil for air working conditions, the S1223 profile has also been tested as a hydrofoil under water conditions operation, showing very good operational qualities [37]. A rotor with S1223 hydrofoil profile keeps the proper balance between lift and drag and maintains an attached flow in the hydrofoil neighborhood. In consequence, this rotor has a better pressure distribution and presents hydrodynamic stability, preventing interference with the rest of the hydrofoils forming the rotor.

2.9.3.2

Simplified Hydrofoil Analytical Pre-Design

For turbine application, hydrofoil must be designed starting from the premise that it has to maintain fluid mechanics parameters (such as angle of attack, homogeneous pressure distribution, etc.) along the whole wingspan, despite the fact that rotary operating _ along the blade axis (which gets higher the nearer the point is from the conditions produce different linear velocity of rotation (u) wingtip). Working with this condition involves the variation of the blade geometry parameters (like camber angle, airfoil chord, etc.) in relation with the wingspan axis. Fig. 2 shows the notation for angles and velocities on the blade profile, where v is the absolute flow velocity in the microscale VC2 volume control, u_ represents the blade’s linear rotational speed and c is the relative flow velocity. The angle of attack a is an aero-hydrodynamic angle defined between c and the airfoil chord, and depends on the airfoil profile and its camber angle b. Instead, camber angle b represents a mechanical angle, defined between the hydrofoil chord and its plane of rotation. By combining hydrodynamics and mechanical angles, the sustentation angle (y) appears which is very useful to obtain the variation camber angle in a rotating blade. Parameters involving the use of a S1223 profile working as a non-twisted, nonrotatory and unturbined designed hydrofoil are explained below [37]. The suitable angle of attack occurs at the optimum angle of attack a0 ¼ 10 degree, which is considered as a starting parameter of the design sequence; it involves lift coefficient cy ¼ 2.2 and drag coefficient cx ¼ 0.046. Lift coefficient can raise until it reaches its maximum at amax ¼15 degrees, but beyond that angle, detachment of the boundary layer will happen, dropping lift coefficient and increasing drag coefficient enormously [38].

240

Composite Materials

The TSR or l is a nondimensional parameter that is defined by taking the relationship between the absolute river flow axial flow _ and it is given by velocity c and the blade speed turbine rotor u, l ¼ ðo
R=vÞ

ð2Þ

where R is the rotor radius. According to Betz’s law [39], turbine mechanical power W specified for axial turbines depends on the flux density r and flow speed v in VC2 volume control; both values are fixed by the river flow, and so these parameters are fixed as initial conditions and will not be modified during the process of the rotor design. According to this, rotor nominal power can be established, and is computed from
 ð3Þ W ¼ 8
r
A
v3 =27 The swept area (A) is the unique variable in Eq. (3), and it depends on R (radius of the rotor). Despite the rotating condition, it is necessary to maintain the angle of attack along the wingspan; this scenario permits to keep the rotor’s fluid-dynamic stability. These commitments are accomplished by varying the geometry parameters of the hydrofoil chord size L and camber angle b, along the wingspan. To achieve this goal the Blade Element Theory can be used; according to Froude [40,41], the airfoil’s total length (X) is split in several segments, and each one is designed individually as x (Fig. 3). Sustentation angle y (Fig. 2) is obtained by means of Eq. (4), as follows: y ¼ arccotððo
xÞ=vÞ

ð4Þ

The chord size of the airfoil is therefore computed for each segment Lx by Lx ¼

SPx x cy nwing

ð5Þ

where nwing is the actual number of airfoils in the rotor, cy is the lift coefficient corresponding to a defined profile section at a certain radius x, and the airfoil shape factor SPx can be computed by a curve approximation given by
 ð6Þ SPx ¼ 2:2762
SRx 1:323 In Eq. (6) the nondimensional parameter SRx is given by

SRx ¼ ðTSR
xÞ=X

ð7Þ

As a result of the chord modification during the process by Eq. (5), the initial attack angle a0 has to be recalculated too through Eq. (9), obtaining a new angle of attack an for each chord Lx in each segment x. For this recalculation, the KL parameter, which represents a relationship between the wingspan and the average of the chord, Lavg is necessary, KL ¼ R=Lavg

ð8Þ

As the length camber and the chord angles have been modified for each wingspan segment, the angle of attack must be verified for each section through the following expression:
 an ¼ a0 þ cy =0:11
ð1 þ ð3=KLÞÞ ð9Þ

Solving from Eq. (2) to Eq. (9), the airfoil parameters can be obtained. These parameters allow the definitive design of the turbine blade. Blade parameters are strictly germane with the rotor composition through Eq. (10), which represents the ideal number of blades nwing included in the rotor according to flow and geometry parameters. A higher torque on turbine will be obtained if a higher number of blades will be included in the rotor. This condition also simplifies the starting of the turbine, ergo is a good design requirement to have the more possible number of hydrofoils in the rotor.
 nwing ¼ ðSPx
xÞ= Lx
cy ð10Þ

2.9.3.3

Brief Comment About the 3D Fluid-Dynamic Numerical Simulation of the Hydrofoil Blade

An ideal representation of the working turbine operation can be performed by a numerical simulation. The model is a confined fluid domain (dx  dy  dz) rendering the underwater operation without free surface (VC2). The hydrofoil rotor is located inside that control volume that is made from a box of appropriate measures, as shown in Fig. 4. The dimensions are chosen so as a steady flow is needed at the boundaries of the box. Flow with x direction will cross from the inlet surface to the outlet surface. The rest of domain surfaces have wall condition. Fluid mechanics governing equations for incompressible flows (r¼cte) involves mass conservation condition (Eq. (4)), ∇
v¼0

ð11aÞ

and Navier-Stokes equation [9,38],   ∂ ∂  rðnÞi ðnÞj ¼ rðnÞj þ ∂t ∂xi where v is the velocity field inside the VC2 control volume.

  ∂p ∂ ∂ðnÞi ∂ðnÞj þ þ m þ rgj þ Fj ∂xj ∂xi ∂xj ∂xj

ð11bÞ

Composite Materials

241

Notice that the governing equations system is constituted by four equations and four unknowns, which are pressure and the three vector components of the field velocity; so numerical techniques are necessary for this treatment. A finite element variational multiscale simulation method (FEVMS) [42,43] can be applied as the resolution method.

2.9.4

Structural Rotor Design of a Hydrokinetic Turbine: A Composite Material Structure as a Solution

Multi-laminated composite structures are an ever-increasingly important topic in the fields of fabrication of mechanical, aerospace, marine, and machinery industries due to their advantages such as durability (no corrosion – lower maintenance cost), survivability (fire resistance, crash energy absorption), excellent resistance against cyclic loading (low fatigue), reparability (restoration and repair), etc. [44]. Multilayered fiber-reinforced material systems can offer versatility in composite design because the stacking sequence of each orthotropic layer can take full advantage of the superior mechanical properties in terms of its strength, stiffness, and total weight. One of the goals in design of multilayered composite structure is to increase its strength while lowering its weight/rotational inertia with a given set of fibrous materials. Laminate of fiber-reinforced composites are very useful when low weight/rotational inertia together high strength/stiffness are required, like the case of axial water turbines. As an additional advantage, it is possible to fit the weight without downgrade the efficiency through the design of the fiber orientation, fraction reinforcement volume, choice of large or short fibers, layer thickness, and stacking sequence. A compact hydrokinetic turbine design, its condition of axial flow and its low rotational inertia due to the composite material rotors, confer the functionality at low speed fluvial beds, avoiding the requirement of great earthworks, and expensive civil constructions. Next section describes the structural design and analysis of a turbine rotor made-up of a laminate of fiber-reinforced composites material using the S/P mixing theory [4,14]. This formulation manages several linear and nonlinear constitutive models simultaneously including damage and plasticity behavior and provides the homogenized damage composite index, taking into account the orthotropic fiber-matrix reinforced and its debonding effect. The composite material used is a laminate composed by epoxy matrix reinforced with long carbon fibers, allowing obtaining better values of stiffness and strength with a smaller weight and rotational inertia. For the structural analysis is necessary to consider a “composition of several single anisotropic materials,” together with “fluid–dynamic interaction” by means of a “staggered technique approach.” These three aspects are depicted in Fig. 5; the proper coupling allows the fiber-reinforced composites rotor design taking into account the successive geometric and mechanical changes of each component materials that forming the composite. Section 2.9.3 describe the procedure for the rotor fluid-dynamic analysis and obtain the state of pressures and speeds will be applied on each point of the rotor blades. Thus, the fluid pressures and speeds distribution in the axial camera of the turbine are input data for the rotor structural analysis, employing composite materials. These operations can be carried out through a “staggered” procedure [13,45], solving each problem per time as shown in Fig. 5.

2.9.5

Numerical Model for the Analysis of Composite Material Rotor

This section presents the fundamental concepts for the structural design of turbine rotor hydrofoil in fiber-reinforced composite material [14]. For this purpose several constitutive damage models [5,14,46] managed by an orthotropic S/P mixing theory [4,14,15,47], anisotropy mapping space [14,26,46], and fiber debonding strategy [14] are explained. Assembling these numerical models provide a powerful formulation and allows the evaluation of a global homogenized laminate damage index, which comes from the damage index provided by the local constitutive damage model. This section contains the presentation and explanation of the following formulations: Orthotropic S/P mixing theory for the laminate composition material, Constitutive plastic damage model for a single material, Constitutive damage model for a single material, Some numerical strategies, and Homogenized laminate damage index definition.

2.9.5.1

Classical Mixing Theory

The classical rule of mixtures, originally developed by Trusdell and Toupin [6,7,11,14,26,44], uses a phenomenological approach based on macroscale continuum mechanics for the composite mechanical analysis, and is suitable for describing the mechanical behavior of each point of a composite solid. This formulation is based on the interaction principle of compounding substances of the composite material. The following basic assumptions are considered: (1) (2) (3) (4)

A set of compounding substances participate in each infinitesimal volume of the composite. Each component contributes in the composite behavior proportionally to their volumetric participation. All the components have the same strain (compatibility or closure equation). Each component volume is much smaller than the total composite volume.

The second hypothesis involves a homogenous distribution of all the component substances at each point of the composite. The different component substance interaction and their corresponding constitutive law determine the composite material’s behavior and it depends on the volume percentage participation of each component and its distribution in the composite.

242

Composite Materials

Materials with different behaviors can be combined (elastic, damage, elastoplastic, etc.), each one representing an evolutionary behavior governed by its own law. The third hypothesis says that in the absence of atomic diffusion the following compatibility condition is satisfied for each of the composite material phases: e ¼ 1e ¼ 2e ¼ … ¼ ne

i

ð12Þ

where e and e (i¼ 1,…n) represent the composite material strain and the ith component strain of such composite material, respectively. The composite material’s free energy [14,48–52] is given by the additive composition of the free energy of each of the component materials considered as a function of its volumetric participation, thus mCðee ; qÞ ¼

n X

k¼1

kk mk Ck ðee ; qk Þ

)

C¼

n X

k¼1

kk

mk Ck m

ð13Þ

where m is the mass density of the composite material, mk is the mass density of the kth component, kk is the volumetric participation coefficient of the kth component, Ck is the free energy of the kth component, qk are the internal variables of each compounding models defining the nonlinear behavior of any generic component [22–24,28,48]. The weighting factor or volumetric participation coefficient kk gives the contribution of each phase and is obtained by the volumetric participation of each of the component materials with respect to the total volume. kk ¼

dVk dV0

ð14Þ

where Vk represents the material kth component volume and V0 is the total volume of the composite material. The volumetric participation coefficients of the different components of the composite material must satisfy the following condition: n X

k¼1

kk ¼ 1

ð15Þ

in which the free energy can be recovered for a single component materials and the mass conservation can be guaranteed. Following a similar procedure using simple materials based on the Clausius-Duhem inequality and applying Coleman’s method [12,14,48–50] in which a positive dissipation is guaranteed, the stress and its constitutive equation can be obtained: r¼m


n n X X ∂Cðee ; ak Þ ∂Ck ðee ; ak Þ kk mk
kk
rk ¼ ¼ ∂e ∂e k¼1 k¼1

ð16Þ

The secant constitutive equation for composite material (Eq. (16)) results: r ¼ ℂSk
ee ¼

n X

k¼1

kk
rk ¼

n X

k¼1

kk
ℂSk :eec ¼

n X

k¼1

kk
ℂSk :ec

ð17Þ

The tangent constitutive tensor is obtained by the stress variation with respect to the strains and is given by: ℂT ¼

n X ∂r ∂2 Ck ðee ; ak Þ ¼ kk
ℂTk ¼m ∂e ∂e2 k¼1

ð18Þ

The classical mixing theory, based on the hypothesis that the strain tensor is exactly the same for all the composite components, is strictly valid only if it is applied to material components working in parallel. These materials are characterized by the fact that their stress state is the results of the sum of the stresses of each component, which are weighted proportionally to their volume in each phase with respect to the total. In order to solve this problem there are two alternatives: to define another closure equation (Eq. (12)), suitable for the material phenomena simulation, or to carry out a modification of each one of the component’s properties and keep the strain equality hypothesis in each one of the composite components. Therefore, the main problem of classical mixing theory is the poor ability to represent the serial behavior of the components in the composite (Fig. 6, iso-stress case).

2.9.5.2

Serial/Parallel Mixing Theory for One Layer

Due to the classic mixing theory limitation, various modifications have been proposed (see Ref. [12]). These allow representing the composite component’s behavior participating in a combination of serial-parallel behaviors. This involves an automatic adjustment of the composite properties taking into consideration each component’s topology and distribution. Thus, each point of a solid can have a different strain. Here, a short presentation of the S/P mixing theory is made [4,8,14]. This rule of mixtures improves the classical mixing theory by modifying its closure equation, replacing the iso-strain hypothesis for an iso-strain condition in the fiber direction and an isostress condition in the transversal one (Eq. (21)). The modeling of all the components distribution in the composite are shown in Fig. 6. This formulation is an alternative to the homogenization technique, based on the multiple scale study [17,18,22,23,25]. The SP formulation [4,8,14] considers that the component materials of the composite act in parallel along a certain direction and in serial in the remaining directions. The main hypothesis in which the numerical model of the S/P mixing theory are based

Composite Materials

m

s

=

2i

s e 2 f



1i 

P e e

s e 1

ei

Pure serial behaviour

243

Pure parallel behaviour m P

= fP

s

m s

Serial−parallel coupling

m p

s = mk ms + fk fs

= fs

p = mk mp + fk fp

= fp

Transversal loading

Longitudinal loading

Fig. 6 Serial/Parallel mixing theory-composition scheme.

on: (1) (2) (3) (4) (5) (6)

The composite is composed by two component materials: fiber and matrix. The component materials have the same strain in parallel (fiber) direction. The component materials have the same stress in the serial direction. The composite material response is in direct relation with the volume fractions of the compounding materials. The homogeneous distribution of phases is considered in the composite. Perfect bounding between components is assumed.

Consequently, it is necessary to define and separate the serial and parallel components of the strain and stress tensors. Defining e1 as the director vector that determines the parallel behavior in the fiber direction, the parallel projector tensor NP can be defined as NP ¼ e1 #e1

ð19Þ

Using NP, the 4th-order parallel projector tensor PP, and the complementary serial projector tensor PS, are defined as PP ¼ NP #NP ;

PS ¼ I

NP

ð20Þ

Both tensors are used to find the parallel and serial part of the strain tensor eP and eS respectively, eP ¼ PP :e;

eS ¼ PS :e

ð21Þ

Hence, the strain and stress tensors are split into its parallel and serial parts ð22Þ e ¼ eP þ eS ; r ¼ rP þ rS where rP ¼ PP:r and rS ¼ PS:r The equations that define the stress equilibrium and establish the strain compatibility between components are obtained from the analysis of the model hypothesis. Thus, c

Parallel behavior:

c

eP ¼ m eP ¼ f eP þ eS

rP ¼ m k m rP þ k k f rP

eS ¼ k k m eS þ f k f eS c rS ¼ m rS ¼ f rS

ð23Þ

c

Serial behavior:

ð24Þ

where, eP and eS are the parallel and serial components of the stress tensor respectively, rP and rS are the parallel and serial components of the strain superscripts, c, m, and f denotes the composite, matrix, and fiber materials, and mk and fk are the volumetric participation of fiber and matrix in the composite, respectively. The S/P mixing theory can use any constitutive equation to describe the behavior of each component material. The constitutive equations chosen can be different for each component (e.g., an elastic law to describe the fiber behavior and a damage formulation to describe the matrix behavior). The constitutive equations for the matrix and the fiber can be expressed in the following form: # "k # "k # "k eP rP ℂSPP k ℂSPS k r ¼ k ℂS :k e where k ð25Þ : k ¼ k S k S eS rS ℂSS ℂSP where kr is the stress tensor of the kth component material, ke is the total strain tensors, k ℂ is the respective damaged secant constitutive tensor and its elements are: k ℂPP ¼ PP :k ℂ:PP ; k ℂPS ¼ PP :k ℂ:PS ; k ℂSP ¼ PS :k ℂ:PP ; k ℂSS ¼ PS :k ℂ:PS .

Composite Materials

me i+1 s

Fibre serial strains:

fe i+1 s

je k

k

=

= 1 fk

je

p

ce i+1 s

−

j

k

+

es

mk fk

me i+1 s

k

(ji+1)* = (jC):(je i+1 − (je p ))

Closing equation

Elastic trial

Constitutive models for each compound (mi −1)*

(fi+1)*

Fibre

Constitutive model

Matrix

No convergence = modify matrix serial strain prediction (k = k+1)

Prediction of matrix serial stains:

(mi +1)

Constitutive model

244

(fi +1)

Check for convergence [Δs ]k = [ms ]k − [fs ]k ≤ toler

Recomposition n

(i +1)c = ∑ kj (i+1)j j −1

Fig. 7 Flow diagram of the serial/parallel mixing theory (S/P-rule of mixtures).

The schematic S/P (or Generalized) Mixing Theory flow diagram of a numerical implementation is shown in Fig. 7.

2.9.5.3

Serial/Parallel Mixing Theory for a Stacking Layers Composites

Laminate composites are formed by different layers with different fiber orientations. The orientation of the fiber can be defined by the engineer or automatically by an optimization process in order to obtain the better performance of the composite according to its application. The S/P-rule of mixtures formulation can be applied to each layer of the composite and, afterwards, the composite behavior is computed by combining the performance of each constituent layer. The classical mixing theory is applied to each layer to obtain the composite laminate behavior. Applying the classical mixing theory onto the different layers of the laminate composite implies the assumption that all laminate are undergoing the same strain. This is a simplified approach, as the performance of a laminate in the direction perpendicular to the layers is in serial. However, as it is stated in Ref. [8], the loss of accuracy is minimal compared to the improvement in the computational effort. This is because the different layers of the laminate usually have fiber orientation distributions disposed in such a way that provide the laminate with an in-plane homogeneous stiffness, and loads are rarely applied perpendicular to the laminate.

2.9.5.4

Additional Formulations Used by S/P Mixing Theory to Simulate Reinforcement Composite Materials

Having defined the main frame of the formulation to deals with composite materials, there are other special formulations considered to obtain a better performance of the numerical simulation and approximation of the mechanical behavior of composite material structures. In this section, a brief description of them is considered.

Composite Materials

Real anisotropic stress space 

ij =

Aijkl

Fictitious isotropic stress space 

kl

SII

SII F

(ij, qsm )

=0

F



(ij, qsm)

ij = Cijkl kle

Real anisotropic strain space e

245

ije e e ije = Aijkl kl

=0 ij = Cijkl kle

ije Fictitious isotropic strain space e

Fig. 8 Space transformation. Real and fictitious stress and strain spaces in small strain. Image obtained from Shafei MAR, Ibrahim, DK, Ali AM, Younes MAA, El-Zahab EA. Novel approach for hydrokinetic turbine applications. Energy Sustain Dev 2015;27:120–6.

2.9.5.4.1

Anisotropy using a mapping space theory

The mapping space theory permits take into account the anisotropy of each single component material [11,14,26]. It is based on the transportation of all the constitutive parameters and the stress and strain states of the structure, from a real anisotropic space, to a fictitious isotropic space. Once all variables are in the fictitious isotropic space, an isotropic constitutive model to obtain the new structure behavior can be used. This theory allows considering materials with high anisotropy, such as single or composite materials, using all the techniques and procedures already developed for isotropic materials. All the anisotropy information is contained in two-fourth order tensors. One of them, As, relates the stresses in the fictitious isotropic space (r) with the stresses in the real anisotropic space (r) and the other one, Ae, does the same with the strains. The relation of both spaces for the strains and the stresses is exposed in Eq. (26). r ¼ As :r;

e ¼ Ae :e

ð26Þ

Once the stresses and strains are transferred to the respective isotropic spaces, the proposed constitutive equation is integrated, and its results are back to the anisotropic (real) spaces by a simple transportation operator (As;Ae) (Eq. (26)). In Fig. 8 a representation of these transformations is displayed. A more detailed description of this methodology, the extension to large strains and its numerical implementation can be followed in Refs. [7,11,14].

2.9.5.4.2

Fiber–matrix debonding

Matrix-reinforced composite materials have a complex nonlinear behavior due to the reinforcement displacement because of the loss of adherence between the matrix and the reinforcement. This relative movement between the reinforcement and the matrix causes a loss of stiffness in the whole set and a decrease of the composite mechanical parameters without fractures in the reinforcement phase is observed. The formulation introduced in previous section is based on the mechanics of the continuum medium to deal with the anisotropy and the mixing theory. It involves introducing an irrecoverable inelastic behavior in the constitutive equation to represent an approximation of the relative rigid movement of the body produced between the fiber and the matrix. The incorporation of the FDM into the constitutive equation must take into consideration two main characteristics: (1) the global loss of stiffness due to the decrease of the fiber collaboration in the matrix and (2) the irrecoverable relative displacement between the fiber and the matrix. The fiber-reinforced composite materials subjected to tension do not satisfy the kinematic condition imposed by the basic theory of basic substances. A direct consequence of this phenomenon is the matrix limitation to transfer the stresses to the fiber. In other words, the fiber cannot increase its tensional state as a result of the limited adherence in the fiber–matrix interface zone. The fiber–matrix debonding constitutive model is based on the assumption that the loading transfer from the matrix to the fiber varies when the matrix is under plastic strains [7,11,14,26]. The relative movement between the fiber and the matrix can be represented by the mechanics of the continuum medium through an irrecoverable inelastic strain in the fiber. The starting point of this phenomenon is determined through a threshold condition of maximum strength, which compares the effective stress on a point with respect to the fiber strength. That is, the fiber participation in the composite depends on its own strength and on the stress transfer capacity of the fiber–matrix interface. Therefore, its strength is influenced by the medium containing it and its

246

Composite Materials

constitutive treatment might involve a nonlocal formulation. Then, the fiber strength contained in a matrix is defined as: n
 h
 io
R
 ð27Þ f fib ¼ min f N fib ; f N mat ; 2
f N fib mat =rfib

in which rfib represents the radius of the long fiber of the transversal section, (fR)fib is the new fiber strength, (fN)fib is the nominal fiber strength, (fN)mat is the matrix nominal strength, and (fN)fib mat is the fiber–matrix interface nominal strength. Eq. (27) shows that the debonding happens when one of the composite constituents reaches its nominal strength (considering the fiber–matrix interface as a constituent). The numerical implementation of this phenomenon is described in the Refs. [7,14,26].

2.9.5.4.3

Tangent constitutive stiffness tensor

Depending on the constitutive equation used in a composite constituent material, the tangent constitutive tensor cannot be analytically obtained. One solution is to use, in these materials, the initial stiffness matrix, which will lead to the equilibrium state but will require a large amount of structural iterations. Thus, in order to obtain a fast and reliable algorithm, the expression of the tangent constitutive tensor is required. To obtain it, when no analytical expression exists, a numerical derivation using a perturbation method is performed [9,14]. The definition of the tangent constitutive tensor is: r_ ¼ ℂt :_e

where,

 r_ ¼ s_ 1 s_ 2 … s_ n ; 3 2 t c11 … ct1n 7 6 ℂt ¼ 4 ⋮ ⋱ ⋮ 5 ctn1 … ctnn

 e_ ¼ e_ 1

ð28Þ e_ 2

…

e_ n



and ð29Þ

The definition of the tangent constitutive tensor, Eq. (28), shows that the variation of stresses due to an increment in the value of the j element of e_ depends on the values of the j column of ℂt . Thus, writing the j column of ℂt as, h i t t t T ð30Þ ctj ¼ c1j c2j … cnj

the stress variation is

r_ j ¼ ctj
e_ j

ð31Þ

being ctj the unknown. The perturbation method consists of applying a small perturbation on to the strain vector and, using the constitutive equation of the material, determines the variation that will be obtained in the stress tensor due to this perturbation. At this point, the j column of the tangent constitutive tensor can be computed as: ctj ¼

s_ j e_ j

ð32Þ

For the smaller applied value to the perturbation, the better approximation is obtained for the tangent constitutive tensor. Having defined a perturbation value e_ j , the perturbed stress is computed using the constitutive equation of the material applying the following input strain: h iT ^ - r ð33Þ e ¼ e1 … ej þ e_ j … en And the stress variation due to the perturbation is obtained subtracting the original converged stress from the computed one: ^ r r_ j ¼ r

ð34Þ

This procedure must be repeated for all strain components in order to obtain the complete expression of the tangent constitutive tensor. Hence, the numerical cost of using a perturbation method is rather high. However, this procedure allows obtaining an accurate approximation to this tensor for any constitutive equation used, ensuring the Newton–Raphson convergence of the numerical process in few steps.

2.9.5.5

Local Plastic Damage Model for a Component Material

The theory of plasticity provides a suitable physical-mathematical framework to formulate the behavior of metallic materials subjected to loading. From the extension of its main basic principles and the reinterpretation of its main variables, the “Plastic Damage Model” has emerged as one of the more general plastic constitutive model [46,51,52]. This plastic model is based on the “plastic damage variable” dp formulated as an internal variable representing a unit normalized dissipated plastic energy ranging from 0rdpr1. For dp ¼ 0 there is no plastic damage and for dp ¼1 the limit of total plastic damage in a solid point is defined. The

Composite Materials





pic

C

pic C

C( p )

oC

C(d p )

oC g pc

oT

Compression curve

Compression curve oT

T( p) ggPpT

247

T(dp) Tension curve

Tension curve P

(Tp )u

(Tp )u

p

I

(A)

(B) C pic cC C v = f()|0 cC(dp) cT(dP )

Compression curve

Tension curve

(C)

I

p

Fig. 9 Transformation of the uniaxial strength measured in the lab into the uniaxial strength used in the plastic damage model. (A) Uniaxial strength according to the plastic strain; (B) uniaxial strength according to the plastic damage variable; and (C) cohesion according to the plastic damage variable.

latter state can be interpreted as a total loss of strength in a point of the solid produced for the accumulated plastic effect. A brief presentation of this model will be given in this section. For further details, check Refs. [46,51,52]. In short, for a plastic mechanical process with no stiffness degradation, the plastic damage model uses the following set of plastic internal variables qp ¼ {ep, ap} ¼{ep, dp, c}, and its evolution laws will be presented as part of the main equations governing the model.

•

A “deformation splits” into an elastic and a plastic part, e ¼ ee þ eP ¼ ℂ 1 :r þ eP

•

ð35Þ

where ℂ is the constitutive elastic tensor. A “plastic yield and potential plastic criteria” are defined by the following two equations, F P ðr; qp Þ ¼ f P ðrÞ GP ðr; qp Þ ¼ g P ðrÞ

cðdp Þr0

ð36Þ

cte ¼ 0

ð37Þ

being FP(r, qp) the plastic damage threshold function, GP(r, qp) the plastic potential, c(dp) the cohesion or uniaxial strength evolution, depending on internal plastic damage variable dp, fP(r) and gP(r) are two scalar functions of tensorial arguments called yield function and plastic potential respectively, and can be represented by any classical limit criteria (von Mises, Mohr Coulomb, Drucker-Prager, etc.) [46,49,53]. The uniaxial strength evolution c(dP) is assumed as a scaled magnitude regarding an ultimate strength to uniaxial compression of the composite sc (stress discontinuity threshold), that is the stress level for which the volumetric deformation eV reaches its maximum value. Therefore, the initial uniaxial strength is defined as c0 psC for dp ¼ 0, setting the initial position of the yield criterion; and the final uniaxial strength of the material totally deteriorated as cu ¼ 0 for dp ¼ 1, defining the final position of the yield criterion (see Fig. 9).

248

Composite Materials

Unlike the classic plasticity formulation with isotropic hardening, the cohesion or uniaxial strength in this case is not a simple function of the plastic hardening variablec(dp), but is an internal variable that depends on the evolution of the elastoplastic process governed by its evolution equation.

•

A “plastic strain, plastic damage, and cohesion internal variables,” 8 9 ( P ) > eP > Plastic strain < = e Plastic damage variable q¼ ¼ dP ap > : > ; Cohesion variable c

ð38Þ

all of them defined by the following evolution equations,

8 9 ∂GP > > > > > > 8 9 8 P9 > > > > ∂r > > e_ > e_ P > > > > > < = < = < p P = p dq ∂G hp :e_ P ¼ q_ p ¼ d_  l_
H ¼ l_
 hp : > ; > > dt ∂r > : h
h :_eP > ; : > > > > > > c p P> c_ > > ∂G > > > > : hc
hp : ; ∂r

•

ð39Þ

where hp and hc are a second-degree tensor and a scalar function respectively that will be defined later, and which depend on the current state of the free variable ee and the rest of the internal variables qp. As observed in this equation, the main internal variable is the plastic deformation ep, and the others are obtained from it. The plastic consistency factor l is obtained from the consistency condition of the plastic yield function [46,51,52]. A “secant and tangent constitutive equation,” defined as the classic plasticity,

r_ ¼

8 > > > < > > > :

ℂ

r ¼ ℂ:ðe ep Þ 9  P  > ∂G ∂F > > :ℂ ℂ: # = ∂r ∂r   :_e P P P P P > ∂F ∂G ∂G ∂F ∂G > > : þ hc hp : þ :ℂ: ck
; ∂g ∂r ∂r ∂r ∂r

P

)

r_ ¼ ℂT :_e

ð40Þ

where Z is the kinematic plastic flow orientation [46].

The constitutive model resulting from these basic definitions shows a very good response during the general behavior composite process. To sum up, the model has the following characteristics:

• • • • •

It defines a constitutive law depending on the internal variables of cohesion and plastic damage to represent non-radial complex loading situations. It deals with complex states of multiaxial stress in a unified way. It admits that materials have different limits of maximum strength and ultimate deformation, depending on the mechanical process in progress. It admits different plastic yield and potential criterion. Although this is not a characteristic of the model, it can be preestablished as one of its variables. It can obtain all the information related to the point deterioration through the point mechanical information post processing.

2.9.5.5.1

Definition of the plastic damage variable

The classical plasticity theory establishes a hardening variable as a function of the effective plastic strain p , or also as a function of the specific plastic work op ¼ s ep ¼ r:ep [46,49,50]. These definitions are suitable for materials whose final deformation is equal in tension as in compression as metal behavior. However, this assumption is not true for many materials as composites one. Therefore, it is necessary to establish an internal variable defining a unit-normalized dissipation, which is the relation between the density of the dissipated energy at a specific time of the process and the maximum dissipation of the point of the solid. Hence, it is said that “the plastic damage variable” is a unit-normalized measure of the dissipated energy during the plastic process. Generally, for a generic loading process, the plastic damage variable is defined for a multiaxial mechanical process as, p d_ ¼ hp :_ep

ð41Þ

where hp is a second-degree tensor that, for uniaxial tension and compression processes, leads to a plastic damage depending on the loading process. To recover the plastic hardening variable of the classic plasticity theory, this tensor becomes equal to the stress tensor hp ¼ r and in the general case it can be defined as a normalized dissipation to the unit for isotropic materials. " # _dp ¼ hp :_ep ¼ rðrÞ þ 1 rðrÞ
Ξ m ð42Þ p p gf gC p

p

p

p

where Ξ m ¼ r:_ep is the plastic dissipation energy, gf ¼ Gf =lf and gc ¼ Gc =lf are the fracture and crushing energy per unit of volume p p respectively, Gf and Gc are the mechanical fracture and crushing energy parameter, lf is a regularization parameter called charP P acteristic length related with the finite element size, rðrÞ ¼ 3I ¼ 1 〈sI 〉= 3I ¼ 1 jsI j is a scalar function defining the tension-

Composite Materials

249

compression behavior states in each point as a function of the stress state, and 〈x〉 ¼ 0.5 [x þ |x|] is the Macaulay bracket. Note the following particular cases, r(r) ¼1 for pure tension problems, r(r) ¼0 for pure compression and r(r) ¼ 0.5 for a pure shear state. Consequently, the plastic dissipation will always be normalized with respect to the maximum energy of the process at every moment. Thus, the plastic damage variable is objective and evolves within the same limits regardless of the mechanical process. Thus, the p p total plastic damage in a point is reached when dp ¼ 1, but the dissipated energy will be gf in a pure tension process, and gC in a pure compression process.

2.9.5.5.2

Definition of the cohesion or uniaxial strength evolution law c

dp

This plastic damage model assumes that micro-plastic damage in most materials is due to the loss of strength. Due to this failure mechanism, a softening in the strain–stress behavior can only be observed as a macroscopic effect (phenomenological model) caused by the average behavior of a set of points. The plastic damage constitutive model carries out the numerical analysis on a finite domain (integration point of the constitutive equation) through the finite element functional approximation technique. Therefore, every point under analysis represents infinite material points contained in its area of influence. Thus, at macroscopic level, the softening phenomenon by deformation can be considered as a material property and, in such a case, a plastic hardening function must be defined taking into account this phenomenon. This hardening function is represented by the cohesion, written as an internal variable to make it more general and its evolution equation for any quasi-static loading process is defined as p

c_ ¼ hc
d_ ¼ hc
hp :_ep

p

p

ð43Þ

where hc(r, d , c) is a scalar function of the current state of the stress-free variable r and of the internal variables d and c. The expression used for the evolution law of the internal variable of cohesion, or uniaxial strength, is obtained from the following expression for hc,

 rðrÞ dcT 1 rðrÞ dcC hc ¼ c
þ ð44Þ d dp cT d dp cC where r(r) is the aforementioned function which sets the type of behavior (tension or compression or tension–compression), developed during the mechanical process in each point of a solid. The cohesion function cC(dp) [32,46] is obtained in explicit form and it represents the cohesion evolution during a uniaxial simple compression test. The relation between cohesion and uniaxial stress of compression is given by the following expression: cC ðκ p Þ ¼

1 sC ðdp Þ ℵ

ð45Þ

such that ℵ is a coefficient depending on the criterion of the discontinuity threshold and represents a scalar factor between [46,51,52]. For example, for Tresca and von Mises its value is ℵ ¼ 1, for Mohrcohesion and the uniaxial stress of compression  pffiffiffiffiffiffi –Coulomb ℵ ¼ 2 R0 , where R0 ¼ fC0 =fT0 ¼ ½sC ðdp ¼ 0Þ=sT ðdp ¼ 0ފis the relation between compression–tension uniaxial strengths, for Drucker–Prager inscribed in the surface of Mohr–Coulomb ℵ ¼ 6 cosðjÞ=ðsinðjÞ 3Þ and for Drucker–Prager circumscribed in the surface of Mohr–Coulomb ℵ ¼ 6 cosðjÞ=ð3 sinðjÞ 3Þ. Thus, for any discontinuity threshold criterion, this coefficient must be defined. The function cT(dp) (see Fig. 9) can be obtained explicitly and represents the cohesion evolution during a uniaxial simple tension test. The relation between cohesion and uniaxial tension stress is given by the following expression: R0 sT ðdp Þ ℵ

ð46Þ

sC ðdp Þ ¼ cte: ) Rðdp ¼ 0Þ ¼ R0 sT ðdp Þ

ð47Þ

cT ðdp Þ ¼

pffiffiffiffiffiffi For Tresca and von Mises its value is R0/ℵ¼ 1, for Mohr–Coulomb R0 =ℵ ¼ R0 =2 , for Drucker–Prager inscribed in the Mohr–Coulomb surface R0 =ℵ ¼ ð3 þ 3 sinðjÞÞ=6 cosðjÞ and for Drucker–Prager circumscribed in the Mohr–Coulomb surface R=ℵ ¼ ð3 þ sinðjÞÞ=6 cosðjÞ . Some materials strength curves in simple tension and compression, obtained in uniaxial experimental tests, have similar shapes; in other words, it can be stated that the scale relationship between them is a constant during the whole quasi-static process and is given by Rðdp Þ ¼

In such case the explicit functions of uniaxial tension and compression cohesion coincide.

2.9.5.6

Local Damage (Elastic Degradation) Model for a Component Material

The local damage constitutive model [46,51,54] used to set the threshold for the initiation of nonlinear elastic modulus degradation process in each point of component material of the composite one and its subsequent evolution is presented in this subsection. This concept allows the new definition of a global homogenized threshold criterion of damage for the entire laminate, resulting from the composition of the local damage index over all involved material in the laminate composite.

250

Composite Materials

Material degradation- or damage- in a simple continuum material component due to a dissipative process can be simulated by means a local damage formulation [24,35,46,51,52]. This model is used at each simple matrix material embedded in the composite, inducing a stiffness degradation and strength reduction in the entire laminate. The isotropic damage formulation is based on a scalar internal variable dd that represents the level of elastic degradation at each simple component material. This variable is bounded between 0 and 1, being zero for an undamaged and one for a completely damaged state of a single component material. The local damage variable dd is used to link the real stress tensor r with the effective undamaged stress tensor r0. Therefore, the relation between the damaged stress and the strain in the matrix component included in each layer depends on the internal damage variable dd and the elastic constitutive tensor ℂ0 ,  d r ¼ ð1 dd Þ r0 ¼ ð1 dd Þ ℂ0 :e ¼ ½ℂ0 :eŠ d ℂ0 :e ¼ r0 rd ð48Þ |fflfflfflfflfflfflfflffl{zfflfflfflfflfflfflfflffl} ℂS0

In the same form of the plastic formulation, the stress condition at which damage starts and the evolution of the damage variable can be described by the following threshold function:

   F d r0 ; qd ¼ f d ðr0 Þ c dd r0; with qd  dd ð49Þ

being Fd(r0;qd) the damage threshold function, fd(r0) the scalar equivalent stress function and c(dd) the uniaxial strength evolution depending of the internal damage variable dd. In the same form of the plastic damage model, this formulation allows the damage onset and evolution using any isotropic limit criteria already defined in literature (von Mises, Mohr–Coulomb, Drucker–Prager, etc.) [23,53], and the anisotropic behavior is included by means the mapping spaces theory previously defined in Section 2.9.5.4. The norm of the principal stresses with a different degradation path for tension and compression loads is also here defined for the widest range of structures in composite materials. That is,

 sc f d ðr0 Þ ¼ rðrÞ þ ð1 rðrÞÞ
‖rI ‖ 8 I ¼ 1; 2; 3 ð50Þ st being sc and st the uniaxial ultimate strength of the material in compression and tension, respectively, rI the principal stress tensor, and r(r) is the same a scalar function defined in the plastic damage model, that take into account the tension–compression behavior states in each point as a function of the stress state, X3 X3 rðrÞ ¼ 〈sI 〉= js j ð51Þ I¼1 I¼1 I and 〈〉 is the Macaulay bracket already defined. The mechanical evolution of the damage inner variable dd or damage index for a simple component material is obtained by means the damage consistency equation [46] and the Kuhn–Tucker load/unload conditions [23,54]. The uniaxial strength function is defined as,   
 and dd ¼ Gd f d ðr0 Þ ð52Þ cðdd Þ ¼ max tc ; max f d ðr0 Þ

The function Gd defines the softening evolution of the material. The behavior evolution of the damage material in the present work uses an explicit exponential softening, which is defined as,  d  f ðr0 Þ  A 1 sc sc d d d e d ¼ G f ðr0 Þ ¼ 1 ð53Þ ; with sc ¼ cmax d f ðr0 Þ

where A is a parameter that depends of the fracture energy of each simple material. This parameter can be obtained for an exponential softening material as  d  1 Gc
C0 1 A¼ ð54Þ 2 lf
sc 2

being C0 the uniaxial Young modulus of the material, Gdc the damage compression energy of the material and lf the geometrically regularization parameter, called “characteristic fracture length” related with the characteristically size of the finite element used. The introduction of this fracture length in the formulation makes the degradation process mesh independent [23,54].

2.9.5.7

Global Composite Homogenized Laminate Damage Index

In this section, the global damage index [13] to ensure that the composite laminate is found within the elastic range is defined. According to the S/P mixing theory previously introduced, fibers only collaborate to the composite strengthening in longitudinal direction of the fibers. Thus, the damage on the composite material is mainly concentrated in the matrix but not in the reinforcement fibers. Thereby, when stress in matrix reach its maximum elastic value (damage threshold), the material component falls according to the “damage constitutive law” or “plastic damage constitutive low” previously presented. At this time, the energy fracture dissipation begins in each component, and the material point cannot support any more the stresses level, the stiffness contribution disappears, and starts a crack evolution producing a delamination phenomenon. In addition, the lack of strength in the matrix in all directions is produced, except in the longitudinal fibers (because fibers do not reach the damaged threshold).

Composite Materials

251

Hence, this mechanism induces localized fracture (delamination) at constitutive level without the computational cost of breaking the mesh and re-meshing the new delaminated area. In the case of laminates, the global composite damage index dL is obtained by the homogenization of de local damage variable dp or dd of each simple component material (see previous sections). This definition of homogenized laminar damage can also be understood as a safety laminate output warning. This new structural damage index dL is also bounded between 0 and 1 as local damage variable and it is defined as, L

d ¼

nGP X

i¼1

Vi

!

1




nGP X

p;d

V i di

i¼1

ð55Þ

p;d

where di are the local damage for a plastic or degradation process, and Vi are the local damage variable and its Gauss point volume for each single material, nGP is the number of Gauss points involved in all materials included in the layers participating in the finite element. The safety laminate factor can be used, for instance, in optimization process to obtain the composite material design that accomplishes with the structural and functional design parameters [13]. The delamination phenomenon stops when a damaged point can provide enough shear strength to equilibrate the shear stresses that appears in the interlaminar zone. A potentiality application of this structural composite formulation, including a WCT rotor is introduced in next section.

2.9.6

“Micromodel” Versus “Mixing Theory”: Conceptual Comparative Behavior Example

First, a simple and conceptual example intent shows the capabilities of the formulation here presented. Here a little sample of a fiber-reinforced matrix is introduced, and shows the capacities of the “mixing theory with anisotropy in large strains” comparing the results obtained with a micromodel where each of the component materials is individualized. The example consists in subjecting a unit size cube of a composite material under tension in which the reinforcement fiber and matrix component are discretized. Then the results obtained by this micromodel are compared to the results obtained by the macro module proposed work. In Fig. 10 the unit size piece is shown, where both phases of the composite material have been discretized. The boundary conditions imposed can also be observed. The finite element mesh is set up by 5701 triangular finite elements of 3 nodes and 2940 nodes. As an alternative to the mesh described before for the numerical simulation through the model previously described, the same unit size piece modeled by only one single finite element of 4 nodes and 2  2 points of integration is analyzed. The sliding phenomenon between the fiber and matrix – Fiber–matrix debonding (FDM) – will not be considered in this example. The mechanical properties of the materials making up the composite are shown in Tables 1 and 2. The micromodel is made-up of three materials: reinforced fiber, epoxy matrix, and fiber–matrix interface. Epoxy matrix and interface material are considered isotropic and homogenous and have mechanical properties that coincide with the mechanical properties of the macromodel components. In Fig. 11 the micromodel materials distribution are shown schematically.

Fig. 10 Micromodel Finite element mesh made-up by 5701 linear triangular finite elements of 3 nodes, and 2940 nodes.

252

Composite Materials

Table 1

Epoxy resin properties for the macromodel and micromodel

Isotropic Young modulus Poisson coefficient Yield or threshold stress Damage-plastic post-yield behavior law Fracture energy Volume participation Vm

1300 MPa 0.325 43,323 MPa Exponential with softening 10 N/m 76%

Table 2 Fiber carbon properties for the macromodel and micromodel Axial fiber Young modulus Transversal fiber Young modulusa Poisson coefficient Yield or threshold stress Damage post-yield behavior law Volume participation Vf

239,551 MPa 1300 MPa 00 3000 MPa Linear with hardening 24%

a

Adopted equal to matrix elastic modulus.

Mat. 2 Mat. 1 Mat. 3 Fig. 11 Micromodel materials distribution: Mat 1: Matrix; Mat 2: Fiber; and Mat 3: Matrix–Fiber interface.

The numerical test consists in imposing displacements on the upper part of the unit size structure producing tension. This stress state on the specimen leads to the fiber reinforcement alignment with the load direction. In Fig. 12 the deformed shape in its final state is shown. It can also be observed that the fibers have aligned themselves with the direction of the applied stress. This alignment of the reinforcement phase with the stress direction makes it necessary to introduce the theory of large strains in the constitutive model. The advantage of using a micromodel is that a detailed analysis of the mechanical processes can be done during the load application. Fig. 13 shows the shear stress on the material for different loading cases. Fig. 13(A) shows the stress states in a loading phase in which the stresses above the elastic limits of the component materials are not verified (see the plasticity internal variable in Fig. 13). It can also be observed in the same figure that the matrix zone among the fibers is the one presenting a higher tensional state. As the displacements increase (Fig. 13(B)–(D)) a homogenization of the matrix stress state is observed. Fig. 14 shows the stresses in the micromodel in the direction of the imposed displacements. Fig. 14(A) corresponds to a stress state in a loading step in which the composite materials stresses above the elastic limit are not verified (see Fig. 15). Fig. 14(B)–(D) shows the stress state in the direction of the imposed displacement as displacements increase. It can also be observed that in the first loading steps the matrix has a homogenous stress state in the direction of the applied stresses. Fig. 14(B) shows that the reinforcement increases considerably as it aligns itself with the direction of the stress applied.

Composite Materials

253

0.29814 0.26237 0.22659 0.19677 0.191 0.13118 0.095406 0.065591 0.029814 0 Fig. 12 Micromodel displacements contours and deformed shape.

263 208 154 99.5 45 −9.44 −63.9 −118 −173 −227 (A)

.501E4 .36E4 .22E4 801 −601 −.2E4 −.34E4 −.481E4 −.621E4 −.761E4 (B)

.515E4 .463E4 .411E4 .359E4 .307E4 .255E4 .203E4 .151E4 996 477

.484E4 .408E4 .331E4 .255E4 .178E4 .102E4 251 −515 −.128E4 −.205E4 (C) Fig. 13 Stress contours sxy for different loading stages.

(D)

254

Composite Materials

.273E4 .246E4 .218E4 .19E4 .163E4 .135E4 .107E4 797 520 243 (A)

.274E5 .246E5 .218E5 .191E5 .163E5 .135E5 .107E5 .792E4 .513E4 .235E4 (B)

.28E5 .252E5 .224E5 .197E5 .169E5 .141E5 .113E5 .851E4 .573E4 .294E4 (C)

.274E5 .247E5 .22E5 .193E5 .166E5 .139E5 .112E5 .846E4 .575E4 .304E4 (D)

Fig. 14 Stress contours syy for different loading stages.

Fig. 14 shows the plasticity contours in each composite component. It also shows that as the displacement increases, the irreversible strains in the matrix are verified in the areas between reinforcements (see Fig. 14(B) and (C)). In Fig. 14(D) it can be observed that the elastic limit has been exceeded, consequently leading to irreversible strains. Fig. 16 shows the micro and macromodels loading-displacement response. Different values of the transversal module of the reinforcement phase are considered. The same figure shows that the value of the transversal elastic module of this phase plays a fundamental role in the macromodel response. When the shear modulus is zero, it is observed that the matrix reaches its limit of proportionality while the stress in the composite decreases until the fibers coincide with the direction of applied stress. Beyond this point, the reinforcement phase provides stiffness to the system. The response corresponding to the small strains assumption can also be observed. In this case, once the matrix’s elastic limit is achieved, the material response decreases and the fibers do not participate in the response. This is because according to the small strain hypothesis the geometry is not updated and consequently the fibers cannot align themselves with the applied stress direction.

2.9.7

“Micromodel” Versus “Mixing Theory”: Conceptual “Fiber–Matrix Displacement” (Debonding) Behavior Example

This example show a simple validation and comparison between the fibers sliding effect into matrix by means a “an isotropic mixing constitutive model formulation” and an “explicit finite element micromodel.” An example of the formulation application combining the mixing theory, the anisotropic model in large strains and the theory that includes the fiber-matrix displacement (FMD) phenomenon analysis is described below. This example compares the numerical simulation of the composite material specimen (reinforced concrete) with a central notch subjected to traction where the reinforced and matrix phases have been discretized (micromodel), with a similar specimen in which only a composite material made-up by a reinforced phase and the matrix (macromodel) exists. The numerical simulations have been carried out using a linear rectangular finite element mesh of 4 nodes with a total of 343 elements, 392 nodes, and 766 degrees of freedom for the micromodel and 291 elements, 336 nodes, and 644 degrees of freedom for the macromodel. Fig. 17 shows the geometry, material assignment, meshes, and boundary conditions used for each case.

Composite Materials

255

.122 .115 .105 .975E−1 .902E−1 .829E−1 .731E−1 .658E−1 .585E−1 .512E−1 .414E−1 .341E−1 .268E−1 .195E−1 .975E−2

0 0 0 0 0 0 0 0 0 0 (A)

(B) .183 .173 .16 .15 .14 .131 .117 .107 .976E−1 .877E−1 .745E−1 .646E−1 .547E−1 .448E−1 .316E−1

.416 .382 .347 .313 .279 .244 .21 .175 .141 .107

(C)

(D)

Fig. 15 Internal plasticity variable contours for different loading stages.

1800.00 1600.00 1400.00

Load

1200.00 1000.00 Micromodelo Micromodelo G=0 Micromodelo G=600 Micromodelo G=6000 Micromodelo G=11,000 Micromodelo G=15,000 Micromodelo pequeñas

800.00 600.00 400.00 200.00 0.00 0.00

0.02

0.04

0.06

0.08

0.10

Displacement Fig. 16 Load-displacement curves comparison in the micro–macromodel.

The micromodel consists of three materials: matrix, fiber–matrix interface zone and reinforcement. The macromodel is madeup of one composite material consisting of three component materials: reinforced fiber, matrix, and fiber–matrix interface. Table 3 shows the mechanical properties of the materials used in the micromodel. The mechanical properties of the composite material phases in the macromodel are identical to the corresponding ones in the matrix and reinforcement of the micromodel.

256

Composite Materials

Materials Mat. 2 Mat. 3 Mat. 1

1.867 cm

2.866 cm 1.867 cm 1 cm

10 cm

10 cm 0.1 cm

1.0 cm 0.1 cm

Concrete matrix

0.1 cm

Steel reinforced

0.1 cm

Fiber−matrix interface

0.66 cm

1.0 cm

0.66 cm 0.1 cm 1.0 cm 0.1 cm 0.66 cm 0.1 cm 1.0 cm

(A)

(B)

(C) Fig. 17 (A) Geometric dimensions and materials: (1) concrete; (2) steel; (3) interface material. Finite element mesh: (B) micromodel and (C) macromodel.

The purpose of this example is to show the loading transfer phenomenon from the matrix to the reinforcement phase. This is achieved first by comparing the “load-displacement” curve obtained by the micromodel and then by the macromodel made-up of composite material, the components of which are not possible to identify physically (mixing theory) Figs. 18 and 19 show the

Composite Materials

Table 3

Mechanical properties used in the micro and macromodel Material 1 matrix of the concrete

Type of material behavior

Mohr–Coulomb isotropic elastoplastic model 3.5  105 Young modulus (kp/cm2) Poisson coefficient 0.2 Internal Friction 30 degree Compression strength (kp/cm2) 200 Tension strength (kp/cm2) 20 Gf , Gc (kp/cm) 0.25, 26.0 Behavior law after the yield point Line function with softening

Material 2 steel reinforcement

Material 3 matrix–reinforcement interface

Isotropic elastic model

Kachanov damage model

2.1  106 0.0 – 2000 2000 – –

3.5  105 0.0 30 degree 20 20 2.0, 2.0 Exponential function with softening

6

Shear stress (kp/cm2)

4

Icremen. 1

2

0

1

2

4

6

8

10

12

14

16

18

20

−2

−4

Icremen. 10 Icremen. 20 Icremen. 30 Icremen. 40 Icremen. 50 Icremen. 60 Icremen. 70 Icremen. 80 Icremen. 90 Icremen. 100

−6 Position (cm) (piece length) Fig. 18 Shear stresses in the fiber–matrix interface. Increments 1–100.

6

Icremen. 110

4 Shear stress (kp/cm2)

Icremen. 120 Icremen. 130

2

Icremen. 140 Icremen. 150 0

0

2

4

6

8

10

12

14

16

18

20

Icremen. 160 Icremen. 170

−2

Icremen. 220

−4

Icremen. 280

Icremen. 250 Icremen. 300 −6 Position (cm) (piece length) Fig. 19 Shear stresses in the fiber–matrix interface. Increments 110–300.

257

258

Composite Materials

140 120 Icremen. 1 Icremen. 10 Icremen. 20 Icremen. 30 Icremen. 40 Icremen. 50 Icremen. 60 Icremen. 70 Icremen. 80 Icremen. 90 Icremen. 100

Axial stress (kp/cm2)

100 80 60 40 20 0 1

2

4

6

8

10

12

14

16

18

20

Position (cm) (piece length) Fig. 20 Longitudinal stresses in the reinforcement. Increments 1–100.

140 120

Axial stress (kp/cm2)

Icremen. 110 100

Icremen. 120 Icremen. 130

80

Icremen. 140 Icremen. 150

60

Icremen. 160 Icremen. 170

40

Icremen. 220 Icremen. 250

20

Icremen. 280 Icremen. 300

0 1

2

4

6

10 12 14 8 Position (cm) (piece length)

16

18

20

Fig. 21 Longitudinal stresses in the reinforcement. Increments 110–300.

shear stress evolution in the fiber–matrix interface zone for different loading increments. The change of sign of the stresses, which are mainly due to the presence of the notch, can be observed in the central zone. Figs. 20 and 21 show the longitudinal stress evolution in the reinforcement phase for different loading increments. It can be observed that for the first loading increment the maximum shear stresses are obtained at the reinforcement end zone, while the longitudinal stresses increase from zero at the end zone to a constant value along the reinforcement. Moreover, a variation of the longitudinal stress in the central zone due to the presence of the notch is observed. Additionally, a decrease of the stress transfer capacity from the matrix to the fibers is observed. This phenomenon also causes a modification of the stress state and, as observed, the stress distribution curve along the reinforcement is no longer constant. Fig. 22 shows the interface zones exceeding the material proportionality limit for different loading stages. It can be noted that the fiber-matrix relative sliding starts at the fiber’s end zone and moves towards the specimen’s center. Fig. 23 shows the displacement contours in the first and final converged loading increments for the micro and macromodels. As observed, in this last increment the displacements are basically in the specimen’s central zone and along the central reinforcement. A displacement is observed between the fiber and the reinforcement at the specimen’s ends.

Composite Materials

259

Increment 2

Increment 10

Increment 20

Increment 30

Increment 40

Increment 50

Fig. 22 Plastic strains in the fiber–matrix interface for different loading increments.

Fig. 24 shows the total forces’ response for micro and macromodels. It can be noted that the micromodel’s results match satisfactorily those of the macromodel. It is important to highlight that the micromodel cannot carry out the simulation of the relative movements between different phases but it can carry out the reinforcement simulation. However, the characterization of the latter would involve a considerable high computational cost of analysis due to the carbon fibers small dimension.

2.9.8

Numerical Simulation of a Structural Analysis of a “Composite Material Rotor-Hydrofoil” of a WCT

The constitutive model described in this chapter is used in the structural numerical simulation of the composite turbine rotor. Reduction of rotational inertia of the WCT rotor is one of the principal aims of the fiber-reinforced composite material application to this kind of structure. This will lower resistance to rotation in front of the river speed changes, allowing more flexibility in starting and stopping of the turbine rotation. The numerical simulation of the multilayered composite structure design of the flow axial turbine rotor by means of finite elements method is presented in this section. A comparative study considering the structural response of the steel turbine rotor versus fibers-reinforced composite material is carried out [13]. The composite material analysis is developed employing the orthotropic mixing theory previously presented, while an isotropic constitutive model is used for the steel rotor.

260

Composite Materials

Micromodel

.91E−4 .819E−4 .728E−4 .637E−4 .546E−4 .455E−4 .364E−4 .273E−4 .182E−4 .91E−5

Macromodel

Micromodel

.278E−2 .25E−2 .223E−2 .195E−2 .167E−2 .139E−2 .111E−2 .835E−3 .557E−3 .278E−3

Macromodel

Fig. 23 Displacement contours of the macro and micromodels in the first and final converged loading step.

140 120

Load (kp)

100 80

Micromodel Macromodel

60 40 20 0 0.000

0.001

0.001

Fig. 24 Force–displacement curve in macro and micromodels.

0.002 0.002 Displacement (cm)

0.003

0.003

0.004

Composite Materials 2.9.8.1

261

Geometry, Boundary Conditions and Finite Element Mesh

The rotor is placed under an axial water flow described in Section 2.9.3 that causes a distribution of pressures on the hydrofoils. These flow pressures are obtained by Computational Fluid Dynamics (CFD) finite element code and, particularly, in the leading edge of the hydrofoils of the rotor. The 8 rotor blades have a hydrodynamic profile with 151 of attack angle (see Figs. 2 and 3 and Table 4). From the geometry of the rotor a mesh of 4100 linear shell triangular finite elements is generated (rotational-free shell triangle [42]), with 2012 nodes (Fig. 25). These shells structure are analyzed firstly made of steel material and then made of laminate composites material with layers of epoxy matrix reinforced by unidirectional carbon fibers. In both cases, the properties of the materials are detailed in Tables 4 and 5.

2.9.8.2

Action on the Hydrofoil’s Rotor

The water pressures obtained by the Section 2.9.3 formulation on the blade of the turbine rotor in the axial axis are applied. These pressures are obtained from CFD code for the fluvial flow at low speed river (see Section 2.9.3) [9,34]. This pressures cause two kinds of loads in the rotor:

• •

Load 1: Rotation loads on the surfaces of the hydrofoils produced by the differential pressures between the up and down surfaces of the wing. This load is obtained by the CFD finite element code to obtain the speed, correct attack angle of hydrofoils, diagrams of pressures on the wing areas, etc. Load 2: Axial loads caused by the directly applied pressures over the attack edge of the hydrofoil, that cause its deformation and the tensional state of the rotor, trending to break it in the perpendicular direction of the plane of the rotor. This reaction forces are studied and analyzed in this example through the previously composite-formulation included in a structural finite element program (Fig. 26).

Kinematic pressures are obtained by CFD finite element procedure which is available on the reference Oller et al. [9,34]. Thus, using this pressure has been obtained an applied load of FRotor ¼ 672 N at the leading edges surfaces of the nwing hydrofoils. This load is distributed over all nodes of the blades, and is applied in one time step, in the rotor at time t ¼ 1  10 5 s. A more complete fluid-dynamic numerical simulation using formulations of Sections 2.9.2 and 2.9.3 can be found in Refs. [9,34]. Table 4 X (m) L (m) b (degree) Z a (degree)

Hydrofoil’s dimensions and geometrical profile 0.175 0.182 63.40 16.23

0.35 0.145 52.23 15.61

0.525 0.128 45.99 15.2

0.7 0.116 38.84 14.89

Detail

Fig. 25 Composite structural Rotor’s finite element model mesh.

0.875 0.108 32.83 14.65

1.05 0.102 27.81 14.46

1.225 0.104 23.59 14.33

1.4 0.093 19.52 14.79

1.575 0.089 17.04 14.17

1.75 0.086 14.59 14.05

Composite Materials

Table 5 material

Mechanical properties of matrix and fibers of the composite e¼ 6  0.3 mm¼1.8 mm

Matrix

r ¼1200 kg/m3 E¼4.0  109 Pa (n)¼0.353 fT ¼610 MPa 60% of volume fraction

Fiber reinforcement

r ¼1800 kg/m3 E¼242.1  109 Pa (n)¼0.22 fTlong ¼ 3800 MPa

1.2 Unitary normalized starting torque

Steel

Composites

1.0 0.8 0.6 0.4 0.2 0 +/−45°

(A)

0°/45° Carbon fiber angle

Unitary normalized maximum displacement

1.4

1.0 0.8 0.6 0.4 0.2 0

(C)

Steel Composites

1.0 0.8 0.6 0.4 0.2 0 +/−45°

(B)

1.2

0°/45° Carbon fiber angle

1.2

0°/90°

Steel Composites

+/−45°

Unitary normalized maximum stress

Composite

Unitary normalized maximum stress

262

(D)

0°/90°

1.2 Steel

1.0

Composites +/-45

0.8 Composites 0/45

0.6

Composites 0/90

0.4 0.2 0 0

0°/90°

0°/45° Carbon fiber angle

1 Unitary normalized maximum displacement

2

Fig. 26 Relative comparisons among the composite rotors with different fiber orientation vs. steel rotor: (A) starting torque; (B) stress field; (C) maximum displacements; and (D) stiffness.

The restrictions of movement are applied to the nodes corresponding to the turbine shaft, representing the sharing points between the rotor and the axis of the turbine.

2.9.8.3

Numerical Simulations of the Rotor Made of Steel and Composite Material

The details of the structural behavior of each rotor conformed in steel and composite laminate are described below. In the structural analysis of the composite laminate, the previously general formulation is used. An additional parametric comparison is also carried out in this case chosen from three pairs of fiber directions and stacking sequences. Numerical simulations involves the turbine steel rotor with the following mechanical parameters: density m ¼ 7850 kg/m3, Young modulus E¼ 210.0  106 Pa, Poisson ratio (n)¼ 0.3, and thickness t¼ 1.2 mm; and the turbine rotor made by composite six layers, each one with e ¼0.30 mm thickness, and three different composite layups, 745, 0–90, and 0–45 degrees (see mechanical properties of components in Table 5), are used in the analysis comparison.

Composite Materials

Table 6

263

Comparison between the four most significant rotors

Type of rotor

T (mm)

Required starting torque (Nm)

Maximum stress (Pa)

Maximum displacement (m)

Steel Comp. 7451 Comp. 0/45 degrees Comp. 0/90 degrees

1.8 6  0.3¼1.8 6  0.3¼1.8 6  0.3¼1.8

217 40 40 40

25.324,00 13.397,00 10.622,00 12.342,00

2,21E-03 1,65E-03 2,73E-03 2,42E-03

It can be observed after the 1s applied load, that is the minimum stress, s¼ 10622 Pa (Table 6 and Fig. 26(A)), corresponds to the non-orthogonal 0–45 degrees layup configuration, which occurs in the blades near to the shaft junction. Fig. 26(B) and (D) shows that the “0/45 degrees” laminate composite rotor works with nearly 40% level of stresses of the steel rotor. The “745 degrees” composite laminate rotor and the “0/90 degrees” composite laminate also work at less stresses than steel rotor (Table 6). All composite laminate rotors have less stiffness than the steel rotor, but particularly the composite with fibers oriented to “745 degrees” has a high stiffness and near to the steel value (see its relative comparison, Fig. 26(D)). However, the composite laminate of “0/45 degrees” has a much lower stiffness than the other (Fig. 26(D)), but is enough for this machine requirements, as its maximum displacement is tolerable in these work functions (see its relative comparison, Fig. 26(C)). The reduced starting torque of composite laminate rotor is a big advantage during the operating work of the water turbine, since composites have 5.5 times less starting torque than the steel rotor (Fig. 26(A)). It means a machine with better performances at low water flux velocities, easier to ship, handle, repair, start, etc. Concluding, the composite laminated rotor of fibers oriented to “745 degrees” is the best suited material for this function, since it has a very low rotational inertia (18.3% of the steel rotor, see Fig. 26(A)), a maximum working stress a 47% lower than the steel rotor (Fig. 26(B)), and finally has a good stiffness (53% of the value corresponding to the steel rotor, see Fig. 26(D)).

2.9.9

Conclusions

Riverbed hydroelectric development is a concept that is currently being successfully explored by several researchers, and the use of composite materials in the design of these turbines adds a significant technical and economic improvement over the use of classical materials as shown in this work. This book chapters presents a numerical formulation for the design and analysis of composite material structures to be used within the energy sector of renewable energies. The numerical formulation is based on the S/P mixing theory as manager of the different constituent models for the composite material components. Also the local damage constitutive model is employed to set the threshold for the initiation of nonlinear damage process (initiation damage) in each point of the composite material. These concepts allow a threshold criterion of damage resulting from the composition of the behaviors of all material involved in the theory of mixtures. The homogenization of local damages obtained in each one of the composite constituents to measure the damage in a shell structure is also presented. Taking advantage of the composite materials compared to the classic ones, a very good performance of the WCT immersed in the rivers is obtained, improving its performance thanks to its smaller rotational inertia. The numerical model for the treatment of the composites of WCT, and its fluid dynamics, shows a way forward and establishes a work base that offers an important tool for the design and analysis of these turbines.

2.9.10

Future Works

Two research sub-lines are proposed below to future research in the electric-power exploitation using hydrokinetic turbines. (1) From the turbine point of view and its location in each particular place, the research progress is currently made on fluid-dynamic design to optimize the rotor and the rest of the machine. In this order of things are proposed new devices and infrastructures of hydraulic flow control, which will allow a speed increase at the entrance to the turbine to improve the electrical performance. Another aspect that is also being studied is the design of sanders and separation of material in suspension to avoid impact of objects on the rotor. All this type of studies is related to the particular behavior of each kind of river. (2) From the point of view of the composite material constituting the turbine, the central theme of this chapter, future research is aimed at obtaining new fabric-reinforced compounds, which are more resistant, rigid and appropriate to withstand impacts, than compounds Reinforced with unidirectional fibers. To this end, the formulation of the S/P Theory of Mixtures presented here is currently being reformulated and generalized to fit these new materials. This, along with the great deformations and the incorporation of the misalignment of the fibers of the fabric, will allow to properly considering the interaction between orthogonal fibers located in the same plane, bases for the numerical simulation of the fabrics-reinforced composites.

264

Composite Materials

References [1] Ismail KAR, Batalha TP. A comparative study on river hydrokinetic turbines blade profile. Int J Eng Res Appl 2015;5(5):01–10. Part 1. [2] Anyi M, Kirke B. Review: evaluation of small axial flow hydrokinetic turbines for remote communities. Energy Sustain Dev 2010;14(2):110–6. [3] Botto A, Claps P, Ganora D, Laio F. Regional-scale assessment of energy potential from hydrokinetic turbines used in irrigation channels. In: Conference proceedings on 4th international conference on sustainability & environmental protection (SEEP2010), Bari, Italy, June 29–July 2, 2010. [4] Rastellini F, Oller S, Salomón O, Oñate E. Composite material non-linear modelling for long fibre-reinforced laminates. Continuum basis, computational aspects and validations. Comput Struct 2008;86:879–96. [5] X. Martinez, S. Oller, E. Barbero, Study of delamination in composites by using the serial/parallel mixing theory and a damage formulation. In: ECCOMAS thematic conference on mechanical response of composites, Portugal, September 12–14, 2007. [6] Trusdell C, Toupin R. The classical field theories. Handbuch der physic 1960;Berlin: Springer-Verlag; 1960. [7] Car E, Oller S, Oñate E. An anisotropic elasto-plastic constitutive model for large strain analysis of fiber reinforced composite materials. Comput Method Appl Mech Eng 2000;185:245–77. [8] Martinez X, Rastellini F, Flores F, Oller S, Oñate E. Computationally optimized formulation for the simulation of composite materials and delamination failures. Compos Part B: Eng Part B 2011;42:134–44. [9] Martinez X, Oller S, Rastellini F, Barbat A. A numerical procedure simulating RC structures reinforced with FRP using the serial/parallel mixing theory. Comput Struct 2008;86:1604–18. [10] Martinez X, Oller S. Numerical simulation of matrix reinforced composite materials subjected to compression loads. Arch Comput Method 2009;16:357–97. [11] Car E, Oller S, Oñate E. A large strain plasticity for anisotropic materials – composite material application. Int J Plasticity 2001;17:1437–63. [12] Oller S, Oñate E. A hygro-thermo-mechanical constitutive model for multiphase composite materials. Int J Sol Struct 1996;33:3179–86. [13] Oller Aramayo SA, Nallim LG, Oller S. An integrated procedure for the structural design of a composite rotor-hydrofoil of a water current turbine (WTC). Appl Compos Mater 2013;20:1273–88. [14] Oller S. Numerical simulation of mechanical behavior of composite materials. Barcelona, Spain: CIMNE-Springer; 2014. ISBN 978-3-319-04932-8. [15] Comellas E, Ivvan Valdez S, Oller S, Botello S. Optimization method for the determination of material parameters in damaged composite structures. Compos Struct 2015;122.417–24. doi:10.1016/j.compstruct.2014.12.014. [16] Escudero C, Oller S, Martinez X, Barbat A. A laminated structural finite element for the behavior of large non-linear reinforced concrete structures. Finite Elements in Analysis and Design 2016;119.78–94. doi:10.1016/j.finel.2016.06.001. [17] Sánchez-Palencia ES. Boundary layers and edge effects in composites. In: Sánchez Palencia E, Zaoui A, editors. Homogenization techniques for composite media. Berlin: Springer-Verlag; 1987. p. 121–92. [18] Oller S, Miquel J, Zalamea F. Composite material behavior using a homogenization double scale method. J Eng Mech 2005;131:65–79. [19] Otero F, Oller S, Martinez X, Salomón O. Numerical homogenization for composite materials analysis. Comparison with other micro mechanical formulations. Compos Struct 2015;122:405–16. [20] Otero F, Martinez X, Oller S, Salomón O. An efficient multi-scale method for non-linear analysis of composite structures. Compos Struct 2015;131:707–19. [21] Otero F, Martínez X, Oller S, Salomón O. Study and prediction of the mechanical performance of a nanotube-reinforced composite. Compos Struct 2012;94:2920–30. ISSN: 0263-8223. [22] Petracca M, Pelà L, Rossi R, Oller S, Camata. G, Spacone E. Regularization of first order computational homogenization for multiscale analysis of masonry structures. Comput Mech 2016;57(2):257–76. doi:10.1007/s00466-015-1230-6. [23] Otero F, Martínez X, Oller S, Salomón O. An efficient multi-scale method for non-linear analysis of composite structures. Compos Struct 2015;131.707–19. doi:10.1016/j. compstruct.2015.06.006. [24] Paredes J, Oller S, Barbat A. New tension-compression damage model for complex analysis of concrete structures. J Eng Mech – ASCE 2016 04016072. doi:10.1061/ (ASCE)EM.1943-7889.0001130. [25] Petracca M, Pela L, Rossi R, Oller S, Camata G, Spacone E. Multiscale computational first order homogenization of thick shells for the analysis of out-of-plane loaded masonry walls. Comput Method Appl Mech Eng 2017;315:273–301. ISSN: 0045-7825. Available from: https://doi.org/10.1016/j.cma.2016.10.046. [26] Oller S, Car E, Lubliner J. Definition of a general implicit orthotropic yield criterion. Comput Methods, Appl Mech Eng 2003;192:895–912. [27] Rango RF, Nallim LG, Oller S. Static and dynamic analysis of thick laminated plates using enriched macroelements. Compos Struct 2013;101:94–103. ISSN: 0263-8223. [28] Rango RF, Nallim LG, Oller S. An enriched macro finite element for the static analysis of thick general quadrilateral laminated composite plates. Mech Adv Mater Struct 2015 1537–6532. doi:10.1080/15376494.2015.1068400. [29] Lago LI, Ponta FL, Chen L. Advances and trends in hydrokinetic turbine systems. Energy Sustain Dev 2010;14:287–96. [30] Ben Elghali S, Benbouzid M, Charpentier J. Marine tidal current electric power generation technology: state of the art and current status. In: IEEE international electric machines & drives conference (IEMDC'07); 2007, p. 2. [31] Shafei MAR, Ibrahim DK, Ali AM, Younes MAA, El-Zahab EA. Novel approach for hydrokinetic turbine applications. Energy Sustain Dev 2015;27:120–6. [32] Khan M, Iqbal MM, Quaicoe J. River current energy conversion systems: Progress, prospects and challenges. Renew Sustain Energy Rev 2008;12:2177–93. [33] Khan M, Bhuyan G, Iqbal M, Quaicoe J. Hydrokinetic energy conversion systems and assessment of horizontal and vertical axis turbines for river and tidal applications: A technology status review. Appl Energy 2009;86:1823–35. [34] Oller Aramayo SA, Nallim LG, Oller S. Fluid dynamic design of an axial rotor for hydrokinetic riverbed turbine-improvement introduced by a high lift foil profile. Environ Progress Sustain Energy 2016;35:1198–206. [35] Selig MS, Guglielmo JJ. High-lift low reynolds number airfoil design. J Aircraft 1997;34:72–9. [36] Kolekar N, Hu Z, Banerjee A, Du X. Hydrodynamic design and optimization of hydro‐kinetic turbines using a robust design method. In: Proceedings of the 1st marine energy technology symposium METS13, Washington, D.C.; April 10–11, 2013. [37] Oller Aramayo SA, Nallim LG, Oller S. The usability of the Selig S1223 profile airfoil as a high lift hydrofoil for hydrokinetic application. J Appl Fluid Mech 2016;9:537–42. [38] Goundar JN, Rafiuddin Ahmed M, Lee Y. Numerical and experimental studies on hydrofoils for marine current turbines. Renew Energy 2012;42:173–9. [39] Betz A. Introduction to the theory of flow machines. Oxford, New York: Pergamon Press; 1966. [40] Froude W. On the elementary relation between pitch, slip and propulsive efficiency. Trans R Inst Nav Arch 1878;19:47–65. [41] Carlton JS. Marine propellers and propulsion. 2nd ed. Oxford: Elsevier; 2012. [42] Hughes T. Multiscale phenomena, Green’s functions, the Dirichletto–Neumann formulation, subgrid-scale models, bubbles and the origin of stabilized methods. Comput Method Appl Mech Eng 1995;127:387–401. [43] Guermond J. Stabilization of Galerkin approximations of transport equations by subgrid modeling. Math Model Num Anal 1999;33:1293–316. [44] van Tooren M. Response 1 – airbus composite aircraft fuselages – next or never. In: Vermeeren C, editor. Around Glare a New Aircraft Material in Context. Netherlands: Springer; 2007. p. 145–58. [45] Ferziger J, Perić M. Computational methods for fluid dynamics. Berlin Heidelberg, New York: Springer-Verlag; 2002. [46] Oller S. Nonlinear dynamics of structures. Barcelona, Spain: CIMNE-Springer; 2014. ISBN 978-3-319-05193-2.

Composite Materials

265

[47] Paredes JA, Barbat AH, Oller S. A compression-tension concrete damage model, applied to a wind turbine reinforced concrete tower. Engineering Structures 2011;33.3559–69. doi:10.1016/j.engstruct.2011.07.020. [48] Lubliner J. Thermomechanics of deformable bodies. Lecture Notes, Berkeley, NY: Department of Civil Engineering, University of California; 1985. [49] Lubliner J. Plasticity theory. New York, NY: MacMillan; 1990. [50] Maugin GA. The thermomechanics of plasticity and fracture. Cambridge: University Press; 1992. [51] Lubliner J, Oliver J, Oller S, Oñate E. A plastic-damage model for concrete. Int J Sol Struct 1989;25:299–326. [52] Oller S, Oñate E, Oliver J, Lubliner J. Finite element non-linear analysis of concrete structures using a plastic-damage model. Eng Fracture Mech 1990;35:219–31. [53] Zienkiewicz OC, Taylor LR. The finite element method. London: McGraw-Hill; 1991. [54] Oliver J, Cervera M, Oller S, Lubliner J. A simple damage model for concrete, including long term effects. In: Second international conference on computer aided analysis and design of concrete structures, vol. 2, Viena; 1990. p. 945–958.

Further Reading Pérez MA, Martínez X, Oller S, Gil L, Rastellini F, Flores F. Impact damage prediction in carbon fiber-reinforced laminated composite using the matrix-reinforced mixing theory. Composite Structures 2013;104:239–48. ISSN: 0263-8223.

Relevant Websites https://www.gidhome.com GiD. http://www.cimne.upc.es/kratos/ KRATOS Multi-physics. http://www.cimne.com/PLCd Plastic Crack Dynamic.

2.10 Semiconductors Franco Gaspari, University of Ontario Institute of Technology, Oshawa, ON, Canada r 2018 Elsevier Inc. All rights reserved.

Nomenclature 2.10.1 Introduction 2.10.2 Background/Fundamentals 2.10.2.1 Energy Bands in Semiconductors 2.10.2.2 Electron and Holes Statistics and Dynamics 2.10.2.3 Doping 2.10.2.4 Transport 2.10.2.5 p–n Junctions 2.10.2.6 Basic Equations for a p–n Junction 2.10.2.7 Metal–Semiconductor Junctions 2.10.2.8 Optical Absorption in Semiconductors 2.10.2.9 Recombination in Semiconductors 2.10.3 Applications 2.10.3.1 Microelectronic Applications 2.10.3.2 Optoelectronic Devices 2.10.3.3 Solar Cells 2.10.3.3.1 Heterojunction with intrinsic thin layer, Tandem and multi-junction cells 2.10.4 Analysis and Assessment 2.10.4.1 The Evolution of Computers 2.10.4.2 The Evolution of Light Emitting Diodes 2.10.5 Illustrative Examples 2.10.6 Results and Discussion 2.10.7 Future Directions 2.10.8 Closing Remarks Acknowledgement References Further Reading Relevant Websites

266 268 269 271 272 273 273 274 276 277 279 281 282 282 284 286 288 290 291 292 294 297 298 299 300 300 301 301

Nomenclature Symbol/Acronym

Definition

Units

a a A b CdSe CdTe CIGS d D De,h DSSC e e h ! E ! Ed Ea,d Ec

Acceleration Absorption coefficient Area of semiconductor slab Recombination coefficient Cadmium selenide Cadmium telluride Copper–indium–gallium–selenide Thickness of depletion layer (p–n junction) Depth of semiconductor slab Diffusion constant for electrons, holes Dye sensitized solar cell Electron charge Electron–hole pair Electric field Electric field in the depletion layer Acceptor, Donor energy level Conduction band edge energy level

m/s m 1 m2

266

Comprehensive Energy Systems, Volume 2

m m m2s

1

C V/m V/m eV eV

doi:10.1016/B978-0-12-809597-3.00221-2

Semiconductors

EF EG Ev F(E) FF FM,S g(E) G,Gph Ga GaAs GaP GaN GaSb Ge Gph h ℏ I0 Iab IL0 Imax Iph ISC InAs InP InSb Je,h ! k k l mi ms m e;h me,h n ne,h NA,D ! p Pin PFOM q QD QW ! ri R ! Rn Si SiC SMD s T tr U Ueff(r) vth VB VC

Fermi energy Energy gap of semiconductors Valence band edge energy level Fermi–Dirac distribution function Fill factor (solar cells) Work function (metal, semiconductor) Density of states per unit energy Carrier optical generation rate Gallium Gallium arsenide Gallium phosphide Gallium nitride Gallium antimonide Germanium Density of photons per unit second Planck’s constant (4.135667662(25)  10 15) h/2p Diode leakage current Absorbed light intensity Intensity of incident light Maximum operating current for a solar cell Diode photocurrent Short circuit current for a diode Indium arsenide Indium phosphide Indium antimonide Current density for electrons, holes Wave vector Boltzmann constant Angular momentum Magnetic quantum number Spin quantum number Electron, hole effective mass Mobility of electrons, holes Orbital, principal quantum number Density of mobile electrons, holes Density of acceptors, donors Crystal momentum Incident optical power Power figure of merit Electron, hole charge Quantum dot Quantum well Lattice electron coordinates Reflectance Lattice nuclear coordinates Silicon Silicon carbide Surface mount light emitting diodes (LED) Conductivity Transmittance Relaxation time Carrier recombination rate Effective potential Thermal velocity External applied bias Diode contact potential

267

eV eV eV % eV m m

3 3

eV 1 s 1

m 2s eV s eV s A Wm2 Wm2 A A A

1

Am 2 m 1 JK 1

Kg m2 V m m

1

s

3 3

Wm2 C

O s m

1

m

3

s

m/s V V V

1

1

1

268

Semiconductors

Vmax VOC χ cj(rjξj)

2.10.1

Maximum operating voltage for a solar cell Open circuit voltage for a diode Electron affinity Particle function

V

Introduction

Semiconductors have played a major role in the 20th century in bringing forth disruptive technological advances. Devices made from semiconductor materials are the foundation of modern electronics, including radio, computers, telephones, and many other devices. Semiconductor devices range from diodes to transistors and integrated circuits (ICs) but also photodetectors and solar cells. Their presence can be found everywhere in our daily lives, in all electronic devices that make our life faster and more productive. We have come to rely on them and increasingly have come to expect higher performance at lower cost. The increase in speed of these electronic components is clearly understood if we consider the constant advances of some of the technological devices we use every day, like personal computers (PC) and lighting devices. Indeed, one has only to think on the low resale value of a 5 years old computer to realize how quickly these devices become obsolete. The complexity and performance of today’s PC vastly exceeds that of an old computer; however, while speed and memory are distinctively superior, prices remain comparable. It is in fact the improved device performance coupled with the corresponding decrease in fabrication costs that makes semiconductors the signature material that has defined the 20th century technological revolution. Through industrial skill and technological advances we have been able to make smaller and smaller transistors, from about 20 mm in the early 1960s to below 1 mm nowadays, capable of delivering better performance while consuming less power and because of their smaller size they can also be manufactured at a lower cost per device. Most semiconductor devices are based on silicon (Si), although germanium (Ge) and compound semiconductors like gallium arsenide (GaAs) have also proven very effective. Indeed, Si, Ge, and GaAs are the main materials used for electronic applications, such as digital circuits and microwave circuits. Digital circuits, based on semiconductor electronics like transistors (bipolar, field effect, etc.), have been used in energy applications in a variety of forms, from AC to DC converters to pure signal amplification, while microwave diodes have been defined “the most powerful solid state sources of microwave energy” [1]. On the other hand, semiconductors have also been employed in the area of energy production (photovoltaics (PV)) and light emission (light emitting diodes, or LEDs) by exploiting the interaction between the semiconductor and light (optical) excitations. Semiconductor lasers are also based on such interaction. When assessing the energy impact of semiconductors technology, there are two major aspects that need to be emphasized: firstly, the most prominent example of application of semiconductors for “pure” energy production can be found in the PV field. In order to assess the energy and economic impact of PV, it is common to analyse the correlation between fabrication cost and efficiency of a PV system. Other considerations might include greenhouse gas (GHG) emission contribution of the fabrication process. On the other hand, as indicated above, semiconductor technology has been having an impact also on energy management since the mid-20th century. One must remember that when we talk about energy efficiency we refer to the extraction of greater value from our energy resources. In short, energy efficiency is about achieving cost-effective reductions in wasted energy. Semiconductors have indeed introduced a new paradigm in this analysis, as their main electronic applications, like sensors, microprocessors, smart grids and virtualization technologies, have revolutionized the relationship between economic production and energy consumption. In summary, integrating semiconductor technology within well-known device processes leads to saving in energy consumption while maintaining the same, or in many cases obtaining a better, performance. There are a great number of excellent textbooks and review articles that cover the fundamental processes linked to semiconductors and the theory behind them (see, for instance, Refs. [2–5] and the list of publications at the end of this chapter in the “further reading” section). Most of the theoretical background presented here is a summary of these publications. Indeed, it is not possible to cover all the relevant aspects related to semiconductors in a single chapter, however, the main goal of this chapter is to provide the reader with a comprehensive view of the uniqueness of semiconductors, when compared to other materials like metals or insulators, and of their flexibility. Therefore, an initial general review of the physical properties of semiconductors is provided at the beginning of this chapter, followed by a review of their applications in microelectronics and PVs. The first, microelectronics, has been completely revolutionized by the introduction of semiconductors, while the second, PVs, owes its very existence to the nature of these materials. Semiconductor physics is covered in all undergraduate physics programs, however, a review of some of the basic ideas is necessary in order to fully appreciate the impact of the optoelectronic properties on device applications. In particular, the concept of semiconductors is introduced in the background Section 2.10.2 with a focus on the definition and properties of gap energy states, transport properties, optical absorption, doping, junctions, and device optoelectronic processes. The main purpose of Section 2.10.2 is to provide the reader with a primer of semiconductors device physics. Although the overall descriptions are by necessity incomplete, they do cover the necessary background to understand how a semiconductor-based device functions. Without the knowledge of band gaps, transport and absorption properties, and doping mechanisms, it would be very difficult to comprehend the dynamics of the devices, and to correlate the functions and processes of novel materials with that of the standard ones.

Semiconductors

269

Section 2.10.3 presents a summary of the major electronic and energy producing semiconductor-based devices. These represent the basic applications (or systems) that will be considered throughout the chapter. The implications of heterojunctions and metal–semiconductor junctions are presented along the standard p–n junctions, and the basic applications in electronics, optoelectronic devices, and solar cells are discussed in this section. It is important to note that Section 2.10.3 also covers material described in a variety of textbooks, but the understanding of the physics principles of semiconductors devices is also crucial in order to appreciate the variations and alternative solutions introduced by novel semiconductor materials and structures. This is particularly relevant when we want to assess the energy aspect of semiconductors; as indicated previously, it is the very nature of semiconductors that has led to the unique, disruptive, technological applications described in this chapter, whether we consider energy production or energy management. Section 2.10.4 will offer an analysis and assessment of the impact of semiconductor-based technology by introducing two iconic examples, namely the evolution of computer technology and of light emitting devices. Section 2.10.5 on the other hand describes a series of examples aimed at illustrating alternative applications, materials, and device structures. The examples will focus in particular on the energy aspect of these applications. For instance, a distinction is made between the energy impact of semiconductors in microelectronics, where the majority of applications are related to energy management, and the pure energy producing devices, like solar cells. Finally, Section 2.10.6 will summarize the major points of the chapter while Sections 2.10.7 and 2.10.8 will illustrate future directions and conclusions, respectively.

2.10.2

Background/Fundamentals

In this section, a review of the basic physics of semiconductors is presented. As mentioned in the introduction, many of these concepts are covered in undergraduate physics classes. However, the impact of semiconductor technology is directly related to the peculiar physical properties of these materials. Furthermore, alternative solutions in energy and electronic applications try to mimic the standard semiconductors’ physical processes. Therefore, a background of the relevant characteristics is necessary in order to appreciate fully the applications of these materials. The analysis presented below is by no means complete, but will serve to provide an understanding of the basic equations and the physics of these materials. Indeed, it is useful to begin from the basic process of crystal formation and bonding, so that a first distinction can be made with metals and insulators. The bonding of atoms in semiconductors is accomplished by electrostatic forces and by the tendency of atoms to fill their outer shells. Interatomic attraction is balanced by short-range repulsion due to strong resistance of atoms against interpenetration of core shells. As two similar atoms approach each other, the wave function of their electrons begins to overlap [6] (see Fig. 1). In Fig. 1(A), the forces acting between the two atoms are plotted as a function of distance. It should be noted that the two main forces are an attractive force, due to attraction between the positive nuclei and the electron cloud of the other atom, and a repulsive force, that becomes more evident as the interatomic distance decreases and is mainly due to the nuclei electrostatic repulsion. At r¼r0 the two forces even out and the two atoms reach equilibrium. Fig. 2(B) shows how this distance corresponds to the minimum of the system potential energy. Pauli’s exclusion principle dictates that all spin-paired electrons acquire energies which are slightly different from their values in the isolated atom. When N isolated atoms assemble into a crystal structure, the overlap increases. This leads to the formation of band states as the number of atoms increases (see, for instance, Refs. [2–4,6] for a more detailed treatment of band formation). r0 +

+

r0

FR = Repulsive force

− (A)

Attraction

Potential energy, E (r)

FN = Net force

ER = Repulsive PE E = Net PE

0 Repulsion

FA = Attractive force

r E0

r0

EA = Attractive PE

−

Attraction

Separated atoms

Interatomic separation, r Repulsion

Force

0

+

+

+

+

Molecule

r=∞

(B)

Fig. 1 Net force (A) and potential energy (B) vs. interatomic distance. Reproduced from Kasap S. Principles of electronic materials and devices. New York, NY: McGraw-Hill. Available from: http://ElectronicMaterials.Usask.Ca; 2006.

270

Semiconductors

Sub shell

s

p

d

f

Value of I

0

1

2

3

Shape

Spherical

Dumb bell

Double-dumb bell

Complex

y

y

y

y

x

Structure x

x

S orbital

x

z P orbital

d orbital

f-orbital

Fig. 2 Electron density and orbital shapes for 1s, 2p, 3d, and 4f orbitals. Reproduced from Yahoo. Available from: https://images.search.yahoo. com/yhs/search?p=electron þ orbital þ shapes&fr=yhs-mozilla-001&hspart=mozilla&hsimp=yhs-001&imgurl=http%3A%2F%2F3.bp.blogspot.com %2F-7WaPCpa9mF4%2FTfgpuFxlFII%2FAAAAAAAAACA%2Fx5_0HjUpn2Y%2Fs1600%2Fch9orbitals1.jpg#id=1&iurl=http%3A%2F%2F3.bp. blogspot.com%2F-7WaPCpa9mF4%2FTfgpuFxlFII%2FAAAAAAAAACA%2Fx5_0HjUpn2Y%2Fs1600%2Fch9orbitals1.jpg&action=click.

Bonding can be ionic, where forces are predominantly electrostatic; or covalent, where valence electrons are shared by neighbouring atoms to fill their outer shells. These attractive forces, leading to the bonding of atoms, are counteracted upon by the repulsive interatomic forces, called Born forces, which are caused by a strong resistance of the electronic shells of atoms against interpenetration. Ionic bonding occurs primarily between elements of group I and group VII of the periodic system of elements, where one electron needs to be exchanged. This is due to the Coulomb attraction between ions. Covalent bonding is caused by two electrons that are shared between two atoms. The analysis of covalent bonding requires a quantum mechanics treatment. Indeed, the quantum states on N electrons in a solid are solutions of the N particle Schrodinger equation: ! N X ℏ2 2 Ze2 X 1 1 X e2 ∇i c c þ c ¼ Ec ð1Þ ! ! 2m 4pe0 Rn j! 8pe0 ij j ri ! rj j ri Rn j i¼1

! where ! ri are electron coordinates and Rn are nuclear coordinates. We have no hope of solving this equation. However, in order to determine the wave-function for atoms and molecules with more than one electron, many approximations can be used to describe realistic physical properties. The Hartree approximation, for instance, associates a potential to each electron and computes the interactions between individual electrons; the Hartree–Fock approximation adds an exchange term to the equation, while other, more sophisticated approximations introduce the effect of periodic potentials due to the crystal lattice and electron screening due to electron clouds that surround the individual atoms [4]. Essentially, one starts from the orbital approximation, which states that each electron in a many-electron system occupies its own one-electron function, which is called an orbital. Fig. 2 shows the electron density and the orbital shape for 1s, 2p, 3d, and 4f orbitals. It must be underlined that the coloured and shaded areas represent the probability of an electron to be found in that region. To overcome the complexity of Eq. (1), one must try to construct an approximate solution by initially defining an effective potential Ueff(r) to be used in the one-particle Schrodinger equation ℏ2 2 ∇ ci ðr Þ þ U eff ðr Þci ðr Þ ¼ ei ci ðr Þ 2m

ð2Þ

In the Hartree approximation, Ueff(r) includes the energy in the field of all the ions in the metal, and also the interaction of the electron in the state ci(r) with the average field of all other electrons. " # Z jcj ðr 0 Þj2 e2 X ð3Þ U eff ðr Þ ¼ U ion ðr Þ þ dr 0 jr r 0 j 4pe0 ja i Of course the Schrodinger equation is nonlinear in the Hartree approximation, so it must be solved by iteration. A form is guessed for Ueff(r), the equation is solved for the wave function ci(r). These are used to calculate a more accurate Ueff(r), and the cycle is repeated until the effective potential has converged. An important flaw of the Hartree theory is that it can be shown to correspond to a many electron wave function that is simply a product of single particle functions cj(rjξj). That is to say, c is of the form: cðr1 ξ1 ; r2 ξ2 …rN ξN Þ ¼ c1 ðr1 ξ1 Þc2 ðr2 ξ2 Þ…cN ðrN ξN Þ

ð4Þ

Semiconductors

271

But such a wave function is not an acceptable form for the many particle wave function, because it does not satisfy the Pauli Exclusion Principle. The simplest way to correct things is to replace the trial wave function by a Slater determinant [4]   c1 ðr1 ξ1 Þc1 ðr2 ξ2 Þ…c1 ðrN ξN Þ   c ðr ξ Þc ðr ξ Þ…c ðr ξ Þ  2 1 1 2 2 2 2 N N cðr1 ξ1 ; r2 ξ2 …rN ξN Þ ¼  ⋮   c ðr ξ Þc ðr ξ Þ…c ðr ξ Þ  N 1 1 N 2 2 N N N

         

ð5Þ

Then by applying the variational theorem [4] it can be shown that the spin-orbitals ci(r) are solutions of the Hartree–Fock equation: 2 6 4

2

ℏ 2 ∇ 2m

X

2

2

Ze e þ 4pe0 jr RN j 4pe0

X ja i

Z

occupied

3 0 2 jc ð r Þj j 7 dr 0 5 ci ðr Þ jr r 0 j

e2 X 4pe0 j

Z

dr 0

occupied

0 0 jc j ðr Þci ðr Þj

jr

r0j

ci ðr Þδqi qj ¼ ei ci ðr Þ

ð6Þ

Further refinements to the Hartree–Fock approximation, as mentioned previously, include the introduction of a “screening” function, where, instead of considering the electrons as part of a “free” electron gas, local atomic potentials are introduced and their effect on the electron gas interaction is included. Examples of screening models include the Thomas–Fermi theory of screening and the Lindhard theory of screening (see Ref. [4] for more detailed descriptions of these approximations).

2.10.2.1

Energy Bands in Semiconductors

If the electron gas is part of a periodic crystal lattice, the models described above can be further simplified by using the periodicity of the lattice via the Bloch theorem [4], which introduces a periodic one-electron wave function. This leads to the so-called ! ! reduced-zone representation of the electrons allowed energy states as a function of a wave-vector k [4]. The quantity ℏ k is usually referred to as the “crystal momentum,” where ℏ is Planck’s constant divided by 2p. In the reduced-zone approximation, at absolute zero, the electrons of the system will occupy the allowed states, one per state, following the Pauli Exclusion Principle. As a consequence, in a one dimensional crystal, some of the allowed bands will be entirely filled, some will be entirely empty, and one band might be partially filled. If we apply an electric field, no current will result from both the completely filled bands and the completely empty ones. The fact that no electric current can be generated by empty or completely filled bands is one of the main reasons why semiconductors are so important for energy applications. That is, any electrical conductivity, due to the motion of free electrons, must come from the motion of electrons in partially filled bands. Therefore, a distinction is made among three categories of materials, based on their energy bands, i.e., insulators, metallic conductors and semiconductors, as shown in Fig. 3. In an insulator, the number of electrons in the crystal is just sufficient to completely fill a number of energy bands. Above these bands is a series of empty bands, with a forbidden gap too wide for electrons to acquire enough energy to overcome, even at high temperatures. If the energy gap is small, there will be enough statistical probability for an electron to be excited from the valence band to the conduction band. These electrons are available for transport and will respond to an external potential. Furthermore, the electrons

Energy Conduction band Conduction band Conduction band Fermi level

Overlap Valence band Valence band

Valence band

Conductor

Semiconductor

Fig. 3 Energy band diagrams for conductors, semiconductors, and insulators.

Insulator

272

Semiconductors

promoted to the conduction band will leave behind empty electronic states near the top of the valence band. These states are called “holes” and can be considered as positive electrons, although with a larger mass, and will also contribute to conduction. In metals, there is an overlap between the valence band and the conduction band, however, contrary to semiconductors, the number of electrons will remain constant regardless of temperature. An interesting contrast between metals and semiconductors is represented by the dependence of the conductivity on temperature. In metals, at higher temperatures, scattering and collision processes will decrease the metal conductivity, while in semiconductors more carriers will be excited in the conduction band with increasing temperatures and the conductivity will indeed increase. There are two major concepts that form the theoretical basis for the description of electronic energy levels: the Pauli Exclusion Principle, mentioned above, and the Fermi–Dirac statistics (see appropriate chapters in Refs. [2–6] for an in-depth description of these concepts). In summary, the Pauli exclusion principle states that no two electrons can have the same set of quantum numbers, n (orbital, principal quantum number), l (azimuthal quantum number – angular momentum), mi (magnetic quantum number), and ms (spin). For example, n ¼2 has eight possible states of n,l,m,i, and ms. In order to conduct, a band must be partially full so that there might be higher states to which electrons can jump. Since only electrons in the conduction band can move and create a current flow in response to an applied voltage, it is convenient to define a function F(E) as the probability that a given energy level E somewhere in a band is actually occupied by an electron, with N(E)dE being the number of electrons per unit volume between energies E and E þ dE. The expression for F(E) is given by: F ðEÞ ¼

1 eðE

EF Þ=kT

ð7Þ

þ1

where EF is the Fermi energy (or Fermi level), k is the Boltzmann constant, and T is the absolute temperature. The Fermi energy is the maximum energy an electron can have at absolute zero. In semiconductors, EF must be somewhere between the valence band and the conduction band. Indeed, it can be shown that EFEEG/2, where EG is the energy gap of the semiconductor [4]. In the following sections the physics of semiconductors will be examined in more detail, however, more in-depth descriptions and analysis can be found in the references cited so far and in the suggested reading reference list.

2.10.2.2

Electron and Holes Statistics and Dynamics

In order to obtain a current we must apply an electric field which will exert an electric force on both electrons and holes. The effects of the periodic forces of the crystal atoms can be accounted for with simple approximations. For an electron lying within the conduction band, Newton’s law becomes: F ¼ m e a¼

dp dt

ð8Þ

! where the asterisk indicates an “effective” mass of the electron which incorporates the periodic force of the lattice, and p is the ! crystal momentum, ℏ k [4]. The number of allowed states per unit volume in a semiconductor is obviously zero for energies in the forbidden gap and nonzero in the allowed bands. The formula that describes the number of states per unit volume and energy, at energy E near the conduction band edge (Ec), is [2–4]: NðEÞ ¼ 8

pffiffiffi 3=2 2pme ðE h3

Ec Þ1=2

ð9Þ

The number of electrons (nc) and holes (nh) that can contribute to transport is given by:   Ec EF or; more generally; nc ¼ Nc f ðEc ; EF ; TÞ nc ¼ Nc exp kT nh ¼ p ¼ Nv exp



Ev

EF kT



or; more generally; nv ¼ p ¼ Nv ½1

f ðEv ; EF ; T ފ

ð10Þ

ð11Þ

Fig. 4 provides a visual summary of the statistics described above. Fig. 4(A) shows the simplified energy band diagram, with the white circles in the valence band denoting the holes created by thermal excitation of electrons in the conduction band (red circles). The relevant energy levels, Ev, Ec, and EF are also indicated. Fig. 4(B) is a plot of the density of allowed states g(E) in the valence band, and in the conduction band, versus energy. The density of possible states at a particular energy must be multiplied by the probability of a carrier (either electron or hole) of occupying that energy level, in order to calculate the number of carriers at that energy. The probability of occupancy for electrons and holes is described by the Fermi–Dirac function, and is plotted in Fig. 4(C). Finally, Fig. 4(D) shows the product g(E)f(E) and the coloured areas show the actual density of electrons in the conduction band (red) and holes in the valence band (blue).

Semiconductors

E Ec +

g(E) ∝ (E−Ec)1/2

E

273

E [1−f(E)]

CB

For electrons

Area = n Ec

Ec

nE(E)

EF

EF

Ev

Ev

pE(E) Area = p

For holes VB 0 (A)

g(E) (B)

(C)

f(E)

(D)

nE(E) or pE(E)

Fig. 4 (A) Energy band diagram. (B) Density of states g(E) (number of states per unit energy per unit volume). (C) Fermi–Dirac probability function f(E) (probability of occupancy of a state). (D) The product of f(E) and g(E) is the energy density of electrons in the conduction band (CB) (number of electrons per unit energy per unit volume). The area under nE(E) vs. E is the electron concentration in the conduction band. Reproduced from Kasap S. Principles of electronic materials and devices. New York, NY: McGraw-Hill. Available from: http://ElectronicMaterials. Usask.Ca; 2006.

2.10.2.3

Doping

Another advantage of semiconductors is represented by the possibility of doping the materials, which is achieved by the introduction of impurity atoms and the consequent increase in free carriers available for transport. Impurity atoms can be incorporated in two ways: interstitial impurities, squeezed between the atoms of the host lattice; and substitutional impurities, which take the place of a host atom. Atoms from groups III (e.g., boron) and V (e.g., phosphorus) act as substitutional dopants for group IV semiconductors (e.g., silicon). All references cited so far give excellent reviews of the physics of semiconductor doping. The effects of doping on the electronic states of the semiconductor can be summarized as follows: doping impurities form a band of energy states very close to the conduction band (for n-type doping) or to the valence band (p-type doping). Note that n-type and p-type refer to the type of extra carriers generated by the ionization of the impurities, which can be either donors (n-type, providing extra electrons to the conduction band), or acceptors (p-type, providing extra holes to the valence band). The small gap in energy between the dopant states and their respective bands allows for almost complete ionization of the dopant at room temperature. We have now sharp allowed energy levels in the band gap, belonging to electrons of the doping atoms, or since electrons cannot be distinguished, to all electrons in the semiconductor. These levels may or may not be occupied by an electron. If it is not occupied by an electron, it is by necessity occupied by a hole; the Fermi distribution will give the probability for occupancy as before. Finally, the Fermi level shifts toward the conduction band (for n-doping) or the valence band (for p-doping). This effect of the doping of semiconductors is crucial for the understanding of all semiconductor-based devices. A visual representation of doping states is given in Fig. 5. In Fig. 5(A) donor states are shown at energy ED near Ec, and their density of states (DOS), g(E) is shown in Fig. 5(B). Fig. 5(C) shows the shift in the Fermi level, and, finally, Fig. 5(D) shows the actual carrier density (electrons, in this case), in red, given by the product of the occupation probability, F(E), and the donor DOS, g(E).

2.10.2.4

Transport

The conductivity of semiconductors and the mobility of carriers are two fundamental parameters that must be accounted for in the analysis of semiconductors. There are two types of transport that occur in semiconductors: 1. Drift, which occurs under the influence of an electric field. In a crystal, electrons will undergo collisions. The average time between collisions is called relaxation time. The average velocity increase between collisions is the drift velocity [2–4]. 2. Diffusion, which is due to concentration gradients. Thermal velocity is the main cause of diffusion [2–4]. These two modes of transport are described by the following equations: For Drift transport, ! ! v d ¼ qme ne E ðDrift current density for electronsÞ J e ¼ qne !

ð12Þ

! ! J h ¼ qnh ! v d ¼ qmh nh E ðDrift current density for holesÞ ð13Þ ! ! where q is the electron charge, E is the electric field, v d is the drift velocity, ne,h are the densities of electrons or holes, and me;h ¼ vEd ¼ etr are the respective mobilities. me;h

274

Semiconductors

E Ec +

g(E) ∝ (E −Ec)1/2

E

E [1−f(E)]

CB

For electrons

Area = n Ec

Ec

nE(E)

EF

ED EF

Ev

Ev

pE (E) Area = p

For holes VB 0 (A)

g(E) (B)

f(E ) (C)

nE(E) or pE(E) (D)

Fig. 5 Band representation for group V doping. The donor state is very close to the conduction band (A) and forms a band of states at that energy level (B). The Fermi level shifts towards the conduction band (C). The consequent occupancy of the impurity levels is shown in (D). Reproduced from Kasap S. Principles of electronic materials and devices. New York, NY: McGraw-Hill. Available from: http://ElectronicMaterials. Usask.Ca; 2006.

The total current flow is the sum of the two. The conductivity is: s¼

1 !! ¼ J E ¼ qðne me þ nh mh Þ r

ð14Þ

Note that as doping level increases, the relaxation time decreases and the mobility decreases. For Diffusion transport we can consider the one dimension expression. For electrons: dn ! where De is the diffusion constant J e ¼ qDe dx

ð15Þ

For holes: ! J h¼

qDh

dp dx

ð16Þ

Drift and Diffusion are linked by Einstein’s relations [2–4]: De ¼

2.10.2.5

kT kT m and Dh ¼ m q e q h

ð17Þ

p–n Junctions

p–n junctions have been the subject of many studies for their applications in rectifiers, transistors, and solar cells [5–9]. In this section the electrostatics of a p–n junction is presented from the point of view of the Fermi level position. One must first introduce the concept of quasi-Fermi energies, which applies when the electrons and holes are not in thermal equilibrium with each other. This occurs when an external voltage is applied to the device of interest. Even though the electrons and holes are not in thermal equilibrium with each other, however, they still are in thermal equilibrium with themselves and can still be described by a Fermi energy which is now different for the electrons and the holes [2–4,6,8]. Fig. 6 represents an idealized build-up to a p–n junction. It is possible to imagine the two sides of a p–n junction separately at first, so that their Fermi levels can be shown also separately. Although this is not a realistic case, where doping occurs on both sides of a single wafer, it does provide a simple picture of the implications of forming a junction. This is shown in Fig. 7. Basic physics and chemistry teaches that a system in thermal equilibrium can have only one Fermi level (also called chemical potential in chemistry), so if the two sides are in contact, there will be a shift between the band levels, and a potential barrier will be formed [2–4]. Furthermore, a depletion region will also appear about the junction. The formation of the depletion region can be explained looking at Fig. 8. In Fig. 8, each side of the p–n junction has an excess of either mobile holes or mobile electrons, which creates a concentration gradient at the junction and initiates a diffusion process. Consequently, near the junction, there will be e–h recombination due to diffusion. However, recombination will affect only the mobile carriers, a “depletion” region (i.e., depleted of mobile carriers) will ! ! be formed, and the fixed charges will produce an electric field, E d . The electric field E d tends to sweep any electron in the depletion region near the junction towards the n-type material and holes towards the p-type material. As the depletion region

Semiconductors

p

n

p = pp0 ≈ NA n2 n = np0 ≈ i NA

n = pn0 ≈ ND n2 p = pn0 ≈ i ND

275

CB EFn

EFp VB Fig. 6 A p-type semiconductor and an n-type semiconductor. The respective Fermi levels are shown in the bottom diagrams.

n

p

Transition region

E2

E1

EF

Fig. 7 The energy band diagram for an idealized p–n junction. The transition region is characterized by a potential barrier, qc0 ¼EG

p-type + −

−

+ −

+ −

−

+ −

+ −

−

E2.

n-type

Junction

+ −

E1

+

− +

− +

+

− +

− +

+

− +

− +

Ed

Depletion region Fig. 8 Schematics of the formation of a depletion layer in a p–n junction. Near the junction, due to diffusion, there will be e–h recombination. A “depletion” region will be formed, and the fixed charges will produce an electric field Ed. Note that the colored circles represents mobile electrons ( ) and holes ( þ ). All signs outside the circles are fixed ionized donors ( þ ) or acceptors ( ).

becomes free of mobile carriers, it can be considered an insulating region. An effective potential VC will be produced (contact potential) where, VC ¼|Ed| d, where d is the thickness of the depletion region. We can connect a battery, i.e., apply an external voltage, to bias the p–n junction. According to the direction of the bias, one can establish either reverse bias or forward bias conditions. These two cases are shown in Figs. 9 and 10. Under reverse bias, the p-type

276

Semiconductors

p-type + −

−−

+ −

−−

+ −

−−

Reverse bias

n-type

+

− +

+

+

− +

+

+

− +

EB −

Ed −

Depletion region −

R

+

Fig. 9 A p–n junction under reverse bias.

p-type

Forward bias

+ −

+ −

+ −

+ −

−

+ −

+ −

−

n-type − +

− +

+

− +

− +

+

− +

− +

Eb Ed

+

−

Fig. 10 A p–n junction under forward bias.

material is made negative with respect to the n-type material. The thickness of the depletion region will increase and there will be very little current flowing. The small current observed will be due to minority carriers. Under forward bias, recombination will take place when electrons reach the p-type material and holes the n-type material, so a (large) current will flow. More electrons will continuously be injected into the n-type material by the wire connected to the negative side of the battery, and electrons will be taken out of the p-type side from the positive side. Note that to achieve conduction we need VB4VC (B0.5 V for Si), where VB in the applied external bias.

2.10.2.6

Basic Equations for a p–n Junction

The electrostatic analysis of a p–n junction (or diode) is of interest since it provides knowledge about the charge density and the electric field in the depletion region. It is also required to obtain the capacitance–voltage characteristics of the diode. It is useful to outline a set of basic equations of semiconductor-device physics, which are: Poisson’s equation, Current-density equations, and the Continuity equation. In depth derivation of these equations can be found in several sources (see, for instance, the appropriate sections in McKelvey [3]). The summary of the equation set needed for a complete analysis of a p–n junction is given below. Table 1 gives an explanation of all parameters used in Eqs. (18) to (22). dE q ¼ ðp dx e

n þ ND

NA Þ ðPoisson’s EquationÞ

ð18Þ

dn ! ! ðCurrent density for electrons; from Eq: ð15Þ and Eq: ð17ÞÞ J e ¼ qme n E þ qDe dx

ð19Þ

dp ! ! J h ¼ qmh p E þ qDh ðCurrent density for holes; from Eq: ð16Þ and Eq: ð18ÞÞ dx

ð20Þ

1 dJe ¼U q dx

G ðContinuity equation for electronsÞ

ð21Þ

Semiconductors

Table 1

277

Essential parameters for p–n junction equation set

q n p e ! E k G U ND NA me mh

Electronic charge Density of mobile electrons Density of mobile holes Permittivity of semiconductor Electric field Boltzmann’s constant Carrier generation rate Carrier recombination rate Density of donors Density of acceptors Electron mobility Hole mobility

Source: Reproduced from Carlson EP, Kizilyalli IC, Heidel TD, Cunningham DW. Current topics in electronic devices based on wide band-gap semiconductors for power applications and energy efficiency. ECS Trans 2016;75(12):3–9.

Metal

Semiconductor

Anode

Ohmic contact Cathode

N-type I

x 0

xd

+

− Va

Fig. 11 A metal–semiconductor junction. Adapted from Sze SM, Lee M-K. Semiconductor devices: physics and technology. 3rd ed. New York, NY: John Wiley & Sons, Inc.; 2012 and Principles of Semiconductor Devices. Available from: http://ecee.colorado.edu/Bbart/book/book/chapter3/.

1 dJh ¼ q dx

ðU

GÞ ðContinuity equation for holesÞ

ð22Þ

Poisson’s Eq. (18) correlates the electric field with the net charge density. The electron and hole current Eqs. (19) and (20) represent the sum of the drift and the diffusion currents. The continuity Eqs. (21) and (22) simply state that, for a given volume, the difference between the number of carriers entering the volume and those exiting the volume must equal the difference between carriers being generated within the volume by an external source and those that undergo a recombination process. These last set of equations applies in particular to optical processes, where light (or, more appropriately, photons) contribute to the generation of extra carriers.

2.10.2.7

Metal–Semiconductor Junctions

When a metal (M) and semiconductor (S) are brought into contact, a potential drop occurs at the interface due to the different work functions. Essentially, all the potential drop occurs on the semiconductor side of the device. This can give rise to a depletion region at the interface as in the p–n junction case. The metal acts similarly to very heavily doped semiconductor material from the point of view of its effects on the electrostatic properties of the depletion layer. These M–S contacts (Fig. 11) are called Schottky diodes and have both rectifying and PV properties. These junctions rely on the position of the respective Fermi energies and the vacuum level for electrons. The barrier between the metal and the semiconductor can be identified on an energy band diagram. We first consider the energy band diagram of the metal and the semiconductor, and align them using the same vacuum level (see Figs. 12 and 13). An excellent background on M–S contacts can be found in Chapter 16 in McKelvey [3].

278

Semiconductors

E Evacuum

Evacuum

Semiconductor

Metal

q

q EC EF

qM Metal

Semiconductor

EFM

EC EF

qM qB

Ei

Ei EFM

Ev

(A)

Ev x

(B)

Fig. 12 Energy band diagram of the metal and the semiconductor before (A) and after (B) contact is made. Adapted from Sze SM, Lee M-K. Semiconductor devices: physics and technology. 3rd ed. New York, NY: John Wiley & Sons, Inc.; 2012 and Principles of Semiconductor Devices. Available from: http://ecee.colorado.edu/Bbart/book/book/chapter3/.

E

qi

qB

EC EF Ei Ev xd Fig. 13 Energy band diagram of a metal–semiconductor contact in thermal equilibrium. Adapted from Sze SM, Lee M-K. Semiconductor devices: physics and technology. 3rd ed. New York, NY: John Wiley & Sons, Inc.; 2012 and Principles of Semiconductor Devices. Available from: http:// ecee.colorado.edu/Bbart/book/book/chapter3/.

Similarly to the procedure used for p–n junctions, we first consider the energy band diagram of the metal and the semiconductor, and align them using the same vacuum level as shown in Fig. 12(A). As the metal and semiconductor are brought together, the Fermi energies of the metal and the semiconductor do not change right away. This yields the flat-band diagram of Fig. 12(B). The barrier height, fB, is defined as the potential difference between the Fermi energy of the metal and the band edge where the majority carriers reside. From Fig. 12(B) one finds that for an n-type semiconductor the barrier height is obtained from: fB ¼ FM

χ for an n-type semiconductor

ð23Þ

where FM is the work function of the metal and χ is the electron affinity. For p-type material, the barrier height is given by the difference between the valence band edge and the Fermi energy in the metal: fB ¼

EG þχ q

FM

ð24Þ

Fig. 13 shows the energy band diagram for a metal–semiconductor junction with no bias. Fig. 14 compares an M–S junction with no external bias (Fig. 14(A)) with the forward and reverse bias cases (Fig. 14(B) and (C), respectively). Fig. 14(D) shows the standard I–V characteristics of an M–S junction. The parameters that are indicative of the different conditions represented in Fig. 14 are indicated in the figure as IM-S and IS-M, that is, the currents from the metal side to the semiconductor side, and vice versa. Under no bias (Fig. 14(A), V ¼ 0 in Fig. 14(D)), the two currents have the same magnitude, I0 , also referred to as the leakage current. Under forward bias (Fig. 14(B), V40 in Fig. 14(D)), IM-S does not change, but carrier injection will occur from the semiconductor side to the metal side, and IS-M becomes an exponential function of V, described by the formula IM-S ¼ I0eqV/RT. Finally, under

Semiconductors

IM→S = −I0

IM→S = −I0 −

IS→M = I0 −

− qB

E − EF = qB

IS→M = I0eqV/kT − > IM→S = I0

(A) V = 0. IS→M = IM→S = I0 IM→S = −I0 −

279

IS→M ≈ 0

qB >qB

I

qV EFn V Reverse bias (C) Reverse bias. Metal is negative wrt Si. IS→M > > > < ð RA ÞP rP 2 = 2 ZðjÞ ¼ { 1:0 ð33Þ > > De CAS > > : ; ðjÞc1; Z ¼ 1=ðjÞðstrong pore diffusionÞ; ZðjÞ2 c1:0

Experimental quantities such as ( RA)P, dP, rP, De, and CAS, the surface concentration, and if external mass transfer effects have been reduced, CAb are used. Internal concentration gradients can be large even when internal temperature gradients are negligible; hence, in practice, the second transport resistance limitation not to be overlooked is intraparticle mass transfer.

2.16.2 2.16.2.1

Catalyst Materials and Synthesis Catalyst Materials

Solid catalysts are rather complex materials containing various components that are carefully selected during the design process. Even though catalyst design set out with an Edisonian approach in the mid-20th century by testing of numerous combinations of

Catalysts

491

possible constituents until an optimal composition is found, a systematic theoretical approach gradually developed in recent decades as relationships between the chemical and physical properties of a catalyst and its catalytic performance became better understood [1,2,12]. Catalyst design and development activities are, therefore, highly multidisciplinary and necessitate knowledge of chemistry, chemical physics, material science, and chemical engineering. Catalyst materials, catalyst properties, and the art/science of catalyst preparation are the major elements of the entire process. The chemical, physical and dynamic properties of catalysts are briefly reviewed in Sections 2.16.1.1–2.16.1.3 of this chapter. In the design and development of heterogeneous catalysts, the principal consideration is chemical composition, but other factors of a physical nature may also have impact on catalyst performance; this is true for catalysts used for fundamental work and industrial applications. A heterogeneous catalyst typically contains three major ingredients: 1. An active catalytic phase for carrying out the desired reaction(s), 2. One or more promoter(s) for increasing catalytic activity and stability, 3. A high surface area carrier for providing dispersion and stability to the active phase. Components of typical solid catalysts are listed in Table 2, which includes some classic examples. Active phases composed of metals or metal oxides/sulfides are dispersed within the pores of carrier materials in the form of nanoparticles that are typically 1–50 nm in diameter, and the surfaces of these crystallites contain the sites (involving atoms or clusters of atoms) that are active for catalyzing particular reactions. Most of the metals in the three transition series and in the immediately following groups of the periodic table, along with their oxides, sulfides, nitrides, and carbides are distinctive active phase components because of the mobility of their electrons and the resulting ability to show variable valence states (see Section 2.16.1.1.2). Some of the major active phases and the reactions they catalyze are presented in Table 3. Commercial solid catalysts also include one or more promoter materials that are added during catalyst synthesis in relatively small quantities (usually 0.5–5 wt%). Generally, promoters are not catalytically active by themselves but they contribute to the overall performance of the active catalytic phase by enhancing its activity, stability, and/or accessibility. Promoters often used are of two types, as indicated in Table 2: chemical and textural. Chemical promoters improve the activity or selectivity of the catalytic phase and typically include alkali metals and alkaline earth metals or metal oxides. Electronic promoters are dispersed within the active phase to affect the nature of chemical binding on the active sites or to stabilize the surface atoms in their valence states. For example, alkali metals Na and K are often used as activity modifiers for promoting the dissociative chemisorption of adsorbate molecules on transition metals by electron donation. Alloying Pt with Sn suppresses some side reactions in alkane aromatization as a result of the ensemble effect. Promoters are included to shield the active phase against poisoning; a typical example is the addition of Mo to Co- or Ni-based catalytic phases to increase their resistance to sulfur poisoning. Textural promoters mainly assist in the preparation of well-dispersed active metal phases and also in their upkeep by preventing loss of active surface through Table 2

Components of typical solid catalysts

Component

Material types

Some examples

Active phase

Metals Metal oxides Metal sulfides

Noble metals: e.g., Pt, Pd, Au, Rh, Ru; base metals: for example, Ni, Fe, Co, Cu Transition metal oxides: e.g., MoO2, CuO Transition metal sulfides: MoS2, Ni3S2

Metal oxides Metal oxides Stable, high surface area metal oxides, carbons

Transition metal and Group IIIA: Al2O3, MgO, SiO2, BaO,TiO2, ZrO2 Alkali or alkaline earth: K2O, PbO Group IIIA, alkaline earth and transition metal oxides: for example, Al2O3, MgO, SiO2, zeolites, activated carbon

Promoters

• •

Textural Chemical

Carrier (support)

Source: Adapted from Bartholomew CH, Farrauto RJ. Fundamentals of industrial catalytic processes. 2nd ed. Hoboken, NJ: John Wiley & Sons, Inc.; 2005. Copyright (2006), with permission of John Wiley & Sons.

Table 3

Examples of active catalytic phases and reactions

Active phase

Elements/compounds

Reactions catalyzed

Metals

Fe, Co, Ni, Cu, Ru, Rh, Pd, Ir, Pt, Au

Oxides Sulfides

Oxides of V, Mn, Fe, Cu, Mo, W, Al, Si, Sn, Pb, Bi, rare earth Sulfides of Co, Mo, W, Ni

Carbides

Carbides of Fe, Mo, W

Hydrogenation, dehydrogenation, steam reforming, hydrocarbon reforming, synthesis (NH3, Fischer–Tropsch), oxidations Total and partial oxidation of hydrocarbons and of CO, methanol synthesis, acidcatalyzed reactions (e.g., isomerization, alkylation, cracking) Hydrotreating (hydrodesulfurization (HDS), hydrodenitrogenation, hydrodeoxygenation) hydrogenation Hydrogenation, FT synthesis

Source: Adapted from Bartholomew CH, Farrauto RJ. Fundamentals of industrial catalytic processes. 2nd ed. Hoboken, NJ: John Wiley & Sons, Inc.; 2005. Copyright (2006), with permission of John Wiley & Sons.

492

Catalysts

sintering; these promoters may also assist in preparing high surface area active phases and/or support materials. Promoters and their functions are discussed in greater detail by Hagen [8], and some examples of promoters used in industrial practice are presented in Table 4. Carriers (supports) are porous, high surface area metal oxides or carbons with ample pore volume, and their major function is to enable the preparation of well-dispersed active phases and also to improve their thermal stability under reaction conditions over long periods of time. More commonly used catalyst carriers and their characteristic physical properties are given in Table 5. Among these carriers, aluminas find extensive use in industrial practice because of their high thermal stability besides the wide spectrum of chemical and physical properties offered by nearly a dozen amorphous or crystalline alumina structures formed, depending on the preparation, dehydration, and thermal pretreatment procedures employed [2]. High surface area silica (SiO2) is used as catalyst support in particular commercial applications, although its use is limited because of its lower thermal stability in reactive environments and its forming/pelleting problems. Silica-supported V2O5 is used in the oxidation of SO2 to SO3. More recently, aerogels are produced from alcohol silica gels and formed into high-porosity, low-density monolithic structures, which may expand the usage of silica as carrier in environmental applications because of its resistance to the formation of surface sulfate species that can cause catalyst deactivation, as in diesel converters. Foremost usage of silica is in acidic solid catalysts (SiO2–Al2O3) as the active catalytic phase. SiO2 is not acidic by itself, but when its matrix is introduced with Al2O3, the SiO2–Al2O3 formed acquires acidic properties arising from the interaction of Si and Al atoms (see Section 2.16.1.1.2). Carbon carriers are available in a wide range of porosities, purities, and cost; they are mostly produced from natural resources such as trees and plants, wood and coal, or nut shells by high-temperature pyrolysis; therefore, their chemical and physical characteristics vary widely with the source and pyrolysis conditions. For instance, treatment at very high temperatures (up to 3300K) in an inert atmosphere results in lower surface area graphitic carbons, while pyrolysis at moderately high temperatures, 1073–1773K, produces carbon materials with medium porosity, which is enhanced by further treatment to obtain activated carbons with high surface areas of 1000–1500 m2 g 1 and micropores of 0.5–1 nm. Activated carbons have high thermal stability in reducing environments and provide large surface areas for high dispersion of metal crystallites; hence, they are used as carriers for precious metal catalysts employed in the fine chemicals industry. Some less frequently used but important catalyst supports include magnesia (MgO), titania (TiO2), aluminosilicates or zeolites, and calcium aluminate. In general, commercial high surface area TiO2 contains two crystal phases, anatase and rutile, and anatase with a much higher surface area of 50–80 m2 g 1 is the preferred phase as catalyst carrier. The rutile phase forms at 873K, and transformation of anatase to rutile results in drastic loss of surface area with subsequent deactivation. Catalyst carriers, their properties, and functions are further discussed in Section 2.16.2.3, and reported in more detail in Ref. [2] and in Ref. [13].

Table 4

Examples of promoters in the chemical industry

Catalyst (use)

Promoters

Function

Al2O3 (support/catalyst) SiO2/Al2O3 (cat cracking) Pt/Al2O3 (cat reforming) MoO3/Al2O3 (hydrotreating HDS, HDN) Cu/ZnO/Al2O3 (low-temperature conversion) Fe3O4 (NH3 synthesis) Ni/ceramic carrier (steam reforming) Ag (ethylene oxide synthesis)

SiO2, ZrO2, P; K2O; MgO Pt Re Ni, Co; P, B ZnO

Increase thermal stability; poisons coking sites; slows sintering Increases CO oxidation Lowers hydrogenolysis activity and sintering Increases hydrogenolysis of C-S and C-N bonds in HDS and HDN; increases MoO3 dispersion Decreases Cu sintering

K2O; Al2O3 K

Electron donor, favors N2 dissoc; textural promoter Improves coke removal

Alkali metals

Increases selectivity, hinders crystal growth, stabilizes oxidation states

Source: Adapted from Hagen J. Industrial catalysis: a practical approach. 3rd ed. Weinheim: Wiley-VCH; 2015. Copyright (2015), with permission of John Wiley & Sons).

Table 5

Typical properties of common support materials

Support material

Sg (m2/g)

Vg (cm3/g)

Pore diameter (nm)

Activated carbon Zeolites (molecular sieves) Silica gels Activated clays Activated Al2O3 Kieselguhr (Celite 296)

500–1500 500–1000 200–600 150–225 100–300 4.2

0.6–0.8 0.5–0.8 0.40 0.4–0.52 0.4–0.5 1.14

0.6–2 0.4–1.8 3–20 20 6–40 2200

Source: Adapted from Bartholomew CH, Farrauto RJ. Fundamentals of industrial catalytic processes. 2nd ed. Hoboken, NJ: John Wiley & Sons, Inc.; 2005. Copyright (2006), with permission of John Wiley & Sons.

Catalysts 2.16.2.2

493

Catalyst Synthesis

The form in which a solid catalyst is produced depends entirely on the purpose for its use and on the type of system it will eventually be used in. It is crucial to make a distinction between catalysts that are used for fundamental research and those that are used on industrial scale. For basic research, it is imperative that the catalyst be of well-defined structure and of known composition; also, its surface must be clean or easily be transformed into a clean state. In the case of metal catalysts, for instance, research is conducted with wires degassed by electrical heating or with foils cleaned by ion bombardment, or with specifically grown single crystals; the metal is obtained in a reproducible and clean state in each of these cases, and powerful characterization techniques may be used to study both the solid surface and the adsorbed species. Suitability for basic work is hardly a measure for practical catalysts, which often have several chemical components and diverse physical structures. For industrial applications, the type of system in which the catalyst will be employed is also important, since the preferred physical form of the solid catalyst is determined by whether the reactants are in the gaseous or liquid state, and by whether the catalyst will be required to work in a packed or fluidized bed or in a slurry reactor [1,8,9]. For industrial catalysts, although chemical composition is the main concern, other factors that are related to physical characteristics are also influential. The important physical properties of solid catalysts are discussed in Section 2.16.1.2. High specific surface area, Sg, is required for high activity per unit reactor volume as heterogeneous catalysis is a surface phenomenon. Hence, the majority of solid catalysts are intentionally made to be porous. Intricate porous structures of macroparticulate catalysts involve significant internal diffusion resistances that affect overall reaction rates and selectivities, and ultimately catalytic reactor design (see Section 2.16.1.3.2.2). A second important factor is catalyst stability, including the ability to withstand high temperatures, poisons, fluctuations in process conditions, reactants or products such as water vapor, or regeneration conditions and cycles. Catalyst lifetime is usually critical for the economics of a process. Stable catalysts change very gradually in time under usual operation and cyclic regeneration conditions. Catalyst stability is affected by many factors (see Section 2.16.2.4). In catalyst design and development priorities are given as: selectivity4stability4activity. Thirdly, mechanical properties such as hardness, resistance to crushing, abrasion, and attrition are of particular concern in reactor systems operating at high pressures and temperatures with high throughput. In regular operation, neither a typical catalyst bed nor the individual catalyst pellets in the bed are subject to harsh stresses arising from the weight of the catalyst or the pressure drop through the bed; but, the thermal shocks or depressurizing shocks that may occur during shut-downs or sudden operational upsets may create extremely severe strains, and the catalytic framework should be able to resist these effects. The fundamentals of mechanical grain properties, short descriptions of attrition and crushing tests, and correlations between tests for mechanical resistance are presented in some detail in Ref. [1]. The process of manufacturing catalysts involves (1) recognition of the chemical and physical properties that are of utmost importance in a particular process, followed by (2) selection and development of suitable preparative methods for attaining these properties economically on a large scale. The processes by which the majority of industrial catalysts are manufactured can be grouped into three broad categories according to their products [2,8,28]: 1. Bulk catalysts and carriers/supports, 2. Impregnated catalysts using preformed carriers, 3. Mixed-agglomerated catalysts. Preparation methods used for industrial solid catalysts are summarized in Fig. 5. Bulk catalysts consist mainly of active phases; some typical examples are SiO2–Al2O3 for catalytic cracking of naphtha, CuO–Cr2O3 for water-gas-shift (WGS) reaction, and Fe3O4 doped with Al2O3–K2O for ammonia synthesis. In impregnated catalysts, the carrier or support material gives texture, physical form, and mechanical resistance; active phases are infused onto the carrier material, which may be inert or active. Mixedagglomerated catalysts constitute a hybrid category and can often be classified as either bulk or impregnated. The method of catalyst preparation can change with the choice of base (starting) materials; and even for a given selection of base materials, several ways of preparing the catalyst can be considered. A representative example is the industrial hydrodesulfurization (HDS) catalyst that contains alumina (Al2O3), cobalt sulfide, and molybdenum sulfide (Co9S8 and MoS2 respectively). Commercially available base materials that can be used for alumina component are aluminum nitrate, sulfate, metallic Al, sodium aluminate, or various hydrates of alumina; cobalt may be used in the form of a nitrate, sulfate, acetate, carbonate, or an amine complex, while molybdenum may be included as molybdic anhydride or ammonium paramolybdate. The sulfide forms of Co and Mo are obtained by the in situ H2S pretreatment of their oxides in the industrial reactor during plant start-up. In general, the starting precursor materials are decided first, and then the best method of manufacture is selected accordingly. The production of heterogeneous catalysts consists of numerous physical and chemical steps, and the conditions in each step have a decisive influence on almost all catalyst properties. Therefore, catalysts must be manufactured under precisely defined and carefully controlled conditions as described in protocols developed and perfected over years of practice. A general flowchart of specific processes and unit operations involved in the preparation of industrial solid catalysts is presented in Fig. 6. The two key processes used for catalyst preparation both for catalyst testing in the laboratory and for economically feasible catalyst manufacture on a large industrial scale are: 1. Precipitation 2. Impregnation

494

Catalysts

Bulk catalysts

Impregnated catalysts

Precipitation (silica/alumina)

Wet impregnation (automotive exhaust catalysts)

Hydrothermal synthesis (zeolites)

Incipient wetness (Pt/Sn/Al2O3)

Fusion /alloy leaching (mixed oxides, raney metals)

Vacuum pore impregnation (Bi/Pb/SiO2)

Sol-gel synthesis (mixed oxides, supports)

Ion-exchange (acidic zeolites)

Flame hydrolysis (fumed oxides, supports)

Anchoring/grafting (supported tm-complexes)

Fig. 5 Preparation methods for industrial solid catalysts. Reproduced from Rothenberg G. Catalysis: concepts and green applications. Weinheim: Wiley-VCH; 2008. Copyright (2008), with permission from John Wiley & Sons.

Precipitation Gelation

Support preparation/ precursor treatment

Fusion/ alloy leaching

Impregnation Active phase preparation

Grafting/ anchoring

Filtration Drying

Ion-exchange

Post-treatment

Calcination Extrusion Forming Pelleting

Activation

Fig. 6 Specific unit operations involved in the preparation of industrial solid catalysts. Reproduced from Rothenberg G. Catalysis: concepts and green applications. Weinheim: Wiley-VCH; 2008. Copyright (2008), with permission from John Wiley & Sons.

Precipitation involves the mixing of two or more solutions or suspensions of base materials, initiating the formation of a solid precipitate, followed by filtration, washing, drying, forming, and heat pretreatment; the latter is mostly conducted to provide homogeneity and compound formation but may, nevertheless, cause some sintering and loss of surface area. Impregnation of porous carrier materials with solutions of active components, on the other hand, is one of the best known, easiest, and

Catalysts

495

most widely applied methods for producing practical solid catalysts. In this process, preformed and evacuated carrier particles are submerged in a solution of active phase components with thermally unstable anions such as nitrates, acetates, carbonates, and hydroxides under well-defined conditions, followed by drying, calcination, and, for metallic catalysts, in situ reduction in the reactor before start-up. The impregnation technique is simpler since filtering and forming operations are totally eliminated, and the shape and size of catalyst particles are the same as those of the support [2,7–9,28].

2.16.2.2.1

Bulk catalysts and carrier materials

The preferred preparation method for bulk catalysts is precipitation, which is used primarily for the production of oxide and mixed-oxide catalysts as well as carrier or support materials, most of which are oxides. Precipitation should be carried out meticulously using carefully selected base materials, since the precipitate is the precursor of the final catalyst or of the final support. It produces either lyophobic crystallized precipitates or lyophilic precipitates in the form of gels, which are either amorphous or loosely organized as hydrogels and flocculates. Bulk catalysts are often called precipitated catalysts, although other methods are also used in their manufacture, such as fusion/alloy leaching, hydrothermal synthesis, and flame hydrolysis as listed in Fig. 5. 2.16.2.2.1.1 Precipitation and coprecipitation Precipitation involves the mixing of two or more solutions of active materials to set off precipitation or coprecipitation for obtaining a solid from a liquid solution. In initial stages of the usual precipitation procedure, an aqueous metal salt solution is contacted with an aqueous alkali, ammonium hydroxide or carbonate, to start the precipitation of an insoluble metal hydroxide or carbonate, which is easily converted to oxides by heating. The base materials are normally selected according to their availability and high water solubility, and sometimes to avoid elements that may afterwards be harmful to the final catalyst. For instance, halogens, mostly chlorine, persisting in the final catalyst impart an undesirable acidity to the catalyst, or leftover sodium compounds may increase sintering. If a support is to be included in the final catalyst, precipitation is generally conducted in the presence of a suspension of the calcined solid support or of a compound that will finally be converted to the support, as in the case of soluble aluminum salts that are converted to aluminum hydroxide during precipitation and eventually to alumina. The latter is not the best method since using a soluble aluminum salt as reactant increases the likelihood of a chemical reaction between the active phase precursors and the carrier material [1,9,28]. For several catalytically relevant materials, particularly for carrier materials, precipitation is the preferred method of preparation; these include primarily silicon and aluminum oxides. Precipitation techniques are also used for manufacturing iron, titanium, and zirconium oxides. Preparation of catalysts and carriers by precipitation is more difficult and costly than several other preparation methods because it requires a solvent and a precipitating agent, necessitating product separation arrangements and large volumes of salt-containing solutions in the process. The major advantage of precipitation is the prospect of producing very pure materials as well as the flexibility of the process to final product quality. Multicomponent catalysts containing two or more active components can be prepared by coprecipitation. Two major examples of supported catalysts that are typically prepared by coprecipitation are Ni/Al2O3 hydrogenation catalysts with high Ni loadings and Cu/ZnO/Al2O3 methanol synthesis catalysts. Crystallized precipitates are made up of organized particles with minor porosity, hence very small specific surface, and often contain attached molecules like H2O and NH3. An example for pure carrier materials is the formation of hydrated aluminum oxide (Al2O3
3H2O) crystals in clusters. Coprecipitation is very appropriate for obtaining a homogeneous distribution of catalyst components, because it can give stoichiometric mixtures of very small and intimately mixed crystallites of active components after a calcination or reduction step. Some examples are cobalt molybdate (CoMoO4, H2O, NH3) for hydrotreatment of petroleum fractions, vanadium (IV) pyrophosphate (VO)2P2O7 for n-butane oxidation to maleic anhydride, and CuO–ZnO–Al2O3 with Cu, Zn, and Al from their nitrates for methanol synthesis, and also for industrial scale low-temperature WGS conversion. Highly homogeneous catalysts are prepared from mixed salts or mixed crystals as starting materials, since the ions are already there; easily decomposable anions like carbonate, oxalate, or formate are used. Examples are Ni6Al2(OH)16CO3
4H2O decomposing to give supported Ni/Al2O3 catalyst or Cu(OH)NH4CrO4 as precursor to versatile copper chromite, CuCr2O4, catalyst for hydrogenation and dehydrogenation, oxidation, alkylation, or cyclization [8,29]. 2.16.2.2.1.2 Sol–gel synthesis Precursor products generated by precipitation may be lyophilic precipitates in the form of gels, which are either amorphous or loosely arranged as hydrogels and flocculates. The latter result from a sol composed of micelles that stay dispersed due to electrical charges on their surfaces and in the surrounding solution. A sol, which is a liquid suspension of solid particles ranging in size from 1 nm to 1 mm, can be produced by the hydrolysis and partial condensation of an inorganic salt or a metal alkoxide precursor such as Si(OR)4 where R is an alkyl group. The further condensation of sol particles into a three-dimensional network produces a hydrogel. The four key steps in the sol–gel process are formation of a hydrogel, aging of the gel, drying of the gel for removal of solvent, and heat treatment to obtain different catalytically useful product forms like powder, thin films, membranes, or monoliths. The sol–gel method offers a flexible preparation procedure where parameter selection in preparation steps and in drying is important in tailoring the porosity of the final product. For instance, supercritical drying forms aerogels with high surface area and a broad mesopore size distribution of 2–10 nm. Slow drying at ambient conditions produces xerogels, microporous materials with a narrow pore size distribution similar to that of zeolites [8,13,28]. Carriers such as silica and alumina or bulk catalysts can be produced by the sol–gel or gelation technique. Relatively pure silica is produced on a large scale, and a first production route is by the gelation of aqueous solutions. Natural sodium silicates as base

496

Catalysts

materials are dissolved in an alkaline solution to form a variety of oligomeric species and cyclic structures, which are soluble at high pH. Lowering the pH by acid addition or by solvent evaporation leads to the condensation of hydroxylated species and formation of Si–O–Si bonds. Larger silica particles are gradually formed with increasing condensation (a sol), followed by gelation. On the other hand, very pure and more expensive silica gels can be generated by metal–organic starting materials such as tetra-ethoxy-silane, Si(OC2H5)4. A hydrolysis step followed by condensation yields a final SiO2 product that does not contain the organic group removed in the hydrolysis step as an alcohol. However, hybrid materials can also be prepared by including the organic groups in the final product; these materials are hydrophobic and more flexible but thermally less stable than pure SiO2. After drying, the gel has a surface area of 100–800 m2 g 1 with primary particle sizes of 2–30 nm and a porosity of B65%. Supercritical drying produces aerogels with porosities up to 98% [13]. 2.16.2.2.1.3 Hydrothermal synthesis Hydrothermal treatments implicate heating of precipitates, gels, or flocculates in the presence of water and are usually conducted at 373–573K in an autoclave to obtain structural and/or textural changes including crystal growth, changes in particle size and crystal structure, or conversion of amorphous solids into crystalline form [7]. The foremost application of the method is the manufacture of aluminosilicate zeolites, synthesized by mixing a silica source such as sodium silicate, an alumina source such as sodium aluminate or aluminum sulfate, and an organic salt acting as template such as alkyl ammonium salts for directing the growth of zeolite crystals. Control parameters are pH of the solution, the molar ratios of (Si/Al) and (OH /SiO2), temperature, silica and alumina sources, nature of organic templates, order of reactant addition, aging and ripening before crystallization, synthesis time, special additives, stirring rate, and so on [8]. When the gel is prepared, it is placed in the autoclave for nucleation and crystallization; the zeolite crystals formed are then filtered, washed, dried, and calcined at 773–873K for several hours. Zeolites are also called molecular sieves because of their micropore structure with nanometric channels and cavities, resulting in shape selectivity. Zeolite structures can also be modified after synthesis by ion exchange, the Si/Al ratio can be changed, and other atoms like B, Ga, Ge, Be, Li, Mg, Mn, Co, Zn, Fe, and Ti can be incorporated into the zeolite framework [2]. There is also interest in mesoporous zeolites for reactions with larger molecules. Examples of mesoporous molecular sieves are MCM-22 used in ethylbenzene production from ethylene and benzene, and MCM-41 proposed for Ti-supported epoxidation of fatty acids [8]. Application of the hydrothermal synthesis method for the preparation of other catalysts such as copper chromite has also been reported [29]. 2.16.2.2.1.4 Fusion and alloy leaching A relatively small number of catalysts such as metallic alloys and some mixed oxides belong to this group of unsupported bulk metal catalysts that are prepared by comelting and fusion of the metal or oxide precursors at high temperatures. Melting facilitates intimate mixing of the precursor atoms and clusters, creating pure and well-dispersed solids [7,8,28]. One example of metallic alloy catalysts is the Pt–Rh gauze catalyst used in ammonia oxidation. Other major applications are the iron oxide catalyst used for ammonia synthesis, and the fused V2O5–K2S2O7 catalyst used for sulfuric acid synthesis by SO2 oxidation. Preparation of metallic glasses is another option, and reactions investigated on various glassy metal alloys have been reviewed in Ref. [28]. The drawback of the melting/fusion alternative is that it is an energy-intensive process compared to the precipitation–calcination protocols and it requires special equipment [7]. The preparation of skeletal or sponge metal catalysts is achieved by alloy leaching, and the first example is Raney nickel obtained by leaching a 50 wt%Ni–50 wt%Al alloy in aqueous NaOH, which forms the basis for the skeletal Ni catalysts currently in use. Other skeletal metal catalysts include Fe, Co, Cu, Pt, Ru, and Pd. The two most commonly used skeletal catalysts are unpromoted or promoted Ni and Cu used in hydrogenation, ammonolysis, and reductive alkylation. These catalysts have high metal surface area and are ready to use, necessitating no reduction or activation. Raney metals are stable in alkaline liquids; however, they are highly pyrophoric and hence are kept immersed in water [7,28]. 2.16.2.2.1.5 Flame hydrolysis The term flame hydrolysis refers to a process where a gaseous mixture of a precursor material (usually a metal chloride such as SiCl4, AlCl3, TiCl4), hydrogen, and air are reacted in a flame located in a continuously operated flame reactor. The precursor is hydrolyzed by the steam produced in the oxyhydrogen reaction and the corresponding oxides are precipitated. Flame hydrolysis is used for preparing high surface area oxides on a large industrial scale. For example, flame hydrolysis of TiCl4 produces anatase with a specific surface of 40–80 m2 g 1, which is the catalytically relevant phase. In the case of more frequently used catalyst carriers like silica and alumina, flame hydrolysis produces highly aggregated nonporous structures with high external surface area and good thermal stability. Control parameters are flame temperature, gas and liquid feed rates, nature and concentration of precursor and solvent, and solvent-to-fuel ratio. Hence, exact control of crystallinity and particle size is hard to achieve [8,28]. Since the establishment of flame hydrolysis as a method for the production of SiO2, any mixed oxide that is a combination of SiO2, Al2O3, TiO2, and ZrO2 can be produced as well. These are available on a commercial scale, including Al2O3/SiO2 mixed oxides. Among these, pyrogenic or fumed silica is undoubtedly the most commonly produced oxide. SiO2 produced by flame hydrolysis is a very pure material defined as an aerosol with very fine primary clusters (B1 nm) that are densely packed to form nanometer-size particles (10–30 nm); specific surface areas of 50–400 m2 g 1can be obtained by varying control parameters such as flame temperature, H2/O2 ratio, and residence time in the reactor. Silica particles are separated from the byproduct HCl in a

Catalysts

497

cyclone, and the acid is removed by steam/air treatment in a fluidized-bed reactor. Amorphous silica supports are weak acids with thermal stabilities lower than alumina [8,13,28]. 2.16.2.2.1.6 Unit operations in bulk catalyst preparation Preparation procedures for bulk catalysts and carriers, particularly by precipitation techniques, require solvents and precipitating agents, introducing product separation and solvent removal steps into the manufacturing process [1,7,28]. Filtration and centrifugation: the unit operations of decantation, filtration, and centrifugation are necessary for separating the solid product from the liquid phase, which is relatively easier for crystallized precipitates, problematic but achievable for flocculated gels, and of no use for hydrogels where water is confined in the solid framework. Washing: this unit operation has to accomplish three major objectives: (1) to replace the solution left in pores with pure water in order to remove the undesirable ionic or molecular solutes, (2) to desorb by dilution some ions and adsorbed molecules lingering on the solid surface or mixed in the solid product after precipitation or ripening, and (3) to exchange undesirable ions with other ions that can easily be decomposed during the calcination step, such as the exchange of Cl or Na þ for NO3 or NH4 þ , respectively. Washing precipitates with small surface areas usually poses no problems, but washing gels with some degree of porosity is much more difficult due to diffusional problems. Drying: this operation is for the elimination of water or solvent from the pores of the solid before or after forming. It is a routine procedure for crystallized precipitates, but critical for gels that may contain up to 90% water; special drying procedures with/without hydrothermal transformation have to be used. Ordinary drying done in conventional ovens at 373–473K produces a dry gel or a xerogel; if there is no hydrothermal transformation, the total volume of the particles remains constant so that the surface changes slightly during drying, but the pore volume decreases and hence density increases. In particulate drying for producing aerogels, capillary forces are excluded by operating at temperatures higher than the critical temperature of water or by evaporating it under vacuum at temperatures of 223–268K, and the highly porous product is an aerogel. Therefore, the important variables of the drying operation are temperature, pressure, water partial pressure, time, size of the particles, and nature of the intermicellar solvent [1]. Calcination: this unit operation may be done before or after the forming operation. It is done in air at temperatures higher than the reaction and regeneration temperatures, 573–1073K. Major aims are (1) to eliminate extraneous materials like binders and lubricants as well as volatile unstable anions and cations not wanted in the final catalyst; (2) to obtain a well-determined structure for active components and supports, with parallel adjustment of the texture regarding the surface and pore volume; and (3) to get good mechanical resistance for the catalyst, if it does not already exist. Calcination procedures can result in different textural modifications for crystalline precipitates and gels [1,7]. Forming operations: the order in which calcination and forming operations are conducted changes for each case. If calcination leads to significant textural alteration, there is no point in prior forming as in the case of crystallized precursors, which are recalcined after forming to increase their mechanical resistance. In certain bead-like xerogels, calcination can precede forming to improve their mechanical properties. Forming microgranules or particulates on the order of millimeters correspond to two extreme types of forming. For microgranules, crushing and grinding are done to prepare the catalyst charge for forming, using cyclonepulverizers, ball-crushers, or mixer-grinders. Spray-drying is used to achieve simultaneous forming and drying and is appropriate for forming small-diameter beads in the range of 7–700 mm; this is the technique used for forming the SiO2–Al2O3 microbeads for catalytic cracking. Another technique used for obtaining microbeads from metastable sols suspended in a liquid is drop-coagulation, which accomplishes simultaneous gelation, ripening, and forming; key control variables are sol composition, pH, temperature, contact time, and properties of the liquid phase including surface tension and density. Catalyst particulates, on the other hand, are used in packed, slurry, and fluidized-bed reactors, and include pellets, pills, rings, extrudates, spheres, granules, and powders. Forming operations for particulates require a standardized powder or a paste with rheological properties suitable for pelletizing, extrusion, and granulation; whatever the forming method, the charge must have some degree of fluidity and adhesiveness. Pellets are formed by compressing dry catalyst or carrier powder in cylindrical dies at high pressures (100–4000 atm), and fluidity of the powder ensures homogeneous filling of the die; a degree of plasticity is necessary for maximum contact between grain boundaries for producing a high-strength pellet of moderate porosity. The ability of a catalyst or carrier to form durable pellets depends on its tensile strength, mesoporosity, and moisture content. If the powder does not have the necessary properties, small amounts of liquid lubricants like water or mineral oil or plasticizers like talc, graphite, or stearates can be added, which may affect catalyst activity. Some of the organic additives may partially burn off during calcination. In extrusion, the powder is first made into a paste, wet powder, or heavy slurry in a hopper and forced through the holes in an end plate, and as the solid ribbon emerges from the hole it is partially dried to keep its shape, cut into chosen lengths by rotating knives, and is sent to the drier on a conveyor belt. The extrudates thus formed have lower strength, higher porosity, and less regular shapes than pellets, but the process is substantially less expensive with higher rates of production. The ease of extrusion and final product quality depend on paste properties such as viscosity, thixotropy, stability, and homogeneity [1,2,7].

2.16.2.2.2

Impregnated catalysts

Catalysts impregnated on preformed support materials are the largest group of heterogeneous catalysts and have economic significance in the chemical industry and especially in refinery technology. Impregnated catalysts are often called supported catalysts, and the methods used in their manufacture are listed in Fig. 5. The procedure of infusing a solid carrier with a precursor solution gives this technique its special character, hence catalyst manufacture by impregnation associates all other relevant unit

498

Catalysts

operations with achieving good dispersion of the active catalytic phase on a carrier that may be inert or catalytically active (as in the case of bifunctional catalysts). Other operations like drying, washing, calcination, and activation are governed by the same principles, depend on similar parameters, and use similar equipment as those mentioned above for bulk catalysts (Fig. 6). The active components are not introduced into the carrier in their final form but as a solution containing the precursor compounds; the choice of base materials is critical for the quality of the final deposit on the carrier surface in terms of structure, grain size, and dispersion. Two types of impregnation may be discerned: impregnation with no catalyst–support interaction and with catalyst–support interaction [1,2,8,9]. 2.16.2.2.2.1 Impregnation with no catalyst–support interaction If the support is inert, the sole function of the carrier material is to give the final catalyst its form, texture, and mechanical strength. Selection of starting materials is vital; they must be adequately soluble in the impregnating solution and also give the best result on the final catalyst. For instance, Ni deposited on SiO2 from the formate, Ni(NH3)4(HCOO)2, created a higher metal surface area than the Ni from the nitrate Ni(NO3)2 [1]. Wetting consists of contacting the carrier with the precursor solution so that it penetrates the pores of the carrier by the influence of capillary forces. This process may be slowed down by the air trapped in the pore structure, but bubbles are released rather rapidly, and theoretically, the solute concentration is uniform in the entire pore volume. Oxide supports such as alumina and silica are quickly wetted by aqueous solutions, as are most activated carbons that have a layer of chemisorbed oxygen on their surface. Capillary forces ensure that the precursor solution is sucked into the pores. Two methods of contacting may be considered for the wetting of the support material: dipping into an excess amount of solution, or impregnation to incipient wetness (dry impregnation). In the former method, the uptake is equal to the sum of the solution trapped in the pore volume and that adsorbed on the pore surfaces. If two or more components are present in the solution, they may be adsorbed on the carrier surface in a ratio different from that in the solution whose concentration also changes with prolonged contacting; in this case, the absence of catalyst–carrier interaction must definitely be verified. Impregnation to incipient wetness is more commonly used industrially as it allows exact control of the quantity of active component incorporated into the catalyst. A batch of the preformed and evacuated support is tumbled and sprayed with a precursor solution of correct concentration equivalent in quantity to the known pore volume of the support. The resulting catalyst is normally dried and calcined. For metallic catalysts, reduction is conducted in the reactor before start-up. In some applications, deposition-precipitation is used whereby the active component is fixed inside the carrier by immersing the impregnated catalyst in a suitable reagent for pH adjustment in order to precipitate the catalyst precursor in the form of a hydroxide or carbonate within the pores and on the surface of the carrier. 2.16.2.2.2.2 Impregnation with catalyst–support interaction If the active component or solute to be impregnated forms a bond with the carrier surface as wetting is done, this interaction can be an ion exchange, an adsorption, or a chemical reaction. Adsorption or ion exchange technique implicates the soaking of predried catalyst particles in a suitable metal salt solution at 298–353K for a time from minutes to hours. The metal salt chosen for generating the ions of the catalytic agent must be compatible with the surface charge of the carrier. The carrier containing ion X þ is dropped into a volume of solution (in excess of the pore volume) containing ion Y þ , which gradually penetrates into the pore structure and replaces ion X þ , pushing it into the solution until an equilibrium is reached. Almost all solid mineral carriers are oxides and they act as ion exchangers when their surface bears electric charges. The two categories are natural exchangers and amphoteric oxides. Natural exchangers have a framework carrying electric charges neutralized by ions of the opposite sign. For zeolites, for instance, these charges are negative due to the environment of aluminum. The AlO4 tetrahedron is an overall bearer of a negative charge distributed over the oxygen atoms, and this charge is neutralized by the presence of different cations like Na þ or K þ , etc. Zeolites are cation exchangers and have a definite number of exchange sites equal to the number of Al atoms in their framework. There are other natural ion exchangers, such as clays and silicates, which are cation exchangers while hydrotalcites are anion exchangers. In any of these natural exchangers, the number of exchange sites is not dependent on pH. In the second category are oxide surfaces that are covered with hydroxyl groups when contacted with water; they can be represented as S–OH where S stands for Al, Si, Ti, Fe, etc. Some of these groups may behave like Brønsted acids while other hydroxyl groups may behave like Brønsted bases. The surface charge is a function of the solution pH in this case. Support surfaces in contact with metal salt solutions adsorb equilibrium amounts of either cations or anions on their active sites having either protons or hydroxyl groups. The amount of adsorbed metal is determined by the number of adsorption sites available on the support, the concentration of ions in solution, pH, and the equilibrium constant for adsorption strength. Silicon adsorbs cations weakly, alumina adsorbs both anions and cations weakly, carbon adsorbs cations weakly, and magnesia adsorbs anions strongly [1,7,8,28]. The term anchoring immobilization is used for the attachment of a central metal atom of a transition metal complex to a solid support. The catalytic active components may be attached to appropriate inorganic or organic carriers like silicones, oxides, polystyrenes, or styrene copolymers or, in special cases, anchored inside the channels of nanosized supports such as zeolites. Grafting, on the other hand, is the technique of equilibrium adsorption of an active catalytic species (metal or precursor) by covalent bonding to a solid support from a solution in which the support is suspended. Silica is the most widely used support, and its surface is composed of siloxane bridges (Si–O–Si) and silanol groups (Si–OH). Starting from silanol groups, it is possible to carry out stoichiometric reactions between molecular organometallic compounds and the support. This is done by grafting a catalytically active complex or a molecular precursor with a reactive ligand to give the related monosiloxy complex and the ligand

Catalysts

499

as byproduct. Mesoporous Ti–SiO2 catalysts with a nonordered pore system were created as very active catalysts in epoxidation of fatty compounds. Titanium active centers are attached to the silica support by grafting titanocene dichloride precursors, thus localizing all Ti sites on the surface to give catalysts that are both mechanically strong and resistant to leaching [8].

2.16.2.3

Porous Materials

Materials characterized by porous structures are heavily involved in many fields of chemical engineering such as heterogeneous catalysis and separations. In the former, porous materials are mainly used as catalysts or catalyst supports. They also form the basis of membranes and adsorbents, both of which are used in separation/purification of gaseous and liquid mixtures. This section is focused on the use of porous materials in catalysis. Readers are directed to [30] for an in-depth discussion of membrane separation and adsorption. Pore size, which is also known as pore width or diameter, is defined as the gap between two opposite walls of a pore, and is an important parameter in defining porous materials. In terms of their sizes, pores are grouped as micro-, meso-, and macropores, which refer to pore sizes below 2 nm (i.e., 2  10 9 m), between 2 and 50 nm, and above 50 nm, respectively (see Section 2.16.1.2.4.3). In micropores, pore size is comparable to that of the molecules and mechanism of activated transport, i.e., micropore diffusion or molecular sieving, dictates the motion of gas molecules within the pores [31]. In mesopores, however, pore size is either similar to or smaller than the mean free path length, which is defined as the average distance that a molecule travels between collisions. As a result, transport mechanisms within mesopores become characterized by Knudsen diffusion, surface diffusion, and laminar flow [31]. Due to sizes larger than the mean free length of the fluid mixture, transport of the gas molecules in macropores is defined by bulk diffusion and viscous flow mechanisms. Porosity is a key property in heterogeneous catalysis, as it affects reactor performance via diffusion limitations. A descriptive example is low-temperature Fischer–Tropsch (LTFT) synthesis (Section 2.16.3.2.7), in which chain length of the synthesized hydrocarbons is strongly linked with pore diffusion that affects the degree of contact of reactive flow with the active sites. Porosity also dictates the surface area of the catalyst, which positively affects catalytic activity. When it has an ordered pattern, porosity can be used to carry out reactions specific to molecules with certain shape and/or size leading to increased selectivity. Porous materials are classified based on the degree of uniformity of their pores and their dimensions (i.e., sizes). Materials such as aluminum oxide (alumina, Al2O3) have irregular textures and involve pores with a variety of sizes that are subject to a distribution. In heterogeneous catalysis, such materials are employed as catalyst supports. c-Al2O3 is a frequently used support material due to its high surface area (B250 m2 g 1), thermal stability (up to B873K), relatively high bulk density, and low price. In high-temperature applications, a–phase of Al2O3, having the highest thermal stability, is preferred as the support material. This advantage of a-Al2O3, however, comes at the expense of reduction in surface area, which is only B5 m2 g 1. An alternative material, titanium dioxide (titania, TiO2), finds use as a catalyst support due to its high thermal and chemical stability, and mechanical strength, together with the possibility of arranging its porous texture. Strong metal–support interactions that lead to improvement in catalytic activity and stability are also reported for TiO2 supports. Among the phases of TiO2, anatase is more suitable for use as a catalyst support rather than rutile, as metal–support interactions are more favored in the former. Transition from anatase to thermally stable rutile phase is reported to occur above 873K [32,33]. Apart from alumina and titania, silica (SiO2), an alternative porous support material, has a lower bulk density and has the risk of being affected by sintering (Section 2.4.3) at temperatures higher than B900K [34]. Activated carbon is another support material that combines distinct benefits such as high mechanical resistance, chemical and thermal stability, high surface area and porosity, and easy dispersion of the active metal particles with low cost. It is manufactured from materials such as coconut shells via their carbonization and subsequent activation steps. Carbonization involves pyrolytic decomposition of the starting material into elemental carbon atoms that are grouped in the form of stacks of aromatic sheets. The resulting structure is then treated with air, CO2, or steam at temperatures between 1073 and 1173K to obtain the resulting porous, high surface area network. Surface area is somewhat related to the carbon source. It is reported to change between 300–1000 and 700–1500 m2 g 1, when wood and coconut shells are used as carbon precursors, respectively [35]. In contrast with those described above, materials such as zeolites, mesoporous silicas, and metal–organic frameworks and carbon nanotubes (CNTs) are characterized by well-defined, ordered structures. A brief discussion of these materials is provided in Sections 2.16.2.3.1–2.16.2.3.3.

2.16.2.3.1

Zeolites and mesoporous silicas

Zeolites are aluminosilicate, crystalline solids having microporous pore structures. Pore dimensions in zeolites, which vary between 0.3 and 1 nm, are well defined and dictated by their particular crystalline, cage-like structure. Due to their highly ordered porous morphologies, zeolites are also known as molecular sieves. The so-called apertures, which act as sieves, permit smaller molecules to enter the crystal cage, but reject larger molecules. This property, together with their high BET surface areas (e.g., 300–400 m2 g 1 for ZSM-5, Fig. 7(A)), ability to exchange cations, and thermal, chemical, and mechanical stabilities, allow zeolites to be used as shape/size selective catalysts for several commercial processes. Zeolites are also used as adsorbents (e.g., for water and odor removal) and as water softeners in detergents. Mesoporous silica materials involve regular arrangement of mesopores with sizes ranging between 2 nm and 20 nm. In contrast with zeolites, pore walls in mesoporous silicas are amorphous. Due to the fact that they offer high surface areas, which can exceed 1000 m2 g 1, and pore volumes together with narrow pore size distributions, mesoporous silicas such as the ones in the M41S

500

Catalysts

(A)

(B)

Fig. 7 Representative structures of porous nanomaterials: (A) ZSM-5 zeolite and (B) MCM-41 mesoporous silica. Reproduced from Lehman SE, Larsen SC. Zeolite and mesoporous silica nanomaterials: greener syntheses, environmental applications and biological toxicity. Environ Sci – Nano 2014;1:200–13. Copyright (2014), with permission from The Royal Society of Chemistry.

family are used as catalysts, catalyst supports, or adsorbents. MCM-41 (Fig. 7(B)), a member of the M41S family, and SBA-15 are well-known examples of mesoporous silicates. These materials are known to have high stabilities, widening the range of their possible practical applications. It is reported that pure silica M41S can preserve its stability up to 1123K in air or up to 1073K in air with low water vapor pressure. The ordered structure, however, is reported to lose its integrity under mechanical compression or by exposure to water vapor at prolonged periods under room temperature [36]. Detailed descriptions of zeolites and mesoporous silica materials can be found elsewhere [37]. Zeolites can be natural or synthetic, with the latter group being used in catalysis. Manufacturing of synthetic zeolites is typically carried out by the sol–gel hydrothemal synthesis routes in which parameters such as composition, temperature program, pH, rate of agitation, and order of mixing determines the properties of the zeolite. During synthesis, zeolite structure is formed around an organic template, which is then removed by high-temperature calcination to obtain the porous texture. Synthesis of mesoporous silica, however, is somewhat different and involves the use of a surfactant, whose type and concentration are known to affect pore dimension and structure, respectively. Similar to template removal in zeolites, high-temperature calcination of the surfactant leads to the formation of the ordered porous texture. Details of zeolite and mesoporous silica synthesis techniques, as well as their recent versions with reduced impact on health and environment are extensively reviewed in Ref. [38]. Fluidized catalytic cracking (FCC), which involves conversion of heavy gas oil molecules into shorter-chain hydrocarbons such as gasoline and olefins, is the process involving the highest use of zeolite catalysts. It is reported that more than 95% of zeolite catalyst utilization is by FCC [39]. Zeolites are also used in operations such as methanol-to-gasoline conversion, hydroxylation, alkylation, and epoxidation [39]. In addition to their use as support materials in catalysis, mesoporous silicas are also considered as catalysts themselves. They are mainly used in acidic catalysis. Alcohol dehydration and phenol conversion are some examples involving the use of mesoporous catalysts [40].

2.16.2.3.2

Metal–organic frameworks

Metal–organic frameworks (MOFs), also known as porous coordination polymers (PCPs), have received enormous interest in the last decade. This interest is due to the many diverse properties of the materials, and, more importantly, to the possibility of customizing the structure of the material for the objective of interest on a rational basis. Some distinct properties of MOFs are as follows [41]:

• • •

They consist of a metal linker and organic ligands (Fig. 8). The extraordinary variability of combinations and additives make these materials adaptable to almost any need. Their porosity can reach up to 90% of free volume, creating very low-density stable structures. They can have surface areas as high as 6000 m2 g 1 while retaining their structure. In MOFs, surface area is highly specific to the material. For example, MOF-2, MOF-5, and MOF-177 have surface areas of 270 m2 g 1, 2900 m2 g 1, and 4500 m2 g 1, respectively [42,43].

MOFs are characterized with crystal structures composed of pores with well-defined pore dimensions. Depending on the types and relative amounts of metal and ligand, solvent, and the source of the balance anions for metal ions involved in synthesis, pore sizes can be customized in a wide range, which can be less than 20 Å or higher than 500 Å . Controlling the pore size, which tends to increase with surface area, remains a challenge in MOF synthesis. When the pore size range of 2–50 nm is combined with the sharp size/shape selectivity offered by their highly ordered textures and with high surface areas, MOFs become attractive candidates for catalysis applications. They are found to be active and selective for a large variety of mechanisms, ranging from acid–base reactions to redox catalysis. HKUST-1 is a distinct example for

Catalysts

501

Fig. 8 Representative structure of MOF-5. Reproduced from Kaye SS, Dailly A, Yaghi OM, Long JR. Impact of preparation and handling on the hydrogen storage properties of Zn4O(1,4-benzenedicarboxylate)(3) (MOF-5). J Am Chem Soc 2007;129:14176–7. Copyright (2007), with permission from American Chemical Society.

Lewis acid catalysis, such as selective oxidation of alkenes. MIL-100 (Fe, Cr) is reported to be active in Friedel–Crafts benzylation due to their Brønsted acid sites. Polymerization reactions are also highlighted due to the mesoporous structures of MOFs, especially when coupled with shape selectivity. Enantioselective catalysis, usually used in the drug industry, is found to occur on MOFs due to their potential homochiral and photoluminescence properties. In addition to their direct use as catalysts, MOFs are also considered as catalysts supports [44]. MOFs with pore sizes less than 2 nm, however, favor interactions between gas molecules and pore walls, and are therefore preferred for storage and separation of gases [45]. As in the case of catalytic applications, properties such as shape selectivity, molecular sieving, and high porosity have led to the extensive investigation of adsorption of gases such as H2, CH4, and CO2 on MOFs [46–48]. Further studies addressing selective adsorption of gases on MOFs is reviewed elsewhere [49]. Use of MOFs as membranes for gas separation purposes is also studied extensively and reviewed in the literature [50]. Despite their desirable functional properties, MOFs are not as stable as zeolites or mesoporous silicates. Organic nature of the ligands in MOFs reduces their stability against temperature and/or moisture. Moreover, some MOFs cannot exist without a guest solvent residing in their pores [44]. These drawbacks hinder the potential use of MOFs in catalytic applications involving hydrothermal conditions. Nevertheless, these problems are minimized for some MOFs. It is reported that MIL-110, MIL-101, MOF-74, MIL-53(Al), and ZIF-8 are proven to be stable up to 573K with at most 50% steam in the feed streams [51]. For example, MIL-53(Al), having surface area and pore volume of 950 m2 g 1 and 0.32 cm3 g 1, respectively, is reported to be stable against water, several organic solvents, and temperatures up to 673K [52]. Higher thermal and chemical stabilities are demonstrated for ZIF-8 (surface area: 1947 m2 g 1, pore volume: 0.66 cm3 g 1), which is shown to preserve its structure until 823K and in the presence of boiling benzene, methanol, water, and aqueous sodium hydroxide [53].

2.16.2.3.3

Carbon nanotubes

A CNT is a tubular material that can be considered as rolled-up sheets of graphene, a form of carbon involving one-atom-thick planar sheets with atoms configured in a honeycomb-shaped structure. If the tube has only one layer, it is defined as a singlewalled carbon nanotube (SWCNT). A multiwalled carbon nanotube (MWCNT) is a structure defined by the existence of smallerdiameter tubes within larger-diameter ones (Fig. 9) [54]. Depending on the particular configuration, the surface area of CNTs can change between 50 and 500 m2 g 1. CNTs can have diameters in the range of 5–100 nm that are accompanied by very high aspect ratios (up to B108), due to their extremely high mechanical strength. In heterogeneous catalysis, CNTs are mainly considered as novel support materials that are superior to their conventional counterparts in several aspects. In addition to their exceptional mechanical and thermal stability, they have very high thermal conductivity that simply eliminates intraparticle heat transfer resistances. Similarly, mass transport limitations are also eliminated due to lack of any microporosity, leading to effective use of the catalyst. Possibility of customizing metal–support interactions in CNTs is another property leading to increased catalytic activity. Tubular structure of CNTs also allows deposition of the active phase onto the inner surface, leading to so-called confinement effects, which are discussed and reviewed by Serp and Catillejos [55]. The same authors also summarize the benefits of CNT-supported catalysts in applications such as Fischer–Tropsch synthesis, ammonia decomposition, hydrogenation, photocatalysis, and electrocatalysis.

502

Catalysts

(A)

(B)

(C)

0.36 nm

(D) Fig. 9 Representative structures of (A–C) single-wall carbon nanotubes and (D) double-wall carbon nanotubes. Reproduced from Iijima S. Carbon nanotubes: past, present, and future. Phys B – Condens Matter 2002;323:1–5. Copyright (2002), with permission from Elsevier.

2.16.2.4

Deactivation of Supported Solid Catalysts

Catalysts used in industrial operations are prone to deactivation, i.e., loss in their functionality over time, measured in terms of either activity or selectivity, or a combination of both. Owing to the high price of catalyst materials, and its negative and notable impacts on the profitability of chemical processes, the possibility of catalyst deactivation, its severity, and duration should be properly addressed at the design and operation stages of reactor operation. Deactivation of supported solid catalysts can occur due to mechanisms that are either physical or chemical in nature. These mechanisms, namely poisoning, fouling, thermal degradation, and mechanical deactivation, are explained in Sections 2.16.2.4.1–2.16.2.4.4. Depending on the nature of the catalytic reaction, time evolution characteristics of deactivation change in a very wide frame ranging from seconds to years. In FCC, in which longer chain hydrocarbons in gas oil are cracked into shorter hydrocarbons such as gasoline, coke is deposited over the zeolite-based catalysts on the order of seconds [17]. On the other hand, deactivation in iron-based catalysts used in ammonia synthesis (Section 2.16.3.2.3) is reported to be very slow to cause notable deactivation only after 10 years of operation [56]. Deactivation can be managed by proper selection of catalyst, reactor, and process conditions. Optimal strategy is based on understanding possible deactivation types and eliminating their root causes. Regeneration is the process of renewal of the catalyst without permanently removing it from the reactor, and is essential to increase the duration of feasible operation of the catalysts. Its way of implementation depends on the specific requirements of regeneration conditions. In most situations, regulating the operating conditions such as temperature and feed composition is usually sufficient for restoring catalytic activity within the reactor unit. As in the case of FCC, however, regeneration calls for the need of using a separate unit, since its conditions are different from those involved in the reactor. Despite successful regeneration cycles and careful selection of the operating conditions elongating their life, the aged catalysts should be removed from the reactor permanently for recycle or disposal. The break-even of catalyst changeover is specific to the type and economics of the particular process.

Catalysts 2.16.2.4.1

503

Poisoning

The fundamental reason for catalyst poisoning is the interaction between a component existing within the feed or product and the active sites of the catalyst. In this respect, poisoning is a chemical phenomenon, whose severity is positively correlated with time (Fig. 10). The most common type of poisoning is known to be due to sulfur, but also heavy metals or their ions can poison metallic catalysts, and organic bases may poison acidic catalysts. Three types of catalyst poisons have been identified, which are (1) Group 5 A and 6 A elements, including N, P, As, and Sb (5 A) and O, S, Se, and Te (6 A); (2) toxic heavy metals such as Pb, Hg, Cd, and Cu; and (3) specific molecules that interact with catalyst surfaces [57]. In the first type of poisons, the elements poison metal catalysts by interaction through their “s” and “p” orbitals. Hence, the number of bonding electrons can be changed by oxidation or reduction, and the degree of poisoning can be modified. For instance, the poisoning effect of sulfur increases in the order of SO42 oSO2oH2S. It was reported that H2S present at parts per billion levels is sufficient to result in significant surface coverage [58]. Depending on the types of poison, active metal, support material, operating conditions, and duration of contact of poison with the catalyst, poisoning can be reversible or permanent. In the former case, weak chemisorption of the poison helps its removal from the metal surface to restore original functioning of the catalyst. For example, oxidation at high temperatures is proposed to regenerate catalysts poisoned by sulfur, since the sulfates, which can react with supports such as alumina, break up to give SOx that can be removed from the system in the gas phase [57]. Moreover, temperature can affect the adsorption strength of the poison. In catalytic combustion, for example, temperatures in excess of 1273K are reported to prevent adsorption of poisons such as sulfur and halogens on the catalyst surface [56]. In case of permanent deactivation, however, chemisorption occurs strongly, leading to irreversible poisoning of the catalyst bed and requiring its replacement with a fresh batch. In contrast with first type of poisoning, removal of the second type of toxic heavy metal poisons is more complex, and involves processes known as abrasion or leaching, which are extensively reviewed in Ref. [58]. Rather than regeneration of the deactivated catalyst, preventing the contact of the poisons with the catalyst is a more common strategy employed to manage poisoning. In this respect, it is a common processing practice to remove sulfur compounds before they reach the reactor. This process is known as hydrodesulfurization and involves the conversion of sulfur compounds to H2S at temperatures between 623 and 673K on alumina-supported cobalt and molybdenum oxides. The resulting H2S can then be removed by adsorption on a ZnO catalyst [57]. In some cases, standalone guard beds involving ZnO catalysts are often used to minimize H2S slip into the catalytic reactors, such as in the cases of steam reforming (Section 2.16.3.2.1) and LTFT synthesis (Section 2.16.3.2.7). Poisoning of the Pt-based catalysts within the PEM fuel cells by CO can be presented as an example for describing the third type of poisoning. Existence of very small amounts of CO as low as 10 ppm in H2-rich streams is a serious threat for PEM fuel cell operation. At low temperatures, CO molecules adsorb on the active Pt sites so strongly that almost no active sites are left for the dissociation of H2 into its protons and electrons [59]. In this respect, elimination of CO from the H2-rich reformate stream and development of CO-tolerant Pt-based catalysts are being studied extensively as strategies to minimize the adverse effects of CO poisoning in PEM fuel cell operation. In the context of the first approach, CO, which is at a concentration of 1–2% by volume at the exit of the water–gas shift converters, is minimized to concentrations allowable by the PEM fuel cell (o10 ppm) by techniques such as its selective oxidation over Au- or Pt-based catalysts [59,60]. Bimetallic fuel cell catalyst formulations such as Pt/Ru, Pt/Mo, Pt/W, and Pt/Fe are demonstrated to increase CO tolerances up to 100 ppm [61].

A

A

P

B

B

B

P

P

P

Support

P

A

P Support

P

P

P

Support

A

Reactant

B

Product

P

Poison

Fig. 10 Mechanism of poisoning of supported metal catalysts (A: reactant, B: product, P: poison).

504

Catalysts

2.16.2.4.2

Fouling

Fouling is defined as the physical blocking of the active sites due to the deposition of a substance on a catalyst surface. Root causes of fouling can be different. Fine particles that undesirably penetrate into the catalyst bed can accumulate on the surface of the catalyst pellets and block their pores to prevent contact of the active sites with the reactants. Deposition of process dirt or dust that can be produced by mechanical erosion of equipment or by combustion of hydrocarbons are examples of fouling. Despite the fact that the impact of fouling is physical, the substances that lead to fouling are usually generated by means of metal catalyzed reactions. Coke (or carbon), the most commonly known type of fouling agent, is a collective description of various kinds of carbonaceous deposits characterized with high molecular weights and low hydrogen-to-carbon ratios (Fig. 11) [56,57]. It is usually generated in high-temperature catalytic reactions involving hydrocarbons such as catalytic steam reforming of natural gas (Section 2.16.3.2.1) in which coke formation is more pronounced at higher temperatures and at lower quantities of steam via the following reactions [57,62]: CH4 ¼ C þ 2 H2 ; DHro ¼ 74:9 kJ mol 2CO ¼ C þ CO2 ;

DHro ¼

1

172:4 kJ mol

CO þ H2 ¼ C þ H2 O; DHro ¼ Cn H2nþ2 ¼ nC þ ðn þ 1ÞH2 ;

ð34Þ 1

135:6 kJ mol

ð35Þ 1

DHro 40

ð36Þ ð37Þ

The reactions are reversible and it is possible to minimize coke formation by adjusting operating conditions such as increasing the amounts of carbon dioxide and steam that will shift the directions of Reactions (35) (also known as Boudouard reaction) and (36) in the direction of coke minimization. The carbon limit diagrams that present the tendency of carbon formation over nickel catalyst as functions of O/C and H/C are presented in Ref. [63]. Operating temperature defines the significance of the reactions listed above. Reactions (35) and (36) are exothermic, hence favored at lower temperatures. On the other hand cracking reactions ((34) and (37)) are likely to occur at higher temperatures due to their endothermic nature. There are three common types of carbon that have been identified in Ni-catalyzed steam reforming. These are pyrolytic carbon, whisker-like carbon, and encapsulating carbon [64]. Pyrolytic carbon forms mainly as a result of the thermal cracking of hydrocarbons, which is favored at temperatures above B923K [65]. The process is further facilitated at lower steam-to-hydrocarbon ratios, at higher pressures and in the case of high surface acidity of the catalysts. This mode of carbon can also form as a result of higher residence times and of overheated gas films at the tube walls acting as a source of radicals and coke precursors. Therefore, the nature of catalyst packing, i.e., void fraction, affects the carbon formation by influencing residence time and the gas film volume. Pyrolytic carbon both deactivates the Ni-based catalyst by encapsulating it and blocks the reactor by accumulation within the voids between the particles [64]. Suppressing the factors that facilitate it can minimize the formation of pyrolytic carbon. The formation of whisker carbon and encapsulated carbon can be explained by the coking mechanism over nickel [65]. The process is believed to start with the dissociation of hydrocarbons into highly reactive monatomic carbon (Ca), which can easily be gasified to form carbon monoxide. However, if Ca is formed in excessive quantities or if gasification is slow, then polymerization to Cb is facilitated. It has been shown that Cb is much less reactive than Ca and the gasification step is fairly slow [62]. As a result, Cb may either accumulate on the catalyst surface or dissolve in the nickel leading to encapsulated carbon and whisker carbon,

Carbon

Support particle

Metal crystallite

Fig. 11 Schematic presentation of carbon deposition on a supported metal catalyst. Reproduced from Bartholomew CH. Mechanisms of catalyst deactivation. Appl Catal A: Gen 2001;212:17–60. Copyright (2001), with permission from Elsevier.

Catalysts

505

C H2

C

CnHm H H

Metal crystallite

C

Support

Growing filament

Fig. 12 Mechanism of formation of carbon whiskers on a supported metal catalyst. Reproduced from Moulijn JA, van Diepen AE, Kapteijn F. Deactivation and regeneration. In: Ertl G, Knozinger H, Schüth F, Weitkamp J, editors. Handbook of heterogeneous catalysis. Weinheim: Wiley-VCH; 2008. p. 1828–46. Copyright (2008), with permission from John Wiley & Sons.

respectively. Noble metal catalysts such as Rh-based ones, on the other hand, do not dissolve carbon, i.e., do not favor whisker mechanism, and therefore are resistant against coking [66]. In the case of formation of its whisker pattern, carbon dissolves or forms a compound with nickel, leading to diffusion through the metal particle to a grain boundary. As a result, carbon precipitates out and lifts the nickel particle at the tip of a growing whisker (Fig. 12). Nickel is still active in this mode of carbon formation, but accumulation of carbon whiskers leads to breakdown of the catalyst and blocks the catalyst bed, which increases pressure drop dramatically. In general, the process is favored at temperatures higher than B723K [64]. However, some of the carbon formed on the catalyst surface remains undissolved and eventually encapsulates the nickel. At temperatures lower than B773K, encapsulation becomes more pronounced with time by the polymerization of CnHm radicals on the nickel surface, as explained above, and leads to progressive deactivation of the catalyst [64]. It is mentioned above that polymerization of Ca species to Cb is an intermediate step in the formation of dissolved or deposited carbon. It is obvious that polymerization is possible by the presence of more than one Ca species, whose formation requires high number of surface sites. Therefore, it has been reported that it is possible to reduce Ca formation by controlling the number of sites in an ensemble, which will minimize carbon formation but maintain steam reforming. The process is called ensemble size control (or sulfur passivation) and is achieved by addition of minimal quantities of sulfur (e.g., H2S/H2o7.5  10 7) in order to facilitate controlled sulfur adsorption on nickel. As a result, the rate of steam reforming is somewhat reduced but coke formation is almost eliminated [62,64]. Carbon formation can be minimized by use of various promoters. It was reported that addition of small amounts of molybdenum compounds (MoO3) to nickel with quantities less than 0.1% by weight significantly reduced coke deposition and increased the catalytic activity in n-butane steam reforming. Coking resistance was found to increase at elevated steam-to-carbon ratios, since water vapor was thought to stabilize surface molybdenum oxides that limited carbon formation [67]. Addition of SnO2 and WO3 as well as oxides of K, Na, Mg, Ca, and Ba was also found to suppress coking over nickel [67,68]. The presence of these promoters basically prevented methane adsorption, and hence its decomposition on the nickel surface. The type of catalyst support is also found to affect coke formation. It was shown that 1%Pd/ceria works well under low steam-to-carbon ratios (between 1.0 and 2.0) and gives high activities without carbon formation during n-butane steam reforming [69]. Apart from the regulation of catalysts and operating conditions, carbon formation can be minimized at the process design stage. Higher hydrocarbons, particularly olefins, which may appear as impurities in natural gas, can be decomposed to deposit carbon more easily than methane. In order to eliminate this, catalytic prereformer units are usually employed in industrial practice to reform higher hydrocarbons at lower temperatures before they enter the main, multitubular reformer unit. Therefore carbon formation due to the uncontrolled coking of higher hydrocarbons at the conditions of methane steam reforming in the multitubular reformer is minimized [70].

2.16.2.4.3

Thermal degradation

Thermal degradation refers to physical deactivation by means of sintering, chemical transformations, and evaporation, and is positively correlated with temperature. Sintering is the process of transformation of catalysts with high surface area, which is not thermodynamically favored, to lower surface area agglomerates, which are thermodynamically favored, at high temperatures and in the presence or absence of a suitable chemical environment. Sintering can cause reduction in the accessible areas of either the active phase or the support. In the former, dispersed metal nanoparticles migrate on the catalyst surface to diffuse and merge into each other to form bigger clusters (Fig. 13). In this case, metal particles that remain within the clusters become inaccessible to the reactants. In some cases, pores of the high surface area supports merge into each other, and the diffusion of reactants into the active

506

Catalysts

Metal crystallite (B) (A)

Support

Fig. 13 Mechanisms of crystallite growth due to sintering by (A) atomic migration and (B) crystallite migration. Reproduced from Bartholomew CH. Mechanisms of catalyst deactivation. Appl Catal A: Gen 2001;212:17–60. Copyright (2001), with permission from Elsevier. Copyright (2001), with permission from Elsevier.

sites existing in the deeper zones of the pores is simply blocked. For example, g–Al2O3, a commonly used catalyst support, starts to lose its high surface area (ca. 250 m2 g 1) at temperatures greater than 873K. As the temperature is increased, g-phase transforms into δ- and y-phases, respectively. The final transformation occurs at ca. 1373K to a highly stable a-phase, which has a surface area as low as B5 m2 g 1. Both surface migration and pore closure mechanisms eventually lead to decline in catalytic activity. The effect of temperature on sintering of solids is well known. In general, thermal rearrangement of most of the solids occurs at around 0.3–0.5 times the melting point of the material (Tmelting). These points, known as the Hüttig (0.3  Tmelting) and Tamman (0.5  Tmelting) temperatures, correspond to points where the mobility of the atoms at defects and at the bulk, respectively, start [12,56]. These temperatures can be used as references for checking the probability of sintering and for the proper choice of catalyst materials to prevent sintering at the particular range of operating conditions. Melting points as well as Hüttig and Tamman temperatures for various metals and their compounds are provided by [56]. It is worth noting that variations in the nature of the active phase during the reaction can change the sintering characteristics of the catalyst. A descriptive example of such a scenario is the Pt/Rh gauzes used in oxidation of ammonia to nitric oxide. During the process, Pt tends to exist in oxide forms rather than its metallic form. In contrast with metallic Pt, which has a melting point of 2028K, PtO and PtO2, oxide forms of Pt, melt at 823K and 723K, respectively. Considering the high temperatures involved in the conversion (4B1173K), surface mobility of the oxidic species becomes excessive, leading to their evaporation and loss [56]. High temperatures may also lead to unwanted reactions between the metal and the support. Formation of the inactive nickel aluminate phase as a result of the interaction between nickel and alumina demonstrates undesired component interactions at high temperatures. The possibility of alloy formation or phase separation is also reported [59]. In addition to temperature, existence of particular atmospheres accelerates sintering. To illustrate, hydrothermal conditions lead to catalyst deactivation due to sintering in steam reforming. Moist atmospheres are also known to favor sintering of oxide supports [59]. Unlike carbon formation and some cases in sulfur poisoning, sintering leads to irreversible deactivation of the catalysts. In addition to temperature, it is a function of time, which is dictated by the magnitude and duration of the conditions causing sintering. However, sintering resistance of the catalysts can be increased at the stage of their synthesis. For example, sintering of Al2O3 is found to be inhibited by the addition of oxides such as BaO, La2O3, SiO2, Li2O, and K2O, which act as stabilizers filling the vacancies in the lattice structure of alumina [71]. Among these additives, SiO2 is reported to introduce additional resistance against moist atmospheres. The change in the preparation method of alumina, which would have affected the rate of sintering due to the modified particle morphology and grain structure, was proposed as another method to improve thermal resistance [71].

2.16.2.4.4

Mechanical deactivation

Mechanical deactivation refers to a loss in catalytic activity as a result of undesired change in the size and form of the catalyst. As the name indicates, the effect is physical and more commonly observed in fluidized-bed and slurry reactors (Sections 2.16.3.1.4 and 3.1.5, respectively) in which the catalyst bed is mobile. In these units, collision of the catalyst particles with the reactor internals accelerates their mechanical attrition and breakdown into even smaller sizes. Even though reduction in catalyst size increases surface area available for reaction, some fine particles cannot be separated by the catalyst–fluid separation systems (such as cyclones in fluidized beds) and are lost with the product stream. The end effect is a loss in reactor performance, which can only be reversed by online addition of fresh catalyst. It is worth noting that mechanical resistance is determined by the geometric shape and porosity of the pellet, the latter being determined by the support material. Catalysts with spherical shapes and low porosities are reported to have higher mechanical strength [56]. Different scenarios such as coke formation may also initiate breakdown of catalyst pellets. A descriptive example is hightemperature Fischer–Tropsch synthesis (Section 2.16.3.2.6) in which the unsaturated and aromatic nature of the synthesized hydrocarbons in the pores of the Fe-based catalysts can be transformed into coke at the reaction conditions. Low density of coke reduces mechanical strength of the catalyst pellets and accelerates their mechanical destruction into fines with dimensions below ca. 2  10 5 m. Such small fines can hardly be filtered by the cyclone systems and are usually lost with the gaseous product stream.

Catalysts

507

Operations involving high pressures may also accelerate mechanical deactivation. In packed-bed reactors (PBRs) operating at high pressures, mechanical strength is one of the primary prerequisites for the catalysts. If the catalyst damage occurs in a particular section of the bed, the reaction mixture will bypass the damaged zone with lower void fraction, and per-pass conversion will decrease. If all of the bed is affected by mechanical breakdown, smaller particles will fill the voids in the bed and lead to excessive pressure drop. Moreover, residence time of the reactive mixture will increase, which may be dangerous in case of exothermic reactions such as partial oxidation of ethylene to ethylene oxide (Section 2.16.3.2.2). In such cases, increased conversion will cause excessive release of exothermic heat that may not be removed effectively by external cooling. The end result of such a scenario is either the formation of hotspots that cause thermal sintering of the particular section of the bed or, in extreme cases, thermal runaway. Rapid change in operating temperature is another factor that may cause mechanical damage of the catalysts. This phenomenon, however, is a more important concern in monolith and microchannel reactors (Section 2.16.3.1.3). Differences in thermal expansion coefficients of the reactor material and the washcoated catalyst can cause damage and, in extreme cases, delamination, i. e., separation of the catalyst layer. Automotive applications, such as oxidation of soot from diesel engines, are typical examples involving sudden changes in temperature. An excellent review of the mechanical failure of the catalysts can be found in Ref. [57].

2.16.3

Catalytic Reactors and Applications

2.16.3.1

Types and Operation of Catalytic Reactors

Reactors are the processing units that transform raw materials to desired products at specified ranges of operating conditions such as temperature, pressure, and contact time. Due to the fact that catalysts enable transformations to be carried out at reduced energy requirements, most of the commercial reactors involve the use of solid catalysts. Identification of reactors is usually based on the geometry of the catalyst bed and its pattern of contact with the reactive mixture. In the first group of reactors, such as in packedbed, trickle-bed, and in monolithic units, reactive flow contacts with a fixed bed of catalyst. In the second group, however, the catalyst bed is mobile, and the reactive flow is mixed with catalyst particles within the reactor volume. Fluidized-bed, moving-bed, and slurry reactors are typical examples in which catalyst particles are in motion. Description and main operating characteristics of these reactors are provided in the following sections.

2.16.3.1.1

Packed-bed reactors

A PBR is defined by random packing of the catalyst pellets into an enclosed volume, which is either a tube or a vessel (Fig. 14) [17]. The feed stream that is usually in gas phase is passed through the voids between stationary catalyst particles for being converted into products. When the feed mixture involves two phases, the unit is defined as a trickle-bed reactor, which is discussed separately in Section 2.16.3.1.2. PBRs are the most widely used reactor types due to their relatively simple architecture and ease of operability. Classification of PBRs is based on the thermal nature of the catalytic reaction of interest. In this respect, tubular PBRs are used for highly endothermic or exothermic reactions that require extensive heating or cooling, respectively. Vessel-type PBRs, on the other hand, are preferred mostly in providing adiabatic conditions. In this respect, they are also known as adiabatic PBRs. Brief descriptions of tubular and adiabatic PBRs are provided in Sections 2.16.3.1.1.1 and 2.16.3.1.1.2, respectively. 2.16.3.1.1.1 Tubular packed-bed reactors Tubular PBRs are configured by collection of multiple tubes, each of which is packed with a bed of catalyst pellet. As demonstrated in Fig. 15, the tubes are enclosed in a volume, and the resulting geometry is analogous with that of a shell-and-tube heat exchanger [72]. Depending on the heat of reaction, the catalyst packed tubes are either heated or cooled by means of circulation of either a hot or a cold fluid, respectively, on the outside, that is within the shell side of the reactor. In this configuration, the mechanism of heat transfer is primarily by convection. In case of endothermic reactions, hot fluid can either be a gas or a liquid, whereas the

 = Rp

Catalyst particles

Pores r

Rt z Flow L Catalyst bed

Fig. 14 Schematic presentation of a packed-bed reactor (PBR). Reproduced from Onsan ZI, Avci AK. Reactor design for fuel processing. In: Shekhawat D, Spivey JJ, Berry DA, editors. Fuel cells: technologies for fuel processing. Amsterdam: Elsevier Science; 2011. p. 451–516. Copyright (2011), with permission from Elsevier.

508

Catalysts

Feed gas

Saturated steam

Steam Steam drum Feed water

Feed water

(A)

(B)

Circulation turbine

Steam generator

(C)

Fig. 15 Various heat management configurations in multitubular PBRs. (A) Cross-flow, (B) parallel flow, and (C) boiling water cooling. Reproduced from Eigenberger G, Ruppel W. Catalytic fixed-bed reactors. In: Elvers B, editor. Ullmann’s reaction engineering. Weinheim: Wiley-VCH; 2013. p. 305–70. Copyright (2013), with permission from John Wiley & Sons.

phase of the coolant is liquid only for exothermic processes. As presented in Fig. 15, cooling can also be established by means of partial evaporation of water outside the tubes where the vapor bubbles improve the extent of mixing and increase the convective heat transfer coefficient, hence the effectiveness of cooling. In this configuration, circulation of water takes place by means of a positive density difference from the bottom to the top of the tubes. However, in cases where it is designed to remain in liquid phase, coolant is circulated by means of a pump or a mixing/circulating device. In case of some strongly endothermic processes, such as steam reforming of natural gas (Section 2.16.3.2.1), tubes mounted in a fired furnace are externally heated by energy released upon homogeneous combustion of a fuel (mostly natural gas). This configuration, which involves radiative heating of the catalyst tubes, is preferred when the reaction is strongly endothermic and higher production capacities are needed. The extent to which external heat is transferred into or removed from the catalyst bed depends on a number of factors such as reactant flow rate, and geometric and structural properties of the tube and catalyst bed. These factors require careful optimization as the packed-bed structure has inherently weak heat transfer properties. The network of voids existing between the randomly packed catalyst pellets and in the inner tube wall–catalyst bed interface acts as a resistance against uniform distribution of heat within the reactor. This causes unequal temperatures in the bed, which reduces the effectiveness of the catalyst, and, depending on the magnitude of the reaction enthalpy, may lead to hot and cold spots in the case of exothermic and endothermic reactions, respectively. Improving heat transfer rates in PBRs is possible by increasing the intensity of mixing of the reactive flow. Due to the lack of mechanical mixing equipment such as an impeller, the degree of fluid mixing in PBRs can only be increased by the use of smaller tube diameters and catalyst pellet dimensions. These design properties reduce the cross-sectional area of reactive flow, and, in turn, increase its linear velocity thus promoting turbulence and mixing. The end result is increased rate of heat transfer both in axial and in radial directions of the packed tube. Smaller tube diameters are preferred also due to (1) their shorter radial distances corresponding to reduced resistance against heat flow in the radial direction and (2) the possibility of bundling more tubes in the reactor shell to increase the surface area subject to heat exchange. Use of smaller pellet dimensions, another design property for favoring turbulent conditions, offers higher catalyst surface area per unit bed volume that increases the rate of catalytic reactions. Moreover, reduction in the fraction of voids by closer packing of smaller particles enhances conduction that further improves heat transfer along the reactor tube. Despite the operational benefits that they offer, reducing the reactor tube and catalyst pellet dimensions elevates pressure drop, which is the loss of mechanical energy of the reactive fluid due to its friction with the narrow and randomly distributed voidage of the catalyst bed. Pressure drop should be avoided due to high operating expenses of gas compression, and is typically limited to a tolerable level determined by the requirements of the specific operation. As a result, design of tubular reactors, i.e., determination of internal diameter, length and wall thickness of the tube, and geometric properties of the catalyst calls for balancing heat transfer requirements and pressure drop limitations. At this stage, availability of a variety of pellet shapes and dimensions (Fig. 16) with different heat transfer and fluid mechanical properties offers the possibility of wider operating windows in terms of enabling more turbulent conditions and better heat transfer properties at acceptable pressure drop values [73,74]. Some examples of commercial applications of tubular PBRs are provided in Sections 2.16.3.2.1 and 2.16.3.2.2. 2.16.3.1.1.2 Adiabatic packed-bed reactors In contrast with the tubular PBRs, the heat exchange function is not integrated with the catalytic reactions carried out in adiabatic PBRs. These units are configured by the packing of catalyst pellets into a pressure vessel (Fig. 17(A)), and uniformity of catalyst

Catalysts

509

Fig. 16 Various geometric configurations of solid catalysts. Reproduced from Gallei EF, Hesse M, Schwab E. Development of industrial catalysts. In: Ertl G, Knozinger H, Schüth F, Weitkamp J, editors. Handbook of heterogeneous catalysis. Weinheim: Wiley-VCH; 2008. p. 57–66. Copyright (2008), with permission from John Wiley & Sons.

(A)

(B)

(C)

(D)

Fig. 17 Various types of adiabatic PBRs. (A) Single-bed adiabatic PBR, (B) adiabatic PBR with interstage gas injection, (C) multiple adiabatic beds with interstage heat exchange, and (D) multiple adiabatic beds with external interstage heat exchange. Reproduced from Walas SM. Chemical process equipment. Boston, MA: Butterworth-Heinemann; 1990. Copyright (1990), with permission from Elsevier.

packing is critical to ensure uniform distribution of the feed stream to the catalyst bed. Trade-off between heat transfer and pressure drop along the catalyst bed, which is explained for tubular PBRs in Section 2.16.3.1.1.1, is also valid for the design and operation of the adiabatic PBRs. Heat management in adiabatic PBRs depends mainly on the magnitude of the reaction enthalpy. In case of mildly exothermic or endothermic reactions, temperature can be regulated by means of direct injection of a portion of cold or hot feed, respectively,

510

Catalysts

at various locations of the catalyst bed (Fig. 17(B)). Success of this strategy depends on the extent to which the injected feed is mixed with the main reactive flow and redistributed to the catalyst bed. This factor together with the amount of injected feed has a strong impact on the change in adiabatic reaction temperature, which must be regulated carefully to ensure desired reaction rates. Direct injection becomes insufficient for processes involving high heats of reaction. In such cases multiple adiabatic stages (beds) with interstage heat exchange are employed (Fig. 17(C) and (D)) [75]. Dimension of each bed is determined by the combination of a number of factors including feed conditions, allowable adiabatic temperature change, and pressure drop. When compared with a tubular PBR, multistage adiabatic PBR configuration offers less control over reaction temperature as heat exchange occurs continuously in the former, whereas it takes place in definite sections, i.e., only between the stages of the latter. Despite this fact, multistage configuration is preferred over the tubular PBRs due to the possibility of sectional replacement of the catalyst bed, packing different pellets to each section for customizing product distribution, side feeding of part of the feed, and side drawing a product for reasons such as improving conversion in equilibrium-limited reactions [74]. A commercial example, ammonia synthesis, which involves the use of multistage adiabatic PBR, is explained in Section 2.16.3.2.3.

2.16.3.1.2

Trickle-bed reactors

Trickle-bed reactors are similar to the PBRs described in Section 2.16.3.1.1.2 in terms of geometry and adiabatic operation. The main difference is the coexistence of concurrently fed gas and liquid species in the feed of trickle-bed units. This condition requires the use of somewhat sophisticated equipment, such as bubble-cap trays, which uniformly mix gas and liquid species and distribute them evenly to the cross-section of the catalyst bed. The uniformity of flow distribution along the catalyst bed depends also on the reactor diameter, which is usually designed to remain below 5 m, as well as on the catalyst pellet sizes, which are in the range of 1–5  10 3 m to give bed void fractions between 0.35 and 0.40 [17]. In multistage trickle-bed reactors, trays are located prior to each catalyst bed to ensure remixing of the main reactive stream together with the quench feed, and to unify flow distribution in each stage. Height-to-diameter ratios of trickle-bed reactors are reported to be between 5 and 25 [17]. Trickle-bed reactors are widely used in petroleum refining and petrochemicals processing operations. Typical examples of refining operations involving trickle-bed reactors are hydrocracking of vacuum gas oil and residues (see Section 2.16.3.2.4 for further details), HDS of crude oil, and hydroisomerization of C4–C6 alkanes to branched isomers. These reactors are also used in hydroprocessing of the waxy hydrocarbon product mixture of Fischer–Tropsch synthesis, which is a key step of gas-to-liquid complexes. Comprehensive description, modeling, design, and operation of trickle-bed reactors can be found elsewhere [17,76].

2.16.3.1.3

Monolith and microchannel reactors

Despite being categorized as a fixed-bed reactor, monolith reactors differ significantly from PBRs and trickle-bed units by means of their geometry. In monolith reactors, fluid flow occurs through ordered, parallel channels, whose inner walls are coated with a porous catalyst layer (Fig. 18). Small channel diameters that range typically between 3  10 4 m and 6  10 3 m lead to a notable increase in the surface area-to-volume ratio, which is on the order of B104 m2 m 3. This metric, however, is limited to B103 m2 m 3 in conventional PBRs [27]. The resulting intensification leads to up to ca. 2 times higher void fractions (B0.75 in monoliths vs. B0.4 in PBRs) that reduce pressure drop limitations significantly in monolith units [77], and allows the monolith reactors to remain as the only option for space-limited applications such as catalytic posttreatment of exhaust gases in vehicles. In addition to their unique geometric properties, monolith reactors have operating characteristics that are different from those of PBRs. Due to the small channel diameters varying in the micro-to-millimeter range, flow regime is laminar. Therefore, in contrast with PBRs, transport coefficients within the channel remain almost unaffected by the flow rate. Instead, heat and mass transfer coefficients are determined by Nusselt (Nu) and Sherwood (Sh) numbers, respectively, both of which have constant values specific to cross-sectional shape of the channels provided that they are straight and flow is fully developed. Due to the definitions of Nu and Sh, transport coefficients become inversely proportional to the channel diameter [27]. Another operational difference of monolith reactors is based on their residence time distribution, which is narrower than those of packed-bed units. Presence of catalyst in coated form and existence of well-defined void space in monoliths lead to reactive flows that follow similar streamlines

Flow

L

Pores Fig. 18 Schematic presentation of a monolith reactor. Reproduced from Onsan ZI, Avci AK. Reactor design for fuel processing. In: Shekhawat D, Spivey JJ, Berry DA, editors. Fuel cells: technologies for fuel processing. Amsterdam: Elsevier Science; 2011. p. 451–516. Copyright (2011), with permission from Elsevier.

Catalysts

511

in the channel and have almost the same duration of contact with the catalyst layer. The random nature of the voids in PBRs, however, randomizes the flow streams, causing occurrence of different residence times. In this respect, monolith reactors are more suitable for carrying out reactions whose extent depends strongly on the residence time, which can be reduced down to the order of milliseconds by the contact of fast reactive flows, allowed by reduced pressure drop limitations, with small amounts of coated catalysts. A well-known example of monolith reactor–driven, contact time–dependent reaction is the direct partial oxidation of methane to synthesis gas, as described in Section 2.16.3.2.5. The primary mechanism of transfer of heat within a monolith unit is conduction. Even though convective heat transfer exists due to fluid flow within the channels, its magnitude is not as strong as in packed beds. At this stage, material of construction plays an important role in determination of heat transfer rates. For processes requiring temperature control by means of external heating or cooling, metallic monoliths are preferred. However, for adiabatic reactions, monoliths made of low thermal conductivity materials such as cordierites are employed. Despite its advantages, the risk of plugging of the channels and subsequent flow maldistribution leading to lower conversions and disturbed product selectivities, as well as the inflexibility in replacing the catalyst in case of any irreversible deactivation, remain as the main practical drawbacks of monolith reactors. Moreover, care should be taken to prevent delamination of the catalyst layer, which is caused by the difference in the coefficients of thermal expansion of the monolith and catalyst materials. At this point, mechanical stability of the catalyst coatings, which depends on the monolith material, composition of the catalyst and the coating method, should be well studied prior to running monolith reactors. A review of catalyst coating techniques can be found in Ref. [78]. An alternative to the monolith reactors is microchannel units, which have geometric and operational properties similar to those of monoliths. Microchannel reactors, however, involve smaller hydraulic channel diameters that are mainly in the micrometer range, leading to even higher surface area-to-volume ratios up to B5  104 m2 m 3 [79–81]. The differences in the method of construction of microchannel reactors, which mainly involve high-precision machining of the channels onto plates and their subsequent bonding (Fig. 19), allow versatility in obtaining different longitudinal channel shapes over a wider range of construction materials such as copper, aluminum, stainless steel, iron–chromium alloys, silicon carbide, ceramics, and polymers [82]. Compared to straight channels existing in monolithic units, nonstraight (e.g., sinusoidal, zig-zag shaped) channel shapes are reported to induce mixing and improve transport coefficients. Presence of higher number of options for reactor material, on the other hand, offers the chance of improving functional properties such as thermal conductivity. In addition to coating, the possibility of packing the channels with properly sized catalyst particles, which allows easier replacement of the catalyst bed, is another structural flexibility existing in microchannel reactors.

H S

Bonding (C)

(A)

Bonding

(B)

Pores Catalyst washcoat Line of symmetry Microchannel

H/2 s Wall

Microchannel L

Ls

yw

H/2 Line of symmetry

s y x

Fig. 19 Schematic presentation of a microchannel reactor. (A) Machined plates with microchannels, (B) reactor block obtained after bonding the plates, (C) characteristic section of the multichannel unit with porous catalytic washcoats. Reproduced from Onsan ZI, Avci AK. Reactor design for fuel processing. In: Shekhawat D, Spivey JJ, Berry DA, editors. Fuel cells: technologies for fuel processing. Amsterdam: Elsevier Science; 2011. p. 451–516. Copyright (2011), with permission from Elsevier.

512

Catalysts

In addition to volumetric intensification, microchannel reactors can be intensified in terms of function. The versatility in terms of their construction allows integration of heat-exchange or separation functions to the microchannel units. In the context of heatexchange integrated microchannels, coupling of (1) endothermic and exothermic reactions [83–85], (2) exothermic reactions with coolant flow [86–89], and (3) endothermic reactions with heating fluid flow [79] can be realized to offer excellent temperature control that boosts catalyst effectiveness and reactor productivity. Moreover, integration of membranes into the microchannel units is reported to offer effective execution of simultaneous reaction and separation [90]. Similar to monolithic units, microchannel reactors offer realization of catalytic reactions at very short contact times being almost unaffected by pressure drop. Owing to their flexible properties and higher number of options of intensification strategies, microchannel units find use in applications such as fuel processing [17,91] and onsite conversion of raw materials to valuable products (e.g., conversion of offshore natural gas-to-liquid fuels) [79,85,92]. They are also considered in transforming batch-wise producing units to small-scale continuous reactors in particular industries such as production of pharmaceuticals and fine chemicals. The reader is directed to references [17,79] for in-depth information on the operation, design, and analysis of monolith and microchannel reactors.

2.16.3.1.4

Fluidized-bed reactors

Fluidized-bed reactors involve carrying out gas–solid type, two-phase reactions over a mobile catalyst bed. In this respect they differ from PBRs in which the catalyst bed is fixed. Fluidization of the catalyst particles with the reactive flow offers intense mixing, leading to efficient heat exchange with the heat transfer surfaces/coils immersed into the reactor vessel. This allows running very exothermic reactions, such as high-temperature Fischer–Tropsch (HTFT) synthesis, in a very narrow temperature range, which, as described in Section 2.16.3.2.6, is essential for obtaining the desired product distribution. Mobility of the catalyst bed also helps in reducing pressure drop limitations and using smaller-sized catalyst pellets, the latter of which increases the available surface area of the catalyst to improve reaction rates and conversions. In addition to such operational benefits, catalyst mobility allows continuous regeneration, which is essential for operations such as FCC, and online addition of fresh pellets [17]. Despite its unique advantages, capital costs of fluidized-bed reactors are higher than those involved in PBRs due to the need for catalyst separation and circulation systems. Catalyst pellets should be mechanically strong to preserve their integrity against collisions, as their breakdown will increase the risk of catalyst loss with the product stream and will cause further increase in reaction rates where heat exchange configuration may have difficulties controlling the temperature within the specified limits. Accelerated erosion of the reactor internals and defluidization as a result of merged catalyst pellets are other possible side effects of pellet collision. Mobility of the bed also makes the reactor sensitive against poisoning by agents such as sulfur-containing compounds in the feed. Once they enter into the reactor including an active bed of catalyst, poisons interact with the pellets immediately and start to deactivate them due to well-mixed conditions. On the other hand, in a similar situation with a PBR, poisons only affect the upstream portion of the bed, while the rest of it remains active. This phenomenon continues until the upstream section becomes saturated with the poison. In this respect, rate of decrease in conversion is expected to be sudden in fluidized bed reactors, but to be moderate in PBRs. In both configurations, duration of poisoning depends on several factors such as accessible metal surface area of the pellets and concentration of the poison in the feed. In addition to the drawbacks stated above, the scale-up phase involved in fluidized bed units is typically longer and more expensive than those involved in PBRs due to increased uncertainties in predicting hydrodynamics and mixing patterns of mobile gas and solid phases at large scales [17].

2.16.3.1.5

Slurry reactors

Slurry reactors involve coexistence of the mobile catalyst (solid) phase together with gas and liquid phases. The roles of the phases can vary according to the process. Gas phase can either be one of the reactants or products, or an inert used for agitating the liquid–solid mixture. Similarly, liquid phase can be a part of catalytic reaction or can serve as an inert medium for contacting gas and solid catalyst phases. In some configurations, a mechanical stirring equipment can be integrated to the reactor operation [17]. Demonstration of possible configurations of slurry reactors is presented in Fig. 20 [93]. Except for the number of phases involved, slurry reactors possess the operational advantages of fluidized bed reactors explained in Section 2.16.3.1.4. In this respect, efficient mixing brought about by the mobility of the phases allow high rates of heat exchange with the heat transfer surfaces immersed into the reactor volume. Mixing is also critical to improve heat and mass transfer coefficients between the phases. Versatility in catalyst pellet size allows obtaining high reaction rates without being limited by pressure drop. Small pellet dimensions also lead to short diffusion paths, allowing faster intraparticle transport of the reactive mixture within the pores of the catalyst. Costs and complexities associated with the separation of catalysts from the reactor effluent, and the risk of accelerated poisoning due to the intense mixing of the pellets with the fluid phases, remain as the main drawbacks of slurry reactors. Slurry reactors have a wide range of use in the chemical industry. Processes of hydrogenation, oxidation, chlorination, hydroformylation, and bioremediation are examples where slurry reactors are employed. A well-known application, LTFT synthesis, is discussed in Section 2.16.3.2.7. Detailed information regarding description, operation. and design of three-phase slurry reactors can be found in the literature [17].

2.16.3.2

Commercial Applications of Catalysts and Catalytic Reactors

This section outlines examples that describe the use of solid catalysts for commercial production and fuel processing. The examples include the commercial significance of the process, catalyst, and reactor types together with the operating conditions and typical operating characteristics.

Catalysts

513

Gas outlet Gas outlet

G-L separator

Cooling tube Cooling tube

Riser

Down-comer

Gas distributor

Gas distributor

Gas inlet (A)

Gas inlet

(B)

Gas outlet Gas outlet

Gas-liquid separator

G-L separator

Internal

Down-comer

Cooling tube

Riser

Secondary distributor

Downer

Gas distributor

Gas distributor (C)

Gas inlet

(D)

Fig. 20 Various configurations of slurry reactors. (A) Bubble column, (B) internal-loop airlift reactor, (C) external-loop airlift reactor, (D) spherical reactor. Reproduced from Wang T, Wang J, Jin Y. Slurry reactors for gas-to-liquid processes: a review. Ind Eng Chem Res 2007;46:5824–47. Copyright (2007), with permission from American Chemical Society.

2.16.3.2.1

Steam reforming of natural gas to hydrogen

Steam reforming (SR) of natural gas is the conventional technique for large-scale production of hydrogen, which is a critical feedstock for petrochemical and petroleum refining operations. Hydrogen is also the chemical energy source for various fuel cells, but particularly for the polymer electrolyte membrane (PEM) fuel cells used to provide power up to B5 kW for mobile and residential applications. The process can be described by the following reaction given for methane, the major constituent of natural gas: CH4 þ H2 O ¼ CO þ 3H2

DHro ¼ 206 kJ mol

1

ð38Þ

1

ð39Þ

Product distribution is affected by water–gas shift (WGS) that runs together with SR: CO þ H2 O ¼ CO2 þ H2

DHro ¼

41 kJ mol

514

Catalysts

Fig. 21 Different furnace geometries employed in multitubular PBRs. Reproduced from Dybkjaer I. Tubular reforming and autothermal reforming of natural-gas – an overview of available processes. Fuel Process Technol 1995;42:85–107. Copyright (1995), with permission from Elsevier.

Strong endothermic nature of SR and high stability of methane require extensive energy supply to the reactor, which is of multitubular PBR type. In this configuration, the catalyst packed tubes mounted in a furnace are heated externally by radiation, which is obtained by homogeneous combustion of a hydrocarbon fuel such as natural gas or fuel oil. Difference in the location of combustion points in the furnace (Fig. 21) is reported to affect heat distribution into the tubes [94]. A typical example of a natural gas steam reformer, which is also called a radiant reformer, is presented in Fig. 22. The resulting heat flux, which is above B105 kcal m 2 h 1 [63], elevates the feed temperature (723–923K) up to the 1123–1193K range for improving the reaction rate. Extent of heat supply and corresponding upper temperature limit of the process is limited by the tube material whose thermal stability is ensured by the use of microalloys with a typical content of 25Cr 35Ni Nb Ti [95]. The length of the tube subjected to external heating is between 10 and 14 m, while the outer diameter and wall thickness of the tube are in the ranges of 10–18 cm and 0.8–2.0 cm, respectively. Capacity of the tubular reformer unit, which is typically up to 3  105 Nm3 h 1, dictates the number of tubes, which can go up to 1000 [63]. Commercial multitubular natural gas reformers operate at exit temperatures in excess of 1073K and pressures between 25 and 40 atm. Even though the stoichiometry of Reaction (38) is thermodynamically favored at lower pressures, requirement of highpressure hydrogen in downstream applications necessitates compression of the feed mixture, which is economically more feasible than the compression of the product stream. Typical commercial SR catalysts are Ni/a-Al2O3 or Ni/MgO, which are designed to demonstrate sufficient activity together with high thermal and mechanical stability. Even though precious metals such as Rh and Ru exhibit higher SR activities, Ni is preferred due to its lower cost. However, in order to prevent carbon formation (Section 2.16.2.4.2) during Ni-catalyzed SR, steam is fed in excess quantities to obtain inlet molar H2O/C ratios of B3. The H2O/C ratio, however, is limited by the sintering tendency of Ni in steam-rich environments. In addition to its chemical composition, size and shape of the catalyst pellet are important for the improvement of heat distribution with tolerable elevations in pressure drop. Among several options, the 7-hole cylindrical pellet with a diameter of 16 mm, offering sufficient intrapellet voidage, is found to be optimal. Catalyst pellets are expected to preserve their mechanical rigidity, as their destruction causes excessive increase in pressure drop and disturbs flow distribution, which can lead to local heat accumulation and subsequent overheating of the reactor tube to end up with carbon formation and thermal degradation of the catalyst. Natural gas steam reforming can also be carried out in the so-called convective reformers, where heating of the tubes is carried out by means of contacting of the tubes with hot flue gas. Even though the heat fluxes in convective reformers are smaller than those involved in fired reformers, the former is preferred for its compactness and used when lower production capacities are involved [63].

2.16.3.2.2

Partial oxidation of ethylene to ethylene oxide

Ethylene oxide is an important commodity that is primarily used as a raw material in the production of ethylene glycols, glycol ethers, ethanolamines, and surface active agents. Production of ethylene oxide can be carried out by direct catalytic oxidation of ethylene: 2C2 H4 þ O2 -2C2 H4 O

DHro ¼

105 kJ mol

1

ð40Þ

This reaction is accompanied by the total oxidation of ethylene and ethylene oxide described in Reactions (41) and (42), respectively: C2 H4 þ 3O2 -2CO2 þ 2H2 O C2 H4 O þ 2:5 O2 -2CO2 þ 2H2 O

DHro ¼ DHro ¼

1324 kJ mol

1

1219 kJ mol

ð41Þ 1

ð42Þ

The process is conventionally carried out over Ag-based catalysts, which are known to selectively promote Reaction (40) over the undesired total oxidation of ethane and produced ethylene. The catalysts are typically supported on Al2O3. Recent commercial applications involve use of oxygen, with purities in excess of 95%, to improve the reaction rate, conversion of ethylene, and yield of ethylene oxide as well as avoiding postseparation of N2 from the product mixture, which is needed when air is used as the

Catalysts

515

Fig. 22 Side-fired tubular reformer design. Reproduced from Dybkjaer I. Tubular reforming and autothermal reforming of natural-gas – an overview of available processes. Fuel Process Technol 1995;42:85–107. Copyright (1995), with permission from Elsevier.

oxidant. The advantages of using oxygen, however, come at the expense of an air separation unit, which notably increases the capital cost of the overall process. Older processes mainly use air instead of oxygen. The strong exothermic nature of the process requires the use of multitubular reactor structure with external cooling, which is carried out by partial evaporation of cooling water outside the tubes as described in detail in Section 2.16.3.1.1.1. The number of

516

Catalysts

tubes is typically between B1000 and 1500, each of which has an internal diameter and length in the ranges of (2–5)  10 2 m and of 10–15 m, respectively. Wall thickness of the tube is dictated by reactor pressure, which is kept between B15 and 20 atm for higher catalyst productivity. With these dimensions, the reactor can handle ethylene charges between B100–130 t h 1. In case of higher capacities, multiple reactors can be used in parallel to deliver the targeted production rate. The cooling system is operated to keep the catalyst bed temperature between 513 and 543K [96]. Pellets of Ag/Al2O3 catalysts generally have a hollow-cylindrical shape with height-to-diameter ratio close to 1, with each dimension being equal to B9  10 3 m. Deactivation of these catalysts is primarily due to the sintering and agglomeration of Ag particles leading to reduced metal surface area and decreased catalytic performance, which is usually measured in terms of molar yield of ethylene oxide per mole of ethylene consumed. In order to maintain the desired level of catalytic activity in commercial operations, reaction temperature is increased to balance the negative effect of sintering. However, temperature together with oxygen concentration in the feed require careful control as their combined increase is reported to be the primary cause of sintering [97]. Moreover high temperatures are reported to promote CO2 formation. The impact of high oxygen concentrations on product distribution, however, depends on temperature. It is reported that, below 543K, ethylene oxide formation is favored by the increased use of oxygen in the feed. This trend, however, is found to change at temperatures above 543K [97].

2.16.3.2.3

Ammonia synthesis

Ammonia is an important commodity for the chemical industry. It is mainly used in the production of fertilizers as well as in the synthesis of nitrogen-containing compounds such as nitric acid. Ammonia synthesis can be described by the following reaction: N2 þ 3H2 ¼ 2NH3

DHro ¼

92 kJ mol

1

ð43Þ

Process conditions involved in ammonia synthesis are somewhat severe. Synthesis conditions involve temperatures between 673 and 773K. Pressures are typically up to 300 atm, but are reported to be reduced down to 100–150 atm by recent process improvements [98,99]. Such high pressures are needed for favoring ammonia production according to Reaction (43) by lowered thermodynamic restrictions, which also require removal of the exothermic heat of reaction. For this purpose, multistage adiabatic PBRs with interstage cooling are employed. Cooling is carried out by water, which is then evaporated into steam. Product mixture that includes B20% ammonia is passed through a series of condensers, where ammonia is separated. Unreacted N2 and H2 are then recycled back to the reactor together with a make-up stream. Typical production capacities are reported to be up to B2.2  103 t d 1 [99]. An example of a commercial ammonia synthesis converter is shown in Fig. 23. The traditional catalyst packed into ammonia synthesis reactors is of the Fe3O4-based fused iron type. A more active type, Fe1-xO-based fused iron catalyst, has been synthesized in later years. In addition to these iron-based catalysts, some of the recent applications involve the use of Rubased catalysts. Activity and stability of the iron-based catalysts are increased by the use of promoters such as Al2O3, K2O, CaO, SiO2, MgO, and oxides of chromium, manganese, zirconium, and vanadium [98,99]. Deactivation of iron-based catalysts can be due to either thermal sintering or poisoning. Thermal sintering involves irreversible loss in iron surface area. Despite the fact that it is an inevitable phenomenon under reaction conditions of ammonia synthesis, sintering proceeds slowly and does not require replacement before 5–10 years, which is the typical operating life of iron-based catalysts. The other deactivation mechanism, poisoning, occurs as a result of contacting of species such as H2O, CO, and CO2 with the catalyst bed. In this case, the extent of deactivation depends on the level of exposure, i.e., partial pressure of the poisons, and, provided that the duration of exposure is limited to a few days, catalytic activity is restored to its original level. In case of poisoning with sulfur- and chlorine-containing compounds, however, deactivation becomes irreversible, as iron-based catalysts are highly sensitive to very small amounts of sulfur [98]. This risk is somewhat controlled by the use of calcium promoters, which are reported to improve stability against sulfur poisoning. For further details, the reader is directed to references [98,99], which involve comprehensive explanation and in-depth review on the catalysis of ammonia synthesis.

2.16.3.2.4

Hydrocracking

The process of hydrocracking in petroleum refining is used to convert liquid phase vacuum gas oil and residues into smaller-chain hydrocarbons such as kerosene, jet fuel, and diesel fuel in the presence of high pressure H2 over a solid catalyst. Hydrocracking involves cracking, hydrogenation, and isomerization reactions. During cracking, high-chain hydrocarbons are catalytically cracked to shorter-chain saturated (paraffins) and unsaturated (olefins and aromatics) hydrocarbons, with the latter being saturated in the hydrogenation part. Branching of the alkyl groups of the paraffins and opening of the rings of cycloparaffins are carried out in the isomerization stage. Owing to the existence of different reactions, hydrocracking catalysts such as Pt on aluminosilicates or zeolites have multiple functions. The noble metal component drives the hydrogenation reactions while the acidic component is active for cracking and isomerization [100]. Operating conditions of hydrotreating are somewhat severe. Operating temperatures range between B573 and 873K and pressures are well above 100 atm. The requirement of such high pressures of H2 is due to its heavy utilization in cracking and in hydrogenation reactions, as well as in preventing coke formation due to condensation of polynuclear aromatics generated during cracking. The process is carried out in trickle-bed reactors involving multiple adiabatic stages. Due to the exothermic nature of the hydrogenation reactions, heat generated in each stage is absorbed by H2 quenching carried out between the stages. The number of stages in hydrocrackers can go up to six. Hydrocracking is also used as an option for processing of the waxy hydrocarbon mixture produced during LTFT synthesis to obtain a mixture called synthetic crude, which is then distilled to obtain end products such as LPG, naphtha, kerosene, and diesel,

Catalysts

D FG C

517

FE

Temperature (°C) 400

A H

450

500

550

A - Gas inlet B - Gas exit to heat recovery C - Gas exit D - Direct by-pass E - Gas from external start up heater F - Quench gas inlets G - Pyrometer H - Catalyst discharge nozzle

B

Fig. 23 ICI quench-type ammonia synthesis converter. Reproduced from Walas SM. Chemical process equipment. Boston, MA: ButterworthHeinemann; 1990. Copyright (1990), with permission from Elsevier.

the component that is aimed to be maximized. It is worth noting that, due to the LTFT conditions (Section 2.16.3.2.7), diesel is rich in linear hydrocarbons and poor in aromatics, both of which improve quality that is measured in terms of cetane number (B70 in FT diesel vs. 40–50 in conventional refinery diesel). Likewise, the paraffinic content of naphtha makes it an ideal feedstock for ethylene production by steam cracking [101]. In contrast with those involved in refinery hydrocracking, the conditions of FT hydrocracking are somewhat milder due to the absence of species such as polynuclear aromatics and sulfur-containing molecules both of which demand high H2 pressures for preventing their conversion into carbon and for eliminating catalyst poisoning, respectively. As a result H2 pressures below 100 atm are usually sufficient to drive the hydrocracking process. Moreover, longer catalyst lifetimes can be obtained due to the absence of species causing deactivation. Like in the case of refinery hydrocracking, the catalysts should be capable of driving reactions of hydrogenation (via a noble metal component) and cracking (via an acidic component) simultaneously [101].

2.16.3.2.5

Direct partial oxidation of methane to synthesis gas

Synthesis gas (syngas), a mixture of CO and H2, is the feedstock for several industrial operations such as Fischer–Tropsch synthesis (Sections 2.16.3.2.6 and 2.16.3.2.7), and production of methanol, ammonia, and dimethyl ether. Syngas can be produced from natural gas (primarily methane), coal, and biomass. When natural gas is used as the hydrocarbon source, syngas production is carried out by autothermal reforming (ATR), which combines homogeneous, substoichiometric combustion of methane and its

518

Catalysts

catalytic SR (Reaction (38)) in an adiabatic PBR [70,102]. ATR, however, requires cofeeding of methane, oxygen (or enriched air), and steam. An alternative method of producing syngas without requiring steam is by the direct catalytic partial oxidation route: CH4 þ 0:5O2 -CO þ 2H2

DHro ¼

35:7 kJ mol

1

ð44Þ

Apart from its advantage of being capable of operating without steam, the direct route delivers molar H2/CO ¼ 2, the desired syngas composition for LTFT and methanol synthesis, and eliminates any extra processing for the regulation of the syngas mixture. The process is carried out at molar C/O ratios close to 1 and temperatures ranging from 1273 to 1373K [103,104]. In addition to these conditions, realization of Reaction (44) depends strongly on the residence time, which should be between B10 4 and 10 2 s. In case of longer residence times, however, conversion of methane follows the sequential occurrence of its total oxidation and SR (Reaction (38)) [105–108]. Residence times in the order of milliseconds can only be obtained by the contact of high flow rates of reactive flow with small quantities of catalyst. In addition to its level, distribution of residence time is also critical, as it should be narrow enough to eliminate prolonged contact of reactive flow to follow the sequential conversion path. All of these requirements can be obtained in monolith reactors, as high void fractions allow handling of high flow rates with minimal pressure drop, and, compared with a PBR, the amount of catalyst used in the reactor is small (Section 2.16.3.1.3). Moreover, the thickness of the layers of catalyst coated within the channels remains usually below B1  10 4 m, which provides diffusion paths short enough to minimize mass transfer resistances within the catalyst. The catalysts that can drive the direct route under the conditions stated above are Rh-based [103,104]. A number of alternative catalysts involving Pt, Ru, Re, Ir, Co, Ni, and Fe have been investigated. However, it is reported that these catalysts fail to function like Rh due to low activity/selectivity characteristics (Pt, Ir), deactivation caused by aluminate formation (Ni, Co), coke formation (Pd), and problems in driving the autothermal operation (Re, Ru, Fe) [109]. In addition to the active metal, the type of support affects the mechanism of methane-to-syngas conversion. Observation of a sudden elevation in temperature followed by its reduction on SiO2-supported Rh catalyst indicates the initial occurrence of exothermic total oxidation followed by endothermic SR [110]. In the same study, experiments carried out on Al2O3- and TiO2-supported Rh, however, show an initial decrease in temperature followed by its small increase, which is peculiar to the direct oxidation route.

2.16.3.2.6

High-temperature Fischer–Tropsch synthesis

HTFT synthesis is based on the conversion of synthesis gas into a mixture of hydrocarbons that is rich in gasoline and olefins. The reaction is conventionally carried out in a fluidized-bed reactor that is also known as a Sasol Advanced Synthol (SAS) reactor (Fig. 24) at temperatures around 613K and pressures in the range of 20–40 atm [111]. Operating capacity of the SAS reactor is reported to be up to 2  104 bbl d 1, which is equivalent to 8.5  105 t y 1 [17]. Diameter of such a reactor is equal to 10.7 m, allowing enough space for insertion of higher number of cooling coils into the reactor vessel. The coils, in which boiler feed water is used as the coolant, are responsible for providing isothermal conditions by absorbing strong exothermic heat release of the synthesis reaction, which is B145 kJ per CH2 chain added [112]. Upon heat absorption, boiler feed water is evaporated to generate steam with pressures in excess of 40 atm [111]. It is worth noting that temperature is a critical parameter for the regulation of product selectivities, which affect the extent of fluidization. It is reported that excessive condensation in the catalyst pores due to increased selectivity towards longer chain hydrocarbons may lead to wetting of the particles, which can further lead to their agglomeration and end up with defluidization [112]. The catalysts used in HTFT involve Fe as the active phase. The catalyst is prepared by fusing iron oxide at B1773K with Al2O3 or MgO, both of which are structural promoters to act as spacers between iron crystals to inhibit sintering, and with K2O, which is a chemical promoter to improve activity and selectivity towards heavier hydrocarbons [113]. Catalyst pellet dimensions are reported to range from 5  10 5 m to 2  10 4 m [114]. Several reasons such as carbon formation, sulfur poisoning, hydrothermal sintering, and oxidation of active metals/carbides to inactive oxides are stated to be the root causes of deactivation of Fe-based catalysts, with carbon formation being the most significant one [113]. Due to the aromatic nature of heavy hydrocarbons (waxes) produced in the pores of catalyst pellets in the liquid phase, deposition of coke precursors can occur upon evaporation of the hydrocarbons under HTFT conditions. High-temperature conditions also favor elemental carbon formation through the Boudouard reaction (2CO-CO2 þ C(s)). The end effect of carbon deposition is reduction in the density and mechanical strength of the catalyst pellets, which can be crushed into so-called fines, i.e., particles smaller than 2.2  10 5 m by collision effects [113]. This phenomenon, i.e., reduced particle density, has a negative impact on reactor performance due to disturbed fluidization pattern and leads to the loss of fine catalysts from the cyclones. The negative impacts of catalyst breakdown are eliminated by online replacement of a certain fraction of the catalyst bed to obtain stable reactor operation. Detailed information about the synthesis, operation, and deactivation mechanisms of Fe-based HTFT catalysts can be found elsewhere [111–113].

2.16.3.2.7

Low-temperature Fischer–Tropsch synthesis

Even though both routes start from syngas to obtain synthetic hydrocarbons, LTFT is different from the high-temperature route in several aspects. LTFT is carried out at lower temperatures, typically ranging from B463 to 523K, and at pressures between 20 and 40 atm. These conditions are used over Co-based catalysts to obtain the so-called wax, which involves heavier hydrocarbons of a more linear and paraffinic nature to be hydroprocessed to maximize the production of diesel. LTFT can also be carried out by Febased catalysts prepared by precipitation to include SiO2 (that acts as a binder and spacer to improve mechanical strength and minimize metal sintering, respectively), Cu to improve reducibility, and K2O to increase activity. However, at the low-temperature conditions, Co is three times more active than Fe, and is preferred for synthesizing the longer chain hydrocarbons, with diesel

Catalysts

519

Products gases Cyclones

Fluidised Steam

Boiler feed Water

Gas distributor

Total feed

Fig. 24 Sasol Advanced Synthol turbulent fluidized-bed reactor configuration for HTFT. Reproduced from Steynberg AP, Espinoza RL, Jager B, Vosloo AC. High temperature Fischer-Tropsch synthesis in commercial practice. Appl Catal A: Gen 1999;186:41–54. Copyright (1999), with permission from Elsevier.

being the most important one. Co is also preferred to Fe because of its lower tendency towards coke formation, lower selectivities towards olefins and oxygenates, and reduced water–gas shift activity [101]. Despite its comparable activity, Ni is not preferred due to its high (and undesired) methane selectivity. Ru, the most active catalyst, is simply not considered owing to its high cost (440 times more expensive than Co). In order to increase metal dispersion, Co-based catalysts are supported on oxides such as Al2O3, SiO2, and TiO2. Supports also affect the nature of hydrocarbons produced, which tend to be linear when supports are less acidic. High activity of Co-based catalysts increases the release of exothermic heat of reaction, which should be removed efficiently to approach isothermal conditions dictated by the synthesis parameters. This strict requirement can be satisfied in slurry bubble column reactors demonstrated in Fig. 25. As explained in Section 2.16.3.1.5, integration of high-area cooling coils and the vigorously agitated gas–liquid–solid mixture leads to fast transfer of reaction heat from reaction mixture to boiler feed water, which is evaporated to produce steam. In slurry bubble column reactors the rising of the gas, which is sparged into the vessel from the bottom, is realized by the buoyancy gradient between gas and slurry (i.e., mixture of liquid and fine (5  10 5–2  10 4 m sized) catalyst particles). The resulting gradient creates and sustains the circulation leading to intense mixing [17]. An empty volume (without slurry) is foreseen at the upper section of the reactor for a demister, which separates the mist transported by the rising gas. Lighter, vapor phase hydrocarbons leaving the reactor in the gas stream are recovered in a separate unit, whereas heavier, waxy hydrocarbons in liquid phase are separated within the reactor by means of filtration. Co-based LTFT in slurry bubble column reactors is commercially used in the Oryx GTL (gas-to-liquids) complex in Qatar, where the design capacity of one such reactor is between 1.5 and 1.7  104 bbl d 1 (7.5–8.5  105 t y 1), with internal diameter and height of 10 m and 45 m, respectively [114]. Even though small catalyst particles with sizes below B2  10 4 m are used, accumulation of waxy hydrocarbons within their pores can slow down intraparticle diffusion and the overall process. Inhibition in Co-driven LTFT is also caused by sulfurcontaining compounds in the feed and by water vapor produced during FT synthesis. The former acts as a poison and permanently blocks the active sites. Water vapor, however, can cause either sintering or oxidation of Co particles on the surface, or lead to the formation of inactive cobalt-support species such as cobalt silicate and cobalt aluminate [113]. Depending on its partial pressure, deactivation caused by water vapor can either be reversed or remain irreversible. Extent of deactivation is also correlated with the degree of dispersion of Co particles [115]. In contrast with HTFT (Section 2.16.3.2.6), coke formation is unlikely to occur in LTFT due to the absence of aromatics in the waxy hydrocarbon mixture. Moreover, low temperatures together with the suppression of

520

Catalysts

Light products Fresh catalyst

Feed water Steam

To wax/solid separation

Syngas

Fig. 25 Slurry bubble column reactor configuration for LTFT. Reproduced from Wang T, Wang J, Jin Y. Slurry reactors for gas-to-liquid processes: a review. Ind Eng Chem Res 2007;46:5824–47. Copyright (2007), with permission from American Chemical Society.

water–gas shift over the Co catalyst hinder the formation of CO2 and its further conversion into elemental carbon by the Boudouard reaction. An alternative reactor option for Co-based LTFT is the multitubular packed-bed configuration discussed in Section 2.16.3.1.1.1. This option is presently employed by Shell in their Bintulu Plant (Malaysia) and Pearl GTL complex (Ras Laffan, Qatar). The number of tubes exceeds 10,000 in a single reactor, which has a capacity in the range of 7–9  103 bbl d 1 [101,116]. Slurry bubble and packed-bed configurations have several unique advantages and drawbacks. Compared to multitubular PBRs, the slurry reactor is capable of operating with (1) up to 4 times lower pressure drop, (2) higher heat and mass transport coefficients due to intense mixing, (3) faster heat removal providing better regulation of reaction temperature to tune conversion and product distribution, and (4) smaller catalyst particles to obtain higher reaction rates and to reduce diffusion limitations within the pores. In addition to these operational benefits, slurry reactors are less capital intensive. Multitubular PBR operation, on the other hand, is free from the complexities associated with wax–solid separation. Moreover, the impact of sulfur poisoning is not immediate as in slurry bubble units due to the stationary position of the catalyst particles, which also eliminates their motion-based, mechanical erosion. Even though catalyst particle sizes are larger, the corresponding increase in the depth of the pores elongates residence of the reactive flow within them, favoring production of longer-chain hydrocarbons. This feature, together with reduced CO2 selectivity, is reported to improve carbon efficiency in multitubular PBRs [101].

2.16.4

Future Directions and Closing Remarks

Due to stringent environmental regulations and increasing demand in high-quality/low-cost production practices, combination of high efficiency and low emissions of hazardous species is becoming a serious requirement in the manufacturing operations of industrial chemicals. Owing to the fact that the reaction step in chemical processes dictates the efficiency and emission characteristics, design and development of catalysts capable of delivering high activity with desired product selectivity at lower temperatures, and of novel reactor units that enable effective utilization of catalysts, seem to be the strategies for enabling profitable operation with reduced environmental footprints. In this respect, synthesis techniques that give catalysts with welldispersed active phase at very fine, nanometer-scale dimensions will gain importance. This will not only increase the surface area of the active phase and improve activity, but also reduce the cost of catalyst as it becomes possible to obtain the same conversion levels with lower loading of expensive metals. In an effort to increase selectivity based on the shape and size of the molecules, the use of ordered porous materials in catalysis is expected to gain further interest. Among various classes of porous materials, MOFs turn out to be interesting options, as they can easily be synthesized to have desired functional properties with very high surface areas, and the number of studies reporting their applications in catalysis is limited. In addition to the efforts for boosting activity and selectivity, studies addressing stability of the catalysts is expected to gain importance. Considering the high cost of catalysts

Catalysts

521

and costs associated with catalyst changeover (i.e., costs of reduced reactor throughput, and materials and labor associated with removal of the old batch of catalyst from and addition of new batch to the reactor), catalyst stability becomes ultimately important for increasing profitability. It should be noted that the improvements stated above become meaningful when the reactor is capable of providing conditions needed for ideal operation of the catalysts. In this respect, rates of heat addition to/removal from the reactor should be high to approach the state of isothermal operation. Under ideal conditions, this will allow existence of a unique temperature at which the catalyst maximizes the rate of the desired reaction toward the production of target species. Reactor architecture should also ensure good mixing of the reactants and homogeneous contact of the reactive flow at the desired composition with the complete catalyst bed, and should favor conditions that allow minimization of transport limitations that reduce effectiveness of the catalyst. Novel reactors capable of meeting such functional requirements, such as microchannel units that offer very high heat and mass transport rates together with elevated volumetric productivities, are expected to receive increasing interest. Multifunctional, intensified reactors are also anticipated to find increased use in the industry, as they can deliver attractive benefits such as increased product selectivity (such as in membrane reactors) or isothermal operation (such as in heat-exchange integrated reactors). Nevertheless, use of novel reactors at commercial scale should be considered together with the challenges and demands associated with the real operating conditions. The present chapter is aimed at providing the readers with a holistic overview of the solid catalysts. After introducing their significance, the fundamental phenomena involved in the operation of solid catalysts, together with the major metrics used in their quantification, are discussed in-depth in the first part. The second part of this chapter is devoted to the description of materials, methods, and processes involved in the synthesis and bulk production of catalysts, and to the discussion of phenomena that adversely affect catalytic operation. Owing to their critical role in the efficient utilization of solid catalysts, the available and novel reactor architectures are explained in the third section of this chapter by stating their individual operating characteristics, advantages, and drawbacks. This section is concluded with cases including actual and challenging applications of solid catalysts and catalytic reactors in chemicals manufacturing.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35]

Le Page JF. Applied heterogeneous catalysis: design, manufacture, and use of solid catalysts. Paris: Éditions Technip; 1987. Bartholomew CH, Farrauto RJ. Fundamentals of industrial catalytic processes. 2nd ed. Hoboken, NJ: John Wiley & Sons, Inc.; 2005. Hill CG, Root TW. Introduction to chemical engineering kinetics and reactor design. 2nd ed. Hoboken, NJ: John Wiley & Sons, Inc; 2014. Taylor HS. A theory of the catalytic surface. Proc R Soc Lond Ser A 1925;108:105. Thomas JM, Thomas WJ. Principles and practice of heterogeneous catalysis. Weinheim: Wiley-VCH; 1996. Levenspiel O. Chemical reaction engineering. 3rd ed New York, NY: John Wiley & Sons, Inc; 1998. Rothenberg G. Catalysis: concepts and green applications. Weinheim: Wiley-VCH; 2008. Hagen J. Industrial catalysis: a practical approach. 3rd ed Weinheim: Wiley-VCH; 2015. Satterfield CN. Heterogeneous catalysis in practice. 2nd ed New York NY: McGraw-Hill; 1991. Tanabe K, Holderich WF. Industrial application of solid acid-base catalysts. Appl Catal A – Gen 1999;181:399–434. Masel RI. Principles of adsorption and reaction on solid surfaces. New York, NY, Canada: John Wiley & Sons, Inc; 1996. Trimm DL. Design of industrial catalysts. Amsterdam: Elsevier Scientific Pub. Co.; 1980. Beller M, Renken A, van Santen RA. Catalysis: from principles to applications. Weinheim: Wiley-VCH; 2012. Pera-Titus M. Thermodynamic analysis of type VI adsorption isotherms in MFI zeolites. J Phys Chem C 2011;115:3346–57. Pickett G. Modification of the Brunauer–Emmett–Teller theory of multimolecular adsorption. J Am Chem Soc 1945;67:1958–62. Smith JM. Chemical engineering kinetics. 3rd ed New York, NY: McGraw-Hill; 1981. Onsan ZI, Avci AK. Multiphase catalytic reactors: theory, design, manufacturing, and applications. Hoboken, NJ: Wiley; 2016. Brunauer S. The adsorption of gases and vapors: physical adsorption. London: Princeton University Press; 1943. Brunauer S, Deming LS, Deming WE, Teller E. On a theory of the van der Waals adsorption of gases. J Am Chem Soc 1940;62:1723–32. Anderson JR. Structure of metallic catalysts. New York, NY: Academic Press; 1975. Aksoylu AE, Onsan ZI. Hydrogenation of carbon oxides using coprecipitated and impregnated Ni/Al2O3 catalysts,. Appl Catal A: Gen 1997;164:1–11. Akin AN, Kilaz G, Isli AI, Onsan ZI. Development and characterization of Pt-SnO2/gamma-Al2O3 catalysts. Chem Eng Sci 2001;56:881–8. Petersen EE. Chemical reaction analysis. Englewood Cliffs, NJ: Prentice-Hall; 1965. Carberry JJ. Chemical and catalytic reaction engineering. New York, NY: McGraw-Hill; 1976. Froment GF, Bischoff KB, De Wilde J. Chemical reactor analysis and design. 3rd ed. New York, NY: John Wiley & Sons, Inc; 2011. Roberts GW. Chemical reactions and chemical reactors. New York, NY: John Wiley & Sons, Inc; 2008. Onsan ZI, Avci AK. Reactor design for fuel processing. In: Shekhawat D, Spivey JJ, Berry DA, editors. Fuel cells: Technologies for fuel processing. Amsterdam: Elsevier Science; 2011. p. 451–516. Ertl G, Knözinger H, Weitkamp J. Preparation of solid catalysts. Weinheim: Wiley-VCH; 2008. Prasad R, Singh P. Applications and preparation methods of copper chromite catalysts: A review. Bull Chem Reaction Eng Catal 2011;6:63–113. Seader JD, Henley EJ, Roper DK. Separation Process Principles with Applications Using Process Simulators. 4th ed Hoboken, New Jersey: Wiley; 2016. Delange RSA, Keizer K, Burggraaf AJ. Analysis and theory of gas-transport in microporous sol-gel derived ceramic membranes. J Membr Sci 1995;104:81–100. Guo YG, Hu YS, Sigle W, Maier J. Superior electrode performance of nanostructured mesoporous TiO2 (anatase) through efficient hierarchical mixed conducting networks. Adv Mater 2007;19:2087–91. Xu JL, Li K, Shi WY, Li RJ, Peng TY. Rice-like brookite titania as an efficient scattering layer for nanosized anatase titania film-based dye-sensitized solar cells. Journal of Power Sources 2014;260:233–42. Geus JW, van Dillen AJ. Preparation of supported catalysts by deposition–precipitation. In: Ertl G, Knozinger H, Schüth F, Weitkamp J, editors. Handbook of Heterogeneous Catalysis. Weinheim: Wiley-VCH; 2008. p. 428–67. Schlögl R. Carbons. In: Ertl G, Knozinger H, Schüth F, Weitkamp J, editors. Handbook of Heterogeneous Catalysis. Weinheim: Wiley-VCH; 2008. p. 357–427.

522

Catalysts

[36] Hunt JE, Xu LQ, Winans RE, Seifert S. Thermal degradation of M41S-class mesoporous sieves as revealed by small-angle X-ray scattering. Abstr Papers Am Chem Soc 2000;219:U681. [37] Flanigen EM, Broach RW, Wilson ST. Introduction. In: Kulprathipanja S, editor. Zeolites in Industrial Separation and Catalysis. Weinheim: Wiley-VCH; 2010. p. 1–26. [38] Lehman SE, Larsen SC. Zeolite and mesoporous silica nanomaterials: greener syntheses, environmental applications and biological toxicity. Environ Sci – Nano 2014;1:200–13. [39] Yilmaz B, Muller U. Catalytic applications of zeolites in chemical industry. Top Catal 2009;52:888–95. [40] Giraldo LF, Lopez BL, Perez L, et al. Mesoporous silica applications. Macromol Symp 2007;258:129–41. [41] Zhou HC, Long JR, Yaghi OM. Introduction to metal-organic frameworks. Chem Rev 2012;112:673–4. [42] Kaye SS, Dailly A, Yaghi OM, Long JR. Impact of preparation and handling on the hydrogen storage properties of Zn4O(1,4-benzenedicarboxylate)(3) (MOF-5). J Am Chem Society 2007;129:14176–7. [43] Rowsell JLC, Yaghi OM. Metal-organic frameworks: a new class of porous materials. Microporous Mesoporous Mater 2004;73:3–14. [44] Farrusseng D, Aguado S, Pinel C. Metal-organic frameworks: opportunities for catalysis. Angew Chem – Int Ed 2009;48:7502–13. [45] Kuppler RJ, Timmons DJ, Fang QR, et al. Potential applications of metal-organic frameworks. Coord Chem Rev 2009;253:3042–66. [46] Suh MP, Park HJ, Prasad TK, Lim DW. Hydrogen storage in metal-organic frameworks. Chem Rev 2012;112:782–835. [47] Sumida K, Rogow DL, Mason JA, et al. Carbon dioxide capture in metal-organic frameworks. Chem Rev 2012;112:724–81. [48] Ma SQ, Sun DF, Simmons JM, et al. Metal-organic framework from an anthracene derivative containing nanoscopic cages exhibiting high methane uptake. J Am Chem Soc 2008;130:1012–6. [49] Li JR, Kuppler RJ, Zhou HC. Selective gas adsorption and separation in metal-organic frameworks. Chem Soc Rev 2009;38:1477–504. [50] Adatoz E, Avci AK, Keskin S. Opportunities and challenges of MOF-based membranes in gas separations. Sep Purif Technol 2015;152:207–37. [51] Low JJ, Benin AI, Jakubczak P, et al. Virtual high throughput screening confirmed experimentally: porous coordination polymer hydration. J Am Chem Soc 2009;131:15834–42. [52] Huang YBA, Zheng ZL, Liu TF, et al. Palladium nanoparticles supported on amino functionalized metal-organic frameworks as highly active catalysts for the SuzukiMiyaura cross-coupling reaction. Catal Commun 2011;14:27–31. [53] Park KS, Ni Z, Cote AP, et al. Exceptional chemical and thermal stability of zeolitic imidazolate frameworks. Proc Natl Acad Sci USA 2006;103:10186–91. [54] Iijima S. Carbon nanotubes: past, present, and future. Phys B – Condens Matter 2002;323:1–5. [55] Serp P, Castillejos E. Catalysis in carbon nanotubes. ChemCatChem 2010;2:41–7. [56] Moulijn JA, van Diepen AE, Kapteijn F. Deactivation and regeneration. In: Ertl G, Knozinger H, Schüth F, Weitkamp J, editors. Handbook of heterogeneous catalysis. Weinheim: Wiley-VCH; 2008. p. 1828–46. [57] Bartholomew CH. Mechanisms of catalyst deactivation,. Applied Catalysis A: General 2001;212:17–60. [58] Trimm DL. The regeneration of disposal of deactivated heterogeneous catalysts,. Applied Catalysis A: General 2001;212:153–60. [59] Trimm DL, Onsan ZI. Onboard fuel conversion for hydrogen-fuel-cell-driven vehicles. Catal Rev – Sci Eng 2001;43:31–84. [60] Onsan ZI. Catalytic processes for clean hydrogen production from hydrocarbons. Turkish J Chem 2007;31:531–50. [61] Urban PM, Funke A, Muller JT, Himmen M, Docter A. Catalytic processes in solid polymer electrolyte fuel cell systems. Appl Catal A: Gen 2001;221:459–70. [62] Trimm DL. Catalysts for the control of coking during steam reforming. Catal Today 1999;49:3–10. [63] Rostrup-Nielsen JR. Steam reforming. In: Ertl G, Knozinger H, Schüth F, Weitkamp J, editors. Handbook of heterogeneous catalysis. Weinheim: Wiley-VCH; 2008. p. 2882–905. [64] Rostrup-Nielsen JR. Catalytic steam reforming. In: Anderson JR, Boudart M, editors. Catalysis – Science and Technology. Berlin: Springer-Verlag; 1984. p. 1–117. [65] Rostrup-Nielsen JR. Industrial relevance of coking. Catalysis Today 1997;37:225–32. [66] Rostrup-Nielsen JR. New aspects of syngas production and use. Catalysis Today 2000;63:159–64. [67] Borowiecki T, Golebiowski A, Stasinska B. Effects of small MoO3 additions on the properties of nickel catalysts for the steam reforming of hydrocarbons,. Applied Catalysis A: General 1997;153:141–56. [68] Horiuchi T, Sakuma K, Fukui T, et al. Suppression of carbon deposition in the CO2-reforming of CH4 by adding basic metal oxides to a Ni/Al2O3 catalyst, Applied. Catalysis A: General 1996;144:111–20. [69] Wang X, Gorte RJ. Steam reforming of n-butane on Pd/ceria. Catal Lett 2001;73:15–9. [70] Aasberg-Petersen K, Hansen JHB, Christensen TS, et al. Technologies for large-scale gas conversion. Appl Catal A: Gen 2001;221:379–87. [71] Arai H, Machida M. Thermal stabilization of catalyst supports and their application to high-temperature catalytic combustion. Appl Catal A: Gen 1996;138:161–76. [72] Eigenberger G, Ruppel W. Catalytic fixed-bed reactors. In: Elvers B, editor. Ullmann’s reaction engineering. Weinheim: Wiley-VCH; 2013. p. 305–70. [73] Gallei EF, Hesse M, Schwab E. Development of industrial catalysts. In: Ertl G, Knozinger H, Schüth F, Weitkamp J, editors. Handbook of heterogeneous catalysis. Weinheim: Wiley-VCH; 2008. p. 57–66. [74] Eigenberger G. Catalytic fixed-bed reactors. In: Ertl G, Knozinger H, Schüth F, Weitkamp J, editors. Handbook of heterogeneous catalysis. Weinheim: Wiley-VCH; 2008. p. 2075–106. [75] Walas SM. Chemical process equipment. Boston, MA: Butterworth-Heinemann; 1990. [76] Westerterp KR, Wammes WJA. Three-phase trickle-bed reactors. Ullmann’s reaction engineering 2013;Weinheim: Wiley-VCH; 2013. p. 651–63. [77] T. Boger, A.K. Heibel, C.M. Sorensen. Monolithic catalysts for the chemical industry. 2004. [78] Meille V. Review on methods to deposit catalysts on structured surfaces. Appl Catal A: Gen 2006;315:1–17. [79] Delparish A, Avci AK. Intensified catalytic reactors for Fischer–Tropsch synthesis and for reforming of renewable fuels to hydrogen and synthesis gas. Fuel Process. Technol. 2016;151:72–100. [80] Kiwi-Minsker L, Renken A. Microstructured reactors for catalytic reactions. Catalysis Today 2005;110:2–14. [81] Kolb G, Hessel V. Micro-structured reactors for gas phase reactions. Chem Eng J 2004;98:1–38. [82] Ehrfeld W, Hessel V, Löwe H. Microreactors: new technology for modern chemistry. Weinheim: Wiley-VCH; 2000. [83] Karakaya M, Avci AK. Microchannel reactor modeling for combustion driven reforming of iso-octane. Int J Hydrogen Energy 2011;36:6569–77. [84] Karakaya M, Avci AK. Simulation of on-board fuel conversion in catalytic microchannel reactor-heat exchanger systems. Top Catal 2009;52:2112–6. [85] Avci AK, Trimm DL, Karakaya M. Microreactor catalytic combustion for chemicals processing. Catal Today 2010;155:66–74. [86] Tezcan I, Avci AK. Parametric investigation of oxidative coupling of methane in a heat-exchange integrated microchannel reactor. J Chem Technol Biotechnol 2015;90:1827–38. [87] Sen I, Avci AK. Simulation of exhaust gas reforming of propane in a heat exchange integrated microchannel reactor. Int J Hydrogen Energy 2014;39:844–52. [88] Hosukoglu MI, Karakaya M, Avci AK. Modeling and simulation of hydrocracking of Fischer–Tropsch hydrocarbons in a catalytic microchannel reactor. Ind Eng Chem Res 2012;51:8913–21. [89] Gumuslu G, Avci AK. Parametric analysis of Fischer–Tropsch synthesis in a catalytic microchannel reactor. AIChE J 2012;58:227–35. [90] Tan XY, Li K. Membrane microreactors for catalytic reactions. J Chem Technol Biotechnol 2013;88:1771–9. [91] Kolb G. Review: Microstructured reactors for distributed and renewable production of fuels and electrical energy. Chem Eng Process: Process Intensification 2013;65:1–44. [92] Powell JB. Application of multiphase reaction engineering and process intensification to the challenges of sustainable future energy and chemicals. Chem Eng Sci 2017;157:15–25.

Catalysts

523

[93] Wang T, Wang J, Jin Y. Slurry reactors for gas-to-liquid processes: a review. Ind Eng Chem Res 2007;46:5824–47. [94] Dybkjaer I. Tubular reforming and autothermal reforming of natural-gas – an overview of available processes. Fuel Process Technol 1995;42:85–107. [95] Aasberg-Petersen K, Christensen TS, Dybkjaer I, et al. Synthesis gas production for FT synthesis. In: Steynberg AP, Dry ME, editors. Fischer–Tropsch technology. Amsterdam: Elsevier; 2004. p. 258–405. [96] Nawaz Z. Heterogeneous reactor modeling of an industrial multitubular packed-bed ethylene oxide reactor. Chem Eng Technol 2016;39:1845–57. [97] Boskovic G, Wolf D, Bruckner A, Baerns M. Deactivation of a commercial catalyst in the epoxidation of ethylene to ethylene oxide-basis for accelerated testing. J Catal 2004;224:187–96. [98] Jennings JR, Ward SA. Ammonia synthesis. In: Twigg MV, editor. Catalyst handbook. London: Wolfe Publishing Ltd.; 1989. p. 384–439. [99] Liu HZ. Ammonia synthesis catalyst 100 years: practice, enlightenment and challenge. Chin J Catal 2014;35:1619–40. [100] Dingerdissen U, Martin A, Herein D, Wernicke HJ. The development of industrial heterogeneous catalysis. In: Ertl G, Knozinger H, Schüth F, Weitkamp J, editors. Handbook of heterogeneous catalysis. Weinheim: Wiley-VCH; 2008. p. 37–56. [101] Remans TJ, Jenzer G, Hoek A. Gas-to-liquids. In: Ertl G, Knozinger H, Schüth F, Weitkamp J, editors. Handbook of heterogeneous catalysis. Weinheim: Wiley-VCH; 2008. p. 2994–3010. [102] Aasberg-Petersen K, Christensen TS, Nielsen CS, Dybkjaer I. Recent developments in autothermal reforming and pre-reforming for synthesis gas production in GTL applications. Fuel Process Technol 2003;83:253–61. [103] Hickman DA, Schmidt LD. Synthesis gas-formation by direct oxidation of methane over Pt monoliths. J Catal 1992;138:267–82. [104] Hickman DA, Schmidt LD. Production of syngas by direct catalytic oxidation of methane. Science 1993;259:343–6. [105] Basile F, Fornasari G, Trifiro F, Vaccari A. Partial oxidation of methane – effect of reaction parameters and catalyst composition on the thermal profile and heat distribution. Catal Today 2001;64:21–30. [106] Boucouvalas Y, Zhang ZL, Verykios XE. Partial oxidation of methane to synthesis gas via the direct reaction scheme over Ru/TiO2 catalyst. Catal Lett 1996;40:189–95. [107] Tsipouriari VA, Zhang Z, Verykios XE. Catalytic partial oxidation of methane to synthesis gas over Ni-based catalysts – I. Catalyst performance characteristics. J Catal 1998;179:283–91. [108] Dissanayake D, Rosynek MP, Kharas KCC, Lunsford JH. Partial oxidation of methane to carbon monoxide and hydrogen over a Ni/Al2O3 catalyst. J Catal 1991;132:117–27. [109] Torniainen PM, Chu X, Schmidt LD. Comparison of monolith-supported metals for the direct oxidation of methane to syngas. J Catal 1994;146:1–10. [110] Nakagawa K, Ikenaga N, Teng YH, Kobayashi T, Suzuki T. Transient response of catalyst bed temperature in the pulsed reaction of partial oxidation of methane to synthesis gas over supported rhodium and iridium catalysts. J Catal 1999;186:405–13. [111] Steynberg AP, Espinoza RL, Jager B, Vosloo AC. High temperature Fischer-Tropsch synthesis in commercial practice. Appl Catal A: Gen 1999;186:41–54. [112] Steynberg AP, Dry ME, Davis BH, Breman BB. Fischer–Tropsch reactors. In: Steynberg AP, Dry ME, editors. Fischer-Tropsch technology. Amsterdam: Elsevier; 2004. p. 64–195. [113] Dry ME. FT catalysts. In: Steynberg AP, Dry ME, editors. Fischer–Tropsch technology. Amsterdam: Elsevier; 2004. p. 533–600. [114] de Klerk A, Li Y-W, Zennaro R. Fischer–Tropsch technology. In: Maitlis PM, de Klerk A, editors. Greener Fischer–Tropsch processes for fuels and feedstocks. Weinheim: Wiley-VCH; 2013. p. 53–79. [115] Niemela MK, Krause AOI. The long-term performance of Co/SiO2 catalysts in CO hydrogenation. Catal Lett 1996;42:161–6. [116] Suarez Paris R, Lopez L, Barrientos J, et al. Catalytic conversion of biomass-derived synthesis gas to fuels. In: Spivey JJ, Han YF, Dooley KM, editors. Catalysis. Cambridge: Royal Society of Chemistry; 2015. p. 62–143.

2.17 Photoactive Materials Canan Acar, Bahcesehir University, Bes¸iktas¸, Istanbul, Turkey Ibrahim Dincer, University of Ontario Institute of Technology, Oshawa, ON, Canada r 2018 Elsevier Inc. All rights reserved.

2.17.1 Introduction 2.17.2 Background 2.17.3 Photoactive Materials 2.17.4 Photocatalytic Water Splitting 2.17.4.1 Heterogeneous Photocatalysts 2.17.4.2 Addition of Cocatalysts 2.17.4.3 Electron Donors and Electrolytes 2.17.5 Performance Comparison Criteria for Photoactive Materials 2.17.5.1 Technical Comparison Criteria (n_m [µmol/h-gcat], n_A [mmol/h-m2cat], Eg [V]) 2.17.5.2 Cost Comparison Criteria 2.17.5.3 Environmental and Health Impact Comparison Criteria 2.17.5.4 Overall Comparison 2.17.6 Case Study: Comparative Assessment of Photocatalysts 2.17.6.1 Photocatalyst Groups Comparison 2.17.6.2 Individual Photocatalyst Comparison 2.17.6.3 Case Study Conclusions 2.17.7 Photoactive Materials in Photoelectrochemical Systems 2.17.7.1 Photoelectrochemical Hydrogen Production 2.17.7.2 Photoelectrodes 2.17.8 Photoelectrode Coating Methods by Using Photoactive Materials 2.17.8.1 Chemical Vapor Deposition 2.17.8.2 Electrochemical Deposition and Electrodeposition 2.17.8.3 Sol–Gel 2.17.8.4 Spin Coating 2.17.8.5 Spray Pyrolysis 2.17.8.6 Overall Comparison 2.17.8.7 Key Findings 2.17.9 Hydrogen Production Methods 2.17.10 Photoactive Materials in Hydrogen Production 2.17.11 Photoactive Materials in Water Splitting Reactions 2.17.12 Recycling Photoactive Materials 2.17.13 Case Study: Comparative Assessment of Photoactive Materials 2.17.13.1 Titanium Oxide-Based Photocatalysts 2.17.13.2 Cadmium Sulfide-Based Photocatalysts 2.17.13.3 Zinc Oxide and Sulfide-Based Photocatalysts 2.17.13.4 Other Metal Oxide-Based Photocatalysts 2.17.13.5 Overall Comparison 2.17.13.6 Case Study Conclusions 2.17.14 Future Directions 2.17.15 Closing Remarks References Further Reading Relevant Websites

Nomenclature A E0 Eg G h

524

Area (m2) Potential (V) Band gap energy (eV) Gibbs free energy (kJ) Planck’s constant (B6.63  10–34 m2-kg/s)

J n_ P_ V

2 4 5 8 8 10 11 13 13 13 13 14 14 15 16 17 17 19 20 20 20 21 21 24 24 25 26 27 31 32 36 37 41 41 41 42 42 43 45 46 46 48 49

Current density (A/m2) Molar rate of generation (mmol/h) or molar flow rate (mol/s) Power density (W/m2) Potential (V)

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00236-4

Photoactive Materials

Greek Symbols D Finite change in quantity

l n

Wavelength (nm) Frequency (1/s)

Subscript and Superscripts A Per m2 catalyst surface area cat Catalyst g Band gap

m OX RED

Per gram catalyst Oxidization potential Redox potential

Acronyms BET CB CSP CVD e ED hþ HFSS HMIS

IR NHE NREL PEC PV SC SG SP UV VB

Infrared Normal hydrogen electrode National Renewable Energy Laboratory Photoelectrochemical Photovoltaic Spin coating Sol–gel Spray pyrolysis Ultraviolet Valance band

2.17.1

Brunauer Emmett Teller surface Conduction band Concentrated solar power Chemical vapor deposition Electron Electron donor or electrodeposition Hole High flux solar simulator Hazardous materials identification system

525

Introduction

Ever since the industrial revolution, global energy demand has been increasing progressively due to the worldwide growth in population and living standards. Scientists, industrial experts, and governments foresee that the worldwide energy demand, and as a result energy processing capacities, will keep increasing in the future as well. In conventional energy conversion and processing systems, fossil fuels are heavily utilized to meet this demand. However, fossil fuels emit large quantities of polluting exhaust gases into the atmosphere, as well as into land and water resources. In this perspective, environmental and health-related problems as a result of CO2, NOx, and SOx emissions from fossil fuel processes; limited and nonhomogeneous reserves of fossil fuels; and political risks related to dependence on imported fossil fuels highlight the importance of renewable energies to substitute fossil fuels when meeting the increasing global energy demand [1]. Solar irradiation incoming to Earth is a renewable, reliable, and abundant energy supply. Therefore, it can turn out to be a potential sustainable solution to the increasing energy demand of the world. About 30 min of solar radiation incident on the Earth’s surface contains as much energy as the world consumption for 1 year [2]. Another advantage of solar energy is its relatively low gradual system expansion cost compared to conventional fuels. Fig. 1 presents how photoactive materials assist in harvesting solar energy and providing the necessary energy for different end use purposes. In order to effectively capture and utilize solar energy, the following issues need to be addressed. The first issue is the low density of solar radiation per unit of earth surface. The solar radiation incident intensity on the Earth’s surface appears to be between 0.4 and 1 kW/m2 [1]. Therefore, providing energy via an established technology such as photovoltaic (PV) thermal systems would require large not-shaded areas occupied with expensive systems resulting in cost issues related to land occupancy and system hardware requirements. The second issue refers to the intermittent and fluctuant nature of solar radiation. Whence the solar energy is expected as efficiently as possibly converted into a form of energy that can be stored such that a continuous (24 h) energy demand can be met. Hydrogen, as an energy carrier, could be a potential promising candidate to address these issues for the following reasons: (1) it is the most abundant chemical element in the universe, which can be extracted from sustainable sources such as water (both sweet or salty), sulfurous aqueous waters wasted by industry, biomass, and hydrogen sulfide resources (from some oil wells, some geothermal wells, and seas); (2) it has a high energy yield per mass (122 kJ/g) compared to other fuels such as gasoline (40 kJ/g); and (3) if produced by using renewable energies, it is environmentally friendly because its end use will not produce pollutants, greenhouse gases, nor any harmful effect on the environment. Because of all these reasons, hydrogen is considered to be a clean energy carrier that could address the current energy and environmental issues. Last, but not least, hydrogen can be stored in gaseous, liquid, or metal hydride form and can be distributed over large distances through pipelines or via tankers. To date, there have been many studies discussing the potential key role of hydrogen on sustainable development [3]. The abundant supply of water and sun offers an affordable alternative source to generate hydrogen. Its production via solar water splitting generally can be categorized into three types: (1) thermochemical water splitting, (2) photobiological water splitting, and (3) photocatalytic water splitting. Thermochemical water splitting appears to be a simple way to produce hydrogen. However, heat management/control and the search for appropriate heat-resisting materials are the greatest challenges of this

526

Photoactive Materials

Source: Sunlight

., P

olt aic s

e.g lys

e.g .

tro

,P

lec

toe

ho tov

ho

Electricity: carrier

e.g., Fuel cells/electrolysis

is

Photoactive materials

Hydrogen : storage and fuel

Fig. 1 Schematic illustration of harvesting solar energy via photoactive materials.

method. Furthermore, large scale solar concentrator systems are needed in this method to achieve the high temperature requirements, causing high costs and low efficiencies [4]. Photobiological water splitting as an alternative hydrogen production method is divided into two groups. Under light irradiation, hydrogen production in anaerobic conditions is referred to as water biophotolysis, while that in anaerobic conditions is referred to as organic biophotolysis. Organic biophotolysis is capable of decomposing organic wastes to provide a high hydrogen yield. However, it generates CO2 as the byproduct. In water biophotolysis, on the other hand, water is transformed into H2 and O2 in the presence of light by bacteria or green algae with the help of special enzymes. Although it is considered cleaner compared to organic biophotolysis, it still has many problems waiting to be solved, including low H2 yield, the poisoning effect of enzymes under the existence of O2 (generated simultaneously during biophotolysis), and the difficulty in designing and scaling up the bioreactor for the process [5]. Photocatalytic water splitting is one of the candidates for “clean” solar hydrogen production, with potential for application from small scale to large scale hydrogen generators. Hydrogen production via photocatalytic water splitting is similar to the photosynthetic reaction in many ways. However, compared to thermochemical and photobiological water splitting techniques, it has the following advantages: (1) reasonable solar to H2 efficiency, (2) low process cost because reasonable cheap catalysts can be selected, (3) the ability to achieve separate H2 and O2 evolution during reaction, and (4) small reactor systems suitable for residential applications can be mass produced, thus providing for a huge market potential. Photocatalytic water splitting also has the potential to eliminate the need of PV panels, therefore it is seen as a promising candidate to produce inexpensive and sustainable hydrogen [6]. After Fujishima and Honda’s study [2] on photoelectrochemical (PEC) solar energy-based water splitting using TiO2 as photocatalyst, photocatalytic water splitting started to attract the attention of scientists as a clean and sustainable method for producing hydrogen. Overall water splitting using particulate photocatalytic systems, with particle sizes of several hundred nanometers to a few micrometers was first demonstrated unequivocally in 1980. The reaction has been studied extensively by many researchers since then. In 2008, MacDonnell reviewed some past studies on photoreactions generating hydrogen. Also, various PEC and photocatalytic systems for hydrogen production are reviewed extensively by Zamfirescu et al. [3]. Over the past decades, substantial progress has been accomplished on semiconductor (SC)-based photocatalytic hydrogen generation through water splitting and many reviews have been published [7]. In this chapter, numerous photoactive materials that are actively being investigated in the literature are studied, discussed, and comparatively assessed based on criteria of cost, environmental impact, and technical performance. A photocatalyst database is produced to identify possible directions for future research. Various case studies are provided regarding photon based water splitting, photocatalyst activities, and photoelectrode coating. In a photocatalyst comparison case study, 43 heterogeneous photocatalysts and seven photocatalyst groups are ranked and compared based on environmental and health impact, cost, and efficiency criteria. The photoactive materials investigated in this chapter are selected based on the following criteria: publication dates (the more recent publications are followed to identify the research directions), publication numbers (the ones that are heavily studied in the literature have been taken into account), and availability of data such as hydrogen production rates, the

Photoactive Materials

527

amounts of catalysts used and their BET surface areas, operating conditions, and precursors used to manufacture these photocatalysts in order to establish a complete comparative assessment.

2.17.2

Background

Being unlimited, renewable, and free, solar energy is capable of producing electricity or heat without the requirements of having turbines and maintenance. The energy usage of 1 year could be provided by half an hour of solar irradiation on the earth’s surface. However, sunlight is an intermittent source of energy, which limits the amount of solar radiation due to its dependence on geographical position, day, time, and season. Another disadvantage of solar energy is its low density per unit of earth surface. Therefore, developing a source of energy that is storable, clean, continuous, and renewable, is required to meet global energy demand. Hydrogen is an advantageous fuel for (1) being abundant from various sustainable sources (biomass or water), (2) having high energy yield, (3) being environmentally friendly, and (4) having high storage capability, thus it is considered as an ideal alternative source of energy for fossil fuels [8]. Steam methane reforming is a widely used technique to produce hydrogen from natural gas at high temperatures (up to 9001C) and pressures (1.5–3 MPa). Coal gasification is also employed to generate hydrogen through partial oxidation at high temperatures and pressures (up to 5 MPa). Biomass materials such as crops, animal wastes, and plants under thermochemical routes generate hydrogen through pyrolysis and gasification, which produce byproducts of CO, CO2, and methane. Biological processes for hydrogen production from biomass materials are other promising techniques but they are not economically feasible yet. Consequently, current hydrogen generation techniques suffer from a reliance on fossil fuel sources, harsh process conditions, and significant costs. Alternative methods that utilize renewable sources of energy for hydrogen production such as hydropower, wind power, and sunlight must be explored. Among these sustainable energies, solar energy has been considered a more promising source due to its lesser location dependence in comparison to wind and hydropower energy [9]. The combination of solar energy with plentiful water resources provides a reasonable platform for hydrogen generation, which is called solar water splitting. There are three approaches to splitting water using solar energy: (1) thermochemical water splitting, (2) photobiological water splitting, and (3) photocatalytic water splitting. Although the thermochemical approach is the simplest, the requirement for large solar concentrators makes this method highly expensive and less favorable. Biophotolysis can be divided into water biophotolysis and organic biophotolysis depending on the microorganism type, product, and mechanisms of the reaction. Although water biophotolysis is cleaner than organic biophotolysis (regarding CO2 emissions), low yields of hydrogen production, toxic effects of enzymes, and limitations on scaling up the process exist. Photocatalytic water splitting possesses several advantages over thermochemical and photobiological water splitting including: (1) low cost (capable of reducing the PV arrays); (2) relatively high solar-to-H2 efficiency; (3) capability of separating H2 and O2 streams; and (4) flexible reactor size, which is appropriate for small scale usage. The US Department of Energy (DOE) has established the ultimate target of 26% for the solar to hydrogen energy conversion ratio, which will require aggressive research to improve the current status [10]. Fig. 2 shows the classification of photoactive materials. Heterogeneous photoactive systems consist of three components: a catalyst, visible light absorber, and sacrificial electron donor. Although the basic principles of photochemical and PEC systems are identical, they differ in their setup. In photochemical reactions, there is a SC–electrolyte junction at which the water splitting reaction takes place. The required potential for water splitting is generated at the SC–liquid interface. The SC should be stable in

Photoactive materials

Homogeneous materials

Heterogeneous materials

Type I

Type II

Photoelectrochemical

Electrochemical

Molecular catalysts + light absorbers + sacrificial donors (e.g., Rhenium complexes)

Catalysts + electron donors (e.g., phthalocyanin complexes)

Semiconductors (e.g., TiO2)

Photovoltaic materials

Fig. 2 Classification of photoactive materials.

528

Photoactive Materials

the electrolyte to prevent any corrosion. Depending on the band edge position of the SC, they can be active in hydrogen production, oxygen production, or overall water splitting [11]. In PEC systems, a photocatalyst, which is a SC, is irradiated by UV–visible light with energy greater or equivalent to the band gap of the SC. The light energy will be absorbed by the photocatalyst and results in charge separation at the VB and conduction band (CB). The holes are produced at the VB, and the photoexcited electrons are located in the CB. The holes trigger the oxidation of water at the surface of the CB while the photoexcited electrons at the CB reduce the absorbed H þ to H2. Mainly in PEC water splitting, SCs are applied as a photocathode or photoanode depending on the reaction, which is favored. In PEC systems, a SC electrode should be in contact with an electrolyte, which contains a redox couple. In PEC, the overall reaction takes place at two different electrodes. In this method, the potential needed for water splitting is being provided by illuminating the cathode or anode [12]. The most common experimental setup for photoactive materials used by researchers consists of a reaction cell, a gas circulation pump, vacuum pumps, and a gas chromatograph detector. The oxygen and hydrogen produced can also be detected using oxygen and hydrogen sensors, or by volumetric methods. The reaction solution should be purged with inert gases before testing, and the whole setup should be air-free to measure the amount of evolved oxygen accurately. Several light sources can be used to initiate the reaction. For photocatalysts with a UV light response, high-pressure mercury lamps are employed and the reaction cell should be quartz. For the catalysts with band gaps smaller than 3 eV, a 300-W xenon lamp and a filter are used to generate visible light. A solar simulator is also used as incident light when evaluating solar hydrogen evolution. Different types of reaction cells have been reported in the literature. Cells with two simultaneous SCs were employed in the 1970s and 1980s. Single junction cells have been reported to drive the hydrogen evolution reaction. However they are not satisfying for the overall water splitting due to insufficient photovoltage. Multijunction devices coupled with electrocatalysts could provide a large enough photovoltage to drive the process. A monolithic three-junction amorphous silicon PV cell coupled to cobalt phosphate and Ni–Zn–Mo trimetal catalyst has been reported and exhibited an efficiency of 4.7% [13].

2.17.3

Photoactive Materials

Conversion from solar energy to valuable energy carriers (i.e., hydrogen and electricity) by using photocatalytic active materials has been considered as one of the most promising steps toward generating clean and renewable alternatives for fossil fuels. In order to use solar energy more efficiently, different approaches have been employed to shift photocatalyst activity toward the visible range while retaining stability and efficiency. TiO2 as the pioneer photocatalyst also has some limitations such as wide band gap, high hydrogen overpotential, and rapid recombination of produced electron–hole pairs, which have been addressed by various methods including doping, coupling with carbon, noble metal deposition, using dyes, and surface modifications. Other metal oxides such as iron oxide, zinc oxide, and copper oxide also have been discussed, as well as metal sulfides including cadmium sulfides and zinc sulfides. In addition, nitrides and nanocomposite materials that have been used as photocatalysts for water splitting have been reviewed [14]. Since Fujishima and Honda first demonstrated that TiO2 was a promising photoanode for UV light driven photocatalytic water splitting, it has been widely studied in many photocatalytic reactions due to its chemical stability, low cost, environmentally friendly nature, and tunable electronic energy band gap [15]. The other alternative method to extend the photocatalytic activity of TiO2 to visible light region is to dope this material with carbon. Although different mechanisms have been proposed to explain this enhancement, the mechanism of synergic effect of carbon on TiO2 remains unclear. Three mechanisms have been explored to describe the synergetic effect of carbon on TiO2. The first possible mechanism is that carbon can play the role of an electron sink, which can effectively prevent the recombination process. Another mechanism proposes carbon as a photosensitizer, which can pump electrons into the TiO2 CB. Besides the proposed mechanisms, carbon can also act as a template to disperse the TiO2 particles and hinder the agglomeration of TiO2 nanoparticles [16]. Unlike nonmetal ion doping, metallic dopants usually introduce additional energetic levels in the band gap, which reduce the energy barrier and induce a new optical absorption edge. The formation of a SC–SC heterojunction can decrease the charge recombination rate by yielding long-lived electron–hole pairs. Proper band alignments allow charge transfer from one SC to another [17]. Using a sacrificial agent helps TiO2 in performing either water oxidation or reduction. The sacrificial agent reacts with one of the charge carriers while the other carrier is in charge of either oxygen or hydrogen production. Typically, sacrificial agents such as methanol, ethanol, and ethylene glycol, which have lower oxidation potentials than water, are used to inhibit the electron–hole pair recombination in TiO2. In another scenario the VB energy level of one SC is higher than the other while the CB energy level is lower than the other SC. As a result of this band gaps alignment of two SCs, a charge separation occurs and recombination process decreases [18]. In a metal SC heterojunction structure, noble metals, such as Au, Pt, Pd, and Ru, have been reported to trap photogenerated electrons due to their significant role as electron sinks. Among noble metals, Au has been studied as the preferred cocatalyst for photocatalytic hydrogen production due to its high affinity toward photogenerated electrons, high resistance to oxidation, less activity toward the side reactions of hydrogen production, and the existence of surface plasmon resonance [19]. Another method to facilitate the photocatalytic water splitting process in the TiO2 system is by structural modification. The structure of TiO2 has a significant effect on the photocatalysis performance. Other than crystallinity, the mesoporous structure of

Photoactive Materials

529

TiO2 also plays a key role in the study of photoelectrodes. Also, the morphology of the photocatalyst has a major effect on the photocatalytic activity. 1D TiO2 forms such as nanotubes, nanowires, and nanofibers have been studied and show improved photocatalytic activity. Tuning of morphology has attracted considerable attention due to the change in material morphology that can alter charge carrier diffusion pathways. Therefore, to improve the photocatalytic hydrogen evolution efficiency of TiO2, modification of its structure is relevant [20]. Other than TiO2, a number of other representative metal oxides (such as Fe2O3, WO3, ZnO, Cu2O, Al2O3, Ga2O3, Ta2O5, CoO, and ZrO2) have also been widely studied due to their stability in aqueous solution and their low cost. However, most metal oxides suffer from large band gaps limiting their ability to absorb visible light [21]. In a typical metal oxide, the VB and CB have O 2p and metal s character and therefore relatively ionic bonded materials have a large band gap: ZnO (3.4 eV), Ga2O3 (4.5 eV), Al2O3 (8.8 eV). Using transition metal cations with dn electronic configurations may help overcome this issue, Fe2O3 (B2.0 eV) and Co3O4 (B1.3 eV) have increased light absorption but lack efficient charge carrier transfer due to small polaron dominated conductivity and associated high resistivity. Using posttransition metals like PbO (2.1 eV), SnO (2.4 eV), and Bi2O3 (2.5 eV) with occupied s band leads to better photogeneration of charge carriers; however, they are indirect SCs where optical absorption band edges vary with the square root of photon energy leading to a less efficient carrier extraction process. Therefore ternary metal oxide compounds have been investigated to overcome these limitations, such as Bi20TiO32, SnNb2O6, and BiVO4. BiVO4 has been investigated for having both a low band gap (2.4–2.5 eV) and reasonable band edge alignment for the water redox potentials. Both n- and p-type semiconducting properties have been recorded by BiVO4 as well as high photon-to-current conversion efficiencies (440%) [22]. Photoactive materials can be coated on different surfaces such as panels and electrodes via a variety of methods. In this chapter, chemical vapor deposition (CVD), electrochemical deposition (ECD), electrodeposition (ED), sol–gel (SG), spin coating (SC), and spray pyrolysis (SP) are discussed in detail in the next sections. An overview of different types of photoactive material coating methods investigated in this chapter is presented in Fig. 3. Some mechanistic studies also have been conducted for water oxidation and reduction reactions. Similar to Fe2O3, WO3 has been considered as a potential photoanode material for its suitable VB position, which favors a high onset potential for water oxidation. Elsewhere, cobalt oxide (CoO) also shows photocatalytic activity toward H2 evolution. As the morphology of nanostructures can influence the band-edge positions of material, designing CoO with different morphology such as nanotubes or nanowires could provide more efficient photocatalyst materials. SrTiO3 (STO) has also been widely used for hydrogen production as a solid-state photocatalyst with a band gap of 3.2 eV, which has been explored for the overall water splitting under UV light irradiation. Since STO is active toward water splitting only in the UV region, the solar to hydrogen conversion (STH) is low. Doping methods enhance the quantum efficiency of SrTiO3 in the visible light region. Tantalum oxide (Ta2O5) has been an attractive SC for photocatalytic water splitting. Due to the wide band gap of Ta2O5 (about 4 eV), it is required to narrow the band gap with some techniques such as doping with foreign ions [23]. CdS and ZnS are the most studied metal sulfide photocatalysts in the past decades. Compared to metal oxide SCs, CdS with narrower band gap (B2.4 eV) is considered promising as a visible-light-driven photocatalyst for water splitting. However, as a result of rapid recombination of photogenerated electrons and holes, bare CdS SCs usually show low hydrogen production rates. Moreover, high activity of CdS under light irradiation leads to the corrosion of SCs. To circumvent this problem, CdS materials can

Chemical vapor deposition (CVD) Electrochemical deposition (ED)

Spray pyrolysis Photoactive material coating methods Spin coating (SC)

Electrodeposition (ED)

Sol−gel (SG)

Fig. 3 An overview of various types of photoactive material coating methods.

530

Photoactive Materials

be coupled with other noble metals as cocatalyst or form a heterojunction structure with other SCs. In such a case, the photogenerated electrons on the CB of CdS can be transferred to electronic levels of noble metals or be delocalized and transferred between the CBs of SCs. Nickel sulfide, in particular, has proven to be tremendously useful in raising the activity of SCs when used as a cocatalyst along with TiO2, CdS, and g-C3N4. Unlike CdS, wide band gap ZnS (3.6 eV) responds weakly to visible light. Efforts have been made to improve the photoactivity of ZnS for hydrogen evolution. Metal oxides with large band gaps provide higher stability for the composite material. Some of the metal oxides that have been proposed to combine with CdS are TaON, TiO2, and ZnO [24]. Different nanostructures of carbon can also be combined with CdS to promote catalytic behavior toward water splitting by preventing the charge recombination process. Due to the high conductivity of the carbon nanostructures, any contact between CdS and carbon can substantially improve the charge separation and subsequently the catalytic behavior of nanocomposites will be enhanced. Different strategies have been taken to synthesize the carbon-based CdS from simple mixing of carbon and CdS to in-situ growing of the CdS at the surface of graphene oxide using oxygen moieties as the template. WS2–Au–CuInS2 has also been developed for photocatalytic H2 production by insertion of gold nanoparticles between WS2 nanotubes and CuInS2 (CIS) nanoparticles. Introducing Au nanoparticles led to significant enhancement of light absorption. Moreover, H2 evolution efficiency has been reported as highest for WS2–Au–CIS due to the more rapidly photogenerated carrier separation from the type II band structures and the localized surface plasmonic resonance (LSPR) effect from the Au nanoparticles [25]. To efficiently harvest solar light, nitrides and oxynitrides can be applied as photocatalysts for water splitting. The 2p orbitals corresponding to nitrogen in nitrides have higher energy than analogous orbitals of oxygen in metal oxides. Consequently, in nitrides, lower energy is needed to excite electrons to the CB. A solid solution of GaN and ZnO has been considered as a promising photocatalyst for water oxidation. Although GaN and ZnO are poor visible light absorbers due to their large band gaps, after mixing (Ga1 xZnx)(N1 xOx) will have new electronic states that considerably reduce the band gap. The other well-known catalyst for water splitting is a perovskite-like mixed oxide of NaTaO3 and SrTiO3, which has a 50% quantum yield for the 280-nm light. Replacement of one of the oxygens in these oxides with nitrogen leads to a shift in the absorption edge toward higher wavelengths (600 nm), which improves their photocatalytic activity under visible light [26]. Tantalum nitride (Ta3N5) also has been identified as an active photocatalyst for water splitting. Ta3N5 has been modified by partial substitution with Mg2 þ and Zr4 þ , which led to apparent decrease in onset potential for PEC water oxidation. Such compositional modification could apply to other SCs in order to enhance photocatalytic activity [27]. Other than the oxynitrides, graphitic carbon nitrides (g-C3N4) also have been utilized as photocatalysts to produce hydrogen due to their narrow band gap of 2.7 eV that has a CB shallower than hydrogen evolution potential and a valence band potential deeper than the reversible oxygen evolution potential. g-C3N4 could produce hydrogen from water under visible light (o540 nm) in the presence of a sacrificial agent (oxidizing agent) without the aid of any noble metal. However, pristine g-C3N4 shows a low affinity toward photocatalytic reactions. Band gap engineering of g-C3N4 has been reported to enhance the photocatalytic properties through a nonmetal (i.e., S, F, B, P) and metal doping (i.e., Pt, Pd, Fe, Zn, Cu) strategy. Moreover, the charge separation in g-C3N4 can also be enhanced by applying conductive graphene, carbon nanotubes, and reduced graphene oxide at the interface with g-C3N4. Hybridized g-C3N4 with nitrogen doped graphene quantum dots showed a higher photocatalytic activity for the nanocomposite [28]. In the literature, there are several criteria to assess the performance of photoactive materials. In this chapter, the selected performance evaluation criteria for photoactive materials are grouped into three categories (Fig. 4). Technical evaluation criteria includes molar generation rate of electrons or hydrogen (per gram or per square meter of photoactive material), cost criteria indicates the processing and preparation cost of the photoactive materials, and the third criteria (environmental and health impact) estimates the potential health and fire hazard, reactivity, and emissions of photoactive material preparation and processing.

Performance evaluation criteria for photoactive materials

Technical evaluation criteria

Molar generation rate (per gram or m2 material)

Band gap (Eg)

Cost criteria

Environmental and health impact criteria

Health hazard

Fire hazard

Fig. 4 An overview of the selected performance evaluation criteria for photoactive materials in this chapter.

Reactivity

Emissions

Photoactive Materials

2.17.4

531

Photocatalytic Water Splitting

With the introduction of fuel cells, producing H2 in a clean, sustainable, and feasible way gained significant importance in the scientific community. Photocatalytic water splitting, which can also be called “artificial photosynthesis,” is a basic chemical reaction induced by photoirradiation that converts light to chemical energy in the presence of a photocatalyst. The photocatalyst itself facilitates chemical reactions without being consumed or transformed. From a photon energy conversion point of view, photocatalytic water splitting to evolve hydrogen has been attracting considerable attention [29]. Photocatalysis starts with irradiation of light with energy greater than the band gap of a SC-based photocatalyst particle, separating the vacant CB and filled VB, exciting an electron in VB into CB to result in the formation of an electron (e ) and hole (h þ ) pair. These e and h þ reduce and oxidize respectively as chemical species on the surface of photocatalyst, unless they recombine to deliver no net chemical reaction. The original structure (or chemical composition) of the photocatalyst remains unchanged if an equal number of e and h þ are consumed for chemical reaction and/or recombination [30]. The water splitting reaction H2O-H2 þ ½ O2 is a multielectron process that requires a source of energy to meet the Gibbs free energy of the reaction, which is needed to rearrange the valance electrons to make the formation of H2 and O2 possible. This energy is equal to 2.458 eV to produce one molecule of hydrogen in standard conditions, representing the amount required to rearrange electrons under 1.229 eV of potential difference. For a full reaction to occur (for production of 2 mol of H2), four valance electrons of two water molecules are dislocated, requiring an energy input of 4.915 eV. Light could be used as a potential source of energy. Furthermore, utilizing the direct solar radiation could minimize the energy input to the water splitting process. The required energy for the full reaction (4.915 eV) can be produced by one photon of ultraviolet (UV) light with a wavelength shorter than 252.3 nm, or by two photons in the visible spectrum with a wavelength shorter than 504.5 nm, or four infrared (IR) photons of 1.23 eV [31]. The possible methods to produce sustainable hydrogen via water splitting have been studied by Acar et al. [4]. Acar and Dincer [5] have examined thermodynamic and environmental studies of photocatalytic conversion of light energy in terms of energy and exergy efficiencies. In order to evolve hydrogen from liquid water, a photocatalytic system requires photosensitizing, charge transfer, electron donation, and acceptance compounds and catalysts. Photosensitizers are used to make the water splitting system capable of absorbing solar radiation in the visible and UV ranges since pure water does not absorb radiation in these ranges [32]. During the past decade, hydrogen production by solar light driven photolysis of water and photocatalyst development has been studied extensively. The solar radiation driven water photo oxidizing reaction is: 2H2 OðlÞ ⟶hv 4Hþ þ 4e (VOX ¼ þ 0.82 V, normal hydrogen electrode (NHE)) where the active center absorbs two electrons from each water molecule; for complete reaction, two molecules of water is needed. And, the solar radiation driven water reduction reaction is 2H2 OðlÞ þ 2e ⟶hv 2OH þ H2 (VRED ¼ 0.41 V, NHE) [33]. Band gap limitation is a major disadvantage for most of the photocatalysts, causing a low yield of H2 production. Sayama [6] has proposed a novel photocatalytic system that generates H2 and O2 simultaneously (Z scheme) to overcome this problem. In the dual photocatalyst system Z scheme H2 photocatalyst reduces water and O2 photocatalyst oxidizes it [34]. Developing efficient photocatalysts to split water, and as a result, produce hydrogen is an ongoing research endeavor, with significant progress made thus far; the future of photocatalytic systems seem to be bright [35]. Yamashita and Ichihashi [7] have investigated the recent progress of visible light driven TaON-based heterogeneous photocatalysts for overall water splitting. Iwashina et al. [8] have investigated the H2 and O2 evolution under visible light irradiation with oxides in the presence of sacrificial agents. Fujita et al. [9] have reviewed the photocatalytic properties of NiO, Co3O4, Fe3O4, and Cu2O. Solar photo production options including photocatalytic and SC systems are assessed by Kato et al. [10] and Iqbal et al. [11].

2.17.4.1

Heterogeneous Photocatalysts

Overall water splitting using a heterogeneous photocatalyst is an important reaction that offers an ideal method for supplying hydrogen as a clean and renewable energy carrier. Heterogeneous photocatalysts have several advantages compared to homogeneous photocatalysts. One of the major advantages of heterogeneous photocatalysts is that since the reaction occurs in a distinct phase with respect to the medium, the separation and reutilization of heterogeneous photocatalysts are simpler and cheaper than the homogeneous photocatalysts. Recovery of homogeneous photocatalysts from the reaction medium (such as precipitation) is an energy intensive process and such operation may often deactivate the photocatalyst. Another advantage is the good thermal stability of heterogeneous photocatalysts as most of the homogeneous photocatalysts have poor thermal stabilities [36]. In the heterogeneous catalysis process, the photocatalyst is in the solid phase and the water dissociation occurs in the liquid phase. SCs are used with heterogeneous photocatalysts as electrodes. One of the reasons for using SCs is their small band gap between the valence and CBs. Both photosensitization and photocatalytic behavior are observed with some SCs (such as TiO2). With the incident of light, the catalytic sites for water splitting reaction are activated, and photons are absorbed by SCs, which dislocate electrons from the VB and move them to the CB. As a result, redox reactions are aided by the photocatalyst. It is important to note that the photocatalyst is not consumed in the reaction but it is used to reduce the activation energy [37]. In a photocatalytic water splitting reaction, the photocatalyst plays a crucial role. Fig. 5 illustrates the solar light driven water splitting process on heterogeneous photocatalysts schematically. As can be seen from Fig. 6, the CB is separated from the VB by a band gap. The energy of the irradiated light must be larger than the band gap in order to generate electrons and holes in the CBs and VBs, respectively. These electrons and holes cause redox reactions similar to electrolysis: electrons reduce the water molecules

532

Photoactive Materials

+ + + + + + + + + + + Cat+ − + − + − + − + − − + − + − + − − + − − − − − − − − − − − − − − −

Electrode

E

Heterogeneous photocatalysis (control volume)

E

−

−

EC

EC

H2+2OH−

h

2e−

EF

ERed − 2H++2e− →H2

2H2O→O2+4H++4e− − EOX Legend E: energy x: distance Subscripts V: valence band F: Fermi level C: conduction band Red: reduction Ox: oxidation

h

h Photocatalytic reaction: 2H2O+Cat+h→ H2+2OH−+Cat EF

2H2O EV

Helmholtz layer Solvent molecule

+ EV

+

Electrode

Electrolyte 0

(A) Heterogeneous photocatalysis mechanism

Electrode

x

(B) Reduction reaction

Electrolyte 0

x

(C) Oxidation reaction

Fig. 5 Schematic of water splitting on semiconductor (SC) photocatalysts.

Conduction band (CB)

Conduction band (CB) Conduction band (CB)

H+/H2

O2/H2O Valence band (VB)

Valence band (VB)

Valence band (VB) Fig. 6 Schematic of band gap modification via band engineering.

to form hydrogen and holes oxidize the water molecules to form oxygen. Therefore, width of the band gap and energy levels of the CBs and VBs strongly affect the performance of a photocatalyst. Since the bottom level of the photocatalyst has to be more negative than the redox potential of H þ /H2 and the top level of the VB has to be more positive than the redox potential of O2/H2O, the band gap should be wider than 1.23 eV. Furthermore, the photocatalyst must be stable in aqueous solutions under photoirradiation [38]. Novel SC materials to develop efficient heterogeneous photocatalysts have been extensively studied in the literature. One of the challenges of photocatalytic water splitting is preventing backwards reactions. One important issue is to keep the photogenerated electrons and holes separated. Crystallinity and surface properties of photocatalysts strongly affect the electron and hole generation and separation processes. In general, photocatalysts with high crystallinity have higher photocatalytic activities. Photocatalysts with fewer surface defects (increased crystallinity) could prevent electron and hole recombination sites, and this can increase the lifetime and mobility. Higher photocatalytic activity can also be obtained by reducing the particle size of a photocatalyst, because the diffusion length for photogenerated electron–hole pairs can be shortened. Also, surface area of the photocatalyst grain strongly affects the number of active sides [39]. Oxide-based SCs are extensively used as heterogeneous photocatalysts since they are reported to be stable against photocorrosion. However, in 1980, Scaife [12] discussed that it is essentially quite challenging to develop an oxide SC photocatalyst that has both a sufficiently negative CB for hydrogen production and a sufficiently narrow band gap (i.e., o3.0 eV) for visible light absorption because of the highly positive VB (at ca. þ 3.0 V vs. NHE) formed by the O 2p orbital [40]. Most visible light responsive oxide photocatalysts cannot produce hydrogen from water due to their CBs being too low for water reduction. Among the photocatalytic systems that have been reported to be active for overall water splitting (simultaneous

Photoactive Materials

533

generation of O2 and H2), most of them only absorb UV light, which has much shorter wavelength than 400 nm and occupies only 3–5% of solar energy, causing a low efficiency of energy conversion. In their study, Abe et al. have stated that even if all UV light (up to 400 nm) were utilized, the solar conversion efficiency would only be 2% [13]. This is due to the high band gap energy of the photocatalyst. Since almost half of the solar energy incident on the Earth’s surface is within the visible region (with wavelengths between 400 and 800 nm), utilizing visible light up to 600 nm would increase the energy efficiency to 16%, and this amount would go up to 32% if the visible light up to 800 nm could be fully utilized [41]. Although some SCs without any oxides (e.g., sulfide-based) possess appropriate band levels for water splitting under visible light, they are generally unstable and readily become deactivated through photocorrosion or self-oxidation, rather than evolving O2. For example, cadmium sulfide (CdS) has appropriate band levels for water reduction and oxidation as well as a narrow band gap that permits visible light absorption. However, CdS is not stable in the water oxidation reaction to form O2 because the S2 anion is more susceptible to oxidation than water, causing the CdS catalyst itself to be oxidized and degraded [42]. Homogeneous sensitizer molecules (such as organic dyes and metal complexes) are another class of SC photocatalysts that are not oxide-based. Spectral sensitization of wide band gap SCs (e.g., TiO2) by using sensitizer molecules has been studied to achieve effective solar light driven photocatalytic hydrogen production from water [43]. Because of its advantages like stability, resistance to corrosivity, environmentally friendly nature, and cost-effectiveness, titania (titanium dioxide, TiO2) has been a widely used photocatalyst for photocatalytic water splitting reactions. One of the most important advantages of TiO2 is its appropriate energy levels to initiate the water splitting reaction: the CB of TiO2 is more negative than the reduction energy level of water (EHþ =H2 ¼ 0 V), while the VB is more positive than the oxidation energy level of water (EO2 =H2 O ¼ þ 1:23 V). However, despite its many advantages, TiO2 has a quite low photocatalytic water splitting efficiency under solar light. One of the reasons of this low efficiency is that the photogenerated electrons in the CB of TiO2 may recombine with the VB holes quickly to release energy in the form of unproductive heat or photons. Another reason is the large positive Gibbs free energy (DG ¼ 237 kJ/mol) of the decomposition of water into H2 and O2, which makes it easier for the backward reaction (recombination of H2 and O2 into water) to proceed. Last, but not least, the band gap of TiO2 is about 3.2 eV, and as a result, only UV light can be utilized to activate the photocatalyst. Since, as mentioned above, UV light only accounts for approximately 4% of solar energy, the inability to utilize most of the solar light limits the efficiency of TiO2 in solar photocatalytic hydrogen production [44]. In the literature, K4Nb6O17 is extensively reported to be an active water splitting photocatalyst [15]. Even without the aid of a cocatalyst, K4Nb6O17 alone produces H2 and O2 from water under band gap irradiation (4ca. 3.3 eV). Interestingly, the photocatalytic activity of K4Nb6O17 for overall water splitting is almost insensitive to gaseous oxygen, indicating that the photoreduction of oxygen, which sometimes takes place during photocatalytic water splitting, scarcely occurs. Many tantalates have been reported to be highly active photocatalysts since the mid-1990s. However, many tantalates have wide band gaps between around 4.0 and 4.5 eV, and they generally have relatively high activity for overall water splitting under UV irradiation [45]. WO3 functions as a stable photocatalyst for O2 evolution under visible light in the presence of an appropriate electron acceptor, i.e., the bottom of the CB of the material is higher than the potential for water reduction. As a result, WO3 cannot reduce H þ to H2 [46]. Overall, the difficulty in developing a suitable photocatalyst depends on meeting the following three requirements [47]: 1. Band edge potentials suitable for overall water splitting. 2. Band gap energy lower than 3 eV. 3. Stability in the photocatalytic reaction. To date, there are not many stable photocatalysts that are capable of both high visible light absorption and a high potential for water splitting. Thus, achieving water splitting using heterogeneous photocatalysts under visible light has been a major challenge for many years. A possible solution to solar light driven photocatalysis is band engineering, which can be considered as the first step in the design of a photocatalyst [48]. The usefulness of band engineering is suggested schematically in Fig. 6. Table 1 summarizes the photocatalysts selected from the aforementioned literature screening for detailed comparative assessment, along with the references. The effect of cocatalysts and electron donors on photocatalyst performance is discussed in the following sections.

2.17.4.2

Addition of Cocatalysts

The heterogeneous photocatalyst water splitting reaction, as depicted in Fig. 5, consists of three steps. The first step is the energy absorption from photon; this energy is greater than the band gap energy of the photocatalyst. As a result, photoexcited electron and hole pairs are formed. These electron and hole pairs migrate to the surface without recombination (second step). Then electrons and holes are used to produce H2 and O2, respectively (third step). Although the first two steps strongly depend on structural and electronic properties of the photocatalyst, the third step usually requires the use of a cocatalyst [49]. Usually, cocatalysts are noble metals (e.g., Pt, Rh) or transition metal oxides (e.g., NiOx, RuO2, etc.). They are loaded onto the base catalyst as a dispersion of nanoparticles (typically less than 50 nm in size) to (1) extract photogenerated electrons and holes from the photocatalyst, (2) generate and/or introduce active sites, and (3) reduce the activation energy for product

534

Photoactive Materials

Table 1

Selected photocatalyst groups evaluated in this study

Photocatalyst

Advantages

Disadvantages

SrTiO3-based

Stable lattice structure, suitable photocatalytic site, supports the accommodation space for a wide range of cations and valences Environmentally friendly, not toxic, chemically stable, abundant, cost-effective, TiO2 hybrids are ideal for controlling the charge separated states

Wide band gap (3.2 eV), active only under UV light irradiation (limited use)

TiO2-based

Zn/In/S-based Ta/O-based Cd/S/Zn-based K/Ti/O-based Ga/Zn/O-based

High energy conversion efficiency, good crystallinity, appropriate band gap for visible light irradiation High photocatalytic performance, robust and tunable electronic structure High photocatalytic performance, rapid generation of electron and hole pairs, appropriate band gap Proper layered structure Stability, low toxicity, low cost, good catalytic activity

Low photocatalytic efficiency, wide band gap (3.2 eV), low surface area, rapid recombination of electron and hole pairs, fast backward reaction between H2 and O2, large H2 production overpotential Cost and environmental impact issues Less information of photocatalytic properties, limited practical application Cost and environmental impact issues, unstable photocatalyst, photocorrosion Wide band gap Rapid deactivation of the photocatalysts

gas evolution. Consequently, the overall efficiency of a photocatalytic system depends on the type and quality of the loaded cocatalyst [50]. To date, Pt, Pd, Ru, Rh, Cu, Au, and Ni have been applied as a cocatalyst for photocatalytic overall water splitting. Among them, Pt is known to be a promising cocatalyst for the reduction of protons to produce hydrogen molecules. However, a Pt cocatalyst does not necessarily deliver the highest photocatalytic activity among other alternatives (e.g., Ru and Rh) [51]. The efficiency of a photocatalytic system is affected by several factors, such as band gap potential and working mechanism of the photocatalyst and the loaded cocatalyst 1. When designing an efficient cocatalyst for a photocatalytic system, the processes ((1), (2), and (3)) should be taken into account. The structural characteristics and essential catalytic properties of a cocatalyst for H2 (or O2) evolution are important factors affecting the overall photocatalytic system’s performance [52]. Another important requirement of the cocatalyst is to be inactive for water formation from H2 and O2. In order to avoid the backward reaction, transition metal oxides that do not exhibit activity for water formation from H2 and O2 are usually applied as cocatalysts for overall water splitting. Naturally, there are some exceptional cases in which noble metals have been effectively utilized as cocatalysts for photocatalytic water splitting [53]. Despite the fact that some photocatalysts have been reported to exhibit high activity without the presence of a cocatalyst, a cocatalyst usually enhances the overall efficiency of the reaction. Thus, developing a cocatalyst that efficiently promotes photocatalytic water splitting is an important research focus. Pt cocatalysts are widely used in the literature due to their shown catalytic activities for hydrogen production under solar light like irradiation [54].

2.17.4.3

Electron Donors and Electrolytes

The uphill nature of the photocatalytic reaction generally makes the overall water splitting reaction difficult to achieve. Especially due to rapid recombination of photogenerated CB electrons and VB holes, it is difficult to achieve water splitting for hydrogen production using photocatalysts in pure water. Therefore, photocatalytic activities of a compound for water reduction or oxidation reactions are usually studied in the presence of electron donors, in other terms, electrolytes (e.g., methanol, Na2SO3, Na2S, KI, etc.). It should be noted that electrolytes do not undergo reduction or oxidation by CB electrons and VB holes like a sacrificial reagent would. However, when added to the reaction solution, they have a significant effect on the performance of photocatalytic water splitting reaction [55]. In Table 2, some common photocatalyst examples from the literature along with their cocatalysts and electrolytes are provided in conjunction with their required light sources and cutoff filter wavelengths (lfilter) as well as the band gap energies of each photocatalyst (Eg). When the photocatalytic reaction is conducted in the presence of an electron donor, photogenerated holes in the VB irreversibly oxidize the electron donor instead of H2O, thus facilitating water reduction by CB electrons if the bottom of the CB of the photocatalyst is located above the water reduction potential. On the contrary, in the presence of an electron acceptor, photogenerated electrons in the CB irreversibly reduce electron acceptors instead of H þ . As a result, water oxidation by VB holes is promoted if the top of the VB of the photocatalyst is more positive than the water oxidation potential. It is important to note that the ability of a photocatalyst to both reduce and oxidize water separately does not guarantee the capability to achieve overall water splitting without sacrificial reagents. Adding electron donors or sacrificial reagents to react with the photogenerated VB holes is an effective measure to enhance the electron and hole separation, resulting in higher quantum efficiency. However, the drawback of this technique is the need to continuously add electron donors in order to sustain the reaction since they will be consumed during photocatalytic reaction. An attractive solution to this issue is to identify waste materials from industry that can be recovered and used as electron donors in

Photoactive Materials

Table 2

535

Summary of selected common photocatalyst examples from the literature along with some key properties

Photocatalyst

Ref.

Cocatalyst

Electrolyte

lfilter (nm)

Light source

Eg (eV)

(Ru/SrTiO3:Rh)-(BiVO4)-FeCl3 F-TiO2 ZnIn2S4 Bi1.5Zn0.99Cu0.01Ta1.5O7 Ag0.03Mn0.40Cd0.60S CaTa0.8Zr0.2O2.2N0.8 Cd0.8Zn0.2S/S15 K2La2Ti3O10 K2Ti4O9 SrTiO3:Ni/Ta 0.93 TiO2–0.07 ZrO2 SrTiO3:Ni/Ta/La Cr/N doped SrTiO3 In(OH)yS:Ag–Zn Zn(Ag)/In¼0.04 TiO2/SnO2 Rh/Cr2O3/GaZn oxide ZnO/ZnS Cd0.4Zn0.6S TiO2–C-362 CdS N-doped In2Ga2ZnO7 GaFeO2.98S0.02 5% Fe – 4% Ni/TiO2 GO–TiO2 Au–TiO2–AC TiO2–SiO2 TiO2–ZnO 2Au–TiO2 K4Nb6O17/CdS TiO2–NiS g-C3N4–SrTiO3:Rh (0.3 mol%) Au–CdS 25 wt% PbS/K2Ti4O9 0.9 wt% Pd–Gardenia–TiO2 5 wt% In2O3/Ta2O5 CdS/Ta2O5 Cd0.1Zn0.9S Pt–PdS–CdS CdS/Ti K1.025Sr2Nb2.9875Cr0.0125O10 BaZr0.96Ta0.04O3 7.5% Bi doped NaTaO3 (CuAg)0.15In0.3Zn1.4S2

[56] [57] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [58] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40]

– Pt Pt – Pt Pt – CdS CdS Pt Cu Pt Pt – – – – NiS – Pt Pt – – Pt – – – – – – Pt – – – Pt – Cu – – – – Pt Ru

K2SO3 þ Na2S Bisphenol A (BPA) Na2SO3 þ Na2S Na2SO3 þ Na2S Na2SO3 þ Na2S HCOOH Na2SO3 þ Na2S Na2S Na2SO3 þ Na2S þ KOH CH3OH CH3OH CH3OH CH3OH Na2SO3 þ Na2S CH3OH CH3OH Glycerol Na2SO3 þ Na2S CH3OH (NH4)2SO3 CH3OH – C2H6O TEOA Ethylenediaminetetraacetic acid (EDTA) Diethylamine (DEA) C2H6O Ascorbic acid Na2SO3 þ Na2S þ KOH Lactic acid CH3OH Na2SO3 þ Na2S Na2SO3 þ Na2S þ KOH – CH3OH Lactic acid Na2SO3 þ Na2S Na2SO3 þ Na2S TEA CH3OH – CH3OH KI

– – 4420 4400 4420 4420 – 4400 4400 4415 – 4420 4420 4400 – – – 300 4280 4420 4400 4395 4400 4420 4400 4400 – 4420 4400 4400 4415 4420 4400 – 4420 4400 4420 4420 4430 4420 – 4390 420

200 W – 300 W 300 W 500 W 300 W 350 W 500 W 500 W 300 W Hg 300 W 300 W 300 W 400 W 125 W 500 W 300 W 300 W 300 W 125 W 300 W 500 W 300 W 200 W 300 W Hg 300 W 500 W 300 W 300 W 300 W 500 W 125 W 300 W 300 W 500 W 300 W 300 W 300 W – 800 W 300-W

– – 2.59 2.6 2.59 2.59 2.23 3.5 2.4 – 3.25 – 2.39 1.65 2.9 2.6 3.4 2.4 – 2.4 2.5 2.65 2.41 – 3.2 3.26 3.06 3.28 3.1 – 3.2 2.4 – – 2.8 2.4 2.78 2.4 – 3.5 – 2.64 1.9

Xe–Hg Xe Xe Xe Xe Xe Xe Xe Xe Xe Xe Xe Hg Hg Xe Xe Xe Xe Hg Xe Xe Xe Xe–Hg Xe Xe Xe Xe Xe Xe Xe Hg Xe Xe Xe Xe Xe Xe Xe Xe

photocatalytic systems. One option is to use aqueous wastes containing sulfide or sulfite, which can well act as electron donors. Many sulfurous waste streams can be recovered from the petroleum and chemical industries. Other types of pollutant streams can be used for the same purpose. For example, Li et al. [16] have reported enhanced photocatalytic hydrogen production when organic pollutants acting as electron donors, such as formic acid (HCOOH), were added into the reaction system. Decomposition of the organic pollutants was reported to be consistent with hydrogen production.

2.17.5 2.17.5.1

Performance Comparison Criteria for Photoactive Materials Technical Comparison Criteria (̇ nm [µmol/h-gcat], ̇ nA [lmol/h-m2cat], Eg [V])

In this section, the rate of hydrogen formation is expressed as the molar rate of hydrogen production per mass unit of catalyst, denoted with n_ m and measured in mmol/h-gcat, which is micromoles of hydrogen production per hour per gram catalyst. In the case study, 43 photocatalysts are compared based on their mass specific molar rate data taken from the recent literature studies; these photocatalysts are presented in Table 2.

536

Photoactive Materials

The hydrogen generation rates of the photocatalysts are also ranked on 0–3 scale to be able to compare their overall performance based on different criteria mentioned later in this study. The values of 0 and 3 are assigned to the poorest and highest performance catalyst, respectively. In between 0 and 3, the following equation is used to rank the photocatalysts evaluated in this study: Rank i ¼

Performancei Performancemin 3 Performancemax Performancemin

ð1Þ

where Ranki is the rate rank assigned to photocatalyst i. It is calculated based on the H2 production rate of photocatalyst i (Performancei) and the lowest (Performancemin) and highest (Performancemax) H2 production rates among selected photocatalysts. According to Eq. (1), the photocatalyst that gives the highest H2 production rate is assigned the rank “3” and the photocatalyst that gives the lowest H2 production rate is assigned the rank “0.” In addition to the mass specific molar rate of hydrogen production (mmol/h-gcat), a second criterion is used to evaluate the hydrogen production rates of selected photocatalysts: this is the area specific molar hydrogen production rate. The unit is mmol/h-m2cat, which represents micromoles of hydrogen production per hour per unit specific area (m2) of the photocatalyst. This amount is calculated by dividing the mmol/h-gcat data mentioned above by Brunauer Emmett Teller (BET) surface area (m2/gcat). The BET analysis provides precise specific surface area evaluation of materials by nitrogen multilayer adsorption measured as a function of relative pressure using a fully automated analyzer. The technique encompasses external area and pore area evaluations to determine the total specific surface area in square meter per gram yielding important information in studying the effects of surface porosity and particle size in many applications. The mmol/h-m2cat results are ranked by using Eq. (1) and the same procedure as mmol/h-gcat. The photocatalytic performance of a photocatalyst strongly depends on its electronic band structure and band gap energy, Eg. For an efficient photocatalyst, the band gap energy should be smaller than 3 eV to extend the light absorption into visible region to efficiently utilize the solar energy. Apart from the maximum band gap requirement, the minimum band gap of SC photocatalysts for water splitting should be 1.23 eV and effective photocatalysts have been shown to exhibit band gaps larger than 2 eV. A relatively arbitrary ranking system is used to rank the photocatalysts based on the band gap energy: the range of energy bands is divided in four subdomains. The photocatalysts with band gaps between below 1.3 and above 3 eV are given the rank “0.” The ones with 1.3rEgo1.6 and 2.7oEgr3 are assigned the rank “1.” The rank “2” is given to the ones with 1.6rEgo1.9 and 2.4oEgr2.7. Band gap range 1.9rEgr2.4 is ranked as “3.”

2.17.5.2

Cost Comparison Criteria

In order to be able to compare the costs of the selected photoactive materials, their commercially available precursors, cocatalysts, and required electron donors are investigated and their costs on ICIS Pricing and Sigma Aldrich Canada are used to rank the cost of each photocatalyst. ICIS Pricing chemical price reports cover all the major chemical markets across different sectors in Europe, the Middle East, Asia Pacific, North America, and Latin America. The pricing data, trends, and market commentary in the ICIS Pricing chemical reports are considered to be one of the most important market references for price assessments. Due to lack of detailed information on required amounts of precursors and photoactive material synthesis reaction yields of selected substances, ranking is done based on averaging each component used in the preparation of each selected photocatalyst. “3” is assigned to the least expensive and “0” is assigned to the most expensive component. In between 0 and 3, the ranking is evaluated based on the following equation: Rank i ¼

Costi Costmin 3 Costmax Costmin

ð2Þ

where Ranki is the cost rank assigned to photoactive material i. It is calculated based on the cost of photoactive material i (Costi) and the least expensive (Costmin) and most expensive (Costmax) of the selected photocatalysts. According to Eq. (2), the photocatalyst that has the highest cost is assigned the rank “0” and the photocatalyst that has the lowest cost is assigned the rank “3.”

2.17.5.3

Environmental and Health Impact Comparison Criteria

When evaluating the environmental impact of the selected photoactive materials, the Hazardous Materials Identification System (HMIS) database is used. HMIS ranks the reactivity, fire hazard, and health hazard of the chemicals available in its database. The ranking is based on a 0–3 scale. The rank 0 could be interpreted as least harmful and 3 is the most harmful. In this study, to this point, 3 is defined as the ideal and/or most preferred option. Therefore, a slight change has been applied to the HMIS rankings. In our study, 0 and 3 imply the least (HMIS: 3), and most preferred ones (HMIS: 0), respectively. The average of fire hazard and reactivity is considered as the environmental impact and the health hazard is considered as the health impact. The fire hazard, reactivity, and health hazard of commercially available precursors of synthesized photocatalysts, their cocatalysts, and required electron donors are ranked based on the HMIS database. The overall health and environmental impact of catalysts are ranked based on the most harmful component of each photocatalyst.

Photoactive Materials 2.17.5.4

537

Overall Comparison

In this section, the n_ m (mmol/h-gcat), n_ A (mmol/h-m2cat), Eg (eV), cost, environmental and health impact ranking criteria for the photoactive materials are introduced. The photocatalysts shown in Table 2 are compared in the next section as a case study. For each material, the average rankings of the individual photoactive materials from the literature are taken into account. The aim of the overall comparison is to evaluate the technical, cost, health, and environmental performance of different photoactive materials, point out the advantages of each group, and to highlight possible improvement potentials for future research.

2.17.6

Case Study: Comparative Assessment of Photocatalysts

The hydrogen production rates per gram of the photoactive materials (Fig. 7) are slightly different than that of per specific area performances (Fig. 8). In the latter case, the highest rates are observed with g-C3N4-SrTiO3:Rh, Ag0.03Mn0.40Cd0.60S, and (CuAg)0.15In0.3Zn1.4S2 as presented comprehensively in Fig. 8. In addition to the hydrogen production rate comparisons presented in Figs. 7 and 8, the selected photocatalysts are evaluated and ranked based on their technical (n_ m (mmol/h-gcat), n_ A (mmol/h-m2cat), Eg (eV)), cost, and environmental and health impact performance. The results are summarized in Table 3. The ranking used in this study is based on 0 to 3 scale. The rank 3 represents the ideal case: maximum hydrogen production rate (both per gcatalyst and m2catalyst ), ideal band gap (Eg around 2.4 eV), minimum cost, minimum environmental, and health impact.

4500 4000 3500

μmol/h−gcat

3000 2500 2000 1500 1000 500

3 5 7 9 10 11 12 13 14 16 17 18 19 20 21 22 23 24 25 26 27 29 30 31 32 33 34 35 36 37 38 39 40 41 43

0

Fig. 7 M hydrogen production per mass unit of catalyst (mmol/h-gcat) of selected photocatalysts along with their reference numbers.

90 80

μmol/h−m2cat

70 60 50 40 30 20 10 [43]

[41]

[38]

[36]

[31]

[27]

[25]

[23]

[22]

[21]

[19]

[17]

[16]

[14]

[13]

[12]

[11]

[6]

[10]

[5]

[4]

[3]

0 Fig. 8 M hydrogen production per unit of surface area (mmol/h-m2cat) of selected photocatalysts along with their reference numbers.

538

Photoactive Materials

Table 3

Selected photocatalyst systems comparison

Ref.

Photocatalyst

[1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40]

Technical

Cost 2

Name

(mmol/h-gcat)

(mmol/h-m )

Eg (eV)

ZnIn2S4 Bi1.5Zn0.99Cu0.01Ta1.5O7 Ag0.03Mn0.40Cd0.60S CaTa0.8Zr0.2O2.2N0.8 Cd0.8Zn0.2S/S15 K2La2Ti3O10 K2Ti4O9 SrTiO3:Ni/Ta TiO2–ZrO2 SrTiO3:Ni/Ta/La Cr/N–SrTiO3 In(OH)yS:Ag–Zn Rh/Cr2O3/GaZn ZnO/ZnS Cd0.4Zn0.6S TiO2–C-362 CdS N-In2Ga2ZnO7 GaFeO2.98S0.02 Fe–Ni/TiO2 GO–TiO2 Au–TiO2–AC TiO2–SiO2 TiO2–ZnO Au–TiO2 K4Nb6O17/CdS TiO2–NiS g-C3N4–SrTiO3:Rh Au-CdS PbS/K2Ti4O9 Pd–Gardenia–TiO2 In2O3/Ta2O5 CdS/Ta2O5 Cd0.1Zn0.9S Pt dS–CdS CdS/Ti K1.025Sr2Nb2.9875Cr0.0125O10 BaZr0.96Ta0.04O3 Bi–NaTaO3 (CuAg)0.15In0.3Zn1.4S2

1 0 2 0 1 0 0 0 1 2 0 3 1 0 1 3 1 2 0 0 3 2 1 1 0 2 0 2 3 0 0 0 3 1 2 0 0 1 0 0

2 2 3 0 0 0 0 1 0 2 0 2 1 0 0 2 0 1 0 0 0 1 0 1 0 0 0 3 0 0 0 0 1 0 1 0 0 1 0 3

2 2 2 2 3 0 2 – 0 – 2 2 2 0 3 – 3 2 2 2 – 0 0 0 0 0 – 0 3 – – 1 3 1 3 – 0 – 2 3

1 2 1 3 3 3 2 3 3 3 3 0 1 3 3 3 3 1 2 3 2 1 3 3 1 3 3 1 1 1 2 1 2 3 1 3 2 2 2 1

Impact

Avg.

Env.

Health

0 3 3 0 1 1 0 0 1 0 0 2 0 0 1 1 0 0 2 1 1 3 1 1 1 0 1 0 3 0 3 0 0 1 3 0 0 0 0 1

0 1 1 0 0 0 0 0 2 0 0 0 0 0 0 2 0 0 1 2 2 0 1 2 2 0 2 0 1 0 1 0 0 0 1 0 0 0 1 0

1.12 1.74 2.04 0.88 1.28 0.67 0.73 0.85 1.18 1.33 0.96 1.45 0.85 0.61 1.30 2.23 1.14 1.05 1.21 1.40 1.51 1.16 0.98 1.23 0.67 0.92 1.29 0.92 1.78 0.28 1.22 0.39 1.50 1.05 1.85 0.61 0.36 0.70 0.84 1.39

Similarly, rank 0 represents the least deserved system. Therefore, if a photocatalyst has a ranking closer to 3, it delivers closer to ideal case results compared to the other catalysts evaluated in this study. Fig. 7 shows that TiO2–C-362 has the highest performance in terms of mmol/h-gcat. Au–CdS has the second highest hydrogen production rate per gram of photocatalyst. Following these two, In(OH)yS:Ag–Zn, GO–TiO2, K4Nb6O17/CdS, and CdS/Ta2O5 have similar mmol/h-gcat performances.

2.17.6.1

Photocatalyst Groups Comparison

When the average rankings in Table 3 are taken into account, it can be seen that TiO2-based catalysts have better results compared to other groups. In order to see the photocatalytic performance of each group mentioned in Table 1, the photocatalyst group comparison table is constructed based on the data shown in Table 3. Table 4 presents the technical, cost, and environmental and health impact of selected photocatalyst groups. The ranking here is again based on a 0–3 scale and 3 represents the ideal case for each criterion. Table 4 shows that except for the K/Ti/O-based photocatalysts group, the average rankings fall in a small interval of between 1.01 and 1.32. Fig. 9 summarizes these findings and compares each photocatalyst group’s performance to the ideal case.

Photoactive Materials

Table 4

539

Comparative assessment of selected photocatalyst groups

Photocatalyst group

SrTiO3-based TiO2-based Zn/In/S-based Ta/O-based Cd/S/Zn-based K/Ti/O-based Ga/Zn/O-based

Technical

Cost 2

(mmol/h-gcat)

(mmol/h-m )

Eg (eV)

0.94 1.07 1.51 0.60 1.39 0.69 1.02

1.60 0.43 2.41 0.83 0.54 0.00 0.86

1.00 0.33 2.33 2.00 1.91 0.50 2.00

2.50 2.60 0.67 2.00 2.42 2.20 1.33

Impact

Average

Environmental

Health

0.00 1.40 1.00 0.50 1.08 0.20 0.67

0.00 1.60 0.00 0.33 0.42 0.00 0.33

1.01 1.24 1.32 1.04 1.29 0.60 1.04

μmol/h−gcat 3.0 2.5 2.0

Health

1.5

μmol/h−m²

1.0 0.5 0.0

SrTiO3-based Environmental

Eg (eV)

TiO2-based Zn/In/S-based Ta/O-based Cd/S/Zn-based

Cost

Ga/Zn/O-based Ideal

Fig. 9 Comparative assessment of selected photocatalyst groups.

The ideal case is represented as the outer shell of Fig. 9. Closer rankings to this ideal case indicate a better performing photocatalyst. For instance, TiO2 delivers closest to ideal case results in the health and environmental impact and cost categories due to its abundant and chemically stable nature. However, due to its wide band gap, the Eg, mmol/h-gcat and mmol/h-m2cat are far away from the ideal case compared to the other photocatalyst groups. Any potential improvement on TiO2’s wide band gap problem could possibly improve its overall performance. Similarly, Zn/In/S and Cd/S/Zn group photocatalysts have better Eg, mmol/h-gcat, and mmol/h-m2cat rankings, however, their environmental and health impact and cost rankings are rather low compared to the TiO2 group. In general, a tradeoff between technical performance and cost and environmental and health impact can be seen from Fig. 9. For that reason, current photocatalytic systems do not usually provide better technical and cost performance along with minimum environmental and health impact. Future photocatalytic systems research should focus on both improving the technical performance and minimizing the health and environmental impact while minimizing the potential costs.

2.17.6.2

Individual Photocatalyst Comparison

In order to compare the technical, cost, health, and environmental impact performances of selected photocatalysts individually, the top five highest average ranked catalysts are compared separately. Fig. 10 presents the comparison results of these top five photocatalysts. Again, the ideal case is represented as the outer shell. Closer rankings to this ideal case represent a better performing photocatalyst. The tradeoff between technical performance and cost and environmental and health impact can be seen in Fig. 10 as well. Any potential improvement on photocatalytic performance (e.g., via band gap engineering) while keeping the cost and environmental and health impact at minimum could be a major accomplishment in the photocatalyst development phase of hydrogen production.

2.17.6.3

Case Study Conclusions

The present case study discusses the current status and possible future research opportunities for developing photocatalysts for visible light driven hydrogen production from water splitting. Technical, cost, environmental and health impact assessment and

540

Photoactive Materials

Health

μmol/h-gcat 3 3 2 2 1 1 0

Environmental

μmol/h−m2

TiO2-C-362 Ag0.03Mn0.40Cd0.60S Pt-PdS-CdS Au-CdS

Eg (eV)

GO–TiO2 Ideal

Cost Fig. 10 Comparative assessment of selected photocatalysts.

ranking of SrTiO3, TiO2, Zn/In/S, Ta/O, Cd/S/Zn, K/Ti/O, and Ga/Zn/O-based photocatalyst groups are carried out and the results are presented. It has been pointed out that each photocatalytic group has different strengths and weaknesses. In order to be able to develop a system with high photocatalytic performance, low cost, and low environmental and health impact, potential improvement opportunities for each photocatalyst group are visually illustrated based on the literature survey of existing photocatalysts. 43 photocatalysts and 7 photocatalyst groups are ranked and compared in this case study, and the following findings are obtained:

• • •

The highest mmol/h-gcat rates are achieved by TiO2–C-362 (9430 mmol/h-gcat), Au–CdS (4000 mmol/h-gcat), and In(OH)yS: Ag–Zn (3876 mmol/h-gcat). Overall, the highest mmol/h-gcat rates are achieved by Zn/In/S, Cd/S/Zn, and Ga/Zn/O-based photocatalyst groups. When mmol/h-m2cat results are compared, Ag0.03Mn0.40Cd0.60S (85.23 mmol/h-m2cat), g-C3N4–SrTiO3:Rh (83.33 mmol/h-m2cat), (CuAg)0.15In0.3Zn1.4S2 (83.29 mmol/h-m2cat) give the highest production rates. In terms of photocatalyst groups, Zn/In/S, SrTiO3, and Ga/Zn/O-based photocatalyst have the highest hydrogen generation rates. When cost, health, and environmental impact criteria are taken into account, TiO2–C-362 (average performance ranking: 2.3/3.0), Ag0.03Mn0.40Cd0.60S (average performance ranking: 2.04/3.0), Pt–PdS–CdS (average performance ranking: 1.85/3.0) individual photocatalysts and Zn/In/S (average performance ranking: 1.32/3.0), Cd/S/Zn (average performance ranking: 1.29/3.0), and TiO2 (average performance ranking: 1.24/3.0)-based groups have the highest rankings.

There is a strong correlation between the photocatalytic performance and the band gap of a catalyst. By engineering the band gaps of potential photocatalysts, their performance can be improved while minimizing costs and adverse effects on environment and human health. The continuous exploration of new phases and materials for efficient, safe, cheap, and simple photocatalytic hydrogen production will most likely provide promising results in the coming years. Future trends are expected to move toward nanoarchitectures and their novel applications. Also, “artificial leaves,” a concept that combines photocatalysis and PVs, is a possible future research direction for solar hydrogen production. The research need of these photocatalytic systems is not only with regard to material development but also is connected to system engineering.

2.17.7

Photoactive Materials in Photoelectrochemical Systems

Due to population growth and rise in standards of living, global energy demand keeps increasing. This demand is expected to continue even more in the future, due particularly to some highly populated developing countries, for example, China and India. With their nonhomogeneous distribution, limited nature, slow formation, and rapid depletion, fossil fuels are not considered to be able to fully meet future generations’ energy needs. Environmental damage caused during processing, transfer, and utilization of fossil fuels is another issue. The emissions caused by fossil fuel combustion, especially CO2, are known as greenhouse gases and considered to be the main cause of climate change. In recent decades, there has been ongoing research on clean and renewable energy systems to meet the world’s increasing energy demand without exhausting existing sources and damaging the environment. Renewable energy resources are treated as promising candidates compared to fossil fuels since they are replenished, and also their end use emits very few or zero pollutants. However, renewables have an intermittent and fluctuating nature, and they require a medium in which to be stored. Hydrogen, as an abundant and clean energy storage medium, has a potential to meet the world’s energy demand. When produced from renewables and water, production and end use of hydrogen does not emit any pollutant. Solar energy has a great potential to produce hydrogen in a clean and efficient way since it is considered as inexhaustible and nonpolluting. There are two major solar energy conversion technologies: (1) solar thermal conversion, which uses the IR part (about 49%) of the solar spectrum (e.g., solar heat collectors); and (2) solar photon conversion, which uses the visible/near UV part (51%) of the solar spectrum. Furthermore, solar photon conversion devices, namely, solar cells, can be classified into three main categories: (1) solid-state PV (p/n junction) solar cells, PV; (2) metal SC-based Schottky barrier (M-S) solar cells; and (3) SC liquid junction-based PEC solar cells.

Photoactive Materials

Table 5

541

Comparison of photovoltaic (PV) cells and photoelectrochemical (PEC)

Material synthesis Material processing Junction formation Routine conversion efficiency In situ storage Cell stability Economic viability

PV cells

PEC

Costly Costly Difficult 410% (4%20 in special devices) Not possible Stable Viable but costly

Relatively cheap Relatively cheap Relatively easy 47% in special cells Possible Stable with proper electrolytes Viable, cheaper than PV

Note that SC liquid junction PEC solar cells are in between solid-state PVs and natural photosynthetic systems. The solid state solar cells are advanced, efficient, and reliable devices with solar to electricity conversion efficiencies of greater than 10%. In some special multilayer solid-state cells the efficiencies can go up to 25–26%. One of the most important factors affecting solar conversion efficiencies is the type of solar absorbing material. Solid state PV cells are suitable for special applications such as providing power to remote locations. PEC cells are in the early research and development phase, but they have a high potential to utilize a broader range of solar spectrum in an efficient manner. In Table 5, the aspects of material synthesis and processing, junction formation, conversion efficiency, in situ storage, stability, and economic viability for PV and PEC cells are compared. From Table 5, it is clear that PEC has significant advantages over PV cells such as convenient and economically viable material synthesis and processing. PEC cells are also easy to manufacture; the process involves immersing the photoactive SC into the electrolyte containing a suitable redox couple. Possibility of in situ storage is another major advantage of PEC cells. The main components of PEC cells are (1) photoactive SC electrode immersed in a (2) suitable redox couple electrolyte and a (3) metal or SC metal as counter electrode. When irritated with light of hnZEg, the SC generates and separates electron (e ) and hole (h þ ) charge carriers. There are several available solar hydrogen routes in the open literature. By taking advantage of direct use of sunlight, PEC systems offer low cost, environmentally benign, and efficient ways to produce hydrogen. PEC systems integrate solar energy collection and water electrolysis in a single photocell. PEC systems are classified based on their Gibbs free energy change as:

• •

Regenerative PEC solar cells (DG ¼ 0) (direct conversion of solar to electrical energy) Photoelectrosynthetic cells (DGa0) (conversion of solar to chemical energy) ○ Photoelectrolysis cells (DG40) (endoergic conversion of solar to chemical energy), for example, 2 H2O-2H2 þ O2 ○ Photocatalytic cells (DGo0) (Exoergic conversion of solar to chemical energy), for example, N2 þ 3H2-2NH3

In the literature, there are many studies on water splitting by using SC photoelectrodes. The first PEC water splitting activity was reported by Fujishima and Honda [2] by using TiO2 electrodes. The photon energy is converted into chemical energy in these systems and as a result, there is a significant increase in Gibbs free energy. This reaction can be written as 1 H2 O þ hn-H2 þ O2 2

DG1 ¼ 238 kJ=mol

ð3Þ

There are over 100 photocatalytic metal oxide-based systems in the open literature that reported to achieve overall water splitting (simultaneous H2 and O2 production). Yet most of these photocatalysts have large band gaps and they require UV light (lo400 nm) irradiation. Fig. 11 shows that almost half of the solar energy on Earth’s surface is within the visible light region (400 nmolo800 nm). Therefore, large scale photocatalytic water splitting-based hydrogen production requires the efficient utilization of visible light. By using standard solar spectrum, maximum solar conversion efficiency for photocatalytic water splitting is calculated with a quantum efficiency of 100%. This amount would be only 2% even if the entire UV region up to 400 nm were exploited, which goes up to 16% when the light spectrum up to 600 nm is used and further to 32% when light up to 800 nm is utilized. For that reason, there are numerous studies in the literature on efficient water splitting by using visible light. The aim of this section is to summarize the current state of photoelectrode coating methods and materials for PEC hydrogen production. In this regard, photocurrent and photovoltage generations of different photoelectrodes are compared based on coating materials and methods for PEC hydrogen production. For each method, several coating materials’ photocurrent and photovoltage are compared as well as their reported light source in the literature. In order to compare different coating methods performances, average values are taken for each method.

2.17.7.1

Photoelectrochemical Hydrogen Production

In PEC cells, photonic energy is directly converted to hydrogen by using two abundant and clean sources, i.e., solar energy and water. In PEC systems, a SC and electrolyte junction integrates solar to electrical energy conversion with photoelectron chemical water splitting. PEC systems work with two electrodes, namely working and counter electrodes. A SC with either n- or p-type conductivity acts as a working electrode and generally platinum (Pt) is used as counter electrode. In addition to working and

542

Photoactive Materials

UV

Visible light ca. 2% (~ 400 nm)

ca. 16% (~ 600 nm) ca. 32% (~ 800 nm)

200

400

600

800

1000

Wavelength / nm Fig. 11 Solar spectrum and maximum solar light conversion efficiencies for water splitting reaction with 100% of quantum efficiency.

counter electrodes, most of the PEC systems have reference electrodes in order to examine half reactions in the cell. Once illuminated, photons with equal and higher energy level than the band gap (Eg) of the SC create electron and hole pairs on the working electrode. If a p type SC is used as working electrode, generated electrons are used to reduce H þ into H2 and the holes accept the electrons transferred from the counter electrode as a result of water oxidization into O2 and H þ . On the other hand, if an n-type SC is used, the holes oxidize water into O2 and H þ and electrons are transferred to the counter electrode in order to reduce H þ into H2. In summary, p-type SCs generate cathodic photocurrent by transmitting electrons toward the electrolyte where the holes leads to flow and in contrast, n-type SCs produce anodic photocurrent where holes are transferred toward the electrolyte. Depending on the conductivity type of the SC (i.e., p- or n-), namely electron and hole concentrations, the relative Fermi energy (Ef) position of the SC can be predicted. In n-type SCs, Fermi energy level is below the CB and in p-type SCs, the Fermi energy level is above the VB. The difference between energy levels drive a redox reaction when the SC electrode is immersed in a proper redox electrolyte. Separation of photogenerated e and h þ pairs in the p- or n-type SC (i.e., photoelectrode) causes photovoltage generation. When the Fermi energy of the SC and the redox potential are at the same level on both sides of the SC and electrolyte interface, an equilibrium is reached, and the transfer of charges stops. As a result, charge separation and photovoltage generation ends even when the SC is still illuminated. Regardless of the type of the SC (p- or n-), electrons are used to produce H2. This reaction happens in the working electrode if a p-type SC is used. H2 generation occurs at the counter electrode in cases where the working electrode is an n-type SC. Conductor and VB edge positions are key parameters in spontaneous water splitting. In order to spontaneously split water in a PEC system, the selected SC must have a CB energy higher than the reduction potential, VB energy lower than the oxidation potential, and a band gap around 2 eV. Because of the positive Gibbs free energy change (Eq. 3), water splitting is not a spontaneous reaction under normal operating conditions (e.g., room temperature and atmospheric pressure). In theory, to drive this reaction, 1.23 eV potential needs to be applied. However, recombination of photogenerated e and h þ pairs, electrode and component resistances, and voltage losses cause an increase in the theoretical energy requirement. In order to drive the spontaneous water splitting reaction, 1.8 eV is required. Therefore, selecting the optimum band gap for the SC (working electrode) is key to an efficient PEC water splitting. National Renewable Energy Laboratory (NREL) reports that by using GaAs PV and Ga-InP PEC junction as the bottom and top cells, they produced hydrogen with 16% solar to hydrogen (STH) conversion efficiency. However, there are long-term stability and cost issues. There are also low-cost, more stable WO3-based PEC systems producing H2 at STH efficiencies around 3–5%. STH efficiency is the ratio of chemical energy output (H2) to the solar energy input. Chemical energy output is calculated by multiplying the molar rate of hydrogen production (n_ H2 ) with the change in the Gibbs free energy per mol of H2 at room temperature (DG1). In order to calculate the solar energy input, illumination power density (P_ i ) is multiplied with the illuminated SC area (A): STHð%Þ ¼

n_ H2  DG1 P_ i  A

ð4Þ

In order to be considered as commercially viable, United States Department of Energy (US DOE) states a PEC system should meet the following criteria:

• •

Conversion efficiency: 10%. Current density (Jpc): 10 to 15 mA/cm2.

Photoactive Materials

• •

543

Material durability: 42000 h. Economic feasibility.

2.17.7.2

Photoelectrodes

Photoelectrode material selection is the most important design aspect of PEC systems since efficient water splitting requires an appropriate SC to convert photon energy to hydrogen. For efficient, clean, and economically feasible PEC-based hydrogen production, a SC photoelectrode should meet at least the following requirements:

• • • • • •

a band gap range of 1.8 and 2.2 eV for efficient visible light conversion; effective e and h þ pair separation; fast transport of separated charges to avoid recombination; suitable valance and CB edge positions; high chemical stability in electrolytic environment and with no corrosive nature; low cost.

In the open literature, there is almost no SC meeting all of the conditions listed above. For instance, ZnO, SnO2, BaTiO3, SrTiO3, WO3, and TiO2 are large band gap SCs, they are stable in PEC reactors, yet they can only harness photon energy within the UV region (4%) of solar spectrum. On the other hand, small and optimum band gap SCs like Cu2O, CuO, CdSe, CdTe, InP, GaAs, Si, etc. can utilize visible light region but they are not chemically stable in strong electrolyte environments. Without an (electro) chemically stable SC, it is not possible to design a long lifetime PEC system. The photoactive SCs should also drive the water splitting reaction only without permitting any side reactions such as electrode corrosion. Compared to n-type SCs, p-type SCs are more stable (and suitable) in PEC systems since SC materials are generally less resistant to oxidation reactions than reduction reactions. Also, photoelectrode SCs must be resistant to photocorrosion, i.e., oxidation caused by the photogenerated hole. Photocorrosion can be avoided if the rate of anodic oxidative reaction does not exceed the rate of charger transfer through the ntype SC/electrolyte interface. Catalytic surface treatments can potentially increase SC/electrolyte interface charge transfer rates and improve both p- and n-type SC photoelectrode stabilities. Since hydrogen production performance of a PEC system is greatly affected by the type of SC photoelectrode, there are many studies in the literature focusing on the design, preparation, processing, and modification of photoactive SCs. Electron transfer process, band gap energy, and band structure of SCs are the factors that significantly affect the efficiency of a photoelectrode. The existing photoelectrodes can be modified to (1) enhance their response to visible light by band gap altering, (2) increase electron transfer rates, and (3) modify VB and CB energy levels. Some of the available photoelectrode modification techniques are doping, dye sensitization, swift heavy ion irradiation (SHI), metal ion loading, ED, etc. Some of the benefits of using modified photoelectrodes are listed as:

• • • •

Harnessing a wider part of the solar spectrum including the visible light region. Efficient e and h þ generation and separation. Prevention of e and h þ recombination. Increased lifetime.

2.17.8

Photoelectrode Coating Methods by Using Photoactive Materials

PEC water splitting efficiency is significantly affected by several factors such as crystalline structure imperfections and SC properties. Bulk and surface properties of the SC photoelectrode depends on the coating methods and conditions.

2.17.8.1

Chemical Vapor Deposition

CVD is suitable for deposition of either single or multicomponent thin films with controlled stoichiometry and morphology. Therefore, CVD is used in fabricating photoactive materials for dye sensitized solar cells, photocatalysis, and PEC cells. The voltage and light source characteristics of selected CdS, TiO2, ZnO, and WO3-based photoelectrodes prepared by CVD are listed in Table 6. Photocurrent generation performances of these photoelectrodes are presented in Fig. 12. Although TiO2 gives low photocurrents alone, combined with Fe, Zn, Fe2O3, or with silicon, TiO2-based photoelectrodes processed with CVD can generate significant photocurrents. A CVD processed, CdS–ZnO–CdSe-based photoelectrode also produces significantly high photocurrent compared to other selected photoelectrodes fabricated by using CVD method.

2.17.8.2

Electrochemical Deposition and Electrodeposition

Among available photoelectrode coating methods, ECD and ED have some significant advantages such as low operating temperature and cheap and simple operation process. Also, ECD and ED methods provide accurate control during the coating process,

544

Photoactive Materials

Table 6

Voltage and light source requirements of selected photoelectrodes coated by using chemical vapor deposition (CVD) method

Photoelectrode

Reference

Voltage

Light source

ZnO/CdS n-TiO2 n-SiNW/TiO2 CdS–ZnO–CdSe ZnO/CdS TiO2 WO3/TiO2 Fe–TiO2 Fe–TiO2/Zn–Fe2O3

[59] [60] [27] [61] [59] [36] [62] [63] [63]

0 V vs. SCE 3 V vs. SCE

1000 W Xenon (Xe) lamp Xe short arc light (100 mW/cm2)

0.4 V vs. Ag/AgCl 0V 0.8 V vs. Ag/AgCl

1000 W Xe lamp (100 mW/cm2) Halogen lamp 300 W Xe lamp (6.7 mW/cm2)

0.95 vs. SCE

150 W Xe lamp (150 mW/cm2)

1800

Photocurrent density (μA/cm2)

1600 1400 1200 1000 800 600 400 200

iO

Fe

-T iO

2 /Z

n-

Fe

3 /T

O

2O 3

2

2

iO

2

O

W

/C O Zn

nO

Fe

C

-T

dS

-Z

Si

Ti

dS

e -C

/T

dS

iO

2

2

O Ti

N

Zn

W

O

/C

dS

0

Fig. 12 Photocurrent densities of selected photoelectrodes coated by using chemical vapor deposition (CVD) method. Table 7

Voltage and light source requirements of selected photoelectrodes coated by using electrochemical deposition (ECD) method

Photoelectrode

Reference

Voltage

Light source

SnO2/CdSe TiO2/TiSi2 TiO2 Cu2O/TiO2 CdS CdSe CdS/CdSe TiO2 TiO2/SrTiO3 TiO2 TiSi2/TiO2

[64] [65] [31] [66] [20] [14] [67] [12] [68] [27] [69]

0 V vs. SCE 0 V vs. Ag/AgCl 0 V vs. SCE

250 W Xenon lamp (300 W/cm2) 150 W Xenon lamp (100 mW/cm2) Simulated sunlight (33 mW/cm2)

0 V vs. SCE

500 W Tungsten lamp (40 mW/cm2)

0.2 V vs. SCE

450 W Xenon lamp

0.2 V vs. Ag/AgCl

300 W Solar simulator with AM 1.5 filter (87 mW/cm2)

which makes these techniques potential candidates in a wide variety of applications both in small and large scales. Voltage and light source characteristics of selected CdS/CdSe, TiO2, and Cu2O-based photoelectrodes prepared by ECD are listed in Table 7. Photocurrent generation performances of selected photoelectrodes prepared by ECD are presented in Fig. 13. Similar to the CVD results, when used alone, TiO2 produces low photocurrents. However, combined with SrTiO3 or silicon; TiO2-based photoelectrodes processed with ECD can generate comparably high photocurrents. ECD processed, CdS and/or CdSe-based photoelectrodes also generate significantly high photocurrents compared to other selected photoelectrodes fabricated by using ECD method.

Photoactive Materials

545

4000

Photocurrent (μA/cm2)

3500 3000 2500 2000 1500 1000 500

2

O rT i

Ti O 2 Si 2 /T iO Ti

3

2

O

2 /S

Ti O Ti

C dS e C dS /C dS e

2O

C dS

2

O

2

/T i

O Ti

C u

iS 2i 2 /T

O Ti

Sn O

2 /C

dS e

0

Fig. 13 Photocurrent densities of selected photoelectrodes coated by using electrochemical deposition (ECD) method. Table 8

Voltage and light source requirements of selected photoelectrodes coated by using electrodeposition (ED) method

Photoelectrode

Reference

Voltage

Light source

Cu2O/TiO2 CdS/TiO2 CdSe ZnSe CdSe/ZnSe Ti/p–CuO/n-Cu2O/Au Cu2O/Al–ZnO/TiO2/Pt TiO2 BiOI BiOI/TiO2 Fe2O3/ZnFe2O4 Al–Fe2O3/ZnFe2O4 Cu2O/CuO

[66] [70] [14] [67] [67] [47] [5] [14] [71] [71] [72] [72] [73]

1 V vs. Ag/AgCl 0.2 V vs. Ag/AgCl 0V

Xenon lamp (700 W/cm2) Xenon lamp (50 mW/cm2) 200 W Tungsten lamp (1950 lx)

0 V vs. SCE 0 V vs. Ag/AgCl 0.5 V vs. Ag/AgCl

100 W Tungsten lamp (90 mW/cm2) AM 1.5 (100 mW/cm2) 300 W Xenon lamp (25 mW/cm2)

0.4 V vs. Ag/AgCl

300 W Xenon lamp with AM 1.5G filter (100 mW/cm2)

0 V vs. Ag/AgCl

AM 1.5G (100 mW/cm2)

The voltage and light source characteristics of selected CdS/CdSe, TiO2, Fe2O3, and Cu2O-based photoelectrodes prepared by ED are listed in Table 8. The photocurrent generation performances of selected photoelectrodes prepared by ED are presented in Fig. 14. ED results agree on CVD and ECD that TiO2 photoelectrodes generate very low photocurrents when used alone. However, by combining TiO2 with Cu2O, CuO, ZnO, and Pt, TiO2-based photoelectrodes processed with ED produce considerably high photocurrents.

2.17.8.3

Sol–Gel

SG is a material processing method that is used to produce solid products from small molecules. The most common application of this method is metal oxide (e.g., silicon and/or titanium oxides) production. In SG, first a colloidal solution (sol) is produced, which is used to create an integrated network (gel). SG is another good SC coating method since SG is suitable for solar cell applications. Some of the advantages of photoelectrode coating by SG method are fabrication of photoelectrodes with efficient charge separation, proper surface properties, and I–V characteristics. The voltage and light source characteristics of selected SnO2, CdSe, TiO2, Fe2O3, and ZnO-based photoelectrodes prepared by SG are listed in Table 9. The photocurrent generation performances of selected photoelectrodes prepared by SG are presented in Fig. 15. Among the selected photoelectrodes coated by using SG method, ZnO/SnO2 gives the highest photocurrent density, followed by TiO2/SnO2, and Zn–Fe2O3/Fe–TiO2. On the other hand, WO3/Fe2O3, SrTiO3/a-Fe2O3, TiO2:N, and Zn–Fe2O3 generate the lowest photocurrent densities compared to other selected SG processed photoelectrodes.

2.17.8.4

Spin Coating

SC is a common method used to coat thin, uniform layers on planar surfaces. During SC, the deposition solution is first placed on the substrate, then the substrate is accelerated rapidly to the desired rotation rate. Due to the centrifugal force, the deposition

546

Photoactive Materials

6000

Photocurrent (μA/cm2)

5000 4000 3000 2000 1000

/C uO

2O

Al -F

e

2O 3 /Z

nF

e

e nF

2O 3 /Z

C u

O I/T i Fe

2O 4

2O 4

2

I Bi O

Bi O

Ti

O

2

Zn Se C dS e/ Zn uO Se /n C u -C 2O u /A 2O l-Z /A u nO /T iO 2 /P t Ti

/p -C

C dS e

O

C dS /T i

C u

2O

/T i

O

2

2

0

Fig. 14 Photocurrent densities of selected photoelectrodes coated by using electrodeposition (ED) method.

Table 9

Voltage and light source requirements of selected photoelectrodes coated by using sol–gel (SG) method

Photoelectrode

Reference

Voltage

Light source

TiO2/SnO2 ZnO/SnO2 WO3/Fe2O3 SrTiO3/a-Fe2O3 Zn–Fe2O3 Zn–Fe2O3/Fe–TiO2 TiO2:N CdSe/TiO2 CdSe/TiO2:N

[58] [74] [75] [76] [63] [63] [64] [61] [64]

0V 0V B0.7 V 0.3 V 0.95 V vs. SCE

400 W Xenon lamp (30 W/m2) 500 W Xenon lamp 400 W Xenon lamp (240 mW/cm2) 150 W Xenon lamp (150 mW/cm2)

0 V vs. Ag/AgCl

1000 W Xenon arc lamp (100 mW/cm2)

900 Photocurrent (μA/cm2)

800 700 600 500 400 300 200 100 2 :N

e/ dS C

C

dS

e/

Ti O

Ti O

2

2 :N

Ti O

2

iO eT

2O 3

Fe

2O 3 /F

-F

e

2O 3

Zn -

Sr

Ti

O

Zn

-F e

2O 3

3 /

3 /F

e

nO

O W

O /S Zn

Ti

O

2 /S

nO

2

2

0

Fig. 15 Photocurrent densities of selected photoelectrodes coated by using sol–gel (SG) method.

solution covers the desired surface area radially. At the end of the coating process, excess solution is drained off. The deposition solution used in SC is usually volatile, so it simultaneously evaporates as the substrate spins. As the angular speed is increased, the coating becomes thinner. The voltage and light source characteristics of selected Fe2O3, WO3, SrTiO3, CdS/CdSe, TiO2, and ZnO-based photoelectrodes prepared by SC are listed in Table 10.

Photoactive Materials

Table 10

547

Voltage and light source requirements of selected photoelectrodes coated by using spin coating (SC) method

Photoelectrode

Reference

Voltage

Light source

Fe2O3 WO3/Fe2O3 a-Fe2O3 SrTiO3 SrTiO3/a-Fe2O3 TiO2/CdS TiO2/CdS TiO2/CdSe TiO2/CdS/CdSe WO3/BiVO4

[63] [75] [63] [76] [76] [61] [67] [61] [67] [77]

B0.7 V vs. Ag/AgCl

500 W Xenon lamp

0.3 V 0.3 V 0.3 V vs. Ag/AgCl 0.5 V vs. Ag/AgCl B0.7 V vs. Ag/AgCl

400 W Xenon lamp (240 mW/cm2)

150 W Xenon lamp (100 mW/cm2) 300 W Xenon lamp with AM 1.5G filter (100 mW/cm2)

1 V vs. SCE

150 W Xenon lamp with AM 1.5G filter (100 mW/cm2)

6000

Photocurrent (μA/cm2)

5000

4000

3000

2000

1000

O

4

e

iV 3 /B

O W

Ti

O

2 /C

Ti

O

dS

2 /C

/C

dS

dS

e

dS Ti

O

2 /C

dS Ti

-F

3 /

Ti

O

O

e

2 /C

2O 3

3

O Ti Sr

2O 3

e F

Sr

W

O

3 /F

Fe

e

2O 3

2O 3

0

Fig. 16 Photocurrent densities of selected photoelectrodes coated by using spin coating (SC) method.

The photocurrent generation performances of selected photoelectrodes prepared by SC are presented in Fig. 16. Among the selected photoelectrodes coated by using SC method, TiO2/CdS/CdSe gives the highest photocurrent density, followed by TiO2/CdS and TiO2/CdSe. On the other hand, SrTiO3/a-Fe2O3, Fe2O3, WO3/Fe2O3, and SrTiO3 generate the lowest photocurrent densities compared to other selected SC processed photoelectrodes.

2.17.8.5

Spray Pyrolysis

SP is an attractive coating method suitable for applying a wide variety of thin films (e.g., noble metals, metal oxides, superconducting compounds, etc.) on different substrates. One major advantage of SP is its simplicity and efficiency in both small and large scales. Other advantages of SP can be listed as (1) ease of coating with almost any compound and element alone or combined with others in any proportion; (2) no high quality substrate, vacuum, or extreme operating temperature and pressure requirements; (3) scale up availability for industrial applications; (4) efficient control of coating parameters such as film thickness; (5) no chemical or pollutant side products; (6) no restrictions on substrate dimensions and profile; and (7) ability to coat a wide variety of film thicknesses on any substrate. The voltage and light source characteristics of selected TiO2, Fe2O3, WO3, CdS/CdSe, and CuO-based photoelectrodes prepared by SP are listed in Table 11. The photocurrent generation performances of selected photoelectrodes prepared by SP are presented in Fig. 17. Among the selected photoelectrodes coated by using SP method, CdS/CdSe has the highest photocurrent density, followed by WO3/Fe2O3 and Fe2O3. On the other hand, TiO2 and ZnFe2O4/TiO2 generate the lowest photocurrent densities compared to other selected SC processed photoelectrodes.

548

Table 11

Photoactive Materials

Voltage and light source requirements of selected photoelectrodes coated by using spray pyrolysis (SP) method

Photoelectrode

Reference

Voltage

Light source

TiO2 ZnFe2O4/TiO2 Fe2O3 WO3/Fe2O3 TiO2 CdS/CdSe TiO2 CuO/WO3/TiO2

[5] [78] [63] [75] [38] [67] [40] [62]

0 V vs. SCE

200 W Xenon lamp

1.43 V 1.43 V vs. Ag/AgCl 1 V vs. Ag/AgCl

450 W Xenon lamp (100 mW/cm2) 150 W Xenon lamp (100 mW/cm2)

0.75 V vs. Ag/AgCl

300 W UV lamp

5000 4500

Photocurrent (μA/cm2)

4000 3500 3000 2500 2000 1500 1000 500

2

iO

2

O

3 /T

O

dS

C

uO

/W

dS

/C

Ti

e

2

O

C

3 /F

O

Ti

2O 3

e

2O 3

Fe

W

Zn

Fe

2O 4 /T

Ti

O

iO

2

2

0

Fig. 17 Photocurrent densities of selected photoelectrodes coated by using spray pyrolysis (SP) method.

2.17.8.6

Overall Comparison

In this case study, several characteristics of photoelectrodes coated by different methods and materials are investigated. Selected coating methods are CVD, ECD, ED, SG, SC, and SP. For each coating method, various major coating materials’ photocurrent density and voltage/light requirements are presented. In Fig. 18, the average photocurrent densities of each coating method are presented for comparison purposes. Tables 4–11 present voltage and light source requirements of the selected photoelectrodes by using each method. The average photocurrent densities for the CVD method are calculated based on the data presented in Ref. [18]. Fig. 18 shows that the photoelectrodes coated with SC produce the highest average photocurrent density. The SC offers several advantages compared to other methods such as the simplicity, relative ease, and the possibility of making very well defined uniform film thicknesses, which could have an impact on tuning photocurrent generation. Higher consistency at both macroscopic and nanoscales can be achieved via SC due to the possibility of achieving high spin speeds and fast drying times. The second highest average photocurrent density is ECD, followed by ED and SP. CVD and SG have the lowest average photocurrent densities compared to selected coating methods. It should be noted that these values are the averages of data presented in Figs. 12–17. Despite the fact that it gives an interesting insight on different coating methods, further investigation should be conducted on different coating materials. For that purpose, CdS, CdSe, TiO2, WO3, Fe2O3, and CuO/Cu2O-based photoelectrode coatings’ photocurrent densities are compared in Fig. 19. Among selected materials, CdS and CdSe-based photoelectrode coatings have the highest average photocurrent densities. This could be explained by superior light harvesting ability of CdS/CdSe-based photoactive materials and their effective charge separation performances. These factors could possibly improve their possible use for photocatalytic hydrogen generation. TiO2, WO3, and CuO or Cu2O-based photoelectrode coatings generate moderate amounts of average photocurrent densities, compared to the selected materials. Fe2O3-based photoelectrode coatings have the lowest average photocurrent density.

Photoactive Materials

549

Average photocurrent (μA/cm2)

2500

2000

1500

1000

500

0 CVD

ECD

ED

SG

SC

SP

Fig. 18 Average photocurrent densities of selected photoelectrode coating methods.

4000

Average photocurrent (μA/cm2)

3500

3000

2500

2000

1500

1000

500

0 CdS-based

TiO2-based

CdSe-based

WO3-based

Fe2O3-based

CuO/Cu2O-based

Fig. 19 Average photocurrent densities of selected photoelectrode coating materials.

2.17.8.7

Key Findings

In this case study, various photoelectrode coating materials for PEC hydrogen production are selected, evaluated under numerous coating methods, and compared. The selected coating methods are CVD, ECD, ED, SG, SC, and SP. First, photocurrent density generation rates and voltage/light requirements of various coating materials are compared for each method. Then, the average photocurrent densities of each method and material group are compared for further evaluation purposes. Selected photoelectrode coating material groups are CdS, CdSe, TiO2, WO3, Fe2O3, and CuO/Cu2O. The key findings of this study can be listed as

• • •

Within photoelectrodes coated by CVD, TiO2 has the lowest (22 mA/cm2) and Fe–TiO2/Zn–Fe2O3 has the highest (1650 mA/cm2) photocurrent densities. SnO2/CdSe has the lowest (37 mA/cm2) and CdS/CdSe has the highest (3600 mA/cm2) photocurrent density among the photoelectrodes coated by ECD method. In the ED group, Cu2O/Al–ZnO/TiO2/Pt has the highest (5600 mA/cm2) and TiO2 has the lowest (9 mA/cm2) photocurrent densities.

550

• • • • • •

Photoactive Materials

Among photoelectrodes coated by SG, ZnO/SnO2 has the highest (892 mA/cm2) and WO3/Fe2O3 has the lowest (2.5 mA/cm2) photocurrent densities. In SC, SrTiO3/a-Fe2O3 has the lowest (52.7 mA/cm2) and TiO2/CdS/CdSe has the highest (5900 mA/cm2) photocurrent density. TiO2 has the lowest (40 mA/cm2) and CdS/CdSe has the highest (5000 mA/cm2) photocurrent density among the photoelectrodes coated by SP method. When average photocurrent densities of each coating method are compared, SC has the highest (2343.57 mA/cm2) and SG has the lowest (335.06 mA/cm2) average photocurrent density. Average photocurrent densities of selected major photoelectrode coating materials show that CdS-based photoelectrodes have the highest (3715.58 mA/cm2) and Fe2O3-based photoelectrodes have the lowest (445.3 mA/cm2) photocurrent density. The highest photocurrent is reached via TiO2/CdS/CdSe and by using SC method.

2.17.9

Hydrogen Production Methods

As an abundant element, hydrogen can be found in many substances in nature (i.e., fresh and sea water, biomass, hydrogen sulfide, and fossil fuels). In order to produce hydrogen with zero or low environmental impact (“green” hydrogen), all CO2 and other pollutants must be processed (i.e., separated or sequestrated) when hydrogen is extracted from fossil fuels. Thermal, electrical, photonic, and biochemical energy systems are the primary energy sources to generate hydrogen. Table 12 shows an overview and brief description of hydrogen production methods assessed in this study along with their primary energy and material sources. Electrical and thermal energies can be generated from fossil fuels (must be processed to be considered as “clean”), renewable energies (i.e., solar, wind, hydro, wave, ocean, and thermal), biomass, nuclear, or recovered energy. Photonic energy comes from solar irradiation, while biochemical energy is recovered from organic matter. In addition to the four major primary sources listed in Table 12 (electrical, thermal, biochemical, and photonic), there are also hybrid forms of energy such as electrothermal, photobiochemical, and PEC. Water, biomass, and fossil fuels are the material sources. In cases where fossil fuels are utilized, hydrogen production process includes CO2 separation and sequestration. The various paths through which various forms of energy sources are used to drive hydrogen production that can be obtained from green energy sources are presented in Fig. 20. For electrolysis, clean energy sources can be used such as solar, geothermal, biomass, wind, ocean, or other renewables. In nuclear energy plants, electricity produced can be used to drive electrolysis reactions for hydrogen production. The heat recovered from any process, system, or application (e.g., waste heat from an industrial process) can also effectively be used to run power generating systems, such as organic Rankine cycle (ORC). Numerous types of heat engines and/or combustion with carbon capture can be used in power generating systems as well. In solar electrolysis, PV or concentrated solar power (CSP)-based systems are used to generate the electricity required for electrolysis. Geothermal electrolysis power sources can be listed as ORC, flash cycle, etc. Wind electrolysis can be coupled to both grid connected and autonomous wind power plants. Other renewables are tides, ocean currents, and wave energy, which can be converted into electricity to power the electrolysis reaction. Electrolysis of the water reaction is generally written as 2H2 OðlÞ -2H2ðgÞ þ O2ðgÞ

ð5Þ

For water thermolysis, also known as single step thermal dissociation of water, the reaction can be written as heat

H2 O ⟶ 2H2 þ O2

ð6Þ

In order to accomplish a reasonable degree of dissociation, the reaction requires a heat source that could provide temperatures above 2500K. The high temperature steam can be produced by using concentrated solar heat to reach the required high temperature levels for thermolysis. Thermochemical processes are essentially driven by heat. If we use such processes for hydrogen production, it refers to thermochemical hydrogen production. One of the most important advantages of thermochemical cycles is the requirement of no catalysts to drive or support any steps of the chemical reactions to produce hydrogen. All chemicals, but not water (which is the main material resource to produce hydrogen), used in thermochemical cycles can be recycled. Other advantages of thermochemical water splitting cycles can be listed as: (1) no membranes needed to separate hydrogen and oxygen, (2) reasonable temperature requirement range of 600 and 1200K, and (3) zero or low electrical energy requirement. This energy requirement can be met by concentrated solar, geothermal, biomass combustion, nuclear electricity, or recovered energy (i.e., from landfill gas combustion). S–I, Cu–Cl, and Mg–Cl are some examples of thermochemical water splitting. Concentrated solar or autothermal biomass combustion can be utilized to support and run biomass conversion and gasification processes to produce hydrogen. When using biomass to extract hydrogen, the moisture content should be kept below a certain level by drying or supercritical steam gasification. Some of the examples of biomass are wood sawdust and sugar cane. The general biomass conversion is: heat

aCl Hm On þ bH2 O ⟶ aH2 þ bCO þ cCO2 þ dCH4 þ eC þ f Tar

ð7Þ

Biomass gasification reaction is generally expressed as  high temperature heat y Cx Hy þ xH2 O ⟶ x þ H2 þ xCO 2

ð8Þ

Table 12

Overview of H2 production methods by primary energy and material sources Source

Electrolysis Plasma arc decomposition Thermolysis Thermochemical processes

Photovoltaic (PV) electrolysis Photocatalysis Photoelectrochemical (PEC) Dark fermentation High temperature electrolysis Hybrid thermochemical cycles Coal gasification Fossil fuel reforming Biophotolysis Photofermentation Artificial photosynthesis Photoelectrolysis

Brief description

Primary energy

Material

Electrical Thermal

Water Fossil fuels Water Water Biomass

Photonic

Water

Biochemical Electrical þ thermal

Biomass Water

Photonic þ biochemical

Biomass þ water

Electrical þ photonic

Water

Water splitting Biomass conversion Gasification Reforming

Direct current is used to split water into O2 and H2 (electrochemical reaction) Cleaned natural gas pass through plasma arc to generate H2 and carbon soot Thermal decomposition of water (steam) at temperatures over 2500K Cyclical reactions (net reaction: water splitting into H2) Thermocatalytic conversion Conversion of biomass into syngas Conversion of liquid biomass (biofuels) into H2 PV panels are used to generate electricity Water is split into H2 by using the electron and hole pair generated by the photocatalyst A hybrid cell simultaneously produces current and voltage upon absorption of light Biological systems to generate H2 in the absence of light Electrical and thermal energy are used together to drive water splitting at high temperatures Electrical and thermal energy are used together to drive cyclical chemical reactions Conversion of coal into syngas Fossil fuels are converted to H2 and CO2 Biological systems (microbes, etc.) used to generate H2 Fermentation process activated by exposure to light Artificial systems mimic photosynthesis to generate H2 Photoelectrodes and external electricity are used to drive water electrolysis

Photoactive Materials 551

552

Photoactive Materials

Green energy

Thermal energy

Electrical energy

Biochemical energy

Tides and waves

Ocean thermal

Hydro

Biomass

Wind

Renewable energy

Solar

Nuclear energy

Geothermal

Energy recovery (Industrial waste heat waste incineration landfill gas etc)

Photonic energy

Green hydrogen Fig. 20 Green H2 production through four energy types and main renewable energy sources.

In photocatalysis, photonic energy (comes from higher energy spectrum of solar irradiation, namely UV and upper portion of visible spectrum) is converted to chemical energy (hydrogen). Photonic energy is proportional to the frequency of the radiation and given by hn where h is the Planck’s constant and n is the frequency. When a photon hits the photocatalyst, an electron and hole pair is generated and the obtained electrical charge is utilized to dissociate water. In order for a photocatalyst to split water and generate hydrogen, it should have an appropriate band gap and properly located conduction and VBs for oxidation and reduction reactions. Furthermore, rapid generation and separation of electron and hole pairs is essential when picking an appropriate photocatalyst. In past literature, SCs (such as TiO2) and metal oxides (such as Fe2O3) are heavily studied as photocatalysts. Also, chemically modified and engineered complex supramolecular devices are utilized to perform photocatalytic reactions. Acar et al. [17] have reviewed and assessed various simple and complex photocatalysts based on their hydrogen production yield, efficiency, and impact on human health and the environment. Photoreduction and photooxidation reactions can be written as hn

Photoreduction: 2H2 O þ 2e ⟶ þ H2 þ 2OH hn

Photooxidation: 2H2 O ⟶ þ O2 þ 4Hþ þ 4e

ð9Þ ð10Þ

A PEC cell essentially contains one or two photoelectrodes (i.e., photoanode and photocathode). In a PEC, there is at least one electrode that is a SC. The photonic energy is absorbed by the photoelectrode(s) in a PEC. PV cells and PEC SC electrodes have analogous operating mechanisms. In both systems, electron and hole pairs are produced by the photons with energies higher than the band gap of the photoactive surface. Such electron and hole pairs are used to either reduce or oxidize water. PEC can use the

Photoactive Materials

553

solar spectrum more efficiently compared to photocatalysis and PV electrolysis since it combines water electrolysis and photocatalysis in one single element. Since PECs do not have the need of additional power supplies, they are more compact, which is one of the major advantages of these systems. There are many kinds of photosensitive SCs investigated in the literature. The most promising option so far is TiO2. In addition to TiO2, several other SCs have been studied, such as ZnO, Fe2O3, BiVO4, and WO3. Metal nitrides and phosphides (such as Ta3N5 and GaP), metal oxynitrides (such as TaON), and n- and p-type silicon have also been investigated in the open literature. Rabbani et al. [20] have coupled PEC with chloralkali cells and tested the system in batch type. Acar and Dincer [5] have combined and enhanced the studies on PEC and chloralkali reactors in a continuous type hybrid system. Biochemical energy, which is found in organic material, can be used by living creatures to obtain hydrogen either in the presence or absence of light. Dark fermentation is the conversion of biochemical energy stored in organic matter to other forms of energy in the absence of light (this case might happen when there is reduced supply of light). The bioreactors used for dark fermentation are simpler and cheaper compared to photofermentation since the process does not require solar input processing. Hydrogen production by dark fermentation has several other advantages such as the ability to produce hydrogen from organic waste and therefore control and stabilize biological waste, which has a potential danger of contamination. For instance, dark fermentation can be integrated into wastewater treatment systems to produce hydrogen from wastewater. Producing hydrogen from organic waste has a potential to reduce hydrogen production costs since organic waste (including wastewater) is cheap and readily available. High temperature electrolysis is a method of electrolysis where steam is dissociated to hydrogen and oxygen at temperatures between 700 and 10001C. This method is generally considered as more efficient than conventional room temperature electrolysis (efficiency increases with increasing temperature). In high temperature electrolysis, water is converted to steam by using thermal energy. The system components are either heated directly by the steam supply or indirectly by heat transfer. Thus the electrical energy need of this type of electrolysis is lower than that of conventional electrolysis methods. Another advantage of this method is the possibility of achieving zero greenhouse gas emissions when a clean heat source (such as solar, geothermal, and/or nuclear) is used as external heat source. However, due to high operating temperatures, the system components have to meet specific requirements for an efficient hydrogen generation. Current challenges of high temperature electrolysis can be listed as (1) chemically stable electrolyte development with high ionic and low electronic conductivity, (2) porous, chemically stable electrode research in highly reducing/oxidizing environments with good electronic conductivity and coefficient of thermal expansion similar to the electrolyte, and (3) engineering chemically stable materials at high temperatures and highly reducing/oxidizing environments. Hybrid thermochemical cycles operate at lower temperatures compared to thermally driven water splitting cycles. External energy needs of the individual electrochemical reactions are met by thermal and electrical energies. Since hybrid cycles operate at lower temperatures, other sustainable thermal sources other than high temperature nuclear, solar, and/or biomass combustion (such as recovered waste heat from any process or system and nuclear and geothermal heats) can be utilized to provide the required energy for the involved processes. Cu–Cl, Mg–Cl, and S–I cycles are some examples of hybrid thermochemical cycles. Operating temperatures of hybrid thermochemical cycles are lower than traditional thermochemical water splitting cycles. Therefore, a major advantage of these cycles is hydrogen generation from low grade temperature sources, especially those that can be considered as sustainable thermal energy. Nuclear heat, industrial heat, waste heat recovered from power plants, concentrated solar heat, heat resulting from municipal waste incineration, and geothermal heat can be listed as sustainable thermal energy sources. In this process, heterogeneous photocatalysts are applied on one or both of the electrodes. In addition to solar irradiation exposure, an electrolysis cell needs to be supported by electrical energy to conduct photoelectrolysis. Therefore, in photoelectrolysis, both photonic and electrical energies are converted to chemical energy (and hence hydrogen). A photoelectrolytic hydrogen production mechanism includes the following steps: (1) generation of an electron–hole pair with the help of a photon that has sufficiently high energy (higher than the band gap of the p and n junction), (2) flow of electrons from the anode to the cathode generating electricity current, (3) decomposition of water into protons and gaseous oxygen, (4) reduction of protons at the cathode to form hydrogen in gaseous form, and (5) separation of the product gases, processing, and storage. The performance of a photoelectrolytic system depends on the type of photon absorbing material, its crystalline structure, surface properties, corrosion resistance, and reactivity. There usually is a tradeoff between photoelectrode stability and photon energy to hydrogen conversion efficiency: high efficiency photoelectrodes generally have poor stability in electrolytes while the chemically stable photoelectrodes show poor water splitting efficiencies. Biophotolysis and photofermentation processes are known as photonic driven biochemical hydrogen generation techniques from water. In biophotolysis, some light sensitive microorganisms are used as biological converters in a specially designed photobiochemical reactor. Among various possible microorganisms, the most suitable ones are microalgae since they can be cultured. They have a potential to generate hydrogen in closed systems, which permits hydrogen capture. Cultured microalgal strains show high hydrogen yields. The major advantage of biophotolysis is the ability to produce hydrogen from water in an aqueous environment at ambient temperature and pressure. However, it is only demonstrated at laboratory scale and not yet fully technologically advanced to support industrial and commercial utilization. A general hydrogen generation reactions with the help of photoactivated enzymes are: hn

6H2 O þ 6CO2 ⟶ C6 H12 O6 þ 6O2 hn

C6 H12 O6 þ 6H2 O ⟶ 6CO2 þ 12H2

ð11Þ ð12Þ

554

Photoactive Materials

Artificial photosynthesis is a biomimetic process mimicking the natural photosynthesis process to accomplish the following steps:

• •

•

PV-based electricity generation: to support the grid system. Dry agriculture: with this method, carbohydrates (food), liquid fuels, chemical feed stocks, and polymers for fiber manufacture can be produced with near or absolute chemical minimum water usage. This amount is thousands of times lower than the conventional agriculture water usage. The system has an enzyme bed reactor system that fixes CO2 from the air (or other convenient sources), and it is powered by hydrogen and bioelectric transducers. Hydrogen production: electrolytic decomposition of water can be achieved by mimicking photosynthesis.

Although the technology is not mature enough to be applied to large scale hydrogen production, artificial photosynthesis has a significant potential to lower global water usage and support clean energy systems by generating electricity and hydrogen from photonic energy. There is a significant amount of research underway in terms of development, testing, and analysis of sustainable hydrogen production methods. Most of the options discussed here are not yet commercially available due to high production costs and/or low efficiencies. However, hydrogen holds a key to transition to CO2 free energy systems. Cetinkaya et al. [21] have conducted a comprehensive life cycle assessment (LCA) by including all of the major steps for every method, comparing energy consumption and CO2 equivalent emissions to examine their environmental impact for natural gas reforming, coal gasification, water electrolysis by wind and solar electricity, and thermochemical water splitting by the Cu–Cl cycle. Dincer [19] has discussed hydrogen production options based on some potential driving energy sources such as electrical, thermal, biochemical, photonic, electrothermal, photothermal, photoelectric, photobiochemical, and thermobiochemical. Hacatoglu et al. [22] have used LCA to comparatively assess hydrogen production from natural gas, wind, solar, and nuclear-based Cu–Cl hybrid thermochemical cycle and compared their results to gasoline. Ozbilen et al. [23] have investigated and compared the environmental impact and efficiencies of various hydrogen production methods, namely steam gas reforming, coal gasification, wind and solar-based electrolysis, and nuclear-based hybrid thermochemical Cu–Cl and S–I cycles. The primary focus of this study is solar-based hydrogen, therefore, light-based hydrogen production is discussed in detail in the following sections.

2.17.10

Photoactive Materials in Hydrogen Production

Solar energy can be used to convert photonic and or thermal energy that is stored in the sunlight into hydrogen. Thermochemical, photobiological, and photocatalytic water splitting are some of the examples of hydrogen production from solar water splitting. Light (solar energy)-based hydrogen production is generally classified as PV, solar thermal, photoelectrolysis, and biophotolysis. Among these options, PV, photoelectrolysis, and biophotolysis are considered as low temperature applications. Solar thermal can be used in either low or high temperature applications. High temperature applications are generally referred to as concentrated solar energy; some of these applications are thermolysis, thermochemical cycles, gasification, reforming, and cracking. Concentrated solar energy can also be used to generate electricity to drive electrolysis and produce hydrogen. Fig. 21 presents the major solar hydrogen production pathways. Solar thermochemical processes have first been demonstrated in an indoor environment by using a high flux solar simulator (HFSS) to understand the process dynamics and reactor performance. Pitz-Paal et al. [25] have proposed an algorithm to maximize annual solar to chemical energy conversion efficiency of solar thermochemical process. Their methodology had been illustrated for solar reduction of zinc oxide and coal gasification processes. There are review papers on various solar thermochemical processes. These papers review a particular process like separation of solar thermolysis products, methane reforming and decomposition, and gasification of solid carbonaceous feedstocks. Meier and Steinfeld [26] have discussed fuel production and industrial applications of solar thermochemical technology. These review papers address the issues of fundamental science and technology development in the field of solar thermochemical processing. It can be observed that the requirement of high temperature is an impeding force in development of solar thermolysis and with the current state of technology this is not a viable option. High temperature electrolysis, where the temperature required for dissociation of water is brought down by supplying a part of energy in the form of electricity, might be a viable alternative. The cost of hydrogen obtained by high temperature electrolysis, with a parabolic trough solar field for steam generation and PV plant for electrolysis, is estimated at 5.23 USD/kg. Net energy analysis is another tool for assessing the viability of a process. Net energy analysis involves estimating the total material and energy consumed in construction and operation of a plant. Based on net energy analysis, a process can be defined as net energy positive or negative. If the process is net energy positive, the energy payback period can also be estimated. More detailed information on a net energy analysis was reported by Bullard et al. [27]. A past net energy analysis of a parabolic trough system can be found in Krishnamurthy and Banerjee [28]. However, net energy analysis of solar thermochemical processes is missing in the open literature. Yadav and Banerjee [29] have reviewed and compared solar thermochemical hydrogen production methods from experimental demonstration, thermodynamic, cost and life cycle analyses points. Processes like solar methane reforming have been successfully demonstrated on a pilot scale worldwide and operated for longer duration of time. Long-term suitability of this technology, reactor stability and lifespan with daily heating and cooling cycles, transient behavior with cloud cover, and annual energy output can only be evaluated with availability of annual performance data.

Photoactive Materials

555

Solar energy

Concentrated solar thermal energy

Photovoltaic

Biophotolysis

Mechanical to electrical energy H2O

Solar thermolysis

Fossil fuels (natural gas, oil, coal)

Solar thermochemical cycle Solar gasification

Solar cracking

Electrolysis

Photoelectrolysis

Solar reforming

CO2/C sequestration

Solar hydrogen Fig. 21 Major solar hydrogen production pathways.

After this is done and sufficient confidence in the technology is developed, then the process can be taken to the next step of commercialization. Open ended, single reaction processes like upgrading of carbon feed and production of industrial commodities are less complicated as compared to thermochemical cycles and may be given priority for commercialization. Efforts should be directed toward development of processes that involve production of two commodities, viz. zinc and syngas, hydrogen and carbon black, etc. The prospect of fuel production in thermochemical cycles will to a large extent depend on their competitiveness with solar electrochemical approaches. Sustainable methods of solar hydrogen production have been discussed and comparatively assessed by Joshi et al. [30]. Their study has shown that PV-based hydrogen production has a low exergy efficiency due to low PV efficiencies. Wagar et al. [31] have proposed a method for reforming fuels to hydrogen using solar energy at distributed locations and conducted thermodynamic analysis of their proposed system. Ngoh and Njomo [32] have made an inventory of solar hydrogen production methods by conducting technological, environmental, and economic assessments of various methods including electrolysis at different temperatures, gasification, reforming, and decomposition. In this section, solar hydrogen production options are explained briefly. In the next section, a detailed review of photocatalytic water splitting methods is presented.

2.17.11

Photoactive Materials in Water Splitting Reactions

Photocatalytic water splitting has been attracting significant interest in the recent literature since it uses two of the most abundant, clean, renewable, and natural energy sources available to us. Photocatalytic hydrogen production, therefore, has been named as a potential solution to address excessive utilization, limited reserves, and negative environmental impact issues related to fossil fuels. Prospective benefits of photocatalytic water splitting into hydrogen and oxygen can be listed as economic and environmental advantages since the process is using solar energy and produces hydrogen in a clean way with no GHG emissions. In order to use solar irradiation directly for photocatalytic water splitting, a photosensitizer capable of absorbing solar radiation must be dissolved in solution, because water itself is transparent to the visible spectrum. Photonic-based hydrogen production systems are categorized in Fig. 22. In photocatalytic hydrogen production, photons with energies greater than the photocatalyst band gaps are critically required to generate electron and hole pairs to split water. The photons with less energy than the band gap of the photocatalyst cannot produce these electron and hole pairs, and hence, they are not useful in photocatalytic hydrogen production. Because of their band gap and the requirements of water splitting, most catalysts can only utilize photons in high energy portions of the solar spectrum (only UV and in some cases upper part of visible spectrum). Most of the existing photocatalysts in the literature can only use about 4% of the total solar irradiation arriving to Earth’s atmosphere.

556

Photoactive Materials

Photonic hydrogen production systems

Heterogeneous photocatalysis systems

Photoelectrochemical cell

Photoanode

Photocathode

Hybrid photocatalysis systems

Dye synthesized tandem cell

Homogeneous photocatalysis systems

Multicomponent systems

Supramolecular photocatalysis devices

Single electron

Both

Multi electron

Fig. 22 Classification of photocatalytic water splitting methods.

Reactor

Light source

Photocatalyst

Photocatalytic hydrogen production

Electron donor

Electrolyte

Fig. 23 Components affecting photocatalytic hydrogen production performance.

In order to take advantage of the solar spectrum more efficiently, it is essential to be able to utilize the energy stored in the visible light portion. For that reason, one of the major challenges of photocatalytic water splitting research focuses on discovery of cheap, active, abundant, efficient, and stable photocatalysts that are capable of taking advantage of both the visible and UV portions of the solar spectrum. Jang et al. [34] have claimed that the requirements for photocatalytic materials include stability and durability in aqueous electrolyte solutions, competitive cost, better crystallinity with little defects, and excellent conductivity. Crystallinity and particle size, which is determined by the synthesis method and conditions, have a significant impact on the photocatalytic water splitting performance of a photocatalyst. During the past few decades, there has been extensive research on development and synthesis of novel photocatalysts that are capable of splitting water into hydrogen and oxygen by using both UV and visible light spectrum of solar energy. Large scale, clean, relatively low cost, and efficient hydrogen production can be achieved via direct photocatalytic water splitting. Photocatalytic water splitting has the simplicity of using particles in solution with incident sunlight to produce H2 and O2 from water. It can provide a clean, renewable source of hydrogen without producing greenhouse gases or having adverse effects on the atmosphere. Some of the critical factors affecting the performance of a photocatalytic hydrogen production process are illustrated in Fig. 23. In past studies, there have been numerous photocatalysts that were tested under UV radiance and some of these

Photoactive Materials

557

Recyclability Low cost

Stability

Appropriate band gap

Suitable for large scale

Photocatalyst Abundancy

Long lifetime

Corrosion resistance

Suitable CB + VB positions

Efficiency

Fig. 24 Photocatalyst requirements for efficient hydrogen production.

materials have high quantum efficiencies. However, these existing photocatalysts are often inefficient in the process of water splitting. The factors needed in an efficient process include (1) a suitable band gap to support a maximum possible amount of solar spectrum utilization, (2) appropriate conduction and VB placement to drive oxidization and reduction reactions of overall water splitting, (3) stability in the redox environment, (4) low cost of production and operation, (5) recyclability, (6) abundance, (7) corrosion resistance, and (8) suitability for large scale production. Photocatalyst requirements for efficient hydrogen production are listed in Fig. 24. Effective photocatalysts commonly contain principle cation components and empty d or sp orbitals to establish the bottom of their CBs. The principal cation components can either be characteristic metal cations with d10 electronic configuration (such as Ga3 þ , Ge4 þ , In3 þ , Sb5 þ , Sn4 þ , etc.) or transition metal cations with d0 electronic configuration (such as Mo6 þ , Nb5 þ , Ta5 þ , Ti4 þ , W6 þ , Zr4 þ , etc.). In addition, (oxy)nitrides containing d0 transition metal cations, for example, Ta3N5, TaON, and LaTiO2N, are investigated as candidates of efficient photocatalytic hydrogen production. In overall photocatalytic water splitting; solar energy is used to convert water into hydrogen and oxygen, which is a promising candidate to support the hydrogen economy: 1 H2 O- O2 þ H2 ; 2

DG ¼ þ 237 kJ=mol;

lmin ¼ 1100 nm

ð13Þ

A literature review shows that nanomaterials have better photocatalytic water splitting and hydrogen production performance. Cui et al. [33] have compared the photocatalysts with larger particle sizes. However, photocatalytic performance essentially depends on the characteristics of these nanomaterials. Not all nanosized photocatalysts yield maximum hydrogen production rates, yet they still have higher hydrogen production compared to their corresponding micromaterials under similar experimental conditions. Various sizes of CdS crystals between 1 and 5 nm have been experimentally investigated by Deshpande and Gupta. [34]. Their results have showed that the greatest hydrogen production rates (within the given particle size interval) can be obtained by using 2.5 nm CdS particles. Particles of 1.5 nm in diameter have demonstrated poor crystallinity and had a substantial Q size influence on their absorption edge. Deshpande and Gupta [34] have delivered promising indication of the structural sensitivity of photocatalytic water splitting over supported SC nanocrystallites, in which the particle size and interfacial microstructural defects play a vital role. Yao et al. [37] have conducted some experimental investigations of various types of photoreactors and reported that for costeffective design and suitable large scale implementation, passive mixing has a potential to offer a potentially viable and cheap reactor. The composite photocatalysts containing CdS nanowires of high crystallinity decorated with TiO2 nanoparticles have successfully been manufactured by using solvothermal and SG methods. Such materials have been shown to be highly active for hydrogen production from water containing sulfide and sulfite ions as hole scavengers under visible light illumination. A CdS/ TiO2 ratio of 0.2 has shown the highest hydrogen production yield and rates. This configuration has proved effective charge separation as a result of fast diffusion of photoelectrons generated from CdS toward TiO2. Nanostructured ZnO has exceptional physical and chemical properties, which makes it particularly significant in terms of photocatalytic water splitting. ZnO is stable and cheap compared to other binary nanosized metal oxides, therefore, it is a

558

Photoactive Materials

promising photocatalyst option in the literature. Du et al. [38] have studied particle morphology and surface properties of ZnObased photocatalysts and their impact on surface reactivity and photocatalytic water splitting activity of ZnO. Kowsari [39] has reviewed nanocrystalline photocatalysts such as ZnO, CuO, and SrCO3 and discovered the substantial influence of morphology on photocatalytic activity of a photocatalyst. Liu and Syu [40] have investigated some potential ways to advance the optical response of TiO2-based catalysts by doping them with C or N, and S. Mesoporous TiO2 nanostructures can possibly reduce or eliminate the Pt loading on TiO2 for photocatalytic hydrogen production. Kandiel et al. [41] have presented that hydrogen production by using 0.2 wt% Pt/TiO2 calcined at 4501C is three times higher than that evolved on Pt/TiO2–P25 and 12 times higher than that evolved on Pt/TiO2 calcined at 3501C. Ismail et al. [42] have investigated metal cocatalysts (Ru, Ir, Pd, Pt, Os, Re, Co) and their impacts on photocatalytic water splitting performance of a catalyst. There is also ongoing research focusing on various metal oxides. For example, Yan et al. [43] have synthesized K4Nb6O17 as a photocatalyst. Their sample has been the first niobate photocatalyst model with high photocatalytic activity and stability during hydrogen production without the requirement of codoping or another material (i.e., noble metals). Photocatalytic water splitting activities of suspended KCa2Nb3O10 nanoscale and bulk particles have been investigated and comparatively assessed by Sabio et al. [44]. Vaneski et al. [45] have investigated the photocatalytic hydrogen production performance of colloidal noble metal doped photocatalyst nanocrystals. Berr et al. [46] have shown that size, shape, and composition of photocatalyst nanocrystals can be controlled by taking advantage of colloidal chemistry. With colloidal chemistry, it is possible to simultaneously control not only the bulk properties, but also the optical and electronic properties of photocatalysts. For that reason, they provide a better understanding of photocatalysts for hydrogen generation, which is then used as a storage medium for solar energy. Despite the fact that there is noteworthy development in the field of photocatalytic water splitting and synthesis of novel photocatalysts, photocatalysis efficiency (especially under visible light) still needs to be improved in order to make photocatalytic hydrogen production meet today’s energy demands. In addition to efficiency related issues, photocatalysts should be stable, cheap, recyclable, and environmentally benign. The ultimate goal of photocatalytic water splitting research can be summarized as “extensive research to design, develop, synthesize, and test photocatalysts that are abundant, cheap, efficient, reusable, environmentally friendly, stable, simple, and safe.” Band gap configuration and width of a photocatalyst as a direct impact on its efficiency since it controls (1) photon absorption, (2) photoexcitation, (3) electron and hole pair generation, (4) charge carrier movement, and (5) redox capacities of excited electrons and holes. Thus, band gap engineering is an essential aspect of novel photocatalyst design and synthesis. Photocatalysts with direct and narrow band gaps generally have higher optical absorbance properties, for that reason, they are appropriate for effective low energy photon collecting. Nevertheless, photoexcited electron and hole recombination possibility is quite high with these photocatalysts. Also, their band edge positions are generally not compatible with electrochemical potential requirements of the water splitting reaction. TiO2 and WO3 are indirect band gap photocatalysts reported in the literature; what is interesting about them is their absorption characteristics are quite similar to direct band gap photocatalysts. Existing photocatalysts are doped with different element in order to modify their band gap width and band edge positions. Solid solution is another strategy used in band gap engineering. Tong et al. [47] have found that interparticle electronic coupling of SC nanocrystals can be an efficient band gap engineering strategy for narrowing and reconstruction of existing photocatalysts’ band gaps. In the literature, there is extensive effort to increase the efficiency and quantum yield of photocatalytic water splitting. The mail goal is to harness solar energy and use incident photons more efficiently. Band gap engineering is a heavily studied solution to address efficiency issues in photocatalytic systems. In addition to band gap engineering, light sensitization enriched by adding quantum dots, the plasmon–exciton coupling between anchored noble metal nanoparticle cocatalysts and the host SC, and the photon coupling in photonic SCSC crystals are studied in the literature. Surface morphology, crystal structure, and particle size are other key issues affecting water splitting activity of a photocatalyst under solar irradiation. These properties impact surface energy and chemisorption properties, which determine the characteristics of redox reactions on the photocatalyst surface (such as selectivity, rate, overpotential, etc.), electron transfer at the interface, and the resistance toward photocorrosion. Greater surface energies are usually reported to cause better photocatalytic activities. Recent research is also focusing on SC crystals with morphologies providing large percentages of highly reactive facets. Thermodynamic behavior of photocatalysts is generally accepted as a significant challenge when synthesizing novel photoactive materials. Some of the examples of novel photocatalysts developed in the literature are nanostructured materials with surfaces comprising nearly 100% high energy facets with the examples including ultrathin sheets and highly symmetric polyhedral particles.

2.17.12

Recycling Photoactive Materials

One of the most important aspects of photocatalysis is the recycling of chemicals, including catalysts as well. In order to minimize any potential waste, it is preferred to design photocatalysts that have about the same photoactivity during each cycle. In cases where photocatalysts lose their activity with each run, these materials turn into waste at some point in their life cycle. Also, it is important to design photocatalysts that are easier to separate and recycle in order not to lose any material in the waste stream. Most of the photocatalysts have unique properties and specific advantages and disadvantages. However, the drawbacks of photocatalysts in general include the requirement of a large amount of catalysts, difficulty in catalyst recycling, and problematic

Photoactive Materials

559

agglomeration of small particles into large particles. Although nanosized powder photocatalysts present superior properties in aqueous photocatalytic reactions because of their large surface area, good dispersion, and abundant active sites, powderlike nanoparticles still often suffer from deactivation, agglomeration, and difficulty in settling for catalyst recycling after the initial run. Nonreusable photocatalytic materials will considerably increase the operating cost, which significantly limits their actual applications. The separation of photocatalysts from the aqueous phase is of great significance from an economic standpoint. The operating cost of the photocatalytic reaction mainly originates from the photocatalyst being once employed without recycling. TiO2-based photocatalysts are reported to have many advantages over existing photocatalysts in the literature. As a result, they are heavily studied in the literature. However, visible light activity of TiO2-based photocatalysts is significantly lower compared to their UV irradiation photocatalytic activity. Besides, photostability of these photocatalysts should be investigated and improved. In addition to photocatalytic activity, photostability and recyclability are two very important criteria when selecting an appropriate photocatalyst. In the literature, only a few papers have been concerned with the reusability of the photocatalysts and the incident light dependence of the photocatalytic process. By combining the solvothermal method with a calcination process, Song and Gao [48] have synthesized NiO photocatalysts with hierarchical architectures, and controllable morphologies and sizes. They have compared the photocatalytic performance of NiO rods, solid spheres, hollow tubes, and hollow hierarchical structures and showed that the NiO hierarchical architectures they synthesized have significantly higher photocatalytic activities under UV irradiation. A major activity of all these NiO-based photocatalysts is their ability to be recycled easily under an external magnetic field. Sun et al. [49] have synthesized Sb2S3 nanorods with a simple wet chemical method under refluxing condition. They have concluded that Sb2S3 is stable and more active than TiO2 and CdS in terms of photocatalytic water splitting with less than 3% loss in photocatalytic activity in each run. They have also showed that photocatalytic performance of Sb2S3 under visible light irradiation is better than that of TiO2–P25 under UV irradiation. Furthermore, they have recycled and reused their Sb2S3 photocatalyst five times with only insignificant photoactivity loss. By using coprecipitation reactions at temperatures up to 801C, Zhang et al. [50] have prepared crystalline Ag3VO4 photocatalysts with particle sizes varying between 100 nm and 5 mm. They have further reduced particle sizes to between 50 and 100 nm by adding polyethylene glycol in the reactant mixture. These Ag3VO4 photocatalysts, especially the nanosized samples, have showed significant photoactivity under visible light irradiation. Zhang et al. [50] have reported no significant photocatalytic activity loss for both micro- and nanoscale Ag3VO4 photocatalysts after four complete photocatalytic water splitting cycles. Pardeshi and Patil [51] have confirmed that ZnO can be reused five times, as it undergoes photocorrosion only to a negligible extent. Hence, they have reported ZnO to be more appropriate than TiO2 for photocatalytic water splitting under sunlight. Lei et al. [52] have synthesized TiO2 nanoparticles in polyvinyl alcohol by using both heat treatment and solution casting methods. The major outcome of their study is the minimal activity loss of TiO2 photocatalysts even after 25 cycles. Their photocatalyst has showed great potential to be applicable for multicycle use purposes, and has been reported as having outstanding recyclability performance. This performance has been associated with the strong Ti–O–C bonds formed between TiO2 and polyvinyl alcohol during the heat treatment process causing titania particles to be fixed on the polymer matrix, preventing loss of active particles. Recyclability and multiple use of TiO2 and polyvinyl alcohol hybrid photocatalyst with minimal losses is a great step to commercialization of this photocatalyst. Ce loaded ZnO nanoparticles have been studied and characterized by Kuzhalosai et al. [53]. The authors have concluded Ce loaded ZnO photocatalysts are more efficient than bare ZnO, TiO2–P25, and TiO2 at pH 12, and the catalyst has been found to be reusable. Amoozadeh and Rahmani [54] have prepared nano WO3 supported sulfonic acid to be used as a photocatalyst which has shown fast reaction kinetics and high yields in terms of hydrogen production. Also, the used catalyst was easily separated by the authors and reused for 10 runs without appreciable loss of its catalytic activity. A Bi–Au–ZnO composite photocatalyst prepared by Senthilraja et al. [55] showed a higher photocatalytic activity than those of individual Bi–ZnO, Au–ZnO, and bare ZnO and TiO2–P25 at pH 11. Also, Bi–Au–ZnO heterojunction photocatalyst has been more stable and could be easily recycled several times, opening a new path for potential industrial applications. For photocatalytic water splitting to be commercially applicable both in small and large scales, efficient photocatalyst recycling without the loss of activity and/or material is an important factor to be considered. Consequently, photocatalyst efficiency needs to be preserved with no or minimal losses during each use. As a result, there is a need to provide a cheap, simple, and efficient method to apply and recover photocatalysts. Maintaining low cost, simple, and efficient photocatalyst synthesis and application along with the ability to recover and reuse with lowest possible cost, losses, and environmental impact would greatly reduce operating costs associated with photocatalytic water splitting. In general, microsphere shaped photocatalysts have several advantages over powderlike equivalents in terms of recyclability, including ease of recovery and minimal loss of activity.

2.17.13

Case Study: Comparative Assessment of Photoactive Materials

In this case study, 49 photocatalysts are compared based on their band gap, quantum yield, hydrogen production (both per unit mass and unit surface area of photocatalyst) results reported in the literature. When selecting the photocatalysts, some recent studies with the aforementioned data are considered. The photocatalysts are essentially selected from studies using similar reactor vessels, light sources, filters, and temperatures for comparison purposes. Further information on the selected photocatalytic experimental conditions can be found in Table 13.

560

Photoactive Materials

Table 13 this study Catalyst #

Information on selected photocatalysts evaluated in

Photocatalyst Name

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49

Ag0.03Mn0.40Cd0.60S Ag2O/TiO2 Au–CdS Au–TiO2 Au–TiO2–AC BaZr0.96Ta0.04O3 Bi–NaTaO3 Bi1.5Zn0.99Cu0.01Ta1.5O7 Bi2S3/Pt/ZnO Bi2S3/TiO2 CaTa0.8Zr0.2O2.2N0.8 CdS CdS/Cr/Pd CdS/Ta2O5 CdS/Ti Cd0.1Zn0.9S Cd0.4Zn0.6S Cd0.8Zn0.2S/S15 Cr/N–SrTiO3 (CuAg)0.15In0.3Zn1.4S2 Fe–Ni/TiO2 GaFeO2.98S0.02 g–C3N4–SrTiO3:Rh GO–TiO2 In(OH)yS:Ag–Zn In2O3/Ta2O5 K1.025Sr2Nb2.9875Cr0.0125O10 K2La2Ti3O10 K4Nb6O17/CdS K2Ti4O9 N-In2Ga2ZnO7 PbS/K2Ti4O9 Pd–Gardenia–TiO2 Pt/LaOF Pt–PdS–CdS Rh/Cr2O3/GaZn RuO2/MgFe2O4/Pt Sb2TiO5 SrTiO3:Ni/Ta SrTiO3:Ni/Ta/La TiO2–C-362 TiO2–NiS TiO2–SiO2 TiO2–SnO2 TiO2–ZnO TiO2–ZrO2 ZnIn2S4 ZnO/ZnS Zn0.9Ti0.1S

The photocatalysts in Table 13 are divided into four groups, namely titanium oxide, cadmium sulfide, zinc oxide and sulfide, and other metal oxide photocatalysts. The band gap energy (Eg) and electronic band structure strongly affect a photocatalyst’s activity and water splitting performance. In order to provide an efficient photocatalytic water splitting, the band gap energy of a photocatalyst should be higher than 1.23 eV and lower than 3 eV. This criterion is required to utilize the visible light portion of the solar spectrum and provide the necessary energy to split water. In the literature, effective photocatalysts are reported to show band gaps larger than 2eV. The most competent band gap interval for effective photocatalytic water splitting using visible portion of the solar spectrum is stated to be between 2 and 2.4 eV.

Photoactive Materials

561

In order to provide a quantifiable comparison between existing photocatalysts, various criteria are used in the literature. Band gap is an essential indicator of a photocatalyst’s water splitting capacity; rate of photocatalytic hydrogen production is expressed in different units in the literature such as mmol/h, mL/h, mmol/h-gcat, mmol/h-m2cat , etc. Photocatalytic water splitting activity is also assessed by using quantum yield, which is defined based on overall quantum efficiency expression of homogeneous photocatalytic systems. In this study, overall quantum efficiencies are calculated based on Eq. (14), which is developed by Serpone et al. [56] and Fang et al. [57]. Overall quantum yield ð%Þ ¼

2  Number of evolved H2 molecules  100 Number of incident photons

ð14Þ

Two different hydrogen production rates are taken into consideration while conducting comparative assessments in this study. The first expression is the rate of hydrogen production per mass basis (mmol/h-gcat) that represents hourly micromolar hydrogen production rate per unit mass (gram) of photocatalyst. The second rate criterion is hydrogen production per area basis (mmol/h-m2), meaning hourly hydrogen production rate per unit surface area (m2) of photocatalyst. Most of the production rates in the literature are either reported as amount of hydrogen production per unit time or amount of production rate per unit time and unit mass of photocatalyst. In order to keep consistency for the sake of a reliable comparative assessment, necessary conversions are carried out. In cases where hydrogen production rate is not provided per unit mass photocatalyst, reported photocatalyst amounts in corresponding sources are taken into account. Hydrogen production per unit area of catalyst (mmol/h-m2) is calculated by dividing per mass (mmol/h-gcat) by BET surface area of the photocatalyst (m2/gcat). For overall comparisons, the band gap, quantum yield, and hydrogen production rate data of selected photocatalysts are ranked and normalized. The aim of ranking and normalization is to have a consistent scale of measurement. In this case study, hydrogen generation rates (both per unit mass and per unit area) and quantum yields of the selected photocatalysts are ranked on a 0–3 scale to be able to compare their overall performance based on different criteria. A ranking of “0” is assigned to the poorest band gap, quantum yield, and hydrogen production rate. Conversely, a ranking of “3” is assigned to the most appropriate band gap, highest quantum yield, and highest yields among selected photocatalysts. Quantum yield and hydrogen production rates (either per mass or per area) of selected photocatalysts are ranked based on the following equation: Rank i ¼

datai datamin datamax datamin

ð15Þ

here, the quantum yield or hydrogen production rate (either per mass or per area), datai, of a photocatalyst (i) is ranked as Ranki. The lowest and highest quantum yields or production rates among selected photocatalysts are denoted as datamin and highest datamax, respectively. For instance, the photocatalyst with the lowest hydrogen production rate has a ranking of “0” based on Eq. (15). Conversely, the highest production rate is given “3.” A relatively arbitrary ranking system is used to rank the photocatalysts based on the band gap energy: the range of energy bands is divided in four subdomains (Table 14). Band gap data of selected photocatalysts are presented in Fig. 25. In the literature, the most appropriate band gap for photocatalytic water splitting is reported to be between 1.9 and 2.4 eV. According to this information, among selected photocatalysts, the following has the most suitable band gaps: Au–CdS (presented as 3, with band gap of 2.40 eV), CdS (presented as 12, with band gap of 2.40 eV), CdS/Ta2O5 (presented as 14, with band gap of 2.40 eV), Cd0.4Zn0.6S (presented as 17, with band gap of 2.40 eV), Cd0.8Zn0.2S/S15 (presented as 18, with band gap of 2.23 eV), Cr/N–SrTiO3 (presented as 19, with band gap of 2.39 eV), K2Ti4O9 (presented as 30, with band gap of 2.40 eV), Pd–Gardenia–TiO2 (presented as 33, with band gap of 2.30 eV), Pt–PdS–CdS (presented as 35, with band gap of 2.40 eV), RuO2/MgFe2O4/Pt (presented as 37, with band gap of 2.00 eV), SrTiO3: Ni/Ta/La (presented as 40, with band gap of 2.16 eV), and TiO2–NiS (presented as 42, with band gap of 2.06 eV). Fig. 26 presents the quantum yield data of selected photocatalysts. According to the data taken from recent literature, among selected photocatalysts, the highest quantum yields belong to SrTiO3:Ni/Ta (presented as 40, with quantum yield of 17.40%), Cd0.1Zn0.9S (presented as 16, with quantum yield of 17.23%), and K1.025Sr2Nb2.9875Cr0.0125O10 (presented as 27, with quantum yield of 17.19%). And the photocatalysts with lowest quantum yields are RuO2/MgFe2O4/Pt (presented as 37, with quantum yield of 7.42%), TiO2–SnO (presented as 45, with quantum yield of 7.70%), and Bi1.5Zn0.99Cu0.01Ta1.5O7 (presented as 8, with quantum yield of 7.75%). Hydrogen production rates are presented in Figs. 27 and 28. In Fig. 27, hydrogen production rates of selected photocatalysts are compared in terms of mmol/h-g catalyst, while this rate is mmol/h-m2 catalyst in Fig. 28. By selecting two different rate expressions, this distinguishes photocatalyst performance based on both per mass and per active surface area basis. Table 14

Band gap ranking information

Band gap

Rank

Ego1.3 1.3rEgo1.6 1.6rEgo1.9 1.9rEgo2.4 2.4oEgr2.7 2.7oEgr3 3oEg

0 1 2 3 2 1 0

562

Photoactive Materials

4.0

3.5

3.0

Band gap (eV)

2.5

2.0

1.5

1.0

0.5

0.0 1

3

5

7

9

11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49

Fig. 25 Band gap data comparison of selected photocatalysts.

18 17 16

Quantum yield (%)

15 14 13 12 11 10 9 8 7 1

3

5

7

9

11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49

Fig. 26 Quantum yield data comparison of selected photocatalysts.

Photoactive Materials

8000

7000

H2 production (μmol H2/g catalyst)

6000

5000

4000

3000

2000

1000

0 1

3

5

7

9

11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49

Fig. 27 Hydrogen production rate (per unit mass of photocatalyst) data comparison of selected photocatalysts.

500 450

H2 production (μmol H2/m2 catalyst)

400 350 300 250 200 150 100 50 0 1

3

5

7

9

11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49

Fig. 28 Hydrogen production rate (per unit area of photocatalyst) data comparison of selected photocatalysts.

563

564

Photoactive Materials

Fig. 27 shows that, among selected photocatalysts, Au–TiO2 has the highest hydrogen production rate per mass basis (presented as 4, with 7200 mmol H2/h-g catalyst), followed by TiO2–C-362 (presented as 41, with 7430 mmol H2/h-g catalyst). The lowest production rate belongs to Bi–NaTaO3 (presented as 7, with 0.86 mmol H2/h-g catalyst). From Fig. 28, it can be seen that among selected photocatalysts, Au–CdS has the highest hydrogen production rate per area basis (presented as 3, with 472.26 mmol H2/h-m2 catalyst), followed by Ag2O/TiO2 (presented as 2, with 135.90 mmol H2/h-m2 catalyst). Lowest production rate belongs to Sb2TiO5 (presented as 38, with 0.33 mmol H2/h-m2 catalyst), CdS/Ti (presented as 15, with 0.34 mmol H2/h-m2 catalyst), K2La2Ti3O10 (presented as 28, with 0.44 mmol H2/h-m2 catalyst), and Bi–NaTaO3 (presented as 7, with 0.59 mmol H2/h-m2 catalyst). In this case study, the photocatalytic performances of photocatalysts are comparatively assessed and discussed based on various groups, namely titanium oxide, cadmium sulfide, zinc oxide/sulfide, and other metal oxide-based photocatalysts. In the end, average performances of the selected photocatalyst groups based on the ranking information are discussed in order to highlight opportunities and challenges of each group.

2.17.13.1

Titanium Oxide-Based Photocatalysts

Table 15 presents past literature data of titanium oxide-based photocatalysts among the selected ones listed in Table 13. Titanium oxide-based photocatalysts have numerous advantages in terms of photocatalytic water splitting. Therefore, they are heavily studied in the literature. Among the selected ones from the recent literature, Cr/N–SrTiO3, K2Ti4O9, Pd–Gardenia–TiO2, SrTiO3: Ni/La/Ta, and TiO2–NiS have the most appropriate band gaps for water splitting. SrTiO3:Ni/La/Ta and TiO2–ZnO have the highest and lowest quantum yields respectively. In terms of hydrogen production rate per gram catalyst, Au–TiO2 has the highest and Sb2TiO5 has the lowest rates. Production rate per unit area data shows Ag2O/TiO2 has the highest and Sb2TiO5 has the lowest rates. On average, titanium oxide-based photocatalysts show higher production rates per unit mass compared to other photocatalysts. With further band gap engineering, these photocatalysts, and possibly new combinations of titanium oxide, are expected to perform better in terms of quantum yield and production rates.

2.17.13.2

Cadmium Sulfide-Based Photocatalysts

Among the selected photocatalysts, cadmium sulfide-based ones have various advantages such as more appropriate band gaps to split water photocatalytically, and higher production rates both in terms of unit mass and area. These results are displayed in detail in Table 16. From Table 16, it can be said that most of the photocatalysts have appropriate band gaps (between 1.9 and 2.4 eV) except Ag0.03Mn0.40Cd0.60S, CdS/Cr/Pd, CdS/Ti, Cd0.1Zn0.9S, and K4Nb6O17/CdS. Highest and lowest quantum yields belong to Cd0.1Zn0.9S and CdS/Ti, respectively. Au–CdS has both the highest hydrogen production rate per unit area and mass within this group while CdS has the lowest performance in both rate categories. Table 15

Reported performances of titanium oxide-based photocatalysts in the literature

Catalyst

Ref.

Eg (eV)

Quantum yield (%)

Name Ag2O/TiO2 Au–TiO2 Au–TiO2–AC Bi2S3/TiO2 Cr/N–SrTiO3 Fe–Ni/TiO2 g–C3N4–SrTiO3:Rh GO–TiO2 K2La2Ti3O10 K2Ti4O9 PbS/K2Ti4O9 Pd–Gardenia–TiO2 Sb2TiO5 SrTiO3:Ni/Ta SrTiO3:Ni/Ta/La TiO2–C-362 TiO2–NiS TiO2–SiO2 TiO2–SnO2 TiO2–ZnO TiO2–ZrO2

[79] [25] [22] [80] [11] [20] [28] [21] [6] [7] [30] [31] [19] [8] [10] [16] [27] [23] [81] [24] [9]

2.86 2.76 3.20 3.22 2.39 2.41 3.20 2.58 3.50 2.40 2.60 2.30 1.87 2.61 2.16 2.60 2.06 3.26 2.90 3.06 3.25

8.55 12.06 12.51 13.75 10.53 12.67 15.25 12.18 11.00 14.56 16.39 16.09 14.82 17.40 12.99 9.86 8.23 11.30 16.92 7.70 12.15

Results 2

(mmol/h-gcat)

(mmol/h-m )

3350 7200 2750 106 427 362 2233 3800 20 520 617 120 16 561.2 2305 4430 698 1258 1468 1290 1150

135.90 93.34 31.81 3.15 12.57 3.68 85.13 88.95 0.44 11.29 13.14 2.02 0.33 24.62 58.92 61.24 14.46 25.48 29.78 15.02 8.92

Brunauer Emmett Teller (BET) surface area (m2/g) 24.65 77.14 86.46 33.69 33.97 98.35 26.23 42.72 45.07 46.05 46.94 59.46 48.58 22.79 39.12 72.34 48.27 49.38 49.30 85.91 128.95

Photoactive Materials

Table 16

Reported performances of cadmium sulfide-based photocatalysts in the literature

Catalyst

Eg (eV)

Ref.

Quantum Yield (%)

Name Ag0.03Mn0.40Cd0.60S Au–CdS CdS CdS/Cr/Pd CdS/Ta2O5 CdS/Ti Cd0.1Zn0.9S Cd0.4Zn0.6S Cd0.8Zn0.2S/S15 K4Nb6O17/CdS Pt–PdS–CdS

Table 17

2.59 2.40 2.40 2.66 2.40 2.63 2.78 2.40 2.23 3.10 2.40

[3] [29] [18] [59] [33] [36] [34] [15] [5] [26] [35]

12.02 16.05 9.62 9.05 14.91 8.12 17.23 11.90 16.77 13.66 9.56

Reference

Eg (eV)

Quantum yield (%)

Name Bi1.5Zn0.99Cu0.01Ta1.5O7 Bi2S3/Pt/ZnO (CuAg)0.15In0.3Zn1.4S2 In(OH)yS:Ag–Zn N-In2Ga2ZnO7 ZnIn2S4 ZnO/ZnS Zn0.9Ti0.1S

2

(mmol/h-gcat)

(mmol/h-m )

3400 4000 1200 1815 3750 47 1930 1200 1000 3680 2833

83.27 472.26 48.84 72.98 40.81 0.34 45.22 25.93 23.27 52.24 33.07

Brunauer Emmett Teller (BET) surface area (m2/g) 40.83 8.47 24.57 24.87 91.89 138.41 42.68 46.28 42.98 70.45 85.67

2.60 2.71 1.90 1.65 2.50 2.59 3.40 3.02

[2] [30] [40] [12] [18] [1] [14] [82]

7.75 15.63 12.39 16.11 16.03 11.02 14.42 14.21

Results (mmol/h-gcat)

(mmol/h-m2)

45 23 633 3873 2800 2204.5 388.4 400

68.18 0.84 82.75 59.15 39.19 63.26 11.20 7.96

Brunauer Emmett Teller (BET) surface area (m2/g) 0.66 27.45 7.65 65.48 71.45 34.85 34.67 50.24

Reported performances of other metal oxide-based photocatalysts in the literature

Catalyst

Ref.

Eg (eV)

Quantum yield (%)

Name BaZr0.96Ta0.04O3 Bi–NaTaO3 CaTa0.8Zr0.2O2.2N0.8 GaFeO2.98S0.02 In2O3/Ta2O5 K1.025Sr2Nb2.9875Cr0.0125O10 Pt/LaOF Rh/Cr2O3/GaZn RuO2/MgFe2O4/Pt

2.17.13.3

Results

Reported performances of zinc oxide and sulfide-based photocatalysts in the literature

Catalyst

Table 18

565

[38] [39] [4] [19] [32] [37] [83] [13] [4]

2.76 2.64 2.59 2.65 2.80 3.50 2.62 2.60 2.00

14.26 15.11 12.60 13.49 10.20 17.19 11.32 16.11 7.42

Results (mmol/h-gcat)

(mmol/h-m2)

900 0.86 52 233 533 243 202 1488 55

24.92 0.59 7.14 3.02 12.18 3.22 4.15 30.81 1.12

Brunauer Emmett Teller (BET) surface area (m2/g) 36.12 1.46 7.28 77.12 43.75 75.43 48.63 48.3 49.00

Zinc Oxide and Sulfide-Based Photocatalysts

Zinc oxide and sulfide-based photocatalysts, in general, have higher quantum yields compared to the other selected materials evaluated in this study. These findings are presented in Table 17, which shows none of the catalysts in this group have a band gap falling in between 1.9 and 2.4 eV. In(OH)yS:Ag–Zn has the highest quantum yield and hydrogen production rate per unit mass of photocatalyst. Bi1.5Zn0.99Cu0.01Ta1.5O7 has the lowest quantum yield in this category. Bi2S3/Pt/ZnO has the lowest hydrogen production rates both per unit mass and area of photocatalyst. The highest hydrogen production rate per unit area of photocatalyst belongs to (CuAg)0.15In0.3Zn1.4S2.

2.17.13.4

Other Metal Oxide-Based Photocatalysts

Band gaps, quantum yields, and hydrogen production rates of various metal oxide-based photocatalysts are listed in Table 18.

566

Photoactive Materials

Among selected options in Table 18, only RuO2/MgFe2O4/Pt has the most appropriate band gap for photocatalytic water splitting. In terms of quantum yield, K1.025Sr2Nb2.9875Cr0.0125O10 has the highest and RuO2/MgFe2O4/Pt has the lowest. Rh/Cr2O3/GaZn has both the highest hydrogen production rate per gram and area of catalyst. CaTa0.8Zr0.2O2.2N0.8 has the lowest rate per unit mass while Bi–NaTaO3 has the lowest rate per unit area.

2.17.13.5

Overall Comparison

In order to perform an overall comparison of selected photocatalyst groups, the band gap, quantum yield, and hydrogen production rate results are ranked and normalized on a 0–3 scale. Here 0 indicates poorest performance (lowest yield, rate, and least appropriate band gap) while 3 means closest-to-ideal performance (most appropriate band gap, highest yield, and production rates among selected options). These rankings are presented in Table 19. Averages of normalized rankings are taken among the photocatalysts in each group and the results are presented in Fig. 29. Compared to other groups, cadmium sulfide-based photocatalysts have better performance in almost every category except quantum yield. With proper band gap adjustments/engineering, zinc sulfide/oxide-based materials can potentially increase their performance as their low production rates are generally linked to their band gap limitations. Despite many advantages, titanium oxide-based photocatalysts show a poor performance in overall comparison. This is mainly due to the large number of TiO2-based materials reported in the literature. Further investigation can be conducted on materials synthesized in similar conditions. Similarly, due to the large variety in samples, other metal oxides also have poor overall performance. Only one of the selected ones in this group has the appropriate band gap, and the reported ones in the literature have lower production rates compared to titanium oxides, cadmium sulfides, and zinc oxide/sulfides. This issue could also be potentially addressed by band gap modification.

2.17.13.6

Case Study Conclusions

In this case study, current status of photocatalysts available in the literature are comparatively assessed for hydrogen generation through visible light-based water splitting. The comparative assessment criteria considered in the study are (1) band gap, (2) quantum yield, (3) hydrogen production rate both per unit mass of catalyst, and (4) hydrogen production rate both per unit surface area of catalyst. The titanium oxide, cadmium sulfide, zinc oxide/sulfide, and other metal oxide-based photocatalyst groups are the photocatalyst groups that are examined. It has been emphasized that each photocatalytic group has different advantages and challenges. In order to be able to develop a system with high photocatalytic performance, low cost, and low environmental and health impact, potential improvement opportunities for each photocatalyst group are should be developed. The 49 photocatalysts and 4 photocatalyst groups are ranked and compared in this study, and the following findings are obtained:

• • • • •

Among selected photocatalysts, Au–CdS (2.40 eV), CdS (2.40 eV), CdS/Ta2O5 (2.40 eV), Cd0.4Zn0.6S (2.40 eV), Cd0.8Zn0.2S/S15 (2.23 eV), Cr/N–SrTiO3 (2.39 eV), K2Ti4O9 (2.40 eV), Pd–Gardenia–TiO2 (2.30 eV), Pt–PdS–CdS (2.40 eV), RuO2/MgFe2O4/Pt (2.00 eV), SrTiO3:Ni/Ta/La (2.16 eV), and TiO2–NiS (2.06 eV) have appropriate band gaps for photocatalytic water splitting. SrTiO3:Ni/Ta (17.40%), Cd0.1Zn0.9S (17.23%), and K1.025Sr2Nb2.9875Cr0.0125O10 (17.19%) have highest quantum yields while RuO2/MgFe2O4/Pt (7.42%), TiO2–SnO (7.70%), and Bi1.5Zn0.99Cu0.01Ta1.5O7 (7.75%) have the lowest quantum yields. Au–TiO2 has the highest hydrogen production rate per mass basis (7200 mmol H2/h-g catalyst), followed by TiO2–C-362 (7430 mmol H2/h-g catalyst) and Bi–NaTaO3 (0.86 mmol H2/h-g catalyst) has the lowest production rate. Au–CdS has the highest hydrogen production rate per area basis (472.26 mmol H2/h-m2 catalyst), followed by Ag2O/TiO2 (135.90 mmol H2/h-m2 catalyst). Sb2TiO5 (0.33 mmol H2/h-m2 catalyst), CdS/Ti (0.34 mmol H2/h-m2 catalyst), K2La2Ti3O10 (0.44 mmol H2/h-m2 catalyst), and Bi–NaTaO3 (0.59 mmol H2/h-m2 catalyst) have the lowest production rates. Photocatalyst group comparison shows that cadmium sulfide-based photocatalysts have the highest average normalized ranking (1.32/3.00). On average, cadmium sulfide-based photocatalysts perform better compared to other selected catalysts in all categories except quantum yield. In the quantum yield category, zinc oxide and sulfides perform better.

2.17.14

Future Directions

New materials with high photocatalytic activity in the visible range are essential for the development of commercially viable technologies. Of particular interests are metal organic framework (MOF) compounds, mesoporous materials (aluminosilicates), and polyoxometalates (POM). Despite being in an early stage of research, these materials already demonstrate comparable or greater efficiency in the degradation of organic compounds than does TiO2. Metal organic frameworks are two dimensional or three dimensional crystalline porous materials consisting of metal ions or clusters coordinated to rigid organic ligands. The photocatalytic activity of MOF compounds is due to their SC-like properties with band gaps between 1.0 and 5.5 eV. Compared to classical photocatalysts such as TiO2, MOFs preserve a variety of pore sizes with a maximum surface area of approximately 6000 m2/g. Such a large surface area is highly desirable, as this increases the probability for adsorption of reactants and carrier trapping on the catalyst surface. Another advantage of MOFs is that their electronic structures can be manipulated by bridging distinct chemical constituents with unique organic linkers to achieve synergistic photocatalytic activities.

Photoactive Materials

Table 19

567

Normalized rankings of selected photocatalysts evaluated in this study

Catalyst

Band gap

Quantum yield (%)

Name Ag0.03Mn0.40Cd0.60S Ag2O/TiO2 Au–CdS Au–TiO2 Au–TiO2–AC BaZr0.96Ta0.04O3 Bi–NaTaO3 Bi1.5Zn0.99Cu0.01Ta1.5O7 Bi2S3/Pt/ZnO Bi2S3/TiO2 CaTa0.8Zr0.2O2.2N0.8 CdS CdS/Cr/Pd CdS/Ta2O5 CdS/Ti Cd0.1Zn0.9S Cd0.4Zn0.6S Cd0.8Zn0.2S/S15 Cr/N–SrTiO3 (CuAg)0.15In0.3Zn1.4S2 Fe–Ni/TiO2 GaFeO2.98S0.02 g-C3N4–SrTiO3:Rh GO–TiO2 In(OH)yS:Ag–Zn In2O3/Ta2O5 K1.025Sr2Nb2.9875Cr0.0125O10 K2La2Ti3O10 K4Nb6O17/CdS K2Ti4O9 N-In2Ga2ZnO7 PbS/K2Ti4O9 Pd–Gardenia-TiO2 Pt/LaOF Pt–PdS–CdS Rh/Cr2O3/GaZn RuO2/MgFe2O4/Pt Sb2TiO5 SrTiO3:Ni/Ta SrTiO3:Ni/Ta/La TiO2–C-362 TiO2–NiS TiO2–SiO2 TiO2–SnO2 TiO2–ZnO TiO2–ZrO2 ZnIn2S4 ZnO/ZnS Zn0.9Ti0.1S

2 1 3 2 0 1 2 2 1 0 2 3 2 3 2 1 3 3 3 3 2 2 0 2 2 1 0 0 0 3 2 2 3 2 3 2 3 2 2 3 2 3 0 1 0 0 2 0 0

1.38 0.34 2.59 1.40 1.53 2.05 2.31 0.10 2.47 1.90 1.56 0.66 0.49 2.25 0.21 2.95 1.35 2.81 0.93 1.49 1.58 1.82 2.35 1.43 2.61 0.84 2.94 1.08 1.87 2.15 2.59 2.70 2.60 1.17 0.64 2.61 0.00 2.22 3.00 1.67 0.73 0.24 1.17 2.85 0.08 1.42 1.08 2.10 2.04

Results

Average

Rate per mass

Rate per area

1.42 1.40 1.67 3.00 1.15 0.37 0.00 0.02 0.01 0.04 0.02 0.50 0.76 1.56 0.02 0.80 0.50 0.42 0.18 0.26 0.15 0.10 0.93 1.58 1.61 0.22 0.10 0.01 1.53 0.22 1.17 0.26 0.05 0.08 1.18 0.62 0.02 0.01 0.23 0.96 1.85 0.29 0.52 0.61 0.54 0.48 0.92 0.16 0.17

0.53 0.86 3.00 0.59 0.20 0.16 0.00 0.43 0.00 0.02 0.04 0.31 0.46 0.26 0.00 0.29 0.16 0.15 0.08 0.52 0.02 0.02 0.54 0.56 0.37 0.08 0.02 0.00 0.33 0.07 0.25 0.08 0.01 0.02 0.21 0.19 0.01 0.00 0.15 0.37 0.39 0.09 0.16 0.19 0.09 0.05 0.40 0.07 0.05

1.33 0.90 2.57 1.75 0.72 0.90 1.08 0.64 0.87 0.49 0.91 1.12 0.93 1.77 0.56 1.26 1.25 1.59 1.05 1.32 0.94 0.98 0.96 1.39 1.65 0.53 0.76 0.27 0.93 1.36 1.50 1.26 1.42 0.82 1.26 1.36 0.76 1.06 1.35 1.50 1.24 0.91 0.46 1.16 0.18 0.49 1.10 0.58 0.56

In addition to MOFs, another structural motif known as POMs represents an emerging class of materials that appears promising for photocatalytic applications. POMs are a group of molecular clusters that consist of three or more transition metal oxyanions (usually group 5 or group 6 transition metals) linked together by shared oxygen atoms to form a large, yet enclosed, threedimensional framework. POMs have well-defined structures with a typical size of a few nanometers. They are very stable and can be easily deposited onto organic substrates for catalytic reaction via the formation of a preassociated complex. POMs represent an attractive class of photoactive materials due to their highly tunable band gap, which can be tailored by adjusting the size or composition of the metal oxide particulates. In general, decreasing the size of the particulates will increase the band gap, and

Photoactive Materials

568

Titanium oxide Band gap 3.0

Cadmium sulfide

2.5

Zinc oxide/sulfide

2.0 Other metal oxides

1.5 1.0

Ideal

0.5 Molar rate per area

Quantum yield (%)

0.0

Molar rate per mass Fig. 29 Average normalized rankings of selected photocatalyst groups.

New surface ligands

In situ generation

New IR absorbers

Photon up and down conversion

Flexible substrate transparent electrodes

Painting and spray fabrication

Novel material synthesis ↔ Integrated system design and development

Appropriateness for portable and stationary applications

Scalability

Integration with energy storage systems

Future directions of photoactive materials

Fig. 30 Future directions for photoactive materials.

allow for precise modulation of electronic structure to achieve tunable photocatalytic properties across the UV to near IR spectrum. Two other good candidates filling the above requirement are graphitic carbon nitride sheets and graphene nanoribbons. They both have a two-dimensional rigid structure that is slightly different from the traditional conjugated polymers. They both exhibit significant stability under acidic, basic, and high temperature conditions, and their band structures can be tuned by adjusting their size and edge structures. Although studies on both are still in early stages due to the fact that they are still relatively new materials, this field is growing rapidly because of the interest generated by these favorable properties. Polyoxometalate carbon nanotubes show remarkably improved efficiency and operational stability. This result opens new avenues for innovative materials designed for efficiency and cost optimization of photocatalysis. Advanced photoactive material designs with improved efficiency include third generation PVs, PEC systems, and novel photoactive materials with improved efficiencies such as thermophotovoltaics, tandem cells, multijunction cells, hot carrier cells, up/down converters, silicon nanostructures, and polymer and dye sensitized cells. Future directions include fourth generation PVs, multiquantum well structures, quantum dots, nanowire organic cells, thermophotonic conversion, polycrystalline thin films, molecular organic devices, and transparent PVs. Future directions for photoactive materials are presented in Fig. 30.

2.17.15

Closing Remarks

In closing, the requirements for an effective photocatalyst include (1) stability, (2) low cost, (3) abundance, (4) efficiency, (6) appropriate band gap, (7) corrosion resistance, and (8) large-scale production ability. Some technical challenges related to photocatalytic hydrogen production include stability, efficiency, and energetics. Most stable materials have either too large of a

Photoactive Materials

569

band gap for efficient light absorption (B3 eV), or poor SC characteristics. In terms of efficiency, the band gap must be less than 2.2 eV. However, most photoactive materials with appropriate band gaps are unstable in electrolytic environments (and in water). Also, the energetic position of conduction (CB) and valence (VB) bands may prevent photocatalytic water splitting. For spontaneous water splitting, half reactions must be between VB and CB edges. The aim of more efficient photocatalytic water splitting is to address these challenges through band gap engineering. With future developments in material science and the introduction of novel photocatalysts, it is expected that photocatalytic systems will increasingly be stable with appropriate band gaps and CB and VB locations. When these challenges are addressed, a photocatalytic system will become capable of producing hydrogen at promising rates with higher efficiencies. The quantum efficiency for wide bandgap oxide catalysts is worse in the UV range than in the visible range for catalysts with a low bandgap. Thus, it is necessary to narrow the bandgap of photocatalysts to harvest visible light of longer wavelengths and enhance photogenerated charge separation in photocatalysis. High-efficiency and cost-effective H2S splitting systems based on these photocatalysts should also be constructed. Photocatalysts free of noble metals are highly preferable considering that they can be readily used in a more economical way. It is anticipated that this demonstration of solar photocatalytic H2 production will attract attention for further studies in a promising direction. From the literature reviewed so far, it is evident that liquid-phase recovery yields greater H2 production than gas-phase recovery. The current lack of industrial applications of this technology can mainly be attributed to two reasons: the low photocatalytic efficiency due to large-size photocatalysts and the resulting lack of agreement on how to quantify this efficiency, in particular for nanosized photocatalysts and reactor configurations. Therefore, for nanosized catalysts, reactor design and cost reductions for large scale applications have to be given special priority. The other challenge is successful scale-up of laboratory scale photocatalysis to an industrially relevant scale. These two issues have been the object of our research and will also be the focus of our study in the future.

References [1] Dincer I, Zamfirescu C. Sustainable energy systems and applications. New York, NY: Springer Science and Business Media; 2011. [2] Fujishima A, Honda K. Electrochemical photolysis of water at a semiconductor electrode. Nature 1972;238:37–8. [3] Zamfirescu C, Naterer GF, Dincer I. Solar light-based hydrogen production systems. In: Anwar S, editor. Encyclopedia of Energy Engineering and Technology. New York, NY: Taylor and Francis; 2012. [4] Acar C, Dincer I, Zamfirescu C. A review on selected heterogeneous photocatalysts for hydrogen production. Int J Energy Res 2014;38(15):1903–20. [5] Acar C, Dincer I. Analysis and assessment of a continuous-type hybrid photoelectrochemical system for hydrogen production. Int J Hydrog Energy 2014;39(28): 15362–15372. [6] Sayama K. Visible-light-responsive photocatalysts and photoelectrodes using WO3 semiconductors for degradation of organics and water splitting. In: Yamashita H, Li H, editors. Nanostructured Photocatalysts. Switzerland: Springer International Publishing; 2016. p. 429–42. [7] Yamashita H, Ichihashi Y. Application of metal ion implantation for the design of visible light-responsive titanium oxide photocatalysts. Nanostructured Photocatalysts 2016;Switzerland: Springer International Publishing; 2016. p. 173–86. [8] Iwashina K, Iwase A, Ng YH, Amal R, Kudo A. Z-schematic water splitting into H2 and O2 using metal sulfide as a hydrogen-evolving photocatalyst and reduced graphene oxide as a solid-state electron mediator. J Am Chem Soc 2015;137(2):604–7. [9] Fujita SI, Kawamori H, Honda D, Yoshida H, Arai M. Photocatalytic hydrogen production from aqueous glycerol solution using NiO/TiO2 catalysts: effects of preparation and reaction conditions. Appl Catal B: Environ 2016;181:818–24. [10] Kato T, Hakari Y, Ikeda S, et al. Utilization of metal sulfide material of (CuGa)1–xZn2xS2 solid solution with visible light response in photocatalytic and photoelectrochemical solar water splitting systems. J Phys Chem Lett 2015;6(6):1042–7. [11] Iqbal M, Purkait TK, Goss GG, et al. Application of engineered Si nanoparticles in light-induced advanced oxidation remediation of a water-borne model contaminant. ACS Nano 2016;10(5):5405–12. [12] Scaife DE. Oxide semiconductors in photoelectrochemical conversion of solar energy. Sol Energy 1980;25(1):41–54. [13] Abe R, Sayama K, Arakawa H. Dye-sensitized photocatalysts for efficient hydrogen production from aqueous I solution under visible light irradiation. J Photochem Photobiol 2004;166(1):115–22. [14] Abe R. Recent progress on photocatalytic and photoelectrochemical water splitting under visible light irradiation. J Photochem Photobiol C: Photochem Rev 2010;11 (4):179–209. [15] Kudo A, Sayama K, Tanaka A, et al. Nickel-loaded K4Nb6O17 photocatalyst in the decomposition of H2O into H2 and O2: structure and reaction mechanism. J Catal 1989;120(2):337–52. [16] Li Y, Lu G, Li S. Photocatalytic production of hydrogen in single component and mixture systems of electron donors and monitoring adsorption of donors by in situ infrared spectroscopy. Chemosphere 2003;52(5):843–50. [17] Acar C, Dincer I, Naterer GF. Review of photocatalytic water‐splitting methods for sustainable hydrogen production. Int J Energy Res 2016;40(11):1449–73. [18] Acar C, Dincer I. A review and evaluation of photoelectrode coating materials and methods for photoelectrochemical hydrogen production. Int J Hydrog Energy 2016;41 (19):7950–9. [19] Dincer I. Green methods for hydrogen production. Int J Hydrog Energy 2012;37(2):1954–71. [20] Rabbani M, Dincer I, Naterer GF. Efficiency assessment of a photo electrochemical chloralkali process for hydrogen and sodium hydroxide production. Int J Hydrog Energy 2014;39(5):1941–56. [21] Cetinkaya E, Dincer I, Naterer GF. Life cycle assessment of various hydrogen production methods. Int J Hydrog Energy 2012;37(3):2071–80. [22] Hacatoglu K, Rosen MA, Dincer I. Comparative life cycle assessment of hydrogen and other selected fuels. Int J Hydrog Energy 2012;37(13):9933–40. [23] Ozbilen A, Dincer I, Rosen MA. Comparative environmental impact and efficiency assessment of selected hydrogen production methods. Environ Impact Assess Rev 2013;42:1–9. [24] Joshi AS, Dincer I, Reddy BV. Exergetic assessment of solar hydrogen production methods. Int J Hydrog Energy 2010;35(10):4901–8. [25] Pitz-Paal R, Botero NB, Steinfeld A. Heliostat field layout optimization for high-temperature solar thermochemical processing. Sol Energy 2011;85(2):334–43. [26] Meier A, Steinfeld A. Solar energy in thermochemical processing. Encycl Energy 2012;12(1):9588–619. [27] Bullard CW, Penner PS, Pilati DA. Net energy analysis: handbook for combining process and input-output analysis. Resour Energy 1978;1(3):267–313.

570

Photoactive Materials

[28] Krishnamurthy P, Banerjee R. Energy analysis of solar thermal concentrating systems for power plants. In: International conference on future electrical power and energy systems, Lecture Notes in Information Technology 9; 2012. p. 509–514. [29] Yadav D, Banerjee R. A review of solar thermochemical processes. Renew Sustain Energy Rev 2016;54:497–532. [30] Joshi AS, Dincer I, Reddy BV. Solar hydrogen production: a comparative performance assessment. Int J Hydrog Energy 2011;36(17):11246–57. [31] Wagar WR, Zamfirescu C, Dincer I. Thermodynamic analysis of solar energy use for reforming fuels to hydrogen. Int J Hydrog Energy 2011;36(12):7002–11. [32] Ngoh SK, Njomo D. An overview of hydrogen gas production from solar energy. Renew Sustain Energy Rev 2012;16(9):6782–92. [33] Dincer I, Zamfirescu C. Sustainable hydrogen production options and the role of IAHE. Int J Hydrog Energy 2012;37(21):16266–86. [34] Jang JS, Kim HG, Lee JS. Heterojunction semiconductors: a strategy to develop efficient photocatalytic materials for visible light water splitting. Catal Today 2012;185(1):270–7. [35] Cui W, Liu L, Ma S, Liang Y, Zhang Z. CdS-sensitized K2La2Ti3O10 composite: a new photocatalyst for hydrogen evolution under visible light irradiation. Catal Today 2013;207:44–9. [36] Deshpande A, Gupta NM. Critical role of particle size and interfacial properties in the visible light induced splitting of water over the nanocrystallites of supported cadmium sulphide. Int J Hydrog Energy 2010;35(8):3287–96. [37] Yao W, Huang C, Muradov N, Ali T. A novel Pd–Cr2O3/CdS photocatalyst for solar hydrogen production using a regenerable sacrificial donor. Int J Hydrog Energy 2011;36(8):4710–5. [38] Du S, Liu H, Chen Y. Large-scale preparation of porous ultrathin Ga-doped ZnO nanoneedles from 3D basic zinc carbonate superstructures. Nanotechnology 2009;20 (8):085611. [39] Kowsari E. Sonochemically assisted synthesis and application of hollow spheres, hollow prism, and coralline-like ZnO nanophotocatalyst. J Nanoparticle Res 2011;13 (8):3363–76. [40] Liu SH, Syu HR. One-step fabrication of N-doped mesoporous TiO2 nanoparticles by self-assembly for photocatalytic water splitting under visible light. Appl Energy 2012;100:148–54. [41] Kandiel TA, Ismail AA, Bahnemann DW. Mesoporous TiO2 nanostructures: a route to minimize Pt loading on titania photocatalysts for hydrogen production. Phys Chem Chem Phys 2011;13(45):20155–61. [42] Ismail AA, Robben L, Bahnemann DW. Study of the Efficiency of UV and visible‐light photocatalytic oxidation of methanol on mesoporous RuO2–TiO2 nanocomposites. ChemPhysChem 2011;12(5):982–91. [43] Yan L, Zhang J, Zhou X, et al. Crystalline phase-dependent photocatalytic water splitting for hydrogen generation on KNbO3 submicro crystals. Int J Hydrog Energy 2013;38(9):3554–61. [44] Sabio EM, Chamousis RL, Browning ND, Osterloh FE. Photocatalytic water splitting with suspended calcium niobium oxides: why nanoscale is better than bulk–a kinetic analysis. J Phys Chem C 2012;116(4):3161–70. [45] Vaneski A, Susha AS, Rodríguez‐Fernández J, et al. Hybrid colloidal heterostructures of anisotropic semiconductor nanocrystals decorated with noble metals: synthesis and function. Adv Funct Mater 2011;21(9):1547–56. [46] Berr MJ, Vaneski A, Mauser C, et al. delayed photoelectron transfer in Pt‐decorated CdS nanorods under hydrogen generation conditions. Small 2012;8(2):291–7. [47] Tong H, Umezawa N, Ye J, Ohno T. Electronic coupling assembly of semiconductor nanocrystals: self-narrowed band gap to promise solar energy utilization. Energy Environ Sci 2011;4(5):1684–9. [48] Song X, Gao L. Facile synthesis and hierarchical assembly of hollow nickel oxide architectures bearing enhanced photocatalytic properties. J Phys Chem C 2008;112 (39):15299–305. [49] Sun M, Li D, Li W, et al. New photocatalyst, Sb2S3, for degradation of methyl orange under visible-light irradiation. J Phys Chem C 2008;112(46):18076–81. [50] Zhang Y, Ding H, Liu F, et al. Synthesis of nanosized Ag3VO4 particles and their photocatalytic activity for degradation of rhodamine B under visible light. Chin J Catal 2008;29(8):783–7. [51] Pardeshi SK, Patil AB. A simple route for photocatalytic degradation of phenol in aqueous zinc oxide suspension using solar energy. Sol Energy 2008;82(8):700–5. [52] Lei P, Wang F, Gao X, et al. Immobilization of TiO2 nanoparticles in polymeric substrates by chemical bonding for multi-cycle photodegradation of organic pollutants. J Hazard Mater 2012;227:185–94. [53] Kuzhalosai V, Subash B, Shanthi M. A novel sunshine active cerium loaded zinc oxide photocatalyst for the effective degradation of AR 27 dye. Mater Sci Semicond Process 2014;27:924–33. [54] Amoozadeh A, Rahmani S. Nano-WO3-supported sulfonic acid: new, efficient and high reusable heterogeneous nano catalyst. J Mol Catal A: Chem 2015;396:96–107. [55] Senthilraja A, Subash B, Dhatshanamurthi P, Swaminathan M, Shanthi M. Photocatalytic detoxification of Acid Red 18 by modified ZnO catalyst under sunlight irradiation. Spectrochim Acta Part A: Mol Biomol Spectrosc 2015;138:31–7. [56] Serpone N, Sauve G, Koch R, et al. Standardization protocol of process efficiencies and activation parameters in heterogeneous photocatalysis: relative photonic efficiencies ζr. J Photochem Photobiol 1996;94(2-3):191–203. [57] Fang J, Cao SW, Wang Z, et al. Mesoporous plasmonic Au–TiO2 nanocomposites for efficient visible-light-driven photocatalytic water reduction. Int J Hydrog Energy 2012;37(23):17853–61. [58] Desai UV, Xu C, Wu J, Gao D. Hybrid TiO2–SnO2 nanotube arrays for dye-sensitized solar cells. J Phys Chem C 2013;117(7):3232–9. [59] Yu ZB, Xie YP, Liu G, et al. Self-assembled CdS/Au/ZnO heterostructure induced by surface polar charges for efficient photocatalytic hydrogen evolution. J Mater Chem A 2013;1(8):2773–6. [60] Han C, Wang Y, Lei Y, et al. In situ synthesis of graphitic-C3N4 nanosheet hybridized N-doped TiO2 nanofibers for efficient photocatalytic H2 production and degradation. Nano Res 2015;8(4):1199–209. [61] Eley C, Li T, Liao F, et al. Nanojunction‐mediated photocatalytic enhancement in heterostructured CdS/ZnO, CdSe/ZnO, and CdTe/ZnO nanocrystals. Angew Chem Int Ed 2014;53(30):7838–42. [62] Reyes-Gil KR, Stephens ZD, Stavila V, Robinson DB. Composite WO3/TiO2 nanostructures for high electrochromic activity. ACS Appl Mater Interfaces 2015;7(4):2202–13. [63] Reddy JK, Lalitha K, Reddy PVL, et al. Fe/TiO2: a visible light active photocatalyst for the continuous production of hydrogen from water splitting under solar irradiation. Catal Lett 2014;144(2):340–6. [64] Kumar V, Rajaram P, Goswami YC. Sol gel synthesis of SnO2/CdSe nanocomposites and their optical structural and morphological characterizations. Optik-Int J Light Electron Opt 2016;127(5):2490–4. [65] Xie J, Yao X, Madden IP, et al. Selective deposition of Ru nanoparticles on TiSi2 nanonet and its utilization for Li2O2 formation and decomposition. J Am Chem Soc 2014;136(25):8903–6. [66] Liu L, Yang W, Sun W, Li Q, Shang JK. Creation of Cu2O@ TiO2 composite photocatalysts with p–n heterojunctions formed on exposed Cu2O facets, their energy band alignment study, and their enhanced photocatalytic activity under illumination with visible light. ACS Appl Mater Interfaces 2015;7(3):1465–76. [67] Zhou R, Zhang Q, Uchaker E, et al. Mesoporous TiO2 beads for high efficiency CdS/CdSe quantum dot co-sensitized solar cells. J Mater Chem A 2014;2(8):2517–25. [68] Chen CH, Shieh J, Shyue JJ. Characterization of the microstructure and photoelectrical properties of TiO2–SrTiO3 and TiO2–CeO2 Nanocomposites. J Multifunct Compos 2014;2(1): [69] Sun Y, Chemelewski WD, Berglund SP, et al. Antimony-doped tin oxide nanorods as a transparent conducting electrode for enhancing photoelectrochemical oxidation of water by hematite. ACS Appl Mater Interfaces 2014;6(8):5494–9. [70] Qin N, Liu Y, Wu W, et al. One-dimensional CdS/TiO2 nanofiber composites as efficient visible-light-driven photocatalysts for selective organic transformation: synthesis, characterization, and performance. Langmuir 2015;31(3):1203–9.

Photoactive Materials

571

[71] Di J, Xia J, Yin S, et al. Preparation of sphere-like g-C3N4/BiOI photocatalysts via a reactable ionic liquid for visible-light-driven photocatalytic degradation of pollutants. J Mater Chem A 2014;2(15):5340–51. [72] Karpova SS, Moshnikov VA, Mjakin SV, Kolovangina ES. Surface functional composition and sensor properties of ZnO, Fe2O3, and ZnFe2O4. Semiconductors 2013;47 (3):392–5. [73] Wang P, Wang J, Wang X, Yu H, Yu J. Cu2O–rGO–CuO Composite: an effective Z-scheme visible-light photocatalyst. Curr Nanosci 2015;11(4):462–9. [74] Tian W, Zhai T, Zhang C, et al. Low‐cost fully transparent ultraviolet photodetectors based on electrospun ZnO–SnO2 heterojunction nanofibers. Adv Mater 2013;25 (33):4625–30. [75] Gao W, Wu G, Ling Y, Sun J. H2 sensing properties of Pd modified WO3–Fe2O3 nanostructured composite films prepared by amorphous W–Fe dealloying. J Nanosci Nanotechnol 2013;13(2):1190–3. [76] Zhu Y, Schultz AM, Rohrer GS, Salvador PA. The orientation dependence of the photochemical activity of a‐Fe2O3. J Am Ceram Soc 2016;99(7):2428–35. [77] Rao PM, Cai L, Liu C, et al. Simultaneously efficient light absorption and charge separation in WO3/BiVO4 core/shell nanowire photoanode for photoelectrochemical water oxidation. Nano Lett 2014;14(2):1099–105. [78] Liang P, Zhang L, Zhao X, et al. Synthesis of ZnFe2O4/TiO2 composite nanofibers with enhanced photoelectrochemical activity. Sci Adv Mater 2015;7(2):295–300. [79] Kerkez Ö, Boz ˙I. Photodegradation of methylene blue with Ag2O/TiO2 under visible light: operational parameters. Chem Eng Commun 2015;202(4):534–41. [80] Zhu L, Oh WC. Novel Bi2S3/TiO2 heterogeneous catalyst: photocatalytic mechanism for decolorization of texbrite dye and evaluation of oxygen species. J Korean Ceram Soc 2016;53(1):56–62. [81] Sakai G, Tanaka A, Sueda T, et al. Peculiar crystal growth of the trivalent titanium derived TiO2–SnO2 precursor under hydrothermal conditions. Chem Lett 2016;45 (3):318–20. [82] Lee H, Park Y, Kang M. Synthesis of characterization of ZnxTiyS and its photocatalytic activity for hydrogen production from methanol/water photo-splitting. J Ind Eng Chem 2013;19(4):1162–8. [83] Ahmad H, Kamarudin SK, Minggu LJ, Kassim M. Hydrogen from photo-catalytic water splitting process: a review. Renew Sustain Energy Rev 2015;43:599–610.

Further Reading Balzani V, Ceroni P, Juris A. 2014. Photochemistry and photophysics: concepts, research, applications. New Jersey, NJ: John Wiley & Sons; 2014. Bhaskaran M, Sriram S, Iniewski K. 2013. Energy harvesting with functional materials and microsystems. Nottingham: CRC Press; 2013. Borchert H. 2014. Solar cells based on colloidal nanocrystals. Switzerland: Springer; 2014. Brabec CJ, Dyakonov V, Parisi J, Sariciftci NS, editors. 2013. Organic photovoltaics: concepts and realization. vol. 60. New York, NY: Springer Science & Business Media; 2013. Caironi M, Noh YY. 2015. Large area and flexible electronics. New York, NY: John Wiley & Sons; 2015. Chen B. 2014. In: Qian G, editor. Metal-organic frameworks for photonics applications. vol. 157. Berlin: Springer; 2014. Clerici MG, Kholdeeva OA. 2013. Liquid phase oxidation via heterogeneous catalysis: organic synthesis and industrial applications. Hoboken, NJ: John Wiley & Sons; 2013. Dincer I, Joshi AS. 2013. Solar-based hydrogen production systems. New York, NY: Springer; 2013. Dincer I, Midilli A, Kucuk H, editors. 2014. Progress in exergy, energy, and the environment. Switzerland: Springer; 2014. Dincer I, Rosen MA. 2012. Exergy: energy, environment and sustainable development. Oxford: Newnes; 2012. Dincer I, Zamfirescu C. 2016. Sustainable hydrogen production. Amsterdam: Elsevier; 2016. Gogotsi Y, Presser V, editors. 2013. Carbon nanomaterials. Boca Raton, FL: CRC Press; 2013. Gratzel M, editor. 2012. Energy resources through photochemistry and catalysis. Orlando, FL: Elsevier; 2012. Guczi L, Erdôhelyi A, editors. 2012. Catalysis for alternative energy generation. New York, NY: Springer Science & Business Media; 2012. Klauk H. 2012. Organic electronics II: more materials and applications, vol. 2. Hoboken, NJ: John Wiley & Sons; 2012. Li ZR, editor. 2015. Organic light-emitting materials and devices. Boca Raton, FL: CRC press; 2015. Marshall JM, Dimova-Malinovska D, editors. 2012. Photovoltaic and photoactive materials: properties, technology and applications. vol. 80. New York, NY: Springer Science & Business Media; 2012. McEvoy AJ, Castaner L, Markvart T. 2012. Solar cells: materials, manufacture and operation. Amsterdam: Academic Press; 2012. Mohanty SP. 2015. Nanoelectronic mixed-signal system design. New York, NY: McGraw-Hill Education 2015 (No. 9780071825719). Rand BP, Richter H. 2014. Organic solar cells: fundamentals, devices, and upscaling. Boca Raton, FL: Pan Stanford Publishing; 2014. Schneider J, Bahnemann D, Ye J, Puma GL, Dionysiou DD, editors. 2016. Photocatalysis: fundamentals and perspectives. London: Royal Society of Chemistry; 2016. Thomas JM, Thomas WJ. 2014. Principles and practice of heterogeneous catalysis. New York, NY: John Wiley & Sons; 2014. Van de Krol R, Grätzel M. 2012. Photoelectrochemical hydrogen production, vol. 90. New York, NY: Springer; 2012. Vogelbaum HS, Sauvé G. 2017. Recently developed high-efficiency organic photoactive materials for printable photovoltaic cells: a mini review. Synth Met 2017;223:107–21. Yu H. 2015. Dancing with light: advances in photofunctional liquid-crystalline materials. Boca Raton, FL: CRC Press; 2015.

Relevant Websites http://bahcesehir.edu.tr/ Bahcesehir University. https://www.youtube.com/watch?v=qlj7q5ZINVo Experiments with photoactive materials. http://www.gcgw.org/ Global Conference on Global Warming. http://www.infectioncontroltoday.com/news/2011/04/understanding-photoactive-materials-role.aspx ICT. http://www.ich2p-2017.org/ International Conference on Hydrogen Production. http://www.ieees9.fesb.unist.hr/ International Exergy, Energy, and Environment Symposium. https://www.isof.cnr.it/content/photonics-research-activities-photoactive-materials-biomedical-applications ISOF.

572

Photoactive Materials

http://www.aibn.uq.edu.au/Download/NSF/John_Bell_QUT.pdf Photoactive nanomaterials: contributing to sustainable futures. https://phys.org/news/2015-04-pseudoparticles-photoactive-material.html Pseudoparticles travel through photoactive material. http://www.sigmaaldrich.com/materials-science/organic-electronics/opv-tutorial.html Sigma Aldrich Organic Photovoltaics. http://www.epr.unito.it/Research_index/Photoactive%20Materials.htm Unito EPR Lab. https://www.uoit.ca/ University of Ontario Institute of Technology. http://cerl.uoit.ca/ UOIT Clean Energy Research Laboratory. http://dornsife.usc.edu/labs/prezhdo-group/photoactive/ USC.

2.18 Hydrides Robert A Varin, University of Waterloo, Waterloo, ON, Canada r 2018 Elsevier Inc. All rights reserved.

2.18.1 Introduction 2.18.2 Background 2.18.3 Selected Complex Hydrides 2.18.3.1 LiAlH4 With Nano-Additives 2.18.3.2 Metal Borohydrides 2.18.3.2.1 The (LiBH4/NaBH4-MnCl2) systems 2.18.4 Summary Acknowledgments References Relevant Websites

Nomenclature

573 574 575 575 580 580 590 591 591 592

B2H6 DSC EA G H2 DH

Diborane gas Differential scanning calorimetry Apparent activation energy (kJ/mol) Gibbs free energy (kJ) Hydrogen Enthalpy change (kJ/molH2)

MS m P R DS t T

Mass spectrometry Mass (g) Pressure (bar) Universal gas constant (8.314 J/molK) Entropy change (J/molH2K) Time (h) Temperature (1C or K)

Acronyms CO2 D.O.E. Et2O FC FCV

Carbon dioxide Department of Energy USA Diethyl ether Fuel cell Fuel cell vehicle

n PCT PEM QTR rGO TPD

Nano Pressure–concentration–temperature curve Proton exchange membrane Total injected milling energy (kJ/g) Reduced graphene oxide Temperature programmed desorption

2.18.1

Introduction

For the industrialized world, the past 300 years of developments were based on burning fossil fuels. First, it was mainly coal and later it became crude oil and natural gas. Regarding crude oil, the exponential increase in the use of oil to provide power for transportation and industry in the last 100 years, led to a substantial depletion of oil fields in oil-rich countries. It is predicted that the remaining oil reserves will be, most likely, largely consumed within roughly the next 50–70 years [1–3]. In addition, consumption of fossil fuels leads to increasing levels of CO2 (so called “greenhouse gas”) in the atmosphere, which in turn, may lead to the average temperature increase on earth [4]. The ideal situation would require that alternative energy sources and fuels be developed to promote future world energy security. Hydrogen has been perceived as a future energy carrier that may allow a gradual transformation from a fossil fuel-based economy to a hydrogen economy. The full implementation of hydrogen economy requires overcoming three obstacles: an inexpensive hydrogen production that would not use fossil fuels, development of highly efficient fuel cells (FCs), and efficient storage of hydrogen, particularly for mobile/automotive applications where hydrogen gas would be supplied to FCs that, in turn, will power a variety of transport vehicles in a clean, inexpensive, safe, and efficient manner [5]. For the transportation sector (automotive) gaseous and liquid hydrogen storage techniques have been known for quite a long time but they have a number of serious drawbacks, such as relatively low volumetric densities, substantial energy losses occurring during pressurizing or liquefaction and various safety issues. Nevertheless, high pressure (70 MPa) gaseous storage is the most viable solution for the next 20–30 years. The first commercial Toyota and Hyundai Fuel Cell Vehicles (FCV), coming to market in 2016–2017 store H2 in a more conventional, gaseous form under a high pressure of 70 MPa [6,7]. The most attractive storage method is the one in solid hydrides but it is also the most challenging. A proton exchange membrane fuel cell (PEM FC) stack (sometimes also named a polymer electrolyte fuel cell (PEFC)), which has been a primary candidate for an automotive power plant, imposes severe constraints on a solid state H2 hydride storage system such as a low dehydrogenation temperature range of 70–801C at 1.1–1.8 bar H2 pressure [8]. Furthermore, the Department of Energy

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00241-8

573

Hydrides

574

US (D.O.E) 2017 targets for the driving distance of, at least, 300 miles (480 km) require a storage system with the gravimetric system capacity of at least 5.5 wt% H2 which includes storage medium, tank, and some auxiliary devices. That translates to the storing hydride material capacity of roughly 11 wt% H2 [8,9]. In addition, the convenience of operating a fuel cell vehicle (FCV) calls for reversible and quick “on board” refueling at a hydrogen station. In summary, these constraints, both technical and manmade, require the potential hydrogen storage material for fueling a PEM FC to desorb at o1001C in 1 bar H2, have a practical achievable capacity B11 wt% H2, exhibit reversibility under reasonable conditions of pressure/temperature, desorb/absorb rapidly, and be relatively inexpensive.

2.18.2

Background

Unfortunately, no solid hydride system that could meet all those requirements has been discovered yet. Despite a very extensive research on solid hydrides in the past decade, a breakthrough has not been achieved yet and there is not even one hydride system that is close to commercialization for application as a storage medium for the most important automotive market. However, there is a number of other potential market applications for simple H2 generation systems rather than “on board” reversible storage systems, where some of them could be even recharged “off board,” supplying H2 at the ambient and slightly elevated temperatures, in the commercial, non-automotive sectors of the economy. For example, they could be utilized for supplying FCs in such applications as stationary auxiliary power systems, off-road vehicles (forklifts, street sweepers etc.), locomotives, submarines, drones, coastal and international shipping, auxiliary devices in air transportation, lawn mowers, disposable cartridges for longduration, low power military devices, civilian portable electronic devices, bulk hydrogen storage, and many others [10–18]. A screening of various hydrides from the standpoint of their suitability of desorbing H2 in 1 bar pressure and at temperatures, not exceeding 1001C, is very important. This can be achieved by knowing a thermodynamic property of hydrides such as the enthalpy of decomposition/formation change, DH. If the natural logarithm of desorption/absorption H2 pressure is plotted versus wt% H2 desorbed/absorbed at constant temperature then so-called pressure–composition–temperature (PCT) curve is obtained which usually exhibits plateau [19]. Then the equilibrium mid-plateau H2 pressures for the hydrogen desorption/absorption plateau is related to temperature through the Van’t Hoff equation [19]: lnðp=p0 Þ ¼

DH=RT þ DS=R

ð1Þ

3.0 2.8 2.6 2.4 2.2 2.0 1.8 1.6 1.4 1.2 1

295°C (2.6 bar) 285°C (2.1 bar) 275°C (1.6 bar)

0 (A)

10

20

1.1 1.0 ΔH = 62.4 kJ/mol 0.9 ΔS = 117.8 J/(mol•K) 0.8 0.7 0.6 0.5 0.4 y = −7.4951x + 14.163 0.3 R2 = 1 0.2 1.75 1.77 1.79 1.81 (B) 1000/T (1/K) In p

Pressure (bar)

where p is the recorded pressure (atm or bar), p0 is the atmospheric pressure (1 atm or 1 bar), DH and DS are the enthalpy and entropy changes in kJ/molH2 and J/molH2K, respectively, R is the gas constant (8.314472 J/molK) and T is absolute temperature (K). This approach requires time-consuming measurements of pressure and corresponding H2 capacity for one PCT and at least for three selected temperatures. More recently, we simplified this approach by using step-wise desorption curves (pseudo-PCT curves), where we measure pressure versus desorption time at a constant temperature and then after obtaining satisfactory equilibrium plateau, temperature is rapidly ramped to a higher value until equilibrium plateau is achieved and again temperature is ramped to yet higher value and equilibrium plateau pressure recorded. Then the term ln (p/p0), where p is the equilibrium plateau pressure for each temperature, is plotted versus 1000/T according to Eq. (1). This method is shown in Fig. 1(A) and (B) as applied to a ball milled mixture of lithium amide (LiNH2) and lithium hydride (LiH) in the molar ratio (LiNH2 þ 1.2LiH) [20]. A value of the dehydrogenation enthalpy obtained in Fig. 1(B) agrees well with the values quoted in the literature [21]. Furthermore, according to the Van’t Hoff Eq. (1) a certain hydride (or a hydride composite system) would be thermodynamically capable of desorbing H2 only if desorption pressure (e.g., 1 bar) is lower than its equilibrium plateau pressure at experimental desorption temperature (T ¼ const.). The larger the difference between desorption pressure and the PCT equilibrium pressure, the larger is the driving force for decomposition. However, these thermodynamic considerations do not tell us about the rapidity of H2 release during dehydrogenation which is a kinetic property.

30 40 Time (min)

50

60

1.83

Fig. 1 (A) Step-wise pseudo-pressure-concentration-temperature curves at varying temperatures and (B) corresponding Van’t Hoff plot (Eq. (1)) for the (LiNH2 þ 1.2LiH) mixture milled for 25 h under a high energy milling mode. Reproduced from Varin RA, Jang M. The effects of graphite on the reversible hydrogen storage of nanostructured lithium amide and lithium hydride (LiNH2 þ 1.2LiH) system. J Alloys Compd 2011;509:7143–51.

Hydrides

575

The earlier status of solid state hydrides and their H2 storage behavior has been reported in our book Nanomaterials for Solid State Hydrogen Storage that was published in 2009 [19] and a book chapter that appeared in 2013 [22]. The aim of the present chapter is to review the progress made in the past few years in the research on complex, metal-nonmetal hydrides and their composites which have the greatest potential as a viable source of generation/storage H2 for non-automotive applications, particularly, for being applied in disposable containers/cartridges. In this respect a system based on LiAlH4 with nanometric additives, may be close to near-commercialization for applications for disposable H2 storage cartridges for long-duration, low power demand devices that use H2. Newly discovered manganese borohydride exhibiting very interesting dehydrogenation properties is also discussed.

2.18.3

Selected Complex Hydrides

In the broadest sense complex hydrides are composed of an anionic metal-hydrogen complex or non metal-hydrogen complex bonded to a cationic alkali or transition metal (TM) [19,23,24]. Thus complex hydrides can be roughly subdivided into two categories: group I and II salts of [AlH4] , [NH2] , [BH4] , i.e., alanates, amides, and borohydrides [19,23,24], and TM complex hydrides that have anionic (TMHx) complexes such as [FeH6]4 attached to a cationic light metal, for example, Mg2 þ , in Mg2FeH6 [19,23]. Their bonding is usually an ionic-covalent mix [24]. Similar TM ternary complex hydrides exist in the Mg–Co, Mg–Ni, and Mg–Mn systems forming Mg2CoH5, Mg2NiH4, and Mg3MnH7 (the latter was synthesized under 20 kbarH2 at B8001C [19,25]). Some references classify certain complex hydrides, for example, NaBH4, as “chemical hydrides” because they can easily react with water solution of KOH or NaOH (or water steam) releasing hydrogen [19].

2.18.3.1

LiAlH4 With Nano-Additives

One of the most interesting hydrides for solid state hydrogen generation/storage, for non-automotive applications, is a complex metal hydride LiAlH4 (lithium alanate), since it can liberate a relatively large theoretical quantity of 7.9 wt% H2 much below below 2501C [19]. Graetz and Reilly [26] classified LiAlH4 as belonging to a group of hydrides called “kinetically stabilized metal hydrides” which also include AlH3, Mg(AlH4)2, and Ca(AlH4)2, among others. All these hydrides are characterized by an equilibrium H2 pressure of their respective PCT curves at room temperature (298 K), much higher than 1 bar. This feature alone creates a large driving force for decomposition. However, they are quite stable at near room temperature, most likely, due to kinetic limitations as their name indicates. The mechanisms responsible for their stable behavior are not well understood although slow/hydrogen metal diffusion and surface barriers that hinder the easy formation of molecular H2 are quite likely [26]. It is well established [19,27] that pure (additive-free) LiAlH4 decomposes releasing H2 according to the following steps: LiAlH4 ðsÞ-LiAlH4 ðlÞ

ð2aÞ

LiAlH4 ðlÞ-1=3Li3 AlH6 ðsÞ þ 2=3AlðsÞ þ H2 ðgÞ

ð2’bÞ

1=3Li3 AlH6 ðsÞ-LiH þ 1=3Al þ 0:5H2

ð2cÞ

LiH-Li þ 0:5H2

ð2’dÞ

LiH þ Al-LiAl þ 0:5H2

ð2”dÞ

or where s – solid, l – liquid, and g – gas. In differential scanning calorimetry (DSC), reaction (2’a) is endothermic, (2’b) is exothermic, and (2c) and (2’d) are both endothermic reactions. Reactions (2’b), (2c) and (2’d) proceeds with a theoretical hydrogen release of 5.3, 2.6, and 2.6 wt%, respectively [19]. Obviously, these values will be lower for purity-corrected capacity. Decomposition of additive-free LiAlH4 always occurs from a molten state through reaction (2’b). The DSC thermal peak maxima for the above reactions are not affected by ball milling (BM) as shown for a DSC curve in Fig. 2(A) [27]. The first exothermic peak at 143.81C in Fig. 2(A), before the peak (2a) was assigned by Block and Gray [28] to the reaction of the surface aluminum-hydroxyl groups owing to the presence of impurities. We reported in Refs. [29–31] that the nano-additives substantially modified the thermal DSC response of LiAlH4. Adding 5 wt% of the nano-nickel (n-Ni) additive [29] by a simple mixing with LiAlH4, using mortar and pestle, does not change the DSC curve as compared to that for ball milled LiAlH4 without additives and LiAlH4 still decomposes in a molten state (Fig. 2(B)) through reaction (2’b) forming Li3AlH6 (s). In contrast, a highly energetic BM of LiAlH4 with either 5 wt% of the n-Ni or n-Fe additive, completely eliminates melting of LiAlH4 and the LiAlH4 decomposition peak (2’b) in Fig. 2(A) and (B) merges with the first peak for reaction of the surface aluminum-hydroxyl groups [28] and becomes now a single peak corresponding to reaction (2”b): LiAlH4 ðsÞ-1=3Li3 AlH6 ðsÞ þ 2=3AlðsÞ þ H2 ðgÞ

ð2”bÞ

The (2”b) peak is now shifted to slightly lower temperatures (Fig. 2(C) and (D)). Remarkably, this behavior is identical for both types of nanometric metallic additive, either n-Ni or n-Fe. The other two peaks corresponding to reaction (2c) and (2’d) appear unaffected except that they are shifted to slightly lower temperatures after high energy BM with a nanometric metal additive

Hydrides

576

(2a) DSC (mW/mg) 174.4°C exo 4 10°C/min 2

(2d) 430.6°C (2c)

0

234.7°C

−2 143.8°C (2′b)

−4 50 (A)

LiAIH4-milled IMP68-R40-15min

DSC (mW/mg) 169.8°C exo 3 2 1 0 −1 −2 153.8°C

100 150 200 250 300 350 400 450 Temperature (°C)

DSC (mW/mg) exo 4

(2d)419.0°C (2c)209.0°C

242.8°C

(LiAIH4+5 wt% n-Ni)-mixed-15min

100 150 200 250 300 350 400 450 Temperature (°C)

DSC (mW/mg) 3 exo

(2d) 418.5°C

(2c)

2

3

203.2°C

1

2

0

1

(LiAIH4+5 wt% n-Ni)-IMP68-R132-15min

0

(2″b) 50

(C)

50 (B)

436.1°C

100

150

200 250 300 350 Temperature (°C)

(2″b)

−1

(LiAIH4+5 wt% n-Fe)-IMP68-R132-15min

132.1°C

138.6 °C

−2 400

50

450 (D)

100 150 200 250 300 350 400 450 Temperature (°C)

Fig. 2 Differential scanning calorimetry (DSC) curves for (A) additive-free, ball milled LiAlH4, (B) (LiAlH4 þ 5 wt% n-Ni) mixed without ball milling, (C) (LiAlH4 þ 5 wt% n-Ni) ball milled under a high energy mode, and (D) (LiAlH4 þ 5 wt% n-Fe) balled milled under a high energy mode. IMP68-a high impact energy mode of milling, R40 and R132 means ball-to-powder-weight ratio ¼40 and 132, respectively. n-Ni: nano-nickel additive. Reproduced from Varin RA, Zbroniec L. Decomposition behavior of unmilled and ball milled lithium alanate (LiAlH4) including long-term storage and moisture effects. J Alloys Compd 2010;504:89–101; Varin RA, Zbroniec L. The effects of nanometric nickel (n-Ni) catalyst on the dehydrogenation and rehydrogenation behavior of ball milled lithium alanate (LiAlH4). J Alloys Compd 2010;506:928–39; and Varin RA, Parviz R. The effects of the micrometric and nanometric iron (Fe) additives on the mechanical and thermal dehydrogenation of lithium alanate (LiAlH4), its selfdischarge at low temperatures and rehydrogenation. Int J Hydrog Energy 2012;37:9088–102.

(Fig. 2(C) and (D)). The merging of peaks attributed to surface hydroxyl reaction and the decomposition of LiAlH4 in reaction (2’b) into one peak (2”b) for solid state decomposition is a very peculiar phenomenon which we reported for the first time [22,29]. As pointed out in Ref. [22], it is still unclear if this is just a coincidence or indeed, a new physical phenomenon. We hypothesized in Refs. [22,29,30] that reaction of the surface aluminum-hydroxyl groups may trigger a spontaneous decomposition of ball milled LiAlH4 containing nanometric metallic additives at temperatures much lower than the melting point of either unmilled or ball milled LiAlH4 without additives. However, it is unknown if the origin of the exothermic nature of the first exothermic peak in Fig. 1(A) and (B) is related to the inherent exothermic nature of reaction of the surface aluminum-hydroxyl groups or, indeed, reaction (2”b) is simply exothermic in nature, despite the fact that melting has been eliminated. It should also be pointed out that in contrast to the nanometric Fe additive, the DSC curves for LiAlH4 ball milled with the 5 wt% micrometric size Fe additive, were exactly the same as those for additive-free LiAlH4 in Fig. 2(A), in a sense, that the decomposition of LiAlH4 occurred after melting (reaction (2’b)) [30]. The DSC observations clearly show that the elimination of the melting stage of LiAlH4 and its decomposition from solid state can only occur if two conditions are met: (1) the metallic additive must be nanometric and (2) high energy BM is employed for processing of LiAlH4 with a nanometric additive. It seems that an important condition is that the n-Ni and n-Fe additives must be very intimately embedded within the LiAlH4 particles in contrast to mixing where metal nanoparticles are just loosely distributed at the surfaces of LiAlH4 powder particles. In addition, (LiAlH4 þ 5 wt% n-Fe) nanocomposite powder was subjected to a thermal sectioning in DSC by stopping the test at varying pre-determined temperatures and after stopping the test at selected temperature, the powder sample was immediately extracted from a crucible for an X-ray diffraction (XRD) test [22,30]. At 415 and 5001C just before and after peak (2d), an intermetallic compound LiAl appeared in the microstructure which confirmed the occurrence of reaction (2”d). Another phenomenon that we discovered during our studies on the effect of additives in LiAlH4, which has never been reported before in the literature, is that some nanometallic [22,30] and nano-carbide [31] additives trigger a profound mechanical dehydrogenation phenomenon. We have found that LiAlH4 ball milled together with the n-Fe additive, under a high energy milling mode, starts decomposing rapidly during milling releasing large quantities of H2 (Fig. 3(A)) [30]. Interestingly, no H2 desorption was observed during low energy milling of LiAlH4 containing n-Fe, for example, under a low energy shearing mode (Fig. 3(A)). Similarly, no H2 desorption occurred during high energy BM for LiAlH4 containing micrometric Fe (m-Fe) and, for comparison, containing the micrometric and nanometric Ni (m-Ni and n-Ni) additive. In order to identify reactions occurring as the milling progresses, the XRD patterns were obtained from the powders milled for 15 min, 1 and 5 h, which are compared to the XRD pattern for a just mixed powder in Fig. 3(B). After barely

Hydrides

6000

4.0

LiAIH4+5 wt% n-Fe-IMP68-4B-R132

∗ LiAIH4 (12-0473)

∗

3.0

Li3AIH6 (27-0282)

4000

2.5

Counts

H2 desorbed (wt %)

3.5

2.0 1.5

∗

∗

∗

∗ ∗ ∗

∗∗

∗ ∗

∗

∗

∗ Mix 15 min

∗

2000

1.0

1h

LiAIH4+5 wt% n-Fe-IMP68-4B-R132 LiAIH4+5 wt% n-Fe-LES6-4B-R132

0.5

5h 0

0 0 (A)

577

100

200 300 Milling time (min)

20

400 (B)

30 40 Degree 2-theta

50

Fig. 3 (A) The quantity of H2 desorbed during ball milling under the high impact energy (IMP68; 4B: 4 balls [19]) mode and the low shearing energy (LES6) mode as a function of milling time. (B) X-ray diffraction patterns after milling under high energy (IMP68) for various times. Milling under high energy mode (IMP68) was repeated twice (solid circles and triangles). ICDD file numbers are shown for peak identification. 4B – four balls in the milling vial. Ball-to-powder weight ratio R ¼132. Reproduced from Varin RA, Wronski ZS. Progress in hydrogen storage in complex hydrides. In: Gandia LM, Arzamendi G, Diéguez PM Renewable Hydrogen Technologies. Production, Purification, Storage, Applications and Safety. Elsevier; 2013. p. 293–332 [chapter13] and Varin RA, Parviz R. The effects of the micrometric and nanometric iron (Fe) additives on the mechanical and thermal dehydrogenation of lithium alanate (LiAlH4), its selfdischarge at low temperatures and rehydrogenation. Int J Hydrog Energy 2012;37:9088–102.

15 min of milling, the diffraction peaks of Li3AlH6 are clearly seen as opposed to the LiAlH4 peaks observed for the mixed powders. The intensity of the Li3AlH6 peaks increases after 1 and 5 h (300 min) of milling while the intensity of LiAlH4 peaks gradually decreases and they eventually disappear (Fig. 3(B)). Simultaneously, more H2 is desorbed as can be seen in Fig. 3(A). It is obvious that during high energy milling, there is a gradual decomposition of LiAlH4 in solid state according to reaction (2”b). Furthermore, XRD studies have shown that BM resulted in a varying degree of the lattice expansion of LiAlH4 for both the n-Fe and n-Ni additives [22,30]. It seems that a volumetric lattice expansion larger than 1% can trigger the accelerated decomposition of LiAlH4 accompanied by continuous H2 desorption during milling according to reaction (2”b). It was hypothesized [8,22,30] that the Fe and Ni ions are able to dissolve in the lattice of LiAlH4 by the action of mechanical energy, replacing the Al ions and forming a substitutional solid solution of either the LiAl1 xFexH4 or LiAl1 xNixH4 type [8,22,30]. This is similar to a well-known mechanical alloying phenomenon. The ionic radius of 77 pm for Fe2 þ is much larger than 53 pm for Al3 þ [32] and the LiAlH4 lattice expansion increases. In contrast, the ionic radius of Ni2 þ is 69 pm [32] and the volumetric lattice expansion is smaller. It is interesting that the profound dissolution of the Fe2 þ ions in the LiAlH4 matrix can only occur when the Fe additive is nanometric (n-Fe) while it does not occur if the additive is micrometric in size [8,22,30]. Furthermore, the experimental results in Refs. [8,22,30] showed that n-Fe is more efficient in enhancing the mechanical dehydrogenation of LiAlH4 during BM than n-Ni. This is a puzzling finding because the Ni2 þ ionic radius is smaller than that of the Fe2 þ ion and thus, one would expect more Ni2 þ ions dissolved in the LiAlH4 lattice. Apparently, for the same milling conditions, the incorporation of Ni2 þ ions into the ionic-covalent crystal of LiAlH4 is less efficient than the Fe2 þ ions. Therefore, we suggested [33] that an efficient incorporation of metal ions in an ionic-covalent lattice, like that for LiAlH4, could be driven by chemical reactions with oxide forming on the additive metal nanoparticles, instead of the metal itself. Indeed, the n-Fe particles of received n-Fe (NANOFER 25S from the Nano Iron, s.r.o., The Czech Republic) had a high surface content of Fe3O4 (about 15 wt.%) [30] that was much higher than that for n-Ni [34]. The NANOFER 25S n-Fe containing Fe3O4, could be more efficient in enhancing the dehydrogenation reaction during BM. Since the iron oxide associated with n-Fe is Fe3O4 [30] which, in essence is a partly reduced form of Fe2O3 such as Fe2 þ OFe32 þ O3, we postulated the following reaction sequence [33]:
 ð1 þ 2xÞLiAl3þ H4 þ xFe2þ O
Fe3þ 2 O3 -LiAl3þ Fe3þ x H4 þ 2xFe2þ O þ 2xLiOH þ ð1 þ 2xÞAlH3 ð3aÞ LiAl3þ Fe3þ x H4 -1=3 Li3 AlFex H6 þ 2=3AlðFex Þ þ H2

ð3bÞ

Li3 AlFex H6 2 3LiH þ AlðFex Þ þ 3=2H2

ð3cÞ 3þ

In reaction (3a) the additive is metal oxide Fe3O4. A fraction of ions in Fe3O4 exhibit valence Fe the same as the valence of Al3 þ . The ionic radius of Fe þ 3 is smaller than that of Fe2 þ (63 vs. 77 pm [32]) and the charge is the same as Al. So, the substitution of Fe3 þ for Al3 þ (53 pm [32]) in the LiAlH4 alanate lattice sites is more favorable. Both the substituted LiAlFexH4 and Li3AlFexH6 might be easier to dehydrogenate through (3b) and (3c). Finally, it must be pointed out that the mechanism of accelerated mechanical dehydrogenation of LiAlH4, containing n-Fe additive, seems not to be a typical catalytic mechanism in which a catalyst is supposed to accelerate the surface formation of molecular H2 from the atomic hydrogen escaping from the bulk.

Hydrides

578

6000

(LiAIH4+5 wt% TiC)

Injected energy QTR132= 72.8 kJ/gh

7

n-TiN n-TiC n-ZrC

6 5

LiAIH4 Li3AIH6 TiC Al

4000 Counts

Hydrogen desorbed (wt%)

8

4

(12-0473)/(73-0461) (27-0282) (01-1222) (85-1327)

AR LiAIH4

3

2000

15 min BM (QTR=18.2 kJ/g)

2

5h BM (QTR=364 kJ/g)

1

25h BM (QTR=1820 kJ/g)

0

0 0 (A)

5

10 15 20 Milling time (h)

25

20

30 (B)

30

40 50 60 Degree 2-theta

70

80

90

Fig. 4 (A) The quantity of H2 desorbed during ball milling (mechanical dehydrogenation) under the high impact energy mode IMP68-4B-R132 (QTR132 ¼72.8 kJ/gh [35]) as a function of milling time for LiAlH4 containing 5 wt% n-TiC, n-TiN, and n-ZrC. (B) X-ray diffraction patterns after milling of (LiAlH4 þ 5 wt% n-TiC) under high energy milling mode IMP68-4B-R132 (QTR132 ¼72.8 kJ/gh [35]) for various times. ICDD file numbers are shown for peak identification. AR, as received; BM, ball milled. Reproduced from Varin RA, Parviz R. The effects of the nanometric interstitial compounds TiC, ZrC and TiN on the mechanical and thermal dehydrogenation and rehydrogenation of the nanocomposite lithium alanate (LiAlH4) hydride. Int J Hydrog Energy 2014;39:2575–86.

A similar mechanical dehydrogenation phenomenon was observed for LiAlH4 containing nanometric interstitial compounds, such as nanocarbides n-TiC and n-ZrC and nanonitride n-TiN [31]. Fig. 4(A) shows a trend of increasing amount of H2 desorbed during BM under milling energy intensity of QTR132 ¼ 72.8 kJ/gh (the methodology of milling energy intensity for a certain milling mode is described in Ref. [35]) with increasing milling time (h). After about 18 h of milling (total injected milling energy QTR ¼ 1310 kJ/g [35]) the amount of mechanical dehydrogenation for the n-TiN additive reaches B6.6 wt% H2. Finally, after 25 h of milling (QTR ¼ 1820 kJ/g [35]) the amount of mechanical dehydrogenation reaches B7 wt% H2 for the n-TiC additive, and B3 wt% H2 for the n-ZrC additive. Therefore, the mechanical dehydrogenation curves in Fig. 4 confirm that the mechanical dehydrogenation rate of the samples containing the nanometric interstitial compound additives increases noticeably during high energy BM, with increasing quantity of injected energy, in the order of n-TiN4n-TiC4n-ZrC, while no mechanical dehydrogenation occurs during milling when the quantity of total injected energy is very small, QTR ¼ 18.2 kJ/g [31]. The XRD patterns in Fig. 4(B) were taken for (LiAlH4 þ 5 wt% n-TiC), milled for 15 min, 5 and 25 h [31]. For comparison, an XRD pattern for as received (AR) LiAlH4 is also included. As can be seen, after 15 min of BM of a nanocomposite (LiAlH4 þ 5 wt% n-TiC) (QTR ¼ 18.2 kJ/g) no peaks of Li3AlH6 are observed. The diffraction peaks of Li3AlH6 appear after 5 h of milling (QTR ¼364 kJ/g) accompanied by decreasing intensity of the LiAlH4 peaks and more H2 desorption (Fig. 4(A)). This trend continues until a complete disappearance of the LiAlH4 and Li3AlH6 peaks after 25 h BM (QTR ¼1820 kJ/g) occurs. Only diffraction peaks of elemental Al and the n-TiC additive are visible on the XRD pattern after 25 h of BM (Fig. 4(B)). It is obvious that during high energy BM with intensity QTR132 ¼ 72.8 kJ/gh [35] there is a gradual decomposition of LiAlH4 in solid state according to reactions (2”b) and (2c). It was found that there was no measurable change in a unit cell volume of LiAlH4 containing nanointerstitial compounds after BM [31]. That means that accelerated mechanical dehydrogenation for LiAlH4 containing nanometric interstitial compounds in Fig. 4(A) is unrelated to the lattice expansion that might have been induced by, for example, diffusion of the Ti or Zr ions from the nanometric interstitial compounds into the LiAlH4 crystal lattice. The absence of a measurable lattice expansion is in agreement with the experimental fact that the nanometric interstitial compounds are very stable during high energy BM and do not decompose releasing the Ti or Zr ions. The crystal structures of interstitial compounds like TiC and TiN belong to the general family of electron compounds whose crystallographic structures depend on the average number of valence electrons per atom (valence electron concentration (VEC)) [36]. However, for both TiC and ZrC VEC ¼ 8 and TiN has VEC¼ 9 [36] but the n-ZrC additive is the weakest one then their catalytic activity for mechanical dehydrogenation seems not to be related to their VEC number. We postulated in Ref. [31] that the primary factor responsible for a strong catalytic effect of the nanometric interstitial compounds during mechanical dehydrogenation is a very small particle size on the order of 20 nm or less. The secondary factor, at an equal particle size, seems to be a stronger catalytic effect for Ti than that for Zr in an interstitial compound. This hypothesis is in agreement with the work of Rafi-ud-din et al. [37] who suggested that the very stable nanometric TiC particles, strongly embedded in the LiAlH4 matrix during BM, produced deformed interfacial regions surrounding them which exhibited an increased density of lattice defects which, in turn, enhanced dehydrogenation capability. Since the n-TiN had the smallest average particle size of 20 nm work [31], most likely, it produced the highest density of lattice defects in the surrounding deformed matrix leading to considerably enhanced mechanical dehydrogenation, compared to the other two additives, n-TiC and n-ZrC, exhibiting larger particle sizes. Table 1 summarizes the propensity to mechanical dehydrogenation for LiAlH4 with metallic and nonmetallic nano-additives.

Hydrides

Table 1

579

Summary of the high energy ball milling behavior of the LiAlH4 systems with nano-additives Reaction between additive and the LiAlH4 matrix and formation of metal/intermetallic product

H2 desorption during ball milling

References

(LiAlH4 þ 5 wt% n-Ni) (LiAlH4 þ 5 wt% n-Fe) (LiNH4 þ 5 wt% n-TiC/n-TiN/n-ZrC)

Small Excessive None

No No No

No Yes Yes

[29] [30] [31]

Hydrogen desorbed (wt%)

9 Ball milled 8 Dehydrogenation 100°C 7 1 6 3 4 5 4 1-n-TiC 3 2-n-TiN 2 3-n-ZrC 4-n-Ni 1 5-n-Fe 0 0 5 10 Time (h) (A)

2 5

15

9 8 7 6 5 4 3 2 1 0

n-Fe-BM n-Fe-BM n-Ni-BM ~5 Wt% H2

n-Ni-desorbed at 250°C n-Fe-desorbed at 140°C 0

20 (B)

Hydrogen desorbed 200° C (wt %)

LiAlH4 lattice parameter change

Hydrogen desorbed (wt%)

LiAlH4 system

10 20 30 40 50 60 70 80 Time at 40°C (days) (C)

9 8 7 6 5 4 3 2 1 0

LiAIH4+5 wt% n-TiC LiAIH4+5 wt% n-TiN LiAIH4+5 wt% n-ZnC

~6 Wt% H2

0

5 10 15 20 25 30 35 40 Time at 40°C (days)

Fig. 5 (A) Thermal H2 desorption curves at 1001C under 1 bar H2 pressure for LiAlH4 BM with 5 wt% of nano additives. Slow desorption of LiAlH4 BM with 5 wt% (B) nanometallic and (C) nano-interstitial compound additives. BM, ball milled. Adapted from Varin RA, Zbroniec L. Decomposition behavior of unmilled and ball milled lithium alanate (LiAlH4) including long-term storage and moisture effects. J Alloys Compd 2010;504:89–101; Varin RA, Zbroniec L. The effects of nanometric nickel (n-Ni) catalyst on the dehydrogenation and rehydrogenation behavior of ball milled lithium alanate (LiAlH4). J Alloys Compd 2010;506:928–39; Varin RA, Parviz R. The effects of the micrometric and nanometric iron (Fe) additives on the mechanical and thermal dehydrogenation of lithium alanate (LiAlH4), its selfdischarge at low temperatures and rehydrogenation. Int J Hydrog Energy 2012;37:9088–102; and Varin RA, Parviz R. The effects of the nanometric interstitial compounds TiC, ZrC and TiN on the mechanical and thermal dehydrogenation and rehydrogenation of the nanocomposite lithium alanate (LiAlH4) hydride. Int J Hydrog Energy 2014;39:2575–86.

As shown in Fig. 5(A) the reduction of thermal dehydrogenation temperature for LiAlH4, which is accompanied by increasing dehydrogenation rate, can be achieved by incorporating nano additives. In general, isothermal decomposition of LiAlH4 occurs in two stages: quite fast Stage I and much slower Stage II (Fig. 5(A)) which corresponds to reaction (2”b) and (2c), respectively. Those stages were confirmed by XRD after completion of desorption at each temperature [27,29–31]. It was found [29–31] that for thermal dehydrogenation in Stage I the average apparent activation energy, EA, for the nano-interstitial compound additives, n-TiC/n-TiN/n-ZrC, was within the range of 87–96 kJ/mol whereas, for comparison, the nanometric metallic additives, n-Fe and n-Ni, exhibited drastically smaller apparent activation energy on the order of 55–70 kJ/mol. The average apparent activation energy for thermal dehydrogenation in Stage II was in the range of 63–80 kJ/mol in the order of EA(n-ZrC)oEA(n-Ti ¼ n-TiC) and was lower than that for the nanometric metal additives n-Ni and n-Fe. In summary, the nanometric interstitial compounds did not substantially affect the apparent activation energy of Stage I but were able to reduce the apparent activation energy of thermal dehydrogenation in Stage II. Fig. 5(B) and (C) show that ball milled nano-additive bearing LiAlH4 is capable of slowly desorbing H2 at a low temperature of barely 401C. The nano-interstitial additives (n-TiN, n-TiC, and n-ZrC) seem to be slightly more effective in inducing lowtemperature H2 desorption because after 20 days the desorbed H2 quantity is about 6 wt%, as compared to about 5 wt% H2, desorbed after the same number of days from LiAlH4 containing nano-metal (n-Ni and n-Fe) additives. The slow desorption phenomenon at such a low temperature makes LiAlH4 with nano-additives a nearly ideal hydride for filling in disposable cartridges for long-term, low-volume applications. The PCT equilibrium plateau pressure for the decomposition of LiAlH4 has been reported to be very high which makes LiAlH4 irreversible under practical conditions of temperature/pressure. The plateau pressure for Stage I dehydrogenation reaction of LiAlD4, containing a catalytic precursor TiF3, into Li3AlD6, Al and H2 ((2’b) or (2”b)) was reported by Brinks et al. [38] to be higher than 99 and 87 bar at 53 and 801C, respectively. Furthermore, Mulana and Nishimiya [39] estimated the enthalpy (DH) and entropy (DS) for Stage I dehydrogenation ((2’b) or (2”b)) and the second stage dehydrogenation reaction (2c) (Stage II) as being equal to DH¼ 17.5 kJ/molH2 and DS¼121.6 J/molH2 K, and DH¼ 11.1 kJ/molH2 and DS¼ 62.6 J/molH2 K, respectively. Assuming DH ¼17.5 kJ/molH2 and DS ¼ 121.6 J/molH2 K, the equilibrium pressure for Stage I dehydrogenation, calculated from Eq. (1), would amount to 1.88  103 and 19.6  103 atm. at room temperature (241C ¼297K) and 1701C (443K), respectively. Apparently, the first dehydrogenation reaction appears to be completely irreversible due to extremely high pressures required for rehydrogenation. Assuming DH ¼ 11.5 kJ/molH2 and DS¼ 62.6 J/molH2 K for Stage II in Eq. (1), it is calculated that at 1701C the

580

Hydrides

equilibrium H2 pressure is around 82 atm. In contrast, the computed stability diagrams for LiAlH4/Li3AlH6/LiH show very high pressures, on the order of 103 atm. at 1701C, needed for the rehydrogenation of LiH/Al into Li3AlH6 [19,40,41].

2.18.3.2

Metal Borohydrides

2.18.3.2.1

The (LiBH4/NaBH4-MnCl2) systems

Metal borohydrides are very interesting materials for potential solid state hydrogen generation/storage due to their very high theoretical capacities of hydrogen [19,24,42–46]. Most of borohydrides crystallize in complex lattices that can be found in Ref. [46]. Two notable borohydrides, sodium borohydride (NaBH4) and lithium borohydride (LiBH4), are easily commercially available with a theoretical gravimetric hydrogen capacity of 10.6 and 18.4 wt%, respectively. They also have quite high volumetric densities of hydrogen [44]. Unfortunately, their relatively high enthalpy change for dehydrogenation (Eq. (1)) results in the dehydrogenation temperatures in excess of 4001C [44–46]. Metal catalysts, such as Ni [47], nonmetal catalysts (e.g., fullerenes [48]) and metal chloride catalytic precursors (FeCl2, CoCl2, and NiCl2 [49]), accelerate the dehydrogenation rate of LiBH4 but do not change its unfavorable thermodynamics and in certain cases led to the formation of diborane gas (B2H6). It has been proposed [49] that the stronger the binding between the metal and boron, the less possible diborane was produced. Nakamori et al. [50,51] in their seminal publications proposed that the unfavorable thermodynamics of LiBH4 could be overcome by converting it to another borohydride with better thermodynamics. They employed mechanochemical activation synthesis (MCAS) by BM of complex hydrides, LiBH4 and NaBH4, mixed with appropriate metal di-or tri-chlorides, MCln, for synthesizing a large spectrum of new metal borohydrides according to the following reactions: MCln þ nLiBH4 -MðBH4 Þn þ nLiCl

ð4Þ

MCln þ nNaBH4 -MðBH4 Þn þ nNaCl

ð5Þ

The reaction which occurs during MCAS is a “metathesis reaction” which occurs in solid state similar to a reaction which uses diethyl ether as a solvent [22]. Nakamori et al. [50,51] reported that for both LiBH4 or NaBH4, the XRD patterns after completion of MCAS, showed the absence of the diffraction peaks of crystalline M(BH4)n which indicated a formation of amorphous hydride. In contrast, the Raman spectra confirmed the presence of M(BH4)n after MCAS. Furthermore, they reported [50,51] that only the LiCl peaks were observed when LiBH4 was used as a reactant while when NaBH4 was used with MCln (M ¼ Cr, Mn, Ti, V, Zn) then the NaCl peaks were observed shifted to lower diffraction angles. For the NaBH4 and MCln (M ¼ Ca, Sc, and Al) systems the diffraction peaks of NaBH4 were still present after BM. Nakamori et al. [50,51] concluded that reaction (4) with LiBH4 was much easier than reaction (5) with NaBH4 due to similar ionic radii of Li þ (0.076 nm) and Mn2 þ (0.067 nm) in solid-solid cation exchange reaction as compared to that of Na þ (0.102 nm). They also reported that the hydrogen desorption temperature, Td, of disordered M(BH4)n decreased with increasing values of the Pauling electronegativity χP of M in borohydride. They noted that the desorbed gas for M ¼ Ca, Sc, Ti, V and Cr (χPr1.5) was hydrogen only, while that for M ¼ Mn, Zn and Al (χPZ1.5) contained a mix of diborane and hydrogen. Later, Choudhury et al. [52] reported that mechanochemical synthesis of LiBH4 with MnCl2 in the ratio of 3:1 resulted in the synthesis of new complex metal borohydride LiMn(BH4)3. Following Choudhury et al. [52], we also attempted to synthesize LiMn (BH4)3 using a (3LiBH4 þ MnCl2) mixture with a presumption that LiMn(BH4)3 would indeed be formed [53]. However, it was firmly established by Cerny et al. [54,55] and Severa et al. [56] that Mn(BH4)2 was always synthesized during MCAS of the LiBH4 and MnCl2 constituents and not LiMn(BH4)3. The theoretical maximum H2 capacity of Mn(BH4)2 is quite high being close to 9.5 wt% which makes it a very interesting borohydride for hydrogen generation/storage. Liu et al. [57] reported that dehydrogenation of Mn(BH4)2 occurred with the release of diborane (B2H6). However, they did not conduct any isothermal dehydrogenation tests at 1 bar H2 pressure as required by the Van’t Hoff law (Eq. (1)) for supplying H2 to a FC stack. In order to study the details of the synthesis of Mn(BH4)2 and to understand thoroughly its dehydrogenation behavior, we carried out studies for a wide range of molar ratios, n, in the mixture (nLiBH4 þ MnCl2) where n¼ 1, 2, 3, 5, 9, and 23 [58] and later we focused on n¼ 2, 3 [59] and the effect of milling energy input which was calculated from a semi-empirical model as reported in Ref. [35]. We observed a high stability of the (nLiBH4 þ MnCl2; n¼ 2, 3) mixtures since their mechanical dehydrogenation was limited to 0.3 wt% H2 (n¼ 2) and 0.2 wt% H2 (n¼ 3) up to 2 h of BM. For the n ¼2 nanocomposite there was a gradual acceleration of mechanical dehydrogenation to the extent that the quantity of desorbed H2 increases from about 0.3 after 2 h to 0.7 wt% after 5 h of BM [59]. A typical XRD pattern after mechanochemical synthesis of the (nLiBH4 þ MnCl2; n¼2, 3) mixtures is shown in Fig. 6 [59]. Only peaks of LiCl and Mn(BH4)2 are observed confirming that the following metathesis reaction occurred during milling [58,59]. 2LiBH4 þ MnCl2 -MnðBH4 Þ2 þ 2LiCl

ð6Þ

With increasing molar ratio n Z2 the reaction is modified to the following: nLiBH4 þ MnCl2 -MnðBH4 Þ2 þ 2LiCl þ ðn

2ÞLiBH4 ðn ¼ 3 to 23Þ

ð7Þ

Hydrides

581

(nLiBH4+MnCl2) (BM; QTR=145.6kJ/g)

2000

LiCl (74-1972) LiAIH4 (Standard)

Counts

1600

Mn(BH4)2 n=3

1200

800

n=2

400

0 20

30

40

50 60 Degree 2-theta

70

80

90

Fig. 6 X-ray diffraction patterns after ball milling (BM) with an energy input, QTR ¼145.6 kJ/g [35], for the (nLiBH4 þ MnCl2; n¼2 and 3) mixtures. Adapted from Varin RA, Shirani Bidabadi AR. The effect of milling energy input during mechano-chemical activation synthesis (MCAS) of the nanocrystalline manganese borohydride (Mn(BH4)2) on its thermal dehydrogenation properties. Int J Hydrog Energy 2014;39:11620–32. The diffraction peaks of Mn(BH4)2 were indexed after data in Černý R, Penin N, Hagemann H, Filinchuk Y. The first crystallographic and spectroscopic characterization of a 3d-metal borohydride: Mn(BH4)2. J Phys Chem C 2009;113:9003–7; Černý R, Penin N, D’Anna V, Hagemann H, Durand E, Růžička J. MgxMn(1 x)(BH4)2 (x¼0-0.8), a cation solid solution in a bimetallic borohydride. Acta Mater 2011;59:5171–80; Severa G, Hagemann H, Longhini M, Kaminski JW, Wesolowski TA, Jensen CM. J Phys Chem C 2010;114:15516–21; and Varin RA, Zbroniec L, Polanski M, Filinchuk Y, Černý R. Mechano-chemical synthesis of manganese borohydride (Mn(BH4)2) and inverse cubic spinel (Li2MnCl4) in the (nLiBH4 þ MnCl2) system and its dehydrogenation behavior. Int J Hydrog Energy 2012;37:16056–69.

H2/B2H6=4667 (up to 200°C) 5E−05

1.0E−02 2E−05

8.0E−00 6.0E−03

1E−05

4.0E−03 2.0E−03

−2 −1.0 −4 Endo

−1.5

−6 −2.0 −8

−2.5

0.1 0.0 −0.1 −0.2

Heat flow (mW/mg)

3E−05

1.2E−02

0.2 −0.5

Mass change diff. (%/min)

1.4E−02

0.0E+03

(A)

4E−05

1.6E−02

0.3 0.0

0

Mass change (wt%)

1.8E−02

Diborane relative partial pressure (μTorr/mg)

Hydrogen relative partial pressure (μTorr/mg)

2.0E−02

−0.3 −0.4

0E+00 50 100 150 200 250 300 350 400 450 500

Temperature (°C)

−3.0

−10 100

(B)

200

300

400

500

Sample temperature (°C)

Fig. 7 (A) Mass spectrometry (MS) gas desorption spectra and (B) Thermogravimetric and differential scanning calorimetry curves for a synthesized ((Mn(BH4)2)/LiCl) þ 5 wt% Ni) sample. Adapted from Shirani Bidabadi AR, Varin RA, Polanski M, Biglari M, Stobinski L. The effects of filamentary Ni, graphene and lithium amide (LiNH2) additives on the mechano-chemical synthesis of crystalline manganese borohydride (Mn(BH4)2) and its solvent extraction. Mater Res Bull (submitted).

A special case occurred for n ¼ 1 where after the MCAS synthesis instead of LiCl an inverse cubic spinel ionic conducting compound Li2MnCl4 was formed [58] according to the reaction: LiBH4 þ MnCl2 -0:5MnðBH4 Þ2 þ 0:5Li2 MnCl4

ð8Þ

The same compound Li2MnCl4 was found in the LiAlH4–MnCl2 system [22]. Apparently, whether the reacting hydride is lithium alanate (LiAlH4) or lithium borohydride (LiBH4) at the molar ratio n ¼1 in the ball milled mixture with MnCl2 there is always formation of an inverse cubic spinel Li2MnCl4 [22]. Formation of a very similar inverse cubic spinel Li2MgCl4 was also reported for the LiBH4–MgCl2 system after BM for 36 h in a planetary ball mill [60]. This compound was formed together with the Mg(BH4)2 hydride as a result of metathesis reaction. However, we carried out similar synthesis for the (3LiBH4 þ MgCl2) mixture but instead of Li2MgCl4 the regular salt LiCl was formed [22].

Hydrides

582

The XRD studies showed that the crystallite (grain) size of the synthesized nanocrystalline Mn(BH4)2 hydride attains 2175.0 nm for the energy input QTR ¼ 36.4 kJ/g and then it is further reduced to 1871.0 nm for QTR ¼ 145.6 kJ/g and finally to 1470.5 nm for QTR ¼ 364 kJ/g [59]. The crystallite (grain) size of LiCl was close to 30 nm [59] regardless of the milling energy input, QTR [35]. TEM selected area electron diffraction patterns (SAEDPs) clearly confirmed the presence of the Mn(BH4)2 and LiCl phases in the synthesized nanocomposite [61]. No other phases were detected. Bright field high-resolution TEM imaging of the synthesized composite powder particles revealed the presence of nanograins consistent with LiCl and Mn(BH4)2 within the powder particles. Their respective grain sizes, estimated as the equivalent circle diameters (ECD) from the high-resolution TEM micrographs, with the corrected sample standard deviations, were within the range of 14.173.7 nm and 10.072.9 nm for LiCl and Mn(BH4)2, respectively. The XRD patterns of the thermally dehydrogenated (Mn(BH4)2 þ 2LiCl) nanocomposite did not exhibit any Bragg diffraction peaks belonging to either crystalline Mn or B. In contrast, the SAED patterns and energy-dispersive X-ray spectroscopy (EDS) elemental maps provided strong evidence that both Mn and B existed in the dehydrogenated powder as

5

5 100°C Hydrogen desorbed (wt%)

Hydrogen desorbed (wt%)

100°C 4 3 2 (Mn(BH4)2/LiCl)

1

(QTR=36.4 kJ/g)

0

4 3 2 (Mn(BH4)2/LiCl)

1

(QTR=145.6 kJ/g)

0 0

4

(A)

8

12 Time (h)

16

20

0

2

4

(B)

6

8 10 Time (h)

12

14

16

20

Fig. 8 Dehydrogenation curves at 1001C for the Mn(BH4)2/LiCl nanocomposite mixture synthesized by ball milling (BM) from (2LiBH4 þ MnCl2) with the energy input (A) QTR ¼36.4 kJ/g (0.5 h BM) and (B) QTR ¼145.6 kJ/g (1 h BM). Adapted from Varin RA, Shirani Bidabadi AR. The effect of milling energy input during mechano-chemical activation synthesis (MCAS) of the nanocrystalline manganese borohydride (Mn(BH4)2) on its thermal dehydrogenation properties. Int J Hydrog Energy 2014;39:11620–32.

LiCl (74−1972) [(Li(Et2O)2)Mn2(BH4)5] (F) (Mn(BH4)2/LiCl)-1st EXT 42°C-des.100°C/21.3 h

intensity (a.u.)

(E) (Mn(BH4)2/LiCl)-powder in the filter after 1st EXT 42°C (D) (Mn(BH4)2/LiCl)-2nd EXT 42°C (C) (Mn(BH4)2/LiCl)-1st EXT 42°C (B) (Mn(BH4)2/LiCl)-2nd EXT RT (A) (Mn(BH4)2/LiCl)-1st EXT RT

20

30

40

50 Degree 2-theta

60

70

80

90

Fig. 9 X-ray diffraction patterns of synthesized (Mn(BH4)2)/LiCl) after (A) first solvent extraction at room temperature (RT) in diethyl ether (Et2O), (B) second solvent extraction at RT, (C) first solvent extraction (1st EXT) at 421C, (D) second solvent extraction (2nd EXT) at 421C, (E) powder in the filter after first solvent extraction, and (F) after isothermal dehydrogenation at 1001C of the first solvent extracted sample (1st EXT) at 421C. Adapted from Shirani Bidabadi AR, Varin RA, Polanski M, Biglari M, Stobinski L. The effects of filamentary Ni, graphene and lithium amide (LiNH2) additives on the mechano-chemical synthesis of crystalline manganese borohydride (Mn(BH4)2) and its solvent extraction. Mater Res Bull (submitted).

Hydrides

583

1.0E−07

2.0E−09

5.0E−08

0.0E+00

0.0E+00

5.0E−07

8.0E−09 6.0E−09

3.5E−07 3.0E−07 2.5E−07 2.0E−07 1.5E−07 1.0E−08 0.5E−08

2.0E−09 0.0E+00

(C)

0.0E+00 50 100 150 200 250 300 350 400 450 500 Temperature (°C)

2 3

−40

(B)

4.5E−07 4.0E−07

4.0E−09

−3

−30

0

50 100 150 200 250 300 350 400 450 500 Temperature (°C) [{Li(Et2O)2}Mn2(BH4)5]/LiCl-1st EXT

−2

endo

−50 100 200 300 400 500 Sample temperature (°C)

−4 −5 600

1 0

0 Mass change (wt%)

1.0E−08

−20

−1

−1

1

−10

−2 −3

−20

−4 −30

−5

endo

−6

−40

2

−7

−50

−8 0

(D)

100 200 300 400 500 Sample temperature (°C)

600

0.0 −0.2 −0.4 −0.6 −0.8

Heat flow (mW/mg)

4.0E−09

1

−10

−1.0

0.2 0.0 −0.2 −0.4 −0.6

Heat flow (mW/mg)

1.5E−07

0

Mass change diff. (%/min)

6.0E−09

0.2 0

Mass change diff. (%/min)

2.0E−07

Mass change (wt%)

2.5E−07

8.0E−09

(A) Diborane partial pressure (Torr)

[{Li(Et2O)2}Mn2(BH4)5]/LiCl-1st EXT

Hydrogen partial pressure (Torr)

Diborane partial pressure (Torr)

1.0E−08

Hydrogen partial pressure (Torr)

crystalline phases, a-Mn and b-B, respectively. The results showed that the lack of XRD Bragg diffraction peaks was insufficient evidence that the Mn and B elemental products of Mn(BH4)2 thermolysis could be classified as being amorphous [61]. Gas mass spectrometry (MS) during temperature programmed desorption (TPD) up to 4501C showed the release of hydrogen as principal gas with a maximum intensity around 130–1501C accompanied by a miniscule quantity of borane B2H6. The intensity of the B2H6 peak was 200–600 times smaller than the intensity of the corresponding H2 peak [58]. Very recently, we found for the first time, that during isothermal dehydrogenation up to 2001C of the synthesized (Mn(BH4)2/LiCl) mixture with filamentary Ni, LiNH2, and graphene (reduced graphene oxide (rGO)), each of additives reduced the released quantity of B2H6, increasing the released H2 gas intensity ratio, H2/B2H6, from 493 for an additive-free sample, to 4667 for 5 wt% filamentary Ni (Fig. 7), to 3213 for 5 wt% LiNH2 and to 722 for 5 wt.% graphene. Apparently, the filamentary Ni additive is the most effective suppressor of B2H6 released from crystalline Mn(BH4)2 [62]. In a DSC during continuous heating, Mn(BH4)2 usually decomposes in an endothermic event (Fig. 7). However, the DSC curve of the BM ((Mn(BH4)2/LiCl) þ 5 wt% LiNH2) sample showed a pronounced exothermic peak with a maximum at 1201C preceding the characteristic endothermic peak with a maximum at 1331C corresponding to the decomposition of Mn(BH4)2 [62]. The synthesized nanocrystalline Mn(BH4)2 hydride, coexisting with a nanocrystalline LiCl salt, is capable of desorbing up to B4.5 wt% at 1001C (Fig. 8) [59]. The values of the apparent activation energy for dehydrogenation obtained in Ref. [59] were very low. The apparent activation energy for the n¼ 3 nanocomposite decreased monotonically from B70 to B59 kJ/mol with increasing milling energy input whereas the apparent activation energy for the n¼2 nanocomposite decreased from about 65 kJ/mol for QTR ¼ 36.4 kJ/g to about 53 kJ/mol for QTR ¼ 145.6 kJ/g and then again increased to B59 kJ/mol for the QTR ¼364 kJ/g. These changes closely followed the variations in the average powder particle size obtained with the varying milling energy input. For the milling energy input QTR ¼36.4 and 145.6 kJ/g the average powder particle size decreased to 14.976.6 and 7.572.6 mm, respectively, and subsequently increased reaching the average size of 16.176.3 mm for the milling energy input QTR ¼364 kJ/g. On the other hand, the apparent activation energy for dehydrogenation did not exhibit any dependence on the average crystallite (grain) size. The LiCl salt that is formed during mechanochemical synthesis according to Eq. (6) is a “dead weight” that reduces the gravimetric H2 capacity of the Mn(BH4)2/LiCl mixture. For the first time, we attempted to remove LiCl by solvent filtration and extraction in diethyl ether (Et2O) [62]. The most striking result of the solvent extracted sample was that a dimetallic borohydride solvate [{Li(Et2O)2}Mn2(BH4)5] crystallized during extraction instead of expected crystalline c-Mn(BH4)2. The new [{Li(Et2O)2} Mn2(BH4)5] borohydride solvate was identified from the respective XRD patterns (Fig. 9(C) and (D)), based on the data reported in

−0.8 −1.0

Fig. 10 ((A) and (C)) Mass spectrometry gas desorption spectra (not normalized by the mass of the powder sample) and ((B) and (D)) Thermogravimetric and differential scanning calorimetry curves for ([{Li(Et2O)2}Mn2(BH4)5]/LiCl) obtained after first extraction (1st EXT) at 421C. Adapted from Shirani Bidabadi AR, Varin RA, Polanski M, Biglari M, Stobinski L. The effects of filamentary Ni, graphene and lithium amide (LiNH2) additives on the mechano-chemical synthesis of crystalline manganese borohydride (Mn(BH4)2) and its solvent extraction. Mater Res Bull (submitted).

584

Hydrides

Ref. [63], which show that it has a monoclinic space group C2/c (see Table 2 in Ref. [63]). After isothermal desorption at 1001C for 21.3 h, the [{Li(Et2O)2}Mn2(BH4)5] peaks that formed due to the solvent extraction completely disappeared, indicating that the newly formed hydride is fully decomposed (Fig. 9(F)). Fig. 10(A)–(D) shows the MS and thermogravimetric analysis (TGA)/DSC curves, respectively, for two repeated samples without additive, containing [{Li(Et2O)2}Mn2(BH4)5] as a principal hydride phase and some retained LiCl (Fig. 9(A)–(D)), which were obtained after 1st EXT. The MS curve for the first sample in Fig. 10(A) shows a narrow H2 peak doublet while the second sample in Fig. 10(C) shows a single, smooth, H2 peak. Both samples also show peaks of released B2H6. Remarkably, the calculated MS peak area intensity ratio H2/B2H6 is nearly identical for both samples being equal to 147 and 148 for samples in Fig. 10(A) and (C), respectively. This ratio is lower than that for the additive-free, crystalline, Mn(BH4)2 sample [58] which indicates a slightly higher quantity of B2H6 released from solvent extracted ([{Li(Et2O)2}Mn2(BH4)5]) than that released from mechanochemically synthesized crystalline Mn(BH4)2 in dynamic decomposition experiments, in a TG apparatus under a continuous flow of helium gas. The DSC curves in Fig. 10(B) and (D) show endothermic doublet peak for both samples with the respective DSC peaks at 100 and 1501C. A first derivative of the TG line (dTG/dT-broken red line) confirms the exact positions of DSC peaks. The TG analysis shows that the first sample exhibits at least three distinctive mass loss steps designated 1, 2, and 3 in Fig. 10(B). In step 1, the mass loss reaches about 39% up to 1501C. In the following steps 2 and 3 up to 3001C, the mass loss amounts to about 6%. The TG curve for the second sample in Fig. 10(D) contains only two steps of mass loss where in step 1 the mass loss reaches about 40% up to 1501C and about 5% in step 2 up to 3001C. In the middle of step 1 there seems to appear a slight inflection point on the TG line (at B1201C) which might be related to a beginning of H2 desorption, simultaneous with decomposition of Et2O.

(200)

(2NaBH4+MnCl2) NaCl (05-0628) NaBH4 (09-0386)

Intensity (au)

5 h BM (QTR=364 kJ/g)

1 h BM (QTR=72.8 kJ/g)

0.5 h BM (QTR=36.4 kJ/g)

20

30

40

(A)

50 60 Degrees 2-theta

70

(200)

80

90

0.5-0624 make,sys

(2NaBH4+MnCl2)

Intensity (au)

5 h BM (QTR=364 kJ/g) 1 h BM (QTR=72.8 kJ/g) 0.5 h BM (QTR=36.4 kJ/g)

30.5 (B)

31

31.5

32 32.5 Degrees 2-theta

33

33.5

34

Fig. 11 (A) X-ray diffraction patterns for the (2NaBH4 þ MnCl2) mixture ball milled for 0.5 h (QTR ¼36.4 kJ/g), 1 h (QTR ¼72.8 kJ/g) and 5 h (QTR ¼364 kJ/g). (B) Enlargement of the (200) NaCl peak (100% intensity) in comparison to the (200) standard diffraction line from the ICDD PDF#(05–0628). Adapted from Varin RA, Mattar DK, Shirani Bidabadi AR, Polanski M. Synthesis of amorphous manganese borohydride in the (NaBH4–MnCl2) system, its hydrogen generation properties and crystalline transformation during solvent extraction. J Energy Chem 2017;26:24–34.

Hydrides

585

2359

55 50

1158

60

Amorphous borohydride (2NaBH4+MnCl2) 2272

65

1227

Tumanov et al. [63] synthesized [{Li(Et2O)2}Mn2(BH4)5], accompanied by LiCl, from the wet metathesis reaction between MnCl2 and LiBH4 in Et2O. They studied its thermal properties during dynamic decomposition in a TG apparatus under argon atmosphere. They found that [{Li(Et2O)2}Mn2(BH4)5] decomposed in three defined steps. The first step up to about 1101C, exhibiting a mass loss of B26.8% was assigned to the evolution of Et2O while the second mass loss of B5.2% up to about 1601C, was assigned to the decomposition of Mn(BH4)2. The nature of the third step up to B3251C was not explained by the authors.

35

1104

40 1353

Kubuda (min)

45

30 25 20

3510

10

2564

15

5 3500

3000

2500

2000

1500

1000

Wavenumber (cm−1)

4.5

c-Mn(BH4)2

4.0

(2LiBH4+MnCl2)

2274

(A)

1229

4000

3.5 3.0

2.0

1620

0.0

2562

0.5

1352

1.0

1113

1.5

3530

Kubuda (min)

2.5

−0.5 −1.0 −1.5 4000 (B)

3500

3000

2500

2000

1500

1000

Wavenumber (cm−1)

Fig. 12 (A) Fourier-transform infrared (FT-IR) spectrum for a (2NaBH4 þ MnCl2) sample ball milled for 5 h (QTR ¼ 364 kJ/g). (B) Reference FT-IR spectrum for a (2LiBH4 þ MnCl2) sample ball milled for 2 h (QTR ¼145.6 kJ/g) containing a synthesized, crystalline Mn(BH4)2 hydride. Reproduced from Varin RA, Mattar DK, Shirani Bidabadi AR, Polanski M. Synthesis of amorphous manganese borohydride in the (NaBH4–MnCl2) system, its hydrogen generation properties and crystalline transformation during solvent extraction. J Energy Chem 2017;26:24–34.

586

Hydrides

It is clear that our TG results in Fig. 10(B) and (D) are in disagreement with those reported by Tumanov et al. [63]. One of the discrepancies is the quantity of mass loss observed in step 1 in Fig. 10(B) and (D) which amounts to B40% versus only B27% reported by Tumanov et al. [63]. It can be calculated that the theoretical capacity of Et2O in [{Li(Et2O)2}Mn2(BH4)5] is 43.7 wt% and the theoretical capacity of H2 is B5.9 wt% (the practical values are slightly lower due to the presence of LiCl, coexisting with [{Li(Et2O)2}Mn2(BH4)5] after solvent extraction, as can be seen in Fig. 9). As such, the observed mass loss of B40% in step 1 in Fig. 10(B) and (D) agrees quite well with a theoretical capacity of Et2O in [{Li(Et2O)2}Mn2(BH4)5], corrected for the presence of LiCl. Furthermore, the mass loss in step 2 and 3 in Fig. 10(B) and step 2 in Fig. 10(D), can be assigned to the desorption of H2 (mixed with B2H6) from [{Li(Et2O)2}Mn2(BH4)5] according to a hypothetical decomposition reaction:
   LiðEt2 OÞ2 Mn2 ðBH4 Þ5 þ ½retained LiClŠ-Li þ 2Et2 O þ 2Mn þ xB2 H6 þ yB þ zH2 þ ½retained LiClŠ ð9Þ

(2NaBH4+MnCl2) 0.5 h BM H2/B2H6=948 (up to 200°C)

5.0E−03

3E−05

4.0E−03 2E−05

3.0E−03 2.0E−03

1E−05

1.0E−03 0E+00

0.0E+03 50 100 150 200 250 300 350 400 450 500

Temperature (°C)

(A)

0.2

0.0

−0.2

−0.4

−0.6 Endo −0.8

0.1 0.0 −0.1 −0.2 −0.3

Heat flow (mW/mg)

4E−05 6.0E−03

−0.5 −1.0 −1.5 −2.0 −2.5 −3.0 −3.5 −4.0 −4.5 −5.0 −5.5 −6.0 −6.5

Mass change diff. (%/ min)

7.0E−03

(2NaBH4+MnCl2) 0.5 h BM

0

Mass change (wt%)

5E−05

8.0E−03

Dibrane relative partial pressure (μTorr/mg)

Hydrogen relative partial pressure (μTorr/mg)

We also conducted studies of the BM behavior and subsequent dehydrogenation of the (2NaBH4 þ MnCl2) system [64]. The mixture of (2NaBH4 þ MnCl2) was ball milled in a magneto-mill for varying time durations (varying milling energy input) [64].

−0.4 −1.0 50

−0.5

100 150 200 250 300 340 400 450 500

(B)

Sample temperature (°C)

Fig. 13 (A) Mass spectrometry (MS) gas desorption spectra for a (2NaBH4 þ MnCl2) sample ball milled for 0.5 h (synthesized a-Mn(BH4)(2x)/NaCl mixture). (B) Thermogravimetric and differential scanning calorimetry curves. Heating rate 51C/min. BM, ball milled. Reproduced from Varin RA, Mattar DK, Shirani Bidabadi AR, Polanski M. Synthesis of amorphous manganese borohydride in the (NaBH4–MnCl2) system, its hydrogen generation properties and crystalline transformation during solvent extraction. J Energy Chem 2017;26:24–34.

NaCl (05-0628) Mn(BH4)2 NaBH4 (09-0386) LiCl (04-0664)

Intensity (au)

(2LiBH4+MnCl2) 0.5 h BM

(2NaBH4+MnCl2) 0.5 h BM and extracted at 42°C-20 min

(2NaBH4+MnCl2) 0.5 h BM

20

30

40

50 2 Theta (degree)

60

70

80

Fig. 14 X-ray diffraction (XRD) patterns of the 0.5 h BM sample, solvent extracted at 421C for 20 min (center) in comparison to the XRD patterns of the 0.5 h BM (2NaBH4 þ MnCl2) (bottom) and (2LiBH4 þ MnCl2) (top) samples. BM, ball milled. Reproduced from Varin RA, Mattar DK, Polanski M, Shirani Bidabadi AR, Stobinski L. The effects of additives on the dehydrogenation of amorphous manganese borohydride and its crystalline form after solvent filtration/extraction. Energies 2017;10:1741.

Hydrides

1.6E−04

6.0E−03 1.4E−04 5.0E−03

1.2E−04 1.0E−04

4.0E−03

8.0E−05

3.0E−03

6.0E−05 2.0E−03 4.0E−05 1.0E−03

2.0E−05

0.0E+03

0.0E+00 50 100 150 200 250 300 350 400 450 500

200

0.0

180

−2 Endo

−0.4 −0.6

−4

−0.8 −6

−1.0 −1.2

−8

−10

Temperature (°C)

140 120 100 80 60 40 20

−1.4

0

−1.6

−20

−1.8

(A)

160

−0.2

−12

Heat flow (mW/g)

H2/B2H6 ratio = 1626

0.2

Mass change diff. (%/min)

7.0E−03

0

Mass change (wt%)

Hydrogen relative partial pressure (μTorr/mg)

1.8E−04

Dibrane relative partial pressure (μTorr/mg)

2.0E−04

8.0E−03

587

−40 −60

100

200

(B)

300

400

500

Sample temperature (°C)

Fig. 15 Mass spectrometry (MS) of gas released from extracted Mn(BH4)2 during combined MS/thermogravimetric analysis (TGA)/differential scanning calorimetry (DSC). Reproduced from Varin RA, Mattar DK, Polanski M, Shirani Bidabadi AR, Stobinski L. The effects of additives on the dehydrogenation of amorphous manganese borohydride and its crystalline form after solvent filtration/extraction. Energies 2017;10:1741.

2000

(2NaBH4+MnCl2)

NaCl (05-0628) Mn(BH4)2 NaBH4 (09-0386)

1600

Counts

(2NaBH4+MnCl2) −1 h BM and extracted at 42°C-25 min 1200

800 1 h BM 400

0 20

30

(A)

40

50 60 Degree 2-theta

8 100 °C 6 4 2

1 h BM (QTR=72.8 kJ/g) EXT 42°C

0 (B)

80

10 1 h BM (QTR=72.8 kJ/g) EXT 42°C

Hydrogen desorbed (wt%)

Hydrogen desorbed (wt%)

10

70

8 120°C 6 4 2 0

0 10 20 30 40 50 60 70 80 90 100 Time (h)

0 (C)

8

16 24 32 40 48 56 64 72 Time (h)

Fig. 16 (A) X-ray diffraction (XRD) patterns for a-Mn(BH4)2 after solvent filtration/extraction at 421C, compared to the XRD after ball milling for 1 h. Desorption curves for a solvent extracted c-Mn(BH4)2 at (B) 1001C and (C) 1201C. Adapted from Varin RA, Mattar DK, Polanski M, Shirani Bidabadi AR, Stobinski L. The effects of additives on the dehydrogenation of amorphous manganese borohydride and its crystalline form after solvent filtration/extraction. Energies 2017;10:1741.

588

Hydrides

No gas release was detected during BM. The XRD patterns of the ball milled mixture exhibited only the Bragg diffraction peaks of the NaCl-type salt (Fig. 11(A)). Fig. 11(B) shows the enlargement of the (200) NaCl peak (100% intensity) in comparison to the (200) standard diffraction line from the ICDD PDF#(05–0628). The peak is heavily broadened indicating the formation of nanograins within the NaCl particles or, alternatively, formation of nanosized, monocrystalline NaCl particles as a fraction of the total particle population. Furthermore, the peak maximum is always shifted from the (200) NaCl standard diffraction position given by the ICDD PDF#(05–0628) file. This peculiar shift is not an artifact but has been observed after mechanochemical synthesis of every (2NaBH4 þ MnCl2) mixture we processed. Llamas-Jansa et al. [65] reported a formation of the Na(BH4)(1 x)Clx solid solutions possessing cubic NaCl-type structure after mechanochemical synthesis of NaBH4 with various metal chlorides. By analogy, based on the observed (200) NaCl diffraction peak shift (Fig. 11(B)), we postulated in Ref. [64] the formation of a solid solution of NaBH4 in NaCl with a possible stoichiometry of Na(Cl)x(BH4)(1 x) (0oxr1). As shown in Fig. 12, the Fourier transform infrared (FT-IR) spectrum of the ball milled (2NaBH4 þ MnCl2) was very similar to the FT-IR spectrum of crystalline manganese borohydride, c-Mn(BH4)2, synthesized by BM, which strongly suggested that the amorphous hydride mechanochemically synthesized during BM, could be an amorphous manganese borohydride. Combining this observation with the formation of the Na(Cl)x(BH4)(1 x) solid solution, the following mechanochemical reaction during BM was postulated in Ref. [64]: h i ð10Þ 2NaBH4 þ MnCl2 -BM-a‐MnðBH4 Þð2xÞ þ 2 NaðClÞx ðBH4 Þð1 xÞ þ ½ð2‐2xÞClŠ

H2 4.0E−03

3E−05 3.0E−03 2E−05 2.0E−03 1.0E−03 0.0E+00

(A)

4E−05

1E−05 B2H6

0E+00 50 100 150 200 250 300 350 400 450 500 Temperature (°C)

TG 0.0 0.2

0.0

−0.5 −1.0 −1.5

−0.2

DSC

−2.0 −2.5

−0.4

−3.0 −3.5 −4.0

0.1

0.0

Heat flow (mW/mg)

Hydrogen relative partial pressure (μTorr/mg)

5.0E−03

5E−05

Mass change diff. (%/min)

H2/B2H6=1844 to 200°C

Mass change (wt%)

6.0E−03

Diborane relative partial pressure (μTorr/mg)

where 0oxr1 and a-Mn(BH4)(2x) is an amorphous manganese borohydride mechanochemically synthesized during BM. For x¼ 1 reaction (10) is reduced to reaction (5) with a NaBH4 reactant for n ¼ 2. Owing to the presence of heavier Na, the theoretical H2 capacity of reaction (10) for the NaBH4 reactant is about 4.0 wt% at x¼ 1 which is 0.76 wt% H2 lower than that of reaction (6) for the LiBH4 precursor. For 0oxo1 Eq. (9) would require the amorphous manganese borohydride, a-Mn(BH4)(2x), to be nonstoichiometric. It is hypothesized that this non-stoichiometry could prevent attaining a stoichiometric, crystalline structure during a formation of the borohydride by MCAS during BM. This aspect requires more in depth studies possibly by neutron diffraction (PND) techniques as those reported by Llamas-Jansa et al. [65] in order to verify the value of x in Eq. (9). In addition, since an x in Eq. (9) is expected to be close to 1, the Cl quantity would be miniscule. Fig. 13(A) shows MS gas desorption spectra for a (2NaBH4 þ MnCl2) sample ball milled for 0.5 h which contains the synthesized a-Mn(BH4)(2x)/NaCl mixture. They show mainly hydrogen (H2) with a miniscule quantity of diborane gas, B2H6. Fig. 13(B) shows the TG and DSC curves confirming an endothermic character of desorption. In order to remove the NaCl salt from the synthesized a-Mn(BH4)(2x)/NaCl mixture, we carried out a solvent filtration/ extraction procedure [64] in diethyl ether (Et2O) at 421C. Several modifications with respect to the solvent filtration/extraction published in Ref. [64] were implemented [66]: the mass ratio of powder mixture to diethyl ether was increased from 1:4 to 1:28, the BM bottle in Ref. [64] was replaced with a flat bottom flask that was swirled, a rubber stopper was used when Et2O was dissolved and the extraction flask was heated at 421C for 25 min. Remarkably, the process of solvent filtration/extraction at 421C, resulted in the transformation of mechanochemically synthesized amorphous manganese borohydride to a nanostructured, crystalline, c-Mn(BH4)2 hydride (Fig. 14) [66]. The final yield of the crystalline c-Mn(BH4)2 hydride was 49.6% [66]. MS of gas released from the solvent filtrated/extracted sample during combined MS/TGA/DSC experiments, confirms mainly hydrogen (H2) with a negligible quantity of diborane gas, B2H6 (Fig. 15) [66]. The quantity of desorbed H2 from the solvent extracted, c-Mn(BH4)2 at 1001C/1201C under 1 bar H2 pressure is slightly over 7 wt% (Fig. 16). We also reported in Ref. [66] that a non-stoichiometric, amorphous a-Mn(BH4)(2x) hydride, accompanied by a NaCl-type salt, could be synthesized from the (2NaBH4 þ MnCl2) mixture containing the additives of ultrafine filamentary carbonyl nickel (Ni), graphene (reduced graphene oxide-rGO), and LiNH2. It is shown in Figs. 17 and 18 that both graphene and LiNH2 suppressed the

−0.1

−0.6

Endo

−4.5 −0.2 50

(B)

100

150 200 250 300 350 Sample temperature (°C)

400

450

500

Fig. 17 (A) Mass spectrometry (MS) (gas desorption spectra) of a (H2 þ B2H6) gas mixture for a ((2NaBH4 þ MnCl2) þ 5 wt% graphene) sample, ball milled (BM) for 0.5 h (QTR ¼36.4 kJ/g). (B) thermogravimetric analysis (TGA) and differential scanning calorimeter (DSC) curves during desorption. Heating rate 51C/min. Reproduced from Varin RA, Mattar DK, Polanski M, Shirani Bidabadi AR, Stobinski L. The effects of additives on the dehydrogenation of amorphous manganese borohydride and its crystalline form after solvent filtration/extraction. Energies 2017;10:1741.

Hydrides

2.0E−04

1.2E−04 1.0E−04

4.0E−03

8.0E−05

3.0E−03

6.0E−05 2.0E−03 4.0E−05 1.0E−03

2.0E−05

B2H6

−0.2

−2

−3 DSC

H2/B2H6 ratio = 5001

1.6E−04

6.0E−03 1.4E−04 5.0E−03

1.2E−04 1.0E−04

4.0E−03

8.0E−05

3.0E−03

6.0E−05 2.0E−03 4.0E−05 1.0E−03

2.0E−05

0.0 −1 Endo −0.2

−2

−3 −0.4

6.0E−03 1.4E−04 5.0E−03

1.2E−04 1.0E−04

4.0E−03

8.0E−05

3.0E−03

6.0E−05 2.0E−03 4.0E−05 1.0E−03

2.0E−05

Temperature (°C)

100

200

300

400

500

Sample temperature (°C)

0.2

400 350

0.0 −1 Endo −0.2

−2

−3 −0.4

300 250 200 150 100 50

−4

0 −0.6

−5

50 100 150 200 250 300 350 400 450 500

(E)

150

0

0.0E+00

0.0E+00

200

−0.6

1

Mass change (wt%)

1.6E−04

250

0

2.0E−04

H2/B2H6 ratio = 4456

300

50

100

Diborane relative partial pressure (μTorr/mg)

Hydrogen relative partial pressure (μTorr/mg)

7.0E−03

400

−4

(D)

1.8E−04

500

−5

Temperature (°C)

8.0E−03

400

0.2

50 100 150 200 250 300 350 400 450 500

(C)

300

350

0.0E+00

0.0E+00

200

0

Mass change (wt%)

7.0E−03

100

Sample temperature (°C)

1 Diborane relative partial pressure (μTorr/mg)

Hydrogen relative partial pressure (μTorr/mg)

100

(B)

1.8E−04

150

0

2.0E−04

8.0E−03

200

−0.6

−5

Temperature (°C)

250

50

50 100 150 200 250 300 350 400 450 500

(A)

−0.4

−4

0.0E+00

0.0E+00

Endo

Heat flow (mW/mg)

H2

300

Heat flow (mW/mg)

5.0E−03

0.0 −1

Heat flow (mW/mg)

1.4E−04

Mass change diff. (%/min)

6.0E−03

350 0

Mass change diff. (%/min)

1.6E−04

400

Mass change diff. (%/min)

1.8E−04

TG

Mass change (wt%)

7.0E−03

H2/B2H6 ratio = 5198

0.2

1 Diborane relative partial pressure (μTorr/mg)

Hydrogen relative partial pressure (μTorr/mg)

8.0E−03

589

100

(F)

200

300

400

500

Sample temperature (°C)

Fig. 18 Mass spectrometry (MS) (gas desorption spectra) of a (H2 þ B2H6) mixture, and the thermogravimetric (TG) and differential scanning calorimetry (DSC) curves during gas desorption for a ((2NaBH4 þ MnCl2) þ 5 wt% LiNH2) sample, ball milled for 0.5 h (QTR ¼36.4 kJ/g). Heating rate 51C/min. (A,B) sample 1, (C,D) sample 2, and (E,F) sample 3. Reproduced from Varin RA, Mattar DK, Polanski M, Shirani Bidabadi AR, Stobinski L. The effects of additives on the dehydrogenation of amorphous manganese borohydride and its crystalline form after solvent filtration/ extraction. Energies 2017;10:1741.

release of B2H6 during thermal gas desorption, with the LiNH2 additive being the most effective suppressor of B2H6 for an amorphous a-Mn(BH4)(2x) hydride. Fig. 18 shows the MS results of a (H2 þ B2H6) gas mixture and the corresponding TGA and DSC curves during gas desorption for three samples of a ((2NaBH4 þ MnCl2) þ 5 wt% LiNH2) composition, ball milled for 0.5 h (QTR ¼ 36.4 kJ/g). MS of gas desorption is quite remarkable, since it clearly shows nearly no presence of B2H6. The intensity ratios of the MS H2/B2H6 signal up to 2001C are 5198, 5001, and 4456 for samples 1, 2, and 3, respectively. Apparently, the LiNH2 additive is a very potent suppressor of B2H6. During solvent filtration and extraction from diethyl ether (Et2O), the amorphous aMn(BH4)(2x) hydride residing in the additive-bearing (Ni and graphene) samples, transformed into a crystalline c-Mn(BH4)2

590

Hydrides

3000 ((2NaBH4+MnCl2)+5 wt% LiNH2) −0.5h BM-solvent filtered/extracted

Counts

2000 EXT 60°C

1000

EXT 42°C

0 20

30

40

50 Degree 2-theta

60

70

80

90

Fig. 19 X-ray diffraction (XRD) pattern for the ((2NaBH4 þ MnCl2) þ 5 wt% LiNH2) sample, ball milled for 0.5 h (QTR ¼36.4 kJ/g) and subsequently filtered/extracted (EXT) at 42 and 601C using an evaporation time of 6–7 min. BM, ball milled. Reproduced from Varin RA, Mattar DK, Polanski M, Shirani Bidabadi AR, Stobinski L. The effects of additives on the dehydrogenation of amorphous manganese borohydride and its crystalline form after solvent filtration/extraction. Energies 2017;10:1741.

hydride, exhibiting a microstructure containing nanosize crystallites (grains). In contrast, the LiNH2 additive most likely suppressed the formation of a crystalline c-Mn(BH4)2 hydride during solvent filtration/extraction. Fig. 19 shows an XRD pattern for the 0.5 h BM ((2NaBH4 þ MnCl2) þ 5 wt.% LiNH2) sample (QTR ¼36.4 kJ/g), after filtration/extraction at 42 and 601C. No diffraction peaks for crystalline c-Mn(BH4)2 are present in Fig. 19, which is similar to an XRD pattern for an additive-free sample (Fig. 11(A)) albeit without any NaCl-type peaks beacuse NaCl-type salt was filtrated during processing. This result is quite remarkable, since it strongly suggests that the LiNH2 additive suppressed the transformation of an amorphous a-Mn(BH4)(2x) hydride into a crystalline c-Mn(BH4)2 hydride during solvent extraction. In a DSC, the thermal decomposition peaks of both amorphous a-Mn(BH4)(2x) and crystalline c-Mn(BH4)2 were endothermic for both the additive-free samples, as well as the samples with added graphene and Ni. The samples with LiNH2 exhibited an exothermic DSC decomposition peak [66].

2.18.4

Summary

The present chapter contains a thorough discussion on the complex metal-nonmetal hydrides and their composites which have the greatest potential as a viable source of hydrogen (H2) generation/storage for non-automotive applications, particularly, for disposable H2 storage cartridges for long-duration, low power demand devices that are powered by FC stacks, and use H2 gas as a fuel. The hydride system based on lithium alanate, LiAlH4, which is ball milled with nanometric additives, such as, nano-Ni, nanoFe, and nano-interstitial carbides, exhibits a relatively fast and efficient dehydrogenation behavior within the temperature range 40–1001C. Apparently, it is a very viable system for disposable cartridge applications. Another hydride system is a novel manganese borohydride, Mn(BH4)2, that has been recently synthesized by mechanochemical synthesis during BM using either the LiBH4 or NaBH4 precursors mixed with MnCl2. It is observed that during BM the LiBH4 precursor results in the formation of crystalline Mn(BH4)2 (c-Mn(BH4)2) with a theoretical H2 capacity of 9.5 wt% coexisting with a LiCl salt. In contrast, BM with the NaBH4 precursor results in the formation of amorphous Mn(BH4)2 (a-Mn(BH4)2) coexisting with a NaCl salt. It is found that during isothermal dehydrogenation up to 2001C of the synthesized (Mn(BH4)2/LiCl) mixture with filamentary Ni, LiNH2, and graphene rGO, each of additives has reduced the released quantity of B2H6, increasing the released H2 gas intensity ratio, H2/B2H6, from 493 for an additive-free sample, to 4667 for 5 wt% filamentary Ni (Fig. 7), to 3213 for 5 wt% LiNH2 and to 722 for 5 wt% graphene. Apparently, the filamentary Ni additive is the most effective suppressor of B2H6 released from crystalline Mn(BH4)2. Furthermore, during solvent extraction in diethyl ether (Et2O), a crystalline Mn(BH4)2 transformed into another hydride {Li(Et2O)2}Mn2(BH4)5 while the amorphous Mn(BH4)2 hydride transformed into crystalline Mn(BH4)2. BM of an additive-free as well as the mixtures containing the additives of ultrafine filamentary carbonyl nickel (Ni), graphene and LiNH2, resulted in the mechanochemical synthesis of a non-stoichiometric, amorphous a-Mn(BH4)(2x) hydride, accompanied by a NaCl-type salt. It is found that both graphene and LiNH2 suppressed the release of B2H6 during thermal gas desorption, with the LiNH2 additive being the most effective suppressor of B2H6 from a non-stoichiometric, amorphous a-Mn(BH4)(2x) hydride. During solvent filtration and extraction of additive-free, as well as additive-bearing (Ni and graphene) samples from diethyl ether (Et2O), the amorphous a-Mn(BH4)(2x) hydride transformed into a crystalline c-Mn(BH4)2 hydride, exhibiting a microstructure

Hydrides

591

containing nanosize crystallites (grains). In contrast, the LiNH2 additive, most likely suppressed the formation of a crystalline c-Mn (BH4)2 hydride during solvent filtration/extraction. In a DSC, the thermal decomposition peaks of both amorphous a-Mn(BH4)(2x) and crystalline c-Mn(BH4)2 were endothermic for the additive-free samples, as well as for the samples with added graphene and Ni. The samples with LiNH2 exhibited an exothermic DSC decomposition peak.

Acknowledgments The research cited in this chapter was funded by the NSERC Discovery and Hydrogen Canada (H2CAN) Strategic Research Network grants which were awarded to Prof. R.A. Varin and are gratefully acknowledged.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42]

Solar Nations Blog. The coming global oil crisis. Available from: http://www.oilcrisis.com. Rath BB, Marder JM. Does the fuel gage read empty? Adv Mater Process 2007;165(1):28–9. Ball M, Wietschelb M. The future of hydrogen – opportunities and challenges. Int J Hydrog Energy 2009;34:615–27. Scott DS. Smelling land: the hydrogen defense against climate catastrophe. Available from: www.h2.ca.Ottawa, ON: The Canadian Hydrogen Association; 2007. Yang J, Sudik A, Wolverton C, Siegel DJ. High capacity hydrogen storage materials: attributes for automotive applications and techniques for materials discovery. Chem Soc Rev 2010;39:656–75. Headlines & Global News. Toyota hydrogen fuel cell vehicles to be available in 2016. Available from: http://www.hngn.com/articles/33584/20140612/toyota-hydrogen-fuelcell-vehicles-to-be available-in-2016.htm. CNW Group. Hyundai first to offer hydrogen fuel cell vehicles to Canadian public. Available from: http://www.newswire.ca/en/story/1453489/hyundai-first-to-offer-hydrogenfuel-cell-vehicles-to-canadian-public. Varin RA, Shirani Bidabadi AR. Nanostructured, complex hydride systems for hydrogen generation. AIMS Energy 2015;3:121–43. Technical system targets: onboard hydrogen storage for light-duty fuel cell vehicles. Available from: http://energy.gov/sites/prod/files/2014/03/f12/ targets_onboard_hydro_storage.pdf. Zimmer R, Welke J. Let’s go green with hydrogen! the general public’s perspective. Int J Hydrog Energy 2012;37:17502–8. Steele BCH, Heinzel A. Materials for fuel-cell technologies. Nature 2001;414:345–52. Haile SM. Fuel cell materials and components. Acta Mater 2003;51:5981–6000. Tüber K, Zobel M, Schmidt H, Hebling C. A polymer electrolyte membrane fuel cell system for powering portable computers. J Power Sources 2003;122:1–8. Scott DS. Contrails against an azure sky. Int J Hydrog Energy 2004;29:1317–25. Yvon K, Lorenzoni J-L. Hydrogen-powered lawn mower. Int J Hydrog Energy 1993;18:345–8. Yvon K, Lorenzoni J-L. Hydrogen-powered lawn mower: 14 years of operation. Int J Hydrog Energy 2006;31:1763–7. Andrews J, Shabani B. Re-envisioning the role of hydrogen in a sustainable energy economy. Int J Hydrog Energy 2012;37:1184–203. Ardica. Freedom from the grid. Available from: http://www.ardica.com/ Varin RA, Czujko T, Wronski ZS. Nanomaterials for solid state hydrogen storage. New York, NY: Springer Science & Business Media; 2009. Varin RA, Jang M. The effects of graphite on the reversible hydrogen storage of nanostructured lithium amide and lithium hydride (LiNH2 þ 1.2LiH) system. J Alloys Compd 2011;509:7143–51. Kojima Y, Kawai Y. IR characterizations of lithium imide and amide. J Alloys Compd 2005;395:236–9. Varin RA, Wronski ZS. Progress in hydrogen storage in complex hydrides. [chapter13] In: Gandia LM, Arzamendi G, Diéguez PM Renewable hydrogen technologies. production, purification, storage, applications and safety. Amsterdam: Elsevier; 2013. p. 293–332. Bogdanović B, Sandrock G. Catalyzed complex metal hydrides. MRS Bull 2002;27(9):712–6. Orimo S-I, Nakamori Y, Eliseo JR, Züttel A, Jensen CM. Complex hydrides for hydrogen storage. Chem Rev 2007;107:4111–32. Bortz M, Bertheville B, Yvon K, Movlaev EA, Verbetsky VN, Fauth F. Mg3MnH7, containing the first known hexahydridomanganese (I) complex. J Alloys Compd 1998;279: L8–10. Graetz J, Reilly JJ. Kinetically stabilized hydrogen storage materials. Scripta Mater 2007;56:835–9. Varin RA, Zbroniec L. Decomposition behavior of unmilled and ball milled lithium alanate (LiAlH4) including long-term storage and moisture effects. J Alloys Compd 2010;504:89–101. Block J, Gray AP. The thermal decomposition of lithium aluminum hydride. Inorg Chem 1965;4:304–5. Varin RA, Zbroniec L. The effects of nanometric nickel (n-Ni) catalyst on the dehydrogenation and rehydrogenation behavior of ball milled lithium alanate (LiAlH4). J Alloys Compd 2010;506:928–39. Varin RA, Parviz R. The effects of the micrometric and nanometric iron (Fe) additives on the mechanical and thermal dehydrogenation of lithium alanate (LiAlH4), its self-discharge at low temperatures and rehydrogenation. Int J Hydrog Energy 2012;37:9088–102. Varin RA, Parviz R. The effects of the nanometric interstitial compounds TiC, ZrC and TiN on the mechanical and thermal dehydrogenation and rehydrogenation of the nanocomposite lithium alanate (LiAlH4) hydride. Int J Hydrog Energy 2014;39:2575–86. WebElements. The periodic table of the elements. Available from: http://webelements.com Varin RA, Parviz R, Zbroniec L, Wronski ZS. Fundamental aspects of mechanical dehydrogenation of Li-based complex hydride nanocomposites and their self-discharge at low temperatures. Energy Procedia 2012;29:644–53. Carpenter GJC, Wronski ZS. The characterization of nanostructured CVD Ni powders using transmission electron microscopy. J Nanopart Res 2004;6:215–21. Parviz R, Varin RA. Combined effects of molar ratio and ball milling energy on the phase transformations and mechanical dehydrogenation in the lithium amidemagnesium hydride (LiNH2 þ nMgH2)(n ¼ 0.5-2.0) nanocomposites. Int J Hydrog Energy 2013;38:8313–27. Girgis K. Structure of intermetallic compounds, Part I. [chapter 5] In: Cahn RW, Haasen P. Physical metallurgy#,. 3rd, revised and enlarged ed. North Holland: Elsevier Science Publishers, BV; 1983. p. 220–69. Rafi-ud-din , Lin Z, Li P, Qu X. Catalytic effects of nano-sized TiC additions on the hydrogen storage properties of LiAlH4. J Alloys Compd 2010;508:119–28. Brinks HW, Fossdal A, Fonneløp JE, Hauback BC. Crystal structure and stability of LiAlD4 with TiF3 additive. J Alloys Compd 2005;397:291–5. Mulana F, Nishimiya N. Application of mechanical milling to synthesize a novel quarterly hydride. J Alloys Compd. 2006;413:273–80. Ke X, Chen C. Thermodynamic functions and pressure-temperature phase diagram of lithium alanates by ab initio calculations. Phys Rev B 2007;76:024112. Jang JW, Shim JH, Cho YW, Lee BJ. Thermodynamic calculation of LiH2Li3AlH62LiAlH4 reactions. J Alloys Compd 2006;420:286–90. Züttel A, Rentsch S, Fischer P, et al. Hydrogen storage properties of LiBH4. J Alloys Compd 2003;356–357:515–20.

592

Hydrides

[43] Orimo S, Nakamori Y, Kitahara G, et al. Dehydriding and rehydriding reactions of LiBH4. J Alloys Compd 2005;404–406:427–30. [44] Li H-W, Yan Y, Orimo S-I, Züttel A, Jensen CM. Recent progress in metal borohydrides for hydrogen storage. Energies 2011;4:185–214. [45] Lang J, Gerhauser A, Filinchuk Y, Klassen T, Huot J. Differential scanning calorimetry (DSC) and synchrotron X-ray diffraction study of unmilled and milled LiBH4: a partial release of hydrogen at moderate temperatures. Crystals 2012;2:1–21. [46] Rude LH, Nielsen TK, Ravnsbaek DB, et al. Tailoring properties of borohydrides: a review. Phys Status Sol 2011;208:1754–73. [47] Xia GL, Guo YH, Wu Z, Yu XB. Enhanced hydrogen storage performance of LiBH4-Ni composite. J Alloys Compd 2009;479:545–8. [48] Wellons MS, Berseth PA, Zidan R. Novel catalytic effects of fullerene for LiBH4 hydrogen uptake and release. Nanotechnology 2009;20:4. 204022. [49] Zhang BJ, Liu BH. Hydrogen desorption from LiBH4 destabilized by chlorides of transition metal Fe, Co, and Ni. Int J Hydrog Energy 2010;35:7288–94. [50] Nakamori Y, Miwa K, Ninomiya A, et al. Correlation between thermodynamical stabilities of metal borohydrides and cation electronegativities: first principle calculations and experiments. Phys Rev B 2006;74:1–9. 045126. [51] Nakamori Y, Li H-W, Kikuchi K, et al. Thermodynamical stabilities of metal-borohydrides. J Alloys Compd 2007;446-447:296–300. [52] Choudhury P, Srinivasan SS, Bhethanabotla VR, Goswami Y, McGrath K, Stefanakos EK. Nano-Ni doped Li-Mn-B-H system as a new hydrogen storage candidate. Int J Hydrog Energy 2009;34:6325–34. [53] Varin RA, Zbroniec L. The effects of ball milling and nanometric nickel additive on the hydrogen desorption from lithium borohydride and manganese chloride (3LiBH4 þ MnCl2) mixture. Int J Hydrog Energy 2010;35:3588–97. [54] Cˇerný R, Penin N, Hagemann H, Filinchuk Y. The first crystallographic and spectroscopic characterization of a 3d-metal borohydride: Mn(BH4)2. J Phys Chem C 2009;113:9003–7. [55] Cˇerný R, Penin N, D’Anna V, Hagemann H, Durand E, Ru˚ˇzicˇka J. MgxMn(1 x)(BH4)2 (x ¼ 0 0.8), a cation solid solution in a bimetallic borohydride. Acta Mater 2011;59:5171–80. [56] Severa G, Hagemann H, Longhini M, Kaminski JW, Wesolowski TA, Jensen CM. Thermal desorption, vibrational spectroscopic, and DFT computational studies of the complex manganese borohydrides Mn(BH4)2 and [Mn(BH4)4]2 . J Phys Chem C 2010;114:15516–21. [57] Liu R, Reed D, Book D. Decomposition behaviour of Mn(BH4)2 formed by ball-milling LiBH4 and MnCl2. J Alloys Compd 2012;515:32–8. [58] Varin RA, Zbroniec L, Polanski M, Filinchuk Y, Cˇerný R. Mechano-chemical synthesis of manganese borohydride (Mn(BH4)2) and inverse cubic spinel (Li2MnCl4) in the (nLiBH4 þ MnCl2) system and its dehydrogenation behavior. Int J Hydrog Energy 2012;37:16056–69. [59] Varin RA, Shirani Bidabadi AR. The effect of milling energy input during mechano-chemical activation synthesis (MCAS) of the nanocrystalline manganese borohydride (Mn(BH4)2) on its thermal dehydrogenation properties. Int J Hydrog Energy 2014;39:11620–32. [60] Chłopek K, Frommen C, Léon A, Zabara O, Fichtner M. Synthesis and properties of magnesium tetrahydroborate, Mg(BH4)2. J Mater Chem 2007;17:3496–503. [61] Shirani Bidabadi AR, Korinek A, Botton GA, Varin RA. High resolution transmission electron microscopy (TEM), energy-dispersive X-ray spectroscopy (EDS) and X-ray diffraction studies of nanocrystalline manganese borohydride (Mn(BH4)2) after mechano-chemical synthesis and thermal dehydrogenation. Acta Mater 2015;100:392–400. [62] Shirani Bidabadi AR, Varin RA, Polanski M, Biglari M, Stobinski L. The effects of filamentary Ni, graphene and lithium amide (LiNH2) additives on the mechano-chemical synthesis of crystalline manganese borohydride (Mn(BH4)2) and its solvent extraction. Mater Res Bull (submitted). [63] Tumanov NA, Safin DA, Richter B, et al. Challenges in the synthetic routes to Mn(BH4)2: insight into intermediate compounds. Dalton Trans 2015;44:6571–80. [64] Varin RA, Mattar DK, Shirani Bidabadi AR, Polanski M. Synthesis of amorphous manganese borohydride in the (NaBH4–MnCl2) system, its hydrogen generation properties and crystalline transformation during solvent extraction. J Energy Chem 2017;26:24–34. [65] Llamas-Jansa I, Aliouane N, Deledda S, et al. Chloride substitution induced by mechano-chemical reactions between NaBH4 and transition metal chlorides. J Alloys Compd 2012;530:186–92. [66] Varin RA, Mattar DK, Polanski M, Shirani Bidabadi AR, Stobinski L. The effects of additives on the dehydrogenation of amorphous manganese borohydride and its crystalline form after solvent filtration/extraction. Energies 2017;10:1741.

Relevant Websites http://www.expouav.com/news/latest/hydrogen-fuel-cells-drones/ Commercial UAV News – Will Hydrogen Fuel Cells Help Drones Stay in the Air. https://en.wikipedia.org/wiki/Hydrogen_economy Hydrogen Economy. http://www.iop.org/publications/iop/2016/file_67429.pdf IOP – Next Steps for Hydrogen: Physics, Technology and the Future. http://www.popularmechanics.com/cars/hybrid-electric/news/a26867/hydrogen-cars-toyota-murai-honda-clarity/ Meet the Next Generation of the Hydrogen Car. http://sputniknews.com/us/20160316/1036424754/lockheed-mach-6-hypersonic.html Sputnik – Lockheed Closes in On Mach 6 Hypersonic Aircraft Technology. http://sputniknews.com/middleeast/20160718/1043214722/drones-test-syria.html Sputnik – Russia Tests Hydrogen-Powered Drones in Syria – Senior Military Source. https://uk.motor1.com/news/222633/toyota-mirai-50k-miles/ The Hydrogen Fuel Cell Car has Been Untroubled on the Streets of London. https://www.thestar.com/news/queenspark/2017/06/15/ontario-looks-at-hydrogen-powered-trains-for-go-transit.html Toronto Star Newspapers Ltd. – Ontario Looks at Hydrogen-Powered Trains for GO Transit. http://www.globalbioenergy.org/uploads/media/0601_UNEP_-_The_hydrogen_economy.pdf UNEP – The Hydrogen Economy: A non-technical review.

2.19 Solid Oxides Sabit Horoz and Omer Sahin, Siirt University, Siirt, Turkey r 2018 Elsevier Inc. All rights reserved.

2.19.1 Overview to Fuel Cells 2.19.1.1 Definition and Working Principle of a Fuel Cell 2.19.1.2 Fuel Cell History 2.19.1.3 Types of Fuel Cells 2.19.1.3.1 Alkaline fuel cells 2.19.1.3.2 Phosphoric acid fuel cells 2.19.1.3.3 Proton exchange membrane fuel cells 2.19.1.3.4 Molten carbonate fuel cells 2.19.1.3.5 Direct methanol fuel cells 2.19.1.3.6 Solid oxide fuel cells 2.19.1.4 Applications of Fuel Cells 2.19.1.4.1 Manageable power 2.19.1.4.2 Backup power 2.19.1.4.3 Carry application 2.19.1.4.3.1 Automobiles 2.19.1.4.3.2 Buses 2.19.1.4.3.3 Utility vehicles 2.19.1.4.3.4 Scooters and bicycles 2.19.1.4.4 Immobile applications 2.19.2 Solid Oxide Fuel Cells 2.19.2.1 Chronological Background 2.19.2.2 Basic Operation and Design of Solid Oxide Fuel Cells 2.19.2.2.1 Basic operation of solid oxide fuel cells 2.19.2.2.2 Design of solid oxide fuel cells 2.19.2.2.3 Stack design of solid oxide fuel cells 2.19.2.2.4 Fabrication of solid oxide fuel cells 2.19.2.3 Components of the Solid Oxide Fuel Cells and Requirements for Components of Solid Oxide Fuel Cells 2.19.2.3.1 Cathode 2.19.2.3.2 Electrolyte 2.19.2.3.3 Anode 2.19.2.3.4 Interconnect 2.19.2.3.5 Sealing materials 2.19.2.4 Applications and Technological Aspects of Solid Oxide Fuel Cells 2.19.2.4.1 Solid oxide fuel cell cogeneration: Collective heat and power (solid oxide fuel cell-collective heat and power) 2.19.2.4.2 Cogeneration of solid oxide fuel cells: Gas turbine (solid oxide fuel cell-gas turbine) 2.19.2.4.3 Cogeneration of solid oxide fuel cell: Chemical fabrication method 2.19.2.5 Thermodynamics of the Solid Oxide Fuel Cell 2.19.2.5.1 Electromotive force and Gibbs free energy change (∆G) 2.19.2.6 Effect of Concentration of the Electromotive Force 2.19.2.6.1 Heat effects in a galvanic cell 2.19.2.6.2 The temperature coefficient of the electromotive force 2.19.2.6.3 The pressure coefficient of the electromotive force 2.19.2.6.4 The thermal and chemical expansion coefficients 2.19.2.7 Advantages and Limitations of Solid Oxide Fuel Cells 2.19.2.8 Environmental Impact of Solid Oxide Fuel Cells 2.19.2.9 Case Studies on Solid Oxide Fuel Cells 2.19.2.10 Future Directions and Prospects for Solid Oxide Fuel Cells 2.19.3 Conclusions Appendix 1: Thermodynamic Data of Selected Chemical Reactions and Substances References Further Reading Relevant Websites

594 594 594 595 595 596 596 597 597 598 598 598 598 599 599 599 599 599 599 600 600 600 600 601 602 602 602 605 606 606 608 608 608 608 609 610 612 612 612 613 615 616 617 618 619 619 620 621 621 623 627 627

Comprehensive Energy Systems, Volume 2

593

doi:10.1016/B978-0-12-809597-3.00242-X

594

Solid Oxides

2.19.1

Overview to Fuel Cells

Fuel cells are presently generating great attention due to their growing need on behalf of sustainable energy sources with possible power generation in fixed, transportable, and vehicle needs [1,2]. The high efficiency of the change of chemical energy to direct electrical energy and the radical decline of the radiations of sulfur and nitrogen oxides and hydrocarbon contaminants to a very low level and the substantial reduction of CO2 emissions provide major ecological benefits over fuel cells on predictable energy production. In spite of the fact that the fuel cell was invented 160 years ago and that fuel cells offer extraordinary efficiency plus green characteristics, fuel cells are approaching commercial reality. The underlying reason behind this is the price of fuel cell skill. Nevertheless, important growth in the development of both materials with improved properties and in manufacturing processes in the last two decades have made fuel cells a realistic proposition to compete on a commercial footing with conventional power generation. As mentioned before, fuel cells are far from being an innovative technology; the idea of a fuel cell was first shown via William Grove in 1839 [3]. Grove observed that when electrolysis of water was investigated, the electrolysis products catalyzed by the platinum electrodes when the stream was closed had a small flow of current through the association between hydrogen and oxygen. Grove realized the possibility of bringing together a majority of these sequences to produce a gas-filled voltaic battery, and in addition made a vital remark that there must be a “significant domain” among the gas, electrode, and electrolyte phases in the cell [4]. The three-segment boundary maximizes the contact area between the gas reactant, the electrolyte, and the electrode (electrocatalytic conductor). It persists at the center of fuel cell investigation and advancement.

2.19.1.1

Definition and Working Principle of a Fuel Cell

It is an electrochemical tool that vigorously changes the potency of the chemical reaction between hydrogen and oxidant into electric energy. Fuel cells work like a battery with a simple principle: two electrodes are separated by an electrolyte. They are different from batteries because they are intended for constant filling of used reactants. In contrast to restricted interior energy packing capability of a battery, they yield electricity with an exterior fuel and oxidant source (usually oxygen or air, even with chlorine or chlorine dioxide). The corporeal configuration of a fuel cell comprises of an electrolyte sheath separated by two electrons. Today, fuel cells are being investigated and continuously developed for good achievement and utilization aims. Supposedly, a fuel cell works similar to a battery, involving an electrolyte positioned between two electrodes: an anode and a cathode. In contrast with a battery, a fuel cell does not expire or need to be charged. Provided that fuel is delivered, it will produce electricity in the form of electricity and heat. Oxygen dismisses an electrode and passes hydrogen through the other, thus producing electricity, water, and heat. Hydrogen fuel is served to the "anode" of the fuel cell. Oxygen (or air) comes in the cathode fuel cell. This response happens along a chemical agent; the hydrogen atom is split into a proton and an electron that lead to the cathode in different ways. The proton crosses the electrolyte. Electrons form an isolated stream that can be used beforehand returning to the cathode to be combined with hydrogen and oxygen in a molecule of water [5].

2.19.1.2

Fuel Cell History

The fuel cell has actually been known for over 150 years in science, in spite of the modern high-tech atmosphere. Although they were of general interest in general in the 1980s, fuel cells have turned into an intense topic of exploration and improvement in the 1900s. British professors W. Nicholson and A. Carlisle defined the use of power to separate hydrogen and oxygen in water in 1800 [6]. This process is called electrolysis. In 1832, M. Faraday stated that the amount of eluted components was proportional to the amount of electrical charge transferred by passing an electrically charged salt. From these tests two basic regulations of electrolysis have been assumed. These are: First Law: the quantity of a material formed at an electrode through electrolysis is relative to the amount of moles of electrons (the number of electricity) transmitted at that electrode. Second Law: the quantity of Faradays of electric charge required to discharge one mole of element at an electrode is identical to the number of "additional" uncomplicated charges on that ion. W.R. Grove, conversely, advanced this knowledge one stage more, or rather a phase backward in 1838. Grove discovered that one end of each of the two platinum electrodes would be dipped in a sulfuric acid flask and the other end sealed discretely in oxygen and hydrogen containing cups to provide a constant flow of current between the electrodes. The closed flasks detained water along with the gases, and he stated that the water level ascended in both cylinders as the current run. By uniting numerous groups of these electrodes in a series circuit, he produced the initial fuel cell [6]. Consequently, it is known that Grove is the inventor of the fuel cell. L. Mond and his assistant C. Langer conducted their experimentations with a gas-motorized battery using "Mond-gas" from coalin 1889. With thin platinum electrodes, they obtained 6 amps (determining the surface area of the electrode) at 0.73 V per square meter. This system was also called the fuel cell by Ludwig. F.W. Ostwald, the founder of the field of physical chemistry, provided much of the theoretical understanding of how fuel cells work [6,7]. In 1893, the interrelated parts of several apparatuses of the fuel cell (electrodes, electrolytes, oxidizing and reducing

Solid Oxides

595

agents, anions and cations) were experimentally determined by Ostwald. Grove said that the action in the gas cell was taking place at the contact point between the electrode, gas and electrolyte, and a loss of further clarification. Ostwald used his inventive work on physical properties and chemical reactions to solve Grove’s gas puzzle. His investigation of the fundamental chemistry of fuel cells formed the basis for later fuel cell scientists [7]. The major applied hydrogen–oxygen fuel cells transform air and fuel unswervingly into electricity through electrochemical procedures, were developed by F.T. Bacon. In the late 1930s, he began investigating alkaline electrolyte fuel cells. He manufactured a cell that worked with a nickel gas electrode and functioned under high pressure of 3000 psi in 1939. During the Second World War, Bacon tried to develop a fuel cell that could be utilized in the Royal Navy (RN) submarines, and in 1958, exhibited an alkaline cell by means of a 10-in. diameter electrode stack for the British National Research Development Company. F.T. Bacon’s fuel cells proved to be reliable enough to draw Pratt and Whitney’s consideration even though they are expensive. The corporation approved F.T. Bacon’s work for the Apollo spacecraft fuel cells. Stimulatingly, these important mechanisms have revolutionized their time, and today they have made hydrogen a worthy candidate as the upcoming fuel carrier for the world’s budget. As a remarkable message, the A. Chalmers Fuel Cell Tractor was a trial tractor industrialized in 1959. The 20-horse power (HP) tractors had 1008 single-fuel cells, which were powered by a combination of gases (mostly propane) that generated a movement of current.

2.19.1.3

Types of Fuel Cells

A number of dissimilar fuel cells are now being developed that vary in the structure of the electrolyte. Nevertheless, the elementary functioning standard of any fuel cell is the equivalent and is presented in Fig. 1. At the anode, a fuel, such as hydrogen, is oxidized into protons and electrons, whereas at the cathode, oxygen is condensed to oxide types, and these then rejoin to form water. Resting on the electrolytes, the protons or oxide ions are passaged by ionic conductivity, but electronically, by an insulating electrolyte as the electrons pass through an external circuit that provides electrical energy. Five main categories of fuel cell are applicable, described below, which all have similar elementary working norms, specifically two electrodes detached by an electrolyte. Ions travel in one way, which counts on the electrolyte, through the electrolyte to the opposing electrode, where the reaction arises, whereas the electrons run curved an external circuit, generating electric power. Every single variety of fuel cell is categorized by an electrolyte. In general, it is thought that the two most likely fuel cells to achieve good market share are the polymer electrolyte sheath and solid oxide fuel cell (SOFC).

2.19.1.3.1

Alkaline fuel cells

Britain’s Francis Thomas Bacon (1904–92) began experimenting with alkaline electrolytes in the late 1930s and placed on potassium hydroxide (or KOH) instead of acid electrolytes known from Grove’s early discoveries. KOH acid electrolytes in addition to the electrodes did not show as corrosive. F.T. Bacon’s cell, however, consumed “gas diffusion electrodes,” which are porous, different from the solid electrodes Grove used. The gas diffusing electrodes have enlarged the exterior part of the reaction concerning the electrode, electrolyte, and fuel. In addition, F.T. Bacon used pressurized gases to ensure that the electrolytes “flooded” the small holes in the electrodes. F.T. Bacon built sufficient growth to make large-scale demonstrations with plenty of alkaline cells for the next 20 years. In the early 1960s, airplane machine maker Pratt and Whitney certified F.T. Bacon’s patents and earned the National Aeronautics and Space Administration (NASA) contract to power the Apollo spacecraft in space missions. These cells are able to attain energy-producing efficiency of as much as 70%. Apollo was required to deliver both electricity and drinking water on the spacecraft. Alkaline fuel cell (AFCs) generally use KOH as the electrolyte and function at 1601F. But, they are extremely sensitive to carbon pollution, hence they require rich hydrogen and oxygen [8]. Hydrogen + Oxygen → Water (+ electrical power + heat) Hydrogen

Porous electrode (anode Electrolyte

Electrons e−

Porous electrode (cathode)

Oxygen Fig. 1 A representation illustration displaying the overall functioning standards for a fuel cell. Reproduced from Ormerod RM. Solid oxide fuel cells. Chem Soc Rev 2003;32:17–28.

596

Solid Oxides

At low temperature, the aqueous alkaline electrolyte cells can easily start from cold and have the advantage that the electrolyte is suitably high at 60–801C for a measured cleaning scale. At these temperatures, favorably energetic catalysts are generally obliged from the platinum species. However, silver and extreme surface nickel were used as catalysts in this scheme; nickel is traditionally used as a conductive essential substance. Low-cost catalysts customarily oblige more advanced functioning temperatures; the F.T. Bacon cell is an example of a nickel catalyst operated at 200–2501C. At these temperatures either high pressures should be applied to the structure or high concentration liquids should be used to avert water loss. High pressure systems are not suitable for the operation of the air due to the high pumping energy, while extreme concentrations are able to initiate deterioration, which limits the variety of building materials. The inability to endure this kind of cell carbon dioxide (CO2) is a major problem. The fuel limits the choice of hydrazine to uncorrupted hydrogen and needs the air filter to remove 0.04% of the airborne CO2. The internal reforming cell is an attempt to dispose of this issue: the fuel is constructed from palladium plus silver, and its hydrocarbons reformed by vapor on a nickel catalyst on one side of the fuel, alcohol, or electron. The resulting hydrogen moves across the electrode and rejoins with the electrolyte, but the palladium prevents CO2 from passing into the electrolyte [9]. Alkaline fuel cell – abbreviation of electrode reactions and comprehensive cell reaction. Anode: H2 þ 2OH -2H2 O þ 2e Cathode:

1 O2 þ H2 O þ 2e -2OH 2

1 Overall: H2 þ O2 -H2 O 2

2.19.1.3.2

Phosphoric acid fuel cells

Acidic electrolyte cells are extra lenient to CO2 and allow the use of normal air and impure hydrogen. However, the deterioration obstruction limits the selection of building of materials particularly for electrodes and catalysts. Electrodes can consist of gold, tantalum, titanium, and carbon, and only platinum group metals can be employed as catalysts. The acid used as the electrolyte should not be unstable, such as sulfuric and phosphoric acids, so only water evaporates. The electrolyte in the PAFC has a paper background packed with phosphoric acid, which transports hydrogen ions. The functional temperature is about 2001C. Working temperature requires platinum as catalyst tested on graphite material. However, this temperature is susceptible to platinum CO damage. Cells using hydrocarbons around 1501C as direct fuel have depleted productivity and current density, which is why they are limited to research studies. Alcoholic fuels and contaminated hydrogen have been used by innumerable elements. The performance of acid cells is much less than that of alkaline cells, probably owing to the inferior performance of the air electrode due to the intensified immovability of the peroxides shaped in the acid medium. However, there are many concessions that are able to complete between alkaline and acid fuel cells, taking into account the structure and the working temperature, and the likely use of the preferred cell [9]. PAFC fuel cell – abbreviation of electrode reactions and comprehensive cell reaction. Anode: H2 -2Hþ þ 2e Cathode:

1 O2 þ 2Hþ þ 2e -H2 O 2

1 Overall: H2 þ O2 -H2 O 2 Phosphoric acid fuel cells (PAFCs) are one of the few fuels present in the market today. Hundreds of fuel cell schemes were appointed in 19 countries in infirmaries, retirement homes, offices, office buildings, institutions, power stations, storage areas, and wastewater treatment plants. PAFCs produce electricity with an efficiency of over 40% – this is about 85% if the fuel cell producing steam is sourced for cogeneration – which corresponds to about 35% of the main electricity in the United States. PAFCs accept watery phosphoric acid by way of the electrolyte and operate at about 4501F. One of the key benefits of this sort of fuel cell is its aptitude to use polluted hydrogen as fuel as well as cogeneration efficiency of about 85%. PAFCs may neutralize a CO intensity of about 1.5%, which magnifies the choice of fuels that they can use. If gasoline is used, the sulfur needs to be uninvolved [10,11].

2.19.1.3.3

Proton exchange membrane fuel cells

The proton exchange membrane fuel cell (PEMFC), also termed polymer electrolyte fuel cell (PEFC) or solid polymer fuel cell (SPFC), handles a solid proton-conducting sheath as electrolyte and the electrode resources are ordinarily made of carbon coated with platinum catalyst.

Solid Oxides

597

PEM fuel cell – abbreviation of electrode reactions and comprehensive cell reaction. Anode: 2H2 -4Hþ þ 4e Cathode: O2 þ 4Hþ þ 4e -2H2 O Overall: 2H2 þ O2 þ -2H2 O þ DE Hydrogen ions move across the sheath and the electrodes run through the external circuit. The most suitable sheaths are perfluorosulfonic acid; for example, nafion tolerates working temperatures up to 1201C. The sheath needs the gas to be humidified, and its curb and heat transfer have crucial pressure for the optimum functioning of this sort of fuel cell. The necessity for excessive water matter virtually controls the operating temperature to 801C. An extra complication is that at reduced temperatures the reforming gas demands decontamination for carbon monoxide (CO), poisoning the catalyst. Danish power systems (DPS) has industrialized a new PEMFC sheath that is able to perform at temperatures up to 2001C. At this condition, the fuel cell is able to drive at a CO concentration of at least 3%. The gas need not be dampened and current densities of up to 1 A/cm2 at 0.5 V are obtained. The DPS style of PEMFC can be united with a methanol–hydrogen innovator capable of operating at equivalent temperatures. In this circumstance, the innovator is able to take a bulky portion of the required warmth directly from the surplus heat generation of the fuel cell, and the total energy efficiency is increased [9].

2.19.1.3.4

Molten carbonate fuel cells

The electrolyte consists of a combination of molten potassium carbonate and lithium carbonate to convey the carbonate ions from the cathode to the anode. The CO32 movement necessitates a CO2 reserve on the cathode part of the cell, which is commonly attained by reprocessing the anode side gas. The temperature is around 8501C, which allows the use of nickel as the catalyst material. The progression ensuing in a hydrogen–oxygen fuel cell operating at higher temperatures deprived of a watery electrolyte is believed to produce oxide ions in the air electrode. İt is then transported to the fuel electrode to oxygenate the hydrogen, thus a molten ionic oxide will afford the preeminent electrolyte to promote this method. Uncomplicated ionic oxides, however, ensure melting points more than 10001C, which is why concentration is emphasized on soluble salts at low temperatures. These are mostly salts with oxygen-holding anions, for instance nitrates, sulfates, and carbonates. At high temperatures, the uninterrupted reaction of the hydrocarbons with the fuel is exceedingly poisonous and consequently it is preferable to switch the coal yields to hydrogen or methane. Consideration should be paid to the influence of hydrocarbon oxidation on fuel choice. Carbon dioxide is an important product that can be troublesome with some salts. That’s why it is of utmost importance to regard it as a mixture of carbonate or carbonate dispersed as an electrolyte. A salt mixture can ensure a considerable benefit since it will have a lower melting point than both of its apparatuses. A convenient approach to withstand the carbonate configuration of the electrolyteimmiscible material is to eradicate CO2, which is a product of the fuel gas, and convey it to the oxidant electrode in the air or oxygen stream. Accordingly, the transfer of carbonate ions in the electrolyte can be counterbalanced by the transfer of CO2 exterior. An analogous system can work for cells that use hydrogen as fuel [9]. Molten carbonate fuel cells (MCFC) – evaluation of electrode reactions and comprehensive cell reaction. Anode reaction: H2 þ CO3 2 -CO2 þ H2 O þ 2e Cathode reaction:

1 O2 þ CO2 þ 2e -CO3 2 2

1 Overall reaction: H2 þ O2 -H2 O 2

2.19.1.3.5

Direct methanol fuel cells

The direct methanol fuel cell (DMFC) was realized and established by NASA’s Jet Propulsion Laboratory and the University of Southern California (USC), and has been produced worldwide with 56 patents and over 62 patents pending. This variety of fuel cell is constructed on solid polymer technology, but utilizes methanol directly in the function of a fuel. If possible, the production of hydrogen, one of the major difficulties in the placement of fuel cells, will be a major step forward in the automotive field. There are essential problems with methanol compared to hydrogen, such as lower electrochemical activity, lower cell voltages, and therefore higher efficiencies. In addition, methanol can be miscible in water, so some are obliged to pass through the water drenched sheath and cause corrosion and deplete gas problems on the cathode side. Methanol fuel cells are replacing customary batteries and are assumed to recommend longer operating times contrasted to existing lithium ion batteries and gain significant market share due to instant recharging by replacing nonrefundable fuel cartridges. DMFCs are like PEM cells in terms of using a polymer covering layer as the electrolyte. In spite of this, the anode catalyst itself does not need a fuel reformer by drawing hydrogen from the liquid methanol. Typically, about 40% efficiency is expected in such fuel cells operating at 120–1901F. This situation is a reasonably low range, which runs mobile phones and laptop computers, making the fuel cell of interest for small and medium-sized functions. Higher efficiency is obtained at higher temperatures. Moreover, researchers are working on DMFC prototypes to be utilized by the army to power electronic field equipment [12].

598

Solid Oxides

DMFC – evaluation of electrode reactions and comprehensive cell reaction. Anode reaction: MeOH þ H2 O-CO2 þ 6Hþ þ 6e
 1 Cathode reaction: 3 O2 þ 6Hþ þ 6e -3H2 O 2
 1 Overall reaction: MeOH þ H2 O þ 3 O2 -CO2 þ 3H2 O 2

2.19.1.3.6

Solid oxide fuel cells

SOFCs are made completely of solid state components using an ion-conducting oxide ceramic as electrolyte and work at 900– 10001C. SOFCs offer some advantages over other sorts of fuel cell. They create several problems in electrolyte management (often compared to liquid electrolytes that are difficult to process) and have the highest efficiency in all fuel cells. Existing industrial science uses quite a lot of ceramic materials for active fuel cell components. The anode is typically formed from zirconium cermet (Ni/YSZ) stabilized with an electronically conductive nickel/yttria. The cathode is grounded on a mixed conductor perovskite, lanthanum manganate (LaMnO3). As the oxygen ion-conducting electrolyte, yttria-stabilized zirconia (YSZ) is used. To produce an appropriate voltage, the fuel cells in the unaffected stack are interconnected by a doped lanthanum chromate (e.g., La0.8Ca0.2CrO3), which is connected to the anodes and cathodes of neighboring sections. While a number of stack formats are accepted around the world, the most common structure is planar SOFC. SOFC operation is required for the YSZ electrolyte to afford plentiful oxygen ion conductivity. But, the price for the fabrication of devices is substantially high, demonstrating that pricy high temperature alloys ought to be used, especially for factory balancing purposes. If the operating temperature is reduced to 600–8001C, these costs are significantly reduced if low-cost structural apparatuses, such as stainless steel, are allowed to be used. A reduced working temperature will result in a longer system life and a more general scheme efficiency and a drop in thermal stresses in active ceramic structures. In order to reduce the working temperature of the SOFC the conductivity of the YSZ should be enhanced. A coordinated effort is being made by scientists around the world to acquire such materials. The ceramics currently under investigation include Gd-doped CeO2, Ba2In205, and (Sr, Mg)doped LaGaO3. Nevertheless, these original materials address severe drawbacks compared to YSZ, and it is likely that the first economic SOFC units will use zirconia-based ceramics as electrolytes [13]. One kind of SOFC utilizes a sequence of tubes in meter length, and other dissimilarities comprise a compacted disk that simulates the top of a soup can [14]. SOFC – evaluation of electrode reactions and comprehensive cell reaction. Anode reaction: H2 þ O2 -H2 O þ 2e CO þ O2 -CO2 þ 2e Cathode reaction: O2 þ 4e -22 Overall reaction: H2 þ O2 þ CO-H2 O þ CO2

2.19.1.4

Applications of Fuel Cells

Fuel cells have been advanced for use in a variety of applications since the early 1990s. Fuel cells can be operated for manageable, backup, portable, and immobile power applications. In this section, some of these uses for fuel cells are briefly described.

2.19.1.4.1

Manageable power

Manageable fuel cells are light, durable, manageable power supplies that extend the life of a device without being charged. Comparatively, rechargeable batteries have charger arrangements that are made up of alternating current (AC) chargers that require the charger to be charged, or direct current (DC) chargers that charge other batteries. Rechargeable batteries are not feasible for moveable and armed forces electronic applications because they can be weighty and cannot accept the power supplies. Some manageable fuel cell applications include laptop computers, phones, tools, army equipment, battery chargers, sensors, and aerial and underwater vehicles. A vital difference between rechargeable batteries and fuel cells is that a fuel cell needs a continuous source of fuel. Some fuel types employed with fuel cells include metal hydrides, methanol, formic acid, ethanol and, of course, hydrogen. For manageable fuel cells, methanol or ethanol may be delivered to the fuel cell, while fuel or a fuel reformer may be added to the fuel cell pack.

2.19.1.4.2

Backup power

Backup power structures afford power when the foremost power supply weakens. Fuel cells used for backup power come in many sizes and types and usually use hydrogen as fuel. Substitution fuel cells can be commercialized faster than other fuel cells because they are not reliant on the execution of a hydrogen infrastructure. Some backup power applications consist of computer systems, manufacturing facilities, homes, and auxiliary installations. Compressed hydrogen fueled PEM fuel cell is the most accepted fuel cell type employed for backup power applications. There are many fuel possibilities; for example, compressed hydrogen, liquid hydrogen, propane, natural gas, etc. An electrolyzer system is

Solid Oxides

599

a good alternative for backup power applications because it is able to generate hydrogen on request. The electrolyzer is able to be employed with some sources, such as solar energy panels, wind, nuclear, or electricity produced by the local electric company.

2.19.1.4.3

Carry application

Fuel cells can be employed for many portable applications, including automobiles, buses, utility vehicles, and scooters and bicycles. Many fuel cell prototype vehicles were fashioned for each of these vehicle types. 2.19.1.4.3.1 Automobiles Most automobile producers have been using fuel cell vehicles for at least several decades. Automotive producers are concerned with fuel cell technology as it is a "next generation" technology that can be formed from native sources and has low or zero emissions. Fuel cell automobiles ordinarily utilize propellant-compressed hydrogen; however, quite a lot of producers have also introduced a fuel cell vehicle that uses methanol. Automotive fuel cells may possibly have one or more of the following:

• •

A fuel pump that can only be utilized to provide the full power of a vehicle. A fuel cell provides a constant amount of power, so additional equipment is connected to the fuel cell for vehicle acceleration and other power-increasing power requirements.

The temperature is generally between 60 and 801C owing to the usage of a polymer sheath that decreases the temperature below 1001C. 2.19.1.4.3.2 Buses Buses have displayed effective application of fuel cells for marketable vehicle transportation. The difference between buses and automobiles include power needs, space availability, operating regimes, and fueling areas. Buses necessitate more power than automobiles and wear out more because of stops and starts. Owing to the size of a bus, enough hydrogen can easily be stored on board. Fuel cell buses are advantageous compared to diesel buses owing to their zero emissions. This is indispensable in densely occupied and polluted cities. Many bus manufacturers began showing the first fuel cell buses at the beginning of the 1990s. Methanol and zinc are shown as they are in fuel cell cars, but the type of fuel most often used is compressed hydrogen. Fuel cell buses are located in British Columbia, California, Amsterdam, Barcelona, Hamburg, London, Luxembourg, Madrid, Porto, Reykjavik, Stockholm, and Stuttgart. 2.19.1.4.3.3 Utility vehicles Utility vehicles have been a popular early adopter of fuel cell technology because the technology used for these vehicles is typically lead–acid batteries that necessitate upkeep and charging. The advantages of fuel cell auxiliary tools exhibit lower functioning costs, diminished preservation, lower interruption, and wider variety. Utility vehicles operating with a fuel cell are also able to function in an enclosed area as there are no emissions. Fuel cells can be used in such useful tools as forklifts, golf carts, lawn care vehicles, airport haulers, wheelchairs, unmanned vehicles, boats, small planes, submarines, and small military vehicles. The primary fuel cell supports were introduced at the beginning of the 2000s. While methanol, metal hydrides, and sodium borohydride are also used, as is the case in fuel cell cars, the most commonly used type of fuel is compressed hydrogen. The most commonly used fuel cell is the PEMFC, but DMFCs and AFCs have also been explored. 2.19.1.4.3.4 Scooters and bicycles In countries with huge populations, bicycles and scooters are popular modes of transport. With compressed hydrogen and methanol, fuel cells for these applications have been shown favorably. Hydrogen storage for these vehicles is still a question; consequently, metal hydrides and electrolyzers have been explored in the last decade. Many industrialists introduced their first fuel cell scooters and bikes after the fuel cell car and bus demonstrations at the beginning of the 2000s. Methanol, metal hydrides, and zinc are also shown in fuel cell cars, but the most commonly used type of fuel is compressed hydrogen. The most popular sort of fuel cell is PEMFC, but DMFCs and zinc-air fuel chillers have been explored too.

2.19.1.4.4

Immobile applications

Fuel cells for immobile applications have been economically available for more than 20 years. The central variable in these systems is the selection of fuel cell, the heating and cooling of the stacks. Immobile fuel cells can be employed as the initial power source. It is usually used to strengthen houses that are not tied to the grid or to provide additional power. Fuel cells can be associated with photovoltaics, batteries, condensers, or wind turbines that bring initial or secondary power. Moreover, fuel cells can function as a backup or energy generator that provides power when the network fails. An independent system might need another supply of energy for these times. These may be combinations of batteries and supercapacitors. Many industrialists began to show off station plants in the 1990s. Contrasting with other applications, the most common sort of fuel is natural gas. Other common types of fuel are propane, compressed hydrogen, biogas, methanol, fuels based on oil, city gas, synthesis gas, digester gas, and landfill gas. The most frequently used immobile fuel cell type is the PEMFC but other types are SOFCs, AFCs, and PAFCs. The United States, Germany, and Japan have the most fixed fuel cell power plants [15].

600

Solid Oxides

2.19.2

Solid Oxide Fuel Cells

SOFCs have appeared as a crucial high temperature fuel cell technology in recent times [16]. They promise to be enormously functional in large-scale; for example, in full-scale industrial stations and large-scale electricity generating stations. An SOFC is characterized by a solid ceramic electrolyte, which is a metallic oxide. The SOFC operates especially at high temperatures between 800 and 10001C. This temperature tolerates internal reforming, promotes rapid electrocatalyst with nonprecious metals, and yields high excellence byproduct heat for coproduction. For such fuel cells, efficiency can reach 70 and 20% for heat recovery (HR). SOFCs are the most suitable solution for power supply. They are useful applications because of the crucial time required to reach operating temperatures [17]. SOFC has received increased attention in recent years because of its efficiency and permanency. SOFCs are thought to be very adaptable with regard to fuels in other fuel cells, which is a major benefit over other fuel cells. Existing studies on SOFCs focus on the description and synthesis of modern materials that are able to diminish the cost of fabrication and extend the life of the SOFC [18–24].

2.19.2.1

Chronological Background

The SOFC was originally presented by Nernst in 1899 [25] after the invention of solid oxide electrolytes. Nernst stated that the conductivity of metal oxides ascended very gradually with temperature and stayed only comparatively low, while metal oxide mixtures could have higher conductivities. He noted that the conductivity values of water and salt are very low, as it is perfectly compatible with the identified behavior of liquid electrolytes and its behavior is comparable to aqueous salt solutions with extremely high conductivity. In 1905, Haber staged his first official document on fuel cells with a solid electrolyte that used glass and porcelain as electrolyte materials, working temperature, and platinum and gold electrode materials. In 1916, Baur and Treadwell published a patent on fuel cells containing metal oxides as electrolytes and on the ceramic solids that are salt-soluble in the pores as electrolytes. Until 1935, Schottky recommended that YSZ might be employed as a solid fuel cell electrolyte. E. Baur and H. Preis tested solid oxide electrolytes in the late 1930s using materials such as zirconium, yttrium, cerium, lanthanum, and tungsten oxide. The paramount ceramic fuel cell to be worked on by Baur and Preis at 10001C was accomplished in 1937 [26]. In the 1940s, O.K. Davtyan supplemented monazite sand, sodium carbonate, tungsten trioxide, and soda glass to raise conductivity and mechanical strong point. Unfortunately, Davtyan’s project also had undesirable chemical reactions and short life duration. Wagner noticed the presence of cavities in the anion subdivision of mixed oxide solid solutions and clarified the Nernst cluster’s conduction mechanism, that is, oxide ion conductors, in 1943 [27]. At the end of the 1950s, research in solid oxide technology began speed up at the Central Technical Institute in the Hague, the Netherlands; the Consolidation Coal Company, Pennsylvania; and General Electric in Schenectady, New York. A debate on fuel cells in 1959 showed that difficulties with solid electrolytes include moderately high internal electrical resistance, melting and short circuit owing to the semiconductivity. After 1960 various factors increased curiosity on fuel cell technology. Progress in the research and fabrication of ceramic materials has guided to an increased importance on the SOFCs. At the beginning of the 1960s, a growing quantity of patent applications were processed considering the enhancement of SOFC technology. One of the problems that SOFCs are experiencing now is that they have dense electrolyte layers and the efficiency is low because they meet with high losses due to internal resistance. For example, investigators at Westinghouse, in 1962, tested with a cell that used zirconium oxide and calcium oxide. Lately, the rise in energy charges and developments in material technology have prompted resurgent work on SOFCs, and a recent report covers about 40 corporations working on these fuel cells, such as the Global Thermoelectric’s Fuel Cell Division, which grows cells fabricated at the Julich Research Institute in Germany. Continued developments in research and fabrication approaches during the 1970s led to the development of thinner electrolytes in electron benches. Over the past 20 years, many SOFC forms have been analyzed, containing different tubular and planar designs. Some progress and improvements have been made in SOFCs, and predominantly the electrode and electrolyte materials used. Cermatec Advanced Ion Technologies is operating on diesel-powered components for capacities up to 10 kW, which will be consumed for power generation. The US Department of Energy has stated that an SOFC microturbine cogeneration element has been assessed by the National Fuel Cell Research Center and Southern California Edison since April 2000. The fuel cell was manufactured by S. Westinghouse and the microturbine by the Northern Research and Engineering Corporation. In a year of real working circumstances, the natural gas powered 220 kW SOFC achieves an efficiency of 60%. In addition, a world record for SOFC operation for about 8 years still stands, and the prototype cells show two major successes: more than 100 thermal cycling durability and less than 0.1% voltage degradation in a 1000 h. Moreover, an SOFC cogeneration system with a power rating of 140 kW provided by S. Westinghouse is still functioning in the Netherlands. This design took more than 16,600 h and operated the longest-lasting fuel cell in the world [28]. The initial display of saleable prototype cells in a full-scale SOFC component is also important at the same time.

2.19.2.2 2.19.2.2.1

Basic Operation and Design of Solid Oxide Fuel Cells Basic operation of solid oxide fuel cells

The fuel cell operating at high temperatures works from 800 to 10001C. The key process comprises the REDOX system. In SOFCs, conductive types are oxygen ions. Oxygen is downgraded by removing electrons in the cathode and ionized. The YSZ electrolyte is

Solid Oxides

Fuel (H2 hydrocarbons)

601

H2O, CO2

Porous fuel electrode (anode)

e− Solid electrolyte O−2

O−2

O−2

O−2

External load O−2

O−2 e−

Oxidant air

Porous oxidant electrode (cathode)

Fig. 2 Concept diagram of a solid oxide fuel cell (SOFC).

frequently utilized for the transmission of oxygen ions. Oxygen is given to the permeable cathode, which is downgraded to oxygen ions transported to the anode by solid electrolyte. At the anode, the oxygen ions merge with hydrogen and produce water together with CO2 as revealed in Fig. 2. Hard materials are needed for SOFC components owing to their high operating temperature. Taking into account long-term proficient cell operation, chemical and thermal permanency is at the highest level, along with high-grade electrical and electrochemical properties, and high chemical and thermal perception in fuel cell functioning surroundings [18,21–23,29–32]. Two sorts of concentration are dealt with during fabrication: cost-effective and mechanical. Cost-effective concentration incorporates dropping manufacturing costs using inexpensive materials. Mechanical concentrations mean that the design should be manufacturing-friendly and cost-effective. Cost-friendliness in technical organizations means that materials must be charge sensitive with maximum results, effortlessly produced to the need of mechanics, and must be reduced with inactivity and less savings [29]. A lone primitive fuel cell boosts the open-circuit voltage marginally above 1 V and falls to about 0.5–0.7 V when the stack is linked [11,31–33]. The current density during operation can vary from 200 to 1000 mA/cm2 (DC). To obtain boosted current and voltage output, many cells are combined in series or parallel to form a series. The SOFC’s power plant includes different subunits, such as refueling the stack, an oxygen supply unit, a fuel processing unit for DC to AC conversion, a unit for achieving the mandatory temperature, and control/security units. The components (anodes, cathodes, and electrolytes) used in SOFC are all solid materials [29].

2.19.2.2.2

Design of solid oxide fuel cells

Since the 1960s, a large number of SOFCs and stack designs have been advised by many various collections [34]. Many of these constructions have been unified for a variety of applications ranging from a few watts to very kilowatts [18,34–36]. Various changes can be accomplished by changing the construction of the cell and changing the sizes of the SOFC elements despite the fact that the materials utilized are identical. Relatedly, changing the stack design affects not only the performance of the SOFC components but also the lifetime. Stack design depends directly on cell design [34,36,37]. Success of the fuel cell can be restricted by three important parts: loss of ohmics resulting from the transport of charges, catalytic activity, or losses due to mass transport, gas flow, or the transport of ions through cell elements. The boundaries of mass and ohmic transport are precisely dependent on the thickness of the diverse cell elements, which signifies that they depend on direct fuel cell design, while affecting the catalytic activity of the material electrodes used in the components of the cell. In addition to performance, other factors considered in fuel cell design include material cost and mechanical strength [34,36,38]. Commonly, four various sorts of cell designs are applied: cathode assisted, electrolyte assisted, anode assisted, and externally supported designs. A cell element is generally about 1–5 mm thicker than the other elements to mechanically support the cell. To increase the performance of the cell, and reduce the mechanical strength and the cell cost, the thickness of the different elements of the cell outside the support material is in the range of 5–50 mm. The mechanical strength of the cell can be intensified by decreasing the thermomechanical stress caused by cracks or porosity, or incompatible thermal expansion coefficient (TEC) of various elements of the cell [18,29,34,38,39]. The electrolyte supported fuel cell design is extensively used. The electrolyte-assisted design has advantages in terms of ease of manufacture and mechanical strength compared to other designs. In this structure, the nonporous electrolyte has a thickness of 50–500 mm. Wet methods, such as dip coating or screen printing, are used to deposit electrodes on both sides. The electrolyte material is cast at a higher temperature than the precipitation of the electrodes since it is ignited at lower temperatures. These methods help to prevent the material grouping from reacting with the dense electrolytes at high temperatures, as in the case where La1 xSrxMnO3 (LSM) and La1 xSrxFe1 yCoyO3 (LSCF) constituents (cathode) are reacting at higher temperature with more widely employed materials. The difficulty of this sort of scheme is low overall performance owing to the thicker electrolyte mandatory for the mechanical strength of the cell [38,40]. Fuel cell schemes built on cathode support are thought to be an outstanding design in terms of developed system size and working hours with insignificant degradations. Owing to high price support matters and industrial cost [41], this design was

602

Solid Oxides

aborted at the beginning of 2000. Rare earth oxide cathode materials, in other words, lanthanum oxide, are the key reason for the high price. Due to the low mechanical strength and high density, the high cost with the cost of LSM-supporting supports caused a very high industrial price compared to the structure made from NiO/YSZ or pristine zirconia. In addition, at high temperature, which is a requirement for conventional ceramic processing, lanthanum oxide reacts with zirconia [34,40]. Elementary temperature progression approaches, such as vapor deposition, are used to avoid reacting with lanthanum oxide zirconia, while avoiding processing at elevated temperatures, but increasing production costs [42]. Nowadays, though there is not a large-scale study on the origin of this design, new developments, such as low-cost cathode production, techniques, such as low rare earth content and cathode infiltration of cathode materials, may cause the cathode-supported design to be looked at again in the near future [41,43]. Fuel cell schemes are built on anode support, which is intensely attractive for NiO–YSZ cermets because of their high strength, NiO’s low solubility in zirconia, ease of manufacturing, and high electrical conductivity with small ohmic losses. However, when NiO is compared with zirconia, there is a large alteration in volume because of the drop of the anode material during high thermal expansion and the operation of the cell becomes prone to electrolyte cracking during manufacturing and cell operation. In consideration of maintaining the performance of the cell, the electrolyte cracking problem is able to be organized by controlling the electrolyte thickness, supporting the microstructure and cautiously selecting the NiO content in the NiO–YSZ composition [34,38,44,45]. Supported cell designs are also divided into two subcategories: ceramic and metal-supported designs. These two subcategories are generally preferred for ceramic based fuel cells because of ceramic support materials, inertia, and low TEC. Cathode degradation is restricted because of the diminished use of high chromium in high temperature alloys [46]. Cell designs built on ceramic backing materials are not thought appropriate for a system that is subject to fast and continuous thermal cycling, typically because of their high hardness and lower toughness [29,34,38]. Metal support base cell designs, metal pieces, normally cannot be heated to the high temperature required to treat ceramic pieces, because they are a bit difficult to manufacture. In addition, high temperature metal alloys are chemically unbalanced in the cathode oxidant condition, causing a decrease in the lifetime of the cell. In addition to the lack of metal support design, it has a very appealing design because of its very high electrical conductivity, high strength, and thermal shock resistance, which helps its use in the automotive and residential micro-collective heat and power (CHP) sectors [34,38,44].

2.19.2.2.3

Stack design of solid oxide fuel cells

A quantity of stack schemes have been established over a period of time, but are now focused on tubular and hybrid schemes with different designs for the fuel and air supply system. Planar designs have a low manufacturing cost, but because of the changeable TEC of the bulk constituents during thermal cycling, sealing is difficult in the planar design. An appropriate candidate for interconnecting material is not well defined in the case of planar design. Further, the tubular design does not require complicated sealing, but the manufacturing cost is very high in this design [18,34,38]. Hybrid designs based on modified or flattened tubes alleviate manufacturing cost issues using techniques such as screen printing [34,38,47]. As a result, there is no precise design that plays an important role at this point.

2.19.2.2.4

Fabrication of solid oxide fuel cells

Dip coating, tape casting, reinforcement rolling, screen printing, etc. are generally utilized in the production of SOFCs at elevated temperatures due to the number of layers in thicknesses varying from 10 to 100 mm [18,34,38]. Other techniques, such as chemical vapor deposition (CVD), pulsed-laser deposition (PLD), plasma spraying, and electrophoretic deposition are also available to avoid chemical interaction of the compounds, but are still not popular due to their difficulty and extra costs [11,38]. Table 1 provides a brief explanation of the different deposition procedures and properties used for the production of SOFC apparatuses [11,38,52].

2.19.2.3

Components of the Solid Oxide Fuel Cells and Requirements for Components of Solid Oxide Fuel Cells

SOFCs place considerable demands on materials utilized as electrolytes, anodes, cathodes, interconnects, and sealing materials. Each element must face specific necessities and have multiple functions. All components must have chemical and physical permanency in the proper chemical situation, be chemically friendly with other apparatuses, have appropriate conductivity, and have TECs similar to other components to prevent cracking during production. However, it is also essential that those SOFC elements are inexpensive and durable, but easy to produce [18,19,29,36,53–65]. Certain requirements for each component [66]:

• • • • • • • •

appropriate solidity appropriate conductivity chemical suitability with other elements cogeneration of thermal expansion to evade cracking during the cell process dense electrolyte to avoid gas mixture suitable for special manufacturing conditions compatibility in high temperatures where ceramic structures are manufactured low-priced

Table 1

The different deposition techniques to fabricate solid oxide fuel cells (SOFCs)

Deposition technique

Concise description

Common applications

Features

References

Screen printing

The organized interruption is located on the screen and is enforced under pressure for its path

Electrolyte, anode, and cathode

[18,34,48]

Tape casting

The state of ceramic film is made with an impermanent support made of moveable plate. A scalpel is used to achieve the preferred thickness Plasma jet (B10,000K) is employed to dissolve the particles and then sprayed onto the substrate with rapid solidification

Electrolyte and anode

Scaling can be done simply. The construction of cracks occurs in some ceria-based electrolytes and leads to unsuitable condensation Calcination is easy to apply, multilayered cells can be produced, and electrolytes can be produced in various thicknesses. Not suitable for large cells Rapid deposition, films with different microstructures and alignments are able to be formed, SOFC layers can be precipitated into metallic substrates without sintering, scaling can be done easily Thin and leaky electrolytes are able to be manufactured by changing the degraded layer solution Low cathode ASR (area-specific resistance), increase in power density Electrolyte thin film, low deposition, high temperature costs, abrasive products The dense film can be deposited on the porous substrate; can be used as a tubular substrate, homogeneous foil with good mechanical properties can be obtained, low sintering temperature, high cost of SOFC, high temperature required for rapid accumulation

Electrolyte

Dense and thin electrolyte can be obtained

[18,34,48]

Electrolyte, anode, and cathode

It is a low-cost method, but time consuming

[18,34,50,51]

Anode and electrolyte

Electrolytes of variable thickness are possible, multilayered cells can be produced

[18,34,48]

Electrolyte, anode, and cathode

Thin electrolytes can be obtained; morphology and composition can be controlled; low temperature for storage. Ceria films can be broken; high price; techniques, such as radio frequency (RF) injection and direct current (DC) injection are time consuming Easy to use; homogeneous films, complex forms, effortless accumulation on the lower layers; easily scaled to control film thickness; pipe-shaped cathode is a cheaper option for accumulating electrolyte

[18,34]

Atmospheric plasma spray (APS)

Spray pyrolyze Colloidal sprayed position (CSD) Chemical vapor deposition (CVD) Electrochemical vapor deposition (EVD)

Spin coating

Dip coating or slurry coating

Tape calendaring

Sputtering

An electric field is applied to force charged particles suspended in a liquid to move toward a countercharged electrode

Electrolyte Electrolyte and cathode Cathode and electrolyte Tubular cells, electrolytes, interconnectors

Electrolyte and cathode

[18,34]

[18,34,49] [18] [18,50] [18,34]

[18,34,50,51]

Solid Oxides

Electrophoretic deposition (EPD)

A suspension of dust is sprayed onto a hot substrate followed by sintering to deposit a film A colloidal solution is pumped on the hot substrate to a liquid dispersion device, such as an ultrasonic nozzle The deposition occurs by a gas phase reaction between the metal halide precursors and the hot substrate Metal chloride vapor and water vapor or oxygen are placed on both sides of the bottom layer in a chamber. Closure of the reaction pores between the metal chloride and the water vapor occurs and then film growth occurs due to the formation of an electrochemical potential gradient The film can be produced by spinning a sol–gel precursor on a porous or dense layer. Film thickness can be controlled by mixing ratio The surface is immersed in alcoholic or aqueous suspension and then dried at room temperature. Thereafter, preheating is performed and this is followed by sintering. The process is repeated Although similar to strip casting, the gap between the rollers is used to control the thickness. The deposited suspension is a thermoplastic material Noble ions usually use argon ions to bombard the target material. After that, the atoms or ions of the targeted material are released and the substrate is deposited

Electrolyte, anode, cathode, and interconnector

[18,34,48]

(Continued )

603

604

Table 1

Continued Concise description

Common applications

Features

References

Pulsed-laser deposition (PLD) or laser ablation Sol–gel

Laser ablation of the material is done in vacuum and then placed on the substrate at a temperature of 7001C The salts of the required cations are dissolved to form the solution. The colloid is then dried to obtain a powder precipitated by conventional methods or partially dried to give viscous slurry precipitated by a wet process The suspension is left on the base with a brush

Electrolyte and cathode

Miniature SOFCs can be produced, have potential for automation, and nanostructures can be created No need for high sintering temperature

[18] [18]

Easy method, difficult to scale, not repeatable

[18]

Painting

Electrolyte

Electrolyte, anode, and cathode

Solid Oxides

Deposition technique

Solid Oxides 2.19.2.3.1

605

Cathode

The drop of oxygen in the SOFC happens in the cathode. The reaction arising from the cathode is specified by [18,35] 1 2 O ðgasÞ þ 2e ðcathodeÞ-O2 ðelectrolyteÞ 2

ð1Þ

The cathode ought to have high electron and oxygen ion conductivity, TEC with electrolyte, chemical immovability at working temperature, adequate permeability, and improved catalysis for reduction reaction [18,35]. In general, it is expected that the electrochemical reaction ensues at the triple stage limit, that is, at the contact point of the electron conductor, the oxygen ion conductor and the gas [18,35,67,68]. It is often responsible for the momentous loss of voltages in the SOFC by reason of the large amount of polarization losses because of the reduction reaction [35,67,68]. Overemployed cathode material is strontium-doped lanthanum manganite (LSM) in SOFC at high temperature [18,35,69–76]. Lanthanide is associated with a manganite perovskite family, which is moderately switched by strontium. Perovskite structures provide a large amount of change in the compositional and oxygen stoichiometry, whose benefits optimize the catalytic and electrical properties. This cathode material performs better at higher temperatures as it is expressively better at a temperature over 8001C. If there is a surplus of lanthanum or strontium oxide in the stoichiometry, it causes the realization of an insulator stage, such as La2Zr2O7 and SrZrO3, by reason of its contact with the YSZ, which leads to lesser performance. This trouble can be solved by raising the manganese concentration and keeping the temperature of the process below 13001C [18,35,72,75–80]. LSM has unimportant oxygen ion conductivity at 9001C with high electronic conductivity. Therefore the reaction zone is restricted only by the electrode and the electrolyte interface. This causes to the need for the cathode to be leaky enough to dissipate oxygen at the electrode/electrolyte interface to reduce oxygen. A composite layer or graded structure of LSM/YSZ is applied to increase triple phase boundaries (TPBs) [18,69,71,76]. The cathode reactions in electron-conducting perovskite-type materials are able to come true in many stages and in dissimilar ways depending on the nature of the electrode material. The most accepted mechanism in pristine electronic conductive materials is the surface path. The bulk pathway is an accepted mechanism for mixed ionic electronic conduction (MIEC) and the electrolyte surface pathway is thought to be in composite materials, such as LSM/YSZ [35,68]. Extensive efforts are being made to improve electrode performance for low and medium temperature SOFCs [35,67,69,70,72–77]. A amount of cathode supplies with better electronic/ionic conductivity and enlarged oxygen exchange rate have been explored. It will allow the oxygen reduction effect to be extended in the cathode structure or the fuel/cathode connection, and as a result will increase the reaction rate [18,35,67,69,70]. To that end, other perovskite materials, such as LSCF, are being investigated. The electronic and ionic conductivity values of some cathode materials (MIEC) are indicated in Table 2 [18,73,77,78]. These materials perform well but exhibit weak chemical permanency caused by reactions with YSZ electrolytes (solid state). To avoid this problem, a double layered electrolyte containing gadolinium and either samarium-doped seria/YSZ is used [18,29,77–82]. Researchers have investigated LSM–YSZ and LSCF cathodes, which are YSZ–Ce0.9Gd0.1O1.95 (CGO) electrolytes, by current and impedance spectroscopy and detected that LSM–YSZ has higher field-specific polarization resistance (1.8 O-cm) than LSCF (0.4 O-cm) at 8501C [81]. The solid state reaction between the YSZ electrolyte and the MIEC materials during synthesis can be evaded by impregnating the porous material with a leaky electrolyte structure requiring a low sintering temperature [18,29,35,83]. The impregnation of the perovskite can be achieved by filtration of the aqueous solution of the molten nitrate salt or the nitrate salts of the ionic species or by the nanoparticle suspension of the material. Parallel consequences have been stated using any of the methods mentioned above [18,35,84]. The other approach is the use of chemically steady perovskite material with good ionic and electronic flow at lower temperatures. Moreover, scientists have worked on cathode material, such as the Pr1 xSrxFeO3 (PSF) group [35,85]. It was found that Pr0.8Sr0FeO3 does not react with YSZ solid state and that it has a precise polarization unwillingness of 0.204 and 0.164 W-cm2 at 800 and 8501C, respectively [85]. Another material, which is the main oxygen split-up sheath owing to its high oxygen surface Table 2 Electronic and ionic conductivities of various solid oxide fuel cell (SOFC) cathode compounds Cathode

se (S/cm)

si (S/cm)

LSM LSC LSF LSCF PSF BSCF LSCu

o200 1600 450 230 300, 78 45, 20 500

o4  10 0.4 5  10 3 0.2 – – –

T (1C) 8

800 800 800 900 550, 800 500, 800 800

Source: Reproduced from Badwal SPS, Giddey S, Munnings C, Kulkarni A. Review of progress in high temperature solid oxide fuel cells. J Aust Ceram Soc 2014;50:23–37. Abbreviations: BSCF, Ba0.5Sr0.5Co0.6Fe0.4O3–5; LSC, La1 xSrxCoO3; LSCF, La1–xSrxFe1 yCoyO3; LSCu, La0.75Sr0.25CuO2.5 δ; LSF, La1 xSrxFeo3; LSM, La1 xSrxMnO3; PSF, Pr1–xSrxFeO3.

606

Solid Oxides

modification and diffusion characteristics, is Ba0.5Sr0.5Co0.6Fe0.4O3–5 (BSCF) [29,35,86]. Both materials have a small electronic conductivity that increases with dropping temperature at 8001C [87]. For this reason, such materials are useful in SOFCs at low temperature and intermediate temperature [44,86–116]. In contrast, La6.4Sr1.6Cu8O20 þ 8 provides very good electrical conductivity, but counters with YSZ at temperatures above 6001C [88].

2.19.2.3.2

Electrolyte

An electrolyte is an element of a SOFC functioned for ion transport between electrodes. High oxygen conductivity is necessary for an extensive range of oxygen-limited pressures with the electrolyte. If a material meets the versatile property criteria, it can be a good electrolyte. To prevent cross diffusion of fuel and oxidizing components, the electrolyte must be completely dense without open porosity. Electrolyte must be chemically inert because it is a contact with these materials according to electrodes and sealing materials. The electrolyte must be chemically and structurally unchanging to bring the number of ionic carriers together as they are subjected to extreme oxidizing and decreasing surroundings in the electrons. It must have good tensile strength and toughness to endure high mechanical and thermal stresses throughout process and production [18,35,44,90–118]. Four sorts of electrolyte, constructed on completely stained zirconia, doped ceria, doped LaGaO3, and doped Bi2O3 have given a great deal of consideration [18,35,36,89,91,92]. Each system contains different subcategory materials with one or more additives that support phase durability. The type and concentration of the dopant is influential in terms of phase combination, chemical durability, ionic conductivity, thermal and mechanical properties. It should be considered that most substances are unbalanced in SOFC functioning circumstances, reveal low ionic conductivity, and acquire conductivity in a reducing surrounding [18,36,80,89,91]. The electrolyte that satisfies most of the requirements for the operation of SOFC is based on Y2O3–ZrO2 containing Y2O3 (3–10 mol%). ZrO2 (YSZ) compound with cubic structure, 8 mol% Y2O3, is known to have the highest ionic conductivity during SOFC operation at about 10001C with a mechanical strength of about 225–300 MPa at room temperature and 150 MPa above 5001C [91]. On the other hand, 3YSZ exhibits high mechanical strength at room temperature of about 1000 MPa and 5001C to 400 MP, while at the same time exhibiting high mechanical strength at the expense of ionic conductivity, unsteady and complex stage conjugation [91]. Much has been emphasized to diminish the working temperatures, costs and severity of material degradation of SOFCs [92,93]. However, if the operating temperature is reduced from 10001C to 8001C, the ionic conductivity of 8YSZ drops from 0.178 to 0.052 S/cm [91]. And so, supplementary materials with higher ionic conductivity at lower temperatures have been examined [18,35,36,88–93]. Other electrolyte substances that are accepted by reason of their higher ionic conductivities contain Sc2O3–ZrO2, doped ceria, doped Bi2O3, lanthanum strontium magnesium gallate perovskite (LSGM) [18,35,36,73,91–97]. The thermal and mechanical properties of scandia-added zirconia are very comparable to yttria-zirconia, and the ionic conductivity of 9 mol% Sc2O3–ZrO3 is virtually twice that of 8YSZ at 10001C. Conversely, Sc2O3 is restricted and luxurious for essential use as an electrolyte in large-scale SOFCs [89,91]. In addition to Sc2O3–ZrO2, other electrolyte species, added cerium, and added Bi2O3 improve electronic conductivity above 6001C, and thus regularly require the fortification of the YSZ thin layer [35,36,91,93]. The ionic conductivity of the ceria (Ce0.9Gd0.1O1.95 (CGO)) electrolyte at 6001C is similar to that of YSZ at 8001C. But, the decrease of Ce4 þ ions to Ce3 þ in CGO causes electron short circuit due to electron conduction. This reduction also results in unwanted lattice expansion [35,36,93]. It has been described that CGO electrolytic SOFC can operate at lower temperatures of about 5001C. At this temperature range, the contribution of the electronic conductivity will be insignificant. In addition, low temperature ferrite provisions the apparatuses operated in this temperature limit and sealing materials [35,36,93]. Alternative electrolyte substance that has ionic conductivity compared to CGO is LSGM with a larger ionic window [18,93]. In spite of the fact that LSGM forms superior electrolyte material at medium and low temperatures, achieving a single period and permanency is a significant barrier when using LSGM [18,35,36]. The LSGM, CeO2, and Bi2O3 electrolytes are considered to be too delicate for applied devices owing to their weak mechanical strength [18,89,91]. Other system to reducing ohmic losses is to employ thin electrolyte in the SOFC. The necessary ionic conductivity is able to be obtained at temperatures o8001C and the electrolyte is less than 50 mm in thickness. If an electrolyte thickness of 10–15 mm is synthesized by conventional ceramic routes, it is considered to be free of emissions. With this thickness, an ionic conductivity value of 40.01 S/cm is able to be attained to obtain a certain resistance in the field below 0.15 Ocm2. Production techniques, such as electrochemical vapor deposition (EVD), sputtering, sol–gel laser sputtering, magnetron sputtering, and colloidal precipitation, can be used to obtain an electrolyte thickness of 10–15 mm using a cell-assisted electrode [36,93]. Both anode and cathodesupported cell designs are being studied. Other materials with structures, such as pyrochlore, perovskite, and brownmillerite, have also been investigated and evaluated with limited use in the fuel cell [18,35,91–97]. Researchers perceived that the ionic conductivities of the different substances were changed not only by the changing additive and concentration but also by the temperature, such as the ionic conductivities of 3, 8, and 10 mol% YSZ at 10001C are 0.058, 0.178, and 0.136 S/cm, respectively, while the ionic conductivities at 8001C are 0.018, 0.052, and 0.037 S/cm, respectively [29].

2.19.2.3.3

Anode

The most important purpose of the anode is to oxidize the fuel and afford a pathway to the electrons manufactured in the oxidation reaction so that it is able to extend the existing collector. The SOFC anode should have properties, such as good electroconductivity, adequate penetrability, good electrocatalytic action, phase relationship, with current collector and electrolyte,

Solid Oxides

607

good microstructure resistance to work at SOFC operating temperature, and TEC similar to electrolyte [18,36,44,73]. 1 O2 ðelectrolyteÞ- O2 ðgasÞ þ 2e ðAnodeÞ 2

ð2Þ

YSZ is the most ordinarily employed anode with electrolyte, Ni, and YSZ, while in the example of cerium-based electrolytes, Nidoped seria is used. Ni is a good catalyst for separating hydrogen at the operating temperature of SOFC and at the same time it has excellent electrical conductivity of about 2  104 S/cm [44,98–101]. NiO also has a higher melting temperature (19551C) than the NiO/YSZ anode fabrication temperature. Since both phases of the Ni–YSZ anodes are a major contributor to their performance, there should be a balance of quantities for sufficient results. Generally, the phase to volume ratio of Ni to YSZ is from 35:65 to 55:45 [101]. The coarse formation of Ni particles can be prevented by adding anode YAZ phase and affording extra TPBs by extending the active reaction zone. Besides the anode configuration, properties such as the particle size and the size of the precursor oxide powders significantly influence anode performance. Different deposition methods, processing steps, and heat treatment are the factors that depend on optimum particle size and shape. It has been informed that the ideal particle size for perfect screen printing is 0.1 mm for YAZ and 0.2 mm for NiO [101–103]. With these starting materials, an anodic excess potential of only 90 mV was perceived with a deterioration of only 1.7% at a current density of 350 mA/cm2 at a temperature of 8001C and after 4300 h of continuous operation at 9001C [103]. The impure substance of the starting oxide material NiO–YSZ is also another factor affecting the anode performance. It has been discovered that when the content of joint pollutants, such as Na2O and SiO2, increases from 100 ppm, especially the glass/sodium silicate phases interfere at the anode/electrolyte interface [104], it diminishes the activity of the anode within 100 h. Anode-assisted SOFCs are undergoing widespread study because they deliver better performance than predictable electrolyte-assisted SOFCs [18,35,44,98–103]. In anode-supported SOFCs, a quantity of thin-film deposition procedures are utilized to form a thin electrolyte layer on the anode, while band casting or die pressing methods are used to prepare the anode. In anode-supported SOFCs, attention is paid to the anodic microstructure, porosity, and thickness. The anode should be sufficiently compressed that it is able to be placed on the electrolyte layer, but it must be leaky enough to permit diffusion of gas therein so that the anode is able to access the electrolyte interface. For some commercial systems, it has been reported that power densities greater than 1 W/cm2 are required in Ni–YSZ anode-supported SOFCs [101,103,105]. Some of the problems encountered by the Ni–YSZ anode throughout operation in SOFCs include the degradation of Ni particles with time channel, low permanency to nonimpurity fuel, and problems causing coking when fuels, such as natural gas and hydrocarbons, are employed [106]. It has been described that the coking problem can be diminished by the addition of metals, such as Ru, Au, or Rh (1–4 mol%), which can be further decreased by consuming catalyst and ceria. Other possible nominees for Ni–YSZ conductive anode are strontium titanates, lanthanum chromite (LaCrO3), and fluorite-based perovskite materials based on cerium with insignificant metal proportion. Excellent performance has been reported in the literature with a La0.75Sr0.25Cr0.5Mn0.5O3 anode at 9001C using humidified hydrogen fuel. With using dry methane fuel, no coking is detected when La0.8Sr0.2Cr0.97V0.03O3 anode is continuously functioned for 100 h. Some cermet anodes display electronic and ionic conductivity recognized as MIEC and raise the reaction area [35,36,101,108]. To advance SOFC functioning, the improvement of new supplies is ongoing, as LSGM supported cells [111] show marked performance (power density 4800 mW/cm2) with dual perovskite anode SrMg1–xMnxMoO6 δ (SMMO). It has also been stated that the above substance reveals a power density of 338 mW/cm2 when there is no substantial carbon buildup when methane is used at 8001C [113]. The catalytic activities of these anodes may be close to the Ni–YSZ anode, but the suitability with YSZ and CGO electrolytes still needs to be investigated. The electrical conductivity of the cermet anodes is smaller than that of the Ni–YSZ anode. This conductivity depends on the production methods and structure of the substances employed. Using ceramics and metal composites, the conductivity of low-conductivity ceramic anodes is able to be increased. Two-layer designs are also able to improve current collection and transmission. Table 3 gives us the best other anode supplies with the best electrode conductivity and polarization resistance for Ni–YSZ [18,36,101–113]. Table 3

Polarization resistance for different materials

Anode

se (S/cm)

Rp (Ω cm2)

Temperature (C)

Y0.3Ce0.7O2 δ La0.4Sr0.6TiO3 La0.35Sr0.65TiO3–Ce0.7La0.3O2 (7:3 mol ratio) La0.25Sr0.75Cr0.5Mn0.5O3 Sr0.88Y0.08TiO3 Pr0.4Sr0.6Co0.2Fe0.7Nb0.1O3 δ Y0.2Zr0.62Ti0.18O1.9 La0.6Sr0.4Fe0.9Mn0.1O3 þ 2 wt% Pd YSZ–40 wt% CeO2 þ 1 wt% Pd

10 360 40 3 64 – 0.15 – –

0.1 0.7 0.2–0.4 0.1–0.3 1–10 0.16 1–10 0.8 in CH4 0.26

900 800 800 800 800 900 800 800 900

Source: Reproduced from Badwal SPS, Giddey S, Munnings C, Kulkarni A. Review of progress in high temperature solid oxide fuel cells. J Aust Ceram Soc 2014;50:23–37.

608

Solid Oxides

2.19.2.3.4

Interconnect

The electrical connection between adjacent cells in the SOFC station is afforded by the interconnect. The interconnect substances should have properties, such as high thermal conductivity, high electrical conductivity, and TECs, that match other elements of the cell. In addition, it should have great gas mechanical strength and phase stability in the fuel cell’s intense working circumstance [36,116,117]. The interconnects used in SOFC operating at a temperature range of 500–10001C typically consist of metals, ceramics, or their composites. A frequently employed ceramic interconnect material is perovskite LaCrO3, which is basically doped with elements, such as Sr and Ca, to accomplish the desired thermal and electrical properties with the phase permanency of the SOFC in extreme working circumstances [36,118]. Nickel and calcium-doped yttrium chromites are other substances that indicate their suitability with YSZ in terms of zirconate configuration in air and resistance to sinterability. But, low thermal and electrical conductivities have been examined in the case of ceramic interconnects. The high manufacturing price of ceramic interconnects confines it to the tubular geometry that necessitates less substance [119]. For this reason, metallic interconnects are thought to be a better choice than ceramic interconnects. The disadvantage of metallic interconnects is that they are exposed to corrosion, cathode pollution, chromium evaporation, and SOFC components due to incompatibility of TEC with other ceramic substances. The most commonly utilized substances with TECs matching ceramic elements are frequently chromite and ferrite stainless steel alloys [120]. Plansee Ducralloy (94 CrSFe/Y2O3) was the most prominent chromium alloy utilized as ceramic material replenishment [120,121]. This alloy exhibits better resistance to oxidation, exhibits good mechanical strength and displays a TEC that matches other components of the SOFC. But, the manufacturing cost for this material is very high. In contrast, even though ferrite stainless steel presents a low production cost with a matching TEC, the oxide characteristics are weak in the cathode circumstances and cause the oxide scale to grow. This growth leads to area-specific high resistance [120]. The vapor phase of the Cr6 þ species reacts with the LSM in the presence of oxygen/water to improve new stages. It also spreads through the electrolyte/cathode interface and deposits chromium and other species and causes destructive cell worsening [116]. Reactive elements, such as oxides, spinel, nitrides, and conductive perovskites, have been proposed by scientists to prevent this protective layer of material. The best of all coatings on ferritic stainless steel is a chrome-free spinel coating (Mn, Co)3O4 [36,120–128].

2.19.2.3.5

Sealing materials

In SOFCs, the mixture of fuel and oxidant is prevented with the aid of sealing materials. Sealing in the tubular geometry of SOFCs is not a big problem. But, it is one of the significant disputes in planar SOFCs. SOFC seals need to have TECs matched to the electrolyte, should not be responsive with the elements of the SOFCs, and have thermal permanency in the SOFC medium. SOFC seals are able to be split into two kinds: compressor seals and rigid seals (glass/ceramic or glass composites). Inexpensive and air-tight seals are usually composite glass ceramic seals, such as CaO–SiO2 and BaO–Al2O3–SiO2 [18,80,129,130]. Furthermore, when contrasted to plain borosilicate-type glasses, these materials exhibit chemical inertness to the elements that contact them directly, and the glass seals function well in the SOFC working medium for short-to-medium-term transactions, but the durability (420,000 operating hours) for long-term operations is still unknown. Moreover, the ability of these materials to provide thermal cycling for the commissioning and shutdown of SOFCs is an important problem. Moreover, in the case of a rigid seal, the entire SOFC stack will turn out to be rigid, so that if the single cell is cracked, the entire stack will turn out to be impractical [80,129]. In other respects, the tightening seal is much more resistant to thermal shocks and remains stable for long periods of operation, but metal seals can cause a short circuit in the seal. In the case of mica sheath gaskets, intensive compaction needs to be hermetic [129]. For this reason, a hybrid approach can be applied, in other words, a solid seal can be used with a compression seal. Hybrid seals [131,132] are used, and minimal leakage is observed with superior thermal cycle performance. It is seen that there is no single approach to the use of sealing technology by R&D groups and companies in the literature [18,36,129,131,132].

2.19.2.4

Applications and Technological Aspects of Solid Oxide Fuel Cells

Generally, the SOFC design for spread power generation involves of three central pieces [133]. These are a fuel processor that transforms hydrocarbon fuels into hydrogen-rich gas, a stack of SOFCs in which electric energy is produced, and a power regulator to convert DC power to available AC power. Because the cell voltage falls by decreasing the fuel density in the cell stack, the SOFC operates with 80–95% fuel use for better energy efficiency, so that the effluent gas leaving a cell stack encloses useful residual fuel [134,135]. Moreover, by reason of the high temperature process and the unalterable electrochemical procedure of the SOFC, unwanted heat is also generated in the cell energy zone [136]. Accordingly, the use of high temperature and residual SOFC fuel is able to further increase the efficiency of the SOFC system. This brings the SOFC system together with other main cycles, such as a gas turbine (GT), for the production of additional electrical energy [137,138]. Additionally, the unwanted heat is able to be used as a heat supply for the system through a series of heat exchangers. Electric heat cogeneration (a combined heat and power system) can also increase SOFC system performance.

2.19.2.4.1

Solid oxide fuel cell cogeneration: Collective heat and power (solid oxide fuel cell-collective heat and power)

The high characteristic of unwanted heat and electric energy can be subjected to cogeneration in SOFC designs. For SOFC-CHP, the heat substance of the output gas from the combustion process of an unemployed fuel leaving a fuel cell stack in the combustion chamber makes progress and delivers to other components in the SOFC design.

Solid Oxides

609

In an applied cell process, beneficial heat from the SOFC is generally used in pieces that preheat the require energy to produce steam and hot water, i.e., preheaters and reformers. Some researchers have concentrated on the combination of SOFC with CHP and investigated different alignments of the SOFC design. Chan et al. [139] studied an SOFC power system fed with hydrogen and methane. The H2 feed SOFC system includes two preheaters, one SOFC stack, and one combustor; while the CH4 feed SOFC design exhibited a marginally more complicated plant consisting of a mixer, an evaporator, two preheaters, an external reformer, a SOFC stack, and an afterburner. In the CH4 feed SOFC design, unreacted fuel from the cell stack was burned in a burnt burner, and the heat produced was provided to the reformer, evaporator, and preheaters. Model consequences indicated that the efficiency of the CH4 feed system was higher than that of the H2 feed system. Fontell et al. [140] have worked on a 250-kW natural gas-fired SOFC plant. The desulfurization piece was involved in a design for the purification of natural gas, which is dangerous to the SOFC stack. In the suggested SOFC design, the output anode gas is utilized for preheating a reforming fuel and then divided into two pieces; while the first piece is burned in the combustor by air through the cathode, the others are recirculated and diversified with the input fuel flow before the prereformate is fed. The consumed gas from afterburner is used to vaporize the water stream and preheat the natural gas feed stream. The results show that the system efficiency can reach 85%, while the heat cogeneration is thought to be strong in the SOFC design. Since hydrogen is able to be generated by a gasification procedure other than classic reforming progressions, Omosuna et al. [141] explored the combination of SOFC with biomass gasification for power and heat generation. Two different systems of 200-kW SOFC-CHP were compared: a hot gas scrubbing first and the other cold gas scrubbing. As stated by the cold progression, the anode and cathode waste are operated to preheat the input anode and cathode currents before burning in the afterburner. On the other hand, high temperature wastes from the SOFC stack are not employed to the system for hot working. The marks display that the system efficiency of the hot progression is higher than the cold development due to better heat controlling and high gasification temperature in the scrubbing process. But, the price of the hot progression is higher than the cost of the cold development. In addition to the use of a CHP concept in SOFC systems, CHP can be designed to combine SOFC with absorption heat and cooling system. Zink et al. [136] offered an integrated SOFC absorption heating and cooling system utilized in construction. The opportunity of the combined scheme to afford heating/cooling and/or hot water for the construction has been debated. The initial energy and mass balance investigation was achieved and the results showed that such a system had electrical energy, heating and/ or cooling capacity for construction, and reported a high total system efficiency of 87% in different approaches. Another use of CHP technology in housing uses is reported by Braun et al. [142]; they assessed the performance of an anode-assisted SOFC with micro-CHP for housing uses. The effects of internal reforming on anode circulation, cathode circulation, oxidant processing, and system recycling and system performance of the fuel type and methane reforming progression have been studied. The results showed that the cathodic and anodic gas circulation reached maximum efficiency together with the internal reforming of methane. The electrical efficiencies of SOFC and SOFC-CHP are 40% HHV (45% LHV) and 79% (LHV 88%), respectively. Recently, Jamsak et al. [143] offered a hybrid procedure involving an SOFC system with ethanol and a distillation column (SOFC-DIS). The design comprises of three heat exchangers (for preheating supplies), a reformer, and an afterburner. The concentration column was integrated with the SOFC design to cleanse the bioethanol to the preferred ethanol concentration and recapture stages. It was burned in the waste incineration furnace from the anode and the cathode, and then a reformer and three heaters were supplied. The net heat from the SOFC system was supplied to the recovery of the condensation row. The combined design makes available maximum electrical energy without requiring an additional heat source when net energy is zero. The effects of ethanol concentration and ethanol recovery on electrical performance in self-sustaining conditions for different fuel use were studied. Total electrical efficiency and power density were 33.3% (LHV) and 0.32 W/cm2 in 41% ethanol, 80% fuel use, and 80% ethanol recapture concentration, respectively.

2.19.2.4.2

Cogeneration of solid oxide fuel cells: Gas turbine (solid oxide fuel cell-gas turbine)

The idea of a GT in combination with an SOFC has been identified for many years as a concept. But, though the pressurized setup of SOFC stacks has been around for a long time, the SOFC/GT combined cycle design has only become practical in recent times [137]. Commonly, GTs can be linked to SOFCs in two different ways: indirect and direct integrations. In the past, the combustion of the GT was substituted by a heat exchanger where the air from the compressor was heated by the fuel cell exhaust. In the indirect SOFC-GT hybrid design, the SOFC can work under atmospheric conditions. Even though the SOFC stack lessens the need for sealing, the heat exchanger must operate at very high temperature and pressure differentials. The material needs in the indirect combination of SOFC-GT are actually a problem and so, it is not commonly used. The compressed air from the compressor is directed to the cathode and responds with hydrogen in the renewed gas by means of electrochemical reactions. Unburned fuel is burned with air that is dumped in the afterburner. The high temperature consumed is carried to operate a GT as a down period. Exhausted heat content can also be provided to other components in the SOFC design. In this method, the SOFC functions at high pressure. While the performance of the SOFC is enhanced because of this high-pressure process, high-pressure gradients between the anode and the cathode are shown. This can cause SOFC fragility [137]. It has been found that the combined SOFC design and GT scheme will increase the total electricity efficiency by up to 70% [144,145]. In addition, the use of SOFC exhaust gas ensures that the conversion of the fuel in the SOFC stack is completed and leads to a reduction in the dimension and price of the SOFC stack [146]. The SOFC scheme is also able to be combined with a steam turbine as a bottom-up cycle to generate additional energy. Even though the SOFC and steam turbine hybrid system are less complex it proposes lower efficiency than the SOFC-GT system [137].

610

Solid Oxides

In the past few years, there have been a number of reports dedicated to researching and improving the performance of SOFC-GT hybrid systems with different configurations. A SOFC-GT system with external prereforming and anode gas recycling was investigated by Palsson et al. [145]. Steam and heat to the external reformer that was provided by the effluent from the anode was partially recycled. The remainder of the anode waste was burned with the cathode ray in the combustion chamber. The hot gases were then sent to the GT to provide additional power, and the exhaust gas from the gas species was used to preheat the fuel feed stream. They examined the effects of working factors, such as pressure, air flow rate, and fuel flow rate and air inlet temperature on system performance. It has been indicated that the pressure ratio is a big influence on the system performance and that the electric efficiency is higher than 65% at low pressure ratios. Costamagna et al. [147] presented the design and off-design performances of a hybrid system built on the combination of a recovered micro-gas turbine (MGT) with a solid oxide fuel cell reactor (SOFCR) at high temperature. Regulation of an SOFC plant is made up of MGT, centrifugal compressor, flow expander, combustion chamber, recuperator, electric generator, and natural gas compressor with reformer and SOFC stack. The effects of the working line, MGT, and partial load efficiency behavior were studied. The results show that the hybrid system provides thermal efficiency above 60% at the design point and above 50% at the load. Moreover, a combination of SOFC on the net with a GT was analyzed by Selimovic and Palsson [45]. The SOFC-GT system comprised of two subcomponents in which the fuel feed was serially organized and the air fed in series and parallel into the stack. A significant increase in system efficiency was observed when the reactant and air were serially fed to the SOFC stacks. The reason is that the cooling of the cell is better, the bulk of the temperature profile is uniform, and the total fuel consumption is higher. But when the air currents were fed in parallel, the system efficiency decreased by 1.5%. The joint power generation system of SOFC and GT using steam and HR using ASPEN Plus were determined by Kuchonthara et al. [148]. The effect of steam recovery (SR) on total efficiency of the combined system was investigated during heat and vapor recovery by comparing SOFC-GT against the system only during heat accumulation. The system with heat and vapor recovery was found to provide higher overall efficiency than the system with only HR. An optimization of a hybrid system joining SOFC and GT with and without CO2 imprisonment was studied by Möller et al. [149]. Before the prereformer in the SOFC-GT scheme comprising of two SOFC stacks, the desulfurizer unit was placed. The reformed gas was equivalently split between the two stacks and fed anodes in parallel. Stacks on the air side are connected in series. The remaining fuel in the anode outlet gas was burned in the GT burner with additional fuel source. In the SOFC-GT system with CO2 capture capability, the flue gas was cooled and dried in an exhaust gas condenser. The marks present that the SOFC-GT system combined with CO2-capture exhibits an electrical efficiency of over 60%. A model of full and incomplete load exergy examination of a hybrid SOFC-GT power plant was provided by Calise et al. [150] in 2006. At full load, the maximum electrical efficiency value reached 65.4%. Built on the changes in fuel and air mass flow rate, three different partial load strategies were introduced. Model of the incomplete load process of the SOFC-GT hybrid plant has revealed that the best performance is able to be attained if the fuel to air ratio is constant. But, a lower net electric energy value is able to be attained by reducing the fuel flow rate at a constant air flow rate. Haseli et al. [151] studied the performance of an SOFC at high temperature with a predictable curative GT (SOFC-GT) plant. Model consequences demonstrate that rising the turbine inlet temperature improves the net specific power output, while causing the thermal efficiency of the cycle to decrease. Furthermore, a growth in turbine inlet temperature results in a higher rate of entropy formation in the plant. It has been shown that flammable and SOFC mostly influence the entire irreversibility of the scheme. The SOFC-GT plant and the customary GT cycle built on the same working conditions were also compared. It was found that the SOFC-GT plant had a 27.8% higher efficiency than the original GT plant.

2.19.2.4.3

Cogeneration of solid oxide fuel cell: Chemical fabrication method

As an additional method, the SOFC can be utilized for chemical fabrication. Most of the SOFC efforts have pointed at electricity production, but some scientists thought that SOFC was another possible application as an integrated combination of electrical energy production and chemical production. Farr and Vayenas [153] showed such an operation for the first time in the case of NH3 conversion to NO. This scheme comprises of a fuel cell reactor, an external load and a chemical product recovery piece [152]. Even though cogeneration schemes are able to function at low, medium, or high temperatures, operating at high temperatures has many advantages. Recently, several chemical coproduction methods have been proposed for the fabrication of inorganic and organic composites from high temperature SOFCs [153–158]. For this reason, the investigations are mainly concerned with the chemical and electrical power production of high temperature SOFCs. The financial assessment of this cogeneration process has also been observed. The first investigational indication of chemical cogeneration with SOFCs was examined by Farr and Vayenas [153]. They worked on the production of nitric oxide from ammonia oxidation and the recovery of energy as electricity. YSZ was used as an electrolyte placed between porous Pt electrodes. When the cation side is in contact with the air, NH3 is fed to the anode side at a concentration of 4.59% in helium. Oxygen ions move across the electrolyte and respond with NH3. At a temperature range of 427–9271C, NO was acquired as the main anodic product by an optimum efficiency of above 60%. A noteworthy amount of byproduct N2 was created due to the catalytic reaction between NH3 and NO on the Pt electrode. Another interesting chemical cogeneration procedure is hydrogen cyanide synthesis and electrodeposition, metallurgy, and photography, commonly used in fumigation as insecticides and in the synthesis of adiponitrile [154]. The cogeneration system uses methane and ammonia as fuel. Furthermore, the cell was completed with a YSZ tube surrounded in a quartz tube comprising a leaky Pt electrode as the cathode and a leaky rhodium as the anode and the platinum electrode. Reactions (3) and (4) occur in

Solid Oxides

611

the anode and cathode, respectively. Reaction (5) provides the general reaction of this scheme. CH4 ðgÞ þ NH3 ðgÞ þ 3O2 -HCNðgÞ þ 3H2 OðgÞ þ 6e 2

O2 ðgÞ þ 4e-2O

4NH3 ðgÞ þ 5O2 ðgÞ-4NOðgÞ þ 6H2 OðgÞ

ð3Þ ð4Þ ð5Þ

The reactor was operated at 800–10001C and 1 atm. The selectivity of HCN from the studies exceeded 2 to 75% and power density of approximately 0.01 W/cm2 was achieved. It was also discovered that the cell could work deprived of an external heat source so that the CH4/NH3 feed ratio was above 1.25 [155]. In another interesting example, the SO2 oxidation reaction of H2S has been successfully carried out [156]. At 650–8001C and atmospheric pressure, hydrogen sulfide was rusted on a leaky Pt electrode. Hydrogen sulfide weakened in He was utilized as fuel and fresh air was employed as oxidant. The cation reaction is similar to Reaction (4), whereas Reaction (6) shows the anode reaction. Reaction (7) is the general reaction of this scheme. 2H2 SðgÞ þ 3O2 -SO2 ðgÞ þ 2H2 OðgÞ þ 6e

ð6Þ

2H2 SðgÞ þ 3O2 ðgÞ-2SO2 ðgÞ þ 2H2 OðgÞ

ð7Þ

At low current densities, the selectivity to SO2 leftovers under 15% and the elemental sulfur is the key product. But, at high current densities, product selectivity for SO2 is above 90%. Given the construction of organic composites, the styrene is able to be acquired by electrochemical oxidative dehydrogenation of ethylbenzene in SOFC [157], the anode and general reactions being as follows: C6 H5 2CH2 CH3 ðgÞ þ O2 -C6 H5

CH ¼ CH2 ðgÞ þ H2 OðgÞ þ 2e

2CH4 ðgÞ þ 2NH3 ðgÞ þ 3O2 ðgÞ-2HCNðgÞ þ 6H2 OðgÞ

ð8Þ ð9Þ

As stated by the state of this study, styrene, water, and other byproducts were formed. However, although the anodic current and dehydrogenation rates increase, the total transformation is restricted to only 15% or less. Formaldehyde incomplete oxidation of methanol, a vital composite for fertilizers, dyes, disinfectants, germicides and preservatives, is an additional main industrial reaction successfully produced in SOFC [158]. So, leaky Ag was placed on both sides of the YSZ tube base, and the scheme was operated at 547–6971C and atmospheric pressure. High purity diluted methanol was injected into the interior of the YSZ tube via the steam feed. Anode and general reactions are as follows: CH3 OHðgÞ þ O2 -H2 COðgÞ þ H2 OðgÞ þ 2e

ð10Þ

2CH3 OHðgÞ þ O2 ðgÞ-2H2 COðgÞ þ 2H2 OðgÞ

ð11Þ

Output power density obtained from this work was about 1 mW cm. Methanol transformation was above 30%, while selectivity to formaldehyde was 85–92%. As for methane exchange, oxidative coupling of methane is characteristically investigated in conventional reactors [159–163]. However, there are very few studies that have applied SOFC to directly convert methane to valuable chemicals. For example, Pujare and Sammells [164] investigated the oxidative attachment of methane to SOFC. Despite the relatively low methane conversion, high selectivity to C2 hydrocarbons (490%) was achieved. Otsuka et al. [165] directed comparable reports on numerous catalysts placed on the Au electrode. They propose that the most energetic and selective catalyst was BaCO3 over Au. Jiang et al. [166] studied the oxidative bonding of methane to ethylene and C2 hydrocarbons and stated methane conversion at 8001C up to 97% yield values to 88%. In their work, two types of anodes were employed, namely leaky Ag film and leaky cermet, including CaO and Ag doped Sm2O3. Eqs. (12) and (13) describe the anode and general reactions of the above scheme. 2CH4 ðgÞ þ 2O2 ðgÞ-C2 H4 ðgÞ þ 2H2 OðgÞ þ 4e

ð12Þ

2CH4 ðgÞ þ O2 ðgÞ-C2 H4 ðgÞ þ 2H2 OðgÞ

ð13Þ

Lately, the kinetic factors of the reactions at the anode of the SOFC have been extensively estimated [167] by employing the fuel cell sort temperature encoded desorption method to examine oxygen species species [168,169] and the oxygen infusion under fuel cell processing [170]. The obtained electricity is considered to be a good reactor for the fabrication of the SOFCR C2 for methane oxidative coupling, although it is far from a typical SOFC used to generate electricity [171]. The performance of SOFCR for C2 production has been additionally researched by comparing it with the unoriginal immovable bed reactor, the leaky membrane reactor, and the mixed ionic and electronic conductive membrane reactor. SOFCR obviously developed C2 selectivity compared to the original immovable bed reactor. But, while PMEM was superior to other reactors at low temperatures (o1150K), the mixed ionic and electronic conductive membrane reactor was desirable at high temperatures (41150K) [172]. Ishihara et al. [173] investigated the coproduction of synthase and electricity with SOFC by burning with methane. Syngas can be utilized as a raw material in the production of hydrocarbons and methanol. The maximum power density of 526 mW/cm2 can be attained with a synthesis gas yield of 20%. A comparable study was implemented by Pillai et al. [174]. At 7501C, the cells reported that they generated 0.9 W/cm2 with a methane conversion of about 90% at a flow rate of 30 sccm/cm2. However, methane conversion has been steadily declining in the first 30–40 h of operation, although SOFC electrical performance is stable.

612

Solid Oxides

Even more important, the supplement of a catalyst coating on the anode suggests that it provides more stable methane conversion to the synthesis gas. In line with the financial investigation of SOFC for chemical cogeneration, very insufficient work has been presented until now. Spillman et al. [175] acquired a basic expression to verify the most gainful operation among the several modes: opposite electrolysis, fuel cell only for electricity generation, chemical cogeneration, and advanced electrolysis. On condition that the chemical product is more valuable than electricity, it is concluded that direct electrolysis is more desirable than cogeneration. Vayenas et al. [176] suggested an index for assessing the relative effectiveness of SOFC chemical cogeneration compared to the chemical reactant in the same product volume and configuration. Only the SOFC reactor, which uses cheap resources with an exothermic reaction, may be useful. The results present that H2SO4 and HNO3 cogeneration is probably advantageous in the absence of ethylene oxide and formaldehyde production.

2.19.2.5

Thermodynamics of the Solid Oxide Fuel Cell

Thermodynamics is the hypothetical basis of any electrochemical intensity cell. It determines the essential connection between thermodynamic quantities and electrical quantities; this connection not only helps you determine the thermodynamic properties of substances with the right electrochemical approaches, but also describes the maximum cell voltage and concentration, temperature, and pressure dependence of a given chemical reaction. In addition, volume changes due to changes in temperature, oxygen stoichiometry, and incomplete oxygen pressure are able to be comprehended with thermodynamic principles. The laws of thermodynamics are employed to clarify the electrochemical and mechanical behavior of mechanisms of SOFCs.

2.19.2.5.1

Electromotive force and Gibbs free energy change (∆G)

Once a system enters a rescindable progression underneath isothermal and isobaric circumstances, the reduction in Gibbs free energy (G) of the system is equal to the work (W) completed by the system (excluding expansion work). Such a progression increase is symbolized by [177] 2dG ¼ δW

ð13Þ

For a galvanic cell system, such as a SOFC, the work, δW, is carried out by moving an electric current through a voltage difference, i.e., from an electric potential resulting in the formation of a chemical reaction to the other. The electrical effort completed by the galvanic cell is the product of the charge moved and the electric potential alteration, ∆(j) (volts), and the units are joules. If the dn moles of the valence z ions are carried by the voltage alteration held between the electrodes of a cell, then δW ¼ zF  DðjÞ  dn

ð14Þ

where, F is Faraday’s constant, 96,485 C/mol. If the transfer is reversed, the electromotive force (EMF), E, is called the electrical potential alteration between the electrodes of the cell, δW ¼ zF  E  dn ¼ 2dG

ð15Þ

For the transport of 1 mol of ions, Eq. (18) becomes DG ¼ 2zFE

ð16Þ

Eq. (16) is identified as the Nernst formula and is the basis for the calculation of the E value for a given chemical reaction. Under the Raoultian standard state, Eq. (16) is modified as DG0 ¼ 2zFE0

ð17Þ

0

here, E denotes the EMF under the typical state. The typical state of an element is an orientation state in which the element is compared only in any other case; which means that any situation can be chosen as the typical case and the choice is usually made only on the basis of convenience. For gas systems, the standard condition is customarily taken as pure material at ideal gas circumstances, 1 bar pressure and system temperature. For the condensed substance system, the typical state is taken as pure substance at system temperature and 1 bar pressure.

2.19.2.6

Effect of Concentration of the Electromotive Force

For an overall chemical reaction: aA þ bB ¼ cC þ dD

ð18Þ

here, reactants and products do not appear in the typical states. DG ¼ DG0 þ RT ln

acC adD aaA abB

ð19Þ

And, from Eq. (16), the above reaction is the EMF of the cell in which the reaction occurs electrochemically E ¼ E0 þ

RT acC adD ln zF aaA abB

ð20Þ

Solid Oxides

613

1.100 CH4(g) + 2O2(g) = 2H2O(g) + CO2(g) 1.050

1.000

C(s) + O2(g) = CO2(g)

E° (V)

H2(g) + 0.5O2(g) = H2O(g) 0.950 CO(g) + 0.5O2(g) = CO2(g)

0.900

0.850 500

600

700

800 t (°C)

900

1000

1100

Fig. 3 E0 is a variety of electrooxidation as a function of the temperature of the fuel.

where E0 ¼ of H2.

DG0/zF. To yield an example, think through the singular electrode and general electrochemical oxidation reactions Cathode: 0:5O2ðgÞ þ 2e ¼ O2 Anode: H2ðgÞ þ O2 ¼ H2 OðgÞ þ 2e Total reaction: H2ðgÞ ðIn fuelÞ þ 0:5O2ðgÞ ðin oxidantÞ ¼ H2 OðgÞ ðin fuelÞ

ð21Þ

Eq. (23) for this reaction with z¼ 2 for system pressure Pt ¼ 1 atm. E ¼ E0

RT aH2 O ¼ E0 ln zF aH2 a1=2 O2

RT PH2 O ln zF PH2 P 1=2

ð22Þ

O2

It should be noted that DG0 of a chemical reaction is usually present in the thermodynamic database; for this reason the cell E0 is known. The temperature-dependent changes of E0 due to electrooxidation of a galvanic cell are shown in Fig. 3. As has been pointed out, the E0 of CH4 and C oxidations are virtually free of temperature, while H2 and CO oxidizations produce a more sensitive E0 against temperature. The remark infers that CH4 and C can be more appropriate as fuels for low temperature SOFCs [178]. In a nonstandard case, a given fuel can be calculated by a condition known in E of the oxidation reaction. Fig. 4 demonstrates P an instance of how the E of the H2 electrooxidation reaction alters with T and logarithm of PHH2 O , with air as oxidant P ¼ 0.2 atm. The E equation is advanced from the thermodynamic data as a function of T and E ¼ 1275

0:3171T

0:0430T ln

PH2 O PH2 .

P H2 O ðmV Þ PH2

2

ð23Þ

P

As expected, E is reduced by PHH2 O and T. In addition, Fig. 4 also shows the ENi (Ni oxidation EMF (air charge)), which represents 2 the Ni–NiO balance performed by Ni. According to the plot, for example, a mixture of 99.9% H2O–0.1% H2 will be required to oxidize Ni metal at 10001C. Given that both Ni and NiO are solids, the degree of freedom coexists under isobaric situations in line P with the Gibbs phase regulation. Hence, ENi is T dependent only for a constant oxidant, such as PO2 ¼0.21atm, but PHH2 O is 2 independent. In an SOFC, if the oxidation of Ni occurs, the cell voltage will not change at ENi. Likewise for CO fuel, the E equation PCO2 is inscribed from the thermodynamic data as a function of T and PCO . E ¼ 1462

0:4830T

0:0430T ln

PCO2 ðmV Þ PCO

ð24Þ

P

2 at three representative temperatures. Fig. 5 shows E together with Ni oxidation potentials as a function of PCO CO P 2 PH2 O at a certain temperature, suggesting that a CO2–CO mixture has lower than As can be seen, oxidation of Ni requires PCO PH2 CO more oxidation ability than an H2–H2O mixture.

2.19.2.6.1

Heat effects in a galvanic cell

The alteration in the enthalpy of a system is equal only when the volume change job is the only work completed by the system, while the heat in the system going in or going away from the system during a steady pressure process is equal. When an electrochemical reaction is carried out in a galvanic cell, the consequence is electrical processes, DHaq. For a variation of state at

614

Solid Oxides

1200

E

1000

E (mV)

800

1000 °C 900 °C 800 °C

600

Ni oxidation EMF (vs air) 400

300

0 0.1

1

10

100

1000

100

1000

PH2O /PH2 (atm) Fig. 4 Differences of E of H2 O–H2 fuel with temperature and H2 O/H2 ratio.

1200

1000

E (mV)

800

E 1000 °C 900 °C

800 °C

600 Ni oxidation EMF (vs air)

400

300

0 0.1

1

10 PCO2 /PCO (atm)

Fig. 5 Disparities of E of CO2–CO fuel with temperature and CO2/CO ratio.

constant temperature and pressure: DG ¼ q

w þ pDV

TDS ¼ q

0

q ¼ DG þ TDS ¼ DH if w ¼ 0. If the progression comprises the process and in the case of reversed intensity cells, then Eq. (15)

w0

TDS

w0 ¼

ð25Þ

w0 max ¼ DG, and if the process is reversed as

q ¼ TDS DS (T) depending on the temperature of the chemical reaction can be acquired from: Z T DCP dT DSðT Þ ¼ DS298 þ 298 T

ð26Þ

ð27Þ

Solid Oxides

615

20 C(s) + O2(g) = CO2(g) 0

∆ S° (J/(K mol))

CH4(g) + 2O2(g) = 2H2O(g) + CO2(g) −20

−40 H2(g) + 0.5O2(g) = H2O(g) −60

CO(g) + 0.5O2(g) = CO2(g)

−80

−100 850

950

1050

1250

1150

1350

T (K) Fig. 6 The entropy variations of fuel oxidation reactions as a function of temperature.

DS298 is the entropy alteration of the reaction at 298K; DCp is the alteration in steady pressure molar heat capacity of the reaction. For each element incorporated into the reaction, Cp is naturally: CP ¼ a þ b  T þ c  T 2 þ d  T 3 þ e  T Substitution of Eq. (31) into Eq. (30) provides:
 T DSðT Þ ¼ DS298 þ Da  ln þ DbðT 298

1  298Þ þ Dc T 2 2

2

 1  2982 þ Dd T 3 3

ð28Þ

2983




1 1 De 2 2 T

1 2982



ð29Þ

here, a, b, c, d, and e, and S298 are the coefficients for each constituent of a fuel oxidation reaction. The changes of DS0 with T for the fuel electrooxidation reactions with Eq. (29) are depicted in Fig. 6. The zone between the T axis and each line exemplifies the heat related to the electrooxidation reaction. Obviously, both H2 and CO electrooxidations are exothermic (TDS0o0) and CO oxidation releases more heat than H2 oxidation. The amounts of heat connected directly with CH4 and C electrooxidations are less important as proved by very small regions. The difference is that the old reaction is poor exothermic and the second is weak endothermic. More heat output from electrooxidation refers to the conversion of less chemical energy into electrical energy and thus lower electrical activity for H2 and CO fuels. Put differently, C and CH4 may be better fuels to achieve higher electrical efficiency through direct electrooxidation. A mild endothermic effect for the C electrooxidation reaction revealed in Fig. 6 shows that it may require additional energy input to maintain direct electrooxidation of C. With TDS0 present in Fig. 6, heat production rate HPR (J/s) by electrooxidation of fuels: Q0f ðTDS0 Þ ð30Þ 22:4  60 here, Q0f is the mass flow rate in slpm of the fuel for 100% electrochemical application. For a fuel at the Uf flow rate to be partially used by a galvanic cell, the HPR can be rewritten as HPR ¼

HPR ¼

2.19.2.6.2

 Qf  Uf  TDS0 22:4  60

ð30aÞ

The temperature coefficient of the electromotive force

Eq. (16) is the derivative of DG with regard to the temperature at a steady pressure,

 ∂DG ∂E ¼ zF ¼ DS ∂T P ∂T P

ð31Þ

Thus, for the cell reaction


∂E ∂T



P

¼

DS zF

ð31aÞ

It changes the enthalpy of the reaction DH ¼

zFE þ zFT




∂E ∂T



P

ð32Þ

616

Solid Oxides

0.10 C(g) + O2(g) = CO2(g) 0.00

(dE°/d T )P (mV/K)

CH4(g) + 2O2(g) = 2H2O(g) + CO2(g) −0.10

−0.20 H2(g) + 0.5O2(g) = H2O(g) −0.30

−0.40 CO(g) + 0.5O2(g) = CO2(g) −0.50 850

950

1050

1150

1250

1350

T (K) Fig. 7 Temperature coefficients of E 0 of different electrooxidation reactions of fuels as a function of temperature.

 0 The execution of ∆S0 to Eq. (31a) of the electrooxidation reactions shown in Fig. 6 gives Fig. 7, where ∂E ∂T P is plotted against temperature. ∂E0  0 electrooxidations are negative It is clear that the E0 values of H2 and ∂T P whereas the E values of CH4 and C are virtually ∂ECO ∂E0   0 0 free of temperature or a small positive ∂T P . Even though allfuels seem insensitive to temperature change, the size ∂E ∂T P is  ∂T P ∂E0 obviously different for each fuel. Higher absolute values of ∂T P for CO and H2 radiate more heat and less chemical energy is  0 . Such behavior can also converted to electrical energy. Alternatively, E0 for C electrooxidation reaction has a slightly positive ∂E ∂T P be utilized to assess whether carbon formation occurs for a given hydrocarbon fuel as working conditions in an SOFC change.

2.19.2.6.3

The pressure coefficient of the electromotive force

The variation according to Eq. (16) at a constant temperature gives:

 ∂DG ∂E ¼ zF ¼ DV ∂P T ∂P T

ð33Þ

Rearrangement of Eq. (33) provides,


∂E ∂P



¼

T

DV zF

ð34Þ

This indicates that a cell E increases in the place where volume is reduced in the reaction. In the reaction [20], the value E of the H2 oxidation reaction is increased with respect to the volume reducing reaction result. Assuming the ideal gas behavior for all relevant gases, the partial molar volume V A of the component A in the mixture (e.g., A ¼ H2, H2O, and O2): VA XA RT VA ¼ P ni ¼ PA

ð35Þ

And the general change in the molar volume of Reaction (21) DV ¼ VH2 O

VH2

1 XH2 O RT VO ¼ 2 2 PH2 O

XH2 RT P H2

1 XO2 RT RT ¼ 2 P O2 Pt

RT Pt

1 RT 1 RT ¼ 2 Pt 2 Pt

ð36Þ

Integration of Eq. (34) from atmosphere P0 to an elevated pressure Pt (z ¼2) provides the boost ∆E in E for the H2 oxidation reaction DE ¼ EðP; T Þ

  RT Pt ln E P0 ; T ¼ 4F P 0

ð37Þ

Obviously, high T and Pt values support further development in E. The whole equation of E as a function of T, concentration, and P for Reaction (21) can then be described from Eqs. (23) and (36) as E ¼ 1275

0:3171T

0:0430T ln

P H2 O Pt þ 0:0215T ln 0 PH2 P

ð37Þ

Solid Oxides

For CO fuel, the improvement expression of E is identical to Eq. (37), and so the whole equation of E with T, variables is resulted from Eq. (24) as

 PCO2 Pt E ¼ 1462 0:4830T 0:0430T ln þ 0:0215T ln 0 PCO P

PC02 PCO ,

617

and Pt as

ð38Þ

For direct electrochemical oxidation of C and CH4 fuels, the system pressure Pt ought to have no influence on the value of E because the volumes of the reaction products and the products participating in the reaction stay unaffected. Nevertheless, if the most important fuels are converted to plain fuels, such as H2 and CO, the Pt effect on plain fuels should be well thought-out instead. For this reason, the two basic equations used are to calculate the E values of hydrocarbon fuels under different circumstances.

2.19.2.6.4

The thermal and chemical expansion coefficients

The dimensional stability of SOFC elements is a vital principle to consider when developing a trustworthy and vigorous SOFC product. Thermal expansion and chemical expansion coefficients, both material properties, are two main aspects affecting dimensional stability. The first is employed to label the linear volume change of a substance with temperature, and the second describes the change in volume of a substance by reason of variations in the chemical composition instigated by the surrounding atmosphere. The descriptions of thermal and chemical expansion coefficients appertaining to thermodynamics are discussed. For a multicomponent oxide system in which the molar volume V of one component is a function of temperature T, system pressure P, and oxygen nonstoichiometry δ V ¼ ðT; P; δÞ

ð39Þ

The common differentiation in Eq. (39) dV ¼




 ∂V ∂V ∂V dT þ dP þ dδ ∂T P;δ ∂P T;δ ∂δ T;P

ð40Þ

Rearrangement of Eq. (43) by dividing by V provides


 1 ∂V 1 ∂V 1 ∂V d lnV ¼ dT þ dP þ dδ ð41Þ V ∂T P;δ V ∂P T;δ V ∂δ T;P       1 ∂V 1 ∂V where at ¼ V 1 ∂V ∂T P;δ is expressed as the TEC, g ¼ V ∂P T;δ is the compressibility, and ac ¼ V ∂δ T;P is the chemical expansion coefficient. For a very rigid SOFC system g is insignificant for most of the relevant system pressure range. Development in each of three 3D directions can be determined by considering the relationship of molar volume V to the strain tensor e. For a differential change in expansion, tr ðdeÞ ¼ d lnV is estimated. For this reason, in the nonappearance of mechanical stress a uniaxial strain Eq. (41) is given for an isotropic solid [179]. de ¼

1 1 1 tr ðdeÞ ¼ at dT þ ac dδ 2 3 3

ð42Þ

Think through the total derivative of the strain de relating to temperature T at constant PO2


 ∂e 1 ∂ ln V 1 1 ∂δ ¼ ¼ at þ ac ¼ TEC ∂T PO 3 ∂T 3 3 ∂T PO PO 2

2

ð43Þ

2

Eq. (43) symbolizes an accurate explanation of the TEC. It reveals that under a constant PO2 , the slope of the e–T plot  is ∂δ composed of two terms: (1) “ordinary” thermal expansion (1/3a) and (2) thermally persuaded chemical expansion 1/3ac ∂T . PO2 The second term can become reliant on the thermal properties of the nonoxygen stoichiometric δ at the higher temperature materials that are becoming increasingly widespread. Consequently, the thermal expansion curves of perovskite oxides, which are powerfully reliant on the temperature of the oxygen stoichiometry, show a curvature (deviation from straightness) at higher temperatures. An instantaneous substance is La1–xSrxCoO3 δ (δ40) and here δ changes powerfully with temperature. Alternatively, the total derivative of the expansion with regard to PO2 at constant temperature T,

 ∂e 1 ∂δ ¼ ac ð44Þ ∂PO2 T 3 ∂PO2 T is frequently identified as the isothermal chemical expansion. As can be understood from the equation, the volumetric change of the material may be either shrinkage or expansion, based in part on how the nonoxygen stoichiometry δ changes with PO2 . Fig. 8 indicates the measured thermal expansion curves for the perovskite La0.6Sr0.4Co0.2Fe0.8O3–δ (LSCF) at various PO2 values to graphically illustrate the relationship described in Eq. (44). As has been pointed out in Fig. 8, the volume of the sample is actually wider with the decrease in PO2 . A spin-state transition of a Co3 þ ion, on the other hand, may also contribute to the expansion change. In summary, the frequently utilized TEC of oxides comprises two expressions; one is instigated by the alteration in temperature and the other by consequences from the change in oxygen nonstoichiometry (δ) induced by temperature, PO2 , and/or in the case of cobalt, a change in the spin state of the ion. For δ-sensitive materials, the dimensional variation measured by increasing the

618

Solid Oxides

0.1 bar 13

0.01 bar

12

Chemical expansion

10−4 bar

Uniaxial expansion  (103 ppm)

11

Return to 1.0 bar

10−3 bar

10

Hold at 792 °C

9

Linear ramp 25−792 °C PO2 − 1.0 bar

8 7

Thermal expansion

6 5 4 3 2 1 0 0

1000

2000 3000 Time (min)

4000

5000

Fig. 8 Graph of thermal expansion and chemical expansion for La0.6Sr0.4Co0.2Fe0.8O3–δ (LSCF).

temperature is provided as a function of temperature as a linear deviation. Under isothermal circumstances, the volumetric change of material upon change in PO2 is comparable to ∂P∂δO . 2 Understanding isothermal exchange in δ with PO2 in cathode materials is vital for cathode-supported SOFCs, since the cathode/ electrolyte interface PO2 is significantly reduced by the charging current. If such a change is not appropriately handled, great stresses can occur throughout filling and the totality of the cell may become problematic.

2.19.2.7

Advantages and Limitations of Solid Oxide Fuel Cells

SOFCs have several benefits: they can be sectional, and they can be deployed to eliminate the requirement for diffusion lines, run inaudibly, and vibrate. SOFCs are more able to afford higher system efficiency, advanced power density, and straightforward designs than fuel cells built on liquid electrolytes. They can contest with the joint cycle GT for disseminated uses at a reasonably low cost. The high cell functioning temperature allows for high remote output and thus simplifies rapid electrode kinetics and activation polarization. This is particularly valuable since valuable platinum electrocatalysts are not needed and the electrodes are not able to be infected with CO. Consequently, CO is a potential fuel in SOFCs. In addition, functioning temperatures rise as high as possible, so performance problems relate to osmotic losses due to load transfer over components and component interfaces, not kinetic (activation overpotentials) [180]. Advantages of SOFCs consist of the following:

• • • • •

Energy safety: reducing oil depletion, cutting oil imports, and growing the country’s current electricity needs. Trustworthiness: up to 90% working time and up to 99.99% power. Low functioning and care costs: the efficiency of the SOFC system can greatly decrease the energy costs and reduce repair prices compared to other options. Fixed power generation: it yields continuous power unlike backup generators, diesel engines, or uninterruptible power supply (UPS). Fuel selection: permits choice of fuel; natural gas, propane, butane, methanol, and diesel fuel can be extracted from hydrogen.

Until now, SOFCs were operating at 10001C, which is the most fuel-efficient. Regrettably, the high temperature abbreviates the cell lifespan and raises the material fee, since expensive high temperature alloys are employed to accommodate the cell and costly ceramics are utilized for the connections, which significantly increases the cost of the fuel cells. Low working temperature is universally accepted as a crucial subject for low-cost SOFCs. Lowering the temperature will permit the usage of low-priced internal connections and structural elements, such as stainless steel. A lower temperature provides the system’s effectiveness and a lessening in thermal strains in dynamic ceramic structures, extending the system’s expected lifetime and making it conceivable to employ low-priced interconnect materials, such as ferritic steels, without LaCrO3 protective coatings. Over the years, scientists and researchers around the world have been conducting research to lower the functioning temperatures of SOFCs without

Solid Oxides

619

compromising their performance. The SOFC’s operating temperature of 600–10001C requires considerable commissioning time. The performance of the cell is very perceptive to the working temperature. A 10% decrease in temperature drop causes a 12% decrease in cell performance due to a rise in internal resistance to oxygen ions [181]. Furthermore, high temperatures require the system to include major thermal protections to protect workers and conserve heat.

2.19.2.8

Environmental Impact of Solid Oxide Fuel Cells

As systems for electricity generation, efficiency and ecology combine to surpass SOFCs. In recent years, public utilities have been receiving considerable attention in the industry, especially in the coproduction of heat and energy. The ecological effect of SOFC usage rests on the hydrogen-rich fuel source utilized. If pure hydrogen is employed, the fuel cell has almost no emissions apart from water and heat. Hydrogen is seldom exploited for storage and carriage problems, but some researchers foresee the development of a solar hydrogen budget in the future. In this situation, photovoltaic cells convert sunlight into electricity. This electricity will be utilized to separate water (via electrolysis) into hydrogen and oxygen to store solar energy as hydrogen fuel [182]. In this state, the SOFCs producing the stations will not have a greenhouse. It is thought that mostly atmospheric emissions are left by a fuel cell power plant during the fueling phase. Nevertheless, the high efficiency of the SOFC gives rise to less fuel expended to yield a certain amount of electricity, which means a lower CO2 emission, the main greenhouse gas responsible for global warming. When hydrogen from natural gas is utilized as a fuel, there is no net emission of SOFCs as the unconfined carbons are taken up by photosynthetic plants from atmospheric light. CO2 emissions can be reduced by more than one million kilogram per year. In addition, the emissions from SOFC systems will be very low with almost zero NOx, SOx, and particulate levels, thus eliminating 20,000 kg of acid rain and smoke-induced contaminants. Regardless, SOFCs afford the lowest emissions of nonrenewable energy generation methods, for instance conventional thermal power plants, as indicated in Table 4 [182]. This is very vital in terms of environmental concerns about energy. When joined with a heat engine that utilizes any waste heat, SOFCs are the cleanest and most effective devices that can be used for this purpose. Furthermore, the SOFC is able to deliver high-quality waste heat that can be employed to heat homes or afford cooling and air conditioning without damaging the environment. Armstrong stated that the single emission is steam, less nitrogen oxide and sulfur oxide, and less CO2. If CO2 can be detached from the source for destruction in another place, SOFCs will indeed come to be the ultrahigh efficiency, zero-emission power plant of the 21st century. A specific focus on SOFC-associated ideas was made in April 1999 by an initiative of the NEDO international combined research program. The key impartial of this combined international research team, which is one of the most protuberant substances in the chemical and physical sciences of Korea and functioning at lessened temperatures 500–6001C. The purpose of the program was to exhibit a zero-emission presenter unit by the end of the 3-year period and expose new study parameters for an ecologically friendly energy production system.

2.19.2.9

Case Studies on Solid Oxide Fuel Cells

Researchers focused on SOFCs, whose efforts have been commercially viable for many years, are working together on the growth of appropriate resources and on the production of ceramic structures, which are now the major technical challenges facing SOFCs. In Japan and the United States, programs are being developed that employ a moderately simple ceramic process to grow a thin-film electrolyte that reduces cell resistance, both of which double the power output and significantly reduce the cost of SOFCs. There is a current effort to integrate SOFCs and develop new stack geometry. Indication of low-temperature SOFC directly on methane gives a significant new opportunity signal to make simple, cost-effective power plants [183]. Global SOFCs endures to make noteworthy enhancements in its simple fuel cell design. The degree of accomplishment is a 48.6% upgrading in single cell power densities representing the maximum available power density for profitable-sized SOFCs in the world (see Table 5) [183]. Variations in cell configuration and design have caused this power density to increase. Higher power densities contribute to the weight, dimensions, and price of fuel cell systems. SOFCs can be appropriate for small-scale housing market uses once the final cost targets of $1000 kW are met. Material cost estimates compiled by the Material Science and Research Organization in Salt Lake City are listed in Table 6 [184]. Table 7 shows the medium-term California Energy Commission results for the implementation of SOFCs with a power density of 300 mW/cm2 targets and a capacity of 50,000 units/year from 2005 to 2010 [185]. Table 4

Characteristic solid oxide fuel cell (SOFC) air emissions

Air emissions

SOx

NOx

CO

Particles

Organic compounds

CO2

Fossil fueled plant SOFC system

12,740 0

18,850 0

12,797 32

228 0

213 0

1,840,020 846,300

Source: Data from the International Fuel Cells, a United Technology Company. Fuel cells review; 2000.

620

Solid Oxides

Table 5 Development of single-fuel solid oxide fuel cell (SOFC) performance determined at Watts/ cm at 0.7 V using H2 as fuel Sort and year

Temperatures 6001C

6501C

7001C

7501C

8001C

G1-1998 G2-1999 G3-2000 G4-2000

– 0.156 0.197 0.200

– 0.318 0.382 0.416

– 0.487 0.635 0.723

0.150 0.528 0.900 1.093

0.250 0.594 1.093 1.216

Source: Reproduced from Global Thermoelectric Inc. Annual report; 2001.

Table 6

Materials price per kilowatt

Component

Material

Cost/kg ($)

Thickness (mm)

Weight (g)

Total cost ($) per kilowatt and component

Electrolyte Anode Cathode Interconnect Total

Yttria-stabilized zirconia (YSZ) Ni–YSZ La1 xSrxMnO3 (LSM) Metallic alloy

10 15 25 15

10 1500 100 125

12 1365 60 200 1631

0.12 20.50 1.50 3.00 25.12

Source: Material Science and Research Inc. Fuel cell seminar, Portland, OR. October 30–November 02, 2000; 2000.

Table 7

Preferred performance results and elasticity purposes for solid oxide fuel cell (SOFC) systems

Parameters

Results

Elasticity purposes

Observations

Capital price, installed ($/kW) Power degradation Power density (mW/cm2)

800 o1% per 1000 h 300

400 o0.5% per 1000 h 500

2005–10 at 50,000 units/year For years 2005–10 44 Cell stack and 425 cm2 electrode

Source: California Energy Commission. California renewable technology market and benefits assessment: Final report; 2001.

2.19.2.10

Future Directions and Prospects for Solid Oxide Fuel Cells

Numerous research works have been completed to advance the high performance of the SOFC since Weissbart and Ruka [186] published their scientific paper in 1962. In recent times, SOFCs have been industrialized to generate power. A group of Dutch and Danish researchers have performed testing analysis of a 100-kW Siemens Westinghouse (SWH) SOFC power generation system beginning in December 1997 [187]. The system utilizes 1152 sustained tubular type air electrodes. The 100-kW SOFC has accomplished electrical generation productivity of 46% and at net system output of 106 kW to the grid. Below the standard process situation of 10001C, 80% fuel operation is accomplished. Emission capacities verified a NOx level below 0.2 ppm with imperceptible traces of SOx, CO, and unburnt hydrocarbons. At present, the scheme has been in process for 3700 h at the examination site, and apart from one stack, the parts of the SOFC generator have functioned continuously. The price of this scheme adds up to $10 million. Some companies, such as Mitsubishi Heavy Industry (MHI) and Electric Power Development in Japan [188] have industrialized an additional sort of tubular SOFC. A forced 10-kW type segment was verified that services 414 cells and functions at 9001C at 5 atm. A 15-kW output power was attained at 200 mA/cm2. The segment has functioned for up to 5000 h. More than a few sorts of planar SOFCs have been exploited. SWH in Germany activated a 10-kW group planar structure of SOFC [189]. The compartments were attached between bipolar shields. A mount with 80 compartment layers was accumulated, each layer involving 16 collinear compartments. Nevertheless, at this point in time, SWH has elected not to improve a planar SOFC and is at present engaged on a tubular SOFC. The Chubu Electric Power Company and MHIs have been assessing a 5-kW group planar sort of SOFC [190]. The compartments have utilized a ceramic interconnect of doped LaCrO3. The sustenance layers were invented with the identical substances by way of the electrodes and the layers were crenelated to permit the gas to move across. A power density of 0.22 W/cm2 with fuel exploitation of 38% was attained. Ceramic fuel cells (CFSs) in Australia [191] industrialized a 5-kW planar sort of SOFC with stainless steel interconnects. The 400 compartments were invented from 10 cm  10 cm  100 mm 3YSZ electrolyte layers manufactured by tape-forming. S. Hexis in Switzerland [192] has established a planar type ceramic–metal hybrid mount with metallic interconnects. The existing position of the schemes and number of kilowatts of SOFCs are listed in Table 8. At this point in time, the SWH type is the most enhanced SOFC. A huge amount of air electrode maintained compartments have been electrically verified for up to 10,000 h. The functioning devolution was less than 0.2% per 1000 h. At present, the efficiency of electric power output is 46%, a figure more advanced than that of previous electric power plants. Nevertheless, for applied uses in electric power plants of more than a few megawatt or greater, the system needs additional enhancements. The most vital fact is the price. The price of SWH’s system can be almost

Solid Oxides

Table 8

621

Mount performing a huge balance of solid oxide fuel cell (SOFC)

Designer

Sort

Power (kW)

Interconnect

Observation

Siemens Westinghouse (SWH) SWH Mitsubishi Heavy Industries (MHIs) and Electric Power Development Co. SWH MHIs and Chubu Electric Power Co.

Tubular Tubular Tubular

10 25 15

LaCr0.9Mg0.1O3 LaCr0.9Mg0.1O3 NiAl/Al2O3

Energy conversion efficiency: 46% Working hours: 1300 Working pressure: 5 atm

Planar Planar

10 5

Cr with 5% Fe and 1% Y2O3 Doped LaCrO3

5.4 kW at 9501C, 4.1 kW at 8501C Energy conversion efficiency: 22%

$100,000 kW. The price of predictable electric power must be decreased to below $3000 kW. Numerous tactics are conceivable; for example, utilizing mixed lanthanides in place of high purity element oxides and approving low-cost deposition procedures for compartment production. Toto and Kyushu Electric Power in Japan industrialized an SWH-type tubular SOFC by wet progression method [193]. A thick–thin YSZ film was made on a leaky sustained (La, Sr)MnO3 tube by slurry covering. The (La, Ca)CrO3 was carried out to develop an interconnecting band by slurry covering. Effective results were attained by this technique for a 1.3-kW mount. Furthermore, intermediate gauges of SOFCs are appealing for dispersed power generation. For this use, it is necessary to address several specific necessities: they must be easy to maintain, have high energy conversion efficiency, and they must be cheap. Maintenance-free process is the greatest benefit of SOFC systems. Nevertheless, GT systems are durable candidates. The entire energy productivity of a GT for the combined heat and power generation is up to 60% and price is likely to be below $3000 kW. A minor gage of the SOFC has been established for electric vehicle (EV) applications. The process conditions of fuel cells for the use of EVs vary noticeably from those of immobile applications. For the EV, the fuel cells have to need adequate power density and be satisfactorily low-cost to parallel well with inner burning apparatuses on a financial base. The ideal fuel cell technology for EV applications is a type of polymer electrolyte. It employs hydrogen or rehabilitated hydrocarbon as a fuel and is very susceptible to CO poisoning. Nevertheless, the moderately high working temperature of existing designs would necessitate additional time and energy for operating. By decreasing the functioning temperature to 5001C, these difficulties can turn out to be controllable. In 1999, Murry et al. [194] described the opportunity of a direct-methane fuel cell with 8 m/o Y2O3 doped ZrO2 electrolyte. The electrolyte was made by sputter method for growth of the electrolyte on a leaky La0.8Sr0.2MnO3 cathode. The anode was leaky Ni–YSZ covered Y2O3 doped CeO2. The power density of 0.37 W/cm2 was attained at 6501C. Once Ni substances are employed with hydrocarbon support bonds, they are disposed to deactivation over the construction of carbonaceous places on the surface of the substance. An important result was the nonexistence of carbon deposits in the temperature range studied. Moreover, Baker et al. [195] concluded that no carbon deposits were perceived on the La0.8Ca0.2CrO3 anode throughout direct oxidation of CH4 in the temperature range 400–6001C. These results suggest opportunities for SOFCs in EV applications in the near future.

2.19.3

Conclusions

SOFC technology proposes a cost-efficient and eco-friendly power generation source. Furthermore, the SOFC eliminates the necessity for an expensive metal catalyst and permits internal improvement of fuels, for that reason reducing the price of fuel cell. One more benefit of the SOFC is the adaptability of the fuels that are able to be employed by it. A noteworthy effort has been completed during the last decade to achieve the highest power densities. Various advanced resources have been recognized and verified to construct profitable and enduring fuel cells with the purpose of building SOFCs suitable for commercialization. In addition, these materials substitute for the usual materials in the situation when the SOFC utilizes hydrocarbon as fuel. As well as the materials, various schemes and manufacturer technologies are developing with the intention of addressing the necessity of price decrease for SOFC commercialization. Even now there are numerous difficulties and rigors to face, and this subject may take years to mature.

Appendix 1:

Thermodynamic Data of Selected Chemical Reactions and Substances

Table A1 lists the standard Gibbs free energy changes for selected reactions in the form DG0 ¼ A þ BT Or DG0 ¼ A þ B þ log T þ CT And it lists the temperature ranges for which it is valid.

622

Solid Oxides

Table A2 presents the constant-pressure molar heat capacities in the form Cp ¼ a þ bT þ cT 2 þ dT 3 þ eT

2

ðJ=ðmol KÞÞ

And it includes the temperature ranges at which the expressions are valid. Table A3 lists standard molar heat of formation and molar entropies for selected substances. Table A1

Standard Gibbs free energy changes of selected chemical reactions

Chemical reaction

G0 (J/mol)

T range, (K)

z*

H2(g) þ 0.5O2(g) ¼H2O(g) C(s) þ 0.5O2(g) ¼ CO(g) CO(g) þ 1/2O2(g) ¼CO2(g) 3H2(g) þ CO(g) þ 2O2(g) ¼CO2(g) þ 3H2O(g) C(s) þ O2(g) ¼CO2(g)

G0 ¼ 246,202 þ 54.758T G0 ¼ 111,606 87.57T G0 ¼ 282,150 þ 86.735T G0 ¼ 1,021,482 þ 251.177T G0 ¼ 393,756 0.836T G0 ¼ 817,106T 51.205 TlogT þ 173.97T G0 ¼706,420 þ 53.922 TlogT 519.574T G0 ¼901,208 173.47T Go ¼ 234,122 þ 85.15T G0 ¼ 166,364 þ 70.56T G0 ¼ 195,206 16.38 TlogT þ 142.54T G0 ¼ 168,078 þ 30.31 TlogT 5.06T

298–2500 298–2500 298–2000 298–2000 298–2000 298–2500

2 2 2 8 4 8

298–1700

4

700–1700 298–1725 298–1356 1356–1503

4 2 2 2

298–1750

2

CH4(g) þ 2O2(g) ¼2H2O(g) þ CO2(g) SiO2(s) ¼Si(s) þ O2 Ni(s) þ 0.5O2(g) ¼NiO(s) 2Cu(s) þ 0.5O2(g) ¼Cu2O(s) 2Cu(l) þ 0.5O2(g) ¼Cu2O(s)

H2(g) þ S(s) ¼H2S(g) z*, Number of electrons transferred during the reaction.

Table A2

The constant-pressure molar heat capacities of some substances

Substances

a

b 103

C(graphite) CH4(g) CO(g) CO2(g) H2(g) H2O(g) N2(g) O2(g) Air

0.10868 0.703029 25.56759 24.99735 33.066178 30.092 26.092 29.659 26.65

38.90326 108.4773 6.096130 55.18696 11.363417 6.832514 8.218801 6.137261 39.4

Table A3

c 106 21.53536 42.52157 4.054656 33.6914 11.432816 6.793435 1.97614 1.18652 25.9

d 109

e 10–6

0 5.862788 2.671301 7.948387 2.772874 2.53448 0.159274 0.09578 6.07

0.147972 0.678565 0.131021 0.13664 0.158558 0.082139 0.044434 0.21966 0.747441

The standard molar heat of formation and molar entropies of some substances at 298K

Substances

fH298

(J/mol)

C(graphite) CH4(g) CO(g) CO2(g) H2(g) H2O(g) O2(g) N2(g)

0 74,870 110,500 393,500 0 241,800 0 0

S298 (J/(mol K)) 5.73 186.01 197.5 213.7 130.46 232.9 205.1 191.5

Symbol

Meaning

Usual unit

ac at g δ e Cp

Chemical expansion coefficient Thermal expansion coefficient Compressibility Oxygen nonstoichiometry Strain tensor Constant-pressure molar heat capacity

K–1 K–1 m2/N None None J/(Kmol)

Solid Oxides dn EMF or E ENi E0 F dG DG DG0 DH HPR O2– O2 P Po PCO PCO2 PH2 PH2 O PO2 Pt q Qf Q DS DS298 T TEC Uf DV VA W δW Xm(XO2 ) YSZ z

Incremental moles of ions Electromotive force (EMF) or Nernst potential Nernst potential of the Ni oxidation volt vs. air Electro motive force at standard volt state Faraday’s constant An increment of Gibbs free energy Change of Gibbs free energy Change of Gibbs free energy at standard state Change of enthalpy Heat production rate Oxide ion Oxygen molecule Pressure Atmospheric pressure Partial pressure of CO Partial pressure of CO2 Partial pressure of H2 Partial pressure of H2O Partial pressure of O2 Partial pressure of system Heat Mass flow rate of fuel Mass flow rate of fuel at 100% utilization Change of entropy Standard-state entropy at 298K Absolute temperature Thermal expansion coefficient Fuel utilization Change of volume Partial molar volume of species A The work (other than expansion) An increment of work (other than expansion) Molar fraction of species m (O2)

mole volt volt volt C/mol J/mol J/mol J/mol J/mol J/s None None atm atm atm atm atm atm atm atm J sccm or slpm sccm or slpm J/(K mol) J/(K mol) K K–1 None cm3 cm3/mol J J None

Yttria-stabilized zirconia Valence

None

623

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22]

Carrette L, Friedrich KA, Stimming U. Fuel cells: principles, types, fuels, and applications. ChemPhysChem 2000;1:162–93. Dyer CK. Replacing the battery in portable electronics. Sci Am 1999;281:88–93. Grove WR. XXIV. On voltaic series and the combination of gases by platinum. Phil Mag Ser 1839;3(14):127–30. Grove WR. LXXII. On a gaseous voltaic battery. Philos Mag Ser 1843;21:417–20. Williams MC. Fuel cells and the world energy future. In: Proceedings of IEEE/PES summer meeting, vol.1, Vancouver, Canada; 2001. p. 725. Ortiz-Rivera EI, Reyes-Hernandez AL, Febo RA. Understanding the history of fuel cells. Available from: http://minds2create.ece.uprm.edu/media/pps/Understanding%20the %20History%20of%20Fuel%20Cells.pdf. Smithsonian National Museum of American History. Fuel cell history project. Available from: http://americanhistory.si.edu/fuelcells/index.htm; 2007. Lipman TE. electric vehicles, direct-hydrogen fuel cell vehicles, and direct-methanol fuel cell vehicles. In: Energy conversion engineering conference and exhibit, (IECEC) 35th Intersociety, vol. 2, July, 24–28 2000; 2000. pp. 1352–8. http://education.lanl.gov/resources/fuelcells. Raissi A. Current technology of fuel cell systems. In: Proceedings of the 32nd conference on intersociety energy conversion engineering (IECEC-97), vol. 3; 1997. p. 1953–7. Available from: https://www.scribd.com/document/57306812/Current-Technology-of-Fuel-Cells-Systems. Laughton MA. Fuel cells. Eng Sci Educ J 2002;11(1):7–16. Narayanan SR, Valdez TI, Kindler A, Witham C, Surampudi S, Frank H. Direct methanol fuel cells-status, challenges and prospects. In: Battery conference on applications and advances, the 15th annual January 11–14 2000; 2000. pp. 33–6. Haile SM. Fuel cell materials and components. Acta Mater 2003;15:5981–6000. Hirschenhofer JH. 1996 Fuel cell status. In: Energy conversion engineering conference, IECEC 96. Proceedings of the 31st Intersociety, vol. 2, August 11–16, 1996; 1996. pp. 1084–9. FuelCellStore. Available from: http://www.fuelcellstore.com/blog-section/intro-fuel-cell-applications. Kreith F, Goswami DY, editors. Handbook of energy efficiency and renewable energy Boca Raton, FL: CRC Press; 2007. Stambouli AB, Travesa E. Fuel cells, an alternative to standard sources of energy. Renew Sustain Energy Rev 2002;6:297–306. Fergus J, Hui R, Li X, Wilkinson DP, Zhang JJ. Solid oxide fuel cells: materials properties and performance. London: CRC Press; 2009. Weber A, Ivers-Tiffee E. Materials and concepts for solid oxide fuel cells (SOFCs) in stationary and mobile applications. J Power Sources 2004;127:273–83. Nancy LG, Dimitrios CP, Joseph MS. Hydrogen and fuel cell technology: progress, challenges, and future directions. Energy Procedia 2012;28:2–11. Adams II TA, Nease J, Tucker D, Barton PI. Energy conversion with solid oxide fuel cell systems: a review of concepts and outlooks for the short- and long-term. Ind Eng Chem Res 2013;52:3089–111. Bhattacharyya D, Rengaswamy RA. A review of solid oxide fuel cell (SOFC) dynamic models. Ind Eng Chem Res 2009;48:6068–86P.

624 [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72] [73] [74] [75] [76] [77] [78] [79] [80]

Solid Oxides Kilner J, Burriel M. Materials for intermediate-temperature solid-oxide fuel cells. Mater Res 2014;44:365–93. Singhal SC. Solid oxide fuel cells for power generation. WIREs Energy Environ 2014;3:179–94. Nernst W. Uber die elektrolvtische Leitung Fester Korper bei sehr hohen Temperaturen. Z Electrochem 1899;6:41–3. Minh NQ. Ceramic fuel cells. J Am Ceram Soc 1993;76:563–88. Möbius H-H. On the history of solid electrolyte fuel cells. J Solid State Electrochem 1997;1:2–16. Mogensen M, Sammes NM, Tompsett GA. Physical, chemical and electrochemical properties of pure and doped ceria. Solid State Ion 2000;129:63–94. Badwal SPS, Giddey S, Munnings C, Kulkarni A. Review of progress in high temperature solid oxide fuel cells. J Aust Ceram Soc 2014;50:23–37. Badwal SPS. Stability of solid oxide fuel cell components. Solid State Ion 2001;143:39–46. Lee JH, Lee HW. High performance anode supported solid oxide fuel cell. U.S. Patent No. 20070015045 A1, 2007. Recent Acres G. Recent advances in fuel cell technology and its applications. J Power Sources 2001;100:60–6. Singhal SC. Advances in solid oxide fuel cell technology. Solid State Ion 2000;135:305–13. Singhal SC, Kendal K. High-temperature solid oxide fuel cells: fundamentals, design and applications. Philadelphia, PA: Elsevier; 2003. Sun C, Hui R, Roller J. Cathode materials for solid oxide fuel cells: a review. J Solid State Electrochem 2010;14:1125–44. Jacobson AJ. Materials for solid oxide fuel cells. Chem Mater 2010;22(2010):660–74. Holtappels P, Vogt U, Graule VT. Ceramic materials for advanced solid oxide fuel cells. Adv Eng Mater 2005;7:292–302. Stolten D. Fuel cell science and engineering: materials, processes, systems and technology. Hoboken, NJ: Wiley; 2012. Timurkutluk B, Timurkutluk C, Mat MD, Kaplan Y. Development of high‐performance anode supported solid oxide fuel cell. J Energy Res 2012;36:1383–7. Badwal SPS, Foger K. Solid oxide electrolyte fuel cell review. Ceram Int 1996;22:257–65. Lee S, Bevilacqua M, Fornasiero P, Vohs JM, Gorte RJ. Solid oxide fuel cell cathodes prepared by infiltration of LaNi0.6Fe0.4O3 and La0.91Sr0.09Ni0.6Fe0.4O3 in porous yttria-stabilized zirconia. J Power Sources 2009;193:747–53. Huang K, Singhal SC. Cathode-Supported tubular solid oxide fuel cells technology – a critical review. J Power Sources 2013;237:84–97. Marozau IP, Kharton VV, Viskup AP, Frade JR, Samakhval VV. Electronic conductivity, oxygen permeability and thermal expansion of Sr0.7Ce0.3Mn1 xAlxO3 δ. J Eur Ceram Soc 2006;26:1371–8. Islam S, Hill JM. Chapter 4 – anode materials development. In solid oxide fuel cells: from materials to system modeling: the royal society of chemistry. London: RSC Publishing; 2013. Roehrens D, Han F, Haydn M, et al. Advances beyond traditional SOFC cell designs. Int J Hydrogen Energy 2015;40:11538–42. Badwal SPS, Deller R, Foger K, Ramprakash Y, Zhang JP. Stability of solid oxide fuel cell components. Solid State Ion 1997;99:297–310. Agnew GD, Collins RD, Jorger M, Pyke SH, Travis RP. The components of a Rolls-Royce 1 MW SOFC system. ECS Trans 2007;7:105–11. EG&G Technical Services. Fuel cell handbook. 5th ed. Morgan, WV: Parsons, Inc.; 2000. Perednis D, Gauckler LJ. Solid oxide fuel cells with electrolytes prepared via spray pyrolysis. Solid State Ion 2004;166:229–39. Santillán MJ, Caneiro A, Quaranta N, Boccaccini AR. Electrophoretic deposition of La0.6Sr0.4Co0.8Fe0.2O3 δ cathodes on Ce0.9Gd0.1O1.95 substrates for intermediate temperature solid oxide fuel cell (IT-SOFC). J Eur Ceram Soc 2009;29:1125–32. Matsuda M, Hosomia T, urata MK, Fukui T, Miyake M. Fabrication of bilayered YSZ/SDC electrolyte film by electrophoretic deposition for reduced-temperature operating anode-supported SOFC. J Power Sources 2007;165:102–7. Taroco HA, Santos JAF, Domingues RZ, Matencio T. Ceramic materials for solid oxide fuel cells. In: Costas S, editor. Advances in ceramics – synthesis and characterization, processing and specific applications. Rijeka: INTECH Open Access Publisher; 2011. Okkay H, Bayramoglu M, Öksüzömer MF. Ce0.8Sm0.2O1.9 synthesis for solid oxide fuel cell electrolyte by ultrasound assisted co-precipitation method. Ultrason Sonochem 2013;20:978–83. Li C, Hu S, Tay K, Fu Y. Electrochemical characterization of gradient Sm0.5Sr0.5CoO3 δ cathodes on Ce0.8Sm0.2O1.9 electrolytes for solid oxide fuel cells. Ceram Int 2012;38:1557–62. Giddey S, Badwal SPS, Kulkarni A, Munnings C. A comprehensive review of direct carbon fuel cell technology. Prog Energy Combust Sci 2012;38:360–99. Zheng Y, Zhou M, Ge L, Li S, Chen H, Guo L. Effect of Fe2O3 on Sm-doped ceria system solid electrolyte. J Alloy Compd 2011;509:546–50. Zhang X, Chan SH, Li G, Ho HK, Li J, Feng Z. A review of integration strategies for solid oxide fuel cells. J Power Sources 2010;195:685–702. Quaschning V. Renewable energy and climate change; CH-13. Hoboken, NJ: JohnWiley Sons & Ltd.; 2010. Guo Y, Shi H, Ran R, Shao Z. Performance of SrSc0.2Co0.8O3 δ þ Sm0.5Sr0.5CoO3–δ mixed-conducting composite electrodes for oxygen reduction at intermediate temperatures. Int J Hydrogen Energy 2009;34:9496–504. Fu YP, Wen SB, Lu CH. Preparation and characterization of samaria‐doped ceria electrolyte materials for solid oxide fuel cells. J Am Ceram Soc 2008;291:127–31. Zhu B, Liu X, Zhu Z, Ljungberg R. Solid oxide fuel cell (SOFC) using industrial grade mixed rare-earth oxide electrolytes. Int J Hydrogen Energy 2008;33:3385–92. Gu H, Chen H, Gao L, Zheng Y, Zhu X, Guo L. Effect of Co doping on the properties of Sr0.8Ce0.2MnO3 δ cathode for intermediate-temperature solid oxide fuel cells. Int J Hydrogen Energy 2008;33:4681–8. Kordesch KV, Simader GR. Fuel cells and their applications. Weinheim: Wiley-VCH Verlag GmbH & Co. KGaa; 1996. Skinner SJ, Laguna-Bercero MA. Advanced inorganic materials for solid oxide fuel cells. In: Duncan WB, Walton R, O’Hare D, editors. Energy Materials. Weinheim: Wiley-VCH Verlag GmbH & Co. KGaA; 2011. p. 33–94. Orera A, Slater PR. New chemical systems for solid oxide fuel cells. Chem Mater 2010;22:675–90. Nguyen QM. Ceramic fuel cells. J Am Ceram Soc 1993;76:563–88. Ding D, Li X, Lai SY, Gerdes K, Liu M. Enhancing SOFC cathode performance by surface modification through infiltration. Energy Environ Sci 2014;7:552–75. Tao SW, Wu QY, Peng DK, Meng GY. Electrode materials for intermediate temperature proton-conducting fuel cells. J Appl Electrochem 2000;30:153–7. Ding C, Hashida T. High performance anode-supported solid oxide fuel cell based on thin-film electrolyte and nanostructured cathode. Energy Environ Sci 2010;3:1729–31. Chroneos A, Parfitt D, Kilner JA, Grimes RW. Anisotropic oxygen diffusion in tetragonal La2NiO4 þ d: molecular dynamics calculations. J Mater Chem 2010;20:266–70. Hauch A, Ebbesen SD, Jensen SH, Mogensen M. Highly efficient high temperature electrolysis. J Mater Chem 2008;18:2331–40. Belardi RM, Deseure J, Brant MC, Matencio T, Domingues RZ. Electrical study of cathodic activation and relaxation of La0.80Sr0.20MnO3. Ionics 2009;15:227–32. Laguna-Bercero MA. Recent advances in high temperature electrolysis using solid oxide fuel cells: a review. J Power Sources 2012;203:4–16. Jiang SP. Development of lanthanum strontium manganite perovskite cathode materials of solid oxide fuel cells: a review. J Mater Sci 2008;43:6799–833. Brant MC, Matencio T, Dessemond L, Domingues RZ. Electrical and microstructural aging of porous lanthanum strontium manganite/yttria-doped cubic zirconia electrodes. Chem Mater 2001;13:3954–61. Brant MC, Matencio T, Dessemond L, Domingues RZ. Electrical degradation of porous and dense LSM/YSZ interface. Solid State Ion 2006;177:915–21. Dutta A, Mukhopadhyay J, Basu RN. Combustion synthesis and characterization of LSCF-based materials as cathode of intermediate temperature solid oxide fuel cells. J Eur Ceram Soc 2009;29:2003–11. Marina OA, Pederson LR, Williams MC, et al. Electrode performance in reversible solid oxide fuel cells. J Elctrochem Soc 2007;154:452–9. Uhlenbruck S, Moskalewicz T, Jordan N, Penkalla HJ, Buchkremer HP. Moskalewicz, Element interdiffusion at electrolyte–cathode interfaces in ceramic high-temperature fuel cells. State Ion 2009;180:418–23. Holtappels P, Vogt U, Graule VT. Ceramic materials for advanced solid oxide fuel cells. Adv Eng Mater 2005;7:292–8.

Solid Oxides

625

[81] Bebelis S, Kotsionopoulos N, Mai A, Tietz F. Electrochemical characterization of perovskite-based SOFC cathodes. J Appl Electrochem 2007;37:15–20. [82] Adijanto L, Küngas R, Bidrawn F, Gorte RJ, Vohs JM. Stability and performance of infiltrated La0.8Sr0.2CoxFe1 xO3 electrodes with and without Sm0.2Ce0.8O1.9 interlayers. J Power Sources 2011;196:5797–802. [83] Vohs JM, Gorte RJ. High‐performance SOFC cathodes prepared by infiltration. Adv Mater 2009;21:943–56. [84] Huang Y, Anh K, Vohs JM, Gorte RJ. Characterization of Sr-doped LaCoO3-YSZ composites prepared by impregnation methods. J Electrochem Soc 2004;151:1592–7. [85] Piao J, Sun K, Zhang N, Chen X, Xu S, Zhou D. Preparation and characterization of Pr1 xSrxFeO3 cathode for IT-SOFCs. J Power Sources 2007;172:633–40. [86] Zhao H, Shen W, Zhu Z, Li X, Wang Z. Preparation and properties of BaxSr1 xCoyFe1 yO3 δ cathode material for intermediate temperature solid oxide fuel cells. J Power Sources 2008;182:503–9. [87] Richter J, Holtappels P, Graule T, Nakamura T, Gauckler LJ. Materials design for perovskite SOFC cathodes. Monatshefte Chem 2009;140:985–99. [88] Skinner SJ, Munnings CN. Electrical properties of iron-substituted La6.4Sr1.6Cu8O207δ. Mater Lett 2002;8:594–7. [89] Kharton VV, Marques FMB, Atkinson A. Transport properties of solid oxide electrolyte ceramics: a brief review. Solid State Ion 2004;174:135–49. [90] Chroneos A, Yildiz B, Tarancón A, Parfitt D, Kilner JA. Oxygen diffusion in solid oxide fuel cell cathode and electrolyte materials: mechanistic insights from atomistic simulations. Energy Environ Sci 2011;4:2774–89. [91] Badwal SPS, Ciacchi FT. Oxygen-ion conducting electrolyte materials for solid oxide fuel cells. Ionics 2000;6:1–21. [92] Rushton MJD, Chroneos A, Skinner SJ, Kilner JA, Grimes RW. Effect of strain on the oxygen diffusion in yttria and gadolinia co-doped ceria. Solid State Ion 2013;230:37–42. [93] Wachsman ED, Lee KT. Lowering the temperature of solid oxide fuel cells. Science 2011;334:935–9. [94] O’Brien JE, Stoots CM, Herring JS, Hartvigsen J. Hydrogen production performance of a 10-cell planar solid-oxide electrolysis stack. J Fuel Cell Sci Technol 2006;3:213–9. [95] Laguna-Bercero MA, Skinner SJ, Kilner JA. Performance of solid oxide electrolysis cells based on scandia stabilised zirconia. J Power Sources 2009;192:126–31. [96] Ishihara T, Jirathiwathanakul N, Zhong H. Intermediate temperature solid oxide electrolysis cell using LaGaO3 based perovskite electrolyte. Energy Environ Sci 2010;3:665–72. [97] Ishihara T, Kannou T. Intermediate temperature steam electrolysis using. LaGaO3-based electrolyte. Solid State Ion 2011;192:642–4. [98] Fergus JW. Oxide anode materials for solid oxide fuel cells. Solid State Ion 2006;177:1529–41. [99] Zhu WZ, Deevi SC. A review on the status of anode materials for solid oxide fuel cells. Mater Sci Eng 2003;362:228–39. [100] Martins RF, Brant MC, Domingues RZ, Paniago RM, Sapag K, Matencio T. Synthesis and characterization of NiO–YSZ for SOFCs. Mater Res Bull 2009;44:451–6. [101] Sun C, Stimming U. Recent anode advances in solid oxide fuel cells. J Power Sources 2007;171:247–60. [102] Lee JH, Moon H, Lee HW, Kim J, Kim JD, Yoon KH. Quantitative analysis of microstructure and its related electrical property of SOFC anode, Ni–YSZ cermet. Solid State Ion 2002;148:15–26. [103] Jiang SP, Callus PJ, Badwal SPS. Fabrication and performance of Ni/3 mol% Y2O3–ZrO2 cermet anodes for solid oxide fuel cells. Solid State Ion 2000;132:1–14. [104] Liu YL, Primdahl S, Mogensen M. Effects of impurities on microstructure in Ni/YSZ–YSZ half-cells for SOFC. Solid State Ion 2003;161:1–10. [105] Tietz F, Buchkremer HP, Stover D. Components manufacturing for solid oxide fuel cells. Solid State Ion 2002;152:373–81. [106] McIntosh S, Gorte RJ. Direct hydrocarbon solid oxide fuel cells. Chem Rev 2004;104:4845–65. [107] Kulkarni A, Ciacchi FT, Giddey S, et al. Mixed ionic electronic conducting perovskite anode for direct carbon fuel cells. J Hydrogen Energy 2012;37:19092–102. [108] Goodenough JB, Huang YH. Alternative anode materials for solid oxide fuel cells. J Power Sources 2007;173:1–10. [109] Vulliet J, Morel B, Laurencin J, et al. First results on a (La, Sr)CrO3 anode fed with methane. Electrochem Soc Proc 2003;2003:803–11. [110] Tao S, Irvine JTS. A redox-stable efficient anode for solid-oxide fuel cells. Nat Mater 2003;2:320–3. [111] Vernoux P, Guillodo M, Fouletier J, Hammou A. Alternative anode material for gradual methane reforming in solid oxide fuel cells. Solid State Ion 2000;135:425–31. [112] Huang Y, Dass RI, Xing Z, Goodenough JB. Double perovskites as anode materials for solid-oxide fuel cells. Science 2006;312:254–7. [113] Yang C, Yang Z, Jin C, Xiao G, Chen F, Han M. Sulfur‐tolerant redox‐reversible anode material for direct hydrocarbon solid oxide fuel cells. Adv Mater 2012;24:1439–43. [114] Atkinson A, Barnett S, Gorte RJ, et al. Advanced anodes for high-temperature fuel cells. Nat Mater 2004;3:17–27. [115] Li X, Zhao H, Xu N, Zhou X, Zhang C, Chen N. Electrical conduction behavior of La, Co co-doped SrTiO3 perovskite as anode material for solid oxide fuel cells. Int J Hydrogen Energy 2009;34:6407–14. [116] Badwal SPS, Deller R, Foger K, Ramprakash Y, Zhang JP. Interaction between chromia forming alloy interconnects and air electrode of solid oxide fuel cells. Solid State Ion 1997;99:297–310. [117] Fontana S, Amendola R, Chevalier S, et al. Metallic interconnects for SOFC: characterisation of corrosion resistance and conductivity evaluation at operating temperature of differently coated alloys. J Power Sources 2007;171:652–62. [118] Fergus JW. Lanthanum chromite-based materials for solid oxide fuel cell interconnects. Solid State Ion 2004;171:1–15. [119] Yoon KJ, Cramer CN, Stevenson JW, Marina A. Advanced ceramic interconnect material for solid oxide fuel cells: electrical and thermal properties of calcium- and nickel-doped yttrium chromites. J Power Sources 2010;195:7587–93. [120] Shaigan N, Qu W, Ivey DG, Chen W. A review of recent progress in coatings, surface modifications and alloy developments for solid oxide fuel cell ferritic stainless steel interconnects. J Power Sources 2010;195:1529–42. [121] Yang Z. Recent advances in metallic interconnects for solid oxide fuel cells. Int Mater Rev 2008;53:39–54. [122] Wu J, Liu X. Recent development of SOFC metallic interconnect. J Mater Sci Technol 2010;26:293–305. [123] Gaillard F, Joly J, Boreave A, Vernoux P, Deloume J. Intermittent temperature-programmed desorption study of perovskites used for catalytic purposes. Appl Surf Sci 2007;253:5876–81. [124] Zhu J, Zhang Y, Basu A, et al. LaCrO3-based coatings on ferritic stainless steel for solid oxide fuel cell interconnect applications. Surf Coat Technol 2004;65:177–8. [125] Larring Y, Norby T. Spinel and perovskite functional layers between plansee metallic interconnect (Cr–5 wt% Fe–1 wt% Y2O3) and ceramic (La0.85Sr0.15)0.91MnO3 cathode materials for solid oxide fuel cells. J Electrochem Soc 2000;147:3251–6. [126] Chen X, Hou P, Jacobson C, Visco SJ, de Jonghe LC. Protective coating on stainless steel interconnect for SOFCs: oxidation kinetics and electrical properties. Solid State Ion 2005;176:425–33. [127] Yang Z, Xia G, Simner S, Stevenson J. Thermal growth and performance of manganese cobaltite spinel protection layers on ferritic stainless steel SOFC interconnect. J Electrochem Soc 2005;152:1896–901. [128] Yang Z, Xia G, Li X, Stevenson J. (Mn,Co)3O4 spinel coatings on ferritic stainless steels for SOFC interconnect applications. Int J Hydrogen Energy 2007;32:3648–54. [129] Weil KS. The state-of-the-art in sealing technology for solid oxide fuel cells. J Mater 2006;58:37–44. [130] Rautanen M, Pulkkinen V, Tallgren J, Himanen O, Kiviaho J. Effects of the first heat up procedure on mechanical properties of solid oxide fuel cell sealing materials. J Power Sources 2015;284:511–6. [131] Stolten D. Fuel cell science and engineering: materials, processes, systems and technology. Hoboken, NJ: Wiley; 2012. [132] Singhal SC, Kendal K. High-temperature solid oxide fuel cells: fundamentals, design and applications. Philadelphia, PA: Elsevier; 2003. [133] Hoogers G. Fuel cell technology handbook. Boca Raton, FL: CRC press; 2003. [134] Campanari S. Thermodynamic model and parametric analysis of a tubular SOFC module. J Power Sources 2001;92:26–34.

626

Solid Oxides

[135] Riensche E, Stimming U, Unverzagt G. Optimization of a 200 kW SOFC cogeneration power plant: part I: variation of process parameters. J Power Sources 1998;73:251–6. [136] Zink F, Lu Y, Schaefer L. A solid oxide fuel cell system for buildings. Energy Convers Manag 2007;48:809–18. [137] Larminie J, Dicks A. Fuel cell systems explained. 2nd ed. Chichester: John Wiley & Sons; 2003. [138] Singhal SC, Kendall K. High-temperature solid oxide fuel cells: fundamentals, design and applications. Oxford; New York, NY: Elsevier Science; 2004. [139] Chan SH, Low CF, Ding OL. Energy and exergy analysis of simple solid-oxide fuel-cell power systems. J Power Sources 2002;103:188–200. [140] Fontell E, Kivisaari T, Christiansen N, Hansen JB, Palsson J. Conceptual study of a 250 kW planar SOFC system for CHP application. J Power Sources 2004;131:49–56. [141] Omosun AO, Bauen A, Brandon NP, Adjiman CS, Hart D. Modelling system efficiencies and costs of two biomass-fuelled SOFC systems. J Power Sources 2004;131:96–106. [142] Braun RJ, Klein SA, Reindl DT. Evaluation of system configurations for solid oxide fuel cell-based micro-combined heat and power generators in residential applications. J Power Sources 2006;158:1290–305. [143] Jamsak W, Assabumrungrat S, Douglas PL, et al. Thermodynamic assessment of solid oxide fuel cell system integrated with bioethanol purification unit. J Power Sources 2007;174:191–8. [144] Dokiya M. SOFC system and technology. Solid State Ion 2002;152–153:383–92. [145] Palsson J, Selimovic A, Sjunnesson L. Combined solid oxide fuel cell and gas turbine systems for efficient power and heat generation. J Power Sources 2000;86:442–8. [146] Granovskii M, Dincer I, Rosen MA. Performance comparison of two combined SOFC-gas turbine systems. J Power Sources 2007;165:307–14. [147] Costamagna P, Magistri L, Massardo AF. Design and part-load performance of a hybrid system based on a solid oxide fuel cell reactor and a micro gas turbine. J Power Sources 2001;96:352–68. [148] Kuchonthara P, Bhattacharya S, Tsutsumi A. Energy recuperation in solid oxide fuel cell (SOFC) and gas turbine (GT) combined system. J Power Sources 2003;117:7–13. [149] Möller BF, Arriagada J, Assadi M, Potts I. Optimisation of an SOFC/GT system with CO2-capture. J Power Sources 2004;131:320–6. [150] Calise F, Palombo A, Vanoli L. Design and partial load exergy analysis of hybrid SOFC-GT power plant. J Power Sources 2006;158:225–44. [151] Haseli Y, Dincer I, Naterer GF. Thermodynamic modeling of a gas turbine cycle combined with a solid oxide fuel cell. Int J Hydrog Energy 2008;33:5811–22. [152] Alcaide F, Cabot PL, Brillas E. Fuel cells for chemicals and energy cogeneration. J Power Sources 2006;153:47–60. [153] Farr RD, Vayenas CG. Ammonia high temperature solid electrolyte fuel cell. J Electrochem Soc 1980;127:1478–83. [154] Kiratzis N, Stoukides M. The synthesis of hydrogen cyanide in a solid electrolyte fuel cell. J Electrochem Soc 1987;134:1925–9. [155] McKenna EA, Othoneos A, Kiratzis N, Stoukides M. Synthesis of hydrogen cyanide in a solid-electrolyte-cell reactor. Electrochem Soc 1987;134:1925–9. [156] Yentekakis IV, Vayenas CG. Chemical cogeneration in solid electrolyte cells. J Electrochem Soc 1989;136:996–1002. [157] Michaels JN, Vayenas CG. Styrene production from ethylbenzene on platinum in a zirconia electrochemical reactor. J Electrochem Soc 1984;131:2544–50. [158] Neophytides S, Vayenas CG. Chemical cogeneration in solid electrolyte cells. J Electrochem Soc 1990;137:839–45. [159] Hugill JA, Tillemans FWA, Dijkstra JW, Spoelstra S. Feasibility study on the co-generation of ethylene and electricity through oxidative coupling of methane. Appl Therm Eng 2005;25:1259–71. [160] Zhang H, Wu J, Xu B, Hu C. Homogeneous, catalytic, oxidative coupling of methane to acetic acid in one step. Catal Lett 2006;106:161–5. [161] Periana RA, Mironov O, Taube D, Bhalla G, Jones CJ. Homogeneous, catalytic, oxidative coupling of methane to acetic acid in one step. Top Catal 2005;32:169–74. [162] Wang J, Chou L, Zhang B, et al. Comparative study on oxidation of methane to ethane and ethylene over Na2WO4-Mn/SiO2 catalysts prepared by different methods. J Mol Catal A: Chem 2006;245:272–7. [163] Amin NAS, Pheng SE. Influence of process variables and optimization of ethylene yield in oxidative coupling of methane over Li/MgO catalyst. Chem Eng J 2006;116:187–95. [164] Pujare NU, Sammells AF. Methane activation to C2 hydrocarbon species in solid oxide fuel cell. J Electrochem Soc 1988;135:2544–5. [165] Otsuka K, Suga K, Yamanaka I. Oxidative coupling of methane applying a solid oxide fuel cell system. Catal Today 1990;6(1990):587–92. [166] Jiang Y, Yentekakis IV, Vayenas CG. Methane to ethylene with 85 percent yield in a gas recycle electrocatalytic reactor-separator. Science 1994;264:1563–6. [167] Kiatkittipong W, Tagawa T, Goto S, Assabumrungrat S, Praserthdam P. Oxidative coupling of methane in LSM/YSZ/LaAlO SOFC reactor. J Chem Eng Jpn 2004;37:1461–70. [168] Kiatkittipong W, Tagawa T, Goto S, Assabumrungrat S, Praserthdam P. TPD study in LSM/YSZ/LaAlO system for the use of fuel cell type reactor. Solid State Ion 2004;166:127–36. [169] Tagawa T, Kuroyanagi K, Goto S, Assabumrungrat S, Praserthdam P. Oxygen transport in an LSM/YSZ/LaAlO solid oxide fuel cell type reactor. Chem Eng J 2003;93:3–9. [170] Kiatkittipong W, Tagawa T, Goto S, Assabumrungrat S, Praserthdam P. Oxygen transport through LSM/YSZ/LaAlO system for use of fuel cell type reactor. Chem Eng J 2005;106:35–42. [171] Kiatkittipong W, Goto S, Tagawa T, Assabumrungrat S, Praserthdam P. Simulation of oxidative coupling of methane in solid oxide fuel cell type reactor for C2 hydrocarbon and electricity co-generation. J Chem Eng Jpn 2005;38:841–8. [172] Kiatkittipong W, Tagawa T, Goto S, Assabumrungrat S, Silpasup K, Praserthdam P. Comparative study of oxidative coupling of methane modeling in various types of reactor. Chem Eng J 2005;115:63–71. [173] Ishihara T, Yamada T, Akbay T, Takita Y. Partial oxidation of methane over fuel cell type reactor for simultaneous generation of synthesis gas and electric power. In: Chemical engineering science. International symposium on multifunctional reactors; vol. 54; 1999. pp. 1535–40. [174] Pillai M, Bierschenk D, Barnett S. Electrochemical partial oxidation of methane in solid oxide fuel cells: effect of anode reforming activity. Catal Lett 2008;121:19–23. [175] Spillman RW, Spotnitz RM, Lundquist Jr. JT. Chemtech 1984;14:176–83. [176] Vayenas CG, Bebelis SI, Kyriazis CC. Solid electrolytes and catalysis. Part 1: chemical cogeneration. Chemtech 1991;21:422–8. [177] Gaskell DR. Introduction to thermodynamics of materials. 3rd ed. Oxford: Taylor & Francis; 1995. [178] Kubaschewski O, Alcock CB. Metallurgical thermochemistry. 5th ed. Oxford: Pergamon; 1979. [179] Adler SB. Chemical expansivity of electrochemical ceramics. J Am Ceram Soc 2001;84:2117–9. [180] Brawn R. Report from the solar energy laboratory. Madison, WI: University of Wisconsin. Available from: http://[email protected]/; 2002. [181] Mc Evoy AJ. Laboratoire de Photonique et des Interfaces. Ecole Polytechnique Fe anne, fuel cell technology status and prospects; 1998. [182] Data from the International Fuel Cells, a United Technology Company. Fuel cells review; 2000. [183] Global Thermoelectric Inc. Annual report; 2001. [184] Material Science and Research Inc. Fuel cell seminar, Portland, OR. October 30–November 02, 2000; 2000. [185] California Energy Commission. California renewable technology market and benefits assessment: Final report; 2001. [186] Weissbart J, Ruka R, Solid A. Electrolyte fuel cell. J Electrochem Soc 1962;109:723–6. [187] van Gerven RJF, van der Veer JHC. Abstracts of 1998 fuel cell seminar, Palm Springs, CA; 1998. p. 76. [188] Mori H, Ohmura H, Ikeda K, Tomida KA. Abstracts of seventh symposium on solid oxide fuel cells in Japan, Tokyo, Japan; 1998. p. 27. [189] Beie HJ, Drenckhahn LW, Greiner H, Rufolf B, Schichl H. In: Stimming U, Singhal SC, Tagawa H, Lehnert W, editors. Proceeding of the fifth international symposium on solid oxide fuel cells; 1997. p. 51.

Solid Oxides

627

[190] Sakaki Y, Esaki Y, Hattori M, et al. Ln1-xSrxMnO3 (Ln ¼ Pr, Nd, Sm and Gd) as the cathode material for solid oxide fuel cells. In: Stimming U, Singhal SC, Tagawa H, Lehnert W, editors. Proceeding of the fifth International symposium on solid oxide fuel cells; 1997. p. 61. [191] Badwal SPS, Foger K. SOFC development at ceramic fuel cells limited. In: Proccedings of third european solid oxide fuel cell forum, Nantes, France; 1998. p. 95. [192] Diethelm R, Schmidt M, Doggwiler B, et al. In: Proceedings of the third european solid oxide fuel cell forum Nante, France; 1998. p. 87. [193] Aizawa M, Kuroishi M, Takeuchi H, Tajiri H, Nakayama T. Abstracts of 1998 fuel cell seminar, Palm Springs, CA; 1998. p. 270. [194] Murray EP, Tsai T, Barnett SA. A direct-methane fuel cell with a ceria-based anode. Nature 1999;400:649–51. [195] Baker RT, Metcalfe IS. Evaluation of perovskite anodes for methane oxidation in solid oxide fuel cells. In: Dokiya M, Yamamoto O, Tagawa H, Singhal SC, editors, Proceedings of fouth international symposium on solid state fuel cells, Nagoya, Japan. Pennington, NJ: The Electrochemical Society, 1995; p. 781.

Further Reading Bansal NP, Kusnezoff M, Shimamura K, Wang J, Kirihara S. Advances in solid oxide fuel cells and electronic ceramics: ceramic engineering and science proceedings, Wiley; 2016. Boaro M, Salvatore AA. 2016. Advances in medium and high temperature solid oxide fuel cell technology. Cham: Springer; 2016. Boudghene Stambouli A, Traversa E. In: Proceeding of the premier seminaire international sur ´veloppement Durable Adrar. Algeria 30– l’Implication de l’Energie Solaire et Eolienne dans le De ; Octobre 31, 2001; 2001. p. 631–8. Bove R, Ubertini S. 2008. Modeling solid oxide fuel cells. Berlin: Springer; 2008. Brandon N. 2017. Solid oxide fuel cell lifetime and reliability: critical challenges in fuel cells. Cambridge, MA: Academic Press; 2017. Dickerson J, He W, Lv W. 2014. Gas transport in solid oxide fuel cells. New York, NY: Springer; 2014. Fergus J, Hui R, Li X, Wilkinson DP, Zhang J. 2008. Solid oxide fuel cells: materials properties and performance. Boca Raton, FL: CRC Press; 2008. Goodenough JB. 2009. Solid oxide fuel cell technology: principles, performance and operations. Oxford: Elsevier; 2009. Gurbinder K. 2015. Solid oxide fuel cell components: interfacial compatibility of sofc glass seals. New York, NY: Springer; 2015. Ishihara T. 2008. Perovskite oxide for solid oxide fuel cells. New York, NY: Springer; 2008. John TS, Irvine PC. 2012. Solid oxide fuels cells: facts and figures. Past Present and Future Perspectives for SOFC Technologies. London: Springer; 2012. Kendall K, Kendall M. 2015. High-temperature solid oxide fuel cells for the 21st century: fundamentals, design and applications. Oxford: Elsevier; 2015. Kusnezoff M, Bansal NP, Gyekenyesi AL, Halbig M. Advances in solid oxide fuel cells X: a collection of papers presented at the 38th international conference on advanced ceramics and composites. In: Ceramic engineering and science proceedings. Hoboken: Wiley; 2014. Minh NQ, Takahashi T. 1995. Science and technology of ceramic fuel cells. New York, NY: Elsevier; 1995. Monjur Murshed AKM, Huang B, Qi Y. 2012. Dynamic modeling and predictive control in solid oxide fuel cells: first principle and data-based approaches. Hoboken, NJ: Wiley; 2012. Ni M, Zhao TS. 2013. Solid oxide fuel cells: from materials to system modeling. London: RSC Publishing; 2013. Selimovic A, Palsson J. 2002. Networked solid oxide fuel cell stacks combined with a gas turbine cycle. J Power Sources 2002;106:76–82. Shao Z, Tade MO. 2016. Intermediate-temperature solid oxide fuel cells: materials and applications. Berlin: Springer; 2016. Singhal SC, Kendall K. 2003. High-temperature solid oxide fuel cells: fundamentals, design and applications. Oxford; New York, NY: Elsevier; 2003. Stimming U, Singhal SC, Tagawa H. In: Proceedings of the fifth international symposium on solid oxide fuel cells (SOFC-V), The Electrochemical Society; 1997. Świrski K, Leone P, Milewski J, Santarelli M. 2011. Advanced methods of solid oxide fuel cell modeling. London: Springer; 2011. Turan A, Ferrari ML, Damo UM, Sanchez D. 2016. Hybrid systems based on solid oxide fuel cells: modelling and design. Hoboken, NJ: Wiley; 2016.

Relevant Websites www.arpa-e.energy.gov Advanced Research Projects Agency. http://coloradofuelcellcenter.org/ Colorado Fuel Center. www.dur.ac.uk Durham University. http://www.fuelcelltoday.com Fuel Cell Today. www.fuelcellworkshop.com Fuel Cell Work Shop. www.fusion-conferences.com Fusion Conferences Limited. http://www.gtu.edu.tr Gebze Technical University. http://www.cism.it International Centre for Mechanical Sciences. http://www.physics.montana.edu Montana State University. http://www.ohu.edu.tr Omer Halis University. http://me.sjtu.edu.cn School of Mechanical Engineering, Shanghai Jiao Tong University. http://www.sofccenter.com/ Solid Oxide Fuel Cell Center at Universty of South Carolina. http://www.electrochem.org The Electrochemical Society. http://www.chemeng.uq.edu.au The University of Queensland.

628

Solid Oxides

www.sheffield.ac.uk The University of Sheffield. www.tugraz.at TU Graz Graz University of Technology. http://www-diva.eng.cam.ac.uk University of Cambridge. http://www.umerc.umd.edu Univeristy of Maryland Energy Research Center. http://chemeng.uwaterloo.ca University of Waterloo. www.yildiz.edu.tr Yıldız Technical University.

2.20 Batteries Hui (Claire) Xiong, Boise State University, Boise, ID, United States Eric J Dufek and Kevin L Gering, Idaho National Laboratory, Idaho Falls, ID, United States r 2018 Elsevier Inc. All rights reserved.

2.20.1 Introduction 2.20.1.1 Grand Challenges in Battery Systems 2.20.1.1.1 Transportation 2.20.1.1.2 Electrical grid 2.20.1.1.3 Portable electronics 2.20.1.2 Current Scientific and Technical Barriers 2.20.1.2.1 Materials challenges 2.20.1.2.2 Safety 2.20.1.2.3 Life 2.20.1.2.4 Performance 2.20.1.2.5 Cost 2.20.2 Background/Fundamentals 2.20.2.1 History of Lithium-Ion Batteries 2.20.2.2 Operating Principle of Lithium-Ion Batteries 2.20.2.3 Electrochemical Cells and Cell Potential 2.20.2.4 Current Flow and Kinetic Parameters 2.20.2.5 Theoretical Capacity 2.20.2.6 Efficiency 2.20.2.7 Thermal Events in Lithium-Ion Batteries 2.20.3 Systems 2.20.3.1 Cathode Materials 2.20.3.1.1 Layered Li metal oxides 2.20.3.1.2 Spinel LiMn2O4 2.20.3.1.3 Polyanionic LiFePO4 2.20.3.2 Anode Materials 2.20.3.2.1 Carbonaceous materials 2.20.3.2.2 Si and Sn 2.20.3.2.3 Ti-based oxide anode materials 2.20.3.3 Li-Ion Electrolytes and Their Impact on Performance and Degradation 2.20.3.3.1 Carbonate electrolytes 2.20.3.3.2 Phosphorus-containing electrolyte additives 2.20.3.3.3 Concentrated electrolytes 2.20.4 Analysis and Assessment 2.20.4.1 Electrochemical Analysis 2.20.4.2 Morphological and Structural Analysis 2.20.4.3 Surface Analysis 2.20.4.4 System Level Analysis 2.20.5 Illustrative Example – Regarding Low-Temperature Limitations of Lithium-Ion Batteries 2.20.5.1 Challenges for Batteries in Cold-Climate Energy Storage 2.20.5.2 Supporting Data 2.20.6 Battery Energy Storage for Grid Applications 2.20.7 Closing Remarks and Future Directions Acknowledgments References Further Reading Relevant Websites

Abbreviations BMS CCES

Battery management system Cold-climate energy storage

Comprehensive Energy Systems, Volume 2

DEC DMC EC

doi:10.1016/B978-0-12-809597-3.00245-5

630 631 631 632 632 632 632 632 633 633 633 633 633 634 634 635 636 636 636 637 637 637 640 640 641 641 641 641 645 645 647 649 650 650 650 650 651 651 651 652 656 657 657 658 662 662

Diethyl carbonate Dimethyl carbonate Ethylene carbonate

629

630

Batteries

EES EMC EV FEC HOMO LCO LFP LIB LUMO

Electrochemical energy storage Ethylmethyl carbonate Electrical vehicle Fluoroethylene carbonate Highest occupied molecular orbital LiCoO2 LiFePO4 Lithium-ion battery Lowest unoccupied molecular orbital

NCA NMC PC SEI USABC USCAR VC

Li[Ni1 y zCoyAlz]O2 Li[Ni1 y zCoyMnz]O2 Propylene carbonate Solid electrolyte interphase United States Advanced Battery Consortium United States Council for Automotive Research LLC Vinylene carbonate

Nomenclature Roman symbol

Meaning

Unit

E

V

Eeq Eg f F DG G0 DH i i0 K k0 M n O P Q R T Vch Vdis VOC

1. Potential of an electrode 2. Emf of a reaction 1. Standard potential of an electrode 2. Standard emf of a reaction Potential at equilibrium Bandgap of electrolyte F/RT The Faraday constant Change of Gibbs free energy in a chemical process Standard Gibbs free energy Change of enthalpy in a chemical process Current Exchange current Equilibrium constant Standard heterogeneous rate constant Molecular weight Stoichiometric number of electrons involved in an electrode reaction Oxidized form of the redox reaction O þ ne2R Pressure Charge passed in electrolysis Reduced form of the redox reaction O þ ne2R Absolute temperature Output voltage at charge Output voltage at discharge Open circuit voltage

K V V V

Greek symbol a Z ma mc

Transfer coefficient Overpotential Chemical potential of anode Chemical potential of cathode

None V kJ mol kJ mol

E0

2.20.1

V

eV V 1 C kJ, kJ mol kJ, kJ mol kJ, kJ mol A A None cm s 1 g mol 1 None

1 1 1

1 1

Introduction

The ever-growing demand in energy consumption fueled by economic growth and population expansion has intensified the research and development in electrochemical energy storage (EES) technologies for a sustainable future to reduce the dependence on imported fossil fuels as well as to reduce greenhouse gas emissions. With the world energy consumption projected to double (28 TW) by 2050 [1,2], it is essential to design new EES systems with consideration of material abundance, environmental benign synthesis and processes, and life-cycle analysis [3]. EES in the form of batteries can be used not only as the power source for small devices, such as portable electronics and power tools, as well as for electrical vehicles (EVs) or military applications, but also for electrical grid applications in “load leveling” of renewable energy (RE) sources, such as wind and solar

Batteries

631

power [4–8]. A battery is a device to convert chemical energy directly into electricity. Batteries are classified as primary (nonrechargeable) or secondary (rechargeable): primary batteries are designed to convert chemical energy to electrical energy only once; secondary batteries are capable of being recharged repeatedly. Since the first introduction of an electrochemical battery by Alessandro Volta in 1800, there have been significant advancements in various battery systems from primary aqueous electrolyte cells, such as the zinc–carbon cells to secondary lead-acid cells, and to rechargeable lithium (Li) and beyond Li batteries. Of the applications listed above the common denominator for the EES device is that it needs to be able to operate in a discharge mode for timescales that range from less than a second to several hours. The ability of EES and specifically batteries to meet these demands are not truly unique. On the shorter response timescale it is feasible to use other electrical storage options such as capacitors to meet demand, while on the longer scale there are other options including mechanical, thermal, and chemical energy storage options that can provide beneficial aspects to the end application. However, in terms of meeting a broad set of demands which includes volumetric energy density and utility it is difficult to replace batteries for vehicle and personal device use. On the larger scale stationary storage side, batteries serve a key role in providing stability over a key timescale and also provide beneficial storage in situations where options such as compressed air and pumped hydro are not feasible. Multiple electrochemical battery technologies exist. Some of the most prominent of the commercially available secondary batteries are lithium-ion batteries (LIBs), nickel metal hydride (NiMH), lead-acid batteries, and nickel cadmium batteries. Each of these batteries have advantages and disadvantages. As an example lead-acid batteries are generally viewed as one of the more cost effective options, yet they also have some of the lowest specific energy densities of the four types listed above. In addition to the more standard battery chemistries there has also been growth in the use of flow battery technologies. The key distinguishing feature between the conventional secondary batteries and flow batteries is that flow batteries store charge in at least one of the electrolyte solutions which can be either aqueous or organic based. For charge storage the electrolytes contain a variety of redox active components usually in the form of transition metals (TMs), such as iron, nickel, and vanadium among others. These electrolytes are flowed passed electrodes and enable the system to be either charged or discharged on demand [9]. The need for electrolyte movement and the fact that electrolytes are separated in flow batteries increases the complexity of the flow battery systems. The added cost of membranes and pumps in addition to the cost of the metals in some flow batteries are currently a component that make flow batteries more expensive than some of the other commercially available battery types [10]. Among the present practical battery technologies, rechargeable LIBs offer the highest energy density. Since the commercialization by Sony in the early 1990s [11], LIBs have transformed portable consumer electronics, such as cell phones and laptops and are considered the most promising EES technologies for the next-generation EVs and RE systems. In addition to LIB, new “beyond Li-ion” chemistries, such as Li-air, Li-sulfur, and sodium-ion batteries have been rigorously researched in the past decade, aiming at higher energy or more sustainable and lower cost alternative options for various applications, such as transportation and electrical grids. However, the increasing demands from the emerging markets have placed several challenging scientific and technical barriers [6,12–19]. Any future energy option must offer increased energy and power density, longer calendar life, and lower cost. Among all properties that must be taken into consideration in evaluation of a rechargeable battery, improvement in energy density is the primary driver for the progress of the technology [14,20]. The quest for rechargeable batteries with higher energy densities has led over the years to the move from aqueous (e.g., Pb-acid batteries) to nonaqueous electrolyte systems (such as LIBs), enabling much higher operating voltages. On the other hand, the pursuit of high energy densities has increased safety risks and concerns. Several well-publicized incidents related to LIBs have resulted in fires and explosions and have fueled public alarm, among them, fires on the Tesla Model S electric car, on the Boeing 787 Dreamliners and most recently fires or explosions by the Samsung Galaxy Note 7 [21–23].

2.20.1.1 2.20.1.1.1

Grand Challenges in Battery Systems Transportation

In March 2012, US government announced the EV Everywhere Grand Challenge – to enable the US to become the first nation in the world to produce plug-in electric vehicles (PEVs) that are as affordable for the average American family as today's gasoline-powered vehicles within the next 10 years [24]. The EV Everywhere battery goals for 2022 include: $125 kWh 1 cost, 400 Wh L 1 energy density, 250 Wh kg 1 specific energy, and 2000 W kg 1 specific power. The United States Council for Automotive Research LLC (USCAR) and United States Advanced Battery Consortium (USABC) have set PEV goals for CY2020 commercialization to be a 15 year calendar life and, depending on application between 1000 and 300,000 cycles for EVs and plug-in hybrid electric vehicles (PHEVs), respectively [25]. These goals call for disruptive innovations in R&D work of industry, national laboratories, and universities focusing on cost reduction and battery performance improvement. “The EV everywhere grand challenge: road to success” report [26] describes the current state and progress in battery technologies for PEVs: battery developers have worked closely with USABC to significantly reduce the cost by using advanced cathodes, improved processings, lower cost cell designs, and pack optimization. Battery energy density has improved from 60 Wh L 1 in 2008, to 150 Wh L 1 in 2013. Despite the continuous improvements, the challenging EV Everywhere goals still

632

Batteries

remain to be accomplished. At present, most battery developers have used conventional electrolytes and graphite anodes. The potential of more advanced LIB materials and “beyond Li-ion” chemistries to achieve the goals has been quantified by the Battery Performance and Cost model (BatPaC) developed by Argonne National Laboratory. The model suggests the combination of high energy cathode and silicon (Si) or Li anodes can meet the cost goals. However, there remain quite a few technical hurdles before affordable PEV battery systems can become a reality that includes cycle life, power, and low-temperature operation.

2.20.1.1.2

Electrical grid

Surging world energy demand, diminishing reserves of fossil fuels and growing concerns with CO2 emission from coal and fossil fuel burning have intensified the pursuit for green, high performance, cost effective, and sustainable energy storage technologies. Solar and wind power are important sources of RE. However, such power sources are variable, which makes them difficult to manage especially at high levels of penetration. Large-scale EES has received wide attention to mitigate the intermittent nature of renewable sources for electrical grid (stationary application) and adapt to the demand. Large-scale EES plays a key role in integration of renewable resources in electrical grid by “peak load-shifting” – storing energy when excess power is generated and releasing it when there is greater demand. The performance requirements for stationary energy storage and transportation are quite different. The cost goal to achieve 20% wind penetration on the grid by 2030 is set to be in the low $100 kWh 1 range, even lower than the $125 kWh 1 goal for transportation [8]. Therefore, cost reduction and scalability are more important for the requirement of EES in electrical grids. Consequently, LIBs have been the main focus in transportation due to their high energy density, while many other EES technologies are considered for stationary storage application. That said the bulk of current large-scale EES installations for grid support are LIB followed by sodium–sulfur and lead-acid chemistries [27]. As of 2013 the largest installations using EES technologies were on the order of 40 MW, however, recent installations and announcements highlight a surge in the use of LIBs for stationary storage with some recent installations having capacity in excess of 100 MWh [28,29]. Thus while the power to energy ratio of LIB for grid support and transportation differ there are many similar needs when it comes to safety and reliability of the systems to ensure safe and extended operation use. EES can help improving grid stability and reliability, for example, serving as a reserve power to avoid catastrophic events, such as the August 11, 2003 blackout in the Northeast, the Midwest, and Canada and the recent September 2011 power failure in Southern California and Arizona, which affected millions of people. Depending on stages of transmission and distribution, EES can be employed to regulate frequency, control power quality, serve as reserve power, and provide load leveling for systems ranging from kilowatts to megawatts, and gigawatts [7,8]. Therefore, it is very difficult to identify one single technology for different applications [7,8]. Performance requirements of EES for stationary use depend on the application markets, which can range from seconds to hours of response time or require a wide power/energy range.

2.20.1.1.3

Portable electronics

LIBs are currently the dominant battery technology and have been widely deployed in the portable electronics market. The volumetric energy density of prismatic LIBs for portable electronics has increased two to three times in the past 15 years [13]. However, the required power for portable electronics is predicted to increase B20% per year, while the improvement in the energy density of LIBs is expected to be B10% per year. There is a lag between the demand and improvement of LIBs. Therefore, massive research and development efforts are underway in design and developing new electrode materials with higher energy density to meet the demand.

2.20.1.2 2.20.1.2.1

Current Scientific and Technical Barriers Materials challenges

The burgeoning markets for EES have placed formidable requirements on battery materials to provide high energy/power and improved safety. Breakthroughs in materials design and engineering are urgently needed to meet the demands. Cathode and anode materials with high energy density and tap density, while maintaining a stable electrode/electrolyte interface, and high thermal stability are desired. Electrolytes with larger electrochemical potential window, high stability, and safety are preferred. All such desired physical and chemical properties of battery components require rigorous materials science and engineering research. On the other hand, the progress in materials development has not been keeping pace with the growing demands. Therefore, fundamental understanding in new electrochemical charge storage mechanisms that might enhance the energy and power is of critical importance to move the field forward.

2.20.1.2.2

Safety

The safety issue remains a major technological barrier of LIBs for their deployment in large-scale for transportation and for grid applications. Safety is the most important criterion for EV batteries [30]. Safety has become an increasing public concern as fueled by several well-publicized incidents related to LIBs, such as fires and explosions. The recent fires on Tesla Model S electric car and Boeing 787 Dreamliners highlight the critical importance of battery safety [21,22]. The safety issues are usually caused by individual component (cathode, anode, electrolyte, separator, or current collector) failure and at system level under abuse

Batteries

633

conditions (overcharge, too high discharge rate, or a short circuit). Significant improvements in areas, such as battery chemistry and thermal management (ThM) are essential to meet the stringent requirements of their applications. Efforts should be made to design electrode materials with high resistance to thermal runaway. At system level for EV and PEV applications, it is urgent to develop a comprehensive and robust battery management system (BMS).

2.20.1.2.3

Life

The USABC has aggressive life and performance goals that span to 15 years calendar life and up to 300,000 charge sustaining cycles for PHEV applications [25]. Battery life can be measured in two ways: cycle stability and overall age. For automotive applications it is expected that the battery will be able to achieve the intended number of cycles without seeing greater than 20% performance fade in either power capability or capacity. The overall age is the total number of years a battery remains operational. Batteries today do meet the EV requirements for cycling stability under test conditions [30]. However, the overall age remains a hurdle for assessment as aging accelerates at higher temperatures, and can also progress under calendar-life conditions where the battery is not undergoing active cycling. Additionally, it is challenging to estimate how quickly batteries will age across a wide range of automotive-specific conditions. Original equipment manufacturers (OEMs) are considering options, such as specifying batteries of sufficient size to account for expected degradation but at the expense of increased cost and decreased system efficiency due to extra weight. Hence, accurate estimation of battery life is critical for OEMs to save cost. At present, the lack of long-term aging data is the main issue for the lifetime study, especially for empirical model-based approaches.

2.20.1.2.4

Performance

For EVs, it is expected that consumers will be able to drive them under all weather conditions including extreme hot or cold weathers. Batteries can be optimized for either low or high temperatures. However, it is difficult to engineer them to function over a wide range of temperatures without degradation. OEMs are likely to choose between high functionality with increased system cost or performance disadvantage with higher mobility [30]. This is especially true when OEMs design the ThM systems for use in EVs. To improve functionality it is reasonable to increase the robustness of the battery TM system. However, the added cost and weight impact overall mobility and price.

2.20.1.2.5

Cost

The cost of batteries plays a critical role in determining the commercial feasibility of EVs. The Boston Consulting Group (BCG) has forecasted the battery cost in their report [30] and estimates that active materials and purchased parts will take B50% of the overall battery costs in 2020. BCG also notes that battery cost is not very sensitive to manufacturing location. Therefore, design and discovering new active materials with low cost or developing low cost manufacturing processes of existing active materials can play a critical role in tackling the cost hurdle. In this chapter, we focus on recent development in rechargeable LIBs. For further reading and information regarding other battery systems, the readers are provided with a list of monographs, reviews, and seminal articles in the Further Reading section. LIBs must overcome safety concerns particular to high energy options. Safety remains a major technological barrier for LIBs, a hurdle that researchers and product developers must address before deploying LIBs in large-scale transportation, grid, or aerospace applications [6,13,14]. LIB performance, life, and safety all strongly depend on the material properties, and significant improvements in the electrochemical and thermodynamic properties of battery materials are essential to meet the stringent requirements of their applications.

2.20.2 2.20.2.1

Background/Fundamentals History of Lithium-Ion Batteries

The concept of rechargeable Li batteries at ambient temperatures was first proposed and demonstrated by M.S. Whittingham in 1976 using a TM sulfide TiS2 as the 2.5 V cathode, a metallic Li anode, and a nonaqueous electrolyte [31,32]. During discharge, Li ions move from metallic Li anode and intercalate into empty sites of the layered TiS2, which is accompanied by a reduction of Ti4 þ to Ti3 þ , and vice versa. Regardless of its voltage, TiS2 is almost an ideal cathode: it is a semimetal therefore there is no need to add carbon additive to increase electrical conductivity; there is no phase transition in TiS2 during charge–discharge process that can alter the crystal structure to induce phase instability; it is a very fast ion conductor for fast rate; it has a soft lattice, which allows for Li ions insertion/extraction repeatedly for good reversibility; and it has good energy density of 480 Wh Kg 1. During the 1970s and 1980s, several other sulfide and chalcogenide cathodes of high capacity were investigated [33]. However, most of these cathodes exhibited a low cell voltage of o2.5 V versus a metallic Li anode. Moreover, using metallic Li as the anode could cause Li dendrites formation during charge. The Li dendrites can eventually penetrate through the polymer separator to short circuit the cell, leading to thermal runaways and explosions. Moli Energy was first to mass produce a rechargeable Li metal batteries (Li/MoS2) in 1980s but after its battery caught on fire, which led to legal action, and the company later declared bankruptcy. In 1980s, J.B. Goodenough’s group at the University of Oxford improved on Whittingham’s invention by focusing on 4 V metal oxide cathodes [34–36]. And researchers at SONY developed a new safe anode to replace metallic Li anode. These inventions led to the

634

Batteries

Discharging

e–

e–

AI

Cu

Anode

Cathode

Fig. 1 Schematic of the operating principle of rechargeable Li-ion batteries (LIB).

commercialization of the first Li-ion battery technology by SONY in 1991 in Japan using LiCoO2 (LCO) as the cathode and graphite as the anode.

2.20.2.2

Operating Principle of Lithium-Ion Batteries

A LIB consists of two electrodes, i.e., anode and cathode, immersed in an ion-conducting and electrically insulating electrolyte and separated by a polymer membrane (Fig. 1). Rechargeable LIBs operate in a “rocking-chair” fashion, i.e., Li ions shuffle between cathode and anode during charge/discharge cycles through a nonaqueous electrolyte. During discharge, Li ions travel from the anode within the nonaqueous electrolyte to the cathode, while electrons move externally from the anode to the cathode to meet electroneutrality while generating electricity, and vice versa. What LIBs share in common with conventional batteries is that the redox reactions at the electrode/electrolyte interfaces are accompanied by the diffusion of ions within the electrolyte. However, the differences are notable as well. In conventional batteries, the redox reactions generally are not accompanied by solid-state ion diffusion within the electrode, while in LIBs the heterogeneous redox reactions are always accompanied by solid-state ion diffusion. The most common type of electrochemical reaction of Li ions storage in rechargeable LIBs is intercalation where Li ions insert/extract into/from vacant space within a host material during the charge/discharge process (Fig. 1).

2.20.2.3

Electrochemical Cells and Cell Potential

A battery is an electrochemical cell, which is a collection of interfaces between conducting phases. The performance of a battery system is affected by the processes and factors that affect the transport of charge across the interfaces, for example, between an electronic conductor (an electrode) and ionic conductor (an electrolyte) [37]. The cell potential of an electrochemical system is a collection of electrical potential difference across all the interfaces in an electrochemical cell, regardless of current passage through the cell. This cell potential is a measure of the energy available to drive charge externally between the electrodes [37]. In an electrochemical cell, the redox reaction, i.e., oxidation and reduction, takes places spatially differently at two electrodes (oxidation at anode and reduction at cathode). In contrast, a chemical redox reaction occurs simultaneously at the same location and is homogenous. Therefore, an electrochemical reaction can be considered as the sum of two “half-reactions” occurring in “halfcells.” In generalized terms, the reaction at one electrode (reduction in forward direction) can be represented by aA þ ne ⇌ cC

ð1Þ

where a moles of A accept n electrons to form c moles of C. At the other electrode, the reaction (oxidation in forward direction) can be represented by bB ⇌ dD þ ne

ð2Þ

the overall redox reaction in the electrochemical cell is given by addition of the two half-cell reactions aA þ bB ⇌ cC þ dD

ð3Þ

Batteries

635

the cell potential or the electromotive force (emf) can be related to a thermodynamic quantity (the Gibbs free energy) using the expression DG ¼

ð4Þ

nFE

where E (V) is the maximum potential between two electrodes, also known as equilibrium potential, and F is the Faraday’s constant (1F¼ 96,485.3 C mol 1). When all substances are at unit activity, DG0 ¼

nFE0

ð5Þ

where E0 is the standard cell potential and ∆G0 is the standard free energy. Based on thermodynamics, the cell potential can be related to concentration through the Nernst equation, RT acC adD ln nF aaA abB

E ¼ E0

ð6Þ

where ai ¼activity of relevant species, R ¼ gas constant (8.314 J K 1 mol 1), and T is the absolute temperature in K. Other thermodynamic quantities can also be derived from electrochemical measurements based on the relationship between the cell potential and free energy (Eq. (4)). For example,
 ∂E ð7Þ DS ¼ nF ∂T P and 
  ∂E E ð8Þ DH ¼ nF T ∂T P The equilibrium constant of the reaction is given by RT lnK ¼ DG0 ¼

nFE0

ð9Þ

Relationship between cell potential and chemical potential is given by E ¼ ðma

mc Þ=F

ð10Þ

where ma and mc are chemical potentials of the anode and cathode, respectively. These relationships can also be used to predict electrochemical properties from thermodynamic data. Direct measurement of absolute electrode potential is considered practically impossible [37]. Experimentally, it is useful to establish a scale by referring electrode potentials and half-cell emfs to a standard reference electrode chosen as “zero” for the reference potential. By convention, the standard potential of the H2 (a ¼ 1)/H þ (a ¼ 1) reaction is taken as zero and all standard potentials are referred to this potential. The values of electrode standard potentials can be found in reference books such CRC handbook of chemistry and physics.

2.20.2.4

Current Flow and Kinetic Parameters

When a battery is under operation, current flows and charge transport must occur at a corresponding rate. Current is a measure of electrode reaction rate. Several processes can affect the current, such as charge transfer reactions at the electrode/electrolyte interface, mass transfer, surface reactions, and chemical reactions preceding, or following the charge transfer reaction. The overall rate of the electrochemical process is determined by the slowest step in the whole sequence. The thermodynamic treatment of the electrochemical process in a battery presented in Section 2.20.2.3 describes the equilibrium conditions of a system. However, most battery systems operate at off-equilibrium conditions, such as polarization when current flows. Polarization associated with an electrode process are termed as “overpotential.” The overpotential Z is defined as Z¼E

ð11Þ

Eeq

where Eeq is the equilibrium potential (zero current) of an electrode and E is potential when a certain current flows. During charge and discharge, polarizations (e.g., activation, concentration, and IR drops) reduce the output voltage Vdis from the open circuit voltage Voc and increase the voltage Vch required to reverse the chemical reaction on charge [24]: Vdis ¼ Voc

Z

ð12Þ

Vch ¼ Voc þ Z

ð13Þ

Depending on the energy barrier for polarization, i.e., the overpotential, the reaction at the electrode can be quasi-reversible or totally irreversible. The Butler–Volmer equation can be used to describe the current–overpotential characteristics for a quasireversible electrode reaction, h i i ¼ i0 e af Z eð1 aÞf Z ð14Þ

ð1 aÞ Ca (A¼ electrode area, k0 ¼standard rate constant, C ; C ¼ bulk concentrations, where i0 is the exchange current i0 ¼ nFAk0 CO R O R a¼ transfer coefficient), f¼F/(RT). Rearrange this equation as suggested by Allen and Hickling [38] log

i 1

ef Z

¼ logi0

af Z 2:3

ð15Þ

636

Batteries

af so that a plot of log 1 ief Z versus Z yields an intercept of logi0 and a slope of 2:3 . Kinetic parameters such as k0 and a can thus be obtained. When the kinetics of an electrode reaction is very slow such as a very high overpotential is needed to be applied and reaction is irreversible (unidirectional), the Tafel equation can be used to describe the electrode process. For example, at large negative overpotentials, e afnce(1 a)fn and Eq. (14) becomes

i ¼ i0 e

af Z

ð16Þ

or Z¼

RT lni0 aF

RT lni aF

ð17Þ

which follow a relation of the Tafel form Z ¼ a þ blogi

ð18Þ

to obtain the empirical Tafel constants a¼

2.20.2.5

2:3RT logi0 aF

2:3RT aF

b¼

ð19Þ

Theoretical Capacity

The theoretical capacity describes the charge stored in a battery cell. The value is expressed in ampere-hour (Ah) or milli-amperehour (mAh). It is typical to represent the capacity of electrode materials in term of specific capacity (i.e., gravimetrical capacity, mAh g 1) or volumetric capacity (mAh L 1). The theoretical capacity of an electrode can be determined from the electrochemical reaction that the electrode is involved and the mass of the active materials through the Faraday’s law: Theoretical capacity ¼

zF M

ð20Þ

where z is the number of electron transferred in the redox reaction involving 1 mol of the active material, F is the Faraday’s constant, and M is the molecular weight of the active material. Theoretically 1-gram-equivalent weight of material will deliver 96,487 C or 26.8 Ah. Similarly, the theoretical capacity can be calculated for the cell, rather than for just one electrode. In this calculation, the coulombs of charge in the two electrodes are made equal and normalized with the masses of both electrodes based on stoichiometry.

2.20.2.6

Efficiency

The energy efficiency of a cell at a fixed current I is given by [24]: Z Qdis V ðqÞdq  100 Z0 Qch V ðqÞdq

ð21Þ

0

Q¼

Z

t

Idt ¼ 0

Z

Q

ð22Þ

dq 0

1

1

where Q is the total charge per unit weight (Ah kg ) or per unit volume (Ah L ) and Q is a function of I as the charge transfer across electrode/electrolyte interface become diffusion-limited at high currents. On charge–discharge cycling, electrode volume changes, chemical reactions between electrode and electrolyte, and/or electrode decomposition can lead to irreversible capacity loss. During the initial charging of a cell, electrode–electrolyte chemical and electrochemical reactions result in the irreversible formation of a passivating solid electrolyte interphase (SEI). The Coulombic efficiency of a single cycle is given by Qch  100 Qdis

2.20.2.7

ð23Þ

Thermal Events in Lithium-Ion Batteries

The risk of thermal events have resulted in heightened safety worries about LIBs. While recent incidents involving batteries used for many applications have focused attention on the high energy density LIBs, there remains a significant quantity of well-manufactured batteries and battery systems, which remain in operation without incident. The total failure occurrence for LIBs is low given the quantity of batteries in use in laptops, cell phones, various forms of battery electric vehicles (BEVs) and in other small electronics and power tools. There are two primary means by which LIBs can proceed to catastrophic failure. In the first, a hard short can develop between the positive and negative electrodes of the cell. This failure mechanism is uncommon and faulty cells are typically identified prior to introduction into the consumer space. The second failure mechanism is related to overheating of the battery, which can also

Batteries

637

result in a catastrophic thermal event. In these instances the heat generation can either be associated with the cell itself or with an external source, such as failed electronics near the battery or in poor system ThM. Regardless of the mechanism of initial heat generation, there is a defined and known path which ensues as the LIB is overheated. Initially heat leads to instabilities in the SEI, which provides stability to the negative electrode of LIBs [39]. As the SEI becomes less stable lithiated and non-lithiated surfaces which are highly reactive become exposed. The reactive surfaces then are free to participate in electrode degradation which increases heat generation. Upon sufficient heat generation most positive electrodes in LIBs release oxygen which is available to react with both the electrolyte and the negative electrode leading to a cascading thermal event [39–41]. Due to worries about the safety of LIBs there have been considerable efforts to address the thermal response of LIBs in aggressive environments. First nonflammable electrolytes and electrolytes which inhibit the early stages of failure have received significant interest. Second the development of different electrode materials, electrode architectures and in cells and pack designs, which are better at heat rejection has occurred in recent years.

2.20.3

Systems

As stated in Section 2.20.2.3 the open circuit voltage (VOC) of a cell is the difference in chemical potential between the anode and cathode as shown in Eq. (10). This working voltage is also limited by the electrochemical potential window of the electrolyte, as illustrated in Fig. 2, which is determined by the energy gap (Eg) between its highest occupied molecular orbital (HOMO) and lowest unoccupied molecular orbital (LUMO) [24]. The selection rule for suitable anode and cathode is such that the ma of the anode lies below the edge of LUMO of the electrolyte and the mc of the cathode sits above the edge of HOMO, otherwise, the electrolyte will be reduced on anode or oxidized on cathode unless an appropriate SEI is formed to prevent such reactions. In addition to the electrochemical window of the electrolyte, the cathode anion-p bands also determine the achievable working voltage. ma cannot be lowered below the top of the anion-p bands, which may have an energy above the electrolyte HOMO, as it may results in an introduction of holes or removal of electrons in anion-p bands so that the anions may undergo oxidation.

2.20.3.1

Cathode Materials

The current state-of-the-art cathode materials are mainly in three types: layered Li metal oxides, spinel-like compounds such as LiMn2O4, and polyanionic compounds such as olivine LiFePO4 (LFP) (Fig. 3). There are quite a few comprehensive in-depth reviews [42–48] on cathode materials, which are suggested for further readings as a full review is outside of the scope of this chapter.

2.20.3.1.1

Layered Li metal oxides

Layered Li metal oxides adopt the a-NaFeO2 structure and are built-up of ordered stacking of MO2 layers with edge-sharing MO6 octahedra with different orientations along the c-axis direction. According to the classification proposed by Delmas et al. [49], the polymorphs of layered oxides can be categorized in terms of (1) a capital letter indicating the surrounding of the interlayer alkali species (O for octahedral, T for tetragonal, and P for prismatic) and (2) a value that equals to the number of MO2 layers needed for periodicity (Fig. 4). For example, using the classification LCO is an O3 type (space group R-3m), where Li ions sit in octahedral sites and 3CoO3 layers are required for periodicity. LCO has long been the benchmark cathode material in commercial LIB, which has a high operating voltage (B4 V), ease of synthesis, and good cycle life. Due to its low molecular weight, LCO has a high theoretical capacity of 274 mAh g 1. However, due to chemical instability and structural instability caused by large electrostatic repulsion between the TM polyhedra, only B50% (140 mAh g 1) of its theoretical capacity can be utilized in practical LIB as the structure tends to collapse at deep charge (x40.5 in

Energy

anode

LUMO

Eg

Voc

HOMO

Anode (–)

Electrolyte

cathode Cathode (+)

Fig. 2 Schematic illustrations of the energy levels in a Li-ion battery. HOMO, highest occupied molecular orbital; LUMO, lowest unoccupied molecular orbital.

638

Batteries

5.0

LiNi0.33Mn0.33Co0.33O2 – Std NCM (Layered)

LiNi0.5Mn1.5O4 (Spinel)

LiCoO2 4.5

Voltage (V)

4.0

3.5

Li2MnO3- LiNi0.33Mn0.33Co0.33O2

Li rich NCM (Layered)

LiFePO4 (Olivines)

3.0

2.5

2.0 0

50

100

150

200

250

Discharge capacity (mAhg–1) Fig. 3 Li-ion battery cathodes: important formulae, structures and voltage profiles during discharge. The potentials are versus Li reference electrodes [48]. NCM, LiNi1/3Co1/3Mn1/3. Reprinted with permission. Copyright (2014), Elsevier.

Li1 xCoO2) [34]. In order to suppress chemical instability or reactivity with the electrolyte, surface coating of nanostructured inert oxides, such as Al2O3, TiO2, SiO2, ZnO, ZrO2, MgO, AlPO4 has been found to enhance the stability and increase reversible capacity of LCO from B140 mAh g 1 to B200 mAh g 1 [50–53]. The long-term stability of such nano-oxide-coated LCO as well as the mechanism of such coating are yet to be evaluated. Although, LCO is widely used for LIB in portable electronics, such as cell phones, due to the high cost and toxicity of Co, other layered oxide materials have been researched in search for higher capacity, lower cost, and better environmental benignity alternatives. LiNiO2 is isostructural with LCO and has been considered as an alternative to LCO due to its low cost and less toxicity [54]. LiNiO2 has slightly lower operating voltage (B3.8 V) than that of LCO, but a higher capacity (B200 mAh g 1). However, due to the difficulty to stabilize Ni3 þ at high synthesis temperatures it is always accompanied by simultaneous reduction of Ni3 þ to Ni2 þ , which leads to the formation of non-stoichiometric Li1 zNi1 þ zO2 with some Ni2 þ [55]. The excess Ni2 þ in non-stoichiometric Li1 zNi1 þ zO2 compounds tend to migrate to the Li þ layers due to smaller difference in charge and size (Table 1), which cause disordering of cations and consequently poor electrochemical performance [44,55,56]. In addition, differential scanning calorimetry (DSC) measurements have indicated that LiNiO2 has a lower decomposition temperature in presence of electrolyte and a greater energy release compared to LCO [44,57,58], which makes this material less competitive to LCO due to safety reasons. Layered LiMnO2 is an attractive material because manganese is less expensive and more environmental friendly compared to cobalt and nickel. Therefore, LiMnO2 has been under extensive studies. However, due to the low size difference between Mn3 þ in high spin configuration (0.645 Å ) and Li (0.76 Å ) (Table 1), LiMnO2 synthesized at high temperatures adopts an orthorhombic structure instead of stabilizing the layered O3-type structure, which shows poor electrochemical activity [59]. Through ion exchange of stable layered a-NaMnO2 with Li þ , metastable layered monoclinic LiMnO2 has been made [60–62]. However, during charge Li1 xMnO2 transforms into the more stable spinel LiMn2O4, which exhibit poor cycling performance [56]. This transformation is facilitated by the fact that both O3-LiMnO2 and spinel LiMn2O4 adopt the same oxygen framework, therefore, less energy is required for the diffusion of cations [44]. In addition, when such materials are in contact with electrolyte, the disproportionation redox process 2Mn3 þ -2Mn2 þ þ 2Mn4 þ is favored, this in consequence generates mobile Mn2 þ that can diffuse from Mn to Li layer [63].

Batteries

639

c a

b

P3 type

O3 type

Octahedral site

Prismatic site

O2 type

P3 type

Fig. 4 Illustration of the polymorphs of layered oxides according to Delmas’s classification.

Table 1

Comparisons of ionic radii of Li-ion and transition metal (TM) ions

Ion species

Li þ

Co3 þ

Ni2 þ

Ni3 þ

Mn3 þ (high spin)

Mn3 þ (low spin)

Mn4 þ

Fe3 þ (high spin)

Fe4 þ

Size (Å )

0.76

0.545

0.69

0.56

0.645

0.58

0.53

0.645

0.585

Li[Ni1 y zCoyAlz]O2 (NCA) partial substitution(s) at the metal site of layered LiMO2 have been intensively explored to achieve higher energy density, higher safety, cycling stability, and lower cost. To suppress spontaneous reduction of Ni3 þ to Ni2 þ and the cation disorder, partial substitution of Co3 þ for Ni3 þ has been investigated. The stability of Co3 þ at high temperatures alleviates the problem of Ni2 þ formation and the radius of Co3 þ (0.545 Å ) is smaller but close enough to that of Ni3 þ (0.56 Å ) to favor the formation of solid solution Li(Ni, Co)O2 [44,64]. Partial doping of non-electrochemically active elements, such as aluminum appears to be the most promising to enhance the electrode thermal stability [44,65]. Al3 þ ions have been shown to act as structural pillars due to stronger bonding between Al-O and Ni-O [66]. In order to stabilize the layered structure as well as to enhance thermal stability, the approach of dual substitution of Co and Al have been pursued, which led to R&D of a group of NCA compounds. Among the family of NCA compounds, the Li[Ni0.8Co0.15Al0.05]O2 [44,67] has shown decent capacity (200 mAh g 1), while maintaining good safety. Li[Ni1 y zCoyMnz]O2 (NMC) compounds in the NMC family have attracted intensive attention due to its promise in high capacity, high thermal stability, and better cycling performance [44,56]. In NMC, Co is partially substituted by Ni and Mn to increase capacity and improve safety. In these oxides, Ni, Co, and Mn exist as Ni2 þ , Co3 þ , Mn4 þ , respectively. Among all the compositions studied, Li[Ni1/3Co1/3Mn1/3]O2 shows the most promising features for commercial values: good capacity (150–200 mAh g 1), good rate capabilities, and enhanced safety [44]. The effects of Ni, Co, and Mn in NMC materials are summarized in Fig. 5 [44] and Table 2. It is noted that Ni governs capacity, Co cycling stability, and Mn thermal stability [68]. Therefore, it is desirable to have Ni-rich NMC for high capacity cathode. However, the penalty associated with the high surface reactivity of Ni-rich NMC, which leads to poor cycling performance [69]. In order to address such issue, several strategies have been explored: (1) surface coating [70]; (2) core/shell structure though a core made of a poorly thermally stable but high capacity Ni-rich NMC phase and a shell made of a highly thermally stable but low capacity Mn-rich NMC phase [71]; and (3) concentration gradient where the Ni concentration decreases linearly, whereas the Mn concentration increases linearly from the center to the outer layer of each particle to address the issue of progressive delamination of core–shell interface upon long cycling [72].

Batteries

320 Co content

95 (3 3 3) (5 3 2)

280 260

Mn content

Thermal stability (°C)

300

100

90 85 80

(6 2 2)

75

(7 1.5 1.5)

240

(8 1 1)

220

NI content

160

170

180

70 65

(8.5 0.75 0.75)

190

200

Capacity retention (%)

640

210

Discharge capacity (mAh g–1) Fig. 5 A map of relationship between discharge capacity (black), thermal stability (blue), and capacity retention (red) of Li/Li[NixCoyMnz]O2 with number in brackets corresponding to the composition (Ni, Mn, Co). Adapted from Noh HJ, Youn S, Yoon CS, Sun YK. Comparison of the structural and electrochemical properties of layered Li[NixCoyMnz]O2 (x¼1/3, 0.5, 0.6, 0.7, 0.8 and 0.85) cathode material for lithium-ion batteries. J Power Sources 2013;33:121–30; Rozier P, Tarascon JM. Review-Li-rich layered oxide cathodes for next-generation Li-ion batteries: chances and challenges. J Electrochem Soc 2015;162:A2490–9.

Table 2

Comparison of effects of Ni, Co, and Mn in Li[Ni1

Advantages Disadvantages

y zCoyMnz]O2

(NMC) materials

Ni

Co

Mn

High capacity High surface reactivity, safety

High voltage, stability Cost, toxic

High safety, cheaper, less toxic Stability

Li-rich layered oxides xLi2MnO3
(1 x)LiMO2 (M ¼ Ni, Co, Mn), Li substitution in the layered oxide systems can be traced back to the early works by the Thackeray’s [73] and Dahn’s group [74]. Their pioneer works have led to materials termed as “Li-rich NMC” which combines the beneficial effects of Ni, Co, Mn with some excess Li in the TM layers, exhibiting capacities exceeding 280 mAh g 1 [44,75]. These materials have been considered as the next-generation cathode materials for high energy density LIB. The material design strategy is to embed a layered Li2MnO3 in a structural compatible layered LiMO2 to stabilize the structure and suppress oxygen evolution at the surface of charged electrode particles. Compared to their conventional layered counterparts, Lirich layered compounds have significantly higher capacity (e.g., almost double that of LCO). The origins of the ultra-high capacity of Li-rich NMC are still under debate. It has been shown that Li2MnO3, which was originally considered to be inactive, is activated in the first charge 44.7 V during Li extraction and O2 evolution [76]. Despite their high capacity, these materials often encounter capacity fading and voltage fading, arising from detrimental structural and surface changes. Lately, the contribution of bulk and surface oxygen on high capacity has been demonstrated by Meng’s group [77] through gas–solid interfacial modification of oxygen activity in Li-rich NMC to achieve a reversible capacity of 300 mAh g 1 after 100 cycles without voltage fading.

2.20.3.1.2

Spinel LiMn2O4

Spinel LiMn2O4 has become popular as Mn is less expensive and more environmental benign compared to Co and Ni. The insertion/extraction of two Li þ ions into/from LiMn2O4 occur at two voltage region [78]. At around 4 V Li insertion/extraction take place at the tetrahedral sites, while maintaining its initial cubic symmetry. At around 3 V Li insertion/extraction occur at the octahedral sites by a two-phase mechanism involving the cubic spinel LiMn2O4 and the tetragonal lithiated spinel Li2Mn2O4 [78]. The structure of spinel LiMn2O4 is more robust during delithiation compared to LCO [43]. However, when it is in contact with common Li nonaqueous electrolytes Mn dissolution has been observed to lead to capacity fading [79]. Surface modification has shown to suppression Mn dissolution and improve capacity retention [80].

2.20.3.1.3

Polyanionic LiFePO4

In polyanionic compounds, where O2 is replaced by PO43 , SO42 , SiO42 , or BO33 anions, offer promise for sustainable cathode materials as they readily form compounds using abundant elements like Fe, as exemplified by LFP [3]. Since the discovery by Goodenough’s group in 1997 [81–83], LFP has been studied extensively and been commercialized. The Li extraction/insertion occurs through a two-phase mechanism with LFP and FePO4 as the two phases without much solid solubility. LFP is an attractive

Batteries

641

cathode material because Fe is abundant, inexpensive and environmental benign as well as LFP has good thermal stability [84]. However, the disadvantages of LFP include its relatively low average potential (B3.4 V) and low electrical and ionic conductivity. Olivine structure LFP is a one-dimensional Li þ conductor with Li þ along edge-shared LiO6 chains [56]. Through reduction in particle size along with surface coating of conductive layers such as carbon, the rate performance of LFP has been significantly improved [56]. Current state-of-the-art LFP exhibits near theoretical reversible capacity (B160 mAh g 1) with good rate capability.

2.20.3.2

Anode Materials

There are three major types of anode materials for LIBs. 1. Intercalation type: similar to intercalation cathode materials, intercalation anode materials such as graphite and Li4Ti5O12 (LTO) act as host materials where Li þ ions are inserted/extracted at empty sites of the host during charge–discharge cycles without pronounced changes in the crystal structure of the hosts. 2. Conversion type: in conversion, anode materials (mostly binary oxides such as Fe2O3, CoO, SnO2) undergo electrochemical reduction according to the following reaction MxOy þ 2yLi þ þ 2ye -xM0 þ yLi2O. the final product of a conversion reaction is a homogeneous nanocomposite made of metal nanoparticles embedded in Li2O matrix. This class of anode materials is the least important as they exhibit huge hysteresis in their charge/discharge voltage profiles and have not been made practically successful in LIB systems. 3. Alloy type: elements such as Si and Tin (Sn) alloy with Li, such as M þ xLi þ þ xe -LixM where x may exceed four. Therefore, this type of anode materials can have very high theoretical capacity as it can react with multiple Li per metal atom. On the other hand, this type of reaction is usually accompanied by huge volume expansion/contraction during alloying/de-alloying, which may induce mechanically instability of the materials.

2.20.3.2.1

Carbonaceous materials

The most common carbon anode is graphite. Together with LCO, graphite enabled the commercialization of LIBs 25 years ago and is still widely used nowadays. In graphite, Li ions intercalate into vacant space within the interlayers of graphene layers stacked in ABABAB (2H graphite) or ABCABC orders (3R graphite), according to the reaction 6C þ Li þ þ e -LiC6. Li þ forms a stage-I (one Li layer and one graphene layer alternate with each other in the stacking) Li-rich phase with graphite with a theoretical capacity of 372 mAh g 1. Due to its low cost, abundance, low operating voltage (B0.1 V vs. Li/Li þ ), graphite is very popular anode material of choice for portable electronics. However, both its gravimetric and volumetric capacity is not high enough to meet the high energy requirement for EV batteries. Carbon materials that can be graphitized by high temperature treatment (2000–30001C) are termed at soft carbon. Upon heat treatment, the turbostratic disorder is removed progressively. Those that cannot be graphitized even at temperature as high as 30001C are terms as hard carbon and contains defects and nanovoids. Nanovoids and defects could provide additional capacity beyond the theoretical capacity of 372 mAh g 1 [85]. Nevertheless, the Coulombic efficiency is worse in hard carbon materials due to a thick SEI layer. In addition, the voids in such material significantly decrease its volumetric capacity.

2.20.3.2.2

Si and Sn

Metal anodes for LIBs, such as Si [86] and Sn [87] have received intensive attention in recent years given their much higher theoretical capacity (4c. 1000 mAh g 1) over commercial graphite anodes. However, they suffer from fast capacity fading due to huge volume change (4300%) at charge/discharge cycles, unstable SEI, and pulverization. Nanostructures can efficiently improve cycling stability by accommodating stress induced from large volume change [47,86,88–96]. Si is one of the most promising anode materials due to its high theoretical capacity (4400 mAh g 1 ) and low operating voltage (B0.3 V vs. Li/Li þ ) [96,97]. However, due to the high volume change (4300%) during charge/discharge process which lead to particle fracturing and loss of electronic contacts, and capacity fading is commonly seen in Si anode. New designs of Si electrode such as using nanostructured Si [96] to improve mechanical stability and ion diffusion as well as new binder designs [98] that embrace the huge volume change have greatly improved the performance of Si anode. Novel electrode architectures such as the yolk–shell design [99] and the pomegranate design [100] have shown to improve cycling stability. Nevertheless, challenges associated with Si anode still include structural degradation and instability of SEI, side reactions with electrolytes, and low volumetric capacity when employing nanomaterials. Sn anode is of significant interest and has been investigated extensively due to its high capacity (993 mAh g 1 or 7262 Ah L 1 for Li4.4Sn), abundance, improved safety, and environmentally benign qualities. However, most Sn-based anodes suffer from rapid capacity decay and poor cyclibility. To overcome these issues, substantial efforts have been made to improve the structural stability and integrity of Sn anodes by synthesizing Sn nanostructures [101,102] and constructing Sn/C composites [103]. A variety of Sn/C composites such as Sn nanoparticles embedded in porous carbon [104–109], and core–shell Sn@C nanoparticles encapsulated in carbon nanotubes [110–112] or Sn@graphene [113–117] anchored on carbon nanotubes [118] have been designed to improve the stability of Sn anode.

2.20.3.2.3

Ti-based oxide anode materials

Ti-based oxide materials have attracted intensive attention as promising anode materials for LIBs due to their excellent cycling stability, low cost, abundance, and environmentally benign nature [90,119,120]. Ti-based oxide is one of the few metal oxide

642

Batteries

b

c (A)

(B)

b

b

a (C)

a

c

(D)

a

Fig. 6 Illustration of crystal structures of (A) anatase, (B) rutile, (C) TiO2-B, and (D) brookite.

materials that intercalates Li ions at relatively low voltages as anodes (B1.5–1.8 V vs. Li/Li þ ) for a reasonable output voltage between cathodes and has been found as a safe alternative to the graphite anode. Researchers have also widely investigated Li titanate spinel LTO as its sufficiently higher working potential versus Li/Li þ (1.55 V) makes it inherently safer than graphite [120]. In addition, LTO exhibits excellent cycling stability arising from the “zero strain” attribute as its crystal structure undergoes negligible volume change upon Li-ion insertion and/or extraction [120]. While LTO is an excellent host for reversible Li insertion and/or extraction, its theoretical specific capacity is limited to 175 mAh g 1 as compared to graphite (theoretical capacity: 372 mAh g 1). In contrast, TiO2 has a relatively larger theoretical specific capacity at 335 mAh g 1 or 1.0 Li per TiO2, which makes it an attractive alternative to LTO. Among various TiO2 polymorphs investigated for their electrochemical properties, researchers have found that rutile (space group P42/mnm), anatase (I41/amd), brookite (Pbca), and TiO2-B (C2/m) shows Li electrochemical reactivity (Fig. 6). Li-reaction with the TiO2 polymorphs is expressed as: TiO2 þ xLiþ þ xe 2 Lix TiO2

ð24Þ

and x varies with different TiO2 polymorphs, morphology, and crystal orientation. The first attempts at using TiO2 as a durable and safe electrode material focused on microcrystalline TiO2 materials such as rutile, anatase and TiO2-B [121]. These electrodes showed moderate specific capacities (the maximum Li uptake of 0.5 Li/Ti for anatase and TiO2-B, and no activity for rutile) [121] due to the limited room temperature reactivity and conductivity at microscale. Recently, rapid development in nanostructuring electrode materials has led to substantial improvement [47,88–93]. There are several benefits to decreasing electroactive material size to the nanometer region: (1) enhanced power and energy densities due to shortened transport lengths for both Li þ and e diffusion, allowing adoption of materials with low electronic and/or ionic conductivity; (2) large electrode/electrolyte contact area, which decreases the current density per unit area and increases accessible sites for Li þ , leading to fast kinetics and higher rate capability; (3) better accommodation of the strain of Li þ insertion/extraction, improving cycle life. Researchers have reported that nanosized TiO2 polymorphs exhibit significantly superior electrochemical properties in LIB as compared to their micro-sized counterparts at room temperature [89,90]. As the thermodynamically most stable form of TiO2, rutile TiO2 in its bulk crystalline form has been known to store negligible Li at room temperature (o0.1 Li per TiO2) [122,123]. In contrast, at 1201C, researchers have reported increased Li-reactivity with bulk rutile TiO2 in Li polymer electrolyte cells [124]. Rutile TiO2 has a tetragonal P42/mnm structure (a¼ b ¼ 4.625 Å , c ¼ 2.959 Å ) comprised of TiO6 octahedra that share edges along the c direction and corners in the ab planes. There are two potential Li intercalation sites between the octahedra with either tetrahedral or octahedral oxygen coordination. Theoretical calculations by Koudrichova et al. showed preferential Li þ insertion into the octahedral sites by B0.7 eV/Li [125]. Despite that, Li intercalation is thermodynamically favorable up to high Li concentrations; intercalation is limited by diffusion of Li þ at room temperature. It is generally accepted based on both theoretical calculations and experimental studies that the diffusion in rutile TiO2 is highly anisotropic. The diffusion coefficients for Li þ diffusion along c axis and in the ab plane are approximately 10 6 and 10–15 cm2 s 1, respectively [126,127]. This means that Li intercalation proceeds through rapid diffusion along the c axis. At low temperature, the

Batteries

643

15 2 theta

10 12 14 16 18 20 2 theta

20

Charge Li + extraction

2.0

Amorphous

1.5

10

15 2 theta

20

Cubic

Intensity (a.u.)

10

2.5 Voltage (V vs. Li/Li+)

Amorphous

Intensity (a.u.)

3.0

Cubic

Intensity (a.u.)

Intensity (a.u.)

very sluggish diffusion in ab plane prevents Li þ from reaching the thermodynamically favorable octahedral sites. Furthermore, screening of the Li–Li interaction along the c axis is not as effective as in the ab plane, which provides a trapping mechanism for Liion pairs [125,126]. The Li-ion pairs formed in the ab plane can effectively block c channels for further intercalation as diffusion in the ab plane is poor. This may explain the low reactivity of bulk rutile TiO2 toward Li at room temperature. Researchers have reported nanometer-sized rutile TiO2 to be able to reversibly accommodate Li up to 0.55 Li/Ti (185 mAh g 1) between 1 and 3 V versus Li/Li þ at room temperature [128]. The improved reactivity results mainly from the drastically decreased diffusion path length. Researchers have also found that when particle sizes approach the nanoscale, the influence of surface and interfaces can no longer be neglected [129–131]. Hu et al. reported that Li surface storage on nanometer-sized particles can be energetically more favorable than bulk insertion [132]. Various experimental and theoretical studies have revealed that Li insertion into rutile lead to a sequence of phase transformations [125,126,132–143]. However, researchers do not yet have a robust understanding of the exact nature of Li insertion and the structural evolution of the final Li titanate phase; they do not unanimously agree upon the structure of the final Li titanate phase in nanostructured rutile upon Li insertion. Using first principles calculations, Koudriachova et al. [136] predicted that Li insertion likely proceeded through two-phase transformations: rutile first transforms to a Li-rich monoclinic phase (P2/m) with x ¼ 0.75; upon further lithiation, the monoclinic phase undergoes irreversible transformation into a layered hexagonal phase (R3 m). The experimental work by Borghols et al. [144] showed similar phenomena with rutile transformed into a monoclinic structure closely resembling the hexagonal phase. In contrast, independent experimental works by Baudrin et al. [138] and Wang et al. [128] showed that a rocksalt type phase LiTiO2 was formed after Li insertion in nanoscale rutile TiO2. Finally, Vijayakumar et al. [145] indicated that Li insertion into rutile TiO2 nanorods causes two consecutive phase transformations to Li titanate phases with spinel (Fd3m) and rocksalt (Fm3m) structures at x ¼0.46 and 0.88, respectively. In summary, researchers need to conduct further studies to elucidate the formation and evolution of Li titanate during Li insertion into nanometer-sized rutile TiO2. Amorphous materials have fascinating properties such as high mechanical hardness [146], chemical inertness [146], wide optical absorption range [147], and improved magnetic properties [148]. Researchers have overlooked the role of amorphous oxide electrode materials in the past due to the presumption that amorphous materials are less electrically conductive than crystalline ones. However, recent research shows that the increased concentration of interfacial regions in amorphous materials may form percolation pathways that facilitate ion diffusion [149–153]. It has been demonstrated that nanoscale amorphous electrode materials can serve as an “open framework” that can be electrochemically altered to form an optimal structure for improved LIB performance [154]. The electrostatic interaction of electrochemically altered materials provides a strong driving force for the diffusion of a large concentration of Li ions into amorphous metal oxide frameworks. This consequently leads to an ordering of the atomic building blocks, Li ions, and metal hosts into a crystalline array (Fig. 7). Inducing crystallization of nanomaterials in operando also enables the realization of the highest possible electrode capacity by optimizing the balance of electrostatic forces. This approach can achieve much higher specific capacities if the system is naturally allowed to choose and optimize its crystalline structure through a process of selforganization. The electrochemically induced structural evolution into high capacity/high power electrodes provides a powerful modular approach to designing improved battery materials with programmable physical and chemical properties. An irreversible phase transformation of nanoscale amorphous TiO2 into a face-centered-cubic phase (Fig. 7) by simply electrochemically cycling Li þ ions in and out of host structure has been observed [154]. It was found that high Li þ concentration in

10 12 14 16 18 20 2 theta

1.0

Discharge Li + insertion

0.05 A/g

0

100

200

300

1st

400

500

Specific capacity (mAh/g)

Fig. 7 Phase transformation of amorphous TiO2 nanotube (NT) electrode upon cycling vs. Li [153].

644

Batteries

Li

0 GPa

5 GPa

20 GPa

60 GPa

0 GPa

11 GPa

11.5 GPa

60 GPa

0 GPa

20 GPa

35 GPa

45 GPa

0 GPa

20 GPa

40 GPa

60 GPa

0 GPa

35 GPa

40 GPa

60 GPa

(A)

0%

(B)

40%

(C)

75%

(D)

85%

(E)

100%

Fig. 8 Molecular dynamic simulations of structural transition of cubic TiO2 under pressure for (A) delithiated, (B) 40% lithiated, (C) 75% lithiated, (D) 85% lithiated, and (E) fully lithiated. Red spheres: O; green spheres: Li; and white spheres: Ti.

the host structure facilitated by fast ion diffusion at nanoscale is the most critical factor for the phase transformation. A unique feature of the transformed cubic structure is that it exhibits long-range order with short-range disorder – well-ordered oxygen layers are separated by layers of mixed Li and Ti in lithiated titania. At delithiated state, the structure maintains its cubic phase without collapse and contains a high concentration of cation vacancies (B50%), which is desirable for Li þ intercalation as the structure can withstand extended cycling without capacity fading. It was found that under extreme conditions such as high pressure, the interstitial Li ions are driven to the cation vacancy sites in the oxide interior, leading to a significant improvement in the structural stability (Fig. 8) [155]. A recent report by Ceder’s group [156] also shows the benefit of cation-disordered oxides, using ab initio computations together with experiments on Li1.211Mo0.467Cr0.3O2 cathode. It shows that Li diffusion could be facile in

Batteries

645

cation-disordered oxides [156]. This unexpected behavior results from the percolation of active 0-TM channels with no facesharing TM ions in disordered Li-excess materials. In contrast, well-ordered oxides, with percolation for 0-TM diffusion well below threshold, tend to collapse and quickly lose their capacity upon disorder due to a significant electrostatic repulsion between TM polyhedral [157–160]. Although the benefits of introducing structural defects, such as cation vacancies in metal oxides for charge storage have been known since the mid-1980s [161,162], research in this area has been limited. There are several possible reasons for this such as battery community preference for well-ordered oxides, as well as analytical technique limitations given that it is quite challenging to elucidate and identify defect presence and distribution. With rapid advances in characterization techniques, elucidating the presence and distribution of defects at cation sites becomes possible, enabling the reexamination of defect-driven materials once overlooked for their potential application in energy storage.

2.20.3.3

Li-Ion Electrolytes and Their Impact on Performance and Degradation

LIBs have seen a dramatic growth in both numbers produced per year and also in applications where they are used. As use and number of applications have expanded, so too has the desire to improve performance and safety. Both can be effectively linked back to the properties of the electrolyte and the manner by which it degrades and impacts the degradation of other components. This section will provide a background on traditional, liquid-based electrolytes, which include organic carbonates before moving into a discussion on degradation and the role that phosphorus-containing electrolyte additives play in enhancing the durability and safety of LIBs.

2.20.3.3.1

Carbonate electrolytes

Over the course of the development of LIBs tens of thousands of different compounds and combinations of compounds have been investigated as electrolytes. A full review is outside the scope of this chapter, however, comprehensive articles have been published by Xu that provide a great deal of information on some of the more prominent electrolyte classes [163,164]. Common battery electrolytes often use organic carbonates as the solvent. These include ethylene carbonate (EC), ethylmethyl carbonate (EMC), dimethyl carbonate (DMC), propylene carbonate (PC), and diethyl carbonate (DEC). Other carbonates which have different reductive stabilities are also used in lower concentrations of a few percent as additives including vinylene carbonate (VC) and fluoroethylene carbonate (FEC). The salt component of the electrolyte has expanded some over the last several years, but LiPF6 remains the prominent salt used due to its ability to adequately meet several metrics including conductivity in organic carbonate solutions, sufficient dissociation constant, adequate thermal stability, and both chemical and electrochemical stability [163]. Generally to be successful a good electrolyte needs to provide a few distinct features. First, it must readily form a stable SEI, which is both ionically conductive and electrically insulating. The electrolyte itself also needs to have both high ionic conductivity and be electrically insulating. Second, the component of the electrolyte needs to effectively solvate Li ions, while maintaining low viscosity and high ionic conductivity. Additionally the electrolyte needs to have an effective operating window based on the desired application and ideally be safe to handle and have few deleterious safety features [163]. A good summation (Table 3) of desired electrolyte properties were released by the USABC as part of the request for proposal information on the development of advanced high performance electrolytes for LIB used in vehicle applications [165]. Table 3

Desired electrolyte properties

Metric

Unit

Goal

Cost

$/kg

o10

Conductivity (

301C)

mS/cm

High voltage stability

V vs. Li/Li

Vapor pressure (301C)

mm Hg

44 þ

5.0 o1

Flashpoint

1C

4100

Purity of individual components

%

499.99

Conductivity (301C)

mS/cm

412

Li

þ

transference number

40.35

Viscosity (301C)

cP

o5

Viscosity (

cP

o20

Water content

ppm

o20

HF content

ppm

o50

Component Purity

%

499.99

301C)

646

Batteries

While meeting the metrics proposed by the USABC for formulated electrolytes is vital, as important is the ability of the electrolyte to repeatedly be exposed to the actual application cycling and environmental conditions. One key to achieve a long-life LIB is the effective formation of a SEI for cells that contain graphite and alloying negative electrodes. SEI formation is a complex area, which has been thoroughly studied over the last few decades. Much of the initial work on the formation of the SEI on Li metal anode was performed by Peled and coworkers in the late 1970s and early 1980s on Li metal [166,167] and is transferable to the realm of LIBs. The key function of the SEI is that it passivates the anode in the LIB, while still allowing ionic (Li þ ) transport. In doing so the SEI provides stability to the thermodynamically hindered LIB. Generally, SEI formation is dictated by the combined electrochemical and chemical reactions, which take place in the first few cycles of the battery. While the SEI formation process is similar regardless of the anode material there are a few distinct differences due to expansion and surface dynamics, which occur when transitioning between the more common anode materials in graphite and Si. That said many of the general reactions in EC-containing electrodes are the same and included in the partial list below [168,169]. Reactions at the positive electrode have been less well characterized. 2EC þ 2e þ 2Liþ -ðCH2 OCO2 LiÞ2ðinsolubleÞ þ CH2 ¼ CH2

ð25Þ

EC þ 2e þ 2Liþ -LiCH2 CH2 OCO2 LiðinsolubleÞ

ð26Þ

H2 O þ ðCH2 OCO2 LiÞ2 -Li2 CO3ðinsolubleÞ þ CO2 þðCH2 OHÞ2

ð27Þ

2CO2 þ2e þ 2Liþ -Li2 CO3ðinsolubleÞ þ CO

ð28Þ

1 H2 O þ e þ Liþ -LiOHðinsolubleÞ þ H2 2

ð29Þ

1 LiOH þ e þ Liþ -Li2 OðinsolubleÞ þ H2 2

ð30Þ

LiPF6 ⇌ LiFðinsolubleÞ þ PF5

ð31Þ

PF5 þ xe þ xLiþ -xLiFðinsolubleÞ þ Lix PF5

xðinsolubleÞ

ð32Þ

While Eqs. (25) and (26) are written for EC, they can be readily modified for both FEC and other molecules which resemble EC, but have different substituent groups [170]. For each case the path is an initial one electron reduction, which generates a reactive radical species that subsequently undergoes a set of chemical reactions to form the end products in the first two equations [170,171]. This initial process typically occurs between the 1.7 and 1.5 V versus Li/Li þ for noble metal electrodes [172–174] and on graphite anodes the value is typically reported as 0.7 V lower with initiation occurring closer to 0.8 V versus Li/Li þ [175,176]. Following the initial reduction of EC and the formation of insoluble products, secondary reactions occur at more negative potentials, which lead to a multicomponent SEI that contains components which are both organic and inorganic in nature [177,178]. Aurbach played a key role in helping in the identification of surface species and in relating the correlation between voltage and reduction products present in the SEI [168,172,179]. Of note is that the inorganic components are viewed as essential to facilitate Li þ transport within the SEI [169]. The presence of different components, the voltage dependence and the partial solubility of different SEI components results in variable thickness and mass of the SEI during early cycling [173,180]. However, for effective electrolyte systems a relatively passivating (though not fully stable) SEI is formed within the first 15–20 cycles [174]. This stability can be characterized by the use of Coulombic efficiency during early cycling of the battery to better understand potential implications to the life of the battery and the role that different electrolyte compositions and additives play [181]. While the ability to form an effective SEI is a key function of the electrolyte in LIBs, it is not the only function of the electrolyte. Following SEI formation the electrolyte still plays a key role in transporting ions between the two electrodes in the battery. However, as a system that is composed of both inorganic salt, such as LiPF6 and organic carbonates the ability to transport ions and to maintain performance can be greatly impacted by degradation of one or more of the electrolyte components. The stability and safety of electrolyte components can be broadly viewed as the sum of the different parts of the electrolyte. From a safety perspective one of the key limitations associated with electrolytes is that some of the carbonates, especially the linear carbonates, such as DMC and EMC have high vapor pressures and low flashpoints and that many of the SEI forming reactions (Eqs. (28)–(30) above) result in the generation of combustible gases. This compounded with the fact that most commercial battery cells are equipped with vents to prevent over pressurization of the battery can lead to unintended consequences, such as fire after a battery vents. Generally one of the key issues with the current electrolyte blends that contain both organic carbonates and LiPF6 is that they begin to degrade at relatively low temperatures between 50 and 851C [182,183]. The degradation is clearly seen Fig. 9 where over the course of 25 h the conductivity of two different electrolytes, which contain LiPF6 and combinations of linear (DMC and DEC) and cyclic carbonates (EC) decreases by roughly 25% in both instances [182]. The impact of this decrease on battery performance would be expected decreased power capability and a diminished operational life.

Conductivity (mS cm–1)

Batteries

20

647

1:1:1 EC:DMC:DEC DMC

18 16 14 12 10 0

5

10

15

20

25

Time at 85 °C Fig. 9 Impact of exposure to high temperature on conductivity of different lithium-ion battery (LIB) electrolytes. Electrolytes were stored at 851C for various lengths of time. DEC, diethyl carbonate; DMC, dimethyl carbonate; EC, ethylene carbonate. Data from Ravdel B, Abraham KM, Gitzendanner R, et al. Thermal stability of lithium-ion battery electrolytes. J Power Sources 2003;119:805–10.

The drop in conductivity, which has a direct link to battery performance that is seen in Fig. 9, can be attributed to the impact that degradation of LiPF6 has on the battery electrolyte. As mentioned above, LiPF6 has many good characteristics as the salt component of the electrolyte, however, it also has a distinct degradation pathway that impacts other components of the electrolyte and which can directly impact the performance of both the positive and negative electrodes. Under standard conditions an equilibrium exists between LiPF6 and the formation of F and gaseous PF5 as shown in Eq. (31) above. In the presence of Li þ this readily results in the formation of solid LiF, which is either incorporated as part of the SEI or which precipitates at other locations in the cell. The impact of LiF formation is that the overall Li available for storage capacity in the battery is diminished. At elevated temperatures the equilibrium for PF6 degradation is shifted to the generation of more F and PF5. In addition to the equilibrium reaction (Eq. (31)) there are other subsequent reactions, which occur with LiPF6 and PF5 in the presence of H2O, as shown in Eqs. (33) and (34). Secondary reactions with the protons associated with HF or H2O can also occur resulting in additional formation of PF5, LiF, PO2F2 , and PO3F2 and in the formation and further reduction in the cell Li inventory [182,184,185]. LiPF6 þ H2 O-LiF þ POF3 þ 2HF

ð33Þ

PF5 þ H2 O-POF3 þ 2HF

ð34Þ

After degradation of PF6 occurs reaction with the other components of the electrolyte have been observed with the formation of OP(OEt)F2 and OP(OMe)F2 being observed in DEC and DMC, respectively [182]. The generation of PF6 degradation products and the resulting chemical degradation of the base solvent molecules in the electrolyte have significant impacts on the ability of the electrolyte to maintain sufficient performance as shown in Fig. 9. In addition, there are impacts to other cell components including some positive electrode materials, such as high voltage LiCoPO4 and LiNi1/3Co1/3Mn1/3 (NCM) [186,187]. In the case of LiCoPO4 the degradation path in the presence of HF can be linked to a nucleophilic attack on the P–O bonds by the F . This results in the generation of the soluble species PO2F2 , which is also a result of the degradation of POF3 described above [187,188]. The result of solubilizing the phosphate in the positive electrode is a significant and rapid degradation in capacity and as a result has limited the use of LiCoPO4. Other positive electrode materials have also shown impacts associated with the presence of HF and LiPF6 degradation products. In a study performed by Gallus et al. it was found that the rate of metal dissolution from NCM electrodes was enhanced by the use of LiPF6 with the extent of dissolution of Ni, Mn, and Co all increasing at the upper cut-off voltage of the cell was increased [186]. The enhanced dissolution of these metals plays a part in overall capacity fade for the positive electrode, but can also significantly impact the negative electrode performance due to the deposition of the metals into the SEI [189,190].

2.20.3.3.2

Phosphorus-containing electrolyte additives

Due to the crucial nature of the electrolyte both the possible safety concerns which arise due to the volatility of linear organic carbonates and the degradation associated with LiPF6 suggest that there remains room for the improvement of electrolyte properties. To help address many of the issues of traditional electrolytes for LIBs, a host of different solvents have been investigated including ethers, esters, nitriles, room temperature ionic liquids (RTILs), sulfones, and sulfoxides [159]. Another class of solvent compounds which have received interest are those that contain phosphorus. This class includes phosphates [191–193], phosphonates [194–196], phosphazenes [197–201], and other smaller molecules with both phosphorus and nitrogen, such as phosphonamidate [202], phosphoranimine [203], and phosphorimidic compounds [204]. Each of these compounds has a similar set of goals. The first is to reduce the flammability of the solvent. This arises from the well-known nonflammable nature of phosphorus and phosphorus-containing compounds. The second is to increase the overall stability of the electrolyte solvents by improving cell performance. Structurally there are several similarities between phosphates and phosphonates with phosphates having four O atoms linked directly to a central P and phosphonates having three O linkages with the fourth position being occupied by a different

648

Batteries

Closed-cup flashpoint (°C)

substituent. In both cases one of the O linkages is through a double bond. Due to the similarity between the molecule types there are significant similarities in the structure of the phosphate and phosphonate molecules, which have been used in LIB electrolytes. Typically both classes rely on small substituents, such as single or two carbon chains attached at each of the non-doubly bound O molecules. Exceptions to this include the use of triphenyl phosphate (TPP) and instances where fluorinated carbon chains are bound to the O atoms. While phosphates and phosphonates are dominated by direct links between a P and O atoms, phosphazenes, phosphoranimines, and phosphorimidic compounds all share a similarity in that they have at least one double bound between a P and an N in the compound. This leaves three available positions for different substitutions on the P that is doubly bound to the N creating a high degree of flexibility when designing electrolyte molecules. Adding additional flexibility is the ability to synthesize compounds, which are either cyclic in nature as is the case for cyclotriphosphazenes (CTPs) [199–201,205] or which are linear, similar to the phosphoranimines and phosphorimidic compounds (PNP) which have been recently reported [203,204]. The key differentiator between the CTPs and the linear molecules is that the phosphoranimines and PNP have lower molecular weights and as such can potentially be used at higher loading levels in the electrolyte. Both phosphoranimines and PNP have a single P ¼ N unit, but PNP has an additional phosphoryl pendant. Total loading level of P-containing compound in the electrolyte is seen as a distinct way to improve overall safety of the electrolyte. Recall from above that EMC and other linear carbonates have high volatility and in part drive many of the fire and flash concerns associated with electrolytes. Thus means to reduce the overall loading of EMC or other high volatility carbonates is key to improving safety. A direct way to reduce content is the inclusion of P-containing or other low volatility additives in the electrolyte blend. The result that this type of substitution has on the safety of the electrolyte can be seen in Fig. 10 where the reported closedcup flashpoints for multiple CTP and phosphoranimine compounds are compared [199,203,205]. As is clear from the figure, while each of the compounds has distinct structure and imparts changes to the electrolyte, such as changes to the conductivity and viscosity, the overall change in safety response when evaluated using flashpoint is still distinctly tied to the level of EMC present in the electrolyte. Another means to evaluate the safety enhancing performance of P-containing electrolytes is to directly compare their selfextinguishing time (SET). This test uses a small amount of electrolyte, which is soaked into a glass-fiber ball to evaluate how long it takes a specific amount of electrolyte to self-extinguish. Due to the wide assembly of different baseline electrolytes used, it is difficult to directly compare SET values across a set of electrolytes. Further complicating the comparison is the fact that there is evidence suggesting that there is a distinct link between SET and the substrate that the SET evaluation is performed on where SET values on watch-glasses fall significantly more with increased additive loading than when the measurement of the same sample is performed on an Al substrate [192]. That said the inclusion of 10–20% P-containing additive typically reduces the SET by 30–50% [192,195,204]. The primary difference in the use of flashpoint and SET when evaluating flammability of electrolyte is that flashpoint tends to report the impact of the most volatile portion of the electrolyte and the safety of the electrolyte with respect to a condition where it could be volatilized, such as during the venting of the battery, while the SET measure encompasses the interactions which occur with the different chemicals in the electrolyte and is more representative of the safety hazard associated with a burning battery. While many groups report one or the other to indicate flammability both, if performed in a reproducible, quantitative manner, are truly necessary when evaluating the entire safety response of an electrolyte. While safety is important for batteries it is inconsequential if the proposed electrolytes do not function effectively in an electrochemical system. As mentioned earlier, a key to the performance of LIBs is the ability to form an SEI, which is both ionically conductive and electrically insulating. Failure to meet these two requirements results in a battery that has shortened life or poor power capability. The ability to form effective SEIs for P-containing electrolytes has been studied by several researchers. In one instance, Sazhin and coworkers developed methods to understand SEI stability and formation dynamics on noble electrodes [174]. In their work the researchers found that some CTPs impacted the formation and stability of the SEI and resulted in a larger EC:EMC only blends CTP from Ref. [199] CTP from Ref. [205] From Ref. [203]

36 34 32 30

0.45

0.50

0.55

0.60

0.65

0.70

Volume % EMC

Fig. 10 Comparison of flashpoints for various electrolytes which contain either cyclotriphosphazenes (CTPs) or phosphoranimines. EC, ethylene carbonate; EMC, ethylmethyl carbonate. Data from Rollins HW, Harrup MK, Dufek EJ, et al. Fluorinated phosphazene Co-solvents for improved thermal and safety performance in lithium-ion battery electrolytes. J Power Sources 2014;263:66–74; Dufek EJ, Klaehn JR, McNally JS, et al. Use of phosphoranimines to reduce organic carbonate content in Li-ion battery electrolytes. Electrochim Acta 2016;209:36–43; Harrup MK, Rollins HW, Jamison DK, et al. Unsaturated phosphazenes as co-solvents for lithium-ion battery electrolytes. J Power Sources 2014;278:794-801.

Batteries

3:7 EC:EMC, 1.2 M LiPF6 (baseline) Baseline with 10% phosphoranimine Baseline with 9% phosphoranimine, 10% VC

1.05 Capacity retained

C/10 C/5

1.00

649

C/3

C/5

0.95 C/1 0.90 0.85 10

20

30

40

Cycle Fig. 11 Impact of the inclusion of the solid electrolyte interphase (SEI) forming additive vinylene carbonate (VC) on the performance of a phosphoranimine. EC, ethylene carbonate; EMC, ethylmethyl carbonate. Figure modified from Tsujikawa T, Yabuta K, Matsushita T, et al. Characteristics of lithium-ion battery with non-flammable electrolyte. J Power Sources 2009;189:429–34.

electrochemical window. Other work looking at SEI stability has found that many P-containing electrolytes benefit from the addition of SEI forming additives such as VC or Li bis(oxalatoborate) (LiBOB) [203,206]. In both instances, one using a phosphoranimine (N-trimethylsilyl-P,P-bis((2-methoxyethoxy)ethoxy)-P-ethylphosphoranimine) [203] and the other using dimethyl methylphosphonate [206], electrolyte blends using just standard carbonates and the P-containing additives had subpar performance, however, with the addition of the SEI forming additives performance improved to near what was observed for baseline electrolyte conditions. The distinct impact for the phosphoranimine containing electrolyte can be seen in Fig. 11. The impact of the inclusion of VC is discretely evident during the initial cycling and formation of the SEI as well as in the overall capacity fade over the course of 50 cycles [203]. One highlight of using smaller molecules such as PNP and phosphoranimines rather than CTPs is that their lower molecular weight should improve transport properties and over power or rate capability of the cells. In comparing the viscosity and conductivity of CTPs and phosphoranimines in solution it was found that the phosphoranimines maintained more desirable conductivities and viscosities than CTPs and that loading levels of nearly two times could be achieved, while keeping a set viscosity or conductivity [203]. The impact of this change in transport properties is that a less profound impact on rate is seen. When comparing rate performance between a C/10 and a C/1 rate for CTPs and the phosphoranimines for baseline and different loading levels the impact of the smaller molecules can be seen where there is negligible loss for the phosphoranimines at a 9% loading but there is a 5–15% reduction in capability with the larger CTPs [199,203]. While there has been some observed impact on rate capability, long-term cycle life is also a key consideration for LIBs where hundreds if not thousands of cycles are expected from consumers and manufacturers alike. Several different P-containing electrolytes have shown that over the course of at least 100 cycles that performance on par or better than baseline electrolytes can be achieved [191,199,204]. In the case of the CTPs and PNP there was little justification given for the long-term performance, however, in the work of Cresce and Xu it was noted that the HFiP additive provided interphasial benefits and led to overall improved high voltage performance [191]. While the reasoning for the improved performance with the phosphazene chemistries is not given it is plausible that interactions between the nitrogen in the molecules, which are Lewis bases, could be inhibiting the degradation of PF6 anions. Supporting this interaction are NMR experiments performed by Lucht et al. which showed a variance in the fluorine spectra with the inclusion of different Lewis bases that contained nitrogen [207]. As a set of different electrolyte components, P-containing electrolytes show possibilities for improving both the durability and safety of LIBs, however, there is still improvement that needs to occur and more understanding and refinement of transport properties remains a key consideration.

2.20.3.3.3

Concentrated electrolytes

Recently other means to improve the safety of LIB have been investigated, which are based on three primary areas: RTILs, concentrated organic, and concentrated aqueous systems. RTILs have been and continue to receive significant research interest. A key portion of the advantage of RTILs is that they can be viewed as a means to improve safety by inhibiting flammability, while still providing sufficient performance. The works of Xu [163,164] provide highlights of RTILs. In the arena of organic electrolytes, Dahn and Yamada have both recently begun to look at concentrated electrolyte systems using ethyl acetate and acetonitrile, respectively [208,209]. In the work of Dahn the researchers investigated the use of different salts in combination with the non-carbonate solvent. At high salt concentrations, including non LiPF6 salts, it was found that noncarbonate solutions can be used, while still achieving desirable cell performance [208]. The work of Yamada used acetonitrile and high concentrations of lithium bis(fluorosulfonyl) amide (LiFSI) to show stabilization of both Li metal and insertion of Li into graphite [209]. The combined results are significant as they provide a possible new route to looking at safer more durable electrolytes, which do not have some of the concerns of carbonates and LiPF6.

650

Batteries

The second set of investigations into high concentration electrolytes shifted the focus from organic solvents to the use of aqueous systems. The water-in-salt [210,211] and hydrate melt electrolytes have shown relatively rapid improvement in performance and have achieved energy densities in excess of 130 Wh kg 1 [212]. These systems rely on high concentrations of single or two salt systems to effectively passivate both the positive and negative electrode of the battery. This passivation much like the formation of the SEI in organic electrolytes allows the aqueous batteries to operate outside the typical electrochemical window for water. The widening of the electrochemical window enables the use of electrodes, such as Li titanates (LTO) and LiMn2O4 as negative and positive electrodes, respectively. While the achievable voltage window for water-in-salt systems falls below that of traditional LIBs, these aqueous batteries are receiving considerable attention due to the improvements to safety that they bring and based on the fact that the environmental impact of aqueous electrolytes minimizes the need for production and handling of organic carbonate electrolytes.

2.20.4

Analysis and Assessment

Various types of analysis techniques are used in the field of LIBs. Each technique provides unique information of the LIB system.

2.20.4.1

Electrochemical Analysis

The performance of a battery system can be assessed by a few cell characteristics, such as charge/discharge profiles, Coulombic efficiencies, rate capabilities as well as cycle-life studies. Slow scan rate cyclic voltammetry (SSCV), electrochemical impedance spectroscopy (EIS), potentiostatic (PITT), or galvanostatic intermitten titration (GITT) [213] can be used to estimate the processes of phase transitions, charge transfer, as well as solid-state diffusion.

2.20.4.2

Morphological and Structural Analysis

In LIBs, Li ions have to insert/extract into/from active materials repeatedly. Ideally, such process is reversible. The loss in reversibility is often related to irreversible processes occurring in the bulk, such as irreversible phase transformation or crystalline structure change in materials during cycling. The kinetics of battery materials is directly related to the size, morphology, and architecture of the active materials, nanosized materials have shown advantages in terms of higher power density and better mechanical strength as compared to their bulk peers [88]. The size and morphology of electrode materials can be monitored by advanced electron microscopy techniques, such as scanning electron microscopy (SEM) and transmission electron microscopy (TEM). Structural defects in electrode materials can be evaluated on an atomic level using analytical TEM capabilities. In addition to classical imaging techniques, chemical analysis by TEM, using techniques such as energy-dispersive X-ray spectroscopy (EDS), electron energy loss spectroscopy (EELS), and electron diffraction is opening another dimension for analysis of complex materials [214]. These techniques can provide not only the difference between components, but also direct information on physical and chemical properties. Z-contrast imaging in scanning transmission electron microscopy (STEM) [215] is able to follow the structural evolution of electrode materials upon electrochemical cycling as well as differentiate the surface/bulk characteristics in electrode particles. In addition, in situ TEM can follow the volumetric changes within electrode upon charge/discharge cycling. Transmission X-ray microscope (TXM) can be used to produce a three-dimensional chemical map of battery materials under working conditions [216–218]. Atom probe tomography (APT) provides nanoscale surface, bulk and interfacial materials analysis of simple and complex structures with atom by atom identification and accurate spatial positioning [219]. APT can be used to reveal the Li distribution in metal electrode, correlating theoretical predictions. Ex situ and in situ synchrotron X-ray microscopy and spectroscopies (X-ray scattering, X-ray, etc.) as well as neutron scattering [220–222] are sensitive powerful techniques to probe the structures of battery materials under synthesis and operations. While XRD provides insight into the crystal structures of battery materials, X-ray absorption near edge structure (XANES) and extended X-ray absorption fine structure (EXAFS) provide information on the long-range and short-range chemical disorder in addition to the oxidation state of the probed elements. Solid-state nuclear magnetic resonance (NMR) is a powerful method for studying the local structures and electronic properties of electrode materials as a function of the state of charge (SoC) [223,224]. 57Fe Mössbauer spectroscopy is a well-developed method to characterize small shifts of Fe element in nuclear energy level with hyperfine precision, which could be used to distinguish the tiny difference of Fe local chemical environment in iron-containing active materials.

2.20.4.3

Surface Analysis

The formation and evolution of SEI on the electrode surface is a key parameter in the stability and safety of a battery. SEI is a very complicated layer consisting of both inorganic components of salt degradation products and organic components of reduction products of the solvent of the electrolyte [225]. A variety of surface techniques can be used for SEI analysis. A comprehensive review of surface analysis techniques of SEI is out of the scope of this book chapter, Novak et al. have summarized available SEI analysis in a great details recently [225]. The most frequently used surface analysis techniques are highlighted here. X-ray photoelectron spectroscopy (XPS) coupled with Fourier transform infrared spectroscopy (FTIR) provide a good picture of SEI. Both techniques are highly surface sensitive and are complementary to each other. FTIR provides information of vibrational energy of

Batteries

651

various bonds, which is characteristic of the corresponding functional group. XPS allows analysis of most elements (except H and He) present in concentration 40.1 atomic percentage in the outermost 10 nm of the surface by measuring the chemical shifts in the binding energies of an element [226]. EIS is able to differentiate processes occurring at different timescale and is used to study the evolution of the interphase impedance. Atomic force spectroscopy (AFM) can be used to provide morphological information regarding the SEI.

2.20.4.4

System Level Analysis

Battery manufacturers are now able to produce high quality Li-ion cells for portable electronics with less than one reported safety incident for every one million Li-ion cells produced [12]. However, such failure rate at cell level is still too high for PEVs and EVs as several hundred Li-ion cells will be needed to power a vehicle [12]. The failure of a single cell can generate a large amount of heat, which can trigger thermal runaway of neighboring cells to cause system failure. Comprehensive and mature BMSs are currently available in portable electronics, but they are not fully developed for EVs [227]. The major hurdles for this lag are (1) battery state evaluation; (2) battery modeling; and (3) cell balancing. Battery prognostics address two critical issues [228]: (1) uncertainty of battery behavior and internal characteristics; (2) safety assurance. Deploying prognostics technologies for batteries are very challenging due to the complex characteristics of electricchemical processes occurred within batteries. In addition, limited information is available from real EV data to establish or validate the battery models. Battery SoC estimate the amount of energy remaining in a cell compared to the energy when it is fully charged, which informs the users of how long it will need recharging. The accurate estimate of SoC significantly affects battery health and safety over time. SoC estimation is challenging and depends on battery chemistry and condition. It is often classified into two categories: direct measurement and indirect measurement [228]. In practice, it is quite challenging to conduct direct measurements due to the need for high accuracy devices and high cost. Therefore, indirect measurements are commonly used. Generally, SoC can be estimated through indirect measurements in two ways: offline and online. Offline measurements give accurate estimation but are time consuming, expensive and interrupt main battery operation. Hence, online measurements, such as Coulomb counting is commonly used for estimation of SoC. State of health (SoH) compares the specified performance and health condition of a used battery with a brand new battery of the same time. Unlike SoC, there is no fixed definition for SoH. Three main approaches are currently used in battery prognostics: physical-model, data-driven, and fusion-model approaches [228]. The physical-model approach considers the physical processes and failure mechanisms but it cannot detect intermittent failures. The data-driven approach is useful when system-specific information is not available and the drawback is that if the data are not available or biased the results can be incorrect. The fusion (or hybrid) approach is desired as it combines the two aforementioned approaches to achieve an optimal solution. Main challenges in developing an effective prognostics approach for battery systems include uncertainty in the analysis of mobility, durability, and safety during battery life [228]. In order to address such challenges, it is necessary to research techniques that can define dynamic conditions such as road slope or driving mode of the vehicle.

2.20.5 2.20.5.1

Illustrative Example – Regarding Low-Temperature Limitations of Lithium-Ion Batteries Challenges for Batteries in Cold-Climate Energy Storage

Conditions within cold climates represent a fundamental roadblock to satisfactory performance of battery energy storage systems, and pose a premier technical challenge for the development and deployment robust battery systems for northern latitudes that will exhibit reliable and economic performance for year-round applications. Deployment of battery energy storage into cold-climate regions is hindered by significant reduction of power at lower temperatures. In northern latitudes, achieving energy availability and affordability within a greater application space calls for improved performance of LIB technologies in terms of power and longevity over a broad and diverse set of conditions. All chemical and electrochemical systems obey fundamental laws and limitations imposed by thermodynamics and other physical laws. In energy storage, aspects of efficiency, robustness, safety/reliability, and longevity are greatly impacted by how the local environment influences thermodynamic behavior, and these metrics suffer in cold-climate energy storage (CCES) scenarios. There is still much to be learned on the scientific level about the root causes for such limits, which can have their origins in multiphysics and multidimensional domains. Added to this is the engineering required to adapt a given technology (e.g., battery) to particular CCES applications, environmental factors, as well as logistical challenges prevalent at remote locations. As a consequence, advanced clean ES is an expensive and challenging endeavor in many northern regions, placing it out of reach for many end-users. If eventually resolved, a stable and increased source of electrical power would enhance the lives of those residing in such regions. In some applications, mitigating limitations in cold-temperature performance also improves the technology for broader use elsewhere, as is the case for battery ES. Previous work at the Idaho National Laboratory (INL) sponsored by the US Department of Energy Vehicle Technologies Program (DOE-VTP) discovered that additional thermodynamic mechanisms emerge at low temperatures that are tied to how the labile electrolyte species interact with electrode surface regions (passivating films, SEI) and porous regions within active electrode hosts, effectively diminishing the surface area available for charge transfer reactions. Low temperatures reduce the energy barrier

652

Batteries

required to form assemblies of solvent molecules (monolayers, bilayers, solid solvates, etc.) that greatly impede ionic transport through the interface and reduce the effective surface area. A fundamental challenge is then centered on the core scientific question of can we alter and optimize Li-ion cell interfaces (chemically, rheologically, physically, electronically) to enhance ionic migration and charge transfer kinetics at low T and hence mitigate several hindrances to deployment of vehicle and grid battery systems at northern latitudes and in other applications (aircraft and aerospace) that have similar constraints. Current LIB technology suffers tremendous power loss at low temperatures, where losses of over 50% are not uncommon at subzero temperatures (centigrade). This technology weakness reduces the extent of deployment of such batteries into various application sectors and geographic locations. Improvements have conventionally involved sacrificing energy storage capacity to permit use of thinner electrodes having quicker cycling rates or adding expensive ThM schemes, but this undercuts ability to economically meet emphasized metrics like extended range of electric drive vehicles or robust battery ES for grid applications. Complications that arise in LIBs operated at low temperatures are manifold:

• • • • • • •

High overpotentials due to polarization, with great loss of power and achievable capacity. Slower charging required to avoid Li deposition. Non-Arrhenius behavior in some performance metrics, with higher Ea at lower T. Extra expense from ThM to keep batteries warm. Extra expense from overdesign of string assembly to compensate for low performance. Added weight from ThM system and battery overdesign undercuts system efficiency. Additional battery aging from stress incurred from daily thermal cycling (DTC). As a start, contributions to reduced performance at low temperatures would include:

• • • • • •

Diminished transport properties of electrolyte (higher viscosity, lower conductivity). Divergence of cathode versus anode electrode rates as temperature is lowered. Polarization within electrochemical double layers can worsen electrolyte transport properties therein. Thermodynamic Consequences: (1) higher Li desolvation energy (affects charge transfer resistance, Rct). (2) phase separation at interfacial regions (e.g., solvent arranging in porous spaces). Thermodynamic and electrochemical: Li deposition, dendrite formation. Mechanical (particle fracturing due to DTC combined with cell cycling).

While some such limitations are mentioned in the literature, there has been little done to adequately diagnose root causes and form a mitigation approach [229–232]. However, early indications and insights on how to diagnose and remedy CCES limits have been provided by one of the authors [233]. We note that porous regions could exacerbate the overall problem due to local surface tension effects within confined regions. And, the polar nature of solvents can lead to self-assembly when conditions are favorable in terms of electrode surface attributes, for example, by providing a locus or epitaxy whereon solvent assembly might occur.

2.20.5.2

Supporting Data

Evidence of CCES limitations in batteries can be seen through several avenues. First, complex impedance spectra (EIS) for 18650type Li-ion cells based on NiCoAl-oxide cathode and graphite anode obtained over a wide range of temperature show activation energies between 50 and 70 kJ mol 1 in Fig. 12, based on the Arrhenius analysis of the interfacial impedance component defined by the sum of semicircle widths from Nyquist plots (“Gen2” cells, Quallion LLC: LiNi0.8Co0.15Al0.05O2 cathode (35 mm laminate), a MAG10 carbon anode (also 35 mm laminate) and a 2300 series Celgard separator.). The electrolyte in these cells is 1.2 M LiPF6 in EC-EMC (3:7, mass). Yet, the Arrhenius analysis of conductivity for this electrolyte system shows activation energies (Ea) that generally only fall within 10–36 kJ mol 1 as seen in Fig. 13 (as predicted by the INL Advanced Electrolyte Model (AEM)), giving Ea values that are roughly half or less than that of Fig. 12. The range of salt concentration in Fig. 13 is meant to account for moderate differences in salt concentrations within double-layer regions. Hence, it is apparent that the electrolyte in contact with the electrode surfaces in a complete cell results in a compounding of poor electrochemical behavior, originating at interfacial regions. Another piece of evidence is Fig. 14, wherein sustained growth of interfacial impedance is seen for another type of Li-ion cell (Sanyo Y chemistry, NMC-spinel/Gr; 18650 configuration) kept at a low temperature for prolonged time. While being held more than 150 h at 301C, the cell’s interfacial impedance grew an extra 25%, indicating a transient (yet reversible) process that impeded Li cation transport through the interfaces. Early trials using DSC also showed that the amount of heat release/gain for samples of battery materials (electrode swatch plus electrolyte) was on the order of three times greater than the thermodynamic latent heat for the materials alone. Collectively, these data speak to the emergence of reversible thermodynamically driven phase behavior of labile electrolyte species (solvent and/or ions) at surface loci that provide lowered interfacial energy for phase separation mechanisms, often referred to as a Gibbs adsorption process. Localized phase behavior always infers an increase in local order of affected species, and a commensurate drop in local entropy. A corresponding heat signature (DSC) or configurational signature (ATR-FTIR) can be used to surmise the presence and magnitude of phase behavior along the temperature coordinate. Such an electrolyte-to-surface phase process in a cell environment would be a highly complicated consequence of numerous factors such as surface morphological features, porous region attributes, chemical and electronic nature of the SEI passivation

Batteries

7 6

Activation energies (kJ mol−1): 60% SOC, low PF A: 59.0 100% SOC, low PF B: 59.2 60% SOC, high PF C: 52.5 100% SOC, high PF D: 54.7

A B

5

C 4

653

A′: 62.6 B′: 66.4 C′: 58.5 D′: 61.9

ln (1/Rint.)

D 3 Transition region 2 occurs at −13.2 °C (50% PF cells)

2 1

A′

–0

B′

–1

C′

Transition region 1 occurs at 3 °C (low% PF cells)

–2 –3 0.0030

0.0032

0.0034

D′

0.0036

0.0038

0.0040

0.0042

Inverse temperature (K–1) Fig. 12 Arrhenius plots of electrochemical impedance spectroscopy (EIS) interfacial impedance measurements derived from Nyquist plot semicircle regions (Gen2 18650 Li-ion cells). For most temperatures, activation energies are a weak function of cell aging or the state of charge (SoC). PF, power fade.

Predictions by AEM Conductivity (mS cm−1)

10

0.5 m 1.0 m 1.5 m

1

Activation energies range from about 10 to 36 kJ mol−1

0.0044

0.0042

0.0040

0.0038

0.0036

0.0034

0.0032

0.0030

0.0028

0.1

Inverse T (K–1)

Fig. 13 Arrhenius plots of conductivity for electrolyte EC-EMC (3:7, mass) þ LiPF6 used in Gen2 Li-ion cells of Fig. 1. Greater activation energies are seen at lower temperatures and higher salt concentrations. AEM, Advanced Electrolyte Model; EC, ethylene carbonate; EMC, ethylmethyl carbonate.

films, chemical, configuration, and electronic nature of electrolyte compounds, surface and electrolyte impurities, system temperature, separator properties, state of battery charge, and other factors. Note that all the above data represent systems that are essentially at rest, i.e., not electrochemically operating. Limitations in CCES are further compounded as the cells experience active charge and discharge excursions. Fig. 15 provides a good overall summary of how cell capacity diminishes at lower temperature as a function of cycling rate (Regarding the ‘C’ nomenclature, the C rate designation signifies the amount of nominal energy storage delivered or extractible in 1 h going from full to no charge (or visa versa); C/10 would be this amount of energy transferred over 10 h, 3C would be performed over 1/3 h, etc.), and how this metric varies between charge and discharge modes. Coin cells (2032 type) were used in this study, having the same Gen2 cell chemistry noted above. Polarization effects at lower temperatures and higher cycling rates cause voltage limits to be reached prematurely in the cells, which result in lower achievable capacity. Also seen is the greater sensitivity of charge capacity to temperature and rate, indicating the need for slower rates of charge at cold conditions to avoid formation of Li dendrites at the anode. In yet another study at INL involving low temperatures, complicated kinetic analyses of DC pulses of full cells showed that both the effective

654

Batteries

Time at low-temperature EIS, Sanyo Y –1 –0.8 4 h EIS

–0.6

8 h EIS 14 h EIS

Z″ (Ohm)

–0.4

24 h EIS –0.2

42 h EIS 0

1

2

3

4

0

76 h EIS 142 h EIS

Z′, Ohm

152 h EIS

0.2 0.4 0.6

Fig. 14 Nyquist plots for a Sanyo Y Li-ion cell (NMC/spinel þ graphite) kept at 301C over prolonged time. The real component to impedance grew by approximately 25% over the course of this test, indicating a dynamic process that affected conductance of interfacial regions. EIS, electrochemical impedance spectroscopy; NMC, Li[Ni1 y zCoyMnz]O2.

Capacity (charge or discharge) (mAh)

10 30 °C

10

0

–10

–20

–30 °C

1

0.1 Discharge 0.01

C/1 rate C/3 C/5 C/10

Charge C/1 C/3 C/5 C/10

0.001

0.0001 0.0032

0.0034

0.0036 1/T

0.0038

0.0040

0.0042

(K–1)

Fig. 15 Polarization plots obtained from Li-ion coin cells containing a Gen2-type cell chemistry (NCA/Gr). An adverse effect on both charge and discharge capacities is seen at temperatures below –101C, and for cycling rates greater than C/3. NCA, Li[Ni1 y zCoyAlz]O2.

surface area for charge transfer (yeff) and the exchange current density (io) decrease at lower temperatures, indicating constrained interfacial transport and greater electrochemical irreversibility [234]. Another consideration is the impact of DTC on materials integrity when a significant cold-start condition is encountered. To date, little is known about Li-ion aging effects caused by thermal cycling superimposed onto electrochemical cycling, and related aging path dependence. This scenario is representative of what a subset of LIBs will experience in vehicle service, where upon the typical start of a HEV/PHEV, the batteries will be cool or cold, will gradually warm up to normal temperature and operate there for a time, then will cool down after the vehicle is turned off. Such thermal cycling will occur thousands of times during the projected life of a HEV/PHEV battery pack. In practice, some ThM will be had in keeping batteries from becoming too hot. However, there appears to be little focus on keeping batteries from residing at colder temperatures at the beginning of their duty cycle (an exception is the Chevy Volt that does provide warming of batteries under cold conditions). To quantify the effects of thermal cycling on LIBs, studies at INL used a representative commercial chemistry, Sanyo ‘Y’ cells of 18650 configuration, which contain a proprietary NMC/Mn-spinel cathode, graphitic anode, and a proprietary electrolyte. In this work DTC was combined with

Batteries

655

Thermal cycling matrix, setting other plug-in hybrid electric vehicle (PHEV) duty cycle conditions constant

Table 4 Test condition

Test type

Number of cells

Temperature range

Time from Tmin to Tmax

Time at Tmax

Time from Tmax to Tmin

Thermal cycle frequency

1 2 3 4 5 6 7 8 9 10

Isothermal Isothermal Isothermal Mild thermal cycling Mild thermal cycling Mild thermal cycling Severe thermal cycling Severe thermal cycling Severe thermal cycling Severe thermal cycling; open circuit voltage (VOC)c

2 2 2 3 3 3 3 3 3 3

01C 20 40 10–40 10–40 10–40 20 to 20 to 20 to 20 to

NA NA NA 30 mina 30 mina 15 mina 30 mina 30 mina 15 mina 30 mina

NA NA NA 1h 1h 1h 1h 1h 1h 1h

NA NA NA 2 hb 2 hb 1 hb 2 hb 2 hb 1 hb 2 hb

NA NA NA Twice daily Continuous Continuous Twice daily Continuous Continuous Continuous

40 40 40 40

a

Linear temperature ramping is assumed. Linear temperature ramping is assumed; conditions are under no electrochemical cycling (vehicle off). c A baseline condition where cells at VOC undergo thermal cycles but no PHEV duty cycles. Note: “Continuous” thermal cycling frequency refers to the scenario whereby a PHEV undergoes repeated commutes throughout a day that each have a thermal cycling period (per above), covering up to 16 h day 1 and within a TBD maximum permissible number of 1 h commutes. Temperature range refers to ambient T surrounding the cells (environmental chamber). Tmax is consistent throughout to avoid biasing the matrix with conditions of higher T that would cause accelerated aging. b

40 Test condition 1 Test condition 2 Test condition 3 Test condition 4 Test condition 5 Test condition 6 Test condition 7 Test condition 8 Test condition 9 Test condition 10

C1/1 capacity loss, Sanyo Y PHEV cells with and without DTC

35

Emergence of other fade mechanism(s)

Capacity loss (%)

30

25 Worsening thermal conditions and DTC

20

15

10 Data averaged over each test group

5

0 0

10

20

30

40

50

60

70

80

90

Weeks Fig. 16 Capacity loss (C/1 basis) for 18650 cells tested under daily thermal cycling (DTC) protocol with a plug-in hybrid electric vehicle (PHEV) duty cycle. The conditions are defined in Table 4. Some test conditions cause aging that is more than twice of the 0 and 201C isothermal baselines, confirming that DTC does accelerate cell aging.

electrochemical cycling to mimic cold-start conditions followed by an hour-long commute, which was repeated once or more each day. Table 4 gives the matrix of conditions investigated at the INL. Electrochemical cycling for this DTC study was based on PHEV-relevant cycle-life protocol consisting of charge depleting and charge sustaining modes that operated between 90 and 35% SOC. For this type of testing an effective thermal cycle will be somewhat dependent on global location, and testing parameters can be assigned to target a specific region of interest (e.g., New York City, Los Angeles, Vancouver, London, Berlin, etc.). Results for capacity loss in the test cells are shown in Fig. 16. Thermal cycling effects on capacity loss are more evident for cells actively undergoing duty cycles. Clearly seen is the outcome that some test

656

Batteries

conditions cause aging that is more than twice of the 0 and 201C isothermal baselines, confirming that DTC does accelerate cell aging. In contrast, cells under calendar-life conditions (no PHEV cycling) with thermal cycling experience slower aging (Condition 10). The results suggest that BEV batteries operated at colder climates will undergo additional degradation possibly due to particle fracturing from excessive stress and/or Li dendrite formation, and should have more sophisticated ThM. Also, the results clearly indicate the need to test past 1 year, as significant new aging mechanisms appear after 50 weeks in three of the cell test groups. One benefit of this study is to provide more realistic and accurate life predictions by accounting for the influence of thermal cycling effects and related path dependence on aging mechanisms.

2.20.6

Battery Energy Storage for Grid Applications

Within the energy policies of the United States and elsewhere there is an understanding that in order for electrical grid stability and RE capture to be effective, there needs to be viable means for intermittent energy storage. There are several options for such energy storage, including compressed air, flow batteries, flywheels, pumped hydro, hydrogen production and storage, various battery types (lead acid, Li-ion, as well as battery chemistries based on sodium and zinc), molten salt thermal energy storage, and other technologies. Each of these options has a fairly well defined expected useful economic life and levelized cost of storage (LCOS) [235]. Within battery energy storage for grid (BESG), Li-ion technology is receiving generous attention and investment for grid reliability and stabilization in terms of RE capture, load leveling, and avoidance of brown-outs and black-outs in critical applications. Municipal and regional electrical grid systems have existed well over a century without the infusion of battery storage as an integral component for grid operation and management. Thus grid applications of Li-ion systems have an element of “frontier” due to the lack of long-term testing of such batteries under grid scenarios. Various ownership models exist for BESG, where the battery installation can be owned by the customer, one or more thirdparties, or a utility. Relative scales of BESG installations generally increase from single or tens of kilowatt hours to tens of megawatt hours according to the following customer segments: residential, commercial and utility. According to the US Department of Energy’s Global Energy Storage Database (Office of Electricity Delivery and Energy Reliability), as of 2016 there were approximately 976 electrochemical-based energy storage projects worldwide for a net power coverage of 3 GW, and within the United States there were 472 projects serving 1.3 GW [236]. On the power scale, this represents a ten-fold growth from levels seen in the 2008–09 timeframe [236]. These numbers are expected to grow by at least another order of magnitude in some segments over the next decade, where the total revenue based on sale of battery cells is estimated between 30 and 40% of the total system revenue for energy storage [237]. A fascinating attribute for the growing BESG sector is the proliferation of grid applications and use cases that have been employed. These can be loosely defined as “in front of the meter” (power grid oriented) or “behind the meter” (distributed applications) cases [235]. Looking at domestic projects that have been contracted, under construction, that are operational, or offline for repair or maintenance, three primary use case categories are [236]: renewables capacity firming (605 MW), electric energy time shift (504 MW), and electric bill management (274 MW). Other use cases include RE time shift, frequency regulation, voltage support, onsite renewable generation shifting, onsite power, grid-connected commercial reliability and quality, ramping, load following, microgrid capability, and others. According to the DOE database, many projects have multiple use cases, which would support the assessment of interdependent factors that impact battery performance and lifetime [236]. A tandem consideration is the ownership of these projects, where third-party ownership dominates for larger BESG projects, followed by utility ownership. Customer-owned BESG constitutes a small fraction of ownership cases in the United States (less than 5% overall on a power basis) [236]. Based on economic factors, the scale of the application tends to drive the choice of energy storage technology. In most BESG cases, there is an economy of scale that benefits larger grid/microgrid and island grid applications, yet BESG becomes more expensive per megawatt hour per installation for smaller industrial, commercial, and residential grid applications [235]. Component costs also vary per unit energy with scale, for example, capital and operation & maintenance (O&M) costs for smaller LIB installations (e.g., residential) can be three to five time higher per megawatt hour than those of larger grid installations [235]. Critical enablers for BESG include increased volumetric energy storage (the need for fundamentally better battery materials to permit a smaller battery form factor), more sophisticated BMSs, and deployment of dedicated computational tools that will provide real-time diagnostics and prognostics of battery cells and their systems. Other related considerations are battery ThM that allows battery operation and survivability in harsher BESG environments (but at increased upfront and operational costs), and the reuse or repurposing of battery components or systems that have aged past their useful life under BESG. Collectively, these enablers will facilitate a more efficient BESG footprint in terms of MW per installation, improved management of batteries to avoid unsafe conditions, and prolonged battery life to optimize the battery investment. BMS cost is estimated at 16% of the total battery pack as opposed to a cell cost of 60%, using an electric drive vehicle basis [238]. Due to economy of scale, the cost of BMS units for BESG would fall beneath this percentage, yet during battery design there needs to be a determination on the maximum number of cells that can be covered by a single BMS. Design and function of BMSs will vary with application, providing opportunity for this niche to grow as battery applications grow. According to a recent report from Frost & Sullivan, multiple BMSs are needed for higher-capacity installations (such as hundreds of KWh storage), while residential storage capacity (less than 50 KWh) uses only one or two BMSs each [238]. Another driver for intelligent BMS design is the need to document battery use conditions that correspond to battery aging so that a clear history of stress factors can be obtained to support the emerging battery secondary

Batteries

657

use industry. Knowing how a battery has aged will promote informed decisions on battery reuse/repurposing versus materials recycling. A healthy secondary use community will help improve the economic cycle of batteries employed in grid applications. Batteries based on sodium-ion (Na-ion) chemistries hold promise to displace more expensive Li-ion counterparts, due to lower cost of some materials. Maturity in Na-ion systems will allow additional options for grid economic models tied to such batteries. However, much testing for reliability and longevity is required to establish Na-ion as a viable long-term alternative for grid usage. Also, the competitive stance for Na-ion systems is hampered by their lower cell voltages, compared to that of many Li-ion cells such as numerous examples that fall within the NMC-graphite group. Thus, Na-ion chemistries need to offer other benefits such as high energy density and fast cycling rates to gain market space. With that said, grid-ready Na-ion battery systems are not expected to be available for widespread use for at least another three to 5 years. In conclusion, Li-ion BESG systems are an important element of the future energy landscape. Less expensive cell materials and good longevity of battery systems are required for deeper investment and widespread use. Much research and development is needed before the levelized cost of Li-ion BESG is less than or equal to alternatives. Some alternatives (e.g., Na-ion batteries) could occupy market space based on reduced costs, but still require extensive validation testing. Tandem to this development path for BESG is the need for continued BMS development and the deployment of scientific computing tools that will touch upon the physics of battery operation and aging and provide real-time diagnostic and prognostic analyses. In concert, this will promote more intelligent management of BESG systems and enable a more seamless integration of BESG with the emerging battery secondary use industry.

2.20.7

Closing Remarks and Future Directions

Economic growth and population expansion are resulting in a rapid rise in energy consumption, intensifying the need for electrical energy storage (EES) technologies. LIB is ubiquitous in applications from portable electronics to electric vehicles and electric grids. Nevertheless, increasing demands require future LIBs to offer increased energy and power density, improved safety, larger temperature range, longer lifetime, and lower cost. The versatility of materials for LIBs makes it possible to vary composition, structure, morphology, architecture, and surface chemistry for enhanced performance to meet future needs. Improved cathode materials that offer higher capacity and better safety are the current foci in the research and development community. Cathode materials with higher Li content are needed to increase both the specific and volumetric energy densities of LIBs. Improving layered oxides and designing new polyanionic compounds will be the further trends for cathode research. On the other hand, electrochemical charge storage mechanism and failure mechanism need to be studied and understood for the high capacity cathode materials. On the anode side, further improvement in mechanical properties and volumetric densities will provide additional performance for the high capacity anode materials to replace the low density graphite electrode. For both anode and cathode materials, higher ionic and electronic conductivity are desired. Study on cation-disordered materials is worth more future research efforts to unlock the potential of these materials for enhanced electrochemical charge storage in LIBs. For electrolytes, study on additive effect, investigation of high concentration and water-in-salt electrolytes, solid-state electrolytes are all promising directions to further improve the electrochemical potential window, stability, and safety of electrolytes. Regarding the future direction of CCES, there is still much to be learned on the scientific level about the fundamental causes for CCES limits of LIBs, which appear to have roots in complex material science, electrochemical, and thermodynamic domains. At the center of improvement on the materials side is the need to reduce thermodynamic processes that compete in terms of detrimental electrolyte behavior (e.g., phase formation) near electrode interfacial regions. Understanding the complex temperature-dependent interplay of electrolyte components with morphology and electronic attributes of affected surfaces and SEI films is crucial for discerning one or more pathways toward mitigation of poor performance. This will doubtless involve greater research in materials selection for both the electrolyte and the electrodes, with special attention given to the resultant SEI films. Early indication of success will be cell chemistries that not only exhibit reduced interfacial impedances at low temperature, but also provide satisfactory battery life over the entire temperature range of the intended application. We recognize that improvements made in low-temperature performance will carry with them potential benefits at higher temperatures, such as less cell polarization, quicker charge transfer kinetics, and improved overall electrochemical efficiency. This calls for balance in improving performance at cold conditions and monitoring battery resilience at upper temperatures (between 40 and 601C), so that one benefit does not obviate another. CCES still contains unfenced frontier R&D where ongoing investigation is needed for existing and future generations of batteries.

Acknowledgments Authors from the INL wish to gratefully acknowledge support from the United States Department of Energy Vehicle Technologies Office and the Laboratory Directed Research and Development Program at the INL. All work at INL was performed under contract DE-AC07-05ID14517. H. Xiong acknowledges assistance from C. Deng for figure preparation.

658

Batteries

References [1] Lewis N. Powering the planet: the future of energy in the world. In: 247th ACS National Meeting & Exposition March 16-20, 2014, Dallas, TX, Abstracts of Papers of the American Chemical Society; 2014. [2] International Energy Agency (IEA). Key World Energy Statistics. Available From: http://www.iea.org/publications/freepublications/publication/key_world_energy_stats-1.pdf; 2011. [3] Larcher D, Tarascon JM. Towards greener and more sustainable batteries for electrical energy storage. Nat Chem 2015;7:19–29. [4] Thackeray MM, Wolverton C, Isaacs ED. Electrical energy storage for transportation-approaching the limits of, and going beyond, lithium-ion batteries. Energy Environ Sci 2012;5:7854–63. [5] Dunn B, Kamath H, Tarascon JM. Electrical energy storage for the grid: a battery of choices. Science 2011;334:928–35. [6] Jeong G, Kim YU, Kim H, et al. Prospective materials and applications for Li secondary batteries. Energy Environ Sci 2011;4:1986–2002. [7] Yang ZG, Liu J, Baskaran S, et al. Enabling renewable energy-and the future grid-with advanced electricity storage. JOM-US 2010;62:14–23. [8] Liu J, Zhang JG, Yang ZG, et al. Materials science and materials chemistry for large scale electrochemical energy storage: from transportation to electrical grid. Adv Funct Mater 2013;23:929–46. [9] Soloveichik GL. Flow batteries: current status and trends. Chem Rev 2015;115:11533–58. [10] Beaudin M, Zareipour H, Schellenberglabe A, Rosehart W. Energy storage for mitigating the variability of renewable electricity sources: an updated review. Energy Sustain Dev 2010;14:302–14. [11] Nagaura T, Tozawa K. Lithium ion rechargeable battery. Prog Batteries Solar Cells 1990;9:209. [12] Choi NS, Chen ZH, Freunberger SA, et al. Challenges facing lithium batteries and electrical double-layer capacitors. Angew Chem Int Ed 2012;51:9994–10024. [13] Tarascon JM. Key challenges in future Li-battery research. Philos Trans R Soc A 2010;368:3227–41. [14] Tarascon JM, Armand M. Issues and challenges facing rechargeable lithium batteries. Nature 2001;414:359–67. [15] Armand M, Tarascon JM. Building better batteries. Nature 2008;451:652–7. [16] Wadia C, Albertus P, Srinivasan V. Resource constraints on the battery energy storage potential for grid and transportation applications. J Power Sources 2011;196:1593–8. [17] Goodenough JB, Kim Y. Challenges for rechargeable batteries. J Power Sources 2011;196:6688–94. [18] Goodenough JB, Kim Y. Challenges for rechargeable Li batteries. Chem Mater 2010;22:587–603. [19] Whittingham MS. Lithium batteries and cathode materials. Chem Rev 2004;104:4271–301. [20] Choi JW, Aurbach D. Promise and reality of post-lithium-ion batteries with high energy densities. Nat Rev Mater 2016;1:16013. [21] Reuters. Boeing 787 suffers a battery fire in Boston, vol. 162. New York, NY: The New York Times; 2013. p. B5. [22] Vlasic B. Car fire a test for high-flying Tesla. New York, NY: The New York Times; 2013. p. B1. [23] Chen BX, Sang-Hun C. Why Samsung abandoned its galaxy note 7 flagship phone. New York, NY: The New York Times; 2016. p. A1. [24] Goodenough JB, Park KS. The Li-ion rechargeable battery: a perspective. J Am Chem Soc 2013;135:1167–76. [25] USCAR. Available From: http://www.uscar.org/guest/article_view.php?articles_id=85; 2016 [accessed 16.10.16]. [26] Energy. The EV everywhere grand challenge: road to success. Available From: http://energy.gov/eere/vehicles/downloads/ev-everywhere-grand-challenge-road-success; 2014. [27] Gyuk I, Johnson M, Vetrano J, et al. DOE OE grid energy storage report. Available From: http://www.sandia.gov/ess/docs/other/Grid_Energy_Storage_Dec_2013.pdf; 2013. [28] Utility Dive. Available From: http://www.utilitydive.com/news/inside-construction-of-the-worlds-largest-lithium-ion-battery-storage-faci/431765/; 2016 [accessed 12.12.16]. [29] DOE Office of Electricity Delivery & Energy Reliability. DOE Global Energy Storage Database, Sandia National Laboratories. Available From: http://www. energystorageexchange.org; 2016. [30] Dinger A, Martin R, Mosquet X, et al. Batteries for electric cars: chanllenges: opportunities, and the outlook to 2020. Available From: http://www.bcg.com/documents/ file36615.pdf; 2010. [31] Whittingham MS. Electrical energy-storage and intercalation chemistry. Science 1976;192:1126–7. [32] Energy Factor. Pioneers of innovation: the battery that changed the world. Available From: https://energyfactor.exxonmobil.com/scienace-technology/battery-changed-world/. [33] Whittingham MS, Jacobson AJ, editors. Intercalation chemistry London: Academic Press; 1982. [34] Mizushima K, Jones PC, Wiseman PJ, Goodenough JB. LixCoO2 (0oxr1) a new cathode material for batteries of high-energy density. Mater Res Bull 1980;15:783–9. [35] Goodenough JB, Mizushima K, Takeda T. Solid-solution oxides for storage-battery electrodes. Jpn J Appl Phys 1980;19:305–13. [36] Thackeray MM, David WIF, Bruce PG, Goodenough JB. Lithium Insertion into manganese spinels. Mater Res Bull 1983;18:461–72. [37] Faulkner LR, Bard AJ. Electrochemical methods: fundamentals and applications. 2nd ed. New York, NY: Wiley; 2001. [38] Allen PL, Hickling A. Electrochemistry of sulphur.1. Overpotential in the discharge of the sulphide ion. Trans Faraday Soc 1957;53:1626–35. [39] Abraham DP, Roth EP, Kostecki R, et al. Diagnostic examination of thermally abused high-power lithium-ion cells. J Power Sources 2006;161:648–57. [40] Richard MN, Dahn JR. Accelerating rate calorimetry study on the thermal stability of lithium intercalated graphite in electrolyte II. Modeling the results and predicting differential scanning calorimeter curves. J Electrochem Soc 1999;146:2078–84. [41] Richard MN, Dahn JR. Accelerating rate calorimetry study on the thermal stability of lithium intercalated graphite in electrolyte I. Experimental. J Electrochem Soc 1999;146:2068–77. [42] Liu CF, Neale ZG, Cao GZ. Understanding electrochemical potentials of cathode materials in rechargeable batteries. Mater Today 2016;19:109–23. [43] Melot BC, Tarascon JM. Design and preparation of materials for advanced electrochemical storage. Acc Chem Res 2013;46:1226–38. [44] Rozier P, Tarascon JM. Review-Li-rich layered oxide cathodes for next-generation Li-ion batteries: chances and challenges. J Electrochem Soc 2015;162:A2490–9. [45] Liu W, Oh P, Liu X, et al. Nickel-rich layered lithium transition-metal oxide for high-energy lithium-ion batteries. Angew Chem Int Ed 2015;54:4440–57. [46] Rousse G, Tarascon JM. Sulfate-based polyanionic compounds for Li-ion batteries: synthesis, crystal chemistry, and electrochemistry aspects. Chem Mater 2014;26:394–406. [47] Wang Y, Cao GZ. Developments in nanostructured cathode materials for high-performance lithium-ion batteries. Adv Mater 2008;20:2251–69. [48] Yoo HD, Markevich E, Salitra G, et al. On the challenge of developing advanced technologies for electrochemical energy storage and conversion. Mater Today 2014;17:110–21. [49] Delmas C, Fouassier C, Hagenmuller P. Structural classification and properties of the layered oxides. Physica B þ C 1980;99:81–5. [50] Cho J, Kim YJ, Park B. Novel LiCoO2 cathode material with Al2O3 coating for a Li ion cell. Chem Mater 2000;12:3788–91. [51] Li C, Zhang HP, Fu LJ, et al. Cathode materials modified by surface coating for lithium ion batteries. Electrochim Acta 2006;51:3872–83. [52] Jung YS, Cavanagh AS, Dillon AC, et al. Enhanced stability of LiCoO2 cathodes in lithium-ion batteries using surface modification by atomic layer deposition. J Electrochem Soc 2010;157:A75–81. [53] Cho J, Lee JG, Kim B, Park B. Effect of P2O5 and AlPO4 coating on LiCoO2 cathode material. Chem Mater 2003;15:3190–3. [54] Dahn JR, Vonsacken U, Michal CA. Structure and electrochemistry of Li17YNiO2 and a new Li2NiO2 phase with the Ni(OH)2 structure. Solid State Ionics 1990;44:87–97.

Batteries

659

[55] Morales J, Perezvicente C, Tirado JL. Cation distribution and chemical deintercalation of Li1 XNi1 þ XO2. Mater Res Bull 1990;25:623–30. [56] Manthiram Arumugam, Muraliganth Theivanayagam. Lithium intercalation cathode materials for lithium-ion batteries. In: Daniel C, Besenhard JO, editors. Handbook of battery materials. Weinheim: Wiley-VCH Verlag GmbH & Co. KGaA; 2011. p. 341–75. [57] Ohzuku T, Ueda A, Nagayama M. Electrochemistry and Structural Chemistry of LiNiO2 (R(3)over-bar-M) for 4 V secondary lithium cells. J Electrochem Soc 1993;140:1862–70. [58] Li W, Reimers JN, Dahn JR. In-Situ X-ray-diffraction and electrochemical studies of Li1 XNiO2. Solid State Ionics 1993;67:123–30. [59] Mishra SK, Ceder G. Structural stability of lithium manganese oxides. Phys Rev B 1999;59:6120–30. [60] Reimers JN, Fuller EW, Rossen E, Dahn JR. Synthesis and electrochemical studies of LiMnO2 prepared of low-temperatures. J Electrochem Soc 1993;140:3396–401. [61] Armstrong AR, Bruce PG. Synthesis of layered LiMnO2 as an electrode for rechargeable lithium batteries. Nature 1996;381:499–500. [62] Capitaine F, Gravereau P, Delmas C. A new variety of LiMnO2 with a layered structure. Solid State Ionics 1996;89:197–202. [63] Thackeray MM. A comment on the structure of thin-film LiMn2O4 electrodes. J Electrochem Soc 1997;144:L100–2. [64] Ueda A, Ohzuku T. Solid-state redox reactions of LiNi1/2Co1/2O2 (R(3)-over-bar-M) for 4-volt secondary lithium cells. J Electrochem Soc 1994;141:2010–4. [65] Ohzuku T, Ueda A, Kouguchi M. Synthesis and characterization of LiAl1/4Ni3/4O2 (R(3)over-bar-m) for lithium-ion (shuttlecock) batteries. J Electrochem Soc 1995;142:4033–9. [66] Park SH, Park KS, Sun YK, et al. Structural and electrochemical characterization of lithium excess and Al-doped nickel oxides synthesized by the sol-gel method. Electrochim Acta 2001;46:1215–22. [67] Robert R, Villevieille C, Novak P. Enhancement of the high potential specific charge in layered electrode materials for lithium-ion batteries. J Mater Chem A 2014;2:8589–98. [68] Noh HJ, Youn S, Yoon CS, Sun YK. Comparison of the structural and electrochemical properties of layered Li[NixCoyMnz]O2 (x ¼ 1/3, 0.5, 0.6, 0.7, 0.8 and 0.85) cathode material for lithium-ion batteries. J Power Sources 2013;233:121–30. [69] Lee KS, Myung ST, Amine K, et al. Structural and electrochemical properties of layered Li[Ni1 2xCoxMnx]O2 (x ¼ 0.1–0.3) positive electrode materials for Li-ion batteries. J Electrochem Soc 2007;154:A971–7. [70] Sun YK, Cho SW, Lee SW, et al. AlF3-coating to improve high voltage cycling performance of Li[Ni1/3Co1/3Mn1/3]O2 cathode materials for lithium secondary batteries. J Electrochem Soc 2007;154:A168–72. [71] Sun YK, Myung ST, Kim MH, et al. Synthesis and characterization of Li[(Ni0.8Co0.1Mn0.1)0.8(Ni0.5Mn0.5)0.2]O2 with the microscale core–shell structure as the positive electrode material for lithium batteries. J Am Chem Soc 2005;127:13411–8. [72] Sun YK, Chen ZH, Noh HJ, et al. Nanostructured high-energy cathode materials for advanced lithium batteries. Nat Mater 2012;11:942–7. [73] Rossouw MH, Liles DC, Thackeray MM. Synthesis and structural characterization of a novel layered lithium manganese oxide, Li0.36Mn0.91O2, and its lithiated derivative, Li1.09Mn0.91O2. J Solid State Chem 1993;104:464–6. [74] Rossen E, Jones CDW, Dahn JR. Structure and electrochemistry of LixMnyNi1 YO2. Solid State Ionics 1992;57:311–8. [75] Johnson CS, Li NC, Lefief C, et al. Synthesis, characterization and electrochemistry of lithium battery electrodes: xLi(2)MnO(3)center dot(1 x) LiMn0.333Ni0.333Co0.333O2 (0 o¼ xo¼ 0.7). Chem Mater 2008;20:6095–106. [76] Lin F, Markus IM, Nordlund D, et al. Surface reconstruction and chemical evolution of stoichiometric layered cathode materials for lithium-ion batteries. Nat Commun 2014;5:3529. [77] Qiu B, Zhang MH, Wu LJ, et al. Gas-solid interfacial modification of oxygen activity in layered oxide cathodes for lithium-ion batteries. Nat Commun 2016;7. doi:10.1038/ncomms12108. [78] Aydinol MK, Ceder G. First-principles prediction of insertion potentials in Li–Mn oxides for secondary Li batteries. J Electrochem Soc 1997;144:3832–5. [79] Jang DH, Shin YJ, Oh SM. Dissolution of spinel oxides and capacity losses in 4V Li/LixMn2O4 coils. J Electrochem Soc 1996;143:2204–11. [80] Sun YK, Hong KJ, Prakash J. The effect of ZnO coating on electrochemical cycling behavior of spinel LiMn2O4 cathode materials at elevated temperature. J Electrochem Soc 2003;150:A970–2. [81] Padhi AK, Nanjundaswamy KS, Goodenough JB. Phospho-olivines as positive-electrode materials for rechargeable lithium batteries. J Electrochem Soc 1997;144:1188–94. [82] Padhi AK, Nanjundaswamy KS, Masquelier C, Goodenough JB. Mapping of transition metal redox energies in phosphates with NASICON structure by lithium intercalation. J Electrochem Soc 1997;144:2581–6. [83] Padhi AK, Nanjundaswamy KS, Masquelier C, et al. Effect of structure on the Fe3 þ /Fe2 þ redox couple in iron phosphates. J Electrochem Soc 1997;144:1609–13. [84] Jiang J, Dahn JR. ARC studies of the thermal stability of three different cathode materials: LiCoO2; Li[Ni0.1Co0.8Mn0.1]O2; and LiFePO4, in LiPF6 and LiBoB EC/DEC electrolytes. Electrochem Commun 2004;6:39–43. [85] Winter M, Besenhard JO, Spahr ME, Novak P. Insertion electrode materials for rechargeable lithium batteries. Adv Mater 1998;10:725–63. [86] Chan CK, Peng H, Liu G, et al. High-performance lithium battery anodes using silicon nanowires. Nat Nanotechnol 2008;3:31–5. [87] Kim MG, Sim S, Cho J. Novel core–shell Sn-Cu anodes for lithium rechargeable batteries prepared by a redox-transmetalation reaction. Adv Mater 2010;22:5154. [88] Bruce PG, Scrosati B, Tarascon JM. Nanomaterials for rechargeable lithium batteries. Angew Chem Int Ed 2008;47:2930–46. [89] Arico AS, Bruce P, Scrosati B, et al. Nanostructured materials for advanced energy conversion and storage devices. Nat Mater 2005;4:366–77. [90] Yang ZG, Choi D, Kerisit S, et al. Nanostructures and lithium electrochemical reactivity of lithium titanites and titanium oxides: a review. J Power Sources 2009;192:588–98. [91] Wagemaker M, Borghols WJH, Mulder FM. Large impact of particle size on insertion reactions. A case for anatase LixTiO2. J Am Chem Soc 2007;129:4323–7. [92] Kavan L, Kalbac M, Zukalova M, et al. Lithium storage in nanostructured TiO2 made by hydrothermal growth. Chem Mater 2004;16:477–85. [93] Maier J. Size effects on mass transport and storage in lithium batteries. J Power Sources 2007;174:569–74. [94] Manthiram A, Murugan AV, Sarkar A, Muraliganth T. Nanostructured electrode materials for electrochemical energy storage and conversion. Energ Environ Sci 2008;1:621–38. [95] Chen JS, Tan YL, Li CM, et al. Constructing hierarchical spheres from large ultrathin anatase TiO2 nanosheets with nearly 100% exposed (001) facets for fast reversible lithium storage. J Am Chem Soc 2010;132:6124–30. [96] Sun Y, Liu N, Cui Y. Promises and challenges of nanomaterials for lithium-based rechargeable batteries. Nat Energy 2016;1:16071. [97] Huggins RA. Lithium alloy negative electrodes. J Power Sources 1999;81:13–9. [98] Liu G, Xun SD, Vukmirovic N, et al. Polymers with tailored electronic structure for high capacity lithium battery electrodes. Adv Mater 2011;23:4679. [99] Liu N, Wu H, McDowell MT, et al. A yolk–shell design for stabilized and scalable Li-ion battery alloy anodes. Nano Lett 2012;12:3315–21. [100] Liu N, Lu ZD, Zhao J, et al. A pomegranate-inspired nanoscale design for large-volume-change lithium battery anodes. Nat Nanotechnol 2014;9:187–92. [101] Kravchyk K, Protesescu L, Bodnarchuk MI, et al. Monodisperse and inorganically capped Sn and Sn/SnO2 nanocrystals for high-performance Li-ion battery anodes. J Am Chem Soc 2013;135:4199–202. [102] Xu LP, Kim C, Shukla AK, et al. Monodisperse Sn nanocrystals as a platform for the study of mechanical damage during electrochemical reactions with Li. Nano Lett 2013;13:1800–5. [103] Qin J, He C, Zhao N, et al. Graphene networks anchored with Sn@graphene as lithium ion battery anode. ACS Nano 2014;8:1728–38. [104] Xu YH, Liu Q, Zhu YJ, et al. Uniform nano-Sn/C composite anodes for lithium ion batteries. Nano Lett 2013;13:470–4. [105] Xu YH, Guo JC, Wang CS. Sponge-like porous carbon/tin composite anode materials for lithium ion batteries. J Mater Chem 2012;22:9562–7.

660

Batteries

[106] Yu Y, Gu L, Zhu CB, et al. Tin nanoparticles encapsulated in porous multichannel carbon microtubes: preparation by single-nozzle electrospinning and application as anode material for high-performance Li-based batteries. J Am Chem Soc 2009;131:15984. [107] Elia GA, Panero S, Savoini A, et al. Mechanically milled, nanostructured Sn–C composite anode for lithium ion battery. Electrochim Acta 2013;90:690–4. [108] Hassoun J, Derrien G, Panero S, Scrosati B. A nanostructured Sn–C composite lithium battery electrode with unique stability and high electrochemical performance. Adv Mater 2008;20:3169–75. [109] Derrien G, Hassoun J, Panero S, Scrosati B. Nanostructured Sn–C composite as an advanced anode material in high-performance lithium-ion batteries. Adv Mate 2007;19:2336. [110] Zou YQ, Wang Y. Sn@CNT nanostructures rooted in graphene with high and fast Li-storage capacities. ACS Nano 2011;5:8108–14. [111] Zheng JW, Nai SML, Ng MF, et al. DFT study on nano structures of Sn/CNT complex for potential Li-ion battery application. J Phys Chem C 2009;113:14015–9. [112] Wang Y, Wu M, Jiao Z, Lee JY. Sn@CNT and Sn@C@CNT nanostructures for superior reversible lithium ion storage. Chem Mater 2009;21:3210–5. [113] Beck FR, Epur R, Hong DH, et al. Microwave derived facile approach to Sn/graphene composite anodes for, lithium-ion batteries. Electrochim Acta 2014;127:299–306. [114] Li N, Song HW, Cui H, Wang CX. Sn@graphene grown on vertically aligned graphene for high-capacity, high-rate, and long-life lithium storage. Nano Energy 2014;3:102–12. [115] Bae EG, Hwang YH, Pyo M. Binder-free Sn/graphene nanocomposites prepared by electrophoretic deposition for anode materials in lithium ion batteries. Bull Korean Chem Soc 2013;34:1199–204. [116] Wang CD, Li Y, Chui YS, et al. Three-dimensional Sn-graphene anode for high-performance lithium-ion batteries. Nanoscale 2013;5:10599–604. [117] Wang GX, Wang B, Wang XL, et al. Sn/graphene nanocomposite with 3D architecture for enhanced reversible lithium storage in lithium ion batteries. J Mater Chem 2009;19:8378–84. [118] Qin J, He CN, Zhao NQ, et al. Graphene networks anchored with Sn@graphene as lithium ion battery anode. ACS Nano 2014;8:1728–38. [119] Deng D, Kim MG, Lee JY, Cho J. Green energy storage materials: nanostructured TiO2 and Sn-based anodes for lithium-ion batteries. Energy Environ Sci 2009;2:818–37. [120] Zhu GN, Wang YG, Xia YY. Ti-based compounds as anode materials for Li-ion batteries. Energy Environ Sci 2012;5:6652–67. [121] Zachauchristiansen B, West K, Jacobsen T, Atlung S. Lithium insertion in different Tio2 modifications. Solid State Ionics 1988;28:1176–82. [122] Takai S, Kamata M, Fujine S, et al. Diffusion coefficient measurement of lithium ion in sintered Li1.33Ti1.67O4 by means of neutron radiography. Solid State Ionics 1999;123:165–72. [123] Ohzuku T, Takehara Z, Yoshizawa S. Non-aqueous lithium-titanium dioxide cell. Electrochim Acta 1979;24:219–22. [124] Macklin WJ, Neat RJ. Performance of titanium dioxide-based cathodes in a lithium polymer electrolyte cell. Solid State Ionics 1992;53:694–700. [125] Koudriachova MV, Harrison NM, de Leeuw SW. Density-functional simulations of lithium intercalation in rutile. Phys Rev B 2002;65:235423. [126] Koudriachova MV, Harrison NM, de Leeuw SW. Effect of diffusion on lithium intercalation in titanium dioxide. Phys Rev Lett 2001;86:1275–8. [127] Johnson OW. 1-Dimensional diffusion of Li in rutile. Phys Rev A-Gen Phys 1964;136:A284. [128] Wang DH, Choi DW, Yang ZG, et al. Synthesis and Li-ion insertion properties of highly crystalline mesoporous rutile TiO2. Chem Mater 2008;20:3435–42. [129] Wagemaker M, Mulder FM, A. Van der Ven . The role of surface and interface energy on phase stability of nanosized insertion compounds. Adv Mater 2009;21:2703. [130] Jamnik J, Maier J. Nanocrystallinity effects in lithium battery materials aspects of nano-ionics. Part IV. Phys Chem Chem Phys 2003;5:5215–20. [131] Milne NA, Skyllas-Kazacos M, Luca V. Crystallite size dependence of lithium intercalation in nanocrystalline rutile. J Phys Chem C 2009;113:12983–95. [132] Hu YS, Kienle L, Guo YG, Maier J. High lithium electroactivity of nanometer-sized rutile TiO2. Adv Mater 2006;18:1421. [133] Koudriachova MV, Harrison NM, de Leeuw SW. First principles predictions for intercalation behaviour. Solid State Ionics 2004;175:829–34. [134] Gligor F, de Leeuw SW. Lithium diffusion in rutile structured titania. Solid State Ionics 2006;177:2741–6. [135] Kavan L, Fattakhova D, Krtil P. Lithium insertion into mesoscopic and single-crystal TiO2 (rutile) electrodes. J Electrochem Soc 1999;146:1375–9. [136] Koudriachova MV, de Leeuw SW, Harrison NM. A new phase of lithiated titania predicted from first principles. Chem Phys Lett 2003;371:150–6. [137] Koudriachova MV, Harrison NM, de Leeuw SW. Open circuit voltage profile for Li-intercalation in rutile and anatase from first principles. Solid State Ionics 2002;152:189–94. [138] Baudrin E, Cassaignon S, Koesch M, et al. Structural evolution during the reaction of Li with nano-sized rutile type TiO2 at room temperature. Electrochem Commun 2007;9:337–42. [139] Stashans A, Lunell S, Bergstrom R, et al. Theoretical study of lithium intercalation in rutile and anatase. Phys Rev B 1996;53:159–70. [140] Koudriachova MV. Mechanism of lithium intercalation in titanates. J Solid State Electr 2010;14:549–53. [141] Koudriachova MV. Effect of particle size on the phase behavior of Li-intercalated TiO2-rutile. J Power Sources 2011;196:6898–901. [142] Koudriachova MV, Harrison NM, de Leeuw SW. Structural deformations in lithium doped titanium dioxide. Comp Mater Sci 2002;24:235–40. [143] Reddy MA, Kishore MS, Pralong V, et al. Room temperature synthesis and Li insertion into nanocrystalline rutile TiO2. Electrochem Commun 2006;8:1299–303. [144] Borghols WJH, Wagemaker M, Lafont U, et al. Impact of nanosizing on lithiated rutile TiO2. Chem Mater 2008;20:2949–55. [145] Vijayakumar M, Kerisit S, Wang C, et al. Effect of chemical lithium insertion into rutile TiO2 nanorods. J Phys Chem C 2009;113:14567–74. [146] Robertson J. Diamond-like amorphous carbon. Mater Sci Eng R 2002;37:129–281. [147] Carlson DE, Wronski CR. Amorphous silicon solar-cell. Appl Phys Lett 1976;28:671–3. [148] Hasegawa R. Advances in amorphous and nanocrystalline materials. J Magn Magn Mater 2012;324:3555–7. [149] Paul H, Sylvio I. Diffusion and ionic conduction in nanocrystalline ceramics. J Phys Condens Matter 2003;15:R1257. [150] Heitjans P, Tobschall E, Wilkening M. Ion transport and diffusion in nanocrystalline and glassy ceramics. Eur Phys J – Spec Top 2008;161:97–108. [151] Kang B, Ceder G. Battery materials for ultrafast charging and discharging. Nature 2009;458:190–3. [152] Huang JJ, Luo J. A facile and generic method to improve cathode materials for lithium-ion batteries via utilizing nanoscale surface amorphous films of self-regulating thickness. Phys Chem Chem Phys 2014;16:7786–98. [153] Sun K, Dillon SJ. A mechanism for the improved rate capability of cathodes by lithium phosphate surficial films. Electrochem Commun 2011;13:200–2. [154] Xiong H, Yildirim H, Shevchenko EV, et al. Self-improving anode for lithium-ion batteries based on amorphous to cubic phase transition in TiO2 nanotubes. J Phys Chem C 2012;116:3181–7. [155] Xiong H, Yildirim H, Podsiadlo P, et al. Compositional tuning of structural stability of lithiated cubic titania via a vacancy-filling mechanism under high pressure. Phys Rev Lett 2013;110:078304. [156] Lee J, Urban A, Li X, et al. Unlocking the potential of cation-disordered oxides for rechargeable lithium batteries. Science 2014;343:519–22. [157] Obrovac MN, Mao O, Dahn JR. Structure and electrochemistry of LiMO2, (M ¼ Ti, Mn, Fe, Co, Ni) prepared by mechanochemical synthesis. Solid State Ionics 1998;112:9–19. [158] Sakurai Y, Arai H, Yamaki J. Preparation of electrochemically active alpha-LiFeO2 at low temperature. Solid State Ionics 1998;113:29–34. [159] Kang K, Carlier D, John J, et al. Synthesis and electrochemical properties of layered Li0.9Ni0.45Ti0.55O2. Chem Mater 2003;15:4503–7. [160] Ozawa K, Nakao Y, Wang L, et al. Structural modifications caused by electrochemical lithium extraction for two types of layered LiVO2 (R-3m). J Power Sources 2007;174:469–72. [161] Hahn BP, Long JW, Rolison DR. Something from nothing: enhancing electrochemical charge storage with cation vacancies. Acc Chem Res 2013;46:1181–91. [162] Turner S, Buseck PR. Defects in nsutite (Gamma-MnO2) and dry-cell battery efficiency. Nature 1983;304:143–6. [163] Xu K. Nonaqueous liquid electrolytes for lithium-based rechargeable batteries. Chem Rev 2004;104:4303–417.

Batteries

661

[164] Xu K. Electrolytes and interphases in Li-ion batteries and beyond. Chem Rev 2014;114:11503–618. [165] United States Advanced Battery Consortium. Development of advanced high performance electrolytes for lithium ion batteries used in electric vehicle applications; 2015. [166] Peled E. The electrochemical-behavior of alkali and alkaline-earth metals in non-aqueous battery systems – the solid electrolyte interphase model. J Electrochem Soc 1979;126:2047–51. [167] Peled E. Film forming reaction at the lithium electrolyte interface. J Power Sources 1983;9:253–66. [168] Aurbach D, Weissman I, Schechter A, Cohen H. X-ray photoelectron spectroscopy studies of lithium surfaces prepared in several important electrolyte solutions. A comparison with previous studies by Fourier transform infrared spectroscopy. Langmuir 1996;12:3991–4007. [169] Aurbach D. Review of selected electrode–solution interactions which determine the performance of Li and Li ion batteries. J Power Sources 2000;89:206–18. [170] Tsubouchi S, Domi Y, Doi T, et al. Spectroscopic characterization of surface films formed on edge plane graphite in ethylene carbonate-based electrolytes containing film-forming additives. J Electrochem Soc 2012;159:A1786–90. [171] Shkrob IA, Zhu Y, Marin TW, Abraham D. Reduction of carbonate electrolytes and the formation of solid-electrolyte interface (SEI) in lithium-ion batteries. 1. Spectroscopic observations of radical intermediates generated in one-electron reduction of carbonates. J Phys Chem C 2013;117:19255–69. [172] Aurbach D, Moshkovich M, Cohen Y, Schechter A. The study of surface film formation on noble-metal electrodes in alkyl carbonates/Li salt solutions, using simultaneous in situ AFM, EQCM, FTIR, and EIS. Langmuir 1999;15:2947–60. [173] Dufek EJ. Evaluation of the SEI using a multilayer spectroscopic ellipsometry model. ECS Electrochem Lett 2014;3:A108–11. [174] Sazhin SV, Gering KL, Harrup MK, Rollins HW. Highly quantitative electrochemical characterization of non-aqueous electrolytes and solid electrolyte interphases. J Electrochem Soc 2014;161:A393–402. [175] Edstrom K, Herstedt M, Abraham DP. A new look at the solid electrolyte interphase on graphite anodes in Li-ion batteries. J Power Sources 2006;153:380–4. [176] Verma P, Maire P, Novak P. A review of the features and analyses of the solid electrolyte interphase in Li-ion batteries. Electrochim Acta 2010;55:6332–41. [177] Niehoff P, Passerini S, Winter M. Interface investigations of a commercial lithium ion battery graphite anode material by sputter depth profile X-ray photoelectron spectroscopy. Langmuir 2013;29:5806–16. [178] Lu P, Harris SJ. Lithium transport within the solid electrolyte interphase. Electrochem Commun 2011;13:1035–7. [179] Schechter A, Aurbach D, Cohen H. X-ray photoelectron spectroscopy study of surface films formed on Li electrodes freshly prepared in alkyl carbonate solutions. Langmuir 1999;15:3334–42. [180] Kwon K, Kong FP, McLarnon F, Evans JW. Characterization of the SEI on a carbon film electrode by combined EQCM and spectroscopic ellipsometry. J Electrochem Soc 2003;150:A229–33. [181] Smith AJ, Burns JC, Trussler S, Dahn JR. Precision measurements of the coulombic efficiency of lithium-ion batteries and of electrode materials for lithium-ion Batteries. J Electrochem Soc 2010;157:A196–202. [182] Ravdel B, Abraham KM, Gitzendanner R, et al. Thermal stability of lithium-ion battery electrolytes. J Power Sources 2003;119:805–10. [183] Arbizzani C, Gabrielli G, Mastragostino M. Thermal stability and flammability of electrolytes for lithium-ion batteries. J Power Sources 2011;196:4801–5. [184] Yang H, Zhuang GV, Ross PN. Thermal stability of LiPF6 salt and Li-ion battery electrolytes containing LiPF6. J Power Sources 2006;161:573–9. [185] Plakhotnyk AV, Ernst L, Schmutzler R. Hydrolysis in the system LiPF6-propylene carbonate-dimethyl carbonate-H2O. J Fluorine Chem 2005;126:27–31. [186] Gallus DR, Schmitz R, Wagner R, et al. The influence of different conducting salts on the metal dissolution and capacity fading of NCM cathode material. Electrochim Acta 2014;134:393–8. [187] Markevich E, Sharabi R, Gottlieb H, et al. Reasons for capacity fading of LiCoPO4 cathodes in LiPF6 containing electrolyte solutions. Electrochem Commun 2012;15:22–5. [188] Sharabi R, Markevich E, Borgel V, et al. Raman study of structural stability of LiCoPO4 cathodes in LiPF6 containing electrolytes. J Power Sources 2012;203:109–14. [189] Delacourt C, Kwong A, Liu X, et al. Effect of manganese contamination on the solid-electrolyte-interphase properties in Li-ion batteries. J Electrochem Soc 2013;160: A1099–107. [190] Abraham DP, Spila T, Furczon MM, Sammann E. Evidence of transition-metal accumulation on aged graphite anodes by SIMS. Electrochem Solid State 2008;11:A226–8. [191] von Cresce A, Xu K. Electrolyte additive in support of 5 V Li ion chemistry. J Electrochem Soc 2011;158:A337–42. [192] Dunn RP, Kafle J, Krause FC, et al. Electrochemical analysis of Li-ion cells containing triphenyl phosphate. J Electrochem Soc 2012;159:A2100–8. [193] Milien MS, Tottempudi U, Son M, et al. Development of lithium dimethyl phosphate as an electrolyte additive for lithium ion batteries. J Electrochem Soc 2016;163: A1369–72. [194] Zeng ZQ, Jiang XY, Wu BB, et al. Bis(2,2,2-trifluoroethyl) methylphosphonate: an novel flame-retardant additive for safe lithium-ion battery. Electrochim Acta 2014;129:300–4. [195] Feng JK, Ma P, Yang HX, Lu L. Understanding the interactions of phosphonate-based flame-retarding additives with graphitic anode for lithium ion batteries. Electrochim Acta 2013;114:688–92. [196] Zeng ZQ, Wu BB, Xiao LF, et al. Safer lithium ion batteries based on nonflammable electrolyte. J Power Sources 2015;279:6–12. [197] Seyoung Ahn, Hyun-Soo Kim, Sunhye Yang, et al. Thermal stability and performance studies of LiCo1/3Ni1/3Mn1/3O2 with phosphazene additives for Li-ion batteries. J Electroceram 2009;23:289–94. [198] Sazhin SV, Harrup MK, Gering KL. Characterization of low-flammability electrolytes for lithium-ion batteries. J Power Sources 2011;196:3433–8. [199] Rollins HW, Harrup MK, Dufek EJ, et al. Fluorinated phosphazene Co-solvents for improved thermal and safety performance in lithium-ion battery electrolytes. J Power Sources 2014;263:66–74. [200] Tsujikawa T, Yabuta K, Matsushita T, et al. Characteristics of lithium-ion battery with non-flammable electrolyte. J Power Sources 2009;189:429–34. [201] Lee CW, Venkatachalapathy R, Prakash J. A novel flame-retardant additive for lithium batteries. Electrochem Solid State 2000;3:63–5. [202] Hu JL, Jin ZX, Hong H, et al. A new phosphonamidate as flame retardant additive in electrolytes for lithium ion batteries. J Power Sources 2012;197:297–300. [203] Dufek EJ, Klaehn JR, McNally JS, et al. Use of phosphoranimines to reduce organic carbonate content in Li-ion battery electrolytes. Electrochim Acta 2016;209:36–43. [204] Wu BB, Pei F, Wu Y, et al. An electrochemically compatible and flame-retardant electrolyte additive for safe lithium ion batteries. J Power Sources 2013;227:106–10. [205] Harrup MK, Rollins HW, Jamison DK, et al. Unsaturated phosphazenes as co-solvents for lithium-ion battery electrolytes. J Power Sources 2014;278:794–801. [206] Dalavi S, Xu M, Ravdel B, et al. Nonflammable electrolytes for lithium-ion batteries containing dimethyl methylphosphonate. J Electrochem Soc 2010;157:A1113–20. [207] Li WT, Campion C, Lucht BL, et al. Additives for stabilizing LiPF6-based electrolytes against thermal decomposition. J Electrochem Soc 2005;152:A1361–5. [208] Petibon R, Aiken CP, Ma L, et al. The use of ethyl acetate as a sole solvent in highly concentrated electrolyte for Li-ion batteries. Electrochim Acta 2015;154:287–93. [209] Yamada Y, Furukawa K, Sodeyama K, et al. Unusual stability of acetonitrile-based superconcentrated electrolytes for fast-charging lithium-ion batteries. J Am Chem Soc 2014;136:5039–46. [210] Suo LM, Hu YS, Li H, et al. A new class of solvent-in-salt electrolyte for high-energy rechargeable metallic lithium batteries. Nat Commun 2013;4:1481. [211] Suo LM, Borodin O, Sun W, et al. Advanced high-voltage aqueous lithium-ion battery enabled by “water-in-bisalt” electrolyte. Angew Chem Int Ed 2016;55:7136–41. [212] Yamada Yuki, Usui Kenji, Sodeyama Keitaro, et al. Hydrate-melt electrolytes for high-energy-density aqueous batteries. Nat Energy 2016;1:129. [213] Aurbach D, Levi MD, Levi E, et al. Common electroanalytical behavior of Li intercalation processes into graphite and transition metal oxides. J Electrochem Soc 1998;145:3024–34. [214] Evans JE, Friedrich H, Bals S, et al. Advanced tomography techniques for inorganic, organic, and biological materials. MRS Bull 2016;41:516–24. [215] Pennycook SJ, Nellist PD, editors. Scanning transmission electron microscopy: imaging and analysis New York, NY: Springer; 2011.

662

Batteries

[216] Bock DC, Pelliccione CJ, Zhang W, et al. Dispersion of nanocrystalline Fe3O4 within composite electrodes: insights on battery-related electrochemistry. ACS Appl Mater Inter 2016;8:11418–30. [217] Chen-Wiegart YCK, Liu Z, Faber KT, et al. 3D analysis of a LiCoO2-Li(Ni1/3Mn1/3Co1/3)O2 Li-ion battery positive electrode using X-ray nano-tomography. Electrochem Commun 2013;28:127–30. [218] Chen WC, Song YF, Wang CC, et al. Study on the synthesis–microstructure-performance relationship of layered Li-excess nickel–manganese oxide as a Li-ion battery cathode prepared by high-temperature calcination. J Mater Chem A 2013;1:10847–56. [219] Miller MK. Atom probe tomography: analysis at the atomic level. New York, NY: Kluwer Academic/Plenum; 2000. [220] Liaw BY, Kostecki R. In Situ characterization of lithium ion battery materials, electrodes, and cells. Electrochem Soc Interface 2011;20:41–2. [221] Shearing P, Wu Y, Harris SJ, Brandon N. In Situ X-ray spectroscopy and imaging of battery materials. Electrochem Soc Interface 2011;20:43–7. [222] Glazer MPB, Okasinski JS, Almer JD, Ren Y. High-energy X-ray scattering studies of battery materials. MRS Bull 2016;41:460–5. [223] Grey CP, Dupré N. NMR Studies of cathode materials for lithium-ion rechargeable batteries. Chem Rev 2004;104:4493–512. [224] Key B, Bhattacharyya R, Morcrette M, et al. Real-time NMR investigations of structural changes in silicon electrodes for lithium-ion batteries. J Am Chem Soc 2009;131:9239–49. [225] Verma P, Maire P, Novák P. A review of the features and analyses of the solid electrolyte interphase in Li-ion batteries. Electrochim Acta 2010;55:6332–41. [226] Vickerman JC. Surface analysis – the principal techniques. Chichester: John Wiley and Sons, Ltd.; 2009. [227] Xing YJ, Ma EWM, Tsui KL, Pecht M. Battery management systems in electric and hybrid vehicles. Energies 2011;4:1840–57. [228] Rezvanizaniani SM, Liu ZC, Chen Y, Lee J. Review and recent advances in battery health monitoring and prognostics technologies for electric vehicle (EV) safety and mobility. J Power Sources 2014;256:110–24. [229] Ding SP, Xu K, Zhang SS, et al. Diminution of supercooling of electrolytes by carbon particles. J Electrochem Soc 1999;146:3974–80. [230] Jansen AN, Dees DW, Abraham DP, et al. Low-temperature study of lithium-ion cells using a LiySn micro-reference electrode. J Power Sources 2007;174:373–9. [231] Belt JR, Ho CD, Miller TJ, et al. The effect of temperature on capacity and power in cycled lithium ion batteries. J Power Sources 2005;142:354–60. [232] Dubarry M, Truchot C, Liaw BY, et al. Evaluation of commercial lithium-ion cells based on composite positive electrode for plug-in hybrid electric vehicle applications III. Effect of thermal excursions without prolonged thermal aging. J Electrochem Soc 2013;160:A191–9. [233] Gering K. Low-temperature performance limitations of lithium-ion Batteries. ECS Transactions 2006;1:119. [234] Gering K. Method for accurate modeling and prediction of kinetic performance of electrochemical devices, with emphasis on low-temperature performance. Manuscript in preparation. [235] LAZAED. Lazard’s Levelized Cost of Energy Storage Analysis, Version 1.0. Available From: https://www.lazard.com/media/2391/lazards-levelized-cost-of-storage-analysis10.pdf; 2015. [236] DOE Global Energy Storage Database. Available From: http://www.energystorageexchange.org/projects. [237] Sam J. Battery systems for stationary storage – global market sizing, market forecasts and market dynamics. Cairn ERA. Available From: http://cairnera.com/research/ stationary-storage-report/; 2016. [238] Global Battery Management Systems. Available From: http://www.frost.com/sublib/display-report.do?id=9AAE-00-20-00-00; 2015.

Further Reading Aurbach D. 2000. Review of selected electrode-solution interactions which determine the performance of Li and Li ion batteries. J Power Sources 2000;89:206–18. Goodenough JB, Park KS. 2013. The Li-ion rechargeable battery: a perspective. J Am Chem Soc 2013;135:1167–76. Goodenough JB, Kim Y. 2011. Challenges for rechargeable batteries. J Power Sources 2011;196:6688–94. Larcher D, Tarascon JM. 2015. Towards greener and more sustainable batteries for electrical energy storage. Nat Chem 2015;7:19–29. Manthiram Arumugam, Muraliganth Theivanayagam. 2011. Lithium intercalation cathode materials for lithium-ion batteries. In: Daniel C, Besenhard JO, editors. Handbook of Battery Materials. Weinheim: Wiley-VCH Verlag GmbH & Co. KGaA; 2011. p. 341–75. Reddy T, Linden D. 2011. Linden’s handbook of batteries. New York, NY: McGraw Hill; 2011. Sun Y, Liu N, Cui Y. 2016. Promises and challenges of nanomaterials for lithium-based rechargeable batteries. Nat Energy 2016;1:16071. Tarascon JM. 2010. Key challenges in future Li-battery research. Philos Trans R Soc A 2010;368:3227–41. Tarascon JM, Armand M. 2001. Issues and challenges facing rechargeable lithium batteries. Nature 2001;414:359–67. Whittingham MS. 2004. Lithium batteries and cathode materials. Chem Rev 2004;104:4271–301. Xu K. 2004. Nonaqueous liquid electrolytes for lithium-based rechargeable batteries. Chem Rev 2004;104:4303–417. Xu K. 2014. Electrolytes and interphases in Li-ion batteries and beyond. Chem Rev 2014;114:11503–618. Yoo HD, Markevich E, Salitra G, et al. 2014. On the challenge of developing advanced technologies for electrochemical energy storage and conversion. Mater Today 2014;17:110–21.

Relevant Websites http://batteryuniversity.com/ Cadex Electronics Inc. http://www.greencarcongress.com/ Green Car Congress. http://www.uscar.org/guest/teams/12/U-S-Advanced-Battery-Consortium United States Advanced Battery Consortium LLC. https://energy.gov/eere/vehicles/vehicle-technologies-office-batteries Vehicle Technologies Office, US Department of Energy.

2.21 Supercapacitors Liang Chang and Yun Hang Hu, Michigan Technological University, Houghton, MI, United States r 2018 Elsevier Inc. All rights reserved.

2.21.1 Introduction 2.21.2 Principle and Structures of Supercapacitors 2.21.2.1 Structures and Operation Mechanisms of Capacitors 2.21.2.2 Measurement and Evaluation of Electrochemical Performance 2.21.2.2.1 Types of test cells 2.21.2.2.2 Electrochemical measurement 2.21.2.3 Electrochemical Parameters 2.21.2.4 Evaluation of Electrochemical Performance 2.21.3 Components of Supercapacitors 2.21.3.1 Current Collectors 2.21.3.2 Separators 2.21.3.3 Binders 2.21.3.4 Electrolytes 2.21.3.5 Electrode Materials 2.21.3.5.1 Electrodes for electric double-layer capacitors 2.21.3.5.2 Electrode materials for pseudocapacitors 2.21.3.5.3 Electrode materials for hybrid supercapacitors 2.21.4 Application of Supercapacitors 2.21.4.1 Small Consumer Electronics 2.21.4.2 Automotive Applications 2.21.4.3 Engine Starts 2.21.4.4 Electric and Hybrid Electric Vehicles 2.21.4.5 Electric Utility Applications 2.21.4.6 Industrial Application 2.21.5 Conclusions 2.21.6 Future Direction Acknowledgment References Further Reading Relevant Websites

Nomenclature CMC CMOS Cn CNTs Cp CT CV CVD 2D 3D DOD DOE EDL EDLCs EIS ES EV GCD GO HEV

Nacarboxymethyl cellulose Complementary metal oxide semiconductor Capacitance of negative electrode Carbon nanotubes Capacitance of positive electrode Overall capacitance Cyclic voltammetry Chemical vapor deposition Two dimensional Three dimensional US Department of Defense US Department of Energy Electric double layer Electric double-layer capacitors Electrochemical impedance spectroscopy Electrochemical supercapacitors Electric vehicles Galvanostatic charge/discharge Graphene oxide Hybrid electric vehicles

Comprehensive Energy Systems, Volume 2

IEC LCD LED MSRE NCE NHE PAA PANI PCs PILs PTFE PVDF PVP q RGO SAE SCE SHE SRAM SSCE

doi:10.1016/B978-0-12-809597-3.00247-9

664 665 665 665 665 666 667 667 668 668 668 668 669 669 670 672 673 676 677 678 679 679 682 683 683 683 683 683 695 695

International Electrochemical Commission Liquid crystal display Light emitting diode Mercury sulfate reference electrode Normal calomel electrode Normal hydrogen electrode Polyacrylic acid Polyaniline Pseudocapacitors Poly(ionic liquids) Polytetrafluoroethylene Poly(vinylidenefluoride) Polyvinylpyrrolidone Charge acceptance Reduced graphene oxide Society of Automotive Engineers Saturated calomel electrode Standard hydrogen electrode Static random access memory Saturated salt calomel electrode (saturated KCl)

663

664

Unit A C Cs d e E E I

2.21.1

Supercapacitors

Active surface of an electrode porous layer (m2) Capacitance (F) Specific capacitance (F/g) Effective thickness of the electric double layer (m) Dielectric constant of electrolyte (F m 1) Gravimetrically energy density (Wh/kg) Volumetrically energy density (Wh/L) Current (A)

m M P P ppm Rs t v V

Mass of an electrode (g) Total mass of two electrodes (g) Gravimetrically power density (W/kg) Volumetrically power density (W/L) part per million The equivalent internal resistance (O) Time (s) Scan rate (V/s) Potential

Introduction

The increase in demand for specialized energy devices is driving research to produce energy storage systems with high efficiency at lower cost [1–7]. Compared with other energy storage devices, electrochemical supercapacitors (ES) can possess fast charge/ discharge feature, long cycling life, excellent reversibility, high power density, and wide temperature-range operation [8–15]. These unique properties make supercapacitors widely applicable for braking systems of automotives, back-up systems of grid, and portable electronic devices [16–18]. There are two types of electrochemical supercapacitors depending on charge-storage mechanisms: (1) electric double-layer capacitors (EDLCs), in which charges are stored by electric double-layer (EDL) formation at the electrode/electrolyte interface [19–28], and (2) pseudocapacitors (PCs), which possess pseudocapacitance from oxidation–reduction reactions [29–35]. Furthermore, the combination between those two types of charge storage mechanism resulted in hybrid capacitors (Fig. 1) [36–46]. Their performances are strongly dependent on electrode materials, particularly their conductivity and accessible surface areas. The EDLCs, which are also called electrostatic supercapacitors with symmetrical electrodes as cathode and anode, were first commercialized by SOHIO and NEC. Furthermore, intensive R&D efforts were made to improve EDLCs [47–49]. The electrode material of traditional EDLCs is activated carbon with capacitance of 40–70 F/g. For PCs that also called faradaic supercapacitors, transition metal oxides, and conducting polymers are employed for electrodes. However, the low electrical conductivity and

Charge + + + + + + + + + + +

− −

Discharge −

− − − − −

+ +

−

+

+

− +

− +

+

+ + + +

− − − − − − − − − − −

+ + + + + + + + + + +

− − − −

−

− − − −

−

−

+ + + + + + + + + + +

−

− − − − − − − − − −

Supercapacitors

Electric double-layer capacitors

Hybrid capacitors

Fig. 1 Types of supercapacitors.

Pseudocapacitors

Supercapacitors

665

insufficient utilization of transition metal oxides and the poor cycling stability of conducting polymers lead to unfavorable rate performance and power density [50–62]. Different from both EDLCs and PCs that have symmetrical configurations, hybrid capacitors are constructed with an asymmetrical configuration, in which an electric double layer carbon material is usually used as an anode and a pseudocapacitive material as a cathode. In this chapter, the structures, materials, and performance of all those three types of supercapacitors will be discussed and their application in consumer electronics, automotives, electric utility and industrial areas will be focused.

2.21.2

Principle and Structures of Supercapacitors

2.21.2.1

Structures and Operation Mechanisms of Capacitors

As shown in Fig. 1, an electrochemical supercapacitor consists of two electroactive electrodes with a separator that electrically isolates the electrodes, metal foil or carbon impregnated polymers are as current collectors to conduct electrical current from each electrodes. Electrolytes impregnate the electrodes and the separators, which allows ions current to flow between the electrodes and prevent electron currents from discharging the cell [50,63–68]. Charges in an EDLC are stored at the interface between the electrode materials and the electrolyte, and the capacitance of the electric double-layer can be calculated by C¼

Ae 4pd

ð1Þ

where A is the active surface of an electrode porous layer, e the dielectric constant of electrolyte, and d the effective thickness of the electric double layer [69,70]. The capacitance of EDLCs is determined by the accumulation of electrostatic charges at the electrode/electrolyte interface [47,48,71–73]. The surface charge generation is based on mechanism as surface dissociation, ion adsorption from electrolyte, and lattice defect. To charge an EDLC, cations need to travel to the negative electrode and anions to the positive one via electrolyte, while electrons move from a negative electrode to a positive one via an external load. In contrast, the reverse processes happen for discharging the cell. During the charging and discharging processes, the electrolyte concentration remains unchanged without net ion exchanges between the electrode and the electrolyte. PCs are based on potential induced faradaic reactions (oxidation–reduction) of electroactive materials or electrosorption [74–82]. Oxidation–reduction reaction (Oad þ ne -Red) happens due to charge exchange across the double layer which is different from static separation of EDLCs. Electrosorption occurs when an electron-donating anion, for example, Cl , I , B and CNS , is chemically adsorbed as M þ A -MA(1 δ) þ δe . Thus, faradaic processes in pseudocapacitive electrodes can be summarized as reversible adsorption, exchanges of ions through electric double layers, and reversible redox reactions. These electrochemical processes occur not only on the surface but also in the bulk near the surface of the solid electrode, leading to much larger capacitance values and energy density for a PC than for an EDLC [83]. Pseudocapacitance thermodynamically originates from the change of the potential (DV) and the charge acceptance (Dq), which can be calculated by C¼

dðDqÞ dðDVÞ

ð2Þ

However, the faradaic processes are usually slower than non-faradaic processes. As a result, a PC has a relatively lower power density than an EDLC [84]. Furthermore, another drawback of a PC is its poor cycling-stability, because redox reactions at its electrodes can cause the degradation of electrode materials. In the recent years, the hybrid supercapacitors with an asymmetrical electrode configuration, in which one is a carbon electrode (EDL materials) and another consists of faradaic capacitive material (Fig. 1), have been extensively investigated to enhance overall cell voltage, energy, and power densities [85–87].

2.21.2.2 2.21.2.2.1

Measurement and Evaluation of Electrochemical Performance Types of test cells

Two types of test cells can evaluate the electrode performance of supercapacitors (Fig. 2). One type is three-electrode cells with working, counter, and reference electrodes, which are widely employed to evaluate single-electrode performance. Working electrodes are often fabricated by active materials. Active materials are mixed with a conductive agent and a binder in organic solution to form a homogeneous paste, followed by rolling to a strip and then pressing it on a current collector [88–97]. Platinum electrode is usually selected as a counter electrode. Various types of electrodes are exploited as a reference electrode, such as saturated calomel electrode (Hg/Hg2Cl2), silver/silver chloride electrode (Ag/AgCl), mercury/mercury oxide electrode (Hg/HgO), and mercury/mercury sulfate electrode. The conversion of aqueous reference-electrode potentials is listed in Table 1 [98,99]. Another type is two-electrode cells, in which a separator is between two working electrodes. This type cell is used to evaluate supercapacitor performance, which can provide reliable parameters (such as energy density, power density, and cell cycling life) for practical applications [100–104]. It should be noted that specific capacitance of a single electrode is different from its cell capacitance [105–111]. Furthermore, specific capacitance of a single electrode in two-electrode cell is not always consistent with that in three-electrode cell due to potential difference and other factors. The difference would be further expanded in PCs. So far,

666

Supercapacitors

(A)

(B)

Fig. 2 (A) Three-electrode cell and (B) two-electrode cell of supercapacitor.

Table 1

Conversion of aqueous reference-electrode potentials

Electrode

NHE/SHE

NHE/SHE SCE SSCE NCE MSRE Hg/HgO (1 M NaOH) Ag/AgCl (saturated NaCl/KCl) Ag/AgCl (3 M NaCl/KCl)

0 þ 0.241 þ 0.236 þ 0.280 þ 0.640 þ 0.098 þ 0.197 þ 0.209

SCE

0.241 0 0.005 þ 0.039 þ 0.399 0.143 0.044 0.032

SSCE

0.236 þ 0.005 0 þ 0.044 þ 0.404 0.138 0.039 0.027

NCE

MSRE

0.280 0.039 0.044 0 þ 0.360 0.182 0.083 0.071

0.640 0.399 0.404 0.360 0 0.542 0.443 0.431

Hg/HgO (1 M NaOH)

Ag/AgCl (saturated NaCl/KCl)

Ag/AgCl (3 M NaCl/KCl)

0.098 þ 0.143 þ 0.138 þ 0.182 þ 0.542 0 þ 0.099 þ 0.111

0.197 þ 0.044 þ 0.039 þ 0.038 þ 0.443 0.099 0 þ 0.012

0.209 þ 0.032 þ 0.027 þ 0.071 þ 0.431 0.111 0.012 0

Abbreviations: NHE: normal hydrogen electrode (aH þ ¼ 1); SHE: standard hydrogen electrode (aH þ ¼ 1); SCE: saturated calomel electrode (saturated KCl); SSCE: saturated salt calomel electrode (saturated NaCl); NCE: normal calomel electrode (1 M KCl); MSRE: Mercury (I) sulfate reference electrode (saturated K2SO4).

various structures of two-electrode cells have been designed to evaluate the electrochemical performance, including sandwiched flexible supercapacitors, fiber supercapacitors, and micro-supercapacitors [112–117].

2.21.2.2.2

Electrochemical measurement

Cyclic voltammetry (CV) is a potentiodynamic electrochemical measurement, in which a linearly changed electric potential was applied between cathode and anode for a two-electrode cell and between working and reference electrodes for a three-electrode cell [118–123]. A linearly changed electric potential is designated as the scan rate (v) and the potential range determined by electrode materials, electrolyte and test cell. CV curves can allow ones to determine capacitances, potential windows, and cycling life for electrodes. Furthermore, typical CV curves for electric double-layer materials are rectangular shapes, while the most CV curves of pseudocapacitive materials possess pairs of redox peaks. In principle, if a material exhibits both EDLCs and PCs behaviors, CV curves can be used to distinguish capacitance contributions of those two behaviors. This is because the responded current is proportional to the scan rate for EDL mechanism, but to the square root of scan rate for pseudocapacitive mechanism. Galvanostatic charge/discharge (GCD) is another type of measurements that are conducted by repetitive charging and discharging at constant current until reaching voltage set point [124–127]. It is important to select a proper current level that can make GCD data comparable and consistent. A typical charge/discharge profile is a symmetrical triangle curve for EDLCs and a platform line for PCs. This technique is generally chosen to evaluate capacitance, rate properties, and cycling stability. Electrochemical impedance spectroscopy (EIS) is widely used for characterization of charge storage mechanism, charge transfer, and mass transports [128–136]. It is a dielectric spectroscopic testing, measuring the impedance of a test cell as a function of frequency by applying a low-amplitude alternative voltage, normally 5 mV. EIS has two types of plots: the Nyquist plot that presents the imaginary and real parts of the cell impedances and the Bode plot that shows the cell response between phase angle and frequency. The contribution of each individual structure component in a cell to the total impedance can be identified by equivalent circuits and models [129,137,138].

Supercapacitors 2.21.2.3

667

Electrochemical Parameters

Toward energy storage devices, energy density, and power density are the most important parameters to evaluate the electrochemical performance and provide the basis for practical application [139–146]. Energy density (E) proves the amount of electrical energy stored and deliverable, that calculated gravimetrically or volumetrically in Wh/kg or Wh/L by [147] E¼

1 1 CV 2 ¼ QV 2 2

ð3Þ

Power density, in W/kg or W/L, describes the efficacy in energy uptake/delivery, can be otained by [148] P¼

1 2 V 4RS

ð4Þ

where V is the voltage built up across the two electrodes, Q the stored total charges, RS the equivalent internal resistance, and C the capacitance. Besides, a simplified equation can also be used to obtain power density, which is reported previously [149,150] P¼

E t

ð5Þ

where t is the discharge time. It is worth noting that capacitance in Eq. (3) is overall capacitance (CT) of a supercapacitor cell which can be calculated by treating it as two capacitors in series as follows [151]: 1 1 1 ¼ þ CT Cp Cn

ð6Þ

where Cp and Cn are the capacitance of positive and negative electrodes, respectively. For a symmetrical supercapacitor cell that contains two same electrodes, Cp and Cn are equal. In contrast to an asymmetrical supercapacitor cell fabricated with two different electrodes, Cp a Cn and the electrode capacitance of two electrodes should be balanced to fully utilize the charge storage ability of two electrodes. To achieve the same capacitance, electrodes with larger capacitance should possess less mass. Specific capacitance (Cs) with a unit of Faraday per gram (F/g) is preferred to describe the charge storage ability of electrode materials. In the three-electrode configuration, Cs can be calculated from cyclic voltammetry or galvanostatic discharge curve as follows: R IðV ÞdV ð7Þ Cs ¼ 2mvDV It ð8Þ mDV where I is current, t discharge time, m the mass of an electrode, V the potential range, and v the scan rate. In the two-electrode configuration, Eqs. (9) and (10) are used to obtain Cs from cyclic voltammetry or galvanostatic discharge curves [152–154]: R IðV Þ dV ð9Þ Cs ¼ mvDV Cs ¼

Cs ¼

4It MDV

ð10Þ

where M is the total mass of two electrodes. Those equations can also be employed to calculate the areal capacitance (in F/cm2) or volumetric capacitance (in F/cm3) by replacing the mass (m or M) of electrode material with area (/cm2) or volume (/cm3) of electrodes [155,156].

2.21.2.4

Evaluation of Electrochemical Performance

To standardize the evaluation for supercapacitor devices, US Department of Defense (DOD), US Department of Energy (DOE), International Electrochemical Commission (IEC), and Society of Automotive Engineers (SAE) have established a series of standards, which are listed in Table 2 [157]. Those have been applied for commercial products with fixed materials, fabrications, and cell designs. However, in research, those evaluation standards are not accommodated due to differences in all parts of supercapacitors, new designs, and different configurations [158,159]. The performance of supercapacitors is mainly evaluated on the basis of acceptably high energy densities (410 Wh kg), substantially greater power density, fast charge discharge process (within seconds), excellent cycle stability (more than 100 times of batteries), low self-discharging, safe operation, and low cost [160,161]. The power density and energy density are determined by cell (total) capacitance, operating voltage, and equivalent series resistance. The capacitance of the cells is mainly dependent on electrode materials. Requirements for an ideal electrode material with excellent performance are high specific capacitance, excellent rate capability, and high cycle stability. Therefore, electrode materials should possess large surface area, suitable pore size, good electrical conductivity, high chemical/electrical stability [162–167]. To further improve the capacitance and accommodate commercial requirements, the introduction of pseudocapacitive materials and the high packing density of electrodes are necessary [168–174]. Furthermore, the cell voltage is determined by the thermodynamic stability of an electrolyte solution.

668

Table 2

Supercapacitors

A chronologic review of supercapacitor evaluation standards

Year

Organization

Title

Document ID

1986 1994 2004 2006 2009

DOD DOE DOE IEC IEC

DOD-C-29501 DOE/ID-10491 DOE/NE-ID-11173 IEC 62391 IEC 62576

2012

IEC

2013

SAE

Capacitors, fixed, electrolytic, double layer, carbon (metric), general specifications Electric vehicle capacitor test procedures manual FreedomCAR ultracapacitor test manual Fixed electric double layer capacitor for use in electronic equipment Electric double-layer capacitors for use in hybrid electric vehicles-Test methods for electrical characteristics Railway applications-Rolling stock equipment-Capacitors for power electronics-Part 3: Electric double-layer capacitors Capacitive energy storage device requirements for automotive propulsion applications

IEC 61881-3 J3051

Source: Reproduced with permission from Zhang S, Pan N. Supercapacitors performance evaluation. Adv Energy Mater 2015;5:1401401.

2.21.3 2.21.3.1

Components of Supercapacitors Current Collectors

As an indispensable part of supercapacitors, a current collector not only provides a robust support for electrode materials, but also connects electrodes to capacitor terminal. The contact impedance between electrodes and current collectors can largely affect the capacitance, energy density, power density, and cycle life of supercapacitors [175–178]. The selection of current collectors mainly depends on electrolytes. Titanium (or Au) foil is usually chosen as current collector for working electrode in an acidic aqueous electrolyte [179,180], nickel foil (or nickel foam) for that in alkaline aqueous solutions [181–183], and aluminum foil for that in nonaqueous electrolyte [184]. Ideal current collectors for nonaqueous solution should possess excellent interface conductivity and electrochemical stability at high potentials [184–188]. Surface treatment of current collectors can be used to decrease ohmic drops at the interface and coatings on it can enhance their electrochemical stability at high potential [189–194]. Nanostructured current collectors are employed to improve the interface between active materials and current collectors by increasing contact area [195–197]. Furthermore, the growing market of wearable electronic devices and flexible supercapcitors has stimulated the development of flexible current collectors [198–208]. Those include carbon in the form of a highly conductive nanotube [209,210] or graphene [211,212], metallic fabrics [213], polymer film [214–217], and Ni foam [218].

2.21.3.2

Separators

Commonly used separators are made from thin and highly porous membranes or films, such as glass fibers, cellulose, and polymer membranes [219–224]. The function of the membranes is to separate cathode and anode in a supercapacitor cell. Qualified separators must possess minimal resistance for ions transfer, excellent electronic insulating feature, high electrochemical stability, and high mechanical strength [225,226]. The choice of separators are also affected by the type of electrodes, working temperature, and operating potential [227,228]. Furthermore, the surface morphology, pore structure, thickness, and chemical composition of separators can generate remarkable influence on specific capacitance, specific energy and power densities, electronic series resistance, and polarizability limits [229–231]. Besides traditional separator materials, new separator materials have been explored for supercapacitors, such as GO films [232], and eggshell membranes [233].

2.21.3.3

Binders

Fluorinated polymetric materials, such as poly(vinylidenefluoride) (PVDF) and polytetrafluoroethylene (PTFE), are widely used as binders to enhance the electrical contact of active materials and their adhesion to current collectors in supercapacitor cells. The binders can also improve the mechanical and electronical properties of electrodes, but sacrifice their diffusion of ions. Therefore, it is important to optimize the binder content in electrodes [234]. Furthermore, the performance of supercapacitors can be greatly affected by binder properties [226,235–237]. PTFE as a binder in a negative electrode layer showed a beneficial effect to achieve higher energy and power performance than PVDF. The replacement of PVDF binder by polycarboxylate can inhibit the initially irreversible capacity of potassium-ion capacitors [238]. Various developments in binders have emerged for better electrochemical performance. Besides mechanically mixing, numerous techniques were explored to improve the contact between binders and active materials. For example, bonds between the carbon additive and electroactive materials can be enhanced via absorption of ionic liquid on particle surfaces [239–242]. Binders may provide additional capacitance for supercapacitors [243]. Poly(ionic liquid)s (PILs) are electrochemically permeable to make active materials more accessible interfacially, enhancing specific capacitance and cyclability [244]. Furthermore, electrochemical performance can be improved by creating dual functionalities in a binder materials as both active materials and binders. A possible method is to attach redox molecules to polymeric skeleton, resulting in redox binders [245]. In addition, environmental-friendly binders have been developed, such as Polyvinylpyrrolidone (PVP), polyacrylic acid (PAA), casein, and Na-carboxymethyl cellulose

Supercapacitors

669

(CMC) [246–251]. Those new types of binders not only reduce environmental pollution but also bring positive effect on electrochemical performance [252–256].

2.21.3.4

Electrolytes

Electrolytes can significantly affect the energy density and power density of supercapacitors [257–259]. An ideal electrolyte for ES should possess the following properties: (a) a wide voltage window, (b) high electrochemical stability, (c) large ionic concentration, (d) small solvated ionic radius, (e) low resistance, and (f) others (such as low viscosity, volatility, toxicity, lowtemperature property and cost) [260–262]. The electrolytes (Table 3) for ES cells can be classified into three types: aqueous electrolytes, organic electrolytes, and ionic liquids. Aqueous solutions of acids, bases and salts (such as H2SO4, KOH, Na2SO4, and NH4Cl) are commonly used as electrolytes to provide a high ionic concentration and low resistance for supercapacitors, leading to large capacitance and high power density. However, aqueous electrolytes have a small voltage window of about 1.2 V and thus limit the improvement of both energy and power densities. Compared with aqueous electrolytes, organic electrolytes can provide a higher voltage window up to 3.5 V. The most commonly used solvents for organic electrolytes are acetonitrile and propylene carbonate [263]. Acetonitrile can dissolve larger amounts of salt than other solvents, but cause environmental and toxic issues. The use of propylene carbonate-based electrolytes can solve those problems. Organic salts (such as tetraethylammonium tetrafluoroborate and tetraethylphosphonium tetrafluoroborate) are also exploited in electrolytes. It was demonstrated that salts with less symmetric structures have a larger solubility due to their lower crystal-lattice energy. However, the impurity of water can cause a serious issue, namely, if the water content in organic electrolytes is above 3–5 ppm, the ES’s voltage will be significantly reduced [264,265]. Ionic liquids can also be exploited as ES electrolytes because of their promising properties, such as low vapor pressure, high thermal and chemical stability, and low flammability [266–268]. They possess a wide range (2–6 V, typically about 4.5 V) of electrochemical operating window [269,270]. Their conductivities are at level of about 10 m S/cm [271]. The ionic liquids (widely explored for supercapacitors) are imidazolium, pyrrolidinium, and aliphatic quaternary ammonium salts [272–275]. It is still a challenge for ionic liquids to obtain a wider potential range together with high conductivity in a wide temperature range.

2.21.3.5

Electrode Materials

Carbon is the most important electrode materials for supercapacitors. It is well known that carbon possesses a large variety of physical properties due to its bonding flexibility. Graphite and diamond are well known as three-dimensional (3D) carbon Table 3

Commonly used electrolytes and solvents

Electrolytes

Ion size (nm) Cation

Inorganic electrolytes H2SO4 KOH Na2SO4 NaCl LiPF6 LiClO4 Organic electrolytes (C2H5)4N BF4(TEA þ BF4 ) (C2H5)3(CH3)N BF4(TEMA þ BF4 ) (C2H5)4P BF4(TEP þ BF4 ) (C4H9)4N BF4(TBA þ BF4 ) (C6H13)4N BF4(THA þ BF4 ) (C2H5)4N CF3SO3 (C2H5)4N (CF3SO2)2N(TEA þ TFSI ) Solvents Acetonitrile (AN) g-Butyrolactone (GBL) Dimethyl kotone (DMK) Propylene carbonate (PC) Water

Anion

0.533 0.26 0.36 0.36 0.152 0.152 0.686 0.654

0.533 0.508 0.474

0.830 0.96 0.686 0.68

0.458 0.458 0.458 0.458 0.458 0.540 0.650

Melting point (oC)

Viscosity (Pa/s)

Dielectric constant, e

0.369 1.72 0.306 2.513

36.64 39 21.01 66.14

43.8 43.3 94.8 48.8

Source: Reproduced with permission from Inagaki M, Konno H, Tanaike O. Carbon materials for electrochemical capacitors. J Power Sources 2010;195:7880–903.

670

Supercapacitors

allotropes. The zero-dimensional carbon fullerenes and one-dimensional carbon nanotubes were discovered within the last 35 years [276–278]. However, two-dimensional (2D) graphene has the shortest history [279]. Graphene, which is a one-atom thick layer made out of carbon atoms in hexagonal ‘‘honeycomb” carbon lattice, constitutes the basic structure of other carbon allotropes, namely, the graphite, fullerenes, and nanotubes can be formed by stacking, wrapping, and rolling-up graphene, respectively. Within the last decade, graphene has become an exciting model system for two-dimensional materials [280–285]. Furthermore, graphene exhibits many unique properties ensuring its promising application in supercapacitors, which will be emphasized in the following sections.

2.21.3.5.1

Electrodes for electric double-layer capacitors

Various carbon forms were explored as electrode materials for EDLCs, including activated carbon, carbon aerogels, carbon nanofibers, and carbon nanotubes (CNTs) [286–298]. Furthermore, commercialized EDLCs are mainly made up of activated carbon. However, the rich micropores and poor conductivity of activated carbon cannot meet the requirements for an ideal EDLC electrode (large ion-accessible surface areas and high electrical conductivity) to obtain desirable specific capacitance, rate performance, high energy and power density, and excellent cycling stability [299,300]. In contrast, graphene is a very promising electrode material for EDLCs due to its high electrical conductivity, large theoretical surface areas of 2630 m2/g (corresponding with 525 F/g theoretical gravimetrical capacitance), large mechanical flexibility, and high chemical stability [301–305]. Ruoff and his coworkers pioneered the application of graphene for EDLCs [306]. They reduced graphene oxide to prepare modified graphene sheets with good electrical conductivity and large surface area, leading to the specific capacitance of 135 F/g and 99 F/g at 10 mA/g in 5.5 M KOH aqueous solution and TEABF4/AN organic electrolyte, respectively. However, the obtained capacitance of graphene electrodes is far below the theoretical value. This happened because the surface area associated with large amounts of micropores was inaccessible to electrolyte ions and the graphene sheets were aggregated during the fabrication of electrodes [307,308]. To solve those issues, various strategies have been employed to develop suitable graphene materials for EDLCs. To enlarge electrolyte ion-accessible surface area of graphene, Zhu et al. synthesized KOH-activated microwave exfoliated graphene oxide (a-MEGO) [309]. The a-MEGO have homogeneously-distributed 3D micropores and mesopores with a large BET surface area up to 3100 m2/g. CV curves showed ideal rectangular shapes at scan rate of 100 mV/s with only slight distortion at a large scan rate up to 500 mV/s, indicating good double-layer performance and rate capacity of the a-MEGO electrodes. The specific capacitance in (EMIMBF4)/AN electrolyte reached 165, 166, and 166 F/g at 1.4, 2.8, and 5.7 A/g, respectively. The excellent performance of the a-MEGO electrodes might be attributed to their enhanced accessible surface areas and electrolyte ion transfer. Similar results were also observed by other groups [310–313]. Doping graphene with surface functional groups (containing O, N, and S) can introduce pseudocapacitance and improve the electroactive surface area and conductivity of graphene to enhance its EDL performance [314–322]. For example, a simple plasma process was employed to synthesize nitrogen-doped graphene, producing a capacitance of 280 F/g that is 4 times larger than the pristine graphene [323]. Even at high current densities, the improved capacitance can remain well with an excellent cycling stability up to 230,000 cycles. In addition to gravimetric capacitance, volumetric and areal capacitances must also be considered for practical supercapacitors. It is important to achieve large areal and volumetric capacitances at a high mass loading of electrode material without sacrificing gravimetric capacitance [324–329]. A useful strategy is to develop the porous yet highly dense graphene electrodes [330–333]. Li and his coworkers synthesized electrolyte-mediated chemically converted graphene (EM-CCG), namely, 2D graphene sheets were self-assembled to porous yet densely packed carbon electrodes with nearly face-to-face stacked graphene sheet morphology using directional-flow-induced bottom-up assembly process [330]. The EM-CCG electrodes possesses a high packing density of 1.33 g/ cm3 with large ion-accessible surface area and low ion transport resistance, leading to only slightly decreased volumetric capacitance even at a large mass loading up to 10 mg/cm2 and volumetric energy densities close to 60 W-h per liter. High density porous graphene macroform (HPGM) could also be prepared using an evaporation-induced drying of a graphene hydrogel, and the EDLCs with HPGM electrodes can achieve a recorded volumetric capacitance of 376 F/cm3 in aqueous electrolyte [331]. Another attracting strategy is to construct three-dimensional graphene network with hierarchical porous structure and functional groups [334–340]. 3D graphene foams can be synthesized by chemical vapor deposition (CVD) process using Ni foam as a template [337]. Furthermore, 3D crumpled paper balls of graphene, which were produced by transforming 2D graphene sheets using an aerosol spray drying process, exhibited promising electrochemical performance at high current density and high mass loading level (150 F/g at 0.1 A/g with mass loading of 20 mg/cm2) [338]. 3D hierarchical porous graphene (HPG), which was prepared by an ion-exchange method, possesses an ultrahigh surface area of 1810 m2/g, generating a large specific capacitance of 305 F/g at 0.5 A/g (which could remain by 75% even at current density of 100 A/g) [339]. Recently, we invented a one-step exothermic approach directly from CO2 to 3D cauliflower-fungus-like graphene (CFG) with hierarchical mesoporous structure and large fully accessible surface area (462 m2/g) (Fig. 3(A) and (B)) [334]. The unique structure makes 3D CFG electrodes can increase mass loading to 11.16 mg/cm2 without sacrificing gravimetric capacitance and obtain an ultrahigh areal capacitance of 1.16 F/cm2 with this very large-efficient-mass-loading (Fig. 3(C) and (D)). Besides, the 3D CFG electrodes exhibited obviously larger CV areas and specific capacitance than commercial-used activated carbon electrodes (Fig. 3(E) and (F)). Even at current density of 5 A/g, the increasing mass loading to commercial requirement (10 mg/cm2) brought a slight decrease on specific capacitance. Therefore, one can conclude that 3D graphene is bringing a bright future for supercapacitors [341,342].

Supercapacitors

(A)

(B) 160

1.6 M1 M2 M3

M1 M2 M3 Specific capacity (F/cm2)

Specific capacity (F/g)

671

120

80

40

0

1.2

0.8

0.4

0.0 2

0

4 6 8 Current density (A/g)

(C)

10

0

2

4 6 8 Current density (A/g)

(D)

10

160 3D graphene Activated carbon Specific capacity (F/g)

Current density (A/g)

8

4

0

−4

120

80

40

3D graphene Activated carbon

−8

0 0.0 (E)

0.2

0.4 0.6 Potential (V)

0.8

1.0

0 (F)

4

8 12 Mass loading (mg/cm2)

16

Fig. 3 (A) FESEM image and (B) TEM image of 3D cauliflower-fungus-like graphene (CFG), (C) mass capacitance and (D) areal capacitance of CFG-based supercapacitors with different graphene-mass-loadings (M1: 4.58 mg/cm2, M2: 7.40 mg/cm2, M3: 11.16 mg/cm2) vs. current density, (E) CV curves at a scan rate of 100 mV/s for supercapacitors with CFG and activated carbon, (F) mass capacity vs. mass loading at current density of 5 A/g. Reproduced with permission from Chang L, Wei W, Sun K, Hu YH. 3D flower-structured graphene from CO2 for supercapacitors with ultrahigh areal capacitance at high current density. J Mater Chem A 2015;3:10183–7.

Supercapacitors

672

2.21.3.5.2

Electrode materials for pseudocapacitors

For PCs, pseudocapacitive materials are introduced into graphene matrix to obtain higher specific capacitance. Transition metal oxides, transition metal sulfates, and conducting polymer possess high theoretical capacitance up to thousands, which are much larger than the theoretical capacitance (525 F/g) of graphene. However, the pseudocapacitive materials often suffer degradation and low conductivity, and limited reversibility and rate capability [343–345]. It has been demonstrated that the introduction of pseudocapacitive materials into a graphene matrix provides a solution for those issues [346–348]. A well-designed electroactive materials/graphene composites can bring synergistic effect to acquire electrochemical performance with high energy density, high power density, excellent cycle stability, and good reversibility [183,335,349–359]. Graphene can be used as a substrate for homogeneous growth of pseudocapacitive transition metal oxides (or sulfates) to improve the conductivity of whole electrodes and mass ratio of pseudocapacitive materials participated in electrochemical reaction, while pseudocapacitive materials can serve as a separator to inhibit the aggregation of graphene sheets and thus maximize electric double-layer formation [350–352]. As a presentative type of composite materials, hydrous ruthenium oxide/graphene sheet composites (ROGSCs) were synthesized by combining a sol–gel method and a low-temperature annealing process [353]. Furthermore, the ROGSCs with 38.3% Ru content exhibited high specific capacitance of 570 F/g, retention of 97.9% after 1000 cycles, and energy density of 20.1 Wh/kg (Fig. 4). The excellent electrochemical performance was attributed to the full utilization of electric double-layer capacitance of graphene sheets and pseudocapacitance of RuO2. Improved capacitances were also observed for similar composite electrodes, such as Co3S4-hollow nanospheres/graphene composites, Mn3O4 nanorods/graphene sheets, and MoSe2/RGO sheets [183,354,355]. Furthermore, high loading of active materials in a novel graphene/MnO2 textile electrode was achieved by electrodepositing MnO2 nanomaterials on graphene nanosheets coated on porous textile fibers, resulting in a high specific capacitance of 315 F/g that is 4–5 times large than barely graphene sheets. (Fig. 5) [356]. The excellent performance would be due to efficient use of the pseudocapacitive MnO2 materials for charge storage and the 3D conductive frameworks for ions/ electrons transfer. It should be emphasized that 3D graphene, which has a large surface area, a high electrical conductivity, and an interconnected macroporous structure, can be used as an ideal scaffold for pseudocapacitive materials [357–359]. For example,

(a) As prepared graphene

Specific capacitance (F/g)

800

(d) Hydrous RuO2 - anchored graphene

Dispersion

150 °C for 2h

0.1 M Ru3+ 1 M NaOH, pH=7

700

Calculated Csp of GSs Calculated Csp of RuO2

600

Cspcal

500

Csp Enhanced Csp

exp

400 300 200 100 0

(b) Dispersion graphene

(A)

0

(c) Formation of ruthenium hydroxide -anchored graphene

(B)

10

20

30

40

50

60

70

Ru content (wt%)

100

GSs

80

80 60 RuO2

60

40

Energy density (Wh/kg)

100 ROGSCs

Coulombic efficiency (%)

Capacitance retentional (%)

120

10

1

40 0 (C)

200

400 600 Cycle number

800

20 1000

0.1 101 (D)

ROGSCs GSs RuO2 102 102 Power density (W/kg)

104

Fig. 4 (A) The preparation of ruthenium oxide/graphene sheet composites (ROGSCs) by combining a sol–gel method and low-temperature annealing. (B) Experimental and calculated specific capacitance of ROGSCs. (C) Cycling stability of the as-prepared GSs, RuO2, and ROGSCs (Ru, 38.3 wt%) measured with a two-electrode cell at a current density of 1 A/g. (D) Ragone plot for the as-prepared GSs, RuO2, and ROGSCs (Ru, 38.3 wt%) supercapacitors. Reproduced with permission from Wu ZS, Wang DW, Ren W et al. Anchoring hydrous RuO2 on graphene sheets for high-performance electrochemical capacitors. Adv Funct Mater 2010;20:3595–602.

Supercapacitors

(i)

Microfibers in textile (A)

673

(ii)

Graphene nanosheets -coated textile fibers

MnO2 deposited graphene-textile

G/MnO2

Specific capacitance (F/g)

300

G-only

200

100

0

0

20

40

60

80

100

Scan rate (mV/s) (B)

(C)

Fig. 5 (A) Schematic illustration of preparing hybrid graphene/MnO2 nanostructured textiles, (B) SEM image of MnO2 deposited graphene textile, (C) specific capacitance of graphene sheets and MnO2 deposited graphene textile at different scan rates. Reproduced with permission from Yu G, Hu L, Vosgueritchian M, et al. Solution-processed graphene/MnO2 nanostructured textiles for high-performance electrochemical capacitors. Nano Lett. 2011;11:2905–11.

CoMoO4-3D graphene (NSCGHs), which was synthesized by combining a CVD and a hydrothermal reaction (Fig. 6), possesses unique properties, including excellent electrical conductivity of 3D graphene, large contact areas and structure stability from strongly coupling between CoMoO4 and 3D graphene film, and full utilization of CoMoO4 materials [357]. The NSCGH electrode exhibited an ultrahigh specific capacitance of 2741 F/g (at 1.43 A/g), 439.7 F/g (at 400 A/g) and an excellent capacity retention of 96.36% after 100,000 cycles. Conducting polymers are another type of pseudocapacitive materials, which possess excellent electrical conductivity and high theoretical capacitance. However, their degradation from shrinkage and swelling during charge/discharge processes and their slow charge transfer reaction cause low rate performance and poor cycling stability, which have greatly limited their practical applications for supercapacitors [52,360–363]. To solve those issues, graphene/polymer composite materials were developed [364]. As shown in Fig. 7(A), a solution process was employed to synthesize polyaniline (PANI)/reduced graphene oxide (RGO) hybrid film [365]. The composite film has a larger electrical conductivity (906 S/cm) than both PANI (580 S/cm) and RGO (45.6 S/cm) components. Furthermore, the nanosized PANI can provide high electroactive regions and short diffusion distances for electrolyte ions. As a result, the capacitance and cycling stability of PANI/RGO electrodes were remarkably improved (Fig. 7(B) and (C)). The performance of conducting-polymer/graphene composites can be further improved by introducing metal oxides as third component [366–370].

2.21.3.5.3

Electrode materials for hybrid supercapacitors

Advanced supercapacitors must operate with a large potential window, a high energy density and power density, and a long cycle life. EDLCs possess high power density and good cycling stability, but their energy densities are low with limited specific

Supercapacitors

674

Ni etching and hydrothermal

CH4 1273 K Ni foams

Ni and 3D graphene

CoMoO4 and graphene

(A)

Current (A)

0.2

0.0 70 mv/s 100 mv/s 150 mv/s 200mv/s 300 mv/s

−0.2 0.0 (B)

0.4 0.8 Potential (V) vs. Hg/HgO

Potential (V) vs. Hg/HgO

1.0 5 mv/s 10 mv/s 20 mv/s 30 mv/s 50 mv/s

1.43 A/g 2.43 A/g 5.00 A/g 9.28 A/g 12.8 A/g 15.71 A/g 22.85 A/g 35.71 A/g 85.71 A/g

0.8 0.6 0.4 0.2 0.0 0

(C)

400

800 1200 Time (s)

1600

Fig. 6 (A) Synthesis procedure of nanohoneycomb-like strongly coupled CoMoO4-3D graphene, (B) cyclic voltammetry (CV) curves of NSCGH electrodes at different scan rates, (C) galvanostatic charge/discharge profiles of NSCGH electrodes at different current densities. Reproduced with permission from Yu X, Lu B, Xu Z. Super long-life supercapacitors based on the construction of nanohoneycomb-like strongly coupled CoMoO4-3D graphene hybrid electrodes. Adv Mater. 2014;26:1044–51

capacitance (100–200 F/g) and narrow operating voltage (about 1 V). Although the introduction of pseudocapacitive materials into electrodes can substantially improve their specific capacitances and energy densities, it must sacrifice their power densities and cycle life. For this reason, hybrid supercapacitors were constructed with a battery-type Faradaic electrode as the energy source and a capacitor-type electrodes as the power source, forming an advanced supercapacitor device with both high energy density and high power density [371–375]. The hybrid supercapacitors with asymmetrical configuration can be divided into two types: redox// double layer and redox//redox. The hybrid supercapacitors with redox//double layer configuration can increase capacitance by the faradic charging/discharging of positive electrodes and maintain high-rate performance by double-layer charge accumulation of negative electrodes. Graphene has been widely explored to improve the electrochemical performance of transition metal oxides/sulfates. For example, to fabricate an asymmetrical supercapacitor, graphene/MnO2 was used as a cathode, activated carbon nanofiber (ACN) as an anode, and aqueous Na2SO4 solution as an electrolyte. The graphene/MnO2 electrode provided high capacitance from redox reaction, while the ACN electrode was responsible for fast ion-transport due to its double-layer property. The hybrid supercapacitor showed a voltage range of 0–1.8 V, energy density of 51.1 Wh/kg, and 97% capacity retention after 1000 cycles [376]. Furthermore, graphene can be used not only as a 3D-skeleton for MnO2 deposition, but also as an active material for an anode [377]. As shown in Fig. 8(A) and (B), 3D embossed-chemically-modified-graphene (e-CMG) film were fabricated by an embossing technique with a sacrificial templates (polystyrene colloidal particles), leading to large surface area and an interconnected structure of 3D porous networks. After amorphous MnO2 was coated on e-CMG surface, the reserved macroporous network provides three dimensionally connected MnO2 for efficient and fast electrolyte-ions diffusion, while interconnected CMG sheets contribute to the double-layer charge storage and constitute a continuous pathway for the electrons transfer to MnO2 (Fig. 8(C)). The e-CMG//MnO2/e-CMG

Supercapacitors

H N

Cr N+ H

H N+ Cr

H N

N n H

N H

N

675

N n

APS NH4OH

Aniline HCl

De-doped PARG

Primary doped PARG

Prepared RGO (A)

CSA

H N

Drop casting

m-Cresol

Annealing

CHCl3

Large-scale Re-PARG film 10

N n H

600

Re-PARG film Re-PANI film RGO Capacitance (F/g)

Current (mA)

H N+ CSA−

Re-doped PARG

Solvent Re-PARG

5

0

−5

400

200

−10

0 0.0

(B)

CSA− N+ H

0.1

0.2 0.3 0.4 0.5 0.6 Potential (V vs. Ag/AgCl)

0.7

0.8

RGO

Re-PANI film

Re-PARG film

(C)

Fig. 7 (A) Schematic illustration of large-scale Re-PARG film fabrication, (B) cyclic voltammetry (CV) curves at a scan rate of 5 mV/s, and (C) gravimetric capacitance at current density of 0.45 A/g. Reproduced with permission from Kim M, Lee C, Jang J. Fabrication of highly flexible, scalable, and high-performance supercapacitors using polyaniline/reduced graphene oxide film with enhanced electrical conductivity and crystallinity. Adv Funct Mater 2014;24:2489–99.

asymmetrical supercapacitor, which is illuminated in Fig. 7(D), can extend potential range to 2 V and achieve a maximum energy density of 44 Wh/kg and a power density of 25 kW/kg. Similar strategies were also applied for various electrode materials of asymmetrical supercapacitors, such as graphene//Ni(OH)2/graphene [378]. In addition to common binary composites, ternary composites have also been evaluated as electrode materials. It was demonstrated that highly porous Ni(OH)2–MnO2–RGO hybrid spheres, which were constructed by a facile one-step hydrothermal codeposition method (Fig. 9), are favorable structures for electrochemical performance: (1) the hybrid spheres possess high surface areas, abundant porous structure, and well-defined sphere morphology for facile electrolyte diffusion and transportation and (2) the intercalation of Ni(OH)2 and MnO2 between graphene sheets can increase the total conductivity and enhance the synergistic effect of ternary composite electrodes [379]. Indeed, the asymmetric supercapacitor of Ni(OH)2–MnO2–RGO//freeze-dried reduced graphene oxide exhibited a high energy density of 32.6 Wh/kg and a large power density of 305 W/kg in 0–1.6 V.

Supercapacitors

676

(A) Deposition of MnO2

Removal of PS

PS-embedded CMG film (B)

MnO2/e-CMG composite film

3D macroporous e-CMG film e-CMG framework Ultrathin MnO2

(C)

(D) Current collector

+

Na cations

3D ion and electron pathways

MnO2/e-CMG film e-CMG film (F) 100

(E)

2

Energy density (Wh/kg)

Current density (A/g)

Separator and electrolyte

0

−2

−4 0.0

0.5

1.0 Voltage (V)

1.5

2.0

ref. 23

ref. 39

ref. 40

ref. 22 10

ref. 16

e-CMG/MnO2/e-CMG MnO2/e-CMG//MnO2/e-CMG e-CMG//e-CMG CMG//MnO2/CMG

1 0.1

1 10 Power density (kW/kg)

100

Fig. 8 (A) Schematic illustration of MnO2/e-CMG film preparation, (B) high-magnified cross-sectional SEM image of e-CMG film, (C) illustration of 3D ionic and electronic transport pathways in the MnO2/e-CMG electrode, (D) schematic diagram of e-CMG//MnO2/e-CMG-based asymmetric supercapacitor device, (E) cyclic voltammetry (CV) curves for asymmetrical supercapacitors with different cell voltages at scan rate of 50 mV/s, and (F) Ragone plot of the symmetric (e-CMG//e-CMG and MnO2/e-CMG//MnO2/e-CMG) and asymmetric (CMG//MnO2/CMG and e-CMG//MnO2/eCMG) supercapacitors compared with data in other literature. Reproduced with permission from Choi BG, Yang M, Hong WH, Choi JW, Huh YS. 3D macroporous graphene frameworks for supercapacitors with high energy and power densities. ACS Nano 2012;6:4020–8.

Asymmetric supercapacitors with the redox//redox configuration are also attractive. Fig. 10(A) shows asymmetric graphene/ MnO2 nanospheres//graphene/MoO3 nanosheets supercapacitors [380]. Even operated with aqueous Na2SO4 electrolyte, the supercapacitors achieved a wide operation voltage of 2 V, a large specific capacitance of up to 307 F/g, and an energy density of 42.6 Wh/kg at a power density of 276 W/kg. The wide operation voltage can be attributed to the large work function difference between MnO2 and MoO3, which was demonstrate by CV measurements (Fig. 10(B)).

2.21.4

Application of Supercapacitors

In recent years, rapid progresses in smart phones and sensors increase consumer demands for fast-charging devices. Electric/hybrid vehicles, large power generation plants, and electricity distribution grids need efficient storage systems. To meet the technical and economical demands for those applications, supercapacitors have experienced scientific and technological breakthroughs

Supercapacitors

0.78 mm

677

Ni(OH)2 PN (z layers) MnO2 PN

0.78 mm

Ni(OH)2 PN (y layers) RGO sheet MnO2 PN Ni(OH)2 PN (x layers)

0.78 mm

0.78 nm corresponds to (003) plane of Ni(OH)2 Ni(OH)2 PN (primary nanosheet) consist of Ni(OH)2 multilayers (A) 5 mV s−1 10 mV s−1 20 mV s−1

30 mV s−1 40 mV s−1 50 mV s−1

Current density (Wh/kg)

Current density (A/g)

20 10 0 −10

100

10 Ni(OH)2-MnO2-RGO Ni(OH)2-MnO2-RGO//FRGO

−20 1 0.0 (B)

0.4

0.8 Potential (V)

1.2

200

1.0 (C)

400

600

800

1000

1200

Power density (W/kg)

Fig. 9 (A) Possible formation mechanism of Ni(OH)2–MnO2–RGO hybrid sphere and the structural schematic diagram of a single nanosheet in a hybrid sphere, (B) cyclic voltammetry (CV) curves of Ni(OH)2–MnO2–RGO//FRGO asymmetrical supercapacitors at different scan rates, (C) Ragone plot of the Ni(OH)2–MnO2–RGO hybrid electrode and the Ni(OH)2–MnO2–RGO//FRGO asymmetrical supercapacitors. Reproduced with permission from Chen H, Zhou S, Wu L. Porous nickel hydroxide-manganese dioxide-reduced graphene oxide ternary hybrid spheres as excellent supercapacitor electrode materials. ACS Appl Mater Interfaces 2014;6:8621–30.

[381–385]. To meet high power/energy requirements, commercially available supercapacitors are combined into a large capacitive systems [386]. As shown in Table 4 and Fig. 11, the supercapacitors in markets (which considerably vary in size and performance) are mainly produced by Murata Manufacturing Co., Ltd. (Japan), Maxwell Technologies, Inc. (U.S.), Panasonic Corporation (Japan), CAP-XX Limited (Australia), Nesscap Energy Inc. (Canada), Axion Power International, Inc. (U.S.), ELNA CO., LTD. (Japan), LS Mtron LTD. (Korea), and Northrop Gurmman Corporation (U.K.). The current global markets are growing explosively in order to meet the demands of efficient and environmental energy storage devices. Supercapcitors are widely applied for consumer electronics, transportation, and industrial power quality with 20% annual growth [387–390]. The market of supercapacitors would increase from $568.2 million in 2015 to $2.18 billion in 2020.

2.21.4.1

Small Consumer Electronics

Supercapacitors are widely used in the consumer electronics that require a burst of power. The features of supercapacitors, such as fast charging/discharging and high power delivery, make them appropriate for smart phones, laptops, emergency kits, toy, radio Tuners, remote power for sensors, LEDs, switches, memory-backup SRAM, etc. [391–401]. To protect memories and simply CMOS circuits, supercapacitors (that require low voltage and limited power and capacitance) generally contain a few unit cells with a total voltage below 11 V and an overall capacitance up to a Farad. Current supercapcitors cannot be used as a continuous power source, which constitutes a main drawback for supercapacitors. However, if high self-discharging rate and low energy density can be solved, the supercapacitors may replace Li-ion batteries as primary energy and power provider. Attempts are conducting in consumer electronic with novel supercapacitor technologies. Supercapcitors show great potential to resolve consumer obsession

Supercapacitors

678

Positive electrode

Collector

Negative electrode Separator Collector

Asymmetric capacitor

Graphene/MnO2

Graphene/MoO3

(A) 104

1.0

5 mV/s

E=1.9V

Power density (W/kg)

Current density (A/g)

0.5 0.0 −0.5 −1.0 −1.5

GrMoO3

−1.2 (B)

−0.8

103 GrMnO2//GrMoO3 MnO2//MoO3 102

GrMnO2

−0.4 0.0 0.4 Potential (V vs. Ag/AgCl)

0.8

100

1.2 (C)

MnO2//AC NiO//PC GrMnO2//Graphene 101 Energy density (Wh/kg)

102

Fig. 10 (A) Schematic illustration of asymmetric supercapacitors with the positive electrode of graphene/MnO2 composite and the negative electrode of graphene/MoO3 in a neutral aqueous Na2SO4 electrolyte, (B) cyclic voltammetry (CV) curves of GrMnO2 and GrMoO3 electrodes performed in a three-electrode cell in 1 M Na2SO4 electrolyte at a scan rate of 5 mV/s, (C) Ragone plot of GrMnO2//GrMoO3 asymmetric supercapacitor operated at 2.0 V in comparison with others in the literature. Reproduced with permission from Chang J, Jin M, Yao F, et al. Asymmetric supercapacitors based on graphene/MnO2 nanospheres and graphene/MoO3 nanosheets with high energy density. Adv Funct Mater 2013;23:5074–83.

that smart phones always suffer power exhaustion within several hours use after overnight charging. If smart phone manufacturers can develop efficient supercapacitors to replace batteries, electricity can maintain for a week after a few seconds of charging. Meanwhile, electricity degradation of full-charged phone can be inhibited by the supercapcitors due to their extremely long cycle life. Although supercapacitor-driven smart phones still stays on the drawings, a supercapacitor charger (Fig. 12(A)) has been produced by Zapgocharger Ltd. The replacement of aluminum foil by graphene can make supercapacitors thinner and lighter and quickly charged (within 5 min). Supercapcitors have been utilized in Lenovo laptop. New Thinkpad Yoga convertible laptops contain a stylus with a tiny capacitors (Fig. 12(B)). In a digital camera, supercapcitors combined with inexpensive alkaline batteries can fit high peak demands of microprocessor activity (disk writing, or LCD operation) and provide long cycle life with constant voltage drop (Fig. 12(C)). Furthermore, cordless hand tools using supercapacitors as energy sources provide benefits of longer lifetime, less temperature-dependent loss of performance, and a much shorter charging time (typically under 3 min).

2.21.4.2

Automotive Applications

Annual markets of high-power energy storage are predicted to grow from about $240 million currently to $2 billion in 2026, of which supercapacitors would capture about $800 million to 1 billion. It has been proposed to apply supercapacitors for vehicles and transport systems from light-duty commercial and passenger cars up to heavy-duty trucks, buses, trams, and trains [402–404]. In a vehicle, supercapcitors can provide very high power and stabilize voltage for engine starting and they can also act as a supplemental energy source.

Supercapacitors

Table 4

679

Performances of some commercial supercapacitors Specific power (W/kg) (95%)

Specific power (W/kg) matched load

Weight (kg)

Volume (L)

4.2 2.35 5.5 5.89 3.6 4.2 4.4 4.9

994 1139 5.695 2574 975 928 982 390

8,836 9,597 50,625 24,995 8,674 8,010 8,728 3,471

0.55 0.20 0.009 0.057 0.38 0.65 0.522 0.210

0.414 0.211 0.045 0.277 0.514 0.38 0.151

1.0

2.3

514

4,596

0.34

0.245

3200 2680 1350

0.25 0.20 1.5

3.7 4.2 4.9

1400 2050 650

12,400 18,225 5,785

0.63 0.5 0.21

0.47 0.572 0.151

3.3 3.3

1800 1500

3.0 1.7

8.0 6.0

486 776

4,320 6,903

0.21 0.23

0.15 0.15

3.8

1800

1.5

9.2

1025

10,375

0.232

0.143

3.8

1000 2000

4 1.9

11.2 12.1

900 1038

7,987 9,223

0.113 0.206

0.073 0.132

Device or producer

Rated voltage VR (V)

C (F)

ESR (mO)

Specific energy (Wh/kg)

Maxwell Maxwell ApowerCap ApowerCap Ness Ness Ness (cyL) Ashi Glass (propylene carbonate) Panasonic (propylene carbonate) LS Cable BatScap Power Sys. (activated argon, propylene carbonate) Power Sys. (graphite carbon, propylene carbonate) Fuji Heavy Industry-hybrid (AC/graphitic Carbon) JSR Micro (AC//graphite carbon)

2.7 2.7 2.7 2.7 2.7 2.7 2.7 2.7

2885 605 55 450 1800 3640 3160 1375

0.375 0.90 4 1.4 0.55 0.30 0.4 2.5

2.5

1200

2.8 2.7 2.7

Source: Reproduced with permission from Burke A, Miller M. In: Proceedings of 24th international electric vehicle symposium EVS-24, Stavanger; 2009. Conte M. Supercapacitors technical requirements for new applications Fuel Cells 2010;10:806–18.

2.21.4.3

Engine Starts

Supercapacitors are placed high expectation to serve start-stop systems, which will be included in more than half of vehicles in USA by 2022. The start-stop systems of some conventional cars suffer a vibration issue when an engine restarts due to the disappointed performance of Pb or Ni-Cd batteries. Supercapacitors can solve this issue by supporting the start-stop system due to their long cycle life, consistent power delivery, and large operation temperature range [405–409]. A supercapacitor can have an excellent cycle stability, which is reflected by 80% capacitance retention even after 10 years. A supercapacitor module added to batteries can start an engine at low temperature, extend the battery life, and decrease the battery size. For example, the supercapacitors applied in Mazda battery charge system can run for 9553 starts, surpassing 981 starts of battery alone (Fig. 13). Other main advantages of adding supercapacitors into the start-stop systems include the very short restarting time of an engine (less than 400 ms) and realized 5–15 þ % savings in fuel and vehicle emissions. Nowadays, more and more motor manufactures are favorable to add a supercapacitor module to their start-stop systems, such as PSA Peugeot Citroen, Lamborghini Aventador, Mazda, General Motors.

2.21.4.4

Electric and Hybrid Electric Vehicles

We are seeing the rapid market growth of hybrid electric vehicles (HEV) and electric vehicles (EV), which can greatly reduce greenhouse gas emissions. Undoubtedly, electrical energy storage components play a critical role in HEV and EV. Batteries attract most attention, but the unique functions of supercapacitors are gradually recognized [410–415]. Although some drawbacks (such as linear discharge voltage, power available for a short duration, and low energy density) render supercapacitors unsuitable as primary power for EV and HEV, they can be used as ideal temporary energy storage systems to capture and store energy from regenerative braking and provide a boost change in response to sudden power demands [416–418]. In general, supercapacitors can possess overall energy content and power varying from 50 up to 2000 Wh dependent on the configuration with or without batteries and the size of vehicles. Those applications of supercapacitors have numerous benefits [419–422]: (1) Supercapacitors can be used for charge/discharge at high rates, which can extend battery life. (2) Supercapacitors possess better charge capacity than batteries at high rates, leading to an improvement for energy efficiency. (3) Supercapacitors have low cost per power unit of electrical generators. Supercapacitors can be used for braking systems of conventional vehicles to capture kinetic energy during deceleration and to improve engine efficiency during acceleration [423–425]. Mazda claims the first production of vehicles equipped with supercapacitors for regenerative braking systems called i-Eloop (Fig. 14(A)). The main component of the supercapacitor used with iEloop is activated charcoal without heavy or precious metals, which makes it very environmental friendly.

680

Supercapacitors

(A)

(C)

(B)

(D)

(E)

(F)

(G)

(H)

Fig. 11 Commercial supercapacitors: (A) Maxwell supercapacitors, (B) ApowerCap supercapacitors, (C) Ness supercapacitors, (D) Panasonic supercapacitors, (E) LS Cable supercapacitors, (F) Batscap supercapacitors, (G) Fuji supercapacitors, and (F) JSR Micro supercapacitors.

Hybrid electric vehicles with small battery packs need supercapacitors as a power buffer to capture brake energy and to return it as a power boost for acceleration, leading to a small engine and substantial fuel savings with drivetrain batteries [426–446]. Toyota has demonstrated that the high-rate and high-power performance of supercapacitors can benefit acceleration and braking. By adding a quick-charging capacitor system to replace the regular rechargeable battery unit, Supra HVR hybrid race car operated a repeated acceleration and deceleration under full system performance and won the Tokashi 24-Hour Race in 2007 (Fig. 14(B)). After the competition, Toyota manufactured a first hybrid vehicle, Lexus GS450h, which finished the 17th overall in Tokashi 24Hour race.

Supercapacitors

(A)

(C)

(B)

(D)

Fig. 12 (A) A graphene supercapacitor charger for mobiles and tablets (Zap &Go), (B) a stylus with a tiny capacitor, (C) supercapacitor-driven double-lens car camera, (D) supercapacitor-driven cordless screwdriver.

Fig. 13 Supercapacitor module built in battery.

681

682

Supercapacitors

(A)

(B)

Fig. 14 (A) Supercapacitor based regenerative braking systems of Mazda, (B) supra HRV hybrid race car with a quick charging capacitor systems.

(A)

(B)

Fig. 15 Supercapacitors applied in: (A) wind grid and (B) solar grid.

To obtain fast charging and starting at low temperature, supercapacitors are applied for electric vehicles with larger battery packs [447–455]. Tesla chose Li-ion batteries before supercapacitors meets the requirements. If 50% power (45 kWh) of a Tesla Model S90D is supplied by supercapacitors and their supercharger stations can deliver 1000A at 550V (550kW), it can fully charge the 45 kWh supercapacitor pack in about 5 min. Therefore, an EV can be fully re-charged in a time shorter than filling a vehicle with a tank of gasoline.

2.21.4.5

Electric Utility Applications

High-power feature makes supercapacitors attract increasing interest in electric utilities, especially for improvement of performance and reliability [456]. For example, 350 FD cell supercapacitors (Fig. 15(A)) were used in emergency power supply systems of medium size wind turbines (250 kW–2 MW). By providing emergency power backup and converting peak power to energy for blade pitching motors, the efficiency of a wind turbine can be increased and the safety of blade pitching can be enhanced by reducing blading damage associated with high wind speeds [457,458]. It is promising to combine solar power and energy storage [456,459–467]. A project at the University of California San Diego’s cutting-edge microgrid, funded with $1.39 million by California Energy Commission grant, tests the costs and benefits of combining the ultracapacitors from Maxwell Technologies and the concentrating photovoltaic (CPV) solar systems from Soitec (Fig. 15(B)). In the Soitec project, a 2.5-kW-hour ultracapacitor array was equipped to a 22-kilowatt CPV system which can add up to roughly five minutes of storage capacity. The supercapacitors, which are faster with discharging in fractions of a second, can perform over a broader temperature range ( 401C to þ 651C).

Supercapacitors

683

Supercapacitors can also provide a perfect solution to solve grid instability for renewable energy generation [468–478]. The instability origins from intermittencies of these sources and around 60 to 70% of the irregularities occurs in less than a minute. The rapid charge/discharge feature without performance degradation of supercapacitors can be exploited to boost power in the short intermittencies. This technology makes back-up generation obsolete, especially in regions (such as Denmark, Ireland, Puerto Rico, and Hawaii) with a high proportion of renewables for energy generation. Furthermore, Maxwell Technologies is developing hybrid systems to provide both short- and long-term energy solutions by pairing supercapacitors with batteries [479,480]. Governments in numerous countries are making great efforts to reduce carbon dioxide emissions, such as replacing aging infrastructure, lowering fossil fuel consumption, and developing efficient renewable energy generators with supercapacitors-based stabilizing systems [481–493].

2.21.4.6

Industrial Application

Energy systems with supercapacitors are suitable for power recovery and high power bursts [494–498]. The new Airbus A380 incorporates supercapacitors in emergency actuation of 16 doors and Slide Management System (DSMS) actuators to operate the aircraft’s heavy doors [499–512]. The similar devices could also be employed to meet peak energy demands on satellites [513–520]. Although supercapacitors cannot totally substitute small rechargeable batteries (such as nickel–cadmium batteries with specific energy of 50 Wh/kg or more advanced lithium-ion batteries with energy density of 150 Wh/kg) to provide energy for small satellites with weigh between 100 and 200 kg, they can deliver high bursts in seconds and keep stable performance during 15 years life of satellites. Uninterruptible power supplies (UPS) is another representative application, in which supercapacitors can provide the short term power back up [521–527]. Furthermore, supercapacitors are employed to provide lift for electric forklifts [528–532]. In addition, the energy storage in the harbor crane and construction/mining systems needs rapid discharge with high duty requirements [533,534]. Supercapacitors integrated in their equipment can deliver up to 20% savings on diesel consumption by both braking/drop energy recovery. If peak power can be fully supplied by supercapacitors, further fuel savings can be obtained, downsizing diesel engines. As a result, CO2 emission can be reduced by 35% for the harbor.

2.21.5

Conclusions

Supercapacitors, which are one type of energy storage devices, can be cataloged into three types based on different operation mechanisms, namely, electric double-layer capacitors, pseudocapacitors and hybrid capacitors. A typical electrochemical supercapacitor consists of two electroactive electrodes, a separator, current collectors, and electrolytes. For EDLCs, an ideal electrode material requires a material with large electrolyte-ions-accessible surface area, fast ion transportation, and good electrical conductivity. However, activated carbon, which is used as electrode material for commercial EDLCs, is not efficient due to its rich micropores. Thus, new types of carbon materials, such as carbon aerogels, carbon nanofibers, carbon nanotubes, and graphene, were developed for EDLCs with higher capacitances. To further enhance capacitance, pseudocapacitive materials (transition metal oxide and conducting polymer) were introduced to carbon matrix. Furthermore, to obtain a large potential window, a high energy density and power density, and a long cycle life, hybrid supercapacitors were developed with a battery-type Faradaic electrode as the energy source and a capacitor-type electrodes as the power source. Nowadays, most of commercialized supercapacitors are EDLCs as a supplementary to batteries for high power output or rapid charge/discharge process. Its applications range from custom electronics, automotives, and industrial and grid fields.

2.21.6

Future Direction

Supercapacitors possess fast charge/discharge and high power density, leading to their applications as supplementary energy sources in consumer electronics, electric vehicles and hybrid electric vehicles, electric utilities, and industrial fields. However, it is very difficult to use supercapacitors as a primary power source due to their rapid self-discharge and low energy densities. To solve those issues, innovations are necessary for configurations, electrode materials, and electrolytes of supercapacitors. Furthermore, it would be attractive to develop efficient battery-supercapacitor hybrid systems, which can combine the advantages of fast charge/ discharge and high power density of supercapacitors with the features of high energy density and low self-discharge of batteries.

Acknowledgment This work was partially supported by U.S. National Science Foundation (CBET-0931587 and CMMI-1661699). Authors also thank Charles and Carroll McArthur for their great support.

References [1] Miller JR, Simon P. Electrochemical capacitors for energy management. Science 2008;321:651–2. [2] Mai L, Tian X, Xu X, Chang L, Xu L. Nanowire electrodes for electrochemical energy storage devices. Chem Rev 2014;114:11828–62.

684

[3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52]

Supercapacitors

Pandolfo AG, Hollenkamp AF. Carbon properties and their role in supercapacitors. J Power Sources 2006;157:11–27. Liu C, Li F, Ma LP, Cheng HM. Advanced materials for energy storage. Adv Energy Mater 2010;22:E28–62. Chang L, Mai L, Xu X, et al. Pore-controlled synthesis of Mn2O3 microspheres for ultralong-life lithium storage electrode. RSC Adv 2013;3:1947–52. Zhang Y, Tao B, Xing W, Zhang L, Xue Q, Yan Z. Sandwich-like nitrogen-doped porous carbon/graphene nanoflakes with high-rate capacitive performance. Nanoscale 2016;8:7889–98. Pham DT, Lee TH, Luong DH, et al. Carbon nanotube-bridged graphene 3D building blocks for ultrafast compact supercapacitors. ACS Nano 2015;9:2018–27. He Y, Chen W, Li X, et al. Free-standing three-dimensional graphene/MnO2 composite networks as ultralight and flexible supercapacitor electrodes. ACS Nano 2013;7:174–87. Roberts AJ, Danll de Namor AF, Slade RCT. Low temperature water based electrolytes for MnO2/carbon supercapacitors. Phys Chem Chem Phys 2013;15:3518–26. Simon P, Gogotsi Y, Dunn B. Where do batteries end and supercapacitors begin? Science 2014;343:1210–1. Zhou G, Li F, Cheng HM. Progress in flexible lithium batteries and future prospects. Energy Environ Sci 2014;7:1307–38. Nishide H, Oyzaizu K. Toward flexible batteries. Science 2008;319:737–8. Wang DW, Zeng Q, Zhou G, et al. Carbon–sulfur composites for Li-S batteries: status and prospects. J Mater Chem A 2013;1:9382–94. Zhang C, Lv W, Tao Y, Yang CH. Towards superior volumetric performance: design and preparation of novel carbon materials for energy storage. Energy Environ Sci 2015;8:1390–405. Reddy ALM, Gowda SR, Shaijumon MM, Ajayan PM. Hybrid nanostructured for energy storage application. Adv Mater 2012;24:5045–64. Wang W, Liu W, Zeng Y, et al. A novel exfoliation strategy to significantly boost the energy storage capability of commercial carbon cloth. Adv Mater 2015;27:3572–8. Xie JL, Guo CX, Li CM. Construction of one-dimensional nanostructure on graphene for efficient energy conversion and storage. Energy Environ Sci 2014;7:2559–79. Liu C, Yu Z, Neff D, Zhamu A, Jang BZ. Graphene-based supercapacitor with an ultrahigh energy density. Nano Lett 2010;10:4863–8. Urita K, Ide N, Isobe F, Furukawa H, Moriguchi I. Enhanced electric double-layer capacitance by desolvation of lithium ions in confined nanospaces of microporous carbon. ACS Nano 2014;8:3614–9. Orita A, Kamijima K, Yoshida M. Allyl-functionalized ionic liquids as electrolytes for electric double-layer capacitors. J Power Sources 2010;195:7471–9. Yoo HD, Jang JH, Ryu JH, Park Y, Oh SM. Impedance analysis of porous carbon electrodes to predict rate capability of electric double-layer capacitors. J Power Sources 2014;267:411–20. Fang B, Binder L. A novel carbon electrode material for highly improved EDLC performance. J Phys Chem B 2006;110:7877–82. Li Q, Jiang R, Dou Y, et al. Synthesis of mesoporous carbon spheres with a hierarchical pore structure for the electrochemical double-layer capacitor. Carbon 2011;49:1248–57. Cai M, Outlaw RA, Quinlan RA, Premathilake D, Butler SM, Miller JR. Fast response, vertically oriented graphene nanosheet electric double layer capacitors synthesized from C2H2. ACS Nano 2014;8:5873–82. Wen S, Mho S, Yeo IH. Improved electrochemical capacitive characteristic of the carbon nanotubes grown on the alumina templates with high pore density. J Power Sources 2006;163:304–8. Chang L, Hu YH. Excellent capacitive deionization performance of meso-carbon microbeads. RSC Adv 2016;6:47285–91. Schutter C, Husch T, Viswanathan V, Passerini S, Balducci A, Korth M. Rational design of new electrolyte materials for electrochemical double layer capacitors. J Power Sources 2016;326:541–8. Lei C, Amini N, Markoulidis F, Wilson P, Tennison S, Lekakou C. Activated carbon from phenolic resin with controlled mesoporosity for an electric double-layer capacitor (EDLC). J Mater Chem A 2013;1:6037–42. Chen K, Yang Y, Li K, Ma Z, Zhou Y, Xue D. CoCl2 designed as excellent pseudocapacitor electrode materials. ACS Sustain Chem Eng 2014;2:440–4. Chen K, Xue D, Komameni S. Colloidal pseudocapacitor: nanoscale aggregation of Mn colloids from MnCl2 under alkaline condition. J Power Sources 2015;279:365–71. Jiang Y, Wang P, Zang X, Yang Y, Kozinda A, Lin L. Uniformly embedded metal oxide nanoparticles in vertically aligned carbon nanotube forests as pseudocapacitor electrodes for enhanced energy storage. Nano Lett 2013;13:3524–30. Zhang X, Shi W, Zhu J, et al. High-power and high-energy-density flexible pseudocapacitor electrodes made from porous CuO nanobelts and single-walled carbon nanotubes. ACS Nano 2011;5:2013–9. Sahoo R, Sasmal AK, Ray C, Dutta S, Pal A, Pal T. Suitable morphology makes CoSn(OH)6 nanostructure a superior electrochemical pseudocapacitor. ACS Appl Mater Interfaces 2016;8:17987–98. Hu Z, Xiao X, Chen C, et al. Al-doped a-MnO2 for high mass-loading pseudocapacitor with excellent cycling stability. Nano Energy 2015;11:226–34. Liu T, Finn L, Yu M, et al. Polyaniline and polypyrrole pseudocapacitor electrodes with excellent cycling stability. Nano Lett 2014;14:2522–7. Yamazaki S, Ito T, Murakumo Y, et al. Hybrid capacitors utilizing halogen-based redox reactions at interface between carbon positive electrode and aqueous electrolytes. J Power Sources 2016;326:580–6. Aida T, Murayama I, Yamada K, Morita M. High-energy-density hybrid electrochemical capacitor using graphitizable carbon activated with KOH for positive electrode. J Power Sources 2007;166:462–70. Haspert LC, Gillette E, Lee SB, Rubioff GW. Perspective: hybrid systems combining electrostatic and electrochemical nanostructures for ultrahigh power energy storage. Energy Environ Sci 2013;6:2578–90. Ding J, Li Z, Cui K, Boyer S, Karpuzov D, Mitlin D. Heteroatom enhanced sodium ion capacity and rate capability in a hydrogel derived carbon give record performance in a hybrid ion capacitor. Nano Energy 2016;23:129–37. Terasawa N, Asaka K. High-performance hybrid (electrostatic double-layer and faradaic capacitor-based) polymer actuators incorporating nickel oxide and vapor-grown carbon nanofibers. Langmuir 2014;30:14343–51. Meng G, Yang Q, Wu X, et al. Hierarchical mesoporous NiO nanoarrays with ultrahigh capacitance for aqueous hybrid supercapacitor. Nano Energy 2016;30:831–9. Wang X, Yan C, Yan J, Sumboja A, Lee PS. Orthorhombic niobium oxide nanowires for next generation hybrid supercapacitor device. Nano Energy 2015;11:765–72. Wang R, Lang J, Zhang P, Lin Z, Yan X. Fast and large lithium storage in 3D porous VN nanowires-graphene composite as a superior anode toward high-performance hybrid supercapacitors. Adv Funct Mater 2015;25:2270–8. Li B, Dai F, Xiao Q, et al. Nitrogen-doped activated carbon for a high energy hybrid supercapacitor. Energy Environ Sci 2016;9:102–6. Zhang F, Zhang T, Yang X, et al. A high-performance supercapacitor-battery hybrid energy storage device base on graphene-enhanced electrode materials with ultrahigh energy density. Energy Environ Sci 2013;6:1623–32. Perera SD, Rudolph M, Mariano RG, et al. Manganese oxide nanorod–graphene/vanadium oxide nanowire–graphene binder-free paper electrodes for metal oxide hybrid supercapacitors. Nano Energy 2013;2:966–75. Rightmire RA. Electrical energy storage apparatus, United States Patent 3288641, 1966. Yu G, Xie X, Pan L, Bao Z, Cui Y. Hybrid nanostructured materials for high-performance electrochemical capacitors. Nano Energy 2013;2:213–34. Xu Y, Chang L, Hu YH. KOH-assisted microwave post-treatment of activated carbon for efficient symmetrical double-layer capacitors. Int J Energy Res 2017;41:728–35. Wang G, Zhang L, Zhang J. A review of electrode materials for electrochemical supercapacitors. Chem Soc Rev 2012;41:797–828. Hou Y, Cheng Y, Hobson T, Liu J. Design and synthesis of hierarchical MnO2 nanospheres/carbon nanotubes/conducting polymer ternary composite for high performance electrochemical electrodes. Nano Lett 2010;10:2727–33. Zhang K, Zhang LL, Zhao XS, Wu J. Graphene/polyaniline nanofiber composites as supercapacitor electrodes. Chem Mater 2010;22:1392–401.

Supercapacitors

685

[53] Banerjee A, Bhatnagar S, Upadhyay KK, Yadav P, Ogale S. Hollow Co0.85Se nanowire array on carbon fiber paper for high rate pseudocapacitor. ACS Appl Mater Interfaces 2014;6:18844–52. [54] Hu L, Chen W, Xie X, et al. Symmetrical MnO2-carbon nanotube-textile nanostructures for wearable pseudocapacitors with high mass loading. ACS Nano 2011;3:8904–13. [55] Zhang C, Xie Y, Zhao M, et al. Enhanced electrochemical performance of hydrous RuO2/mesoporous carbon nanocomposites via nitrogen doping. ACS Appl Mater Interfaces 2014;6:9751–9. [56] Yang L, Cheng S, Ding Y, Zhu X, Wang ZL, Liu M. Hierarchical network architectures of carbon fiber paper supported cobalt oxide nanonet for high-capacity pseudocapacitors. Nano Lett 2012;12:321–5. [57] Huang CJ, Lin P, Tseng TY. Electrophoretic fabrication and pseudocapacitive properties of graphene/manganese oxide/carbon nanotube nanocomposites. J Power Sources 2013;243:594–602. [58] Xiao X, Peng Z, Chen C, et al. Freestanding MoO3-x nanobelt/carbon nanotube films for Li-ion intercalation pseudocapacitors. Nano Energy 2014;9:355–63. [59] Huang ZH, Song Y, Xu XX, Liu XX. Ordered polypyrrole nanowire arrays grown on a carbon cloth substrate for a high-performance pseudocapacitor electrode. ACS Appl Mater Interfaces 2015;7:25506–13. [60] Zhu YG, Wang Y, Shi Y, Huang ZX, Fu L, Yang HY. Phase transformation induced capacitance activation for 3D graphene-CoO nanorod pseudocapacitor. Adv Energy Mater 2014;4:1301788. [61] Yang J, Yu C, Fan X, Ling Z, Qiu J, Gogotsi Y. Facile fabrication of MWCNT-doped NiCoAl-layered double hydroxide nanosheets with enhanced electrochemical performances. J Mater Chem A 2013;1:1963–8. [62] Jiang J, Liu J, Ding R, et al. Large-scale uniform a-Co(OH)2 long nanowire arrays grown on graphite as pseudocapacitor electrodes. ACS Appl Mater Interfaces 2011;3:99–103. [63] Brownson DAC, Kampouris DK, Banks CE. An overview of graphene in energy production and storage application. J. Power Sources 2011;196:4873–85. [64] Ma Y, Chang H, Zhang M, Chen Y. Graphene-based materials for lithium-ion hybrid supercapacitors. Adv Mater 2015;27:5296–308. [65] Sharma P, Bhatti TS. A review on electrochemical double-layer capacitors. Energy Convers Manage 2010;51:2901–12. [66] Khavrus VO, Weiser M, Fritsch M, et al. Application of carbon nanotubes directly grown on aluminum foils as electric double layer capacitor electrodes. Chem Vap Deposition 2012;18:53–60. [67] Huang CW, Wu CA, Hou SS, Kuo PL, Hsieh CT, Teng H. Gel electrolyte derived from poly(ethylene glycol) blending poly(acrylonitrile) applicable to roll-to-roll assembly of electric double layer capacitors. Adv Funct Mater 2012;22:4677–85. [68] Izumi H, Suda Y, Okabe Y, et al. Manufacturing of electric double-layer capacitors using carbon nanocoils and evaluation of their specific capacitances at a high scan rate. Electron Commun Japan 2016;99:660–7. [69] Frackowiak E. Carbon materials for supercapacitor application. Phys Chem Chem Phys 2007;9:1774–85. [70] Huang J, Sumpter BG, Meunier V. A universal model for nanoporous carbon supercapacitors applicable to diverse pore regimes, carbon materials, and electrolytes. Chem Eur J 2008;14:6614–26. [71] Miller JR, Outlaw RA, Holloway BC. Graphene electric double layer capacitor with ultra-high-power performance. Electrochim Acta 2011;56:10443–9. [72] Kannan AG, Samuthirapandian A, Kim DW. Electric double layer capacitors employing nitrogen and sulfur co-doped, hierarchically porous graphene electrodes with synergistically enhanced performance. J. Power Sources 2017;337:65–72. [73] Lu X, Shen J, Ma H, et al. A cost-effective way to maintain metal-doped carbon xerogels and their applications on electric double-layer capacitors. J Power Sources 2012;201:340–6. [74] Zhang Y, Feng H, Wu X, et al. Progress of electrochemical capacitor electrode materials: a review. Int J Hydrogen Energy 2009;34:4889–99. [75] Huang Y, Liang J, Chen Y. An overview of the applications of graphene-based materials in supercapacitors. Small 2012;8:1805–34. [76] Rauda IE, Augustyn V, Saldarriaga-Lopez LC, et al. Nanostructured pseudocapacitors based on atomic layer deposition of V2O5 onto conductive nanocrystal-based mesoporous ITO scaffolds. Adv Funct Mater 2014;24:6717–28. [77] Shao Y, Zhao Y, Li H, Xu C. Three-dimensional hierarchical NixCo1-xO/NiyCo2-yP@C hybrids on nickel foam for excellent supercapacitors. ACS Appl Mater Interfaces 2016;8:35368–76. [78] Liu R, Ma L, Huang S, Mei J, Li E, Yuan G. Large areal mass and high scalable and flexible cobalt oxide/graphene/bacterial cellulose electrode for supercapacitors. J Phys Chem C 2016;120:28480–8. [79] Zhou J, Fang G, Pan A, Liang S. Oxygen-incorporated MoS2 nanosheets with expanded interlayers for hydrogen evolution reaction pseudocapacitor applications ACS. Appl Mater Interfaces 2016;6:33681–9. [80] Cui L, Huang L, Ji M, et al. High-performance MgCo2O4 nanocone arrays grown on three-dimensional nickel foams: preparation and application as binder-free electrode for pseudo-supercapacitor. J Power Sources 2016;333:118–24. [81] Wang K, Zhang X, Zhagn X, Chen D, Lin Q. A novel Ni(OH)2/graphene nanosheets electrode with high capacitance and excellent cycling stability for pseudocapacitors. J. Power Sources 2016;333:156–63. [82] Wang Y, Chai H, Dong H, Xu J, Jie D, Zhou W. Superior cycle stability performance of quasi-cuboidal CoV2O6 microstructures as electrode material for supercapacitors. ACS Appl Mater Interfaces 2016;8:27291–7. [83] Zhi M, Xiang C, Li J, Li M, Wu N. Nanostructured carbon-metal oxide composite electrodes for supercapacitors: a review. Nanoscale 2013;5:72–88. [84] Toupin M, Brousse T, Belanger D. Charge storage mechanism of MnO2 electrode used in aqueous electrochemical capacitor. Chem Mater 2004;16:3184–90. [85] Fan Z, Yan J, Wei T, Zhi L, Ning G, Li T, et al. Asymmetric supercapacitors based on graphene/MnO2 and activated carbon nanofiber electrodes with high power and energy density. Adv Funct Mater 2011;21:2366–75. [86] Amatucci GG, Badway F, Pasquier AD, Zheng T. An asymmetric hybrid nonaqueous energy storage cell. J Elecrochem Soc 2001;148:A930–9. [87] Tang Z, Tang CH, Gong H. A high energy density asymmetric supercapacitor from nano-architectured Ni(OH)2/carbon nanotube electrodes. Adv Funct Mater 2012;22:1272–8. [88] Chang B, Yin H, Zhang X, Zhang S, Yang B. Chemical blowing strategy synthesis of nitrogen-rich porous graphitized carbon nanosheets: morphology, pore structure and supercapacitor application. Chem Eng J 2017;312:191–203. [89] Zhang SW, Yin BS, Liu C, Wang ZB, Gu DM. Self-assembling hierarchical NiCo2O4/MnO2 nanosheets and MoO3/PPy core-shell heterostructured nanobelts for supercapacitor. Chem Eng J 2017;312:295–305. [90] Chen H, Xiong Y, Yu T, et al. Boron and nitrogen co-doped porous carbon with a high concentration of boron and its superior capacitive behavior. Carbon 2017;113:266–73. [91] Yan J, Fan Z, Wei T, Qian W, Zhang M, Wei F. Fast and reversible surface redox reaction of graphene-MnO2 composites as supercapacitor electrodes. Carbon 2010;48:3825–33. [92] Xu J, Wang K, Zu SZ, Han BH, Wei Z. Hierarchical nanocomposites of polyaniline nanowire arrays on graphene oxide sheets with synergistic effect for energy storage. ACS Nano 2010;4:5019–26. [93] Subramanian V, Zhu H, Vajtai R, Ajayan PM, Wei B. Hydrothermal synthesis and pseudocapacitance properties of MnO2 nanostructures. J Phys Chem B 2005;109:20207–14. [94] Wei TY, Chen CH, Chien HC, Lu SY, Hu CC. A cost-effective supercapacitor material of ultrahigh specific capacitances: spinel nickel cobaltite aerogels from an epoxidedriven sol–gel process. Adv Mater 2010;22:347–51.

686

Supercapacitors

[95] Toupin M, Brousse T, Belanger D. Influence of microstructure on the charge storage properties of chemically synthesized manganese dioxide. Chem Mater 2002;14:3946–52. [96] Chen LF, Zhang XD, Liang HW, et al. Synthesis of nitrogen-doped porous carbon nanofibers as an efficient electrode materials for supercapacitors. ACS Nano 2012;6:7092–102. [97] Devaraj S, Munichandraish N. Effect of crystallographic structure of MnO2 on its electrochemical capacitance properties. J Phys Chem C 2008;112:4406–17. [98] Chen Z, Dinh H, Miller E. Photoelectrochemical water splitting – standards, experimental method and protocols. New York: Springer; 2013. http://dx.doi.org/10.1007/ 987-1-4614-8298-7 [99] Bratsch SQ. Standard electrode potentials and temperature coefficients in water at 298.15 K. J Phys Chem Ref Data 1989;18:1–21. [100] Qiu Y, Zhang X, Yang S. High performance supercapacitors based on highly conductive nitrogen-doped graphene sheets. Phys Chem Chem Phys 2011;13:12554–8. [101] Cheng Q, Tang J, Ma J, Zhang H, Shinya N, Qin LC. Graphene and carbon nanotube composite electrodes for supercapacitors with ultra-high energy density. Phys Chem Chem Phys 2011;13:17615–24. [102] Zheng C, Zhou XF, Cao HL, Wang GH, Liu ZP. Edge-enriched porous graphene nanoribbons for high energy density supercapacitors. J Mater Chem A 2014;2:7484–90. [103] Yuan K, Xu Y, Uihlein J, et al. Straightforward generation of pillared, microporous graphene frameworks for use in supercapacitors. Adv Mater 2015;27:6714–21. [104] Xu J, Tan Z, Zeng W, et al. A hierarchical carbon derived from sponge-templated activation of graphene oxide for high-performance supercapacitor electrodes. Adv Mater 2016;28:5222–8. [105] Frackowiak E, Khomenko V, Jurewicz K, Lota K, Beguin F. Supercapacitor based on conducting polymers/nanotubes composites. J Power Sources 2006;153:413–8. [106] Yoo JJ, Balakrishnan K, Huang J, et al. Ultrathin planar graphene supercapacitors. Nano Lett 2011;11:1423–7. [107] Yuan C, Zhang X, Su L, Gao B, Shen L. Facile synthesis and self-assembly of hierarchical porous NiO nano/micro spherical superstructures for high performance supercapacitors. J Mater Chem 2009;19:5772–7. [108] Zhou C, Zhang Y, Li Y, Liu J. Construction of high-capacitance 3D CoO@polypyrrole nanowire array electrode for aqueous asymmetric supercapacitor. Nano Lett 2013;13:2078–85. [109] Khomenko V, Raymundo-Piriero E, Beguin F. Optimisation of an asymmetric manganese oxide/activated carbon capacitor working at 2 V in aqueous medium. J Power Sources 2006;153:183–90. [110] Ji J, Zhang LL, Ji H, et al. Nanoporous Ni(OH)2 thin film on 3D ultrathin-graphite foam for asymmetric supercapacitor. ACS Nano 2013;7:6237–43. [111] Yuan C, Yang L, Hou L, Shen L, Zhang X, Lou XW. Growth of ultrathin mesoporous Co3O4 nanosheet arrays on Ni foam for high-performance electrochemical capacitors. Energy Environ Sci 2012;5:7883–7. [112] Cherusseri J, Kar KK. Hierarchically mesoporous carbon nanopetal based electrodes for flexible supercapacitors with super-long cyclic stability. J Mater Chem A 2015;3:21586–98. [113] Wang G, Wang H, Lu X, et al. Solid-state supercapacitor based on activated carbon cloths exhibits excellent rate capability. Adv Mater 2014;26:2676–82. [114] Yang M, Zhong Y, Bao J, Zhou X, Wei J, Zhou Z. Achieving battery-level energy density by constructing aqueous carbonaceous supercapacitors with hierarchical porous N-rich carbon materials. J Mater Chem A 2015;3:11387–94. [115] Liu F, Song S, Xue D, Zhang H. Folded structured graphene paper for high performance electrode materials. Adv Mater 2012;24:1089–94. [116] Cheng S, Shi T, Huang Y, et al. Rational design of nickel cobalt sulfide/oxide core-shell nanocolumn arrays for high-performance flexible all-solid-state asymmetric supercapacitors. Ceram Int 2017;43:2155–64. [117] Yu P, Zhao X, Li Y, Zhang Q. Controllable growth of polyaniline nanowire arrays on hierarchical macro/mesoporous graphene foams for high-performance flexible supercapacitors. Appl Surf Sci 2017;393:37–45. [118] Stepriak I, Ciszewski A. New design of electric double layer capacitors with aqueous LiOH electrolyte as alternative to capacitor with KOH solution. J Power Sources 2010;195:2564–9. [119] Pell WG, Conway BE. Analysis of power limitations at porous supercapacitor electrodes under cyclic voltammetry modulation and dc charge. J Power Sources 2001;96:57–67. [120] Ding KQ. Direct preparation of metal ions-doped manganese oxide by cyclic voltammetry. J Chin Chem Soc 2008;55:543–9. [121] Beidaghi M, Chen W, Wang C. Electrochemically activated carbon micro-electrode arrays for electrochemical micro-capacitors. J Power Sources 2011;196:2403–9. [122] Karthikeyan K, Amaresh S, Lee SN, An JY, Lee YS. High-power lithium-ion capacitor using LiMnBO3-nanobead anode and polyaniline-nanofiber cathode with excellent cycle life. ChemSusChem 2014;7:2310–6. [123] Stimpfling T, Leroux F. Supercapacitor-type behavior of carbon composite and replica obtained from hybrid layered double hydroxide active container. Chem Mater 2010;22:974–87. [124] Zhang X, Wang W, Lu J, Hua L, Heng J. Reversible heat of electric double-layer capacitors during galvanostatic charging and discharging cycles. Thermochim Acta 2016;636:1–10. [125] Randriamahazaka H, Plesse C, Teyssie D, Chevrot C. Charging/discharging kinetics of poly(3,4-ethylenedioxythiophene) in 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide ionic liquid under galvanostatic conditions. Electrochim Acta 2005;50:4222–9. [126] Sun GH, Li KX, Sun CG. Application of 1-ethyl-3-methylimidazolium thiocyanate to the electrolyte of electrochemical double layer capacitors. J. Power Sources 2006;162:1444–50. [127] Rauhala T, Leis J, Kallio T, Vuorilehto K. Lithium-ion capacitors using carbide-derived carbon as the positive electrode – a comparison of cells with graphite and Li4Ti5O12 as the negative electrode. J Power Sources 2016;331:156–66. [128] Sipahi M, Parlak EA, Gui A, Ekinci E, Yardim MF, Sarac AS. Electrochemical impedance study of polyaniline electrocoated porous carbon foam. Prog Org Coat 2008;62:96–104. [129] Kotz R, Hahn M, Gallay R. Temperature behavior and impedance fundamentals of supercapacitors. J Power Sources 2006;154:550–5. [130] Liu X, Wang Y, Zhan L, Qiao W, Liang X, Ling L. Effect of oxygen-containing functional groups on the impedance behavior of activated carbon-based electric doublelayer capacitors. J Solid State Electrochem 2011;15:413–9. [131] Liu X, Juan L, Zhan L, et al. Effect of conductive filler on the impedance behaviors of activated carbon based electric double layer capacitors. J Electroanal Chem 2010;642:75–81. [132] Itagaki M, Suzuki S, Shitanda I, Watanaba K, Nakazawa H. Impedance analysis on electric double layer capacitor with transmission line model. J Power Sources 2007;164:415–24. [133] Tripathy MC, Mondal D, Biswas K, Sen S. Experimental studies on realization of fractional inductors and fractional-order bandpass filters. Int J Circ Theor Appl 2015;43:1183–96. [134] Wu TH, Hsu CT, Hu CC, Hardwick LJ. Important parameters affecting the cell voltage of aqueous electrical double-layer capacitors. J Power Sources 2013;242:289–98. [135] Momma T, Matsunaga M, Mukoyama D, Osaka T. Ac impedance analysis of lithium ion battery under temperature control. J Power Sources 2012;216:304–7. [136] Iroh JO, Levine K. Capacitance of the polypyrrole/polyimide composite by electrochemical impedance spectroscopy. J Power Sources 2003;117:267–72. [137] Roberts AJ, Slade RCT. Performance loss of aqueous MnO2/carbon supercapacitors at elevated temperature: cycling vs. storage. J Mater Chem A 2013;1:1140–14146. [138] Yoon Y, Lee K, Baik C. et al. Anti-solvent derived non-stacked reduced graphene oxide for high performance supercapacitors. Adv Mater 2013;25:4437–44. [139] Zhai Y, Dou Y, Zhao D, Fulvio PF, Mayes RT. Carbon materials for chemical capacitive energy storage. Adv Mater 2011;23:4828–50. [140] Zheng JP, Jow TR. High energy and high power density electrochemical capacitors. J Power Sources 1996;52:155–9.

Supercapacitors

687

[141] Tomko T, Rajagopalan R, Lanagan M, Foley HC. High energy density capacitor using coal tar pitch derived nanoporous carbon/MnO2 electrodes in aqueous electrolytes. J Power Sources 2011;196:2380–6. [142] Ren JJ, Su LW, Qin X, et al. Pre-lithiated graphene nanosheets as negative electrode materials for Li-ion capacitors with high power and energy density. J Power Sources 2014;264:108–13. [143] Du H, Yang H, Huang C, He J, Liu H, Li Y. Graphdiyne applied for lithium-ion capacitors displaying high power and energy densities. Nano Energy 2016;22:615–22. [144] Ahn CW, Amarsanaa G, Won SS, Chae SA, Lee III DS, Kim W. Antiferroelectric thin-film capacitors with high energy-storage densities, low energy losses, and fast discharge times. ACS Appl Mater 2015;7:26381–6. [145] Nohara S, Asahina T, Wada H, et al. Hybrid capacitor with activated carbon electrode, Ni(OH)2 electrode and polymer hydrogel electrolyte. J. Power Sources 2006;157:605–9. [146] Yu X, Zhan C, Lv R, et al. Ultrahigh-rate and high-density lithium-ion capacitors through hybriding nitrogen-enriched hierarchical porous carbon cathode with prelithiated microcrystalline graphite anode. Nano Energy 2015;15:43–53. [147] Ferrero GA, Fuertes AB, Sevilla M. N-doped porous carbon capsules with tunable porosity for high-performance supercapacitors. J Mater Chem A 2015;3:2914–23. [148] Antiohos D, Romano MS, Razal JM, et al. Performance enhancement of single-walled nanotube-microwave exfoliated graphene oxide composite electrodes using a stacked electrode configuration. J Mater Chem A 2014;2:14835–43. [149] Wang H, Xu Z, Kohandehgham A, et al. Interconnected carbon nanosheets derived from hemp for ultrafast supercapacitors with high energy. ACS Nano 2013;7:5131–41. [150] Jung N, Kwon S, Lee D, et al. Synthesis of chemically bonded graphene/carbon nanotube composites and their application in large volumetric capacitance supercapacitors. Adv Mater 2013;25:6854–8. [151] Zhang LL, Zhao XS. Carbon-based materials as supercapacitor electrodes. Chem Soc Rev 2009;38:2520–31. [152] Qie L, Chen W, Xu H, et al. Synthesis of functionalized 3D hierarchical porous carbon for high-performance supercapacitors. Energy Environ Sci 2013;6:2497–504. [153] He X, Zhang H, Zhang. H, Li X, Xiao N, Qiu J. Direct synthesis of 3D hollow porous graphene balls from coal tar pitch for high performance supercapacitors. J Mater Chem A 2014;2:19633–40. [154] Dong L, Xu C, Yang Q, Fang J, Li Y, Kang F. High-performance compressible supercapacitors based on functionally synergic multiscale carbon composite textiles. J Mater Chem A 2015;3:4729–37. [155] Xu Y, Tao Y, Zheng X, et al. A metal-free supercapacitor electrode materials with a record high volumetric capacitance over 800 F cm-3. Adv Mater 2015;27:8082–7. [156] Liu D, Jia Z, Zhu J, Wang D. A regular, compact but microporous packing structure: high-density graphene assemblies for high-volumetric-performance supercapacitors. J Mater Chem A 2015;3:12653–62. [157] Zhang S, Pan N. Supercapacitors performance evaluation. Adv Energy Mater 2015;5:1401401. [158] Kim ND, Buchholz DB, Casillas G, Jose-Yacaman M, Chang RPH. Hierarchical design for fabricating cost-effective high performance supercapacitors. Adv Funct Mater 2014;24:4186–94. [159] Xing W, Qiao SZ, Ding RG, et al. Superior electric double layer capacitors using ordered mesoporous carbons. Carbon 2006;44:216–24. [160] Zhang LL, Zhou R, Zhao XS. Graphene-based materials as supercapacitor electrodes. J Mater Chem 2010;20:5983–92. [161] Winter M, Brodd RJ. What are batteries, fuel cells, and supercapacitors? Chem Rev 2004;104:4245–69. [162] Zhang H, Cao G, Yang Y. Carbon nanotube arrays and their composites for electrochemical capacitors and lithium-ion batteries. Energy Environ Sci 2009;2:932–43. [163] Zhang M, He C, Liu E, et al. Activated carbon nanochains with tailored micro-meso pore structures and their application for supercapacitors. J Phys Chem C 2015;119:21810–7. [164] Wang Y, Chang B, Guan D, Dong X. Mesoporous activated carbon spheres derived from resorcinol-formaldehyde resin with high performance for supercapacitors. J Solid State Electrochem 2015;19:1783–91. [165] Jin Z, Yan X, Yu Y, Zhao G. Sustainable activated carbon fibers from liquefied wood with controllable porosity for high-performance supercapacitors. J Mater Chem A 2014;2:11706–15. [166] Zhang H, Wang K, Zhang X, et al. Self-generating graphene and porous nanocarbon composites for capacitive energy storage. J Mater Chem A 2015;3:11277–86. [167] Zhang L, Zhang F, Yang X, et al. Porous 3D graphene-based bulk materials with exceptional high surface area and excellent conductivity for supercapacitors. Sci Rep 2013;3:1408. [168] Yu Z, Tetard L, Zhai L, Thomas J. Supercapacitor electrode materials: nanostructures from 0 to 3 dimensions. Energy Environ Sci 2015;8:702–30. [169] Guo Y, Chang B, Wen T, et al. One-pot synthesis of graphene/zinc oxide by microwave irradiation with enhanced supercapacitor performance. RSC Adv 2016;6:19394–403. [170] Han X, Funk MR, Shen F, et al. Scalable holey graphene synthesis and dense electrode fabrication toward high-performance ultracapacitors. ACS Nano 2014;8:8255–65. [171] Wilson E, Islam MF. Ultracompressible, high-rate supercapacitors from graphene-coated carbon nanotube aerogels. ACS Appl Mater Interfaces 2015;7:5612–8. [172] Hu H, Zhao Z, Wan W, Gogotsi Y, Qiu J. Ultralight and highly compressible graphene aerogels. Adv Mater 2013;25:2219–23. [173] Fan X, Yu C, Yang J, Ling Z, Qiu J. Hydrothermal synthesis and activation of graphene-incorporated nitrogen-rich carbon composite for high-performance supercapacitors. Carbon 2014;70:130–41. [174] Xu B, Yue S, Sui Z, et al. What is the choice for supercapacitors: graphene or graphene oxide. Energy Environ Sci 2011;4:2826–30. [175] An KH, Kim WS, Park YS, et al. Electrochemical properties of high-powder supercapacitors using single-walled carbon nanotube electrodes. Adv Funct Mater 2001;11:387–92. [176] Brousse T, Tabema PL, Crosrier O, et al. Long-term cycling behavior of asymmetric activated carbon/MnO2 aqueous electrochemical supercapacitor. J Power Sources 2007;173:633–41. [177] Iwakura C, Fukumoto Y, Inoue H, et al. Electrochemical characterization of various metal foils as a current collector of positive electrode for rechargeable lithium batteries. J Power Sources 1997;68:301–3. [178] Wu HC, Lee E, Wu NL, Jow TR. Effects of current collectors on power performance of Li4T5O12 anode for Li-ion battery. J Power Sources 2012;197:301–4. [179] Oh YJ, Yoo JY, Kim II Y, et al. Oxygen functional groups and electrochemical capacitive behavior of incompletely reduced graphene oxides as a thin-film electrode of supercapacitor. Electrochim Acta 2014;116:118–28. [180] Yu M, Huang Y, Li C, et al. Building three-dimensional graphene frameworks for energy storage and catalysis. Adv Funct Mater 2015;25:324–30. [181] Huang L, Chen D, Ding Y, Feng S, Wang ZL, Liu M. Nickel-cobalt hydroxide nanosheets coated on NiCo2O4 nanowires grown on carbon fiber paper for highperformance pseudocapacitors. Nano Lett 2013;13:3135–9. [182] Wang H, Yi H, Chen X, Wang X. Asymmetric supercapacitors based on nano-architectured nickel oxide/graphene foam and hierarchical porous nitrogen-doped carbon nanotubes with ultrahigh-rate performance. J Mater Chem A 2014;2:3223–30. [183] Wang Q, Jiao L, Du H, Si Y, Wang Y, Yuan H. Co3S4 hollow nanospheres grown on graphene as advanced electrode materials for supercapacitors. J Mater Chem 2012;22:21387–91. [184] Peng C, Yang L, Zhang Z, Tachibana K, Yang Y, Zhao S. Investigation of the anodic behavior of Al current collector in room temperature ionic liquid electrolytes. Electrochim Acta 2008;53:4764–72. [185] Wang X, Yasuikawa E, Mori S. Inhibition of anodic corrosion of aluminum cathode current collector on recharging in lithium imide electrolytes. Electrochim Acta 2000;45:2677–84. [186] Busom J, Schreiber A, Tolosa A, et al. Sputtering of sub-micrometer aluminum layers as compact, high-performance, light-weight current collector for supercapacitors. J Power Sources 2016;329:432–40.

688

Supercapacitors

[187] Peng C, Yang L, Zhang Z, Tachibana K, Yang Y. Anodic behavior of Al current collector in 1-alkyl-3-methylimidazolium bis[(trifluoromethyl)sulfonyl] amide ionic liquid electrolytes. J Power Sources 2007;173:510–7. [188] Cho E, Mun J, Chae OB, et al. Corrosion/passivation of aluminum current collector in bis(fluorosulfony)imide-based ionic liquid for lithium-ion batteries. Electrochem Commun 2012;22:1–3. [189] Wu HC, Lin YP, Lee E, et al. High-performance carbon-based supercapacitors using Al current-collector with conformal carbon coating. Mater Chem Phys 2009;117:294–300. [190] Portet C, Tabema PL, Simon P, Laberty-Robert C. Modification of Al current collector surface by sol–gel deposit for carbon–carbon supercapacitor application. Electrochim Acta 2004;49:905–12. [191] Portet C, Tabema PL, Simon P, Flahaut E, Laberty-Robert C. High-power density electrodes for carbon supercapacitor applications. Electrochim Acta 2005;50:4174–81. [192] Kang SW, Xie HM, Zhang W, Zhang JP, Ma Z, Wang RS. Improve the overall performances of lithium ion batteries by a facile method of modifying the surface of Cu current collector with carbon. Electrochim Acta 2015;176:604–9. [193] Zhao X, Zhang L, Murali S, et al. Incorporation of manganese dioxide within ultraporous activated graphene for high-performance electrochemical capacitors. ACS Nano 2012;6:5404–12. [194] Wu H, Susanto A, Lian K. Thin and flexible Ni-P based current collectors developed by electroless deposition for energy storage devices. Appl Surf Sci 2017;394:63–9. [195] Bo Z, Zhu W, Ma W, et al.Vertically oriented graphene bridging active-layer/current-collector interface for ultrahigh rate supercapacitors. Adv Mater 2013;25:5799–806. [196] Liu D, Wang Q, Qiao L, et al.Preparation of nano-networks of MnO2 shell/Ni current collector core for high-performance supercapacitor electrodes. J Mater Chem 2012;22:483–7. [197] Sumboja A, Wang X, Yan J, Lee PS. Nanoarchitectured current collector for high-rate capability of polyaniline based supercapacitor electrode. Electrochim Acta 2012;65:190–5. [198] Singh N, Galande C, Miranda A, et al. Paintable battery. Sci Rep 2012;2:481. [199] Choi JY, Lee DJ, Lee YM, et al. Silicon nanofibrils on a flexible current collector for bendable lithium-ion battery anodes. Adv Funct Mater 2013;23:2108–14. [200] Wei D, Wakeham SJ, Ng TW, Thwaites MJ, Brown H, Beecher P. Transparent, flexible and solid-state supercapcitors based on room temperature ionic liquid gel. Electrohem Commun 2009;11:2285–7. [201] Ren HM, Ding YH, Chang FH, et al. Flexible free-standing TiO2/graphene/PVDF films as anode materials for lithium-ion batteries. Appl Surf Sci 2012;263:54–7. [202] Zhang D, Ye K, Cheng K, et al. Fabric-based flexible electrode with multi-walled carbon nanotubes@Ni network structure as a novel anode for hydrogen peroxide electrooxidation. RSC Adv 2014;4:17454–60. [203] Hu L, Mantia FL, Wu H, et al. Lithium-ion textile batteries with large areal mass loading. Adv Energy Mater 2011;1:1012–7. [204] Kaempgen M, Chan CK, Ma J, Cui Y, Gruner G. Printable thin film supercapcitors using single-walled carbon nanotubes. Nano Lett 2009;9:1872–6. [205] Hu L, Chen W, Xie X, et al. Symmetrical MnO2-carbon nanotube-textile nanostructures for wearable pseudocapacitors with high mass loading. ACS Nano 2011;5:8904–13. [206] Liu L, Niu Z, Zhang L, Zhou W, Chen X, Xie S. Nanostructured graphene composite papers for highly flexible and foldable supercapacitors. Adv Mater 2014;26:4855–62. [207] Wang H, Zhu B, Jiang W, et al. A mechanically and electrically self-healing supercapacitor. Adv Mater 2014;26:3638–43. [208] He Y, Chen W, Gao C, Zhou J, Li X, Xie E. An overview of carbon materials for flexible electrochemical capacitors. Nanoscale 2013;5:8799–820. [209] Zhou R, Meng C, Zhu F, et al. High-performance supercapacitors using a nanoporous current collector made from super-aligned carbon nanotubes. Nanotechnology 2010;21:345701. [210] Chou SL, Wang JZ, Chew SY, Liu HK, Dou SX. Electrodeposition of MnO2 nanowires on carbon nanotube paper as free-standing, flexible electrodes for supercapacitors. Electrochem Commun 2008;10:1724–7. [211] Ning J, Hao L, Zhang X, Liang M, Zhi L. High-quality graphene grown directly on stainless steel meshes through CVD process for enhanced current collectors of supercapacitors. Sci China Technol Sci 2014;57:259–63. [212] Lin T, Chen IW, Lin F, et al. Nitrogen-doped mesoporous carbon of extraordinary capacitance for electrochemical energy storage. Science 2015;350:1508–13. [213] Yu J, Wu J, Wang H, et al. Metallic fabrics as the current collector for high-performance graphene-based flexible solid-state supercapacitors. ACS Appl Mater Interfaces 2016;8:4724–9. [214] Meng C, Liu C, Chen L, Hu C, Fan S. Highly flexible and all-solid-state paperlike polymer supercapacitors. Nano Lett 2010;10:4025–31. [215] He Y, Chen W, Li X, et al. Freestanding three-dimensional graphene/MnO2 composite networks as ultralight and flexible supercapacitor electrodes. ACS Nano 2013;7:174–82. [216] Peng L, Peng X, Lin B, Wu C, Xie Y, Yu G. Ultrathin two-dimensional MnO2/graphene hybrid nanostructures for high-performance, flexible planar supercapacitors. Nano Lett 2013;13:2151–7. [217] Xu Y, Lin Z, Huang X, Wang Y, Huang Y, Duan X. Functionalized graphene hydrogel-based high-performance supercapacitors. Adv Mater 2013;25:5779–84. [218] Yu D, Goh K, Wang H, et al. Scalable synthesis of hierarchically structured carbon nanotube-graphene fibres for capacitive energy storage. Nat Nanotechnol 2014;9:555–61. [219] Zhang L, Liu Z, Cui G, Chen L. Biomass-derived materials for electrochemical energy storages. Prog Poly Sci 2015;43:136–64. [220] Cao WJ, Zheng JP. Li-ion capacitors with carbon cathode and hard carbon/stabilized lithium metal powder anode electrodes. J Power Sources 2012;213:180–5. [221] Shen J, Yang C, Li X, Wang G. High-performance asymmetric supercapacitor based on nanoarchitectured polyaniline/graphene/carbon nanotube and activated graphene electrodes. ACS Appl Mater Interfaces 2013;5:8467–76. [222] Luo X, Pan W, Liu H, Gong J, Wu H. Glass fiber fabric mat as the separator for lithium-ion battery with high safety performance. Ionics 2015;21:3135–9. [223] He M, Zhang X, Jiang K, Wang J, Wang Y. Pure inorganic separator for lithium ion batteries. ACS Appl Mater Interfaces 2015;7:738–42. [224] Xu Q, Kong Q, Liu Z, et al. Cellulose/polysulfonamide composite membrane as a high performance lithium-ion battery separator. ACS Sustainable Chem Eng 2014;2:194–9. [225] Yoo J, Kim J, Kim YS. Liquid electrolyte-free cylindrical Al polymer capacitor review: materials and characteristics. J Power Sources 2015;284:468–80. [226] Zhong C, Deng Y, Hu W, Qiao J, Zhang L, Zhang J. A review of electrolyte materials and compositions for electrochemical supercapacitors. Chem Soc Rev 2015;44:7484–539. [227] Cao W, Li Y, Fitch B, Shih J, Doung T, Zheng J. Strategies to optimize lithium-ion supercapacitors achieving high-performance: cathode configurations, lithium loadings on anode, and types of separator. J Power Sources 2014;268:841–7. [228] Noorden ZA, Sugawara S, Matsumoto S. Noncorrosive separator materials for electric double layer capacitor. IEEJ Trans 2014;9:235–40. [229] Tonurist K, Thomberg T, Janes A, Kink I, Lust E. Specific performance of electrical double layer capacitors based on different separator materials in room temperature ionic liquid. Electrochem Commun 2012;22:77–80. [230] Lee Y, Park J, Jeon H, et al. In-depth correlation of separator pore structure and electrochemical performance in lithium-ion batteries. J Power Sources 2016;325:732–8. [231] Chen J, Wang S, Ding L, Jiang Y, Wang H. Performance of through-hole anodic aluminum oxide membrane as a separator for lithium-ion battery. J Membr Sci 2014;461:22–7. [232] Shulga YM, Baskakov SA, Simimov VA, Shulga NY, Belay KG, Gulsev GL. Graphene oxide films as separators of polyaniline-based supercapacitors. J Power Sources 2012;245:33–6. [233] Yu H, Tang Q, Wu J, et al. Using eggshell membrane as a separator in supercapacitor. J Power Sources 2012;206:463–8. [234] Tavernor AW, Li HPS, Bell AJ, Stevens R. Improved compaction in multilayer capacitor fabrication. J Eur Ceram Soc 1999;19:1691–5.

Supercapacitors

689

[235] Abbas O, Pajak D, Frackowiak E, Beguin F. Effect of binder on the performance of carbon/carbon symmetric capacitors in salt aqueous electrolyte. Electrochim Acta 2014;140:132–8. [236] Bichat MP, Raymundo-Pinero E, Beguin F. High voltage supercapacitor built with seaweed carbons in neutral aqueous electrolyte. Carbon 2010;48:4351–61. [237] Tsay KC, Zhang L, Zhang J. Effects of electrode layer composition/thickness and electrolyte concentration on both specific capacitance and energy density of supercapacitor. Electrochim Acta 2012;60:428–36. [238] Komaba S, Hasegawa T, Dahbi M, Kubota K. Potassium intercalation into graphite to realize high-voltage/high-powder potassium-ion batteries and potassium-ion capacitors. Electrochem Commun 2015;60:172–5. [239] Komaba S, Yabuuchi N, Ozeki T, et al. Functional binder for reversible lithium intercalation into graphite in propylene carbonate and ionic liquid media. J Power Sources 2010;195:6069–74. [240] Lee J, Sakaushi K, Antonietti M, Yuan J. Poly(ionic liquid) binders as Li conducting mediators for enhanced electrochemical performance. RSC Adv 2015;5:85517–22. [241] Fukushima T, Aida T. Ionic liquid for soft functional materials with carbon nanotubes. Chem Eur J 2007;13:5048–58. [242] Kim G, Jeong S, Joost M, et al. Use of natural binders and ionic liquid electrolytes for greener and safer lithium-ion batteries. J Power Sources 2011;196:2187–94. [243] Takahashi T, Tejedor-Tejedor MI, Wouters JJ, Perez-Roa R, Anderson MA. Performance of metal oxide nanoparticles sols as binders in activated carbon electrodes. J Energy Storage 2016;7:147–58. [244] Eftekhari A, Liu Y, Chen P. Different roles of ionic liquids in lithium batteries. J Power Sources 2016;334:221–39. [245] Benoit C, Demeter D, Belanger D, Cougnon C. A redox-active binder for electrochemical capacitor electrodes. Angew Chem Int Ed 2016;55:5318–21. [246] Varzi A, Raccichini R, Marinaro M, Wohlfahrt-Mehrens M, Passerini S. Probing the characteristics of casein as green binder for non-aqueous electrochemical double layer capacitors’ electrodes. J Power Sources 2016;326:672–9. [247] Choudhury NA, Northrop PWC, Crothers AC, Jain S, Subramanian VR. Chitosan hydrogel-based electrode binder and electrolyte membrane for EDLCs: experimental and model validation. J Appl Electrochem 2012;42:935–43. [248] Kouchachvili L, Maffei N, Entchev E. Novel binding materials for supercapacitor electrodes. J Solid State Electrochem 2014;18:2539–47. [249] Aslan M, Weingarth D, Herbeck-Engel P, Grobelsek I, Presser V. Polyvinylpyrrolidon/polyvinyl butyral composite as a stable binder for castable supercapacitor electrodes in aqueous electrolytes. J Power Sources 2015;279:323–33. [250] Choudhury NA, Sahai Y, Buchheit RG. Chitosan chemical hydrogel electrode binder for direct borohydride fuel cells. Electrochem Commun 2011;13:1–4. [251] Tsao CH, Hsu CH, Kuo PL. Ionic conducting and surface active binder of poly (ethylene oxide)- block-poly(acrylonitrile) for high power lithium-ion battery. Electrochim Acta 2016;196:41–7. [252] Bockenfeld N, Jeong SS, Winter M, Passerini S, Balducci A. Natural, cheap and environmentally friendly binder for supercapacitors. J Power Sources 2013;221:14–20. [253] Choudhury NA, Ma J, Sahai Y, Buchheit RG. High performance polymer chemical hydrogel-based electrode binder materials for direct boronhydride fuel cells. J Power Sources 2011;196:5817–22. [254] Varzi A, Passerini S. Enabling high areal capacitance in electrochemical double layer capacitors by means of the environmentally friendly starch binder. J Power Sources 2015;300:216–22. [255] Lee KT, Tsai CB, Ho WH, Wu NL. Superabsorbent polymer binder for achieving MnO2 supercapcitors of greatly enhanced capacitance density. Electrochem Commun 2010;12:886–9. [256] Sun X, Zhang X, Zhang H, Huang B, Ma Y. Application of a novel binder for activated carbon-based electrical double layer capacitors with nonaqueous electrolytes. J Solid State Electrochem 2013;17:2035–42. [257] Lukatskaya MR, Mashitalir O, Ren CE, et al. Cation intercalation and high volumetric capacitance of two-dimensional titanium carbide. Science 2013;341:1502–5. [258] Torchala K, Kierzek K, Machnikowski J. Capacitane behavior of KOH activated mesocarbon microbeads in different aqueous electrolytes. Electrochim Acta 2012;86:260–7. [259] Narayanan R, Bandaru PR. High rate capacity through redox electrolytes confined in macroporous electrodes. J Electrochem Soc 2015;162:A86–91. [260] Gonzalez A, Goikolea E, Barrena JA, Mysyk R. Review on supercapacitors: technologies and materials. Renew Sust Energy Rev 2016;58:1189–206. [261] Simon P, Gogotsi Y. Capacitive energy storage in nanostructured carbon-electrolyte systems. Acc Chem Res 2013;46:1094–103. [262] Vellacheri R, Al-Haddad A, Zhao H, Wang W, Wang C, Lei Y. High performance supercapacitor for efficient energy storage under extreme environmental temperatures. Nano Energy 2014;8:231–7. [263] Chen XY, Chen C, Zhang ZJ, Xie DH. Gelatin-derived nitrogen-doped porous carbon via a dual-template carbonization method for high performance supercapacitors. J Mater Chem A 2013;1:1093–911. [264] Conway BE. Electrochemical supercapacitors: scientific fundamentals and technological applications. New York: Kluwer Academic Plenum; 1999. [265] Frackowiak E, Abbas Q, Beguin F. Carbon/carbon supercapacitors. J Energy Chem 2013;22:226–40. [266] Armand M, Endres F, MacFarlane DR, Ohno H, Scrosati B. Ionic-liquid materials for the electrochemical challenges of the future. Nat Mater 2009;8:621–9. [267] Beguin F, Presser V, Balducci A, Frackowiak E. Carbons and electrolytes for advanced supercapacitors. Adv Mater 2014;26:2219–51. [268] Zhang K, Man L, Zhang LL, Chan KSO, Zhao XS, Wu J. Surfactant-intercalated, chemically reduced graphene oxide for high performance supercapacitor electrodes. J Mater Chem 2011;21:7302–7. [269] Hall PJ, Mirzaeian M, Fletcher SI, et al. Energy storage in electrochemical capacitors: designing functional materials to improve performance. Energy Environ Sci 2010;3:1238–51. [270] Van Aken KL, Beidaghi M, Gogotsi Y. Formulation of ionic-liquid electrolyte to expand the voltage window of supercapacitors. Angew Chem Int Ed 2015;54:4806–9. [271] Yan J, Wang Q, Wei T, Fan Z. Recent advances in design and fabrication of electrochemical supercapacitors with high energy densities. Adv Energy Mater 2014;4:1300816. [272] Kim TY, Lee HW, Stoller M, et al. High-performance supercapacitors based on poly(ionic liquid)-modified graphene electrodes. ACS Nano 2011;5:436–42. [273] Peng C, Wen Z, Qin Y, et al. Three-dimensional graphitized carbon nanovesicles for high-performance supercapacitors based on ionic liquids. ChemSusChem 2014;7:777–84. [274] Aradilla D, Gentile P, Bidan G, et al. High performance of symmetric micro-supercapacitors based on silicon nanowires using N-methyl-N-propylpyrrolidinium bis (trifluoromethylsulfonyl)imide as electrolyte. Nano Energy 2014;9:273–81. [275] Yang CH, Huang PL, Luo XF, et al. Holey graphene nanosheets with surface functional groups as high-performance supercapacitors in ionic liquid electrolyte. ChemSusChem 2015;8:1779–86. [276] Kroto HW, Heath JR, O'Brien SC, Curl RF, Smalley RE. C60: Buckminsterfullerene. Nature 1985;318:162–3. [277] Iijima S. Helical microtubules of graphitic carbon. Nature 1991;354:56–8. [278] Futaba DN, Hata K, Yamada T, et al. Shape-engineerable and highly densely packed single-walled carbon nanotubes and their application as supercapacitor electrodes. Nat Mater 2006;5:987–94. [279] Novoselov KS, Geim AK, Morozov SV, et al. Electric field effect in atomically thin carbon films. Science 2004;306:666–9. [280] Novoselov KS, Geim AK, Morozov SV, et al. Two-dimensional gas of massless dirac fermions in graphene. Nature 2005;438:197–200. [281] Dikin DA, Stankovich S, Zimney EJ, et al. Preparation and characterization of graphene oxide paper. Nature 2007;448:457–60. [282] Elias DC, Nair RR, Mohiuddin TMG, et al. Control of graphene’s properties by reversible hydrogenation: evidence for graphene. Science 2009;323:610–3. [283] Geim AK. Graphene: status and prospects. Science 2009;324:1530–4. [284] Hu YH, Wang H, Hu B. Thinnest two-dimensional nanomaterial-graphene for solar energy. ChemSusChem 2010;3:782–96.

690

Supercapacitors

[285] Wang H, Sun K, Tao F, Stacchiola D, Hu YH. 3D Honeycomb-like structured graphene and its high efficiency as counter electrode catalysts for dye-sensitized solar cells. Angew Chem Int Ed 2013;52:9210–4. [286] Qu D, Shi H. Studies of activated carbon used in double-layer capacitors. J Power Sources 1998;74:99–107. [287] Barbieri O, Hahn M, Herzog A, Kotz R. Capacitance limits of high surface area activated carbons for double layer capacitors. Carbon 2005;43:1303–10. [288] Kierzek K, Frackowiak E, Lota G, Gryglewicz G, Machnikowski J. Electrochemical capacitors based on highly porous carbons prepared by KOH activation. Electrochim Acta 2004;49:515–23. [289] Fang B, Binder L. A modified activated carbon aerogel for high-energy storage in electric double layer capacitors. J Power Sources 2006;163:616–22. [290] Shiraishi S, Kurihara H, Okabe K, Hulicova D, Oya A. Electric double layer capacitance of highly pure single-walled carbon nanotubes (HiPcoTM BuckytubesTM) in propylene carbonate electrolytes. Electrochem Commun 2002;4:593–8. [291] Barisci JN, Wallace GG, Baughman RH. Electrochemical characterization of single-walled carbon nanotube electrodes. J Electrochem Soc 2000;147:4580–3. [292] Morishita T, Soneda Y, Tsumura T, Inagaki M. Preparation of porous carbons from thermoplastic precursors and their performance for electric double layer capacitors. Carbon 2006;44:2360–7. [293] Xu B, Wu F, Chen R, et al. Highly mesoporous and high surface area carbon: a high capacitance electrode material for EDLCs with various electrolytes. Electrochem Commun 2008;10:795–7. [294] Wang DW, Li F, Liu M, Lu GQ, Cheng HM. 3D Aperiodic hierarchical porous graphitic carbon material for high-rate electrochemical capacitive energy storage. Angew Chem Int Ed 2008;47:373–6. [295] Gao F, Qu J, Geng C, Shao G, Wu M. Self-templating synthesis of nitrogen-decorated hierarchical porous carbon from shrimp shell for supercapacitors. J Mater Chem A 2016;4:7445–52. [296] Biener J, Stadermann M, Suss M, et al. Advanced carbon aerogels for energy application. Energy Environ Sci 2011;4:656–67. [297] Xu Y, Shi G, Duan X. Self-assembled three-dimensional graphene macrostructures: synthesis and applications in supercapacitors. Acc Chem Res 2015;48:1666–75. [298] Oschatz M, Boukhalfa S, Nickel W, et al. Carbon-derived carbon aerogels with tunable pore structure as versatile electrode material in high power supercapacitors. Carbon 2017;113:283–91. [299] Frackowiak E, Beguin F. Carbon materials for the electrochemical storage of energy in capacitors. Carbon 2001;39:937–50. [300] Genovese M, Jiang J, Lian K, Holm N. High capacitive performance of exfoliated biochar nanosheets from biomass waste corn cob. J Mater Chem A 2015;3:2903–13. [301] Zhu Y, Murali S, Cai W, et al. Graphene and graphene oxide: synthesis, properties, and application. Adv Mater 2010;22:3906–24. [302] Stankovich S, Dikin DA, Dommett GHB, et al. Graphene-based composite materials. Nature 2006;442:282–6. [303] Bonaccorso F, Colombo L, Yu G, et al. Graphene, related two-dimensional crystals, and hybrid systems for energy conversion and storage. Science 2015;347. 12465011-9. [304] Wang Y, Shi Z, Huang Y, et al. Supercapacitor devices based on graphene materials. J Phys Chem C 2009;113:13103–7. [305] Xia J, Chen F, Li J, Tao N. Measurement of the quantum capacitance of graphene. Nat Nanotechnol 2009;4:505–9. [306] Stroller MD, Park S, Zhu Y, An J, Ruoff RS. Graphene-based ultracapacitors. Nano Lett 2008;8:3498–502. [307] Yan J, Wei T, Shao B, et al. Electrochemical properties of graphene nanosheet/carbon black composites as electrodes for supercapacitors. Carbon 2010;48:1731–7. [308] Kang D, Liu Q, Gu J, Su Y, Zhang W, Zhang D. “Egg-box”-assisted fabrication of porous carbon with small mesopores for high-rate electric double layer capacitors. ACS Nano 2015;9:11225–33. [309] Zhu Y, Murali S, Stoller MD, et al. Carbon-based supercapacitors produced by activated of graphene. Science 2011;332:1537–41. [310] Kim T, Jung G, Yoo S, Suh KS, Ruoff RS. Activated graphene-based carbons as supercapacitors electrodes with macro- and mesoporous. ACS Nano 2013;7:6899–905. [311] Fan X, Yu C, Yang J, et al. A layered-nanoscope-confinement strategy for the synthesis of two-dimensional porous carbon nanosheets for high-rate performance supercapacitors. Adv Energy Mater 2015;5:1401761. [312] Xu Y, Chen CY, Zhao Z, et al. Solution processable holey graphene oxide and its derived macrostructures for high-performance supercapacitors. Nano Lett 2015;15:4605–10. [313] Sanchez-Sanchez A, Fierro V, Izquierdo MT, Celzard A. Functionalized, hierarchical and ordered mesoporous carbons for high-performance supercapacitors. J Mater Chem A 2016;4:6140–8. [314] Wu YP, Rahm E, Holze R. Effects of heteroatoms on electrochemical performance of electrode materials for lithium ion batteries. Electrochim Acta 2002;47:3491–507. [315] Stein A, Wang Z, Fierke MA. Functionalization of porous carbon materials with designed pore architecture. Adv Mater 2009;21:265–93. [316] Hulicova-Jucakova D, Seredych M, Lu GQ, Bandosz TJ. Combined effect of nitrogen- and oxygen-containing functional groups of microporous activated carbon on its electrochemical performance in supercapacitors. Adv Funct Mater 2009;19:438–47. [317] Zhao L, Fan LZ, Zhou MQ, et al. Nitrogen-containing hydrothermal carbons with superior performance in supercapacitors. Adv Mater 2010;22:5202–6. [318] Wu ZS, Winter A, Chen L, et al. Three-dimensional nitrogen and boron co-doped graphene for high-performance all-solid-state supercapacitors. Adv Mater 2012;24:5130–5. [319] Wang T, Wang LX, Wu DL, Xia W, Jia DZ. Interaction between nitrogen an sulfur in co-doped graphene and synergetic effect in supercapacitor. Sci Rep 2015;5:9591. [320] Ling Z, Wang Z, Zhang M, et al. Sustainable synthesis and assembly of biomass-derived B/N co-doped carbon nanosheets with ultrahigh aspect ratio for highperformance supercapacitors. Adv Funct Mater 2016;26:111–9. [321] Deng W, Ji X, Gomez-Mingot M, Lu F, Chen Q, Banks CF. Graphene electrochemical supercapacitors: the influence of oxygen functional groups. Chem Commun 2012;48:2770–2. [322] Zhang LL, Zhao X, Ji H, et al. Nitrogen doping of graphene and its effect on quantum capacitance, and a new insight on the enhanced capacitance of N-doped carbon. Energy Environ Sci 2012;5:9618–25. [323] Jeong HM, Lee JW, Shin WH, et al. Nitrogen-doped graphene for high-performance ultracapacitors and the importance of nitrogen doped sites at basal planes. Nano Lett 2011;11:2472–7. [324] Maiti UN, Lim J, Lee KE, Lee WJ, Kim SO. Three-dimensional shape engineered, interfacial gelation of reduced graphene oxide for high rate, large capacity supercapacitors. Adv Mater 2014;36:615–9. [325] Wei H, Wei S, Tian W, et al. Fabrication of thickness controllable free-standing sandwich-structured hybrid carbon film for high-rate and high-powder supercapacitor. Sci Rep 2014;4:7050. [326] Pan LZ, Chen TT, Song WL, Li X, Zhang S. High nitrogen-containing cotton derived 3D porous carbon frameworks for high-performance supercapacitors. Sci Rep 2015;5:15388. [327] Wang Z, Han Y, Zeng Y, et al. Activated carbon fiber paper with exceptional capacitive performance as a robust electrode for supercapacitors. J Mater Chem A 2016;4:5828–33. [328] Huang Q, Liu L, Wang D, Liu J, Huang Z, Zheng Z. One-step electrospinning of carbon nanowebs on metallic textiles for high-capacitance supercapacitor fabrics. J Mater Chem A 2016;4:6802–8. [329] Chang L, Wei W, Sun K, Hu YH. Excellent performance of highly conductive porous Na-embedded carbon nanowalls for electric double-layer capacitors with a wide operating temperature range. J Mater Chem A 2017;5:9090–6. [330] Yang X, Cheng C, Wang Y, Qiu L, Li D. Liquid-mediated dense integration of graphene materials for compact capacitive energy storage. Science 2013;341:534–7. [331] Tao Y, Xie X, Lv W, et al. Towards ultrahigh volumetric capacitance: graphene derived highly dense but porous carbons for supercapacitors. Sci Rep 2013;3:2975. [332] Lee JH, Park N, Kim BG, et al. Restacking-inhibited 3D reduced graphene oxide for high performance supercapacitors electrodes. ACS Nano 2013;7:9366–74.

Supercapacitors

691

[333] Xu Y, Lin Z, Zhong X, et al. Holey graphene frameworks for highly efficient capacitive energy storage. Nat Commun 2014;5:4554. [334] Chang L, Wei W, Sun K, Hu YH. 3D flower-structured graphene from CO2 for supercapacitors with ultrahigh areal capacitance at high current density. J Mater Chem A 2015;3:10183–7. [335] Yu M, Qiu W, Wang F, et al. Three dimensional architectures: design, assembly and application in electrochemical capacitors. J Mater Chem A 2015;3:15792–823. [336] Chang L, Stacchiola DJ, Hu YH. An ideal electrode material, 3D surface-microporous graphene for supercapacitors with ultrahigh areal capacitance. ACS Appl Mater Interfaces 2017;9:24655–61. [337] Cao X, Shi Y, Shi W, et al. Preparation of novel 3D graphene networks for supercapacitor applications. Small 2011;22:3163–8. [338] Luo J, Jang HD, Huang J. Effect of sheet morphology on the scalability of graphene-based ultracapacitors. ACS Nano 2013;7:1464–71. [339] Li Y, Li Z, Shen PK. Simultaneous formation of ultrahigh surface area and three-dimensional hierarchical porous graphene-like networks for fast and highly stable supercapacitors. Adv Mater 2013;25:2474–80. [340] Xu Y, Lin Z, Huang X, Liu Y, Huang Y, Duan X. Flexible solid-state supercapacitors based on three-dimensional graphene hydrogel films. ACS Nano 2013;7:4042–9. [341] Cao X, Yin Z, Zhang H. Three-dimensional graphene materials: preparation, structures and application in supercapacitors. Energy Environ Sci 2014;7:1850–65. [342] Yan Z, Yao W, Hu L, Liu D, Wang C, Lee CS. Progress in the preparation and application of three-dimensional graphene-based porous nanocomposites. Nanoscale 2015;7:5563–77. [343] Wang H, Casalongue HS, Liang Y, Dai H. Ni(OH)2 nanoplates grown on graphene as advanced electrochemical pseudocapacitor materials. J Am Chem Soc 2010;132:7472–7. [344] Zhang L, Wang J, Zhu J, Zhang X, Hui KS, Hui KN. 3D porous layered double hydroxides grown on graphene as advanced electrochemical pseudocapacitor materials. J Mater Chem A 2013;1:9046–53. [345] Gawli Y, Banerjee A, Dhakras D, et al. 3D polyaniline architecture by concurrent inorganic and organic acid doping for superior and robust high rate supercapacitor performance. Sci Rep 2016;6:21002. [346] Yang X, Shi K, Zhitomirsky I, Cranston ED. Cellulose nanocrystal aerogels as universal 3D lightweight substrates for supercapacitors materials. Adv Mater 2015;27:6104–9. [347] Wu Y, Gao G, Wu G. Self-assembled three-dimensional hierarchical porous V2O5/graphene hybrid aerogels for supercapacitors with high energy density and long cycle life. J Mater Chem A 2015;5:1828–32. [348] Wang B, Park J, Su D, Wang C, Ahn H, Wang G. Solvothermal synthesis of CoS2-graphene nanocomposite material for high-performance supercapacitors. J Mater Chem 2012;22:15750–6. [349] Cong HP, Chen JF, Yu SH. Graphene-based macroscopic assemblies and architectures: an emerging materials systems. Chem Soc Rev 2014;43:7295–325. [350] Jiang H, Lee PS, Li C. 3D carbon based nanostructures for advanced supercapacitors. Energy Environ Sci 2013;6:41–53. [351] Kong L, Zhang C, Wang J, Qiao W, Ling L, Long D. Free-standing T-Nb2O5/graphene composite papers with ultrahigh gravimetric/volumetric capacitance for Li-ion intercalation pseudocapacitors. ACS Nano 2015;9:11200–8. [352] Zhang J, Liu P, Zhang Y, et al. Enhanced performance of nano-Bi2WO6-graphene as pseudocapacitor electrodes by charge transfer channel. Sci Rep 2015;5:8624. [353] Wu ZS, Wang DW, Ren W, et al. Anchoring hydrous RuO2 on graphene sheets for high-performance electrochemical capacitors. Adv Funct Mater 2010;20:3595–602. [354] Lee JW, Hall AS, Kim JD, Mallouk TE. A facile and template-free hydrothermal synthesis of Mn3O4 nanorods on graphene sheets for supercapacitor electrodes with long cycle stability. Chem Mater 2012;24:1158–64. [355] Balasingam SK, Lee JS, Jun Y. Molybdenum diselenide/reduced graphene oxide based hybrid nanosheets for supercapacitor applications. Dalton Trans 2016;45:9646–53. [356] Yu G, Hu L, Vosgueritchian M, et al. Solution-processed graphene/MnO2 nanostructured textiles for high-performance electrochemical capacitors. Nano Lett 2011;11:2905–11. [357] Yu X, Lu B, Xu Z. Super long-life supercapacitors based on the construction of nanohoneycomb-like strongly coupled CoMoO4-3D graphene hybrid electrodes. Adv Mater 2014;26:1044–51. [358] Lee SW, Kim J, Chen S, Hammond PT, Horn YS. Carbon nanotube/manganese oxide ultrathin film electrodes for electrochemical capacitors. ACS Nano 2010;4:3889–96. [359] Wu ZS, Sun Y, Tan YZ, Yang S, Feng X, Mullen K. Three-dimensional graphene-based macro- and mesoporous frameworks for high-performance electrochemical capacitive energy storage. J Am Chem Soc 2012;134:19532–5. [360] Huang F, Chen D. Towards the upper bound of electrochemical performance of ACNT@polyaniline arrays as supercapacitors. Energy Environ Sci 2012;5:5833–41. [361] Luo J, Zhong W, Zou Y, Xiong C, Yang W. Preparation of morphology-controllable polyaniline and polyaniline/graphene hydrogels for high performance binder-free supercapacitors electrodes. J Power Sources 2016;319:73–81. [362] Yan J, Wei T, Shao B, et al. Preparation of a graphene nanosheet/polyaniline composite with high specific capacitance. Carbon 2010;48:487–93. [363] Wang K, Zhang X, Li C, et al. Chemically crosslinked hydrogel film leads to integrated flexible supercapacitors with superior performance. Adv Mater 2015;27:7451–7. [364] Zhao Y, Liu J, Hu Y, et al. Highly compression-tolerant supercapacitor based on polypyrrole-mediated graphene foam electrodes. Adv Mater 2013;25:591–5. [365] Kim M, Lee C, Jang J. Fabrication of highly flexible, scalable, and high-performance supercapacitors using polyaniline/reduced graphene oxide film with enhanced electrical conductivity and crystallinity. Adv Funct Mater 2014;24:2489–99. [366] Sha C, Lu B, Mao H, et al. 3D ternary nanocomposites of molybdenum disulfide/polyaniline/reduced graphene oxide aerogel for high performance supercapacitors. Carbon 2016;99:26–34. [367] Wu X, Wang Q, Zhang W, Wang Y, Chen W. Nano nickel oxide coated graphene/polyaniline composite film with high electrochemical performance for flexible supercapacitors. Electrochim Acta 2016;211:1066–75. [368] Yin Z, Zhou H, Fu C, Zhang N, Liu D, Kuang Y. Synthesis of curly graphene nanoribbon/polyaniline/MnO2 composite and its application in supercapacitors. RSC Adv 2016;6:41142–50. [369] Lin H, Huang Q, Wang J, et al. Self-assembled graphene/polyaniline/Co3O4 ternary hybrid aerogels for supercapacitors. Electrochim Acta 2016;191:444–51. [370] Yu G, Hu L, Liu N, et al. Enhancing the supercapacitor performance of graphene/MnO2 nanostructured electrodes by conducting wrapping. Nano Lett 2011;11:4438–42. [371] Wu ZS, Ren W, Wang DW, Li F, Liu B, Cheng HM. High-energy MnO2 nanowire/graphene and graphene asymmetric electrochemical capacitors. ACS Nano 2010;4:5835–42. [372] Qu L, Zhao Y, Khan AM, et al. Interwoven three-dimensional architecture of cobalt oxide nanobrush-graphene@NixCo2x(OH)6x for high performance supercapacitors. Nano Lett 2015;15:2037–44. [373] Sumboja A, Foo CY, Wang X, Lee PS. Large areal mass, flexible and free-standing reduced graphene oxide/manganese dioxide paper for asymmetric supercapacitor device. Adv Mater 2013;25:2809–15. [374] Bello A, Barzegar F, Momodu D, et al. Asymmetric supercapacitor based on nanostructured graphene foam/polyvinyl alcohol/formaldehyde and activated carbon electrodes. J Power Sources 2015;273:305–11. [375] Peng YJ, Wu TH, Hsu CT, Li SM, Chen MG, Hu CC. Electrochemical characteristics of the reduced graphene oxide/carbon nanotube/polypyrrole composites for aqueous asymmetric supercapacitors. J Power Sources 2014;272:970–8. [376] Fan Z, Yan J, Wei T, et al. Asymmetric supercapacitors based on graphene/MnO2 and activated carbon nanofiber electrodes with high power and energy density. Adv Funct Mater 2011;21:2366–75. [377] Choi BG, Yang M, Hong WH, Choi JW, Huh YS. 3D macroporous graphene frameworks for supercapacitors with high energy and power densities. ACS Nano 2012;6:4020–8.

692

Supercapacitors

[378] Yan J, Fan Z, Sun W, et al. Advanced asymmetric supercapacitors based on Ni(OH)2/graphene and porous graphene electrodes with high energy density. Adv Funct Mater 2012;22:2632–41. [379] Chen H, Zhou S, Wu L. Porous nickel hydroxide-manganese dioxide-reduced graphene oxide ternary hybrid spheres as excellent supercapacitor electrode materials. ACS Appl Mater Interfaces 2014;6:8621–30. [380] Chang J, Jin M, Yao F, et al. Asymmetric supercapacitors based on graphene/MnO2 nanospheres and graphene/MoO3 nanosheets with high energy density. Adv Funct Mater 2013;23:5074–83. [381] Tan YB, Lee JM. Graphene for supercapacitor applications. J Mater Chem A 2013;1:14814–43. [382] Choe NS, Chen Z, Freunberger SA, et al. Challenges facing lithium batteries and electrical double-layer capacitors. Angew Chem Int Ed 2012;51:9994–10024. [383] Gogotsi Y, Simon P. True performance metrics in electrochemical energy storage. Science 2011;334:917–8. [384] Zhi L, Li T, Yu H, et al. Hierarchical graphene network sandwiched by a thin carbon layer for capacitive energy storage. Carbon 2017;113:100–7. [385] Ma W, Chen S, Yang S, et al. Flexible all-solid-state asymmetric supercapacitor based on transition metal oxide nanorods/reduced graphene oxide hybrid fibers with high energy density. Carbon 2017;113:151–8. [386] Burke A, Miller M. Proceedings of 24th international electric vehicle symposium EVS-24, Stavanger; 2009. [387] Abeywardana DBW, Hredzak B, Agelidis VG, Demetriades GD. Supercapacitor sizing method for energy-controlled filter-based hybrid energy storage systems. IEEE Trans Power Electron 2017;32:1626–37. [388] Dotelli G, Ferrero R, Stampino PG, Latorrata S, Toscani S. Supercapacitor sizing for fast power dips in a hybrid supercapacitor-PEM fuel cell system. IEEE Trans Instrum Meas 2016;65:2196–203. [389] Kupeman A, Mellincovsky M, Leman C, et al. Supercapacitor sizing based on desired power and energy performance. IEEE Trans Power Electron 2014;29:5399–405. [390] APowerCap Technologies. Available from: http://www.apowercap.com/index.php?pg=3&land=eng. [391] Beneito R, Vilaplana J, Gisbert S. Electric toy vehicle powered by a PEMFC stack. Int J Hydrogen Energy 2007;32:1554–8. [392] Sakabe Y. Multilayer ceramic capacitors. Curr Opin Solid State Mater Sci 1997;5:584–7. [393] Yalcinoz T, Alam AS. Improved dynamic performance of hybrid PEM fuel cells and ultracapacitors for portable application. Int J Hydrogen Energy 2008;33:1932–40. [394] Chen SC, Wong KL. Planar strip monopole with a chip-capacitor-loaded loop radiating feed for LTE/WWAN slim mobile phone application. Microwave Opt Technol Lett 2011;53:952–8. [395] Ban YL, Chen JH, Sun SC, Li JLW, Guo JH. Printed wideband antenna with chip-capacitor-loaded inductive strip for LTE/GSM/UMTS/WWAN wireless USB dongle applications. Prog Electromagn Res B 2012;128:313–29. [396] Siami S, Joubert C, Glaize C. High frequency model for power electronics capacitors. IEEE Trans Power Electron 2002;16:157–66. [397] Capozzo L, Attolico G, Cicirelli G. Designing and developing simple low cost mobile robots for autonomous navigation. Intell Autom Soft Comput 2006;173–81. [398] Kwong J, Ramadass YK, Verma N, Chandrakasan AP. A 65 nm sub-Vt microcontroller with integrated SRAM and switched capacitor DC-DC converter. IEEE J Solid-State Circuits 2009;44:115–26. [399] Sinangil ME, Poulton JW, Fojtik MR, et al. A 28 nm 2 Mbit 6 T SRAM with highly configurable low-voltage write-ability assist implementation and capacitor-based sense-amplifier input offset compensation. IEEE J Solid-State Circuits 2016;51:557–67. [400] Lu YL, Chen WJ. A 0.7 V input output-capacitor-free digitally controlled low-dropout regulator with high current efficiency in 0.35 um CMOS technology. Microelectron J 2012;43:756–65. [401] Ogawa S, Watanabe K. Clock-feedthrough compensated switched-capacitor circuits. Electron Lett 1991;27:2045–6. [402] Gonzafez-Gil A, Palacin R, Batty P, Powell JP. A systems approach to reduce urban rail energy consumption. Energy Convers Manage 2014;80:509–24. [403] Gulmann G. Hybrid electric vehicles and electrochemical storage systems – a technology push-pull couple. J Power Sources 1999;84:275–9. [404] Wang L. Realization and analysis on good fuel economy and kinetic performance of a low-cost hybrid electric vehicle. Chinese J Mechan Eng 2011;24:774–89. [405] Becherif M, Ramadan HS, Ayad MY, Hissel D, Desident U, Antonelli M. Efficient start-up energy management via nonlinear control for eco-traction systems. Appl Energy 2017;187:899–909. [406] Catherino HA, Burgel JF, Shi PL, Rusek A, Zou X. Hybrid power supplies: a capacitor-assisted battery. J Power Sources 2006;162:965–70. [407] Catherino HA, Shi PL, Rusek A. Modeling and simulation of capacitor assisted engine starting. In: Proceedings of the summer computer simulation conference, Portland, OR, July 21–25. San Diego, CA: SCS; 1996. p. 549–53. [408] Yuhimenko V, Averbukh M, Agranovich G, Kupeman A. Dynamics of supercapacitor bank with uncontrolled active balancer for engine starting. Energy Convers Manage 2014;88:106–12. [409] Shida Y, Kanda M, Ohta K, Furuta S, Ishii J. Development of inverter and power capacitors for mild hybrid vehicle (MHV)-Toyota “CROWN”. Int J Auto Technol 2003;4:41–5. [410] Castaings A, Lhomme W, Trigui R, Bouscaryrol A. Comparison of energy management strategies of a battery/supercapacitors system for electric vehicle under real-time constraints. Appl Energy 2016;163:190–200. [411] Garcia-Trivirio P, Torreglosa JP, Fernandz-Ramirez LM, Jurado F. Control strategies for high-power electric vehicles powered by hydrogen fuel cell, battery and supercapacitor. Expert Syst Appl 2013;40:4791–804. [412] Capasso C, Veneri O. Laboratory bench to test ZEBRA battery plus super-capacitor based propulsion systems for urban electric transportation. Energy Procedia 2015;75:1956–61. [413] Sun L, Zhang Z. Design, implementation and characterization of a novel bi-directional energy conversion system on DC motor drive using super-capacitors. Appl Energy 2015;153:101–11. [414] González-Gil A, Palacin B, Batty P. Sustainable urban rail systems: strategies and technologies for optimal management of regenerative braking energy. Energy Convers Manage 2013;75:374–88. [415] Burke A. Ultracapacitors: why, how, and where is the technology. J Power Sources 2000;91:37–50. [416] Zhu WH, Tatarchuk BJ. Characterization of asymmetric ultracapacitors as hybrid pulse power devices for efficient energy storage and power delivery applications. Appl Energy 2016;169:450–68. [417] Veneri O, Capasso C, Patalano S. Experimental study on the performance of a ZEBRA battery based propulsion system for urban commercial vehicles. Appl Energy 2017;185:2005–18. [418] Li Q, Chen W, Liu Z, Li M, Ma L. Development of energy management system based on a power sharing strategy for a fuel cell–battery–supercapacitor hybrid tramway. J Power Sources 2015;279:267–80. [419] Zandi M, Payman A, Martin J-P, Pierfederici S, Davat B, Meibody-Tabar F. Energy management of a fuel cell/supercapacitor/battery power source for electric vehicular applications. IEEE Trans Veh Technol 2011;60:433–43. [420] Jung H, Wang H, Hu T. Control design for robust tracking and smooth transition in power systems with battery/supercapacitor hybrid energy storage devices. J Power Sources 2014;267:566–75. [421] Quakenbush T. Portable power system weight reduction with hybrid methanol fuel cells and Li-ion batteries. In: Proceedings of the 40th Power Sources Conference, June 10–13, 2002. p. 88–91. [422] Tarnowski DJ. Hybrid power supplies combining ultracapacitors with direct methanol fuel cells for portable electronics. In: Proceedings of the 40th Power Sources Conference, June 10–13, 2002 p. 92–5.

Supercapacitors

693

[423] Teymourfar R, Asaei B, Iman-Eini H, Fard RN. Stationary super-capacitor energy storage system to save regenerative braking energy in a metro line. Energy Convers Manage 2012;56:206–14. [424] Al-Bahrani AH, Malik NH. Selection of the excitation capacitor for dynamic braking of induction machines. Electric Power Appl 1993;140:1–6. [425] Popescu M, Miller TJE, McGlip M, Strappazzon G, Trivillin N, Santarossa R. Asynchronous performance analysis of a single-phase capacitor-start, capacitor-run permanent magnet motor. IEEE Trans Energy Convers 2005;20:142–50. [426] Hannan MA, Azidin FZ, Mohamed A. Hybrid electric vehicles and their challenges: a review. Renew Sustain Energy Rev 2014;29:135–50. [427] Jorgensen K. Technologies for electric, hybrid and hydrogen vehicles: electricity from renewable energy sources in transport. Utilities Policy 2008;16:72–9. [428] Lachhab I, Krichen L. Impact of ultra-capacitor sizing optimization on fuel cell hybrid vehicle. Int J Renew Energy Res 2015;5:151–9. [429] Azib T, Bethoux O, Remy G, Marchand C. Saturation management of a controlled fuel-cell/ultracapacitor hybrid vehicle. IEEE Trans Veh Technol 2011;60:4127–38. [430] Zheng CH, Oh CE, Park YI, Cha SW. Fuel economy evaluation of fuel cell hybrid vehicles based on equivalent fuel consumption. Int J Hydrogen Energy 2012;37:1790–979. [431] Schaltz E, Khaligh A, Rasmussen PO. Influence of battery/ultracapacitor energy-storage sizing on battery lifetime in a fuel cell hybrid electric vehicle. IEEE Trans Veh Technol 2009;58:3882–91. [432] Kisacikoglu MC, Uzunoglu M, Alam MS. Load sharing using fuzzy logic control in a fuel cell/ultracapacitor hybrid vehicle. Int J Hydrogen Energy 2009;34:1497–507. [433] Thounthong P, Rael S, Davat B. Analysis of supercapacitor as second source based on fuel cell power generation. IEEE Trans Energy Convers 2009;24:247–55. [434] Payman A, Pierfederici S, Meibody-Tabar F. Energy control of supercapacitor/fuel cell hybrid power source. Energy Convers Manage 2008;49:1637–44. [435] Baisden AC, Emadi A. Advisor-based model of a battery and an ultra-capacitor energy source for hybrid electric vehicles. IEEE Trans Veh Technol 2004;53:199–205. [436] Budde-Meiwes H, Drillkens J, Lunz B, et al. A review of current automotive battery technology and future prospects. Proc Inst Mech Eng D J Automob Eng 2013;227:761–6. [437] Hannan MA, Azidin FA, Mohamed A. Hybrid electric vehicles and their challenges: a review. Renew Sustain Energy Rev 2014;29:135–50. [438] Coenen P, Leemans F, Mulder G. Applying large electric double layer capacitor systems. J Appl Electrochem 2014;44:533–42. [439] Erdinc O, Vural B, Uzunoglu M. A wavelet-fuzzy logic based energy management strategy for a fuel cell/battery/ultra-supercapacitor hybrid vehicular power system. J Power Sources 2009;194:369–80. [440] Paldini V, Donateo T, Risi A, Laforgia D. Super-capacitors fuel-cell hybrid electric vehicle optimization and control strategy development. Energy Convers Manage 2007;48:3001–8. [441] Omar N, Daowd M, Hegazy O, et al. Assessment of lithium-ion capacitor for using in battery electric vehicle and hybrid electric vehicle applications. Electrochim Acta 2012;86:305–15. [442] Vinot E, Trigui R. Optimal energy management of HEVs with hybrid storage system. Energy Convers Manage 2013;76:437–52. [443] Cheng Y, Joeri VM, Lataire P. Research and test platform for hybrid electric vehicle with the super capacitor based energy storage. Power Electron Appl 2008; doi:10.1109/EPE.2007.4417332. [444] Aouzellag H, Ghedamsi K, Aouzellag D. Energy management and fault tolerant control strategies for fuel cell/ultra-capacitor hybrid electric vehicles to enhance autonomy, efficiency and life time of the fuel cell system. Int J Hydrogen Energy 2015;40:7204–13. [445] Nafisi H, Agah SMM, Abyaneh HA, Abedi M. Two-stage optimization method for energy loss minimization in microgrid based on smart power management scheme of PHEVs. IEEE Trans Smart Grid 2015;7:1268–76. [446] Amjadi Z, Williamson SS. Power-electronics-based solutions for plug-in hybrid electric vehicle energy storage and management systems. IEEE Trans Ind Electron 2010;57:608–16. [447] Amjadi Z, Williamson SS. Prototype design and controller implementation for a battery–ultracapacitor hybrid electric vehicle energy storage system. IEEE Trans Smart Grid 2012;3:332–40. [448] Chiu HJ, Lin LW, Pan PL, Tseng MH. A novel rapid charger for lead-acid batteries with energy recovery. IEEE Trans Power Electron 2006;21:640–7. [449] Wen H, Xiao W, Wen X, Armstrong P. Analysis and evaluation of DC-link capacitors for high-power density electric vehicle drive system. IEEE Trans Veh Technol 2012;61:2950–64. [450] Long B, Lim ST, Bai ZF, Ryu JH, Chong KT. Energy management and control of electric vehicles, using hybrid power source in regenerative braking operation. Energies 2014;7:4300–15. [451] Wu Y, Gao H. Optimization of fuel cell and supercapacitor for fuel-cell electric vehicles. IEEE Trans Veh Technol 2006;55:1748–55. [452] Fadil HE, Giri F, Guerrero JM, Tahri A. Modeling and nonlinear control of a fuel cell/supercapacitor hybrid energy storage system for electric vehicles. IEEE Trans Veh Technol 2014;63:3011–8. [453] Schroeder JC, Fuchs FW. General analysis and design guideline for a battery buffer system with DC/DC converter and EDLC for electric vehicles and its influence on efficiency. IEEE Trans Power Electron 2015;30:922–32. [454] Amjadi Z, Willamson SS. Digital control of a bidirectional DC/DC switched capacitor converter for hybrid electric vehicle energy storage system application. IEEE Trans Smart Grid 2014;5:158–66. [455] Zhang C, Min H, Yu Y, et al. Using CPE function to size capacitor storage for electric vehicles and quantifying battery degradation during different driving cycles. Energies 2016;9:903. [456] Logerais PO, Riou O, Camara MA, Durastanti JF. Study of photovoltaic energy storage by supercapacitors through both experimental and modelling approaches. J Sol Energy 2013;2013:659014. [457] Wei T, Wang S, Qi Z. Design of supercapacitor based ride through system for wind turbine pitch systems. In: Proceedings of the 2007 international conference on electrical machines and systems, vol. 1, 2007. p. 1524–7. [458] Mahela PP, Shaik AG. Comprehensive overview of grid interfaced wind energy generation systems. Renew Sustain Energy Rev 2016;57:260–81. [459] Thounthong P, Luksanasakul A, Koseeyaporn P, Davat B. Intelligent model-based control of a standalone photovoltaic/fuel cell power plant with supercapacitor energy storage. IEEE Trans Sustain Energy 2013;4:240–9. [460] Martina F, Pugliese M, Serantoni M, et al. Large area self-powered semitransparent trifunctional device combining photovoltaic energy production, lighting and dynamic shading control. Sol Energy Mater Sol Cells 2017;160:435–43. [461] Thounthong P, Chunkag V, Sethakul P, Sikkabut S, Piefederici S, Davat B. Energy management of fuel cell/solar cell/supercapacitor hybrid power source. J Power Sources 2011;196:313–24. [462] Liu W, Wang K, Chung HS, Chuang ST. Modeling and design of series voltage compensator for reduction of DC-link capacitance in grid-tie solar inverter. IEEE Trans Power Electron 2015;30:2534–48. [463] Stauth JT, Seeman MD, Kesarwani K. A resonant switched-capacitor IC and embedded system for sub-module photovoltaic power management. IEEE J Solid-State Circuits 2012;47:3043–54. [464] Mutoh N, Inoue T. A control method to charge series-connected ultraelectric double-layer capacitors suitable for photovoltaic generation systems combining MPPT control method. IEEE Trans Ind Appl 2007;54:374–83. [465] Ahmad MW, Agarwal N, Anand S. Online monitoring technique for aluminum electrolytic capacitor in solar PV-based DC system. IEEE Trans Ind Appl 2016;63:7059–66. [466] Agarwal N, Arya A, Ahmad MW, Anand S. Lifetime monitoring of electrolytic capacitor to maximize earnings from grid-feeding PV system. IEEE Trans Ind Electron 2016;63:7049–58.

694

Supercapacitors

[467] Onar OC, Uzunoglu M, Alam MS. Modeling, control and simulation of an autonomous wind turbine/photovoltaic/fuel cell/ ultra-capacitor hybrid power system. J Power Sources 2008;185:1273–83. [468] Singh M, Khadkikar V, Chandra A, Varma RK. Grid interconnection of renewable energy sources at the distribution level with power–quality improvement features. IEEE Trans Power Delivery 2011;26:307–15. [469] Kayai P, Chanda CK. Strategic approach for reinforcement of intermittent renewable energy sources and capacitor bank for sustainable electric power distribution system. Int J Electr Power Energy Syst 2016;83:335–51. [470] Benson CL, Magee CL. On improvement rates for renewable energy technologies: solar PV, wind turbines, capacitors, and batteries. Renew Energy 2014;68:745–51. [471] Xu Q, Cai H, Tang G, Yukita K, Ichiyanagi K. Charge evaluation of EDLC for autonomous microgrid energy storage. Electr Eng 2011;93:1–8. [472] Wang L, Lee DJ. Coordination control of an AC-to-DC converter and a switched excitation capacitor bank for an autonomous self-excited induction generator in renewable-energy systems. IEEE Trans Ind Appl 2014;50:2828–36. [473] Sadigh AK, Dargahi V, Corzine K. New flying-capacitor-based multilevel converter with optimized number of switches and capacitors for renewable energy integration. IEEE Trans Energy Convers 2016;31:846–59. [474] Muyeen SM, Takahashi R, Tamura J. Wind power and hydrogen generation system with cooperatively controlled three-level NPC-VSC based energy capacitor. Int Trans Electr Energy Syst 2010;20:1071–81. [475] Dargahi V, Sadigh AK, Benayagamoorthy GK, Corzine K. Hybrid double flying capacitor multicell converter and its application in grid-tied renewable energy resources. IET Gener Transm Distrib 2015;9:947–56. [476] Ehteshami SMM, Chan SH. The role of hydrogen and fuel cells to store renewable energy in the future energy network-potentials and challenges. Energy Policy 2014;73:103–9. [477] Parastar A, Candomkar A, Seok JK. High-efficiency multilevel flying-capacitor DC/DC converter for distributed renewable energy systems. IEEE Trans Ind Electron 2015;62:7620–30. [478] Onar OC, Uzunoglu M, Alam MS. Dynamic modeling, design and simulation of a wind/fuel cell/ultra-capacitor-based hybrid power generation system. J Power Sources 2006;161:707–22. [479] Abeywardana DBW, Hredzak B, Agelidis VG. A fixed-frequency sliding mode controller for a boost-inverter-based battery-supercapacitor hybrid energy storage system. IEEE Trans Power Electron 2016;32:668–80. [480] Aktas A, Ethan K, Ozdemir S, Ozdemir E. Experimental investigation of a new smart energy management algorithm for a hybrid energy storage system in smart grid application. Electr Power Syst Res 2017;144:185–96. [481] Becherif M, Ayad MY, Djerdir A, Miraoui A. Electrical train feeding by association of supercapacitors photovoltaic and wind generators. In: Proceedings of the 2007 international conference on clean electrical power, Capri, Italy, May 21–23 2007. p. 55–64. [482] Xu J, Xie S, Zhang B. Stability analysis and improvement of capacitor current active damping of LCL filter in grid-connected applications. J Power Electron 2012;12:1–11. [483] Li X, Wu X, Geng Y, Yuan X, Xia C, Zhang X. Wide damping region for LCL-type grid-connected inverter with an improved capacitor-current-feedback method. IEEE Trans Power Electron 2014;30:5247–59. [484] Liu CM, De Lisio MP, Moussessian A, Rutledge DB. Stability of grid amplifiers. IEEE Trans Microw Theory Techn 1998;46:769–74. [485] Kahrobaeian A, Mohamed YARI. Robust single-loop direct current control of LCL-filtered converter-based DG Units in grid-connected and autonomous microgrid modes. IEEE Trans Power Syst 2014;29:5605–19. [486] Bidram A, Hamedani-golshan M, Davoudi A. Capacitor design considering first swing stability of distributed generations. IEEE Trans Power Syst 2012;27:1941–8. [487] Singh AK, Singh R, Pal BC. Stability analysis of networked control in smart grids. IEEE Trans Smart Grid 2015;6:381–90. [488] Pan D, Ruan X, Bao C, Li W, Wang X. Capacitor-current-feedback active damping with reduced computation delay for improving robustness of LCL-type grid-connected inverter. IEEE Trans Power Electron 2013;29:3414–27. [489] Bao C, Ruan X, Wang X, Li W, Pan D, Wang K. Step-by-step controller design for LCL-type grid-connected inverter with capacitor-current-feedback active-damping. IEEE Trans Power Electron 2013;29:1239–53. [490] Lorzadeh I, Abyaneh HA, Savaghebi M, Bakhshai A, Guerrero JM. Capacitive current feedback-based active resonance damping strategies for digitally-controlled inductive-capacitive-inductive filtered grid-connected inverters. Energies 2016;9:642. [491] Yang WM, Wang XG, Li XY, Liu ZY. An active power sharing method among distributed energy sources in an islanded series micro-grid. Energies 2014;7:7878–92. [492] Darabian M, Jalilvand A. A power control strategy to improve system stability in the presence of wind farms using FACTs devices and predictive control. Int J Electr Power Energy Syst 2017;85:50–66. [493] Khayyam H, Ranjbarzadeh H, Marano V. Intelligent control of vehicle to grid power. J Power Sources 2012;201:1–9. [494] Rufer A. A supercapacitors-based energy-storage system for elevators with soft commutated interface. IEEE Trans Ind Appl 2002;38:1151–9. [495] Belhachemi F, Raël S, Davat B. A physical based model of power electric double-layer supercapacitors. In: Proceedings of the industry applications conference, Rome, Italy, October 8–1, 2000. p. 3069–76. [496] Birx DL, Das PP, Fomenkov IV, Partlo WN, Watson TA. Pulse power generating circuit with energy recovery. US Patent US5729562 A, 1998. [497] Xu Y, Xiang YM, Wan Q, et al. A dual-capacitors type energy recovery power systems for repetitive pulsed high magnetic fields. J Low Temp Phys 2013;170:488–95. [498] Hu Y, Cao W, Finney SJ, Xiao W, Zhang F, McLoone SF. New modular structure DC-DC converter without electrolytic capacitors for renewable energy applications. IEEE Trans Sustain Energy 2014;5:1184–92. [499] Weimer JA, Kazimierczuk MK, Robert CC II. Super capacitor battery clone. US Patent US5850113 A, 1998. [500] Iturriz MF, Richardeau F, Meynard T, Helali H. Power rectifier circuit and systems, associated method, and aircraft comprising such a circuit or systems. US Patent US8467206 B2, 2013. [501] Bauer HA, Zybala A, Rathge C. System for contactless energy transfer, use of a system for contactless energy transfer and vehicle with a system for contactless energy transfer between a first vehicle component and a second vehicle component. US Patent US20130278053 A1, 2013. [502] Foch E, Langlois O. Device and method for generating a back-up electricity supply on board an aircraft. US Patent US20100270858, 2010. [503] Marek AJ. High-speed capacitor discharge circuit suitable for the protection of detonation devices. US Patent US4536693 A, 1985. [504] Okamura M, Mogami A. Electrical storage capacitor system having initializing function. US Patent US6404170 B2, 2002. [505] Rieckmann N Monitoring circuit for a door. US Patent US7414530 B2, 2008. [506] Khatre M, Jack AG. Simulation of PMSM VSI drive for determination of the size limits of the DC-link capacitor of aircraft control surface actuator drives. In: International conference on power electronic, drives and energy systems, 2006, doi:10.1109/PEDES.2006.344430. [507] Langlois O, Foch E, Prisse L, Roboam X. System and method for supplying power for actuators on board an aircraft. US Patent US7923857 B2, 2011. [508] Wisch B, Rowold L. Energy buffer device for an aircraft. US Patent US8368244 B2, 2013. [509] Becker T, Elefsiniotis A, Schmid U. Energy harvester, aircraft component comprising the energy harvester and an aircraft comprising the energy harvester or the aircraft component. US Patent US20150180371 A1, 2015. [510] Lee B. Constant boosted voltage generator circuit for feedback switches in a switched capacitor circuit. US Patent US7372319 B1, 2008. [511] Roux S, Pasquier M. Electronic circuit card and corresponding signal acquisition and generation system, including one or more programmable digital matrix switches. US Patent US20160308699 A1, 2016. [512] Bonnet G, Rousset D. Device for balancing the power supplied by power generators. US Patent US8148850 B2, 2012.

Supercapacitors

695

[513] Abdi B, Alimardani A, Ghasemi R, Mirtalaei SMM. Energy storage selection for LEO satellites. Int J Mach Learning Comput 2013;3:287–90. [514] Abdi B, Alimardani A, Moghani J, Hosseinian SH. A comparative approach within chemical battery, flywheel and supercapacitor base energy storage system for LEO satellite applications. Int Rev Electr Eng 2011;6:739–44. [515] Huang Y, Zhu M, Huang Y, et al. Multifunctional energy storage and conversion devices. Adv Mater 2016;28:8344–64. [516] Shimizu T, Underwood C. Super-capacitor energy storage for micro-satellites: feasible and potential mission application. Acta Astronaut 2013;85:138–54. [517] Uno M, Tanaka K. Spacecraft electrical power system using lithium-ion capacitors. IEEE Trans Aerosp Electron Syst 2013;49:175–88. [518] Neto LPS, Rossi JO, Barroso JJ, Schamiloglu E. High-power RF generation from nonlinear transmission lines with barium titanate ceramic capacitors. IEEE Trans Plasma Sci 2016;44:3424–31. [519] Kobayashi T, Okada H, Masuda T, Maeda R, Itoh T. A digital output accelerometer using MEMS-based piezoelectric accelerometers and arrayed CMOS inverters with satellite capacitors. Smart Mater Struct 2011;20:065017. [520] Lin JF, Chang SJ, Chiu CF, Tsai HH, Wang JJ. Low-power and wide-bandwidth cyclic ADC with capacitor and opamp reuse techniques for CMOS image sensor application. IEEE Sensor J 2009;9:2044–54. [521] Conte M. Supercapacitors technical requirements for new applications. Fuel Cells 2010;10:806–18. [522] Halpin SM, Ashcraft SR. Design considerations for single-phase uninterruptible power supply using double layer capacitors as the energy storage element. In: Proceedings of the industry applications conference, vol. 4, San Diego, CA; 1996. p. 2396–4203. [523] Zhan Y, Guo Y, Zhu J, Li L. Power and energy management of grid/PEMFC/battery/supercapacitor hybrid power sources for UPS application. Int J Electr Energy Syst 2015;67:598–612. [524] Zhang X, Xue H, Xu Y, Chen H, Tan C. An investigation of an uninterruptible power supply (UPS) based on supercapacitor and liquid nitrogen hybridization system. Energy Convers Manage 2014;85:784–92. [525] Maciel RS, de Freitas LC, Coelho EAA, Vieira JB, Gomes d Freitas LC. Front-end converter with integrated PFC and DC-DC functions for a fuel cell UPS with DSPbased control. IEEE Trans Power Electron 2015;30:4175–88. [526] Zhan Y, Guo Y, Zhu J, Li L. Performance comparison of input current ripple reduction methods in UPS applications with hybrid PEM fuel cell/supercapacitor power sources. Int J Electr Power Energy Syst 2015;64:96–103. [527] Tao H, Duarte JL, Hendrix MAM. Line-interactive UPS using a fuel cell as the primary source. IEEE Trans Ind Electron Sci 2008;55:3012–21. [528] Lam LT, Louey R. Development of ultra-battery for hybrid-electric vehicle applications. J Power Sources 2006;158:1140–8. [529] Conte M, Genovese A, Ortenzi F, Vellucci F. Hybrid battery-supercapacitor storage for an electric forklift: a life-cycle cost assessment. J Appl Electrochem 2014;44:523–32. [530] Minav TA, Virtanen A, Laurila L, Pyrhonen J. Storage of energy recovered from an industrial forklift. Automat Constr 2012;22:506–15. [531] Minav T, Hanninen H, Sinkkonen A, Laurila L, Pyrhonen J. Electric of hydraulic energy recovery systems in a reach truck – a comparison. J Mech Eng 2014;60:232–40. [532] Anonymous. ‘Plug and play’ power system cuts forklift energy bill by 50%. Professional Eng 2007;20.18:56. [533] Kim SM, Sui SK. Control of rubber ytred gantry crane with energy storage based on supercapacitor bank. IEEE Power Electron Soc 2006;21:1420–7. [534] Musolino V, Pievatolo A, Tironi E. A statistical approach to electrical storage sizing with application to the recovery of braking energy. Energy 2011;36:6697–704.

Further Reading Bagotsky VS, Skundin AM, Volfkovich YM. Electrochemical power sources: batteries, fuel cells, and supercapacitors. New York: Wiley; 2015. ISBN: 978-1-118-46023-8. Bamford CH, Tipper CFH, Compton RG. Electrode kinetics: principles and methodology. Amsterdam: Elsevier; 1986. ISBN: 9780080868202. Béguin F, Fra˛ckowiak E. Supercapacitors: materials, systems, and applications. New York: Wiley; 2013. ISBN: 9783527328833. Bondavalli P. Graphene and related nanomaterials, properties and applications. Amsterdam: Elsevier; 2017. ISBN: 9780323481014. Brett CMA, Brett AMO. Electrochemistry, principles, methods, and applications. UK: Oxford University Press; 1993. ISBN 0-19-855388-9. Compton RG. Electrode kinetics: reactions. Amsterdam: Elsevier; 1987. ISBN: 9780080868219. Dubal D, Gomez-Romero P. Metal oxides in supercapacitors. Amsterdam: Elsevier; 2017. ISBN: 9780128111697. Kim BK, Sy S, Yu A, Zhang J. Handbook of clean energy systems. New York: Wiley; 2014. ISBN: 9781118991978. Koryta J, Dvorák J, Kavan L. Principles of electrochemistry. Chichester: Wiley; 1993. ISBN 0-471-93838-6. Kularatna N. Energy storage devices for electronic systems, rechargeable batteries and supercapacitors. Amsterdam: Elsevier; 2014. ISBN: 9780124079472. Nguyen MY, Nguyen DH, Yoon YT. A new battery energy storage charging/discharging scheme for wind power producers in real-time market. Energies 2012;5:5439–52.

Relevant Websites http://www.electronicdesign.com/power/can-supercapacitors-surpass-batteries-energy-storage Can Supercapacitors Surpass Batteries for Energy Storage. https://en.wikipedia.org/wiki/Electric_double-layer_capacitor Electric double-layer capacitor. https://en.wikipedia.org/wiki/Electrochemistry Electrochemistry. https://en.wikipedia.org/wiki/Electrode Electrode. https://en.wikipedia.org/wiki/Graphene Graphene. https://www.graphene-info.com/graphene-supercapacitors Graphene-Info: Graphene Supercapacitors: Introduction and News.. http://www.idtechex.com/research/reports/supercapacitor-technologies-and-markets-2016-2026-000486.asp IDTechEx, Supercapacitor Technologies and Markets 2018-2028. https://en.wikipedia.org/wiki/Pseudocapacitor Pseudocapacitors. https://en.wikipedia.org/wiki/Supercapacitor Supercapacitors. https://www.nature.com/subjects/supercapacitors Supercapacitors-Latest Research and Review.

2.22 Piezoelectric Materials Mehmet Sunar, King Fahd University of Petroleum & Minerals, Dhahran, Saudi Arabia r 2018 Elsevier Inc. All rights reserved.

2.22.1 Introduction 2.22.2 Thermopiezoelectricity 2.22.3 Piezoelectricity 2.22.4 Piezoelectric Energy Harvesting 2.22.5 Technology Issues 2.22.5.1 Design and Manufacture 2.22.5.2 PZT Piezoceramics 2.22.5.3 PVDF Piezoelectric Polymer 2.22.5.4 Dynamic Performance and Modeling 2.22.5.5 Various Other Considerations 2.22.6 Piezomagnetism 2.22.6.1 Piezomagnetic/Magnetostrictive Materials 2.22.6.2 Applications 2.22.6.3 Linear Theory of Piezomagnetism 2.22.7 Illustrative Cases for Piezoelectricity 2.22.7.1 Piezoelectric Bimorph 2.22.7.2 Piezoelectric Sensing System 2.22.7.3 Piezoelectro-Magnetic System 2.22.7.3.1 Analytical approach 2.22.7.3.2 Finite element approach 2.22.8 Closing Remarks Acknowledgment References Further Reading Relevant Websites

Nomenclature A B C c D d E EG e F f G H h I i J K L l ℓ M mg mt

696

Magnetic potential vector Vector of magnetic flux density Capacitance Matrix of elastic stiffness coefficients Electric displacement vector Distance of piezoelectric sensor Electric field vector Thermodynamic potential energy Matrix of piezoelectric coefficients Global force vector Natural frequency in Hz Thermodynamic potential Vector of magnetic field intensity Thickness Modulus of elasticity Current Vector of volume current density Thermal conductivity matrix Length or inductance Length Matrix of piezomagnetic stress coefficients Bending moment Magnet mass for electromagnetic energy harvester Tip mass for piezoelectric energy harvester

697 698 700 701 703 703 703 704 705 708 708 708 709 709 711 711 712 714 716 716 717 717 717 719 719

Nu Nf P Pb Ps Pc q R S s T t u V Vp W w Z a e Z l m o f r

Displacement shape function matrix Electric potential shape function matrix Vector of pyroelectric coefficients Vector of body forces Vector of surface forces Vector of concentrated forces Heat flux vector Circuit resistance Strain vector Laplace transformation variable Stress vector Thickness Global displacement vector Applied voltage Voltage generated by piezoelectric bimorph Heat source density Width Complex impedance Thermal constitutive coefficient Permittivity matrix Entropy per unit volume Vector of thermal expansion coefficients Matrix of material permeability Natural frequency in rad/s Global electric potential vector Mass density

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00248-0

Piezoelectric Materials

rv s Θ

2.22.1

Volume charge density Surface charge Absolute temperature

y ζ

697

Temperature change Damping ratio

Introduction

Piezoelectricity defines a relation between mechanical and electric fields. In other words, in certain media, a mechanical effect such as strain results in an electric output like potential. This coupling is known as direct piezoelectricity, which is the basis for piezoelectric sensors and transducers. The reverse coupling, which is a mechanical output due to an applied electric effect such as voltage, is called converse piezoelectricity. The converse piezoelectricity, although existed and known, is “rediscovered” in recent years and has then found many applications to be used as actuators in control science and technologies. The direct piezoelectricity was discovered more than a century ago by Curie brothers [1]. The converse piezoelectricity was then theoretically predicted by Lippman [2]. The direct piezoelectric effect was used earlier to generate acoustic waves from piezoelectric crystal quartz, using a method developed by Langevin [3]. Later in the last 2–3 decades, both piezoelectric phenomena, direct and converse, are extensively utilized in sensing, measurement and actuation of various systems in vastly different fields. Recent survey articles in this area include the work by Ferreira et al. [4] where an overview of multifunctional materials together with piezoelectric media is presented and the corresponding literature is listed. A review of damage detection methods for wind turbine blades is carried out by Li et al. [5]. Four described methods make use of sensing mechanisms via fiber optic and piezoelectric sensors. Recent developments on piezoelectric composite technology for high power and temperature applications are presented by Lee et al. [6] with an outline on expected future trends of research in this direction. A similar review article is compiled by Jiang et al. [7], where high-temperature piezoelectric sensors resisting up to 1250oC for different applications are reported. Among many, piezoelectric sensors are recently employed in modal characterization of flat plates [8], structural acoustic sensing [9], structural health monitoring for temperature assessment [10] and for failure prevention of bridges [11], and structural damage identification [12]. Piezoelectric actuators, on the other hand, are used in active vibration control of plates [13], flexible manipulators [14] and structures [15], and in micro- and nano-positioning applications [16]. Another way of looking at the phenomenon of piezoelectricity is that it can be thought as conversion of energy from one form to another, i.e., from mechanical energy to electric energy or vice versa in this case. Hence, a piezoelectric material can be considered as an energy converter between mechanical and electric fields. In a much broader sense, in suitable media where socalled the phenomenon of magneto-thermopiezoelectricity exists, magnetic, thermal, mechanical and electric fields interact with each other so that various types of energies are transformed into different forms. Thus, energy conversions between the abovestated fields are possible via various phenomena in suitable media, such as piezoelectricity between mechanical and electric fields [17], piezomagneticity between mechanical and magnetic fields [18], thermopiezoelectricity between mechanical, thermal and electric fields [19], and thermopiezomagnecity between mechanical, thermal and magnetic fields [20]. In general, the structures/systems mounted with active (smart) materials are referred to as smart systems/structures. These systems/structures contain active members, which carry loads, and function as sensors and actuators that are distributed in nature. The phenomena in the above-mentioned piezoelectric and piezomagnetic materials with and without thermal effects fall into the category of distributed sensing and actuation utilized in smart structures. The conversion of mechanical energy to electric one at small magnitudes, e.g. from milli- to micro-energy levels, by a piezoelectric material due to mechanical effects is recently called the piezoelectric energy harvesting as in review papers by Kim et al. [21] and Ahmed et al. [22]. The energy harvesting can be used as a natural source for power supply, which also helps to reduce chemical wastes from conventional batteries. Hence, low-power energy harvesters are introduced for collecting energy from different ambient sources. Piezoelectric materials in particular have drawn a great deal of attention in energy harvesting due to their unique abilities to capitalize the ambient vibrations to produce electric voltage. The useful property of energy harvesting via piezoelectricity can be exploited to serve as a self-energy or power source for portable devices and wireless sensor networking systems [23]. Energy harvesting based on various fields including piezoelectric, triboelectric, pyroelectric, thermoelectric, and photovoltaic effects is summarized by Lee et al. [24]. Hybrid cell technologies to simultaneously generate electricity by means of mechanical, thermal and solar energies are discussed. Integration of energy harvesters and energy storage devices is also presented in the context of self-charging power cells. Research activities on vibration-based micro-power generation or energy harvesting are surveyed by Siddique et al. [25]. Piezoelectric and electromagnetic transducer systems by which mechanical vibrations are converted into electrical energy are considered in their work. Nanoscale energy harvesters or nano-generators are investigated by Crossley et al. [26] using piezoelectric materials. Some theoretical and experimental aspects of piezoelectric energy harvesting are discussed using polymer-based nano-generators with an emphasis on mechanical resonance at low-frequency ambient vibrations. Some current piezoelectric materials include lithium niobate (LiNbO3) [27], zinc oxide (ZnO) [28], aluminum nitride (AlN) [29], lead zirconate titanate (PZT) and polyvinylidene fluoride (PVDF or PVF2). Among them, the PZT is a ceramic with high piezoelectric, dielectric and elasticity coefficients [30]. The PVDF, discovered by Kawai in 1969 [31], is a polymer known as being flexible, durable, lightweight, easy to fabricate and lead free and biocompatible [26,32]. Favorable piezoelectric and pyroelectric properties of these two materials properties [33,34] have led to extensive research and development activities in different areas

698

Piezoelectric Materials

including energy harvesting [35]. There have been efforts to integrate them in the form of PZT-PVDF piezoelectric composites in order to obtain properties that may not be attainable by a single piezoelectric material [33]. In the next section, governing equations of linear thermopiezoelectricity, piezoelectricity with thermal effects, are first provided. This write-up is followed by the sections on modeling of piezoelectricity, piezoelectric energy harvesting systems and technology issues regarding design, manufacture and implementations of piezoelectricity. The piezomagnetism as related to piezoelectricity and its implications are then put together in another section, which is proceeded by illustrative case studies on piezoelectricity and piezomagnetism, and closing remarks.

2.22.2

Thermopiezoelectricity

It has long been recognized that mechanical, electrical, and thermal fields are interconnected in many physical problems. The phenomenon of thermopiezoelectricity or piezothermoelasticity stands for the coupling of these mechanical, thermal, and electric fields in appropriate materials. In a way, it can also be defined as the piezoelectricity including thermal effects. These effects may have to be included for the proper modeling of piezoelectric systems used in sensing, control, and energy harvesting. Governing equations of the phenomenon of thermopiezoelectricity were first derived by Mindlin [36]. Two-dimensional equations for high-frequency motions of thermopiezoelectric crystal plates were obtained [37]. General theorems and mathematical models of thermopiezoelectricity were given [38,39]. A generalized thermoelasticity theory taking heat flux as an independent constitutive variable was formulated [40] and was extended to piezoelectric materials [41]. In a recent study, reflection and transmission of plane waves from a fluid-piezothermoelastic solid interface were investigated by Vashishth and Sukhija [42], where the piezothermoelastic solid half-space having some symmetry was assumed to be loaded with water. Expressions for the amplitude and energy ratios corresponding to the reflected and transmitted waves were derived analytically. An illustrative example was taken to study numerically the effects of the angle of incidence, frequency, specific heat, relaxation time, and thermal conductivity on the reflected and transmitted energy ratios. Although thermopiezoelectricity has been long studied and its theory has been well established, most of the investigations have been confined to theoretical studies. Applications of the theory of thermopiezoelectricity to behavior sensing and control of smart systems in general, and energy studies in paticular have been carried out only in recent years. Recently, two electromechanical systems were presented by Lumentut and Howard [43] for formulating the piezoelectric equations in energy/power harvesting system applications. The two voltage and charge-type systems associated with the standard AC–DC circuit interfaces were analyzed using Hamilton’s principle together with the theory of thermopiezoelectricity by making use of the electrical enthalpy and Helmholtz free energies. Benefit of the charge-type system approach was found to be in the compact modeling of the piezoelectric devices and electronic systems, which makes the formulation of the frequency-response functions and time wave-form systems possible. For the voltage-type system approach, however, although convenient, the derivation process was found to be tedious for linking the piezoelectric and electronic circuit parts together. Hence, the voltage-type system needed to be modeled by separating the piezoelectric structure from the electronic system and combining them together after formulating the frequency-response functions. The two approaches, which were validated by the finite element and experimental works, helped to study the power harvester by modeling the piezoelectric structure and electronic system together. In another recent study regarding the concept of thermopiezoelectricity as applied to the energy harvesting, research was conducted on a dynamic thermo-ferroelectric prestressed bimorph energy harvester, where general analytical expressions for the energy conversation coefficients including pyroelectric and thermal expansion effects were presented by Hasanyan and Hasanyan [44]. Being functions of the material properties, location of boundary conditions, vibration frequency and in plane compressive/ tensile follower force, these transformation coefficients were stated to be important on studies of sensors, actuators and energy harvesters in situations when mechanical, electrical and thermal fields co-existed (thermopiezoelectricity). It was shown on a particular example that the strain distribution, charge coefficient and energy generation were strongly influenced by the factors given above, i.e., by the volume fraction, material properties, plain compressive/tensile follower force, location of the boundary conditions and vibrational frequency of the bimorph. For the mathematical modeling of thermopiezoelectricity, it is carried out through a thermodynamic potential G defined as [45] 1 T 1 T 1 2 ST eE ET Py ST ky ð1Þ S cS E eE ay 2 2 2 where S and E are vectors of mechanical strain and electric field intensity; c,  and e are matrices for elastic, dielectric (permittivity) and piezoelectric coefficients, P and k are vectors for pyroelectric and thermal expansion coefficients of the material. Furthermore, y is a small temperature change and a is a scalar of constitutive coefficient given as a¼ rcv Θo 1, where r is the mass density, cv is the specific heat and Θo is the reference temperature. The superscript T in Eq. (1) and in equations below stands for the transpose of a matrix. Using this potential G, linear quasi-static equations are obtained for a thermopiezoelectric medium as [45] G ¼ GðS; E; yÞ ¼

∂G ¼ cS eE ky ∂S ∂G ¼ eT S þ eE þ Py D¼ ∂E ∂G ¼ kT S þ PT E þ ay Z¼ ∂y

T¼

ð2Þ

Piezoelectric Materials

699

where T and D are vectors of mechanical stress and electric displacement or charge per unit area, and Z is the scalar of entropy density. Eq. (2) assumes that mechanical, electric and thermal fields are at static equilibrium at any given instant. Q For differential equations of thermopiezoelectricity, two energy functionals and C are defined as R T R T R R þ ZΘÞ dV V u P b dV S u Ps dV þ V f rv dV þ S f sdS; R C ¼ V ðG m Θ Θ_ W ΘÞ dV þ S Θ hT n dS

P¼

R

V ðG

ð3Þ

R

where Θ is the absolute temperature given as Θ¼ Θo þ y; Pb and Ps are vectors of body and surface forces; u and h are vectors of mechanical displacement and external heat flux; n is the surface normal vector; f, rv, s and W are scalars of electric potential, volume charge density, surface charge and heat source density. The term G in Eq. (3) is given as 1 T ð4Þ ∇ Θ K ∇Θ 2 where ∇ is the gradient vector and K is the matrix of heat conduction coefficients. The generalized Hamilton’s principle has the following forms [46] Z t2 Y δ ðKi Þ dt ¼ 0 ð5Þ G¼

t1

and

δ

Z

t2 t1

C dt ¼ 0

ð6Þ

where Ki, the kinetic energy, is defined and then processed for Hamilton's principle as

Ki ¼

R 1 _T _ V 2 r u u dtδ

Z

t2 t1

Ki dt ¼ δ

Z

t2

t1

Z

1 r u_ T u_ dt dV ¼ V2

Z

t2

t1

dt

Z

€ dt r δuT u

ð7Þ

V

Variation of the thermodynamic potential G(S, E, y) is given by

δG ¼ δST

      ∂G ∂G ∂G þ δET þ δy ∂S ∂E ∂y

ð8Þ

Or, in view of Eq. (2) δG ¼ δST T

δET D

ð9Þ

δy Z

Hence, Hamilton's principle expressed by Eq. (5) now yields δ

R t2 t1

ðKi

R  Rt Q € δST T þ δET D þ δuT Pb Þ dt ¼ t12 dt V r δuT u

 R δfT rv dV þ S δuT Ps dS

R

T S δf s

 dS ¼ 0

ð10Þ

Mechanical strain S and electric field E are given by relations S ¼ Lu u; E ¼

∇f

where Lu is a differential operator matrix of 6  3 in size defined in cartesian coordinate system as 2∂ 3 0 0 6 ∂x 7 6 7 ∂ 6 7 60 0 7 6 7 ∂y 6 7 6 7 ∂ 60 7 0 6 ∂z 7 6 7 Lu ¼ 6 1 ∂ 1 ∂ 7 6 7 6 2 ∂y 2 ∂x 0 7 6 7 6 7 1∂7 61 ∂ 6 7 0 6 2 ∂z 2 ∂x 7 6 7 4 1∂ 1∂5 0 2 ∂z 2 ∂y

ð11Þ

ð12Þ

700

Piezoelectric Materials

Substituting Eq. (11) into Eq. (10) yields R t2 hR  € ðLu δuÞT T r δuT u t1 dt V

ð∇δfÞT D þ δuT Pb

where

R δfT rv dV þ S δuT Ps dS

R T T T V ðLu δuÞ T dV ¼ V δu ðLu TÞ dV R T T V ð∇δfÞ D dV ¼ V ∇ ðD δjÞ dV R R T T ¼ S δfT DT n dS V δf ð∇ DÞ dV

R

R

þ R

R

S δu T

T

R

S δf

T

i s dS ¼ 0

ð13Þ

ðN TÞ dS

T V δf ð∇ DÞ dV

ð14Þ

with N being a matrix consisting of surface normal components. Hence, Eq. (13) now becomes      R t2 R R R R T T T €þLTu T þ Pb dV þ V δfT ð∇T D rv Þ dV þ S δuT ðPs N T Þ dS ru t1 dt V δu V δf s þ D n dS

ð15Þ

Therefore, following equations must be satisfied with appropriate boundary conditions on corresponding surfaces €þLTu T þ Pb ¼ 0 ru ∇ D rv ¼ 0

ð16Þ

T

where the first equation represents equation of motion for the mechanical field [47] and the second equation is Maxwell’s equilibrium equation for the electric field [48]. For the thermal field, Hamilton’s principle on C given in Eq. (3) yields [46]  Z Z t2 Z     dt δy ∇T q W þ ΘZ_ dV þ δy hT qT n dS ¼ 0 ð17Þ t1

V

S

where q is the vector of heat flux. Hence, Hamilton's principle leads to the generalized heat equation expressed as ∇T q

W þ ΘZ_ ¼ 0

ð18Þ

with the boundary condition on the surface given by q¼ h.

2.22.3

Piezoelectricity

As stated before, the piezoelectricity is the phenomenon present in certain materials such as PZT and PVDF, where mechanical and electrical fields exist together (Fig. 1(A)). An electrical field is generated in the direction of poling, which is the process inducing piezoelectric properties in materials. For a piezoelectric sensor, the input is a mechanical effect like force and the output is an electrical quantity like voltage (Fig. 1(B)). It is opposite for a piezoelectric actuator, i.e. the input is electrical and the output is mechanical (Fig. 1(C)). Quasi-static equations of linear piezoelectricity are obtained by removing thermal components in thermopiezoelectric equations in Eq. (2). Hence, following coupled equations between mechanical and electric fields in a piezoelectric medium result T ¼ cS

eE D ¼ e S þ eE

ð19Þ

T

Poling voltage

Eq. (16) presented above can be used for dynamic interactions in a piezoelectric material between mechanical and electric fields. This equation involves partial derivatives and thus often times it is converted into coupled ordinary differential equations using a numerical approach. In this regard, the well-known finite element method (FEM) is frequently used for the solution of variables like mechanical displacement u and electric potential f that appear in these ordinary differential equations. Once computed, these variables are post-processed to eventually obtain other unknown components such as mechanical stress T and electric charge Q. Therefore, due to its importance and frequent use in analyzing the piezoelectric dynamic behavior, the FEM is outlined below for the modeling of piezoelectric materials. Variables of the FEM to be solved for are taken as the nodal mechanical displacement ui and electric potential /i of finite elements. According to the finite element approximation, they are related to corresponding elemental ue and fe through shape

(−)

(−) +

−

(A) Poling direction

+ − (B) Sensing

+ (+)

−

(+)

(C) Actuation

Fig. 1 Piezoelectric material characteristics. (A) Poling direction, (B) Sensing, and (C) Actuation.

Piezoelectric Materials

701

function matrices of Nu and Nf by ue ¼ Nu ui ; fe ¼ Nf /i

ð20Þ

Matrices of elemental mechanical strain Se and electric field intensity Ee are then written as Se ¼ Lu ue ¼ ½Lu Nu Šui ¼ Bu ui Ee ¼ ∇fe ¼ ½∇ Nf Š fi ¼

Bf fi

ð21Þ

Substituting Eqs. (19), (20) and (21) into Eq. (17) yields coupled-field (mechanical and electric) finite element equations for the whole piezoelectric medium as €þKuu u þ Kuf / ¼ F Muu u Kfu u Kff / ¼ G

ð22Þ

where u and / are global vectors of mechanical displacement and electric potential, respectively. Elemental matrices in Eq. (22) are computed from R ½Muu Še ¼ Ve re NuT Nu dV R ½Kuu Še ¼ Ve BTu ce Bu dV R ð23Þ ½Kuf Še ¼ Ve BTu ee Bf dV R T ½Kff Še ¼ Ve Bf ee Bf dV

where coupled matrix Kuf is symmetric, and hence Kfu ¼ KufT. Vectors of the elemental applied mechanical force Fe and electric charge Ge in Eq. (22) are defined as R R F e ¼ Ve NuT Pbe dV þ Se NuT Pse dS þ NuT Pce R T R ð24Þ T Ge ¼ Se Nf se dS Ve Nf rve dV where the concentrated mechanical force Pce is also added to Fe. Eq. (22) can be written in a compact matrix form as #( )

( ) " K Kuf u u€ Muu 0 F uu þ ¼ € Kfu Kff f f 0 0 G

ð25Þ

Finite element Eqs. (22) or (25) can be used to analyze dynamical behavior of piezoelectric materials where mechanical and electric fields are coupled through displacement u and potential /. Furthermore, the thermodynamic potential energy EG is defined as Z G dV ð26Þ EG ¼ V

Substitution of Eq. (1) into this equation without terms corresponding to the thermal field and making use of the finite element matrices in Eq. (23) after some arrangement yields  1 T 1  ð27Þ / Kff / EG ¼ uT Kuu u þ 2Kuf / 2 2

The EG contains the potential energy terms contributed by the coupled mechanical and electrical fields. The magnitude of each term depends on types of applications and severity of existing fields.

2.22.4

Piezoelectric Energy Harvesting

As stated before, the energy harvesting is defined as capturing or scavenging a very small amount of energy from existing mechanical, thermal and other effects in surrounding environment such as vibration, temperature gradient, etc. The energy harvesting even very small is important because systems like portable electronic devices and wireless sensor networks often times require power only at ultra low levels. Furthermore, lives of these systems are usually much longer than batteries used to charge them. In addition, these systems may be placed at hard-to-reach places, where the battery replacement is not inconvenient. Another advantage of utilizing energy harvesters from natural sources around them is that chemical wastes to the environment from conventional batteries can be reduced by this means. Hence, it makes much sense to equip the electronic devices and sensors with energy harvesting capabilities to avoid using traditional batteries for them. There are basically three types of energy harvesting mechanisms available for micro-electromechanical systems (MEMS). These mechanisms are piezoelectric, electromagnetic and electrostatic transductions [49]. A vibration-based piezoelectric energy harvester consists of a piezoelectric structure to convert ambient vibrations into an (AC) electrical signal, an AC to DC converter, and a regulator with storage. Among various types of piezoelectric structures, a cantilever beam-type shown below in Fig. 2 is used most due to its simplicity, low cost and frequency. It contains a host structure, a tip mass (mt) and a piezoelectric bimorph (two piezoelectric layers on both sides of the host structure) sandwiched with electrodes. The base vibration through the host structure imposes a mechanical effect onto the piezoelectric bimorph, which transforms this effect

702

Piezoelectric Materials

Piezoelectric bimorph

z x Base vibration

mt

R

Host structure

+ Vp −

Fig. 2 Piezoelectric bimorph structure for energy harvesting.

z x Moving Host structure magnet

Base vibration

mg

Coil Fig. 3 Electromagnetic energy harvester.

Host structure

z

mt

x Base vibration

Electrodes

Electret

R

+ Vp −

Fig. 4 Electrostatic energy harvester.

into an electrical output in the form of an AC voltage (Vp). This AC voltage is turned into the DC voltage by the converter and is further regulated and stored for the usage. The function of the tip mass is to adjust the fundamental damped natural frequency of the piezoelectric structure at ambient frequencies, which are typically below 1 kHz. The piezoelectric structure produces maximum electrical output at resonance, i.e. when the ambient and natural frequencies match. Schematic of an electromagnetic energy harvester is shown in Fig. 3, where a magnet of mass mg is attached at the free end of a cantilever host structure. The motion of the magnet inside the coil because of the vibration of the host structure changes the magnetic flux, thus generating an AC voltage across the coil. Another electromagnetic energy harvesting mechanism makes use of ambient radio frequency (RF) signals and converts them into usable electrical energy. Today, the RF signals are broadcasted from billions of radio transmitters around the world, including mobile telephones, mobile base stations, and television/ radio broadcast stations. In the electrostatic energy harvesting (Fig. 4), plates of a variable capacitor or two electrodes are moved against each other by mechanical effects such as base ambient vibrations of the host structure to generate charges and hence electric energy. Again, the tip mass of mt is used so that the energy is harvested at mechanical resonance for the maximum energy collection. For the above-stated three energy harvesting methods, the piezoelectric energy harvesting is more advantageous compared with the other two methods, because it is easy to implement and results in large power densities similar to those for lithium-ion batteries and thermoelectric voltage generators [49]. The electromagnetic energy harvesting usually produces voltage outputs much less than those for a comparable piezoelectric setup. Furthermore, unlike magnets and coils used for the electromagnetic transduction, the piezoelectric fabrication techniques are well established, making their energy applications at various scales easily possible. As for the electrostatic energy harvesting, an external voltage is needed for the relative motion of capacitor plates to generate an AC voltage, which is not necessary for the piezoelectric energy harvesting. In addition to vibration to provide input to piezoelectric energy harvesting devices, other types of ambient sources are also possible to harness energy via piezoelectricity. For example, piezoelectric materials are placed in shoes to scavenge energy from walking action [50]. Two methods are exploited, one of which is to place the PVDF bimorph under the insole of the shoe to harness energy from bending the ball of the foot. The second method is to scavenge the foot strike energy by flattening the curved, prestressed metal strips laminated with the PZT layers placed under the heel. Piezoelectric energy harvesters also find many applications in biomedical fields, including devices implanted in the human body like self-powered cardiac pacemakers, vivo energy harvesters and mechanical sensors for detecting nanoscale cellular deflections [51]. Aerodynamic loads acting at remotely-piloted aircraft systems (RPASs) whose wings are mounted with piezoelectric layers can provide input for energy harvesting to power small and wireless electronic devices [52]. Mechanical energy generated due to aeroelastic wing vibrations of these systems can be converted into electric energy via piezoelectric transduction. A

Piezoelectric Materials

703

thermopiezoelectric energy harvester, on the other hand, can capture energy from various mechanical and thermal disturbances due to piezoelectric and pyroelectric couplings [44]. Hence, the ambient mechanical and thermal effects can be added together to collect greater amount of energy from the surrounding environment.

2.22.5 2.22.5.1

Technology Issues Design and Manufacture

Piezoelectric ceramic materials like PZT are made from poly-crystalline ceramics, which are versatile and can easily fit into specific applications [53]. These ceramics are chemically inert, immune to moisture and can be manufactured in different sizes and shapes. They are manufactured by mixing lead, zirconium and titanium oxide powders in certain proportions and heating the mixture to about 1000oC. At this high temperature, the resulting PZT powder is mixed with a binder and is sintered into desired shapes, some of which are: plate, thin disk, ring, and tube. The material is then cooled down, during which the ceramic becomes ferroelectric and its unit cells change from cubic to tetragonal structure. As a result, the unit cells are elongated in one direction and an electric dipole moment is generated within the unit cell. The application of a strong DC electric field has the effect of aligning most unit cells parallel to the applied field. This process is called poling, causing the ceramic to have a permanent net polarization, which leads to both direct and converse piezoelectric effects in the material. After the poling is complete, the ceramic expands in the poling direction and contracts in the perpendicular direction when a voltage less than the poling voltage is applied. When the polarity is changed, the effect is opposite, i.e., the ceramic contracts in the poling direction and expands in the perpendicular one. In the converse action, when a tensile force is applied along the axis normal to the poling one, a voltage of the same polarity of poling is generated in the direction of poling. Piezoelectric materials can be bonded/glued to surfaces of host structures or embedded within them. The advantage in surface bonding is that it is easy to access and maintain, but it brings the disadvantage of being susceptible to damage during service [54]. On the other hand, the embedding leads to a strong mechanical and electrical coupling between the piezoelectric material and the host structure. Besides, the adhesives necessary for surface bonding/gluing are not needed. However, complications may arise in manufacturing and electrical insulation [55]. Instead of bonding the piezoelectric materials directly to the host structures, they can be bonded via thin viscoelastic layers. They can also be bonded to stiff constraining layers that are adhered to host structures by means of thin viscoelastic layers. In both these methods, mechanical/structural vibrational vibrations can be attenuated by damping of energy in constrained viscoelastic layers. The bonded piezoelectric materials act as actuators in this case (converse piezoelectric effect) to help dissipate the vibrational energy. These techniques of vibrational energy damping via converse piezoelectricity, constrained layer techniques, are investigated by many researchers [56–60].

2.22.5.2

PZT Piezoceramics

As mentioned before, the PZT, or lead zirconate titanate, is one of the most widely used piezoelectric ceramic materials [61], due to their favorable properties such as being physically strong, chemically inert and flexible in meeting requirements. The PZT also has a great sensitivity and a high Curie point compared with some other piezoelectric materials. When fired, the PZT has a perovskite crystal structure, whose each unit usually consists of a small titanium or zirconium ion in a lattice of lead ions. A major drawback of the PZT is that its major ingredient is the lead oxide, which is generally a hazardous and toxic material. Hence, research studies are on the way to develop alternative lead-free piezoelectric materials in order to replace the lead-based materials [62]. The PZT piezo ceramic was developed in around 1952 by Yutaka Takagi, Gen Shirane and Etsuro Sawaguchi, physicists at the Tokyo Institute of Technology [63]. A broadband piezoelectric energy harvester with an applied restoring force was presented by Rezaei et al. [64]. The system consisted of a cantilever beam bonded with a piezoelectric PZT layer at the top surface and a tip mass at the free end, which was supported by a spring to model the restoring force. The piezoelectric harvester was subjected to harmonic base excitation and effects of the PZT layer on free vibrations, and those of the tip mass and base excitation on the frequency response of the system were investigated. As expected, the tip mass helped increase the scavenged voltage and tune the resonance frequency. It was also shown that a pure nonlinear restoring force by the spring caused the harvester resonance bandwidth and the output voltage to increase as compared to the energy harvester without the spring. Similarly, a bimorph MEMS made up of a steel cantilever beam with two PZT layers and a tungsten proof mass at the tip was designed by Kuo et al. [65] for vibration energy harvesting. The system was experimented under base excitations of various magnitude with both serial and parallel connections of the PZT layers in order to collect open-circuit output voltages. Several types of piezoceramics of PZT are available in industry [53]. Different types are manufactured according to their uses in applications. Some of these ceramics are given below [30]: PZT-4: This piezoceramic is a hard material and can withstand high levels of mechanical and electrical effects. It has high resistance to depolarization and low dielectric losses under strong electric load, which makes it suitable for deeply-submersed acoustic transducers and electrical power generators. PZT-5 (A, H): These soft piezoceramics are good in applications that require high sensitivity and precision. Thus, they can be used in sensors, piezoelectric energy harvestors, receivers and low-power generators. PZT-5A in particular is recommended for use in hydrophones, accelerometers, vibration sensors and in places where resistance to high temperatures, high sensitivity, dielectric

704

Piezoelectric Materials

permittivity and time stability are needed. PZT-5H is another soft piezoceramic material possessing even higher sensitivity, extremely high permittivity and piezoelectric constant. But, its operation is limited due to its lower time stability and Curie temperature. The PZT-5H piezoceramic was used by Savarimuthu et al. [66] in the vibration energy generator to power a wireless sensor node. A single-stage AC to DC converter integrating the rectification and boosting circuit was designed and implemented. The lithium-ion battery utilized as an energy storage unit was modeled in states of charging and discharging, and the sensor node in both active and sleep states depending upon the power consumption. In another application of the PZT-5H, an experimental study was reported by Vaish et al. [67] for the conversion of thermal radiation via temperature gradient to electrical energy by the use of pyroelectric properties of the PZT-5H. Analytical and experimental works on a piezoelectric energy harvester undergoing a base excitation was performed by Butt et al. [68] to scavenge voltage across its electrical circuit. Numerical simulations were carried out to observe the effect of frequency and loading on the PZT-5A piezoelectric material. The output voltage from the energy harvester was found to be increasing, but the resonance frequency decreasing as the loading increased. Analytical and experimental results were noted to agree with each other. The nonlinear nonconservative dynamic behavior of bimorph piezoelectric cantilevers under low-to-high excitation levels with a focus on soft piezoceramics of PZT-5A and PZT-5H was investigated by Leadenham et al. [69]. The harmonic balance method was used to identify and validate nonlinear system parameters based on a set of experiments for different samples. It can easily be seen in piezoelectric energy harvesting studies that the harmonic excitation is used most. The resonant harmonic excitation is usually preferred over the off-resonant one, since leads to the maximum electrical power output. However, there are some cases where exciting the energy harvester at its resonance frequency may not be possible. For example, the case could be such that the frequency of input excitation is much less than the fundamental resonance frequency of the harvester. As reported above, the soft piezoceramics PZT-5A and PZT-5H have been used by many researchers for energy harvesting under the resonant harmonic excitation mainly because of their high sensitivity and precision. On the other hand, the hard piezoceramic material PZT-4 has much resistance to depolarization and to high levels of mechanical and electrical inputs. Specific advantages of these soft and hard piezoceramics were investigated by Erturk et al. [70] for vibration-based energy harvesting. A piezoelectric energy harvester model, which was validated experimentally, was used to compare power generation performances of the soft and hard piezoceramics, and the soft and hard single crystals.

2.22.5.3

PVDF Piezoelectric Polymer

The piezoelectric PVDF is a semicrystalline polymer, which is available in different forms such as filament, solution, power, granules or semitransparent films. As opposed to piezoelectric ceramics and natural crystals which are usually brittle and stiff, the PVDF is chemically inert, creep-resistant, flexible and tough, and can be manufactured in the form of large flat sheets [71]. Moreover, the PVDF has a low density and dielectric permittivity leading to a high voltage coefficient [72]. The PVDF sensors can be used in a wide range of applications, because they can be manufactured to have rapid response, wide stress range, large signalto-noise ratio, high sensitivity and very small thickness in micrometers [73]. On the other hand, it should be noted that the Curie point of piezoelectricity for a PVDF polymer is around 1501C, but its operation temperature should be limited to temperatures below 80o for the effective use of piezoelectricity. The PVDF exists in a, b, g, and δ phases depending upon the crystalline structure. Out of these four phases, the b-phase is the one that exhibits polarization. Hence, piezoelectricity is only present in b-phased PVDF polymer. A PVDF film without any processes like mechanical stretching or electrical poling yields the a-phase PVDF structure that has zero dipole moment and zero polarization in its crystalline structure, and hence no piezoelectricity as a result [74]. However, the b-phased PVDF structure has fluorine on one side and hydrogen on the other side, forming a net dipole moment in the b-phased PVDF crystalline structure. The mechanical stretching applied to the PVDF polymer together with the poling at a high voltage will then change the local dipole distributions and result in an electric field forming charges and piezoelectric effect in the crystalline structure of the PVDF polymer. The b-phase portion or piezoelectricity in a PVDF polymer can be increased by the mechanical or electrical poling [75]. The piezoelectric polymer PVDF was used by Chakhchaoui et al. [76] in the knee-pad as an energy harvester to produce electricity from the walking action of the knee. The traditional knee-pad is replaced by the knee-pad made up of traditional textile knitted with a PVDF patch to work as an energy harvester. A rotational PVDF energy harvester for collecting wind energy was developed by Zhang et al. [77]. The piezoelectric PVDF beam was subjected to an impact pressure to generate voltage. When the impact frequency was beyond a certain level, the energy output was found to be decreasing with the increase in wind speed. Vortexinduced vibration was analyzed by Pan et al. [78] to convert kinetic energy of flowing water into electricity by an energy harvester. Forces from the shedding vortices made the PVDF beam to vibrate so that the electrical energy was generated by the harvester due to the direct piezoelectric effect. A piezoelectric composite nano-generator was formed by Ding et al. [79] via combining the formamidinium lead halide perovskite nanoparticles (FAPbBr3) nanoparticles with the PVDF polymer to obtain a high piezoelectric power output. The energy generated by the generator could be used to charge a capacitor through a bridge rectifier and light up a light-emitting diode (LED). A PVDF-based piezoelectric energy harvester was demonstrated by Jung et al. [80] for roadway applications. The harvester showed stable performance and durability over many number of bending cycles and provided enough energy density as high as the piezoceramic-based harvester. The copolymer of PVDF, PVDF-TrFE, exhibits a higher temperature range of use and a larger portion of b-phase morphology with a thermal annealing process [74]. Thus, it is a good material for a piezoelectric polymer sensor because it can have a good

Piezoelectric Materials

705

amount of piezoelectric effect without the poling process. A transparent PVDF-TrFE/graphene oxide (GO) ultrathin film was prepared by Wang et al. [81] to improve dielectric properties of the PVDF-TrFE and to enhance its energy harvesting performance in return. The resulting PVDF-TrFE/GO improved copolymer was designed to make the energy harvesting composite films attractive in applications of NEMS or MEMS. Design, fabrication and performance evaluation of cantilever piezoelectric harvesters utilizing the PVDF-TrFE at MEMS scale were carried out by Toprak and Tigli [82]. The objective of their study was to make a piezoelectric energy harvesting system that can be integrated with complementary metal-oxide-semiconductor circuits and pave the way for selfsustained low-power electronics. Effect of micro- and nano-structuration on the piezoelectric properties of polymers having the PVDF-TrFE was reported by Canavese et al. [83]. Measurements showed that the crystallization of the polymer into the ferroelectric b-phase was affected by the confinement size. Uses of the polymeric structures as potential flexible tactile sensors and bendable energy harvesters were demonstrated, indicating the profound effect of the micro- and nano-structuration on device performances. As stated before, the polymer-ceramic composite of PVDF-PZT has received a lot of attention since these materials may combine favorable piezoelectric properties of PZT ceramics with the advantages of PVDF polymers [72]. These combined properties may not be attainable with a single piezoelectric material. A PVDF-PZT nanocomposite film was proposed by Zhang et al. [84] to be used as a piezo-separator to directly convert mechanical energy into electrochemical energy in self-charging power cells. The composite-structured PVDF-PZT was designed to provide a high piezoelectric output, because the PZT would improve the potential of the piezoelectric system as compared to the pure PVDF film. A self-charging power cell based on this nanocomposite PVDF-PZT arrangement could be efficiently charged by the mechanical displacement with no external power source.

2.22.5.4

Dynamic Performance and Modeling

Dynamic performance of a piezoelectric component (sensor or actuator) shows its response to alternating mechanical and electrical fields. A lumped electric circuit model of a piezoelectric element with electrodes is shown below in Fig. 5, where the inductance L1, capacitance C1, and resistance R1 are the lumped parameters for the piezoelectric element itself, and the parameter C2 is the electrical capacitance due to the electrodes. For the complex impedance model of the piezoelectric element shown in Fig. 6, the complex impedances Z1(s) and Z2(s) in Laplace transformation domain are written as Z1 ðsÞ ¼ L1 s þ

1 1 þ R1 and Z2 ðsÞ ¼ C1 s C2 s

ð28Þ

For the parallel circuit arrangement in Fig. 6, the equivalent complex impedance Z(s) is found as Z1 ðsÞ Z2 ðsÞ Z1 ðsÞ þ Z2 ðsÞ

ð29Þ

L1 C1 s2 þ R1 C1 s þ 1 s ½L1 C1 C2 s2 þ R1 C1 C2 s þ ðC1 þ C2 ފ

ð30Þ

Z ðsÞ ¼ Substitution of Eq. (28) into Eq. (29) yields Z ðsÞ ¼ which can be re-arranged to be expressed as

1 1 s2 þ 2ζ1 o1 s þ o21 o22 2 2 o1 s þ 2ζ2 o2 s þ o22 ðC1 þ C2 Þ s ¼ G1 G2 ðsÞG3 ðsÞG4 ðsÞ

Z ðsÞ ¼

L1

C1

R1 i

i C2 V Fig. 5 Lumped dynamic model for a piezoelectric component with electrodes.

Z1 i

i

i ≡

Z2 V Fig. 6 Complex impedance model for the piezoelectric component shown in Fig. 5.

i Z V

ð31Þ

706

Piezoelectric Materials

l

w t Fig. 7 Piezoceramic material with tensile stress in the length (l) direction.

where G1 ¼

1 1 s2 þ 2ζ1 o1 s þ o21 o22 ; G2 ðsÞ ¼ ; G3 ðsÞ ¼ ; G 4 ð sÞ ¼ 2 2 o1 s þ 2ζ2 o2 s þ o22 ðC1 þ C2 Þ s

ð32Þ

In Eqs. (31) and (32), natural frequencies o1 and o2, and damping ratios ζ1 and ζ2 are given by rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi rffiffiffiffiffiffi rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 1 C1 þ C2 R1 C1 C2 ; ζ1 ¼ ; ζ2 ¼ ζ1 o1 ¼ pffiffiffiffiffiffiffiffiffiffi ; o2 ¼ o1 C2 2 L1 C1 þ C2 L1 C1

ð33Þ

Hence, it is clear from the above Eq. (33) that

o2 ζ1 ¼ ¼ o1 ζ2

rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi C1 þ C2 41 C2

ð34Þ

In other words, o24o1 and ζ14ζ2. Here, an example is worked out on a plate-shaped piezoceramic element made up of PZT-5A of Morgan Ceramics, now CramTec, stressed in the length (l) direction as shown in Fig. 7. Dimensions of this piezoceramic plate are taken as: length l¼ 0.01 m, width w¼0.005 m and thickness t ¼0.0005 m. In order to estimate the above lumped parameters, the force-voltage analogy is used between mechanical and electrical systems. Therefore, the following calculations are performed. Mass of Piezoceramic Element ¼ Mass Density  Volume, and hence m ¼ rV¼ r (lwt)¼ 7750(0.01)0.005(0.0005) ¼1.94  10 4 kg, which is analogous to the inductance L1 in Henry (H). Elasticity or Spring Constant in l direction ¼ Cross-Sectional Area  Modulus of Elasticity/ Length, and hence k¼ AE/l¼ twE/l, or 1k ¼ l=ðtwEÞ ¼ 0:01=ð5  10 4  5  10 3  12:1  1010 Þ ¼ 3:3  10 8 m=N, analogous to the capacitance C1 in Farad (F). Viscous Damping Constant ¼ 2  Basic Material Damping Ratio  2p  Resonance Frequency  Mass, or b ¼ 4pζ fr m, where the basic material damping ratio ζ is assumed to be 0.01. The resonance frequency fr of the piezoceramic element in this equation is given by [53] fr ¼

N1 l

where the frequency constant N1 ¼ 1400 Hz
m, and hence fr ¼ 1400/0.01 ¼ 140 b¼ 4p  0.01  1.4  105  1.94  10 4 ¼3.4 N
s/m, which is analogous to the resistance R1 in Ohm (O). Capacitance of the piezoelectric element with electrodes is computed next as [53] Cap ¼

T K33 e0 l w t

ð35Þ kHz.

Therefore,

ð36Þ

T where K33 is the relative dielectric constant or permittivity and e0 is the permittivity of free space in F/m, and hence Cap ¼ 1700  8.854  10 12  0.01  5  10 3 / 5  10 4 ¼ 1.5  10 9 F, which stands for the electrical capacitance of C2. As a result of all the above calculations and assumptions, the following lumped-parameter values are taken for the piezoceramic material with electrodes: the inductance L1 ¼1.94  10 4 H, the capacitances C1 ¼ 3.3  10 8 F and C2 ¼ 1.5  10 9 F, and the resistance R1 ¼ 3.4 O. With these numerical values, the natural frequencies o1 and o2, and damping ratios ζ1 and ζ2 are calculated using the above Eq. (33) as: o1 ¼ 3.95  105 rad/s, o2 ¼ 1.895  106 rad/s, ζ1 ¼ 0.0222 and ζ2 ¼ 0.0046. Bode diagram [85] of the complex impedance Z(s) of this lumped-parameter piezoceramic element with electrodes shown in Fig. 7 is now drawn using Matlab [86] as depicted in Fig. 8. Note that for the magnitude diagram in p the half of Fig. 8, the ffiffiffiffiffiffiupper ffi magnitude of the complex impedance Z(s) is given in dB, which is defined as 20jlog10 ZðjwÞj where j ¼ 1, and the phase angle is given in degrees for the phase diagram in the lower half. As it can be seen in Fig. 8, the magnitude decreases linearly at the beginning with a slope of 20 db/decade due to the effect of the first-order transfer function of G2(s). The effect of the second-order transfer function G3(s) is observed just before the natural frequency o1, where the magnitude curve makes minimum because of the function G3(s). This minimum occurs sharply due to the small amount of damping ratio of ζ1 ¼0.0222. The function G3(s) then causes the magnitude to rise at a slope of 40 db/decade up to the second natural frequency of o2, where the effect of the transfer function G4(s) is now in place as a sharp increase in magnitude, sharper than before at o1, due to the very small value of ζ2 ¼0.0046 less than ζ1. In fact, the magnitude of the complex impedance Z(s) reaches its peak value at o2, after which the magnitude decreases with a slope of 20 db/decade and then remains

Piezoelectric Materials

707

Bode diagram 100

Magnitude (dB)

80 60 40 20 0 90

2

1

Phase (deg)

45 0

−45 −90 104

105

106

107

Frequency (rad/s) Fig. 8 Bode diagram of the complex impedance Z(s) in Fig. 6 for the plate piezoceramic element in Fig. 7.

constant at around 40 dB in this example. As for the phase angle diagram shown in the lower half of Fig. 8, the phase angle is constant at þ 90o between the natural frequencies o1 and o2, and 90o otherwise. The first natural frequency o1, where the magnitude of the complex impedance Z(s) makes a minimum, is related to the resonance frequency of fm by o1 ¼ 2pfm. Similarly, the second natural frequency o2, where the magnitude of the complex impedance Z(s) reaches its maximum, is linked to the so-called anti-resonance frequency of fn by o2 ¼ 2pfn. Since from Fig. 6, in Laplace Transformation domain V(s) ¼Z(s) I(s), and hence jVðjoÞj ¼ jZðjoÞjjIðjoÞj

ð37Þ

indicating that the maximum output, voltage for a piezoelectric sensor, can be obtained when the magnitude of the complex impedance Z(s) is maximum, which occurs when the sensor is excited at its natural frequency of o2. Electric circuit to which a piezoelectric device is connected to should have certain characteristics depending upon the design of the piezoelectric device. For this reason, a piezoelectric device is usually classified into two categories, which are the piezoelectric nonresonant and resonant devices [53]. A nonresonant piezoelectric device operates much below its resonant frequency and a nonresonant one is designed to operate around its resonant frequency. Accelerometers and microphones are good examples for nonresonant piezoelectric devices, which work over a large frequency range well below the resonance in a linear region. For all practical purposes, except operations at very low frequencies, a nonresonant piezoelectric device can be modeled as a lumped electric circuit containing only the equivalent capacitance of the piezoelectric element Ce and the shunt load capacitance of CL due to the wiring, cables and amplifier. In this case, if the operating frequency is much smaller than the first mechanical resonance of the device, the mechanical displacement of the piezoelectric element would be directly proportional to the instantly-applied electrical charge. Electrical modeling of a resonant piezoelectric device is more complicated than the nonresonant piezoelectric device. This device can be better modeled by a capacitor for the static capacitance of the piezoelectric element, which is shunted by the impedance corresponding to the mechanical vibrating system. A unimorph piezoelectric element glued to the surface of a host structure whose action due to external inputs causes the piezoelectric element to displace, which produces an electrical output in return in case of sensing. A bimorph, on the other hand, is made by bonding two pieces of piezoelectric materials together so that changes in lengths of these two pieces yield relatively large displacements when they are subjected to external voltages in the case of piezoelectric actuators. Stacking of several piezoelectric pieces is also possible to obtain even larger displacements. The total displacement would be equal to the summation of individual displacements of stacks.

708 2.22.5.5

Piezoelectric Materials Various Other Considerations

Operation of a piezoceramic material at high stresses and/or voltages has an adverse effect on its useful life. A high mechanical stress can destroy piezoelectric characteristics of the material. The limit on the amount of applied stress depends on the type of piezoelectric material and duration of the stress. A piezoelectric material can also be depolarized if a strong voltage is applied in the opposite direction of poling. Again, the allowable limit for the applied voltage is dependant upon the type of piezoelectric material, duration of the voltage and operating temperature during the application. Typical limits for piezoceramics is between 500 and 1000 V/mm along the poling direction. Hence, for a piezoelectric plate, assuming that the poling is along its thickness, then the applied voltage limit would be 500–1000 V for each mm thickness of the plate. Another consideration for a piezoelectric material is its Curie point, which is the maximum temperature after which the material loses its piezoelectric characteristics. For the above-stated Morgan PZT ceramics, the Curie points are listed as 328oC for the PZT-4, 365oC for the PZT-5A and 193oC for the PZT-5H [30]. Aging is one of the major considerations for a piezoelectric material in that its properties change with time [53]. Aging rate depends on the material composition and the manufacturing process by which the material is made. Because of aging, material properties of a piezoelectric element may have to be specified for a duration of time only. High mechanical stress, elevated temperature close to the Curie point and strong electric field that causes depoling are important factors, which speed up the aging process. Properties of a piezoelectric material differ with respect to the poling axis, which defines direction of action or response related to the property. A certain property is defined as a symbol with a subscript containing numbers that shows directions of corresponding fields. Conventionally, the subscript 3 is the z-axis for the poling direction. Because of this convention, a typical planar piezoelectric element is usually defined in x-z plane (not in x-y plane), where the horizontal x or 1 axis indicates direction of mechanical action or response, and z or 3 axis is the vertical poling direction along which the electrical action or response takes place. Furthermore, the number 2 in the subscript is kept for the property or action (response) corresponding to the y axis. Numbers 4, 5, and 6 refer to various shear modes or related properties for the piezoelectric element. For example, the symbol e31 stands for the piezoelectric coefficient corresponding to the electric field along the z-axis (3) and a mechanical field along the x-axis (1). It is this piezoelectric coefficient that couples the electric field in the z or 3 direction with the mechanical field in the x or 1 direction. Hence, it is given later in this chapter as a case study that application of an applied voltage in the poling direction along the z-axis results in mechanical stress along the x-axis. This stress then generates a uniform bending moment on the bimorph piezoelectric element, which can be used for grabbing an object. A piezoelectric transducer may be having limitations on its efficiency, temperature handling and dynamic capacity [53]. The efficiency may be an issue for a low-frequency piezoelectric transducer used from time to time. On the other hand, a highfrequency piezoelectric transducer utilized in continuous operations may have to be monitored for its temperature. Heat removal may be necessary to reduce and control the temperature.

2.22.6 2.22.6.1

Piezomagnetism Piezomagnetic/Magnetostrictive Materials

Polarized magnetostrictive materials may be considered as piezomagnetic over linear ranges of operation, similar to piezoelectric ceramics [87]. Hence, piezomagnetism is a linear magneto-mechanical effect, which is analogous to the linear electro-mechanical effect of piezoelectricity. In other words, the piezomagnetism is defined as the phenomenon of interaction between mechanical and magnetic fields in a suitable medium. In a way, it describes the energy conversion between these mechanical and magnetic fields. In the light of the above discussion, studies on magnetostriction and piezomagnetism can be carried out along the same line. The phenomenon of magnetostriction was first discovered by Joule in 1842 [88]. Hence it is also known as the direct magnetostriction or Joule effect. It is responsible for generating magnetic output in a suitable medium due to the application of a mechanical field in the form of force, torque, stress, etc. The same medium can give a mechanical output if it is subjected to a magnetic field because of the inverse magnetostriction phenomenon known as the Villari effect. Similar to the poling process in piezoelectric materials, when a magnetic field is applied onto a ferromagnetic material, the internal structure units re-align themselves in the applied magnetic field direction, giving the material a magnetostrictive or piezomagnetic characteristic. Depending upon the direction of the magnetic field, the result could be a mechanical extension or compression for the material. After the process is over, the material exhibits both the direct (Joule) and inverse (Villari) phenomena of magnetostrition, which allows it to be used as a sensor and actuator, respectively. Some magnetostrictive piezomagnetic materials are the cobalt ferrite (CoFe2O4), Nickel (Ni), Alfenol (also known as Alperm or Alfer) and Terfenol-D. The cobalt ferrite CoFe2O4 is known for its highest magnetostrictive coefficient among the oxides-based magnetostrictive materials. It is widely used in electronics and particularly in automotive industry as a stress and noncontact torque sensor [89,90]. Magnetostriction properties of the cobal ferrite can be improved by increasing the sintering temperature or time [90]. The nickel (Ni) is a metal found in many large ore deposits around the world [91]. It was used in sonar transducers as a

Piezoelectric Materials

709

magnetostrictive material during World War II. It has the advantage of high tensile strength and low input electrical impedance. But, the piezoceramic PZT has much lower electrical losses and higher coupling coefficient, and its tensile strength can be further improved by a prestressing process [92]. The alfenol is another magnetostrictive material discovered by Masumoto and Saito [93,94]. It is a magnetically soft alloy consisting of mostly iron and aluminum. It has a large amount of magnetostriction exhibiting high magnetic permeability. This material can be utilized in a wide frequency range due to its aluminum content giving the alloy a good level of electrical resistivity. The Terfenol-D, also called as the giant magnetostrictive material, is an alloy of rare earths Terbium (Tb) and Dysprosium (Dy) with 3D transition metal Iron (Fe), which has by far the largest known magnetostriction among alloys. Hence, it is the most commonly-used magnetostriction material in various areas of science and engineering, especially in actuator making. It is produced in a variety of forms, namely in solid (monolithic), powder and thin film forms. But, its performance is much dependent on the state of the material. Thus, for example, prestressing a Terfenol-D sample greatly increases its total strain capability [95].

2.22.6.2

Applications

There are many research studies carried out in recent years due to advancements in magnetostriction and/or piezomagnetism, nano technologies and energy harvesting. The behavior of a crack in functionally graded piezoelectric/piezomagnetic materials subjected to an anti-plane shear loading was investigated by Zhou et al. [96]. It was assumed that the material properties varied exponentially in the direction parallel to the crack. A nonlinear dynamic model was developed by Cao et al. [97] for a magnetostrictive Terfernol-D energy harvester exhibiting the coupled mechanical-magnetic-electric behavior and eddy-current effects. The model was able to provide reasonable data trends for piezomagnetic coefficients, modulus of elasticity and relative permeability for Terfenol-D. Strain-powered antennas radiating electromagnetic energy by mechanically vibrating a piezoelectric or piezomagnetic material were proposed by Domann and Carman [98]. A closed form analytic model of the electromagnetic radiation from a strain-powered electrically small antenna was formulized. It was shown that a properly designed strain-powered antenna could be more efficient than an equally sized electric dipole antenna. A two-dimensional linear elastic theory of magneto-electroelastic plates covering both the phenomena of piezoelectricity and piezomagnetism was established by Wang et al. [99]. Effects of the poling directions, piezoelectric phase, volume fraction, damping and applied magnetic field on the magnetoelectric coupling were investigated based on the two-dimensional plate theory. A relatively large output power density was also achieved on a magnetic energy nanoharvester using the two-dimensional plate theory obtained. Two hybrid energy harvesters with each employing combinations of piezoelectric, magnetostrictive and electromagnetic technologies were proposed by Ibrahim and Salehian [100]. The two harvesters were fabricated and tested for the voltage, power and displacement outputs. A nanoenergy harvester of piezoelectric/piezomagnetic structure considering the thickness-shear mode by the surface stress model was presented by Fan et al. [101]. As a result, the capability of the energy harvester was found to depend on the residual surface stress, the surface piezoelectric and piezomagnetic elastic constants and the thickness of the piezoelectric layer. The proposed model thus showed the possibility of utilizing nano-composite structures in energy harvesting applications. A magnetic energy harvester using an alternating magnetic field was proposed by Zhao et al. [102]. The energy harvester had functionally graded piezomagnetic and piezoelectric cylindrical layers to capture magnetic energy and to convert that into electric energy. A theoretical analysis was performed using the linear equations of piezoelectricity and piezomagnetism for the corresponding materials. Design and testing of a piezomagnetic energy harvester capable of converting nonharmonic movement of a human head into electric energy was reported by Delnavaz and Voix [103]. The energy was efficiently harvested by a rolling magnet and a piezoelectric element, which is fixed from two sides.

2.22.6.3

Linear Theory of Piezomagnetism

The piezomagnetism equations are written as [45] ∂G ¼ cS ℓB ∂S ∂G ¼ ℓT S þ m 1 B H¼ ∂B

T¼

ð38Þ

where c and ℓ are the matrices of elastic and piezomagnetic stress coefficients, m is the matrix of material permeability; T, S, H, and B are the vectors of stress, strain, magnetic field intensity and magnetic flux density, respectively. It can be shown that [45] the thermodynamic potential G takes on the following quadratic form: G¼

1 T 1 S cS þ BT m 1 B 2 2

ST ℓB

ð39Þ

which is re-written as 1 T 1 S T þ BT m 1 B 2 2 by the use of Eq. (38). Hamilton’s principle given before by Eq. (5) is rewritten here for the ease of readability Z t2  Y dt ¼ 0 Ki δ G¼

ð40Þ

t1

ð41Þ

710

Piezoelectric Materials

where Ki is the kinetic energy and P is the energy functional, which are given by R 1 Ki ¼ V r u_ T u_ dt 2 Z Z Z Z Z P ¼ G dV AT J dV þ AT HE0 n dS uT Ps dS uT Pb dV

ð42Þ

S

V

S

V

V

where Pb and Ps are the vectors of body and surface forces, u, A, J and n are the vectors of mechanical displacement, magnetic 0 potential, volume current density and surface normal, and HE is the matrix of external magnetic field intensity defined in rectangular coordinate system as 3 2 Hy 0 Hz 0 6 Hx 7 HE ¼ 4 Hz 0 ð43Þ 5 Hy Hx 0 E

Variation of G is processed and rewritten in the view of Eq. (38) as

δG ¼ δST cS þ δBT m 1 B δST ℓB δBT ℓT S ¼ δST ðcS ℓBÞ þ δBT ð ℓT S þ m 1 BÞ ¼ δST T þ δBT H

ð44Þ

Variation of the kinetic energy Ki given before by Eq. (7) is re-produced here as Z t2 Z Z t2 € dt Ki dt ¼ dt r δuT u δ t1

t1

ð45Þ

V

Substitution of the above equations in Eq. (41) yields the following: Z t 2 Z Z  Y   Rt € δST T δBT H þ δuT Pb þ δAT J dV þ dt ¼ dt r δuT u δuT Ps δ t21 Ki t1

V

S

Relations for the strain S and magnetic flux density B are expressed as

0 δAT HE n dS ¼ 0

S ¼ Lu u; B ¼ ∇  A ¼ LA A

where Lu and LA are differential operators, and substituting Eq. (47) into Eq. (46) results in the following equation  i R  R t2 hR  0 € ðLu δuÞT T ð∇  δAÞT H þ δuT Pb þ δAT J dV þ S δuT Pb δAT HE n dS ¼ 0 r δuT u t1 dt V

ð46Þ

ð47Þ ð48Þ

Following relations are written for some terms in Eq. (48) Z Z Z   ðLu δuÞT T dV ¼ δuT LTu T dV þ δuT ðN TÞdS V

V

S

and

R

V ð∇

R R  δAÞT H dV ¼ V ∇T ðδA  HÞdV þ V δAT ð∇  HÞdV R R T 0 T ¼ S δA H n dS þ V δA ð∇  HÞdV

where the matrix of normal directions, N, is defined in rectangular coordinates as 3 2 nx 0 0 ny nz 0 7 6 N ¼ 4 0 ny 0 nx 0 nz 5 0 0 nz 0 nx ny

ð49Þ

ð50Þ 0

In the above Eq. (50), nx, ny and nz are the components of the surface normal, n, and H is defined as in Eq. (43). Eq. (48) is rearranged as    R R T 0 R R t2 R 0 T €þLTu T þ Pb dV þ V δAT ð ∇  H þ J ÞdV þ S δuT ðPs N T ÞdS ru H n dS ¼ 0 ð51Þ t1 dt V δu S δA HE Hence, the following equations must be satisfied with appropriate boundary conditions on the corresponding surfaces €þLTu T þ Pb ¼ 0 ru ∇HþJ¼0

ð52Þ

The above equations are the equations of motion for the mechanical field (as in Eq. (16)) and Maxwell’s equilibrium equation for the magnetic field, respectively [48]. For the finite element formulation similar to the phenomenon of piezoelectricity, u and A are chosen variables for the mechanical and magnetic fields, respectively. The finite element approximations are written as ue ¼ Nu ui Ae ¼ NA Ai

ð53Þ

Piezoelectric Materials

711

where, as before, the subscript e and i respectively stand for the element and nodes of the element and N’s are the appropriate shape function matrices. Following relations are also noted Se ¼ Lu ue ¼ ½Lu Nu Šui ¼ Bu ui

and

Be ¼ LA Ae ¼ ½LA NA ŠAi ¼ BA Ai

ð54Þ

€þKuu u KuA A ¼ F Muu u KAu u þ KAA A ¼ M

ð55Þ

and substituting Eqs. (38), (53) and (54) in Eq. (48) yields the finite element equations as

where u, A, F and M are the global vectors of mechanical displacement, magnetic potential, applied mechanical and magnetic excitations, respectively. Element matrices and applied excitations in Eq. (55) are found as R R ½Muu Še ¼ Ve re NuT Nu dV; ½Kuu Še ¼ Ve BTu ce Bu dV R R ½KuA Še ¼ Ve BTu ℓe BA dV; ½KAA Še ¼ Ve BTA me 1 BA dV R R ð56Þ F e ¼ Ve NuT Pbe dV þ Se NuT Pse dS þ NuT Pce R R 0 T Me ¼ Ve NAT J e dV Se BA HAe n dS

where KAu ¼ KuAT and the concentrated elemental applied force of Pce is added to the definition of the elemental applied force of Fe. The above finite element equations of piezomagnetism can be used for dynamic interactions of mechanical and magnetic fields in a suitable medium. Possible applications are sensing and actuation of smart dynamic systems and energy harvesting that involves mechanical and magnetic actions/responses.

2.22.7

Illustrative Cases for Piezoelectricity

This section presents two illustrative examples demonstrating uses of piezoelectricity in sensing, actuation and energy applications. The first example is a bimorph made up of a PVDF piezoelectric polymer actuated by an applied electric field. This bimorph example can also be utilized in piezoelectric energy harvesting. The second example is a cantilever beam structure bonded with a piezoelectric layer, which illustrates the sensing capability of the piezoelectric layer at different locations on the beam. Both the finite element and experimental techniques are used in this example. The sensor presented in this case is an example as to the conversion of the mechanical energy into the electrical one via the phenomenon of piezoelectricity and is the basis for the energy scavenging.

2.22.7.1

Piezoelectric Bimorph

A bimorph or robotic finger shown in Fig. 9 is considered in order to show the actuation capability of piezoelectricity. The problem can be reversed to make the bimorph function as a sensor. Two layers of polymeric PVDF with material properties given in Table 1 are glued together to form a piezoelectric bimorph beam structure. Poling directions of the two layers oppose each other and hence a bending moment is generated through the bimorph when an external voltage is applied. Length and thickness the bimorph are specified as L¼ 100 mm and h¼ 1 mm, respectively. As seen in Fig. 9, the finger is divided into 10 finite elements for the FEM analysis. The theoretical solution for the lateral displacement u3 is calculated as follows [104]. The moment M generated in the finger by the electric field E3 due to the applied voltage V is given by  2     Z h=2 wh V w h2 wh ¼ e31 ¼ e31 V w e31 E3 z dz ¼ e31 E3 ð57Þ M¼ 4 4 h 4 h=2 where w is the width of the finger and e31 is the piezoelectric constant for the PVDF. On the other hand, the lateral displacement u3 at a distance x for the bimorph finger is found from the basic cantilever Euler beam equation expressed as c11 I

∂2 u3 w h3 ∂2 u3 ¼ c11 ¼M 2 12 ∂ x2 ∂x

ð58Þ

z, u3 +V

x h L Fig. 9 Piezoelectric bimorph finger (not to scale).

−V

Electric field, E3

712

Piezoelectric Materials

Table 1

Properties of PVDF

c11 (Pa) e31 (C/m2) e11 (F/m) e33 (F/m)

2

3.8  109 0.046 1.026  10 1.026  10

10 10

× 10−7 Theoretical FEM

1.8 1.6 1.4

u3 (m)

1.2 1 0.8 0.6 0.4 0.2 0 0

0.02

0.04

0.06

0.08

0.1

x (m) Fig. 10 Static deflection of piezoelectric bimorph due to applied unit voltage.

the solution of which yields u3 ðxÞ ¼ 1:5

e31 V  x 2 c11 h

ð59Þ

On the other hand, the finite element solution for the displacement u is obtained by the use of the first of Eq. (22) in static form as from which

Kuu u þ Kuf / ¼ 0

ð60Þ

u ¼ Kuu1 Kuf V

ð61Þ

where V is the applied static voltage. The theoretical and FEM solutions are computed for an applied unit voltage V and results are shown in Fig. 10, which indicates closeness of solutions by these two (theoretical and finite element) approaches. This example can be reversed to make the bimorph work as a sensor rather than an actuator, in which case the system can be subjected to pure bending moment or another mechanical effect to get voltage as an output. This is the basis for the bimorph to function as an energy harvesting device as discussed in the previous section.

2.22.7.2

Piezoelectric Sensing System

A system composed of a cantilever beam and a piezoelectric sensor is taken as an example in Fig. 11. The distance d in this figure is for the location of the sensor, whose values are assumed to be 0.005 and 0.1 m to observe its effect on the sensor output. Table 2 lists material properties for the steel beam and piezoceramic (PZT-5A) sensor [30]. The length, thickness and width for the beam are specified as 0.118m, 0.001m and 0.016m, and for the piezoelectric sensor as 0.01 m, 0.0004 m and 0.016 m, respectively. The piezoelectric sensor and beam are first analyzed using the FEM. The fundamental natural frequency of the system (beam þ sensor) slightly differ for the sensor locations of d¼ 0.005 m and 0.1 m, which are computed as 61.6 Hz and 57.1 Hz, respectively. The system is excited from the tip of the beam by an upward step force of unity as shown in Fig. 6. The sensor output

Piezoelectric Materials

713

Data acquisition

Computer d

Piezoelectric sensor Beam Force

Fig. 11 A piezoelectric sensing system (not to scale).

Table 2

Properties of materials

Piezoceramic (PZT-5A)

Beam (Steel) 10

8

d = 0.005 m

× 10−4

2.07  1011 (c11) 7500

(12.1, 7.52, 11.1, 2.26)  10 7500 ( 5.4, 15.8) (7.35, 8.11)  10 9

(c11, c12, c22, c33) (Pa) r (kg/m3) (e31, e33 ) (C/m2) (e11, e33) (F/m)

6

d = 0.1 m

× 10−5

Voltage (V)

Voltage (V)

6 4 2

4

2

0 −2

0

0.05

0.1

0

0.15

0

0.05

Time (s) × 10−6

8 PSD of voltage (V2/Hz)

PSD of voltage (V2/Hz)

5 4 3 2 1 0

0

25

50

0.1

0.15

Time (s)

75

100

× 10−8

6 4 2 0

0

Frequency (Hz)

25

50

75

100

Frequency (Hz)

Fig. 12 Outputs of the piezoelectric sensor located at d¼ 0.005m (left) and 0.1m (right).

is found from Eq. (22) as / ¼ Kff1 Kfu u

ð62Þ

which is calculated for the middle node at the top of the piezoelectric sensor. A proportional damping is also added to Eq. (22) with an assumption of a damping ratio of ζ ¼ 0.01. Voltage outputs of the piezoelectric sensor both in time and frequency domains are shown in Fig. 12 for the two sensor locations. It is evident from these plots that the sensor output at the location of d¼ 0.005m (Fig. 11) is more clear than the one at d¼ 0.1m. The fundamental natural frequency is hardly noticeable in the frequency spectrum when the sensor is located close to the free end of the beam at d ¼0.1m.

714

Piezoelectric Materials

4

V

3.90 V 20.9 ms

2

0

2

4 50

0 0.5 V/div

100 10 ms/div

150

Amplitude: 0.02 V/div

0.4 Overrange 0.3

0.2

0.1

0.0 0

100

200 300 Frequency: 20 Hz/div

400

500

0257@ 62 Hz

Fig. 13 Experimental voltage output for the location of the sensor at d¼0.005 m.

The experimental setup of the system is briefly shown in Fig. 11. The system is given an upward impulsive excitation at the tip of the beam and the voltage output of the piezoelectric sensor is sent to a data acquisition system and then to a computer for the processing and display. The time history and frequency spectrum plots for the two sensor locations are shown in Figs. 13 and 14. Since the data acquisition system magnifies the voltage by 1000, the actual voltage outputs of the sensor are one thousandth of the values shown in these figures. Similar to the FEM results above, Figs. 13 and 14 also indicate that outputs of the piezoelectric sensor are better visible when it is close to the root of the beam (at d ¼ 0.005m, Fig. 13) than when it is placed near the free end (at d ¼0.1m, Fig. 14). If a thermal field of known temperature profile h is also acting on the system, then voltage output of the piezoelectric sensor given by Eq. (62) is to be modified to include the temperature component by

  / ¼ Kff1 Kfu uþKfy y

ð63Þ

where Kfy is the assembled global finite element matrix made up of pyroelectric P coefficients of the sensor [45].

2.22.7.3

Piezoelectro-Magnetic System

The objective of this study is to test the accuracy and applicability of the piezomagnetic equations of Section 2.22.6.3 on a piezoelectro-magnetic system [45]. The test case is extended to extract the thermodynamic potential energy (strain energy in this case study) using the analytical and finite element approaches. The composite beam shown in Fig. 15 is composed of two layers bonded to each other. The top and bottom layers are made up of magnetostrictive (CoFe2O4) and piezoelectric (BaTiO3) ceramics, respectively, whose material properties are given in Table 3. The dimensions of the beam are taken as: length: L¼0.1 m, height (thickness): h¼ 0.001 m and width (depth): w ¼ 0.005 m. As depicted in Fig. 15, an electrical field is induced in the piezoelectric layer by applying an electrical voltage V to its bottom surface. The magnitude of V is taken as 1 V. The interface between the two layers is grounded.

Piezoelectric Materials

0.185 V 20.9 ms

0.4

0.2

0.0

0.2

0.4 50

0 0.5 V/div

100 10 ms/div

150

Amplitude: 0.01 V/div

0.20

0.15

0.10

0.05

0.00 0

100

200

300

400

Frequency: 20 Hz/div

500

0.010@ 60 Hz

Fig. 14 Experimental voltage output for the location of the sensor at d¼0.1 m.

y

L x

h

h/4 F

−V

Poling direction

Magnetostrictive layer Piezoelectric layer

Fig. 15 Piezoelectro-magnetic system.

Table 3

Properties of materials for piezoelectro-magnetic system CoFe2O4

c11 (Pa) ℓ31 (N/Wb or A/m) e31 (C/m2) r (kg/m3) lt (1/K) n

BaTiO3 11

1.5432  10 2.86  108 – 7600 5  10 6 0.3

Steel 11

1.3793  10 – 10.76 – – 0.3

2.07  1011 – – 7750 1.1  10 5 0.3

715

716

Piezoelectric Materials

2.22.7.3.1

Analytical approach

As shown in Fig. 15, an equivalent tensile force of magnitude e31 E3 wh ð64Þ ¼ e31 Vw 2 is induced in the beam due to the applied voltage of V to the piezoelectric layer. The composite beam is transformed to a beam made up of the magnetostrictive material using the transformed-section method [105]. The bending moment M caused by the above force at the neutral axis of the transformed beam is calculated to be   h M¼F y ð65Þ 4 F¼

where y is the distance of the neutral axis from the bottom surface of the beam. The vertical displacement of the beam due to this bending moment is then calculated as u3 ¼

M 2 x 2Ym I

ð66Þ

where Ym (c11 in Table 3) is the Young’s modulus for the magnetostrictive material and I is the second moment of area of the transformed beam about its neutral axis. The thermodynamic potential energy is computed from EG ¼ Eb þ Et

where Eb ¼

2.22.7.3.2

M2 L 2Ym I

and

ð67Þ

Et ¼

F2L 2Ym A

ð68Þ

Finite element approach

For the finite element analysis, each layer of the composite beam is divided into five equal rectangular elements with a total of 10 elements for the whole composite beam. The static displacement of the beam is found from the first equation of Eq. (22) as u¼

Kuu1 Kuf / ¼ Kuu1 Kuf V

ð69Þ

After finding the beam displacement, the thermodynamic potential energy of Eq. (67) is computed via EG ¼

1 T u Kuu u 2

ð70Þ

For the numerical values, the analytical approach yields the following: the equivalent tensile force F¼ 0.0538 N, distance of the neutral axis from the bottom surface of the beam y ¼ 0:514 h ¼ 5:14  10 4 m, bending moment M ¼ 1.42032  10 5 N
m, second moment of area of the transformed beam I¼ 0.078732bh3 ¼ 3.9366  10 13 m4 and vertical displacement u3 ¼ 1.169  10 4x2 m. Vertical displacements of the beam computed by the analytical and finite element approaches are shown in 1.2

× 10−6 Analytical approach Finite element approach

Vertical displacement (m)

1.0

0.8

0.6

0.4

0.2

0

0

0.02

0.04 0.06 Distance (m)

Fig. 16 Vertical displacement of piezoelectro-magnetic system due to applied voltage.

0.08

0.1

Piezoelectric Materials

717

Fig. 16, which indicates the close agreement between the two approaches for u3. For the thermodynamical potential energy, the analytical approach results in Eb ¼ 1.66035  10 10 J, Et ¼ 1.98058  10 10 J and EG ¼Eb þ Et ¼ 3.64093  10 10 J. On the other hand, the finite element method of Eq. (70) yields that EG ¼ 3.62397  10 10 J, which demonstrates closeness of results for EG by the two approaches.

2.22.8

Closing Remarks

Phenomena of thermopiezoelectricity, piezoelectricity and piezomagnetism are defined, and their quasi-static and dynamic equations are provided. These equations model interactions between thermal, mechanical, electrical and magnetic fields. A thermodynamic potential energy is also defined that has energy components from mechanical and electrical fields. Producing of a small amount of electric output enough to charge a micro-electromechanical device by a piezoelectric material due to a mechanical input like vibration from the existing ambient conditions is known as the piezoelectric energy harvesting, which is a new trend in energy studies. Three illustrative examples provided give a general picture about sensing and actuation capabilities of piezoelectric materials, and interaction of piezoelectric and piezomagnetic phenomena for conversion of different types of energies among them. These examples can be extended to include energy harvesting mechanisms via piezoelectricity and piezomagnetism. Some future research directions in piezoelectric materials as related to energy systems can be listed as: 1) use of thermopiezoelectricity in energy harvesting, 2) optimal design and placement of piezoelectric sensors to increase energy harvesting, 3) investigation of the thermodynamic potential energy as an energy performance indicator, and 4) further studies in coupling of piezoelectric and piezomagnetic mechanisms.

Acknowledgment The author greatly acknowledges the support provided by King Fahd University of Petroleum & Minerals through project RG13171 and RG1317-2 in writing this chapter.

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28]

Curie P, Curie J. Comptes Rendus 1880;91(294). Mason WP. Piezoelectricity, its History and Applications. J Acoust Soc Am 1981;70(6):1561–6. Langevin P. J Physique (Paris). Colloque 1972;c6(11-12). Ferreira ADBL, Novoa PRO, Marques AT. Multifunctional material systems: a state-of-the-art review. Compos Struct 2016;151:3–35. Li DS, Ho S-CM, Song GB, Ren L, Li HN. A review of damage detection methods for wind turbine blades. Smart Mater Struct 2015;24(3). Lee HJ, Zhang SJ, Bar-Cohen Y, Sherrit S. High temperature, high power piezoelectric composite transducers. Sensors 2014;14(8):14526–52. Jiang XN, Kim K, Zhang SJ, Johnson J, Salazar G. High-temperature piezoelectric sensing. Sensors 2014;14(1):144–69. Oliveira EL, Maia NMM, Marto AG, da Silva RGA, Afonso FJ, Suleman A. Modal characterization of composite flat plate models using piezoelectric transducers. Mech Syst Signal Process 2016;79:16–29. Loghmani A, Danesh M, Keshmiri M. Modal structural acoustic sensing with minimum number of optimally placed piezoelectric sensors. J Sound Vib 2016;363:345–58. Kamas T, Poddar B, Lin B, Yu LY. Assessment of temperature effect in structural helath monitoring with piezoelectric wafer active sensors. Smart Struct Syst 2015;16 (5):835–51. Yapar O, Basu PK, Volgyesi P, Ledeczi A. Structural health monitoring with piezoelectric AE sensors. Eng Failure Anal 2015;56:150–69. Lee S, Jeong S, Park J. Damage identification using flexural vibration actuated and sensed by piezoelectric transducers. Proc Inst Mech Eng C – J Mech Eng Sci 2014;228(12):2132–40. Simonovic AM, Jovanovic MM, Lukic NS, Zoric ND, Stupar SN, Ilic SS. Experimental studies on active vibration control of smart plate using a modified PID controller with optimal orientation of piezoelectric actuator. J Vib Control 2016;22(11):2619–31. Lu E, Li W, Yang XF, Fan MB, Liu YF. Modelling and composite control of single flexible manipulators with piezoelectric actuators. Shock Vib 2016; doi:10.1155/2016/ 2689178. Omidi E, Mahmoodi N. Hybrid positive feedback control for active vibration attenuation of flexible structures. IEE-ASME Trans Mech 2015;20(4):1790–7. Islam MN, Seethaler RJ. Sensorless position control for piezoelectric actuators using a hybrid position observer. IEE-ASME Trans Mechatron 2014;19(2):667–75. Lee AJ, Moosavian A, Inman DJ. Control and characterization of a bistable laminate generated with piezoelectricity. Smart Mater Struct 2017;26(8). Xu XJ, Deng ZC, Zhang K, Meng JM. Surface effects on the bending, buckling and free vibration analysis of magneto-electro-elastic beams. Acta Mech 2016;227(6). Vashishth AK, Sukhija H. Inhomogeneous waves in porous piezo-thermoelastic solids. Acta Mech 2017;228(5). Yang J, Jin XY, Jin NG. A penny-shaped crack in an infinite linear transversely isotropic medium subjected to uniform anti-symmetric heat flux: closed-form solution. Eur J Mech A – Solids 2014;47:254–70. Kim HS, Kim J-H, Kim J. A review of energy harvesting based on vibration. Int J Precision Eng Manuf 2011;12(6):1129–41. Ahmed R, Mir F, Banerjee S. A review on energy harvesting approaches for renewable energies from ambient vibrations and acoustic waves using piezoelectricity. Smart Mater Struct 2017;26(8). Sodano HA, Inman DJ, Park GH. A review of power harvesting from vibration using piezoelectric materials. Shock Vib Digest 2004;36(3):197–205. Lee J-H, Kim J, Kim TY, Al Hossain MS, Kim S-W, Kim JH. All-in-one energy harvesting and storage devices. J Mater Chem A 2016;4(21):7983–99. Siddique AM, Mahmud S, Van Heyst B. A comprehensive review on vibration based micro power generators using electromagnetic and piezoelectric transducer mechanisms. Energy Conversion Manage 2015;106:728–47. Crossley S, Whiter RA, Kar-Narayan S. Polymer-based nanopiezoelectric generators for energy harvesting applications. Mater Sci Technol 2014;30(13A):1613–24. Atsushi B, Clifford TS, Bernhard RT. High temperature ultrasonic transducer up to 10001C using lithium niobate single crystal. Appl Phys Lett 2010;97:232901. Wang ZL, Song JH. Piezoelectric nanogenerators based on zinc oxide nanowire arrays. Science 2006;312:242–6.

718

Piezoelectric Materials

[29] Yu H, Matteo R. Fast and high resolution thermal detector based on an aluminum nitride piezoelectric microelectromechanical resonator with an integrated suspended heat absorbing element. Appl Phys Lett 2013;102:093501. [30] Berlincourt D, Krueger HHA, Near C. Properties of Morgan electro ceramic ceramics, Technical Publication TP-226. [31] Kawai H. The piezoelectricity of poly(vinylidene fluoride). Japan J Appl Phys 1969;8(7):975–6. [32] Cauda V, Canavese G, Stassi S. Nanostructured piezoelectric polymers. J Appl Polym Sci 2015;132(13). [33] Jain A, Prashanth KJ, Sharma AK, Jain A, Rashmi PN. Dielectric and piezoelectric properties of PVDF/PZT composites: a review. Polym Eng Sci 2015;55:1589–616. [34] Ramadan KS, Sameoto D, Evoy S. A review of piezoelectric polymers as functional materials for electromechanical transducers. Smart Mater Struct 2014;23(3). [35] Khan A, Abas Z, Kim HS, Oh I-K. Piezoelectric thin films: an integrated review of transducers and energy harvesting. Smart Mater Struct 2016;25(5). [36] Mindlin RD. On the equations of motion of piezoelectric crystals. In: Radok J, editor. Problems of Continuum Mechanics. Philadelphia: SIAM; 1961. p. 282–90. [37] Mindlin RD. Equations of high frequency vibrations of thermopiezoelectric crystal plates. Int J Solids Struct 1974;10(6):625–37. [38] Nowacki W. Some general theorems of thermopiezoelectricity. J Therm Stresses 1978;l(2):l71–182. [39] Nowacki W. Mathematical models of phenomenological piezoelectricity. New problems in mechanics of continua. Waterloo, ON: University of Waterloo Press; 1983. p. 29–49. [40] Lebon G. A generalized theory of thermoelasticity. J Tech Phys 1982;23(1):37–46. [41] Chandrasekharaiah DS. A generalized linear thermoelasticity theory for piezoelectric media. Acta Mech 1988;71:39–49. [42] Vashishth AK, Sukhija H. Reflection and transmission of plane waves from fluid-piezothermoelastic solid interface. Appl Math Mech – Engl Ed 2015;36(1):11–36. [43] Lumentut MF, Howard IM. Intrinsic electromechanical dynamic equations for piezoelectric power harvesters. Acta Mech 2017;228(2):631–50. [44] Hasanyan AD, Hasanyan DJ. Energy harvesting performance of dynamic bimorph thermopiezoelectric benders with arbitrary support location. J Therm Stresses 2015;38:1409–27. [45] Sunar M, Al-Garni AZ, Ali MH, Kahraman R. Finite element modeling of thermopiezomagnetic smart structures. AIAA J 2002;40:1846–51. [46] Mindlin RD. Equations of high frequency vibrations of thermopiezoelectric crystal plates. Int J Solids Struct 1974;10:625–37. [47] Heinbockel JH. Introduction to tensor calculus and continuum mechanics. Bloomington, IN: Trafford Publishing; 2001. [48] Sadiku MNO. Elements of electromagnetics. 4th ed. New York, NY: Oxford University Press; 2006. [49] Erturk A, Inman DJ. Piezoelectric energy harvesting. 1st ed. New York: Wiley; 2011. [50] Shenck NS, Paradiso JA. Energy scavenging with shoe-mounted piezoelectrics. Micro IEEE 2001;21(3):30–42. [51] Hwang GT, Byun M, Jeong CK, Lee KJ. Flexible piezoelectric thin-film energy harvesters and nanosensors for biomedical applications. Adv Healthcare Mater 2015;4 (5):646–58. [52] De Marqui Jr. C, Vieira WGR, Erturk A, Inman DJ. Modeling and analysis of piezoelectric energy harvesting from aeroelastic vibrations using the doublet-lattice method. ASME J Vib Acoust 2010;133(1). [53] Piezoelectrics Ceramics Data Book for Designers, Transducer Products Division, Morgan Matroc Limited, England. [54] Warkentin DJ, Crawley EF, Senturia SD. The feasibility of embedded electronics for intelligent structures. J lntell Mater Syst Struct 1992;3(3):462–82. [55] Wada BK, Fanson JI, Crawley EF. Adaptive structures. In: ASME winter annual meeting, Vol. 15, San Francisco, CA, ASME, AD; 1989. [56] Yang C, Wang J-M. Exponential stability of an active constrained layer beam actuated by a voltage source without magnetic effects. J Math Anal Appl 2017;448 (2):1204–27. [57] Mohammadi S, Hatam S, Khodayari A. Modeling of a hybrid semi-active/passive vibration control technique. J Vib Control 2015;21(1):21–8. [58] Kumar N, Singh SP. Vibration control of curved panel using smart damping. Mech Syst Signal Process. 2012;30:232–47. [59] Providakis CP, Kontoni D-PN, Voutetaki ME, Stavroulaki ME. Comparisons of smart damping treatments based on FEM modeling of electromechanical impedance. Smart Struct Syst 2008;4(1):35–46. [60] Trindade MA. Optimization of active-passive damping treatments using piezoelectric and viscoelastic materials. Smart Mater Struct 2007;16(6):2159–68. [61] APC International, Ltd., https://www.americanpiezo.com/piezo-theory/pzt.html. [62] Panda PK, Sahoo B. Optimization of active-passive damping treatments using piezoelectric and viscoelastic materials. Ferroelectrics 2015;474(1):128–43. [63] SDVOSB, http://www.reade.com/products/lead-zirconate-titanate-pzt. [64] Rezaei M, Khadem SE, Firoozy P. Broadband and tunable PZT energy harvesting utilizing local nonlinearity and tip mass effects. Int J Eng Sci 2017;118:1–15. [65] Kuo C-L, Lin S-C, Wu W-J. Fabrication and performance evaluation of a metal-based bimorph piezoelectric mems generator for vibration energy harvesting. Smart Mater Struct 2016;25(10). [66] Savarimuthu K, Sankararajan R, Murugesan S. Design and implementation of piezoelectric energy harvesting circuit. Circuit World 2017;43(2):63–71. [67] Vaish M, Sharma M, Vaish R, Chauhan VS. Harvesting thermal energy (via radiation) using pyroelectric materials (PZT-5H): an experimental study. Ferroelectr Lett Secion 2017;44(1-3):35–41. [68] Butt Z, Pasha RA, Qayyum F, Anjum Z, Ahmad N, Elahi H. Generation of electrical energy using lead zirconate titanate (PZT-5A) piezoelectric material: analytical, numerical and experimental verifications. J Mech Sci Technol 2016;30(8):3553–8. [69] Leadenham S, Ferri B, Erturk A. Modeling and identification of nonlinear electroelastic and dissipative parameters for PZT-5A and PZT-5H bimorphs: a dynamical systems approach In: Liao WH, Park G, Erturk A, editors. Conference on active and passive smart structures and integrated systems, San Diego, CA; 2015. Proc SPIE 9431. [70] Erturk A, Lee H-Y, Inman DJ. Investigation of soft and hard ceramics and single crystals for resonant and off-resonant piezoelectric energy harvesting. In: Proc of the ASME conference on smart materials, adaptive structures and intelligent systems, Vol. 1; 2010. p. 589–96. [71] Tzou HS, Pandita S. A multi-purpose dynamic and tactile sensor for robot manipulators. J Robotic Syst 1987;4(6):719–41. [72] Anjana J, Prashanth KJ, Asheesh Kr S, Arpit J, Rashmi PN. Dielectric and piezoelectric properties of PVDF/PZT composites: a review. Wiley Online Library Society of Plastics Engineers; 2015. [73] Piezotech Arkema, www.piezotech.eu. [74] Kim TH. Characterization and applications of piezoelectric polymers, Technical Report No. UCB/EECS-2015-253, Electrical Engineering and Computer Sciences University of California at Berkeley; 2015. [75] Bauer S, Bauer F. Piezoelectric polymers and their applications. Berlin: Springer; 2008. Chapter 6. [76] Chakhchaoui N, Ennamiri H, Hajjaji A, Eddiai A, Meddad M, Boughaleb Y. Theoretical modeling of piezoelectric energy harvesting in the system using technical textile as a support. Polym Adv Technol 2017;28(9):1170–8. [77] Zhang J, Fang Z, Shu C, Zhang J, Zhang Q, Li C. A rotational piezoelectric energy harvester for efficient wind energy harvesting. Sensors Actuators A – Phys 2017;262:123–9. [78] Pan F, Xu Z, Jin L, Pan P, Gao X. Designed simulation and experiment of a piezoelectric energy harvesting system based on vortex-induced vibration. IEEE Trans Ind Appl 2017;53(4):3890–7. [79] Ding R, Zhang X, Chen G, et al. High-performance piezoelectric nanogenerators composed of formamidinium lead halide perovskite nanoparticles and poly(vinylidene fluoride). Nano Energy 2017;37:126–35. [80] Jung I, Shin Y-H, Kim S, Choi J-Y, Kang C-Y. Flexible piezoelectric polymer-based energy harvesting system for roadway applications. Appl Energy 2017;197:222–9. [81] Wang Y, Fang M, Tian B, et al. Transparent PVDF-TrFE/graphene oxide ultrathin films with enhanced energy harvesting performance. Chem Select 2017;2(26):7951–5. [82] Toprak A, Tigli O. MEMS scale PVDF-TrFE-based piezoelectric energy harvesters. J Microelectromech Syst 2015;24(6):1989–97.

Piezoelectric Materials

[83] [84] [85] [86] [87] [88] [89] [90] [91] [92] [93] [94] [95] [96] [97] [98] [99] [100] [101] [102] [103] [104] [105]

719

Canavese G, Stassi S, Cauda V, et al. Different scale confinements of PVDF-TrFE as functional material of piezoelectric devices. IEEE Sensors J 2013;13(6):2237–44. Zhang Y, Zhang YJ, Xue X, et al. PVDF-PZT nanocomposite film based self-charging power cell. Nanotechnology 2014;25(10). Ogata K. System dynamics. 4th ed. Upper Saddle River NJ: Pearson Education International; 2004. The MathWorks, Inc., MATLAB, Version 6.5, Release 13. Natick, MA: The MathWorks, Inc.; 2002. IEEE Std 319-1990, IEEE standard on magnetostrictive materials: piezomagnetic nomenclature; 1990. Body C, Reyne G, Meunier G. Modeling of magnetostrictive thin films, application to a micromembrane. J Physics III France 1997;7:67–85. Caltun O, Dumitru I, Feder M, Lupu N, Chiriac H. Substituted cobalt ferrites for sensors applications. J Magn Magn Mater 2008;320(20):e869–73. Lu Y, Yoebin X, Guoxi X, Yong F. Synthesis of cobalt ferrite with enhanced magnetostriction properties by the sol–gel-hydrothermal route using spent Li-ion battery. J Alloys Compounds 2016;680:73–9. McHugh W. Properties of nickel as a magnetostrictive material for ultrasonic conditions, Dissertation for BS Degree, University of Southern Queensland, Australia; 2011. Sherman CH, Butler JL. Transducers and arrays for underwater sound. New York, NY: Springer; 2007. Masumoto H, Saito H. Sci. Rep. RITU A 1952;4:338. Yamashiro Y, Teshima N, Narita K. Magnetic properties of rapidly quenched alperm ribbons. J Magn Magn Mater 1984;41:149–51. Lanza G, Breglio G, Giordano M, Gaddi A, Buontempo S, Cusano A. Effect of the anisotropic magnetostriction on Terfenol-D based fiber bragg grating magnetic sensors Sensors Actuators A: Phys 2011;172 (2):420–7. Zhou Z-G, Wu L-Z, Wang B. The behavior of a crack in functionally graded piezoelectric/piezomagnetic materials under anti-plane shear loading. Arch Appl Mech 2005;4 (8):526–35. Cao SY, Zheng JJ, Wang BW, et al. Mechanical-magnetic-electric coupled behaviors for stress-driven Terfenol-D energy harvester. AIP Adv 2017;7(5). Domann JP, Carman GP. Strain powered antennas. J Appl Phys 2017;121(2). Wang WJ, Li P, Jin F. Two-dimensional linear elasticity theory of magneto-electro-elastic plates considering surface and nonlocal effects for nanoscale device applications. Smart Mater Struct 2016;25(9). Ibrahim M, Salehian A. Modeling, fabrication, and experimental validation of hybrid piezo-magnetostrictive and piezomagnetic energy harvesting units. J Intell Mater Syst Struct 2015;26(10):1259–71. Fan T, Zou GP, Yang LH. Nano piezoelectric/piezomagnetic energy harvester with surface effect based on thickness shear mode. Compost B – Eng 2015;74:166–70. Zhao Y-P, Lu ST, Kong DJ, Zhang CL. A magnetic energy harvester using functional graded multiferroic cylinders. In: Tao X, Zhao X, Wang C, Yu F, editors. Proc. of the 2015 symposium on piezoelectricity, acoustic waves and device applications; p. 485–9. Delnavaz A, Voix J. Piezo-magnetic energy harvesting from movement of the head. In: 15th international conference on micro and nanotechnology for power generation and energy conversion applications (POWERMEMS 2015). J Phys Conf Ser 2015;660. Tzou HS. Development of a light-weight robot end-effector using polymeric piezoelectric bimorph. In: Proc. of the 1989 IEEE international conference on robotics and automation, Scottsdale, AZ, May 14–19. Computer Society Press, Los Angeles, CA; 1989. p. 1704–9. Hibbeler RC. Mechanics of materials. 3rd ed. Piscataway, NJ: Prentice-Hall, Inc; 1997. p. 319–25.

Further Reading Bowen CR, Topolov VY, Kim HA. Modern piezoelectric energy-harvesting materials. Berlin: Springer; 2016. Briand D, Yeatman E, Roundy S, editors. Micro energy harvesting New York: Wiley; 2015. Erturk A, Inman DJ. Piezoelectric energy harvesting. 1st ed. New York: Wiley; 2011. Khan A, Abas Z, Kim HS, Oh IK. Piezoelectric thin films: an integrated review of transducers and energy harvesting. Smart Mater Struct 2016;25(5). Kim HS, Kim J-H, Kim J. A review of piezoelectric energy harvesting based on vibration. Int J Precision Eng Manuf 2011;12(6). Shashank P, Inman DJ, editors. Energy harvesting technologies Berlin: Springer; 2009. Siddique AM, Mahmud S, Van Heyst B. A comprehensive review on vibration based micro power generators using electromagnetic and piezoelectric transducer mechanisms. Energy Convers Manage 2015;106:728–47. Sunar M, Rao SS. Recent advances in sensing and control of flexible structures via piezoelectric materials technology. ASME Appl Mech Rev 1999;52:1–16. Toprak A, Tigli O. Piezoelectric energy harvesting: state-of-the-art and challenges. Appl Phys Rev 2014;1(3). Wei CF, Jing XJ. A comprehensive review on vibration energy harvesting: modeling and realization. Renew Sustain Energy Rev 2017;74:1–18.

Relevant Websites https://www.americanpiezo.com/ APC, International Ltd. https://www.ceramtec.com/electro-ceramics/ CeramTec. http://www.nanomotion.com/piezo-ceramic-motor-technology/piezoelectric-effect/ NANOMOTION. http://www.piezo.com/ PIEZO SYSTEMS, INC.

2.23 Pyroelectric Materials Vincent Ming Hong Ng, Nanyang Technological University, Singapore, Singapore Ling Bing Kong, Nanyang Technological University, Singapore, Singapore and Bengbu University, Bengbu, Anhui, P.R. China Wenxiu Que, Xi’an Jiaotong University, Xi’an, Shaanxi, P.R. China Chuanhu Wang, Bengbu University, Bengbu, Anhui, P.R. China Sean Li, The University of New South Wales, Sydney, NSW, Australia Tianshu Zhang, Anhui Target Advanced Ceramics Technology Co. Ltd., Hefei, Anhui, P.R. China r 2018 Elsevier Inc. All rights reserved.

2.23.1 Introduction 2.23.2 Pyroelectric Effect 2.23.2.1 Definition of Pyroelectricity 2.23.2.2 Pyroelectric Coefficient and Electrocaloric Coefficient 2.23.2.3 Primary and Secondary Pyroelectric Coefficient 2.23.2.4 Tertiary Pyroelectric Coefficient and Other Aspects 2.23.2.5 Pyroelectric Effect Versus Phase Transition 2.23.2.6 Pyroelectric Figure of Merit 2.23.3 Pyroelectric Materials 2.23.3.1 Triglycine Sulphate 2.23.3.2 Polyvinylidene Fluoride 2.23.3.3 Lithium Tantalate 2.23.3.4 Strontium Barium Niobate 2.23.3.5 Perovskite Structured Pyroelectric Ferroelectrics 2.23.3.6 Lead Germanate 2.23.4 Olsen Cycle of Pyroelectric Effect 2.23.5 Thermal Energy Harvesters Based on Pyroelectric With Olsen Cycle 2.23.5.1 Thermal Subsystem 2.23.5.2 Electrical Subsystem 2.23.5.3 Numerical Simulation 2.23.5.4 Governing Equations 2.23.5.5 Initial and Boundary Conditions 2.23.5.6 Material Properties 2.23.5.7 Solution Method 2.23.5.8 Evaluation of Performances of the Pyroelectric Converters 2.23.6 Olsen Harvesters Based on Pyroelectric Polymers 2.23.6.1 Device Assembly and Characterization 2.23.6.2 Performances 2.23.6.3 Heat Conduction 2.23.7 Olsen Harvesters Based on Perovskite Pyroelectric Materials 2.23.7.1 Lead Lanthanum Zirconate Titanate Ceramics 2.23.7.2 Perovskite Relaxor-Ferroelectric Single Crystals 2.23.8 New Pyroelectric Materials and Devices 2.23.9 Future Directions References Further Reading Relevant Websites

Nomenclature d D e E Fi G i p

720

Reduced electric displacement Electric displacement Reduced electric field Electric field Current figure of merit Gibbs free energy Pyroelectric current Reduced polarization

P Ps Ri S t T TC W X

721 722 722 722 724 724 726 727 729 729 729 729 730 730 732 732 733 733 734 734 734 735 736 736 736 737 737 739 742 745 745 746 747 756 756 758 758

Polarization Saturation polarization Current response Entropy Reduced absolute temperature Absolute temperature pyroelectric coefficient Currie temperature Incident power Strain

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00249-2

Pyroelectric Materials

Symbols a e0 e er

Thermal expansion coefficient Dielectric constant of vacuum Dielectric constant of materials Relative dielectric constant

Abbreviations BST DTGS ECE PLZT

2.23.1

Barium strontium titanate Deuterated-triglycine sulfate Electrocaloric effect piezoelectric constant Lead lanthanum zirconate titanate

Z sT tT o

Absorption percentage Thermal relaxation time Thermal conductivity Angular frequency

PVDF PZT TGFB TGS VDF-TRFE

Polyvinylidene fluoride Lead zirconate titanate Triglycine fluoroberyllate Triglycine sulphate Vinylidene fluoride with trifluoroethylene

721

Introduction

Our daily life is highly dependent on oil and coal, but it has been well recognized that they will be used up sometime in the future [1,2]. Another issue, i.e., carbon-dioxide (CO2) emissions and climate change, such as global warming, has drawn much international attention. The dire consequence of global warming and climate change could be the rise of sea level, which will lead to geographical crisis in many countries around the world. In order to address these two issues, there are three ways, i.e., (1) finding new energy, (2) increasing efficiency of current energy utilization, and (3) harvesting various energies in our daily life. Obviously, increasing energy efficiency cannot completely resolve the problem, because it only prolongs the time used to consume the energy resource on Earth. Therefore, discovering new and alternative energy sources is one of the most studied research topic all around the world. Renewable energies include solar energy, wind energy, geothermal energy, and so on. Among them, solar energy is considered to be worth the name to the most extent. Although there is no exact definition, waste energies are usually used to mean the energies that can be or should be used for us, but actually dissipated or wasted. Waste energies are generally classified into two categories, (1) mechanical waste energies and (2) thermal waste energies. Mechanical waste energies are those that are related to all kinds of movement, such as machines in operation, flowing gases or liquids, human body motions, and so on. In this respect, wind resistance can be included as a waste energy. Mechanical energies can be harvested by using piezoelectric, electromagnetic and triboelectric devices, as well as nanogenerators. Thermal waste energies are those related to activities that use thermal energies, mainly due to the low efficiency of energy utilization as mentioned earlier. Waste thermal energy harvesting can be accomplished with pyroelectric, thermoelectric (TE), and phase change effects. In this chapter, we focus on pyroelectric effect and devices for thermal energy harvesting. To introduce pyroelectric materials, it is necessary to first start with piezoelectric materials. Piezoelectricity has close relationships with ferroelectricity and pyroelectricity. Crystallographically, there are 32 distinct classes of dielectric materials; according to the theory of point groups, in which several symmetry elements are involved, including (1) center of symmetry, (2) axis of rotation, (3) mirror planes, and (4) several combinations of them. Out of the 32 point groups, 21 classes are noncentrosymmetric and could result in piezoelectricity. However, due to some reasons, 20 of the 21 groups are of piezoelectric effect. Among these 20 piezoelectric crystal classes, there are 10 crystals that have pyroelectric effect. This class of materials has permanent polarization, over a certain temperature range. Consequential of the variation in polarization as the temperature is changed, the effect is thus named as pyroelectricity. Within a subgroup of the pyroelectric group, there are materials with spontaneous polarization and the polarization may be reversed via an applied external electric field. This phenomenon is known as ferroelectric effect, with the name containing “ferro” to mean iron (Fe), because it is very similar to ferromagnetic property, in various aspects. There is a slight difference even though the polarizations in a ferroelectric and pyroelectric material are seemingly similar. The reversibility of ferroelectric polarization in response to an external electric field, on the condition that the applied field does not exceed dielectric breakdown of the materials. As such, ferroelectric materials must possess spontaneous polarization and reversible polarization under an applied electric field. Therefore, it is concluded that ferroelectric materials possess much enhanced pyroelectric performance as compared to nonferroelectric pyroelectric materials [3–5]. The prefix “pyro-” is originated from the Greek phrase “pyr,” which is to mean “fire,” implying its relation to either fire or heat. Pyroelectric effect, also known as pyroelectricity, is used to relate between thermal and electrical energy. Pyroelectricity was once mainly employed for the application as detectors [6]. More recently, it has been widely explored for applications related to renewable energy generation and environmental protection [7–10]. Different from TE materials, pyroelectric materials require temporal variations in temperature and not a temperature gradient (spatial gradient) [7]. Therefore, they can be used for applications in fields where there are no temperature gradients and the temperatures are not static. This chapter intends to provide an overview on the progress in pyroelectric materials and devices [11]. Principles of pyroelectricity, such as definition, pyroelectric coefficients, phase transition, etc., will be briefly discussed. After that, the types and

722

Pyroelectric Materials

properties of the reported pyroelectric materials will be presented and discussed. Finally, the latest development in waste heat energy harvesters employing pyroelectric materials, such as pyroelectric polymers, perovskite ceramics and single crystals, are comprehensively compiled, with emphasis on the Olsen cycle.

2.23.2 2.23.2.1

Pyroelectric Effect Definition of Pyroelectricity

Pyroelectricity is defined as the effect that has a close association with the variation in polarization of a material, which is caused by the fluctuation in temperature [12,13]. In ferroelectric materials with single domains, there is accumulation of localized charges at the two end of samples’ surfaces, owing to the aligned electric dipoles. As the thermal equilibrium state is reached, equal number of oppositely charged free charges shields all the localized charges. As a consequence, the ferroelectric materials have no net electricity. If the temperature is altered, the polarization in the materials will be changed to respond to the temperature variation. Therefore, some of the localized charges will be devoid of shielding by the free charges. Accordingly, the surfaces will be charged, thus leading to the availability of electric fields. The formation of the electric field can be evidenced by using other charged particles. Depending the sign of the charges, the particles can be either attracted or repelled by the pyroelectric materials. An external circuit can be connected to the surfaces of the pyroelectric materials to allow the electrical current to be flowing out. Obviously, the directions of the electrical currents are different when the pyroelectric materials are heated and cooled. Pyroelectric coefficient is an important factor to characterize pyroelectric effect of pyroelectric materials, which represents pyroelectric efficiency of the materials. As a variation in temperature, DT, is applied to pyroelectric device, a variation in polarization is triggered, which is given by: DP ¼ pDT in which p refers to the pyroelectric coefficient, a vector with three nonzero components, with a unit of C m expressed as: pm ¼

∂Pm ; m ¼ 1; 2; 3 ∂T

ð1Þ 2

K 1, which are

ð2Þ

Depending on the piezoelectric axis of the piezoelectric crystals, the pyroelectric coefficient has different signs. In accordance to the IRE standards, the positive end refers to the end of a crystal axis under tension that possess positive charge. During the heating of a pyroelectric crystal, when positive charges are present on the positive end, the pyroelectric coefficient is determined as positive. Correspondingly, if negative charges are generated on the positive end, the pyroelectric coefficient becomes negative. Generally, because ferroelectric materials experience a decrease in spontaneous polarization as they are heated, pyroelectric coefficient is usually negatively signed. However, in some ferroelectric materials, such as Rochell salt, the spontaneous polarization is increased with an increase in temperature, especially to a temperature close to but not exceeding the Curie point. In view of the piezoelectric effect in all pyroelectric materials, they can also have variation in polarization, when they are subject to geometric deformation that is caused by the change in temperature, which also has contribution to the pyroelectric efficiency. As a result, mechanical boundary conditions should be taken care of when characterizing pyroelectric effect. At given mechanical boundary conditions, there could be two types of pyroelectric effects, as the cooling or heating is sufficiently uniform. For instance, if the sample is subjected to a constant strain, pyroelectric effect is merely contributed by the alteration in polarization induced by the temperature variation. Accordingly, the pyroelectric effect is termed the primary pyroelectric effect or constant strain pyroelectric effect. However, in practice, pyroelectric device cannot be maintained at a condition of constant strain. Therefore, the primary pyroelectric effect always contains the contribution from the change in polarization induced by the thermal expansion of the material and concomitant mechanical action. If the temperature is evenly distributed inside the material, the extra pyroelectric effect will be known as the secondary pyroelectric effect. In other words, the pyroelectric coefficient of a pyroelectric material under constant stress is actually a summation of pyroelectric coefficients, comprising of both primary and secondary coefficients [14]. If the pyroelectric materials are inhomogeneously heated or cooled, there will be stress gradient within the materials, which can contribute to the pyroelectric effect, due to the piezoelectric effect. It is so named the tertiary effect or false effect for it is actually an intrinsic property of piezoelectric materials that are inhomogeneously heated or cooled. In other words, this effect should be observed in all piezoelectric materials, including those that are non-pyroelectric. As a consequence, it is very important to have a uniform heating or cooling during the characterization of pyroelectric materials, in order to minimize or eliminate the false pyroelectric contribution.

2.23.2.2

Pyroelectric Coefficient and Electrocaloric Coefficient

Variables, including temperature (T), entropy (S), electric field (E), electric displacement (D), strain (X), and stress (x), can be employed to characterize the thermodynamic states of elastic dielectrics [13]. Assuming T, E, and X are independent variables, the

Pyroelectric Materials

differential of electric displacement can be written as:       ∂Dm ∂Dm ∂Dm E;T E;X dXi þ dEn þ dT ¼ dmi dXi þ eX;T dDm ¼ mn dEn þ pm dT ∂Xi E;T ∂En X;T ∂T E;X

723

ð3Þ

where the subscripts have values of m ¼ 1–3 and i¼ 1–6, while the superscripts are to imply the constant values of the physical quantities. The first two terms are related to piezoelectric and dielectric properties, respectively, while the third term represents pyroelectric effect. At the conditions of dXi ¼ 0 and dEn ¼ 0, the following equation can be arrived: dDm ¼ pE;X m dT

ð4Þ

Because T, E, and X are independent variables, the Gibbs free energy can be described by the following characteristic function: G¼U

TS

Xi xi

ð5Þ

Em D

Incorporating the first and second thermodynamic laws: dG ¼

xi dXi

Dm dEm

SdT

As a result, the Gibbs free energy will have a differential form, with respect to T, E and X, which is given by:       ∂G ∂G ∂G dXi þ dEm þ dT dG ¼ ∂Xi E;T ∂Em X;T ∂T E;X

ð6Þ

ð7Þ

According to Eqs. (6) and (7), there will be: 

 ∂G ¼ Dm ∂Em X;T   ∂G ¼ S ∂T E;X  2    ∂ G ∂Dm ¼ ¼ pE;X m ∂T E;X ∂Em ∂T X  2    ∂ G ∂S ¼ ∂Em ∂T X ∂Em X;T

ð8Þ ð9Þ

ð10Þ

ð11Þ

Here, Eq. (10) gives the definition of pyroelectric coefficient, while the variation in entropy induced by an applied external electric field is given by Eq. (11), which is the so-called electrocaloric coefficient. In other words, electrocaloric effect (ECE) is inverse pyroelectric effect, which is given by:   ∂S E;X pm ¼ ð12Þ ∂Em X;T This expression implies that under constant electric fields and stresses, the pyroelectric coefficient is the same as the ECE at constant stresses and temperatures. Considering T, E, and x are independent variables, the differential of electric displacement can be rewritten as:       ∂Dm ∂Dm ∂Dm E;T E;x dxi þ dEn þ dT ¼ emi dxi þ ex;T ð13Þ dDm ¼ mn dEn þ pm dT ∂xi E;T ∂En x;T ∂T E;x If both the strain and electric field are constant, we can arrive at the following equation: E;x dT dDm ¼ pm

ð14Þ

Similarly, with respect to independent variables, T, E and x, the Gibbs free energy will have the following characteristic function: dG2 ¼ Xi dxi

Dm dEm

SdT

ð15Þ

Analogy to Eq. (7), we can have: dG2 ¼



∂G2 ∂xi



dxi þ E;T



∂G2 ∂Em





¼

dEm þ x;T

  ∂G2 dT ∂T E;x

ð16Þ

Therefore, there will be the following two expressions: 

∂G2 ∂Em



∂G2 ∂T

Dm

ð17Þ

S

ð18Þ

x;T



¼ E;x

724

Pyroelectric Materials

With partial derivatives, the following equation can be derived:   ∂S E;x pm ¼ ∂Em x;T

ð19Þ

This confirms that under constant electric fields and stresses, the pyroelectric coefficient is actually the ECE at constant stresses and temperatures, as stated above. The change in entropy of the ferroelectric materials can be readily attributed to the variation in the degree of ordering of polarization. A higher degree of ordering corresponds to a low value of entropy and vice versa. Therefore, as a ferroelectric material is depolarized, the value of entropy will be increased. This is simply due to a reduction in the degree of ordering of polarization. At adiabatic conditions, depolarization will induce a drop in temperature, which can be utilized for refrigeration. In this aspect, ECE is in fact a much researched topic globally.

2.23.2.3

Primary and Secondary Pyroelectric Coefficient

Under constant electric field or even without an electric field, electric displacement can be affected by both strain and temperature, where the strain can be determined by the stress and temperature, so that there are following equations:     ∂Dm ∂Dm dDm ¼ dT ð20Þ dxi þ ∂xi T ∂T x dxi ¼



∂xi ∂Xj



dXj þ T



∂xi ∂T



dT

ð21Þ

X

If dXj ¼0, by combining Eqs. (20) and (21), we have:       ∂Dm ∂xi ∂Dm dDm ¼ þ dT ∂xi T ∂T X ∂T x

ð22Þ

Rearranging Eq. (22) gives:         ∂Dm ∂Dm ∂Dm ∂xi ¼ þ ∂T X ∂T x ∂xi T ∂T X

ð23Þ

in which the left hand side’s term refers to the overall pyroelectric coefficient, the first and second terms on the right hand side correspond to the primary and secondary pyroelectric coefficients, respectively. In Eq. (23), there are:   ∂Dm ¼ emi ð24Þ ∂xi T   ∂xi ¼ ai ð25Þ ∂T X where emi refers to a piezoelectric constant and ai refers to thermal expansion coefficient. Accordingly, Eq. (23) can be expressed as: pXm ¼ pxm þ eTmi aXi

ð26Þ

in which the latter term at the right hand side implies the derivation of secondary pyroelectric coefficient by multiplying the piezoelectric constant with thermal expansion coefficient. At the same time, Eq. (23) can be rearranged as:           ∂xj ∂Dm ∂Dm ∂Dm ∂Xi ð27Þ ¼ þ ∂T X ∂T x ∂Xi T ∂xj T ∂T X so that the total pyroelectric coefficient is given by: pXm ¼ pxm þ dTmi cTij aXj

ð28Þ

in which the latter term at the right hand side indicates secondary pyroelectric coefficient can be derived by the multiplication of piezoelectric constant, elastic stiffness, and thermal expansion coefficient. Consequentially, the primary coefficient contributes to most of pyroelectric effect exhibited by a pyroelectric material.

2.23.2.4

Tertiary Pyroelectric Coefficient and Other Aspects

As inhomogeneous heating or cooling is accompanied by mechanical stress, the effective pyroelectric coefficient contains primary, secondary, and tertiary coefficients. The tertiary pyroelectric coefficient is induced by the thermal stress, with a contribution to the variation in polarization to be given by dmnpXnp(r, t), where dmnp is a piezoelectric constant and Xnp is a thermal stress component, whereas r refers to displacement and t refers to time [15]. It is difficult to characterize the thermal stress, due to the fact that it is a function of location and time. Therefore, accurate measurement of the tertiary pyroelectric coefficient is still a challenge.

Pyroelectric Materials

725

When using elastic Gibbs free energy G1 as the characteristic function and assuming that stress and electric field, are one dimensional variables, we will have: dG1 ¼

SdT þ xdX þ EdD

ð29Þ

∂G1 ¼E ∂D

ð30Þ

∂2 G 1 1 ¼ ∂D2 e

ð31Þ

∂G1 ¼ aD þ bD3 þ gD5 ∂D

ð32Þ

The derivative of Eq. (29) gives out:

Close to the Curie temperature, we have:

It can be rearranged as: E ¼ a0 ðT

T0 ÞD þ ðjÞðDÞ

ð33Þ

where (j)(D) includes all the higher order terms and there is a0 ¼ 1/(e0C), in which C is the Curie constant, thus resulting in: ∂E ¼ a0 D ∂T In this case, when E¼ 0 and D ¼Ps, and close to the Curie temperature, we have: a0 ðT

T0 ÞPs þ ðjÞðPs Þ ¼ 0

ð34Þ

ð35Þ

from where there are: ∂G1 ¼ a0 ðT ∂D D ¼ P

T0 ÞPs þ ðjÞðPs Þ ¼ 0

ð36Þ

s

  d ∂G1 ∂2 G1 dD ∂2 G1 ¼ ¼0 þ ∂D2 D ¼ P dT ∂T∂D D ¼ P dT ∂D D ¼ Ps s

ð37Þ

s

Comparison of Eqs. (31) and (34) yields:

∂2 G1 ∂E ¼ ¼ a0 Ps ∂T∂D D ¼ P ∂T D ¼ P s

By combining Eqs. (37) and (38), we have:

ð38Þ

s

p þ a0 Ps ¼ 0 er in which p¼dD/dT refers to the pyroelectric coefficient, so that there is:

ð39Þ

p Ps ¼ C er

ð40Þ

Therefore, the pyroelectric coefficient is linked with the Curie constant, spontaneous polarization and relative dielectric constant of the pyroelectric materials. As a result, pyroelectric coefficient may be roughly derived from the Curie constant, and vice versa. However, a caveat to the formula is that it is only applicable over a relatively narrow temperature range. Interestingly, although the pyroelectric coefficient and dielectric constant of different pyroelectric/ferroelectric materials could be largely different, the ratio, per 1/2, remains almost the same for majority of ferroelectric materials, which is valid in a temperature range that includes room temperature. This observation can be interpreted by various ferroelectric phenomenological theories. According to Devonshire, the elastic Gibbs free energy G1 can be described as follows: ∂G1 ¼ a0 ðT ∂D

T0 ÞD þ bD3 þ gD5 ¼ E

ð41Þ

Close the Curie temperature, if there are E¼0 and D ¼ Ps, we will have: T0 ÞPs þ bPs3 þ gPs5 ¼ 0

a0 ðT

ð42Þ

This equation has the following solutions: Ps2 ¼

b 2g

n

Ps2 ¼

b 2g

n

 1þ 1

4a0 gb 2 ðT

T0 Þ

 1

4a0 gb 2 ðT

T0 Þ

1

1=2 o 1=2 o

ð43Þ ð44Þ

726

Pyroelectric Materials

Eqs. (43) and (44) are for bo0 and b40, respectively. In the case of second order phase transition, where b40, Eq. (43) can be expressed as: " #1=2 b b 4gðT0 TÞ ð45Þ 1þ Ps2 ¼ 2g 2g e0 Cb2 In case of first order phase transition, we have: Ps2

jbj 2g

¼

" jbj 1 2g

4gðT0 TÞ e0 Cb2

#1=2

Therefore, differential of pyroelectric coefficient with respect to temperature is expressed as follows:   4gðT0 TÞ 1=2 p ¼ ð2Ps e0 CjδjÞ 1 1 þ Cb2 According to Eqs. (40) and (47) by timing p/er and Ps/C, respectively, we can have:   1=2 p2 4P 2 g T ¼ P02 ð2e0 CT0 Þ 1 1 þ 0 1 er jbj T0

ð46Þ

ð47Þ

ð48Þ

where Ps2

∂Ps2 2T0 ¼ ¼ T0 ∂T TrT e0 Cjbj

ð49Þ

0

In this case, if there is TET0, Eq. (48) can be rewritten as:

per 1=2 EP0 ð2CT0 Þ

1=2

ð50Þ

This equation well explains the observation discussed above. It is also used as a guide to evaluate the pyroelectric effect of the pyroelectric materials.

2.23.2.5

Pyroelectric Effect Versus Phase Transition

Usually, ferroelectric pyroelectric materials have two distinct phase transition, which are accompanied by a peak in the pyroelectric coefficient curves as a function of temperature [16,17]. The most pronounced phase change in ferroelectric, pyroelectric materials is ferroelectric-paraelectric transition, in which the magnitude of spontaneous polarization suddenly decreases and disappears thereafter. In first order ferroelectrics, an external electric field can be applied to stabilize the ferroelectric phase, at temperatures just above the Curie point TC. In this case, the temperature making the most of the change in polarization, i.e., the peak of pyroelectric coefficient, is increased as the level of the external electric field is increased, which can be understood with Eq. (41) [18]. Consistent with electric displacement d, electric field e, and temperature t, we have: ð2g=jbjÞ1=2 D

ð51Þ

8ð2g3 =jbj5 Þ1=2 E

ð52Þ

d¼ e¼

t ¼ 4a0 b 2 ðT

T0 Þ

ð53Þ

4d3 þ 2td

ð54Þ

As a consequence, Eq. (41) can be rewritten as: e ¼ 2d5

With the definition of pyroelectric coefficient as discussed earlier, it is derived as follows: ∂d  e  1 d ¼ 2 ¼ 4d 4d3 5d4 t ∂t 2d2 6d

ð55Þ

In ferroelectrics with second order phase transition, an external electric field cannot be used to induce phase transition [16,17]. The external electric field will merely flatten the plot of dielectric constant as a function of temperature. Therefore, temperature of the pyroelectric coefficient peak remains independent of an applied external electric field. However, the level of pyroelectric coefficient is decreased as the strength of the electric field is increased. By neglecting the higher order terms, Eq. (41) can be further reduced to: E ¼ a0 ðT

T0 ÞD þ bD3

where T0 ¼ TC. Because the electric field is constant, Eq. (56) can be rewritten as:   0 ¼ a0 ðT T0 Þ þ 3bD2 dD þ a0 DdT

ð56Þ ð57Þ

Pyroelectric Materials

As a result, pyroelectric coefficient can be rewritten as:   ∂D ¼ p¼ ∂T E;X a0 ðT

a0 D0 T0 Þ þ 3bD2

727

ð58Þ

where D0 is the electric filed E induced electric displacement, at the temperature T, which is given by Eq. (56). At three special conditions, there are following equations: T0 ފ 1 ; TcT0

D0 EE½a0 ðT  D0 E a0 b 1 ðT0

1=2

 ΤÞ

;

T { T0

D0 Eða0 =bÞ1=3 ; T ¼ T0

ð59Þ ð60Þ ð61Þ

According to Eq. (58), if p0 is the pyroelectric coefficient at T ¼ T0, following equations can be derived: T0 Þ 2 ; TcT0

p ¼ Ea0 1 ðT p¼

1 1=2 a b 2 0

p0 ¼

1=2

1 a =b 3 0

ðT0 2=3

E

TÞ

1=2

1=3

; T ¼ T0

; T { T0

ð62Þ ð63Þ

ð64Þ

In this case, below Curie temperature, the pyroelectric coefficient is increased with increasing temperature, in the way of (T0–T) 1/2, whereas above Curie temperature, the pyroelectric coefficient is decreased in the manner of (T–T0) 2. At Curie temperature, the peak pyroelectric coefficient is observed. This peak is suppressed when the materials are subjected to external electric fields, while the temperature of the peak remains unchanged. Ferroelectric–ferroelectric phase transition is the latter, during which a change in both the magnitude and direction of spontaneous polarization are most likely. However, in some ferroelectric materials, such a phase transition has no effect on the direction of the polarization. Accordingly, the parameters related to the phase transition will not be changed significantly, while the pyroelectric coefficient curve retains its peak. In this regard, pyroelectric coefficient can be used as an indicator to identify the ferroelectric–ferroelectric phase transition in ferroelectric materials. During a cooling or heating process, approaching the new equilibrium state for the polarization requires time, i.e., the pyroelectric response to the change in temperature is somehow delayed, which is related to several factors, such as thermal conductivity and heat capacity of the pyroelectric materials and dimension and shape of the devices. There is a term known as thermal relaxation time that is used to represent such a time delay, which is given by: tT ¼

L2 c0 sT

ð65Þ

where c0 refers to the heat capacitance per unit volume, with a unit of J K 1 m 3, L refers to the length in the direction of heat conduction with the unit of meter (m) and sT refers to thermal conductivity with a unit of J K 1 m 1 s 1.

2.23.2.6

Pyroelectric Figure of Merit

The response characteristics of the pyroelectric materials are usually evaluated by using pyroelectric FOM, which is given by: t0T ¼

M GT

ð66Þ

where M refers to heat capacitance (J K 1) of the material and GT refers to thermal conductivity (J K 1 s 1) between the material and sample holder. As a result, one way to minimize the response time is to reduce the value of GT, which can be realized by using sample holder that is thermally insulating or separating the sample and the holder with an insulating layer. The item t0 T in Eq. (66) is the time that is required by the pyroelectric element (PE) to achieve an equilibrium state between the ambient environments. This is not the same as the relaxation time tT in Eq. (65), the time required by the PE to attain an equilibrium inside the materials. In practical devices, the pyroelectric unit is generally linked to an impedance amplifier with high input. By correlating the pyroelectric unit, with a capacitance of Cx, the amplifier, with an input resistance of Rg and capacitance of Cg, another parameter, which is known as electrical time constant, will be derived: t0E ¼ Rg ðCx þ Cg Þ

ð67Þ

For different pyroelectric detectors, both thermal and electrical time constants should be properly selected, according to factors, for instance, response time or frequency dependence. Generally, the thermal time constant lies in the range of 0.01–10 s, whereas the electrical time constant a relatively broad range of 10 12–102 s [19].

728

Pyroelectric Materials

As the incident power is represented as W, the variation in temperature DT of the pyroelectric unit can be expressed as: ZW ¼ M

dT þ GT DT dt

ð68Þ

where Z is the percentage of absorption of W. Assuming the incident radiation can be described as: W ¼ W0 expðjotÞ

ð69Þ

DT ¼ ZW0 ðGT þ joMÞ 1 expðjotÞ

ð70Þ

the following expression can be arrived:

Pyroelectric charge generated in the unit is given by: q ¼ pADT

ð71Þ

where p refers to the pyroelectric coefficient while A refers to the unit’s actual area. In addition, the current response as in the induced current by the incident power can be obtained as: Ri ¼

i W0

ð72Þ

in which i is pyroelectric current, i.e., i¼ dq/dt. Therefore, the following equation can be arrived: Ri ¼

ZpAo GT 1 þ o2 t0T 2


1=2

ð73Þ

If the working frequencies are far beyond the value of 1/t0 T, the above equation can be simplified as: Ri ¼

Zp c0 d

ð74Þ

where c0 refers to volumetric heat capacitance while d refers to the pyroelectric unit’s thickness. Inside Eq. (74), a parameter, known as current figure of merit, can be derived: p c0

Fi ¼

ð75Þ

Accordingly, the voltage response is to be obtained from the pyroelectric current and admittance. If the conduction of the pyroelectric unit can be neglected, the admittance may be expressed as: Y ¼ Rg 1 þ joC ¼ Rg 1 þ joðCx þ Cg Þ

ð76Þ

As a result, the voltage response can be obtained as: Rv ¼ As o¼ (t0 Et0 T)

1/2

Rg ZpAo i ¼ 0 2 YW0 GT ð1 þ o tT 2 Þ1=2 ð1 þ o2 t0E 2 Þ1=2

ð77Þ

, Rv reaches its maximum value, i.e., Rv ðmaxÞ ¼ ZpA

Rg GT ðt0T þ t0E Þ

ð78Þ

In this case, a high insulation between the pyroelectric unit and the sample holder implies a low GT, i.e., thus leading to a high voltage response. As oc(t0 E) 1 and oc(t0 T) 1, the following expression can be derived: Rv ¼

Zp c0 dðC0x þ C0g Þo

ð79Þ

Zp c0 dC0x o

ð80Þ

When CxcCg, we will have: Rv ¼

In Eq. (80), another material parameter, which is named as voltage figure of merit, can be obtained: Fv ¼

p c0 e

ð81Þ

In a simple form, either p/e or p/er, is the characteristic pyroelectric performance indicator of a pyroelectric device, which is known as pyroelectric figure of merit.

Pyroelectric Materials

2.23.3 2.23.3.1

729

Pyroelectric Materials Triglycine Sulphate

Triglycine sulphate (TGS), (NH2CH2COOH)3H2SO4, is the pyroelectric material that has the highest figure of merit, among all ferroelectric pyroelectric materials. Single crystal TGS are usually developed through crystal growth from aqueous solutions [20]. TGS has a Curie temperature of TC ¼ 491C. Beyond this point, TGS is centrosymmetric, with class 2/m, while below the Currie point it is in polar state, with point group 2, in which the polar axis is along the monoclinic b axis [21]. In the crystal structure of TGS, the glycine (HN2CH2COOH) groups are polar. The polarization in TGS takes place as the Glycine I group is rotated about the crystallographic a axis, after which it is changed to the mirror image. Attempts have been made to modify TGS in order to develop new pyroelectric materials with higher properties. In this case, the basic crystal structure of TGS should be retained. Currently, various new compounds have been synthesized, such as deuteratedtriglycine sulfate (DTGS), substitution of sulphuric with fluoroberyllic acid to obtain triglycine fluoroberyllate (TGFB) without and with deuterated triglycine fluoroberyllate (DTGFB) [22] and substitution of glycine either with L-alanine or D-alanine (ATGS). In ATGS, an “internal” electric bias was developed in the materials, so that the spontaneous polarization was stabilized, which was attributed to the fact that the additional methyl group in alanine molecules hinders the rotation of molecules inside the lattice [23]. Therefore, the dipoles inside the materials are fixed and are unaltered near the Currie point, so that ATGS crystals can be thermally cycled through Curie temperature, without deterioration in spontaneous polarization. Derivatives through the doping of alanine and phosphoric or arsenic acids have also been reported, such as TGSP or ATGSAS [24,25]. These pyroelectric materials have different optimal operation temperatures [26]. For example, deuteration of TGS could increase the Curie temperature. As a result, the range of working temperature can be largely broadened. TGFB has a Curie point of 611C. After doping, a 30% increase in pyroelectric performance (Fv) was achieved. Although the TGS-based pyroelectric materials have higher pyroelectric properties (Fv) as compared with most oxide-based ferroelectric or polymeric ferroelectric materials, they suffer from high dielectric losses, which results in relatively low values of FD. Due to their special crystal structure, TGS-based pyrorelectric materials exhibit fairly high anisotropic dielectric constant tensor coefficients. This offers an good opportunity to tune the ratio of p'/e, which makes it possible to maximize their pyroelectric performances [27,28]. For instance, if the crystals are cut in such a way that the normal to the electrode faces of the detector is not parallel to the polar axis, a high p'/e ratio could be obtained, which is attributed to the fact that the effective pyroelectric coefficient is dependent of the angle of rotation with a function of cosine and the dielectric permittivity has a cosine squared function. TGSbased pyroelectric materials have encountered crucial issues of water solubility, hygroscopic nature and fragility, so that they are suitable for many applications, even though they have highest pyroelectric voltage responsive performances.

2.23.3.2

Polyvinylidene Fluoride

Polymeric ferroelectric polyvinylidene fluoride (PVDF) has been shown to exhibit high piezoelectric effects. In PVDF, the molecules possess a basic building block, ‒CH2‒CF2‒, with both the carbon–hydrogen bond and the electrically polar carbon–fluorine bond to adopt different configurations, which was closely related to the synthetic process used to prepare the polymers. For example, if the melts are cooled at sufficiently slow rates or the process is conducted in acetone solutions, a-phase is obtained, which is a non-piezoelectric phase. The molecular structure has a trans-gauche-trans-gauche configuration, thus leading to a nonpolar unit cell stacking. The a-phase can be transferred to b-phase, when it is subject to stretching or electric poling. In b-phase, all the molecular groups are arranged in the trans-configuration, resulting in the formation of polar unit cells. Due to its ferroelectric characteristics, b-phase PVDF also has pyroelectric effect. As compared with the TGS group, PVDF has relatively lower pyroelectric coefficients. For instance, its figure of merit is about 1/6 that of DTGS. At the same time, because of its high dielectric loss, PVDF also demonstrates a low level of FD. Although PVDF has a Curie temperature that is higher than its melting point of about 1801C, the polarization begins to decline as the temperature is beyond 801C, thus limiting the working temperatures of this group of pyroelectric materials. As a polymer, PVDF ferroelectrics can be made into large are with high flexibility. Meanwhile, it is very cost-effective as compared with other type of pyroelectric materials. More importantly, lapping, and polishing are not necessary for PVDF-based materials. There have been various derivatives for PVDF, such as copolymers of vinylidene fluoride with trifluoroethylene (VDF-TRFE), which has much improved ferroelectric and thus pyroelectric properties [30–32]. It is worth mentioning that PVDF-TRFE-based ferroelectric can be made into the b-phase directly from melts or solutions, where the stretching process is not required [29]. As a result, it is more convenient to fabricate thin film devices by using the conventional deposition techniques, such as spin-coating and solution coating.

2.23.3.3

Lithium Tantalate

Lithium tantalate, with a chemical formula of LiTaO3, has been acknowledged as an important pyroelectric material, belonging to the oxide family. In its crystal structure, the layers of oxygen ions are closely packed in a near hexagonal configuration, where the Li þ and Ta5 þ ions occupy two out of three octahedral interstices in between the layers. At room temperature, it has a point group of R3m (polar ferroelectric) at room temperature and the point group is changed to R3m (nonpolar paraelectric) at the Curie temperature of 6651C. Due to its very high melting temperature and insolubility in water, it is highly stable and can be used over a

730

Pyroelectric Materials

Fig. 1 Photograph of a LiTaO3 single crystal. Reproduced with permission from Kumaragurubaran S, Takekawa S, Nakamura M, Kitamura K. Growth of 4-in diameter near-stoichiometric lithium tantalate single crystals. J Cryst Growth 2005;285:88–95. Copyright 2005, Elsevier.

broad range of temperature. In most cases, LiTaO3 appears as single crystals, which can be grown through conventional single crystal growth methods, such as Czochralski method. Fig. 1 shows a photograph of a LiTaO3 single crystal [33]. Both dielectric constant and pyroelectric coefficient of LiTaO3 crystal are in the moderate range, and consequently a response figure of merit approximates 25% of TGS. Nevertheless, it has a pretty low dielectric loss tangent of 10 4, resulting a fairly high level of FD, which is about five times the value of DTGS. However, LiTaO3 has a critical problem, due to its high thermal diffusivity. As a consequence, the minimum resolvable temperature difference is diminished, especially at high frequencies, which makes it unsuitable for some applications like arrays. To tackle this issue, it is necessary to substantially isolate the detector elements through cutting slots with ion beam reticulation [34].

2.23.3.4

Strontium Barium Niobate

Strontium barium niobate has a chemical formula of SrxBa1 xNb2O6 (with 0.25rxr0.75), belonging to tungsten bronze ferroelectric family, with promising pyroelectric properties [19]. The crystal structure is constructed through corner-sharing of NbO6 octahedra, which leads to three groups of interstitial sites, with Ba or Sr ions occupying two of the three interstitial sites [35]. The structure remains tetragonal even above and below TC, The Curie temperature is decreased from 195 to 531C, as the composition is changed from x ¼ 0.72–0.25. The composition with equal fraction of Sr and Ba has the optimized pyroelectric performance, demonstrating highest pyroelectric coefficient, largest dielectric constant and lowest dielectric loss tangent. Although it exhibits a relatively low response figure of merit, its D figure of merit is above average, as compared with most of the promising pyroelectric materials. Furthermore, Curie temperature TC of SrxBa1 xNb2O6 can be slightly modified by doping with La, so that it will be close to room temperature. As a consequence, both the pyroelectric coefficient and dielectric constant can be largely increased at room temperature [36]. In addition, it can be operated close to the Curie temperature TC, without significant worsening in pyroelectric figures of merit (FOMs) [37]. Czochralski method is used to grow SrxBa1 xNb2O6 single crystals.

2.23.3.5

Perovskite Structured Pyroelectric Ferroelectrics

Perovskite structured ferroelectric materials have a chemical formula of ABO3, which is formed through the corner-sharing of BO6 octahedra, with the A cation to occupy the twelve-fold coordinated site. Both the A and B sites can accommodate various cations, thus leading a different crystal structures, including rhombohedral, tetragonal, and orthorhombic. Pyroelectric properties of modified lead zirconate (PbZrO3 or PZ) with rhombohedral structure and modified lead titanate (PbTiO3 or PT) with tetragonal structure have been extensively discussed in open literature. Most of the studies focused on ceramic materials, because almost all the properties that predicted in single crystals can be realized in ceramics. More importantly, ceramic process provides more flexibility in introducing dopants without changing the crystal structure. As a result, their pyroelectric properties can be easily optimized, as compared with the single crystal counterparts.

Pyroelectric Materials

731

PZ is an antiferroelectric compound and PT is a typical ferroelectric compound. PZ and PT can be used to formulate solid solution spanning the whole composition, with rich crystallographic phases. In the binary phase diagram of PZ-PT phase diagram, there is a phase in the region of PZ90:PT10, which is known as FR(LT) phase with a rhombohedral structure. It has a relatively low dielectric constant as compared with the samples with other compositions. Therefore, this phase is more suitable for pyroelectric applications, in terms of figure of merit. Furthermore, its pyroelectric properties can be optimized by using various dopants [38]. Additionally, the phase transition from low temperature FR(LT) to high temperature FR(HT) brings about a step-like change in spontaneous polarization, which leads to a high level of dPs/dT, because the dielectric constant is almost unchanged [39]. As a first order phase transition, its effective reversibility in pyroelectric coefficient is slightly deteriorated due to the presence of thermal hysteresis [40]. To address this issue, multi-composition materials are usually used. For instance, a tertiary composition, PbZrO3–PbTiO3–PbFel/2Nbl/2O3, has been reported to have sufficiently high pyroelectric coefficients for real applications [41,42]. Pyroelectric response of these materials may be further enhanced, via either finely tuning the compositions or doping with various dopants, with an attempt to decrease dielectric constant and dielectric loss tangent. By using proper dopants, DC resistivity, rDC, of the materials can be readily controlled, which makes it possible to design pyroelectric devices with more flexibilities. Tetragonal structured PT has a Curie temperature of 4901C, with a pretty high spontaneous polarization of 75 mC cm 2. By using the conventional crystal growth methods, such as flux method and top seeded solution method, single crystals of PT have been reported [43]. Because of the difficulty to achieve single crystals with large dimensions, most of the works have been focused on ceramics. Due to the large c/a ratio, PT experiences a very strong stress during cooling from high sintering temperature crossing the Curie point. As a consequence, the sintered ceramics would be broken at room temperature. Therefore, almost no pure PT ceramics are available. Instead, PT-based ceramics have been modified by using various dopants, so that the c/a ration can be decreased to have minimized stress [44]. Pyroelectric properties PT-based ferroelectrics have been reported [45]. PT ceramics have comparable pyroelectric performances with PZ based materials. However, due to their relatively lower dielectric constant, PT-based pyroelectric materials possess higher values of Fv, while their FD is slightly lower owing to their high dielectric loss tangent. Low dielectric loss can be suppressed by treating the materials with special sintering techniques, such as hot pressing, to have enhanced densification [46,47]. Besides, other perovskite ferroelectrics have also been studied for pyroelectric applications. One example Sr-modified Ba0.85Ca0.15Zr0.1Ti0.9O3 ceramics, with compositions of Ba0.85Ca0.15 xZr0.1Ti0.9O3–Srx (BCZT-Sr) (x¼ 0%, 5%, 10%, and 15%) [48]. It was found that a noteworthy increment in polarization, from 16 to 25 mC cm 2 in the composition with Sr content of 15%, was achieved, i.e., Ca was totally replaced with Sr. Meanwhile, an increase in dielectric constant, from 2743 to 4040, at room temperature at 1 MHz, was also observed, together with a decrease in Curie–Weiss temperature (TCW) from 357 to 308K. The sample with the optimized composition exhibited a pyroelectric coefficient of 25 mC cm 2 K 1 at 308K. Also, pyroelectric FOMs for voltage responsivity (Fv ¼ 0.017 m2 C 1), current responsivity (Fi ¼600  10 12 m V 1), detectivity (Fd ¼17.6  10 6 Pa1/2), 12 m3 J 1) were evaluated. Fig. 2 shows energy harvesting (Fe ¼485  10 12 J m 3 K 2) and new energy harvesting (F e ¼ 10.1  10 dielectric constant (at 1 MHz) and pyroelectric coefficient of Ba0.85Sr0.15Zr0.1Ti0.9O3 sample as a function of temperature.

5500

Dielectric constant ()

4500

20

4000 15

3500 3000

10

2500 5

2000 1500 1000 300

Pyroelectric coefficient (10–4 C/m2, K)

25

5000

0 310

320

330

340

350

360

370

Temperature (K) Fig. 2 Dielectric constant (at 1 MHz) and pyroelectric coefficient of Ba0.85Sr0.15Zr0.1Ti0.9O3 sample as a function of temperature. Reproduced with permission from Patel S, Chauhan A, Vaish R. Large pyroelectric figure of merits for Sr-modified Ba0.85Ca0.15Zr0.1Ti0.9O3 ceramics. Solid State Sci 2016;52:10–8, Copyright 2016, Elsevier.

732

Pyroelectric Materials

Lead-free ferroelectric materials have garnered much attention over the last decades, due to the concern of potential pollution of lead to the environment. As an example, (Ba0.84Ca0.15Sr0.01)(Ti0.90Zr0.09Sn0.01)O3(BCSTZS) lead-free ceramics, co-doped with 1 mol% Sr and 1 mol% Sn, synthesized via conventional ceramic processing method, exhibited a room temperature pyroelectric coefficient of 1116.7 mC K 1 m 2, with FOMs, Fd ¼18.1 mPa 1/2, Fv ¼ 0.013 m2 C 1 and Fi ¼ 479.3 pm V 1, respectively [49]. The pyroelectric FOMs were highly stable over frequency range, from 100 to 2000 Hz. However, they were varied with temperature, which should be further improved. The promising pyroelectric properties of the BCSTZS lead-free ceramics was attributed to the polymorphic phase transition that took place close to room temperature. It is believed that more and more advances on lead-free pyroelectric (ferroelectric) materials will be published in research journals.

2.23.3.6

Lead Germanate

Lead germanate has a chemical formula of Pb5Ge3O11, which possesses a hexagonal crystal structure, with a relatively low Curie temperature of 1781C [50–53]. Pb5Ge3O11 single crystals have been readily fabricated from the corresponding melts with the Czochralski method [54,55]. Normally, barium is used to substitute lead or silicon is employed to replace germanium, so as to modify its ferroelectric and thus improve the pyroelectric properties. Both dopants would result in a decrease in Curie temperature, so as to be shifted close to room temperature. According to literature data, pyroelectric properties of Pb5Ge3O11 (PGO) and Pb4.7Ba0.3Ge3O11 (PGO:Ba3) single crystals are comparable with those of LiTaO3. For example, after doping with barium, p0 and e are increased while tanδ is reduced, thus resulting in an increase in Fv and a decrease in FD.

2.23.4

Olsen Cycle of Pyroelectric Effect

When examining pyroelectric energy harvesting behaviors from heat resources, thermodynamic cycle should be used to identify the maximum efficiency [7]. Although various cycles have been proposed, the Olsen cycle is the most applicable to describe the pyroelectric energy conversion behavior of pyroelectric materials. It is similar to the Ericsson heat engine in the form of charge‒voltage diagram (C–V) [56–66]. Fig. 3 shows a typical q–V diagram of PE, in which the loop (1–2–3–4–1) denotes the Olsen cycle [67]. The PE could be a thin film with two metallic electrodes. The Olsen cycle is started as the PE is charged at a low temperature TL, so that the voltage is increased from VL to VH (i.e., process 1–2). Then, the PE is discharged as it is heated from TL to TH at a constant voltage VH, with TH to be higher than TCurie,↑ (i.e., process 2–3). After that, the PE continues discharging, while the voltage is decreased to VL, at a constant temperature TH (i.e., process 3–4). Last but not least, the PE is recharged as it is cooled from TH to TL, with TL to be lower than TCurie,↓, while the voltage VL is kept constant (i.e., process 4–1). Electrical work output WE with a unit of J L 1 is determined by the shaded area in Fig. 3, which is given by: WE ¼

I

ð82Þ

VPE dqPE

WE = ∫VPEdqPE

2

Tcool

Charge, qPE (C)

Heating 1 3 Thot

Cooling 4

1–2 isothermal charge 2–3 discharge + heating 3–4 isothermal discharge 4–1 charge + cooling VL

VH

Voltage, VPE (V) Fig. 3 Typical charge–voltage (C–V) performance of pyroelectric material at two dissimilar temperatures. The Olsen power cycle can be derived from the shaded area bounded by the loop 1–2–3–4–1. Reproduced with permission from Nguyen H, Navid A, Pilon L. Pyroelectric energy converter using co-polymer P(VDF-TrFE) and Olsen cycle for waste heat energy harvesting. Appl Therm Eng 2010;30:2127–37, Copyright 2010, Elsevier.

Pyroelectric Materials

2.23.5 2.23.5.1

733

Thermal Energy Harvesters Based on Pyroelectric With Olsen Cycle Thermal Subsystem

Prototype direct pyroelectric converters have been reported to evaluate conversion efficiencies of selected pyroelectric materials [68,69]. There are thermal and electrical subsystems in a converter. The function of the thermal subsystem is to generate a temperature oscillation dependent on time, which is used for increasing the PEs to the maximum temperature TH (high temperature) and cooling them to the minimum temperature TC (low temperature). The electrical subsystem is designed to provide electric fields, so as to ensure the PEs to have the Olsen cycles. At the same time, the electrical subsystem is responsible for the collection of the charges generated by the PEs. Fig. 4 shows schematic diagram of a two-stage pyroelectric converter prototype, in which the measurements and related coordinate system are included for the purpose of simulation [69]. Channels with a number of (N þ 1) are formed, as the working z

po

Heating band, Thot

Lcr

Computational domain Oscillating working fluid

L = 15.875

g Lcr Heat exchanger, Tcold x

Lc = 0.635

zp = −S [1+cos(2πft)]

Piston (A)

Heating band, TH

BC 5

LH = 0.635

BC 2, BC 8

Computational domain

BC 3, BC 6

Lcr BC 10

A

PE1

zh

z1 = L − Lcr

BC 1, BC 9 PE2

LPE B zA = zB + LPE

Oscillating working fluid

BC 10 Lcr

zB = z2 + LPE /2 z2 = LC + Lcr

z (0.0)

BC 4, BC 7 x wf = 0.018

ww = 0.0125

(B)

A12O3

Working fluid

Heat exchanger

Pyroelectric elements (PE)

Fig. 4 (A) Schematic diagram of a two-stage pyroelectric converter with a length of L. (B) Schematic of the computational domain, including the coordinate system and the boundary conditions, measured in centimeter. Reproduced with permission from Navid A, Vanderpool D, Bah A, Pilon L. Towards optimization of a pyroelectric energy converter for harvesting waste heat. Int J Heat Mass Transf 2010;53:4060–70, Copyright 2010, Elsevier.

734

Pyroelectric Materials

fluid is oscillated between the N stationary equidistant walls in the vertical direction. All the walls are equivalent, which consist of four plates that are vertical to the flowing direction of the fluid. Two Al2O3 ceramic plates are placed at the bottom and two lead zirconate stannate titanate (PZST) plates are positioned at the center, which is the active region of the device. The working fluid was kept at a constant temperature TC due to the cooling effect by using a heat exchanger that is put at bottom of the test section. Additionally, a thin heating element is inserted in the channel’s center that is on highest of the test section, in order to warm the working fluid if necessary, which is maintained at a constant heat flux of qin00 . Both the flow channel and the wall have a half-width of wf and ww, respectively. The channel has a length of L and the lower and upper Al2O3 plates possess a length of Lcr. The bottom of the heating element is at the position of zh, while the lengths of the heat exchanger and heating element are defined as LC and LH, respectively. The heating element in the middle of the channel is treated as a line source. In addition, the centers of the pyroelectric elements PE1 and PE2 are at A and B, with heights of zA and zB, respectively. The lengths between the test section’s lowest point to highest point of PE1 or lowest point of PE2 are z1 ¼L Lcr and z2 ¼ LC þ Lcr, respectively. The total depth of the pyroelectric converter, d, is measured to be 3.8 cm. By employing a pump of piston-in-acylinder, with variable piston amplitude S as well as frequency f, fluid flow is oscillated across the minichannel. At instant t, the piston’s top surface positions at zp ¼ S[1 þ cos(2pft)]. At the initial state of t¼ 0, the piston is at the bottom of zp(0) ¼ 2S, while the free surface of the liquid is demarcated at the position of z¼L.

2.23.5.2

Electrical Subsystem

As stated earlier, the electric subsystem is a key component that is necessary to maintain the thermodynamic cycles, as shown in Fig. 3 [66]. The pyroelectric unit is linked to the electrical subsystem through electrodes. The voltage drop across the loaded capacitor and resistor is used to measure the charges from both the PEs and the applied voltage. The required voltages for the Olsen cycle are maintained by applying voltages of VL and VH. At the same time, a switch is used to trigger the cycle, so that the temperature and voltage of the PEs can be matched. Upon approaching either the maximum or minimum temperature of the PE, the switch will be triggered, which is realized by using a control system. As shown in Fig. 3, the PE is isothermally charged (1–2), at the constant temperature of TC, through the power supply that is switched from VL to VH. Accordingly, the PE heats (2–3) at the constant voltage of VH, until the temperature approaches TH. After that, isothermal discharge is carried out in the load (3–4), which is facilitated by the voltage change from VH to VL. Lastly, the PE is cooled at the constant voltage of VL, so that the cycle is closed.

2.23.5.3

Numerical Simulation

In order to provide mathematic solutions to the issues, there must be necessary assumptions [68,69]. Firstly, the pyroelectric converter is assumed to be thermally insulated and thus there is no heat losses both to the surroundings and in between the channels. The properties of all the materials are assumedly constant and isotropic. For instance, silicone oil used as the working fluid is considered to be a Newtonian and is not compressible. Also, as compared with the depth of the plate, the interwall spacing is considerable smaller, so that the flow is of two-dimensional characteristic [70]. Because the fluid flow is periodically oscillating, the Reynolds number is below a critical threshold during the simulations and the fluid flow remains laminar. A hysteresis curve is formed, with the correlation between the charge q and the open-circuit voltage V at the constant temperature, as shown in Fig. 3. Moreover, the edge effects are negligible because all the channels are equivalent. As a result, only one of them is used to do the simulation. In addition, it is approximately considered that the electrodes used for the collection of the generated charges have influence on neither the heat transfer nor fluid flow. The PEs at the either stage of the converter are presumed to have the same physical properties.

2.23.5.4

Governing Equations

To calculate each individual components of the velocity vector and the pressure field of the working fluid, it is necessary to solve the two-dimensional mass and momentum conservation equations, typical of an incompressible and Newtonian fluid with constant properties in Cartesian coordinates, given by [68,69]: ∂uf ∂vf þ ¼0 ∂x ∂z ∂uf ∂uf ∂uf þ uf þ vf ¼ ∂t ∂x ∂z ∂vf ∂vf ∂vf þ uf þ vf ¼ ∂t ∂x ∂z

ð83Þ

 2  1 ∂pf ∂ uf ∂2 uf þ vf þ ∂x2 ∂z2 rf ∂x  2  1 ∂pf ∂ vf ∂2 vf þ vf þ ∂x2 ∂z2 rf ∂z

ð84Þ

g

ð85Þ

where rf and vf stand for the density and kinematic viscosity of the working fluid, whereas uf and vf refer to the x- and z- components of the fluid velocity vector. pf represents pressure of the working fluid and g refers to the gravitational acceleration of g¼9.81 m s 2, in the z-axis direction. The two-dimensional energy equation use to solve the temperature distribution inside the

Pyroelectric Materials

735

working fluid is rewritten as follows [80,81], rf cp;f



∂Tf ∂Tf ∂Tf þ uf þ vf ∂t ∂x ∂z



¼ kf



∂2 Tf ∂2 Tf þ 2 ∂x2 ∂z



ð86Þ

where cp,f relates to the heat capacity of the fluid, Tf relates to the local fluid temperature and an assumed constant, kf, is the fluid thermal conductivity. Similarly, for the Al2O3 plates and the pyroelectric PZST plates used to construct walls of the channel, the two-dimensional heat diffusion equation is as follows, ∂Tw ∂2 T w ∂2 Tw ¼ kw;x þ kw;z 2 ð87Þ ∂t ∂x2 ∂z in which the subscript w refers to the walls representing both the Al2O3 and the pyroelectric PZST plates. The thermal conductivities could be anisotropic, as represented by kw,x and kw,z, which are assumedly independent of temperature. rw cp;w

2.23.5.5

Initial and Boundary Conditions

Initially when t ¼ 0, the working fluid is at its origin. In this case, for 0rxrwf and 0rzrL, we have: uf ðx; z; 0Þ ¼ vf ðx; z; 0Þ ¼ 0

ð88Þ

In the beginning, the temperature across the walls and the working fluid in the x-direction is not varied, so that there is 0rxrwf þ ww. Also, the temperature remains linear between that of the heating element at the position of z ¼ zh, the temperature TH of the heat exchanger at the position of z ¼ LC and temperature TC. Furthermore, the temperature profiles of the working fluid are kept consistent along and above the heat exchanger and along the heating element. Therefore, for 0rxrwf þ ww, we have: 8 0rzrLC > < TC ðTH þ TC Þðz LC Þ L T Tf ðx; z; 0Þ ¼ Tw ¼ ð89Þ C rzrzh C þ zh LC > :T zh rzrL H At time t, the no-slip condition becomes principal, in both the wall (x¼wf) and heating element (x¼0). As a result, the velocity components uf(x, z, t) and vf(x, z, t) are obtained as follows: uf ðwf ; z; tÞ ¼ vf ðwf ; z; tÞ ¼ 0;

0rzrL

ð90Þ

uf ð0; z; tÞ ¼ vf ð0; z; tÞ ¼ 0; zh rzrzh þ LH

ð91Þ

At the same time, a symmetry boundary condition is applied at center line of the channel (x¼ 0), given by: ∂vf ð0; z; tÞ ¼ 0; 0rzrzh ; zh þ LH rzrL ð92Þ ∂x Additionally, the velocity is equal for both the working fluid at the piston and that of the piston, which is expressed as vp ¼ 2pfSsin(2pft). According to the mass conservation and incompressible behavior of the fluid, its velocity at the lowermost of the microchannel remains consistent, which follows a sinusoidal function of time, with an amplitude of S0 ¼SAP/(N þ 1)Ac, given by: uf ð0; z; tÞ ¼

uf ðx; 0; tÞ ¼ 0; vf ðx; 0; tÞ ¼ 2pfS0 sinð2pftÞ; 0rxrwf

ð93Þ

where Ap and Ac are the cross-sectional areas of the piston and an individual channel, respectively. The incompressible working fluid above the channel synchronously oscillates with the piston at identical amplitude and frequency. In this case, the free surface oscillates between z¼ L and z ¼ L þ 2S, whereby the piston is moved from the position of zp ¼ 2S to that of zp ¼ 0. Additionally, because the huge cross-sectional area of the working fluid at the top of the reservoir above z ¼L and similarity to that of piston Ap, the pressure at z ¼ L is thus dominated by hydrostatic pressure, which is rewritten as follows: pf ðx; L; tÞ

p0 ¼ rf gS½1

ð94Þ

cosð2pftފ; 0rxrwf

where p0 refers to the ambient pressure at the free surface. The symmetry boundary condition is applied to the temperatures at the center lines of the walls and the flow channels, so that we can arrive at the following equation: ∂Tf ∂Tf ð0; z; tÞ ¼ ðw þ ww ; z; tÞ ¼ 0; 0rzrL ð95Þ ∂x ∂x f Moreover, the working fluid’s temperature at z ¼0 remains unchanged, which is the same as that of the heat exchanger across the channel, the following relationship can be arrived: TC ¼ Tf ðx; 0; tÞ; 0rxrwf

ð96Þ 00

Similarly, the heat flux along the surface of the heating element is also constant and is equal to qin : q00 ð0; z; tÞ ¼ q00in ; zh rzrzh þ LH

ð97Þ

736

Pyroelectric Materials

where q00 can be tuned to reach the desired temperature of TH at the heating element. Furthermore, the heat fluxes that are normal to the flow direction at the interfaces between the working fluid and the PE or Al2O3 plates are given by: kf

∂Tf ðwf ; z; tÞ ¼ ∂x

kw

∂Tw ðwf ; z ; tÞ ¼ 0; ∂x

0rzrL

ð98Þ

As a result, the axial heat conducted between the PEs and the Al2O3 plates is taken into account, which is achieved by assuming that the interfacial axial heat fluxes at positions of z1 and z2 are the same: kcr

2.23.5.6

∂Tcr ðx; z1 =z2 ; tÞ ¼ ∂x

kPE

∂TPE ðx; z1 =z2 ; tÞ ¼ 0; ∂x

wf rxrwf þ ww

ð99Þ

Material Properties

To simplify the simulation, the Al2O3, PZST and the working fluid are considered to have invariable thermophysical properties over the whole temperature range. It has been demonstrated that the variation in properties of all the materials is smaller than 10%, as the temperature is changed between TC ¼1451C and TH ¼1851C, while the viscosity of the working fluid (n)f, has a variation of about 37%. Therefore, the assumption is still reasonable. With a first order approximation, (n)f is 8.29 mm2 s 1, at the arithmetic mean temperature of 1651C. Other properties, for instance, density, specific heat and thermal conductivity of the Al2O3 plates and silicone oil, are measured at this arithmetic mean temperature as well [68,69]. Meanwhile, the values of these physical properties of the PZST plates are those obtained at room temperature.

2.23.5.7

Solution Method

The local velocity, pressure and temperature throughout the computational domain are derived, by resolving the mass, momentum and energy equations of Eqs. (82–85), with the associated initial boundary conditions described in Eqs. (83–95). The mass and momentum conservation equations are resolved concurrently [68,69]. After that, the energy equations are solved for the fluid, the Al2O3 and the PEs. The selection of the internal time step is arbitrary, so that the numerical stability can be ensured, during the solving of the mass, momentum, and energy equations. Meanwhile, the results are obtained at a time interval of Dt, so that the relationship of 2pfDt ¼ p/2 is met at all frequencies. Numerical convergence always ensues when resolving the governing equations by using a coarse grid, 1.3 times larger than the fine grid. The calculated outcomes numerically converge, if the differences throughout the computational domain between two consecutive grid refinements for all frequencies are less than 0.5%, 0.1%, and 1.8%, for the local velocity, pressure, and temperature, respectively. For the purpose of validation, the numerical derivations of velocity and pressure for vertically oscillating adiabatic flow are equated with the analytical solutions for fully developed laminar flow, which can be written as [70]: "  2 # x uf ðx; z; tÞ ¼ 0; vf ðx; z; tÞ ¼ 3pfS0 1 sinð2pftÞ ð100Þ wf The difference between the results of numerical simulation and those of the analytical solutions is less than 1.05%, for 0rxrwf and 0rzrL, at all points, confirming that the validity of numerical simulation as well as application of the governing equations and boundary conditions.

2.23.5.8

Evaluation of Performances of the Pyroelectric Converters

The total electrical power, generated by the two PEs in all channels, is expressed as: I  I _ E ¼ Nf V1 A1 dq1 þ V2 A2 dq2 W

ð101Þ

where N is the total number of internal walls of the channels and f is the frequency in Hertz. For each pyroelectric plate, the surface areas of A1 and A2 are 9.652 cm2. In order to obtain the charge–voltage curve at the temperatures of TC and TH, which are calculated at the points of A and B, as shown in Fig. 4, linear interpolations of the charge–voltage curves for PZST at different temperatures are adopted [83]. It has been assumed that the charge–voltage curves of the PE1 and PE2 are identical. The operation voltages of the devices are VL ¼100 V and VH ¼700 V [77]. It is also assumed that there is only frictional loss inside the channels as the working fluid is pumped through them. Therefore, the total pumping power time averaged over the pumping period, t¼ 1/f, is calculated from the following equation:  Z t Z _ EE N þ 1 rf vf dA dt ð102Þ W t 0 Ac where vf refers to the local instantaneous fluid velocity and rf refers to pressure and they can be obtained by resolving the mass and momentum conservation equations, respectively. The integrations of Eqs. (99) and (100) can be carried out numerically with aid of the trapezoidal rule.

Pyroelectric Materials

737

The heating element has a total heat transfer rate that can be expressed as: _ in ¼ 2ðN þ 1Þqin 00 Ahb Q

ð103Þ 2

where Ahb is the surface area of the heating element in one channel, with a value of 2.413 cm . Accordingly, the system exhibits an average thermodynamic energy efficiency, which can be expressed as [79]: Z¼

_ cycle _p _E W W W ¼ _ _ Qin Qin

ð104Þ

For simplicity, all the other losses, including heat loss to the ambient and current leakage along the electrical circuit, are ignored. The power density of the system, defined as the sum of electrical power per unit volume of the PEs, is given by: PD ¼

_E W 2N8PE

ð105Þ

Because every internal wall supports two PEs, with a volume of 8PE , the total volume of the PEs in the system is 2N8PE , with N refers to the number of the internal walls [65].

2.23.6 2.23.6.1

Olsen Harvesters Based on Pyroelectric Polymers Device Assembly and Characterization

A pyroelectric converter based on pyroelectric polymers (P(VDF-FrFE)) has been reported, which can be used to generate electricity from thermal energy by using the Olsen cycle [67]. Fig. 5 shows a cross-sectional view and a photograph of the device, with a thermal and an electrical subsystems as mentioned above. The thermal subsystem has five components, namely, (1) piston-cylinder assembly, (2) PEs, (3) heating element, (4) cold-heat exchanger, and (5) working fluid [67]. As the piston is oscillated, the working fluid is vertically pumped between the heater (top of the device) and the cold exchanger (bottom of the device). The motion of the working fluid is confined in Teflon cylinders, with inner and outer diameters measured to be 38.10 and 57.15 mm, which acted as the thermal insulator and a support at the same time. They have two holes with a diameter of 3.20 mm that are located at the position away from the center by 25 mm. Two M3 threaded rods with a length of 25 cm are slotted into the holes, so that the device is assembled. Silicone gaskets are put in place in order to prevent the leakage of any working fluid. z 228.0 218.5 Reservoir

Silicone gasket 165.3

Electrical heater

T5 135.6

Mica stack

Cooling coil

10 cm Thermocouples

Heater Cold heat exchanger

T4 T3 T2 89.8 T1

Heater wires

63.5 Connecting rod Cylinder sleeve

Flywheel 0

(A)

(B)

Fig. 5 (A) Schematic diagram illustrating cross-sectional of the P(VDF-FrFE) pyroelectric energy harvester, with individual components labeled. (B) A photograph of the harvester, with millimeter as unit of measurement. Reproduced with permission from Nguyen H, Navid A, Pilon L. Pyroelectric energy converter using co-polymer P(VDF-TrFE) and Olsen cycle for waste heat energy harvesting. Appl Therm Eng 2010;30:2127–37, Copyright 2010, Elsevier.

738

Pyroelectric Materials

Xp Kapton tape

10 40.6

PE

10

(A)

(B)

(C)

Mica plate

(D)

Teflon strip

Working fluid

Mica stack

Teflon cylinder (G)

Thermocouples

(F)

(E)

Fig. 6 Schematic diagram for the assembling of pyroelectric stack: (A) a mica plate with a 10 mm  10 mm window, (B) a mica plate with pyroelectric element (PE), (C) PE in between two mica plates, (D) pyroelectric assembly with Teflon strips, (E) mating of two mica plates, (F) entire PE stack, and (G) a photograph of an actual stack. All measurements are in millimeter. Reproduced with permission from Nguyen H, Navid A, Pilon L. Pyroelectric energy converter using co-polymer P(VDF-TrFE) and Olsen cycle for waste heat energy harvesting. Appl Therm Eng 2010;30:2127–37, Copyright 2010, Elsevier.

To assemble the piston-cylinders, commercially available stainless steel piston and cylinder sleeve are used. The sleeve is 3.5 cm in inner diameter and 6.4 cm in height. The fluid inside the cylinder sleeve remains sealed with a piston ring at the piston head. Stroke length of the piston varies between 2.5 and 4.7 cm, which was controlled by the adjusting the mounting radius of the connection between rod and the flywheel. The individual PE is slotted between two mica plates, to prevent both the horizontal and vertical motions. Fig. 6 depicts a schematic diagram of the pyroelectric assembly and the stack integration process. Firstly, a 310 mm thick mica plate is placed on flat smooth surface, with a square hole of 10 mm  10 mm at the center. The mica plates have a length of 40.6 mm and widths in the range of 20.8–38.1 mm, which can be readily stacked and fixed into the Teflon cylinders. The PE is then attached onto the plate, with the edges to be connected to the mica plate with four pieces of Kapton tape that has a thickness of 89 mm. Then, electrical wires are stuck to the two electrodes of the PE, with 89 mm thick copper tape, which are led out of the pyroelectric convertor to the electrical circuit. After that, a mica plate with the same dimension is mount onto the edges of the PE. The two plates are adhered at the four corners with a thin layer of thermal epoxy. Microchannel with a width of 330 mm is created by sealing the vertical edges of the mica plate at the position it is connected with another pyroelectric assembly, with two pieces of double sided Teflon tape that are 3.18 mm in width and 330 mm in thickness. The step is repeated to form a stack with 38 PEs. A flexible AC Kapton electrical heater with a power of 50 W is used as the heat source, which is stuck to a thin copper sheet, with a dimension of 12.7 cm  2.54 cm  0.20 mm. The flexible heat source is attached to the inner surface of the Teflon cylinder that is positioned above the PE stack. It has a maximum power of 50 W, which is adjustable through varying the applied voltage. The cold heat exchanger is constructed by using a copper tube, which has an inner and an outer diameters of 1.59 mm and 3.18 mm, as well as a length of 34.3 cm. After bending, the helical shaped tube can be installed on the outer cylinder of the cold heat exchanger. Water is used as the coolant, which is pumped into the coil with a DC electrical pump. Desired operating conditions are achieved by controlling the flow rate. A silicone oil is used as the working fluid, which has a dielectric property that is compatible with the electronic components of the pyroelectric system. It has a sufficiently high viscosity, so laminar fluid flow can be maintained in the microchannels. At the same time, silicone oil with a 15.7 MV m 1 dielectric strength at room temperature, and 2001C boiling point, under atmospheric pressure; satisfying all the requirements of the pyroelectric convertor. The convertor has an electrical subsystem that is used to apply voltages to the PE and maintain VL and VH to sustain the Olsen cycle. The subsystem is also employed to obtain the power that is produced by the PE, by measuring the voltage VPE and the charge qPE. Input voltages of VL and VH are provided to the high voltage power supply, by using a triple DC output power supply, which  converts to output voltage according the relation of VL/H ¼ 250V  . connects to two rocker switches. The input voltage VL=H L=H During the operation process, only one switch can be opened. A resistor with resistance of RL ¼ 7.80 MO deployed as a voltage divider, which is utilized to reduce the voltage across another resistor that has a resistance of R2 ¼21.8 kO and is series-connected. By doing this, the maximum voltage input can be controlled at 10 V, for the data acquisition system (DAQ). Meanwhile, a capacitor of C1 ¼1.0 mF is series-connected with the PE. Voltage of V1 across the capacitor and voltage of V2 across the resistor R2 are measured with the DAQ. As a result, both the magnitude of the electric displacement of the PE (DPE) and the electric field (EPE)

Pyroelectric Materials

739

can be calculated. The magnitude of the electric displacement DPE can be expressed as: DPE ¼

qPE C1 V1 ¼ A A

ð106Þ

where A is surface area of the PE. The electric field across the PE may be obtained from the Ohm’s law and the Kirchhoff’s law, which is expressed as: EPE ¼

VPE V ð1 þ RL =R2 Þ ¼ 2 b b

V1

ð107Þ

in which b refers to the pyroelectric film’s thickness. In this case, the D–E plots (equivalent to q–V plots) can be arrived for the PE through the Olsen cycle. Resistivity of the PE without poling may be measured as a function of time prior to commencing the Olsen cycle. As a constant voltage is applied across the PE at 851C, the leakage current (IL) is given by, IL ¼ C1

dV1 dt1

ð108Þ

where V1 refers to the registered voltage across the capacitor C1. Consequently, electrical resistivity of the pyroelectric film, with a unit of ohm meter, can be derived as: rPE ¼

VPE A EPE A ¼ IL b IL

ð109Þ

The frequency of the piston oscillation can be measured with a transmission sensor, which is comprised of an infrared emitting diode and a bipolar negative-positive-negative (NPN) silicon phototransistor. Type J thermocouples have been installed at several essential sites inside the working fluid, so that the local temperatures within the device can be monitored, as shown schematically in Fig. 4. In summary, T1 refers to the central temperature of the cold heat exchanger, whereas T2, T3, and T4 are the temperatures at bottom, middle and top positions of the PE. T5 refers to the central temperature of the heater cylinder. Additionally, Tc,i and Tc,o are the inlet and outlet temperatures of the cold heat exchanger.

2.23.6.2

Performances

The temperature (T1) of the fluid close to the cold heat exchanger to push/pull the piston is varied at 0.025 Hz frequency, with 21.41C as least and 49.71C as most, respectively. The temperature change DTi at the location “i” is given by: DTi ¼ Ti;max

Ti;min

ð110Þ

where Ti,max and Ti,min refer to the maximum and minimum values for one cycle. The arithmetic mean temperature T i at the location “i” can be thus obtained as: Ti ¼

Ti;max þ Ti;min 2

ð111Þ

For instance, T3 is measured at the middle of the PE, which is oscillated between 90.7 and 66.71C, corresponding to a temperature change of DT3 ¼ 241C. Similarly, the values of other temperatures are also obtained. In this case, the maximum temperature stands for the piston to be the lowest, while the minimum temperature implies that the piston is at the highest of the stroke. All the temperatures are oscillated at the same frequency as that of the piston. A piece of P(VDF-TrFE), with a dimension of 1 cm  1 cm  45.7 mm, is used as the PE, which is poled at an electric field of EPE ¼ 201 kV cm 1 for 180 min at the temperature of T3 ¼ 851C. After poling, the system is subject to the Olsen cycle. The device includes a 4.7 cm stroke length, with channels of 330 mm in width, at working frequency of 0.061 Hz. The low and high electric fields applied to trigger the Olsen cycle correspond to EL ¼ 202 kV cm 1 and EH ¼379 kV cm 1, respectively. At low electric field EL, the pyroelectric film is poled during the process of 4 1 in the Olsen cycle. At the same time, the heater provides a heat input at the level of 30.9 W. The low and high temperatures at the PE’s center, T3,min and T3,max, measure as 66.41C and 83.01C, respectively. This corresponds to a temperature change of DT3 ¼16.61C and an arithmetic mean temperature of T 3 ¼74.71C, corresponding to an energy density and a power density of 130 J L 1 and 8.12 W L 1. Also, in consideration of the pyroelectric material, the maximum energy density can be as high as 83 J L 1, at the temperature of T3 ¼ 74711C and the high electric field of EH ¼329 kV cm 1. This maximum energy density can be understood basing on two points. Firstly, as relatively low temperatures, because the temperature change is limited, the achievable energy density is thus small. However, as the temperature is too high, the leakage current of the pyroelectric material is increased, thus leading to a decline in the energy density. As a result, there must be a maximum energy density that is present at an appropriate temperature. The effect of frequency on the temperature change, DT3, at the PE’s center, with channels of 0.33 mm and 1 mm in width, is examined, by varying the frequency in the range of 0.025–0.16 Hz [67]. Generally, the temperature change is decreased as frequency is increased, because of the thermal inertia of the working fluid as well as the PE. With increasing frequency, the time required to exchange thermal energy between the heating element and the oscillating working fluid becomes gradually insufficient, so is that for exchange between the working fluid and the PEs. To tackle the issue, it is necessary to decrease the thermal time

740

Pyroelectric Materials

constant of the PEs, that is given by: tt ¼

rPE cp;PE 8 ho A

ð112Þ

where ho is the convective heat transfer coefficient of the oscillating fluid. The thermal time constant may be shortened through increment of the level of ho and the surface area A of the PE and reducing the PE’s thickness, which is given by b ¼ 8=A. Also, a larger temperature change can be obtained by narrowing down the width of the channels. In the literature, a relationship is available, which is able to link the Strouhal, Reynolds, and Nusselt numbers, if the oscillating fluid is laminar in a circular channel, and can be expressed as follows [85], ho l0 ¼ 0:92St 0:2 StL 0:26 Prf0:4 Re0:44 ð113Þ o kf where l0 and kf are the amplitudes of the displacement of the fluid and the thermal conductivity of the working fluid, respectively. The Strouhal numbers denoted as St and StL, can be expressed as follows [85]: Nuo ¼

St ¼

Dh S

ð114Þ

L ð115Þ S where Dh is the hydraulic diameter of the channel, S is the stroke length of the piston, and L is the length of the channel. Additionally, the Reynolds and Prandtl numbers, denoted as Reo and Prf, are defined as [71]: StL ¼

Reo ¼

u0 l0 u0 2 ¼ ðnÞf ðnÞf

ð116Þ

ðnÞf af

ð117Þ

Prf ¼

where (n)f, o ¼ 2pf, u0 ¼l0o, and af are kinematic viscosity of the fluid, angular frequency of the piston, oscillation amplitude of velocity and thermal diffusivity of the fluid, respectively. When the channel is widened, the hydraulic diameter is reduced and thus increasing the velocity of the working fluid. Consequently, both the Strouhal number St and the Reynolds number Reo are increased, thus resulting in a larger Nusselt number and a higher heat transfer coefficient between the working fluid and the pyroelectric material. As a result, higher temperature change in the PE is achieved. The energy density for an individual cycle (ND) is represented by the area encircled by the Olsen cycle in the D–E diagram, as shown in Fig. 3. It can be derived by the integration of the elementary electrical work, EdD, over the cycle, which is as follows [66]: I ND ¼ EdD ð118Þ Power density, PD, with the unit of mW L 1, refers to the quantity of the energy generated by the PE per unit volume per unit time, given by PD ¼ ND/t, with t to be the cycle length of time. As the pyroelectric material remains unrestrained and the polar axis is in the direction perpendicular to the plain of the electrodes, the elementary change in the electric displacement is obtained from Eq. (3), that is: dD ¼ edE þ pdT

ð119Þ

Dielectric constant is described as e¼ere0, where er refers to the relative dielectric constant of the material and e0 refers to the dielectric constant of vacuum (8.854  10 12 F m 1). p refers to pyroelectric coefficient that is defined as p ¼ ∂D/∂T, with the unit of mC cm 2. As mentioned above, if the PEs are not unclamped, the pyroelectric coefficient is made up of two parts, (1) the primary pyroelectric coefficient due to the change in dipole moment and (2) the secondary pyroelectric coefficient because of variation in the crystal’s dimension. Both are consequential of the change in devices’ temperature. The unconfined pyroelectric coefficient subjected to a constant stress (X) at constant electric field (E) is rewritten as follows:   ∂D d33 a3 p¼ ð120Þ ¼ px;E s33 ∂T X;E

where px,E refers to the primary pyroelectric coefficient subjected to constant strain (x) at constant electric field (E). The latter term at the right hand side refers to the secondary pyroelectric coefficient that is related to thermal expansion of the pyroelectric materials, which is attributed to the piezoelectric effect, as discussed previously. a3 ¼ x3/DT is the coefficient of thermal expansion (CTE), while d33 and s33 refer to the piezoelectric coefficient in the unit of C N 1 and elastic compliance in m2 N 1. If the PE is confined, the secondary pyroelectric coefficient tends to zero, due to the absence of expansion. For isoelectric field processes of 2–3 and 4–1, the primary pyroelectric coefficient is assumed to be zero. For isothermal processes of 1–2 and 3–4, the second term on the right hand side of Eq. (120) is zero. In this regard, the energy density per cycle is obtained, via integrating Eq. (119) over the four processes, given by: Z Thot Z Tcold Z EH Z EL ND ¼ e0 er ðE; Tcold ÞEdE þ EH pðEH ; TÞdT þ e0 er ðE; Thot ÞEdE þ EL pðEL ; TÞdT ð121Þ EL

Tcold

EH

Thot

Pyroelectric Materials

741

In the Olsen cycle, modeling of the energy generation remains challenging, since the properties of the materials have strong dependences on electric field and temperature. Furthermore, because pyroelectric materials have hysteresis behaviors in electrical and thermal responses, the properties experimentally recorded cannot match with the operating conditions. As a result, it is necessary to have further assumption, in order to obtain the accurate values of energy density. Since temperature has a strong effect on the properties of the pyroelectric materials especially near the phase transition region, the dependence of the material properties on temperature must be included. The variation in electric displacement as a function of temperature is the primary pyroelectric coefficient, which is expressed as:   ∂D ∂er ∂Ps px;E ¼ Eþ ð122Þ ¼ e0 ∂T ∂T ∂T x;E This means the primary pyroelectric coefficient consists of dielectric and dipole contributions, corresponding to the two items in Eq. (122). If the secondary pyroelectric coefficient is concluded, the unclamped pyroelectric coefficient is written as: p ¼ px;E

d33 a3 ∂er ∂Ps Eþ ¼ e0 s33 ∂T ∂T

d33 a3 s33

ð123Þ

Combination of Eqs. (119) and (121) gives rise to the energy density for Olsen cycle, i.e.:    e0 d33 x3 ½er ðTcold Þ er ðThot ފðEH þ EL Þ þ Ps ðTcold Þ Ps ðThot Þ þ ND ¼ ðEH EL Þ 2 s33 where x3 ¼ a3(Thot Tcold). As the pyroelectric coefficient remains independent of temperature and electric field, Eq. (124) can be simplified as: Z EH ND ¼ ðThot Tcold Þ pdE ¼ pðThot Tcold ÞðEH EL Þ

ð124Þ

ð125Þ

EL

It has also been observed that the temperature change is decreased with increasing frequency, which is independent on the stroke length [67]. Meanwhile, at a certain frequency, the temperature change is increased with increasing stroke length. As the stroke length is sufficiently long, the fluid particle can be found on the heating element, as the piston reaches its peak, if the thermal energy is transfers to the bottom of the PE, once the piston approaches its lowest. Consequently, a larger temperature change is generated in the PE. Concurrently, the amount of heat required to maintain the favorable temperature of the PE is increased, as the working frequency and the stoke length are increased. Fig. 7 plots energy densities (ND) as a function of high electric field (EH), in the range of 220–415 kV cm 1, at frequencies of 0.035, 0.061, and 0.12 Hz [79]. The piston stroke length, channel width and low electric field are 4.7 cm, 330 mm, and 202 kV cm 1, respectively. In this case, a maximum energy density of 130 J L 1 can be achieved at 0.061 Hz, as the high electric field EH is 379 kV cm 1. As the level of EH is 4379 kV cm 1, the energy density is reduced, which can be ascribed to the increment in leakage current. Conversely, if the values of EH are less than 379 kV cm 1, the relatively low energy density is a result of the small difference in electric field (EH–EL). This is why the EH ¼ 379 kV cm 1 is the optimized electric field in terms of energy density.

Energy density, ND (J/L)

150

100

50

Frequency 0.035 Hz 0.061 Hz 0.12 Hz

0 200

250

300

350

400

450

High electric field, EH (kV/cm) Fig. 7 Energy densities as a function of high electric field EH at distinct frequencies. The channel width, stroke length, low field, and mean temperature T 3 were consistently 330 mm, 4.7 cm, 202 kV cm 1 and 74711C, respectively. Reproduced with permission from Nguyen H, Navid A, Pilon L. Pyroelectric energy converter using co-polymer P(VDF-TrFE) and Olsen cycle for waste heat energy harvesting. Appl Therm Eng 2010;30:2127–37, Copyright 2010, Elsevier.

742

Pyroelectric Materials

As the stroke length of the piston, width of the stack channel and arithmetic mean temperature at the center of the PE are kept to be constant, both the maximum energy and power densities depend on frequency, at certain low and high electric fields. Consequently, a lower energy density at 0.035 Hz is achieved even though the temperature change is larger than that of 0.061 Hz. Comparatively, at 0.035 Hz, the cycle time is relatively longer, which makes it possible for the surface charges of the PE to flow through, i.e., the presence of leakage current. As a result, heat dissipation takes place due to the Joule heating effect. In comparison, as the frequency is 40.061 Hz, a decrease in the energy density is observed, since the temperature change is decreased. Therefore, the temperature change and the cycle time should be well balanced with the leakage current. As discussed before, the heat losses to the ambient by the device should be minimized to ensure the energy conversion efficiency. Therefore, the outer cylinders of the device are usually made of Teflon and the insulation of the device can be further improved with the addition of fiberglass wool. It has been observed that the heat loss of the device in the range of 15–20% of the total heat input; the heat loss will be even more pronounced at elevated operating temperature. Additionally, much room for improvement in the efficiency of the device can be expected for its efficiently currently registered 0.045% and 0.053% at 0.061 and 0.12 Hz, respectively. Therefore, further study is necessary to increase energy efficiency of the pyroelectric converters.

2.23.6.3

Heat Conduction

It has been demonstrated that heat conduction can also be used to increase and decrease the temperature of the PE to maintain the Olsen cycle [72]. Fig. 8 shows photograph and a schematic for the experimental setup with conduction heating and cooling. Similarly, the system contains a thermal subsystem and an electrical subsystem. The thermal subsystem is comprised of two cold and hot aluminum blocks, with a dimension of 3 cm  2.5 cm  1.27 cm.

Temperature controller

J-thermocouple (TPF)

Wires to electrical circuit

Heater wires to controller

Hot source

Cold source 60/40 P(VDF-TrFE) sample

(A)

Process 2−3

Process 1−2 Thermal conductive epoxy (Omegabond-200) Cartridge heater

1.27 cm

TH Process 4−1

3 cm

TC Process 3-4

(B) Fig. 8 (A) Thermal subsystem setup to generate periodic temperature oscillations during the Olsen cycle, connected with the pyroelectric element (PE)-stamp assembly. (B) Schematic diagram of each process in the Olsen cycle during stamping experiments. Reproduced with permission from Lee FY, Navid A, Pilon L. Pyroelectric waste heat energy harvesting using heat conduction. Appl Therm Eng 2012;37:30–7, Copyright 2012, Elsevier.

Pyroelectric Materials

743

They served as the cold and hot reservoirs, at temperatures of TC and TH, respectively. The power supplier is a 50 W Omega CS10150 cartridge heater that sits within the hot block. TH is monitored and moderated by using a temperature controller; TC is set to be the room temperature, made possible through heat loss by convection at the ambient pressure. A layer of thermal conductive epoxy, which has high electrical resistance but relatively high thermal conductivity, covers the top surfaces of both blocks. As a result, the PE’s electrodes are electrically isolated. As the same time, the thermal contact resistance is significantly decreased. The electrical resistivity and thermal conductivity of the item are 1015O cm and 1.384 W m 1 K 1, respectively. The PE is attached to a wooden stamp’s base, which is employed to conveniently and safely handle the film at different pressures. J-type thermocouples are used to monitor and control the temperatures, which are placed centrally of the hot and cold blocks. Additionally, another Jtype thermocouple joins at the top of the film, which is in thermal contact with the electrodes, but electrically isolated. The temperatures read by the thermocouple are assumed the values of the film for its low thickness (in micrometer). The electrical subsystem is identical to the one in the previous example. The P(VDFeTrFE) film is poled at an electric field of 200 kV cm 1. The hot block is set at a temperature of TH ¼ 901C. After that, the stamp assembly that is used to hold the PE is attached to the hot block and kept for 70 min, during which the PE’s resistivity is being observed until a constant value of 5.22  1010 O m is reached. After the PE is successfully poled, the device can be subject to the Olsen cycle for energy conversion. To maintain the time-dependent temperature oscillations for the process of Olsen cycle, cyclic heating and cooling of the PE is possible through alternating contact with the hot and cold blocks, respectively. It takes more time for the film to approach the high and low temperatures of TH and TC, from the hot and cold sources (i.e., processes 2–3 and 4–1), than to fully charge and discharge the PE (i.e., processes 1–2 and 3–4). The Olsen cycle is performed at the high electric fields of EH ¼ 290–475 kV cm 1. Meanwhile, the low electric field of EL remains at 200 kV cm 1, so that the pyroelectric film is not depoled during the Olsen cycle. TC and TH are kept at 251C and 1101C, respectively. Fig. 9 shows variation pattern of temperature of the PE in seven consecutive Olsen cycles, with frequencies across 0.066 to 0.078 Hz, TC at 251C and TH at 1101C. [72]. The PE is experienced average minimum and maximum temperatures in one cycle to be denoted as Tcold and Thot, respectively. The temperature of the PE is oscillated between Tcold ¼ 45.21C and Thot ¼ 94.81C. It is necessary to mention that there is always TH4Thot and TCoTcold, simply because thermal contact resistance is always present between the PE and both the hot and cold sources. Also, the contacting actions are within a limited time duration. Fig. 10 displays D–E plot of the system after one Olsen cycle, at electric fields in the range from EL ¼ 200 to EH ¼350 kV cm 1, whereas the temperatures of the cold and hot sources are TC ¼ 251C and TH ¼ 1101C. In this case, the starting and ending points of the Olsen cycle are not coincide, with the difference that is ascribed to the presence of leakage current in the PE [73]. The charged impurities are migrated at high temperatures and high electric fields. Moreover, the Process 3–4 exhibits a step-like behavior, which is closely related to the incomplete phase transition from ferroelectric to paraelectric, corresponding to the Process 2–3 shown in the figure. An energy density of 155 J L 1·per cycle is recorded, according to the enclosed area of the 1–2–3–4 region in the D–E curve. The 60/40P(VDFeTrFE) polymer ferroelectric material has a Curie temperature of 661C without any applied external electric field [74], which can be increased to 92.51C and 1201C, if it is applied at 300 kV cm 1 and 527 kV cm 1 [75]. Consequently when subjected to high electric field beyond 350 kV cm 1, the hot source temperature TH is lower than 1101C, the operating temperature of the film Thot is actually lower than the Curie point. In this case, the phase transition of ferroelectric to paraelectric remains 100 Thot 90

TPE (°C)

80 70 60 50 Tcold 40 30 0

500

1000

1500

2000

Time, t (s) Fig. 9 Variation in temperature of the pyroelectric element (PE) over seven consecutive Olsen cycles. The operating conditions were TC ¼251C, TH ¼1101C, EL ¼200 kV cm 1, and EH ¼350 kV cm 1 with the cycle frequency in the range of 0.066–0.077 Hz. Reproduced with permission from Lee FY, Navid A, Pilon L. Pyroelectric waste heat energy harvesting using heat conduction. Appl Therm Eng 2012;37:30–7, Copyright 2012, Elsevier.

744

Pyroelectric Materials

0.110

Electric displacement (C/m2)

0.105

EL

2

Tcold

0.100

3 0.095

1

0.090 EH 0.085

Thot 4

0.080 0.075 0.070 0.065

4′ 200

ND = 155 J/L PD = 10.3 W/L 250 300 Electric field (kV/cm)

350

400

Fig. 10 Electric displacement vs. electric field (D–E) diagram obtained for Olsen cycle, with a 60/40P(VDFeTrFE) pyroelectric film of 1 cm  1 cm in area and 60.45 mm in thickness, between temperatures of TC ¼ 251C and TH ¼1101C, at electric fields of EL ¼200 kV cm 1 and EH ¼350 350 kV cm 1. Reproduced with permission from Lee FY, Navid A, Pilon L. Pyroelectric waste heat energy harvesting using heat conduction. Appl Therm Eng 2012;37:30–7, Copyright 2012, Elsevier.

700 Stamping experiments, TH = 110°C Convective heat transfer device (22), TH = 90°C

Energy density, ND (J/L/cycle)

600

Dipping experiments (16), TH = 110°C

500 400 300 200 100 0 200

300 400 500 High electric field, EH (kV/cm)

600

Fig. 11 Energy densities as a function of the high electric field EH for various forms of heat transfer. Operation conditions include: (1) EL ¼200 kV cm 1, EH ¼290–475 kV cm 1, TC ¼251C (present study), (2) EL ¼200 kV cm 1, EH ¼300–600 kV cm 1, TC ¼251C (dipping experiments) [87], and (3) EL ¼202 kV cm 1, EH ¼233–475 kV cm 1, TC ¼251C (convective heat transfer device) [67]. Reproduced with permission from Lee FY, Navid A, Pilon L. Pyroelectric waste heat energy harvesting using heat conduction. Appl Therm Eng 2012;37:30–7, Copyright 2012, Elsevier.

incomplete at the electric fields of EH4350 kV cm 1. Once the temperature of the hot source is raised to 1301C, Thot will be increased, short-circuit will take place, because the dielectric strength of air in the vicinity of the hot block is decreased, because of the increase in temperature. Additionally, if the hot source temperature is above 1101C, excessive leakage current is generated. In this case, the rate of surface charges migration will be higher than that of electrical discharge during the Process 2–3, thus leading to the crossovers between the Process 1–2 and Process 3–4 in the D–E curve. Therefore, the hot source temperature TH of 1101C is the optimal value so as to get the maximum energy density for the specific pyroelectric material of 60/40P(VDFeTrFE). Fig. 11 shows energy density of the system, as a function of the applied high electric field EH, in the range between 290 and 475 kV cm 1. The system is worked at the low electric field EL of 200 kV cm 1, with the cycles are conducted at frequencies in the range of 0.060–0.077 Hz. As the high applied electric field EH raises up to 350 kV cm 1, the energy density first increases and then saturates at 350 kV cm 1. Exceeding this optimal electric field at 350 kV cm 1, the energy density begins to decline. The optimal energy density of 155 J L 1·per cycle is recorded at the high applied electric field of EH ¼350 kV cm 1 at the frequency of 0.066 Hz,

Pyroelectric Materials

745

which corresponds to a power density of 10.3 W L 1. Initially, such increase in high electric field EH implies an extension in the electric field span of (EH–EL), which brings about enhanced energy density and power density. Simultaneously, leakage current is also increased as the electric field is raised, thus leading to the presence of the peak energy and power densities. In order to minimize the thermal contact resistance between the PE and the aluminum blocks, more pressure is applied to the film when it is attached to the blocks. As a result, the time durations required for PE to approach TC and TH will be reduced, which results in an increase in the energy density and power density. On the other hand, the pressure cannot be too high, because a too high pressure could cause electrical short-circuiting of the PE. It is more pronounced in the case of hot block contact. One of the reasons is the generation of microcracks in the materials, due to the roughness of the aluminum blocks. As a result, the applied pressures should not be higher than 200 kPa, which is used in the experiment. In addition, the mechanism of heat transfer has demonstrated a significant influence on the energy efficiency of the pyroelectric thermal converters. For instance, the maximum energy density recorded by this study, standing at 155 J L 1 per cycle, remains relatively low, as compared with those achieved by using other experimental methods [73]. This is mainly because, for instance for the dipping method, both the temperatures (Thot ¼ 1001C) and the applied electric fields are higher (EH ¼ 600 kV cm 1) [73]. Furthermore, silicone oil possess a much higher electric breakdown strength as compared to that of air [75]. When conducting the Olsen cycle, the electric fields applied to the pyroelectric materials should be as high as possible. As a result, it is very difficult to achieve a high energy density and power density, by using bulk materials and thick films. This is because they require even high voltages to ensure the high electric fields and high voltages are not practical for real device applications. With this concern, thin films have to used, so that the applied voltages can be effectively decreased, for a given electric field. Unfortunately, if the thickness of the film is further reduced, leakage current will be increased significantly [76,77]. Therefore, there is always a trade-off when designing pyroelectric devices for practical applications.

2.23.7 2.23.7.1

Olsen Harvesters Based on Perovskite Pyroelectric Materials Lead Lanthanum Zirconate Titanate Ceramics

Ferroelectric materials with perovskite structure have shown the highest pyroelectric properties among all ferroelectric materials [78]. Relaxor ferroelectrics, lanthanum (La) doped 65 mol% lead zirconate and 35 mol% lead titanate solid solutions, Pb1 xLax(Zr0.65Ti0.35)1 x/4O3 or x/65/35 lead lanthanum zirconate titanate (PLZT), have been demonstrated to be the most promising candidates for pyroelectric applications. In reference to the phase diagram of x/65/35 PLZT, at ambient temperature in the absence of an applied external electric field, the materials with compositions in the range of 5–9 mol% are of rhombohedral structure, which is close to the phase boundary of rhombohedral ferroelectric-mixed ferroelectric/cubic [79]. For these materials, their phase transition temperature is known as the Burns temperature, which is TBE3501C, similar to the Curie temperature for those without the presence of La [80–83]. At temperatures of below TB, the materials are at the state of ergodic relaxor, in which the domains are aligned and the domain correlation radius is increased accordingly. Distorted crystal structures, due to the interactions amongst the randomly oriented polar nanodomains, are accompanied by high dielectric constant and large piezoelectric coefficient, as well as high pyroelectric coefficient [84]. With the growth and coalesce of domains, polar clusters are gradually formed, which results in the inactivation of the random fields that are induced in the relaxor state. As a result, the phase transition from ergodic relaxor (nanodomain) to ferroelectric (macrodomain) is started at the Curie temperature TCurie, which is independent on external electric field. In this case, the Cuire point TCurie corresponds to the peak of pyroelectric coefficient (∂D/∂T)s,E vin temperature dependent curve, which is recorded at a given electric field E [85]. Accordingly, if the materials are heated, the phase transition from ferroelectric to ergodic relaxor will take place, as they across the Curie temperature TCurie. Similar behavior is observed when they are depoled as an electric field that is below the critical value Ecr(T). However, the ferroelectric phase of x/65/35 PLZT is not built up spontaneously, if the samples are cooled in the absence of an external electric field [94]. Therefore, the external electric fields E should be higher than the critical field Ecr(T), in order to achieve stable ferroelectric phase [86]. When the temperature is increased above the Curie point TCurie, the isothermal D–E loops of x/65/35PLZT will be seriously narrowed and even become linear ones. Therefore, one the relaxor state is present, the remnant polarization Pr(T) and the coercive field EC(T) are significantly decreases and gradually approach to zero. In contrast, both the remnant polarization Pr(T) and the coercive field EC(T) are pretty high in their ferroelectric state. Usually, a relatively large change in electric displacement D is observed corresponding to the relaxor-ferroelectric phase transition. When using the relaxor-ferroelectric materials for pyroelectric applications, the materials selected should have the relaxor-ferroelectric phase transition that is present during the Olsen cycle, so as to achieve optimal energy density and power output. For example, x/65/35 PLZT ceramics, with x¼ 5, 6, 7, 8, 9, and 10 mol%, have been studied for pyroelectric applications. The ceramics are prepared via the conventional solid state reaction method, by first thoroughly mixing precursory lead carbonate (PbCO3), zirconium dioxide (ZrO2), lanthanum oxide (La2O3), and titanium dioxide (TiO2), followed by calcination at 9001C for 6 h. The as-prepared powders are crashed and re-milled. They are finally sintered at 13001C for 2 h in an Pb rich atmosphere, so as to offset the volatilization of Pb during the high temperature sintering process [87]. The sintered ceramics are directly used for the characterization and pyroelectric device applicaton.

746

Pyroelectric Materials

According to the D–E curves of the 9/65/35 PLZT ceramics, six different Olsen cycles are conducted. The low electric fields are EL ¼ 0, 0.1, 0.2, 0.3, 0.4, and 0.5 MV m 1 and the high electric field is EH ¼ 4.0 MV m 1, while TC and TH are 31C and 1501C, respectively. Energy density and power density as a function of EL have been obtained, indicating optimal energy density of EL ¼ 0.4 MV m 1 has been achieved. When the level of EL is lowered from 0.4 to 0 MV m 1, the average energy density also experiences a decline from 368.2 to 143.0 J L 1·per cycle. This reiterates the fact that the material cannot be repolarized in the absence of an applied external electric field, while the temperature is lowered from Thot to Tcold (Process 4–1). When the material is de-poled at electric fields of below or near the coercive field of EC ¼ 0:09 MV m 1, at Thot ¼ 1501C, the Olsen cycle D–E curve has a cross-over between the Process 1–2 and Process 3–4 (EL ¼0 MV m 1), which leads to low energy density. Meanwhile, if the low electric field EL is raised from 0.4 to 0.5 MV m 1, a decrease in the average energy density from 368.2 to 325.7 J L 1·per cycle is encountered, since the electric field span of (EH–EL) in the Olsen cycle is reduced. The optimal power density is achieved at the low electric field of EL ¼ 0.2 MV m 1. At the same time, at EL ¼ 0.2, 0.3, and 0.4 MV m 1, energy densities recorded are very close one another, and this is also observed in materials with other compositions. The energy densities of the x/65/35 PLZT ceramics have been examined as a function of high electric field EH. The observed increase in energy density is associated with extending electric field span. Also, the limitation of the value of EH is dependent on dielectric breakdown strength of the materials. Generally, suppose the applied electric field is above the dielectric breakdown field, microcracks will be produced inside active materials. The presence of cracks is mainly ascribed to the mechanical stresses accumulated in the materials at the grain boundaries, because of the uneven distribution of the electric fields, which leads to preferential motion of the domain walls [88]. Furthermore, due to the cyclic electric field loading/unloading, propagation of microcracks along the grain boundaries of the materials will first occur, and eventually sample failure will ensue. Similarly, the 7/65/35 PLZT sample has an optimal energy density of 1013.5716.2 J L 1 per cycle, which is equivalent to a power density of 25.970.8 W L 1, which are the highest performances ever reported in the open literature. This outcome is derived from the Olsen cycles that are conducted at the frequency of 0.0256 Hz and the electric field is cycled between the value of EL ¼ EL ¼0:2 MV m 1 and high value of EH ¼ 7.0 MV m 1. Meanwhile, the cold and hot source temperatures are 301C and 2001C, respectively. Energy and power densities of the 9/65/35 PLZT ceramics are also evaluated, at different cycle frequencies. At Tcold ¼ 31C and Thot ¼ 1501C, the four distinct processes in the Olsen cycle are carried out at frequencies of lower than 0.036 Hz. The experiments are conducted under the quasi-equilibrium conditions. An optimal energy density of 509.4729.6 J L 1 per cycle is achieved at the frequency of 0.036 Hz. At the low frequencies of less than 0.036 Hz, the device encounters an increase in ND, which is ascribed to the presence of excessive leakage current in the active materials. This is because if the frequency is sufficiently low, surface charges of the PE have enough time to migrate through the materials, appearing as leakage current. This phenomenon is more pronounced at elevated applied electric fields and working temperatures. In a similar way, the 9/65/35 PLZT ceramics also experience a decrease in energy density related to the leakage current, which is 10% and 20% if the Olsen cycle is conducted at frequencies of 0.036 Hz and 0.02 Hz, respectively. Beyond 0.036 Hz, the energy density declines, because the electric displacement span is reduced, due to the fact that the processes are conducted away from the quasi-equilibrium conditions. In other words, the Olsen cycle does not coincide with the isothermal D–E loops. The power density of the 9/65/35 PLZT ceramics is maximized at 32.470.8 W L 1, at the frequency of 0.096 Hz. The reduction in power output above this optimal frequency can be understood from the point of view of relaxation and heat transfer. The realignment of dipoles during the heating and cooling at the isoelectric fields, corresponding to the Processes 2–3 and 4–1, requires a certain time duration. Therefore, as the frequency is too high, the realignment cannot be fully achieved, because the dipole relaxation is slow, which is the intrinsic characteristic of ferroelectric-relaxor materials. This is more pronounced if the temperature is lower than the Curie point of TCurie, since of the nanodomains are frozen at low temperatures, thus require more energy to realize the reorientation [89]. In addition, there is insufficient time for the materials to approach the thermal equilibrium, once the Olsen cycle is conducted at high cycle frequencies. Consequently, phase transition is not completely fulfilled, which results in reduced electric displacement span and thus lowering energy and power densities. The 9/65/35 PLZT ceramics exhibit a peak power density, which is improved by raising high electric field EH, due to rising energy density ND. This is readily ascribed to the widening of the electric field span and increasing the time rate of change in the electric field at a given cycle frequency. Peak power densities can be obtained at the frequencies of 0.0859, 0.0961, and 0.0709 Hz, corresponding to the EH of 5.0, 6.0, and 7.0 MV m 1, respectively. Energy densities of the 8/65/35 PLZT ceramics have been evaluated as a function of high electric field. The values of energy density are derived from the average of five cycles that are conducted under the quasi-equilibrium conditions [90]. It has been identified that more prominent variation is observed at low temperatures (Thot) and low electric fields (EH). Specifically, the highest variation is observed at the hot temperature of Thot ¼ 1001C and high electric field of EH ¼0.4 MV m 1. In comparison, at the hot temperature of Thot ¼ 1301C and high electric field of EH ¼ 1.5 MV m 1, the lowest variation is encountered, which corresponds to a maximum relative difference of 9.1% among the samples.

2.23.7.2

Perovskite Relaxor-Ferroelectric Single Crystals

As relaxor-ferroelectric single crystals, (1 x)PbMg1/3Nb2/3 xPT (PMN-PT) and (1 x)PbZn1/3Nb2/3O3 xPbTiO3 (PZN-PT), have been employed for pyroelectric applications to harvest waste thermal energies [91–95]. For example, there is a report on 0.75PMN-0.5PT single crystal, which recorded optimal energy density of 186 mJ L 1 per cycle. The device is fabricated with the

Pyroelectric Materials

747

dipping experiments, while the cycle is carried out at temperatures of 35–851C and electric fields of 0–30 kV cm 1 [91]. Another example is PZN-4.5PT single crystals, which have phase transitions of rhombohedral, orthorhombic, and tetragonal during the Ericsson cycle, thus triggering thermal energy harvesting capability. An optimal energy density of 101.8 mJ L 1 per cycle can be realized, as the pyroelectric device is cycled between 100 and 1301C and electric fields of 0–20 kV cm 1 [93]. In addition, relaxor ferroelectric pyroelectric single crystals, 68PbMg1/3Nb2/3O3-32PbTiO3 (PMN-32PT), exhibit a highest energy density of 100 mJ L 1 per cycle, which corresponds to a power density of 4.92 mW L 1, for the device is subject to cycles at temperatures between Tcold ¼ 801C and Thot ¼ 1701C, as well as electric fields between EL ¼ 2 kV cm 1 and EH ¼ 9 kV cm 1 [94].

2.23.8

New Pyroelectric Materials and Devices

In addition to the Olsen cycle that has been employed for pyroelectric thermal energy harvesters, various alternative converters have also been developed [96–98]. One example is Ericsson-based thermodynamic cycles, which have been demonstrated to be viable for pyroelectric thermal energy harvesting [96]. The energy harvesting efficiency is up to 100 times more efficient than with merely simple linear pyroelectric effect. The mechanism of pyroelectric energy harvesting of the Ericsson cycle remains quite similar to that of ECE. As a result, the materials that have high electrocaloric efficiency would have high pyroelectric energy harvesting capability. The harvested energy per unit volume can be determined by the product of the Carnot efficiency and the ECE. Usually, the possible harvested energy efficiency is in the range of 50–600 mJ L 1 with a temperature change of 101C. For example, by using the Ericsson cycle on a 0.90Pb(Mg1/3Nb2/3)O3-0.10PbTiO3 ceramics, an energy density of 186 mJ L 1 at 3.5 kV mm 1, as the temperature change is about 501C. The corresponding pyroelectric thermal energy harvesting cells can be assembled by using a simple method, in which commercially available lead zircornate titanate (PZT) and PVDF are used as the raw materials, while screen printing technique is utilized to fabricate the devices [98]. The PZT cells with a square pyroelectric layer (4 cm  4 cm) are inserted between two PdAg electrodes, which are screen-printed on Al2O3 ceramic substrates. At the same time, PVDF layers are coated on flexible plastic substrates. As the temperature is fluctuated between 300 and 360K, over a time period of about 100 s, the device has a current output of 10 7 A and charge of 10 5 C. A similar behavior is observed, as the pyroelectric harvesters are heated and cooled. It is found that the charge generation is closely related to the difference of temperature between two time instances, while it is unaffected by the temporal evolution of the temperature. The generated current of the PZT cells is increased, as the pyroelectric layer’s thickness and poling electrical field during the fabrication process are increased. A stored energy of 0.5 mJ, sufficient to sustain the operation of an autonomous sensor node, has been achieved in the course of the measurement and transmission cycle [98]. A pyroelectric nanogenerator based on a PZT thin film has been reported [97]. Fig. 12(A) shows a photograph of the pyroelectric nanogenerator, which measures 21 mm in length and 12 mm in width. A representative cross-sectional scanning electron microscope (SEM) image of the device is displayed in Fig. 12(B), indicating that the PZT layer has a thickness of about 175 mm. Fig. 12(C) illustrates a high magnification cross-sectional SEM image, revealing the high polycrystalline nature of the PZT film. Fig. 12(D) depicts output voltage and current of the nano-harvester, which have been recorded between temperatures 295–299K. The peak value is observed at the rate of temperature change at approximately 0.2 K s 1. If there is a rapid change in temperature from 295 to 299K, sharp negative voltage and current pulses (2.8 V and 42 nA) can be generated, for forward connection, whereas positive pulses are detected as the temperature decreases from 299 to 295K, as clearly demonstrated in Fig. 12 (E). As the connection is reversed, i.e., backward connection, the convention of all the signals are reversed, supporting the notion of electricity is indeed generated by the nanogenerator [97]. The nanogenerator can generate electrical energy that can be stored for various applications [97]. Fig. 13(A) displays a schematic diagram, which demonstrates that the charging of a Li-ion battery (LIB) with the electricity generated by the nanogenerator and capability to power a commercial LED. Fig. 13(B) is a photograph of a Li-ion coin cell. This LIB consists of anode made up of TiO2 nanotube array grown on Ti foil. The TiO2 nanotubes have diameter of 150 nm and length of 10 mm, as seen in Fig. 13(C). The constant current charge–discharge curves of the battery are illustrated in Fig. 13(D). The nanogenerator is operated at a frequency of 0.005 Hz. In about 3 h, the battery may be charged to 810 mV, with the electricity generated by the nanogenerator. The voltage peaks are related to the ever-present self-discharge phenomenon of the Li-ion storage. The two adjacent peaks of the charging have a time interval of approximately 200 s. At a given constant current of 1 mA, the discharging of the battery takes 84 s, so that the original state of 650 mV is reached. In this case, the Li-ion storage capacity is estimated to be around 0.023 mAh. The charged LIB can adeptly light up a green LED, as shown in Fig. 13(E). Moreover, the harvested electricity can even be utilized to drive wireless sensors [97]. A pyroelectric–piezoelectric nanogenerator (PPENG) has been combined with triboelectric device to form a triboelectric–pyroelectric–piezoelectric hybrid cell, consisting of a sliding mode triboelectric nanogenerators (TENG), which has demonstrated promising performances [99]. In harvesting the mechanical energy during the sliding motion, the TENG with a size of 63.5 cm2 has an area power density of 0.15 W m 2, at a sliding frequency of 4.41 Hz. The PPENG harvests both the thermal energy induced by the friction and the mechanical energy induced by the normal force. The energy produced with the hybrid cell can be used to power LED and charge supercapacitor. The charging rate is higher that of the individual TENG device by two times. With the frictional motion, the hybrid cell is able to drive self-powered temperature and normal force sensors. In other words, the PPENG can be used not only an energy harvester but also a self-powered device.

Pyroelectric Materials

748

PENG

50 μm

5 mm (B)

(C) 3.0

299 Voltage (V)

Temperature (K)

(A)

298 297 296 295 0

200

400 600 Time (s)

800

0.0 −1.5 0

200

400 600 Time (s)

0

200

400 600 Time (s)

800

1000

50 Current (nA)

dT/dt (K/s) (D)

1.5

−3.0

1000

0.2 0.1 0.0 −0.1 −0.2

2 μm

0

200

400 600 Time (s)

800

1000

25 0 −25 −50

(E)

800

1000

Fig. 12 (A) Photograph of the fabricated pyroelectric nanogenerator (PENG). (B) Cross-sectional scanning electron microscope (SEM) image of the PENG. (C) Enlarged cross-sectional SEM image of the PENG. (D) The cyclic temperature variation of the PENG and its corresponding differential curve. (E) Output voltage and current of the PENG with the temperature variation described in (D). Reproduced with permission from Yang Y, Wang SH, Zhang Y, Wang ZL. Pyroelectric nanogenerators for driving wireless sensors. Nano Lett 2012;12:6408–13, Copyright 2012, American Chemical Society.

Fig. 14(A) displays a schematic diagram of the hybrid cell, which is made up of a sliding-mode TENG on top and a PPENG at the bottom, with a multilayered planar structure [99]. In the TENG structure, a layer of Al foil is used as the sliding part, whereas a piece of polytetrafluoroethylene (PTFE) film is employed as the static part, on which a layer of Cu is coated as electrode. The PPENG is constructed with a piece of polarized PVDF, which is 110 mm thick and with two Cu electrodes on each side. A Kapton film is sandwiched as an insulating layer and a thermal conductor, which is in between the bottom electrode of the TENG and the top electrode of the PPENG. A good contact of Al foil and the PTFE film is maintained and well aligned during the sliding, by tailoring the thicknesses of both the sliding and static parts. The Al foil is controlled by using a linear motor to slide on the PTFE surface periodically, with a period of 0.2267 s and a displacement of 80 mm. The total frictional area was about 63.5 cm2. A thermocouple was attached to the center of the frictional area underneath the bottom side of the PPENG to record the temperature variation during the sliding motion. Fig. 14(B) shows mechanism of the TENG, which relies on the coupling of triboelectrification and electrostatic induction. Specifically, during the contact/friction between the two parts of the TENG, the Al layer was positively charged, whereas the surface of PTFE was negatively charged, because of their respective different surface polarities. Because the net negative charges on the surface of the PTFE were immobile, so that they could be retained for a sufficiently long time, due to the insulating polymer. As the Al foil was sliding away from the aligned position, the triboelectric charges on both surfaces would be separated, which resulted in a difference in potential between the Al and Cu electrodes. The difference in potential corresponded to the voltage for open-circuit, providing the driving force for the charge to flow in the short-circuit condition, as seen in Fig. 14(B-II), as the Al foil sliding entirely away from the top surface of the PTFE film, as illustrated in Fig. 14(B-III). As the Al foil sliding back, the open-circuit potential difference started to drop, so that the electrons flew back in the short-circuit state, reaching an equilibrium state, as demonstrated in Fig. 14(B-IV). Once the Al foil was sliding back to the aligned position (Fig. 14(B-I)), the sliding was conducted for one circle.

Pyroelectric Materials

749

Pyroelectric NG Battery

LED

(A)

(B)

1.2 Discharging

0.6

750 0.0 720 −0.6

690 660 0

Voltage Current 3000

6000 Time (s)

−1.2 9000 12,000

Voltage (mV)

780

780 Current (μA)

Voltage (mV)

810 Charging

(D)

1 μm

400 nm

(C)

5 mm

770 760 750 740 1200

200 s

1400

1600

Time (s)

1800 (E)

Fig. 13 (A) Schematic diagram demonstrating the process of charging the LIB and lighting up a LED. (B) Photograph of the assembled LIB. (C) SEM image of as-synthesized TiO2 nanotubes. (D) The charge–discharge curves of the LIB by using the PENG. (E) Photograph of a green LED powered by the LIB. LIB, Li-ion battery; LED, light-emitting diode; PENG, pyroelectric nanogenerator; SEM, scanning electron microscope. Reproduced with permission from Yang Y, Wang SH, Zhang Y, Wang ZL. Pyroelectric nanogenerators for driving wireless sensors. Nano Lett 2012;12:6408–13, Copyright 2012, American Chemical Society.

Descriptions of the pyroelectric and piezoelectric outputs of the PPENG, are shown in Fig. 14(C) and (D), respectively. The PVDF in the PPENG was placed in such a way that the polarization direction was upward, as illustrated in Fig. 14(C-I) and (D-I). The periodical sliding would apply a normal force to the underlying PVDF film, which generated heat due to the friction. As a result, the temperature would be increased. In this case, the dipole moments were weakened and thus the volume was increased, so that the polarization density was decreased. Because the reduction in polarization density induced extra charges, which would be balanced. Accordingly, a negative potential difference was built up across the PVDF, from the top to the bottom electrodes, under open-circuit condition, so that a current was flowing from the bottom electrode to the top electrode of the PVDF, in the case of shortcircuit condition, as revealed in Fig. 14(C-II). As the sliding motion was stopped (Fig. 14(C-III)), the polarization density of the PVDF film was regained, due to the heat dissipation. As a consequence, the absolute value of the potential difference in the open-circuit condition was reduced, thus leading to a current flowing in the reverse direction in the short-circuit condition, as seen in Fig. 14(C-IV). As the sliding part was moved in, the PVDF film was subject to a normal compressing force, so that volume of the PVDF film was decreased, thus leading to an increased in the polarization density, which corresponded to a current that flew from the top electrode to the bottom electrode, so that the extra polarization density in the short-circuit condition was balanced, as observed in Fig. 14(D-II). As the TENG is in the aligned position, as shown in Fig. 14(D-III), the normal force was maximized. When the sliding part was moved out, the applied force was decreased, so that the current would flow back to balance the charge, due to the absence of the strain (Fig. 14(D-IV)) in the short-circuit condition. The open-circuit voltage and the short-circuit charge owing to the piezoelectric effect were proportional to the magnitude of the applied normal force. A flexible pyroelectric generator based on a thin PVDF film has been developed to harvest thermal energy, which is realized via utilizing the time-dependent fluctuating temperature with spatial uniformity [100]. With a temperature variation of 50K, the PG

Pyroelectric Materials

750

Triboelectric output (1 sliding cycle) Sliding

I

II + + + + − − − − − − − −

+ + + + + + + + − − − −− − − −

1~

+ + ++

IV

III

2~ I

+ + + + − − − − − − − −

− − − − − − − −

+ + ++

+ + + + + + + +

Kapton

(A)

AI

PVDF

Cu

PTFE

Heat transport

Compressive force

(B)

Pyroelectric output I

II +

−

+

−

+

+ −

− +

+

+ − +

−

− +

− + + − + (C)

−

−

+ −

+

−

−

− + −

+

+

−

+

+ −

−

+

+

−

− +

+

− +

+ − +

−

− +

−

III

+ −

+

+ I

II

I −

−

IV

I

Piezoelectric output (1 sliding cycle)

−

+ −

−

− + − +

+ −

+ −

+

+ + −

+

−

− +

−

+

I

−

IV − − − − + + + + + + − − − − − + + + +

−

− − − − + + + + + + I

− +

− − − − − + + +

III − − − − − + + + + + + − − − − − − + + + + +

(D)

Fig. 14 Structure and working mechanism of the hybrid cells. (A) Structure of the hybrid cell, where “1” is the output from triboelectric nanogenerator (TENG) and “2” is the output from pyroelectric–piezoelectric nanogenerator (PPENG). (B) Working mechanism of output in the TENG. (C,D) Working mechanism of pyroelectric and piezoelectric outputs in the PPENG, respectively. PDVF, polyvinylidene fluoride; PTFE, polytetrafluoroethylene. Reproduced with permission from Zi YL, Lin L, Wang J, et al., Triboelectric-pyroelectric-piezoelectric hybrid cell for highefficiency energy-harvesting and self-powered sensing. Adv Mater 2015;27:2340–7, Copyright 2015, WILEY-VCH Verlag GmbH & Co. KGaa, Weinheim.

exhibited an open-circuit voltage of 8.2 V and a short-circuit current of 0.8 mA, corresponding to a maximal output power of 2.2 mW at a load of 0.1 MO. The energy harvested with the PG can capably substantiate an operating liquid crystal display (LCD) and light emitting diodes (LEDs) and charge commercial capacitor as energy storage devices. Furthermore, because the output voltage linearly increases with increasing change in temperature, the pyroelectric generator can be used as an important component in designing self-powered thermosensors. A self-sustaining polymeric pyroelectric nanogenerator has been reported to be assembled with commercially available PVDFs, which generates electricity from water vapor [101]. Because water vapor has a high latent heat, a temperature variation of as large as 231C s 1 can be readily achieved, by using automatic water condensation and evaporation of water on surface of the polymeric pyroelectric nanogenerator. Ultimately, the polymeric pyroelectric nanogenerator devilers an open-circuit voltage of 145 V and a short-circuit current density of 0.12 mA cm 2. The device has a volumetric peak power density of 1.47 mW cm 3, or an areal peak powder density of 4.12 mW cm 2. In addition, the polymeric pyroelectric nanogenerator can provide continuously power for different kinds of low-power electronic devices. At the same time, the electricity can also be stored in capacitors for further energy related applications. PbZr0.53Ti0.47O3/CoFe2O4 (PZT/CFO) multilayered nanostructures (MLNs) were explored in order to achieve enhanced pyroelectric energy harvesting capability, as well as large ECE [102]. Different from the conventional ECEs, the effect of PZT/CFO MLNs was controlled by dynamic magneto-electric coupling (MEC), which could be easily tuned by controlling the combination of the different ferroic layers. The effect of stacking number on ECE was studied, by using stacks of three (L3), five (L5) and nine (L9) alternating PZT and CFO layers. All the configurations showed a negative ECE. The maximum ECE temperature changes were 52.3K, 32.4K, and 25.0K, for the three (L3), five (L5), and nine (L9) layer structures, respectively. More importantly, a maximum pyroelectric energy harvesting, calculated using a modified Olsen cycle, exhibited an energy density of 11,549 kJ m 3 per cycle.

Pyroelectric Materials

751

The enhanced pyroelectric effect was ascribed to the cumulative effect of the multiple layers, which enhanced the overall polarization. The opens a new direction of pyroelectric energy harvesting devices. Porous Ba0.67Sr0.33TiO3 barium strontium titanate (BST) pyroelectric ceramics were fabricated by using carbon nanotubes (CNTs) as the pore agent [103]. The porous BST ceramics exhibited an extremely high pyroelectric coefficient of 9500 mCm 21C and a FOM of 32.0  10 5 Pa 0.5, which were higher than those of its dense counterpart by two and three times, respectively. Due to the presence of high porosity, the BST ceramics had a relatively low dielectric constant and a low heat capacity. The CNTs derived porous BST ceramics also demonstrated higher performances than those prepared by using other pore formers. More recently, a newly designed wind-driven pyroelectric energy harvester has been reported, with a propeller that was set in rotational motion with an incoming wind stream [104]. The speed of shaft of the propeller was reduced by using a gearbox based on a slider-crank mechanism, where a pyroelectric material was attached on the slider. Thermal cycling was realized as the reciprocating slider with the pyroelectric material moved across hot and cold zones alternatively, by using a stationary heat lamp and ambient temperature, respectively. The open-circuit voltage and closed-circuit current were studied in the time domain at various wind speeds. Wind speeds from 1.1 to 1.6 m s 1 were tested. The effectiveness and efficiency for energy harvesting were evaluated by using a 100 nF capacitor with a signal conditioning circuit. Different from the conventional wind turbines, the energy harvested through the pyroelectric material was decoupled from the wind flow, without involving any mechanical power, so that the device could work at relatively low wind speeds, for example, o2 m s 1. Fig. 15 shows a schematic diagram of the newly designed wind-driven pyroelectric energy harvester [104]. A slider-crank mechanism was employed by using a set of gears, together with a wind-driven lightweight propeller. As a result, high-speed rotational motion of the propeller led to a low-frequency reciprocating motion of the slider that was incorporated with a pyroelectric component, so that thermal cycling was enabled. An infrared lamp was used to generate heat. In this case the pyroelectric material passed slowly across the hot zone alternatively. A pyroelectric component, with a sufficiently small ratio of thickness to surface area, was selected to ensure that the main cooling mechanism was natural convection with air [105]. Temperature fluctuation on surface of the pyroelectric material was measured by using a thermocouple that was mounted on it. Experimental data were collected by connecting the pyroelectric component and the thermocouple onto a signal analysis system. Fig. 16 shows a photograph of the setup for experimental study [104]. A Dyson fan with a diameter of 10 in. was placed 1.2 m away from the propeller to generate a near laminar wind flow at certain speeds. The wind speeds used were all below the cut-in speed of the conventional wind turbines. A plastic propeller with a diameter of 12 in. was mounted on a steel shaft, which was connected to a plastic gearbox with a reduction ratio of 13. The entire setup was mounted on a benchtop for testing. A piece of PZT was used as the pyroelectric component, which was attached at the end of a light aluminum slider moved horizontally during the propeller cycle. The temperature of the plate surface was measured by using a K-type thermocouple, which had a response time of 0.5 s, so that it was sufficiently fast to capture the thermal oscillations at frequencies of o0.1 Hz. An external capacitor with a capacitance of 100 nF was charged by the wind-driven pyroelectric harvesting device. A heat lamp was used to simulate the heat source. Fig. 17 shows thermal power that is available for the absorption. To have high thermal efficiency, it was necessary to ensure that the pyroelectric component should have a large surface area, together with a thickness to be as thin as possible, so that the switching between heating up and cooling down would become swift. The PZT layer, with a diameter of 22 mm and a thickness of 0.2 mm, was coated on a circular brass disk with a diameter of 25 mm and a thickness of 0.3 mm. The piezoelectric component was a soft ferroelectric material, with a piezoelectric coefficient (d33) of 490 pC N 1, exhibiting a capacitance of 54 nF at 1 kHz, corresponding to a relative dielectric constant of 3210. It was found that the variation in temperature was very consistent, in the range between 2.9 cycles min 1 at a wind speed of 1.1 m s 1 and 5.7 cycles min 1 at the highest wind speed of 1.6 m s 1. Under this setting, the lower the wind speed, the higher the

Propeller

Heat lamp

Gears Connecting rod

Pyroelectric material

Crank Motion Slider

Fig. 15 Schematic diagram of the newly designed energy harvesting device with pyroelectric materials. Reproduced with permission from Xie MY, Zabek D, Bowen C, Abdelmageed M, Arafa M. Wind-driven pyroelectric energy harvesting device. Smart Mater Struct 2016;25:125023, Copyright 2016, IOP Publishing.

752

Pyroelectric Materials

Fan

Propeller

Heat lamp Slider Connecting rod Pyrolectric material

Fig. 16 Photograph of experimental setup of the wind energy harvester. Reproduced with permission from Xie MY, Zabek D, Bowen C, Abdelmageed M, Arafa M. Wind-driven pyroelectric energy harvesting device. Smart Mater Struct 2016;25:125023, Copyright 2016, IOP Publishing.

9

Lamp irradiance (mW/mm2)

8 7 6 5 4 3 2 1 0

0

20

40

60

80

100

Lamp power (%) Fig. 17 Lamp irradiance intensity vs. electrical power. Reproduced with permission from Xie MY, Zabek D, Bowen C, Abdelmageed M, Arafa M. Wind-driven pyroelectric energy harvesting device. Smart Mater Struct 2016;25:125023, Copyright 2016, IOP Publishing.

change in temperature would be. According to the measured open circuit voltage of the PE at different speeds of wind typical DT values of 4–6K from were equivalent to open circuit voltages of 10–16 V, while the experimental values were in the range of 6–9 V, which was mainly attributed to the temperature difference between the bulk and surface, conductivity losses, and substrate clamping. A nanogenerator based on one-structured multi-effects of multifunctional PZT has been reported [106]. In this case, a pyroelectric nanogenerator (PyENG), a photovoltaic cell (PVC) and a triboelectric-piezoelectric nanogenerator (TPiENG) were integrated into one device with same output electrodes, which could be used to harvest thermal, solar, and mechanical energies individually and/or simultaneously. The TPiENG exhibited low output voltage and high output current, while the PyENG and PVC had much higher output voltage and lower output current. In comparison, the multi-effect nanogenerator well adopted the advantages of individual components, demonstrating a promising electricity generation capability, with a high peak current of 5 mA, a high peak voltage of 80 V and a high platform voltage of 50 V. Fig. 18(A) shows a schematic diagram, describing the coupled technology toward multi-energy harvesting, combining piezotribo-pyro-photoelectric effects, as shown in Fig. 18(B). The nanogenerator was constructed with three parts. PZT block provided

Pyroelectric Materials

753

PyENG/ThENG-PiENG/TENG Th

EN

NG

G

TE

Thermal energy

Mechanical energy (C)

PiE

NG

NG -P

G/ EN

Solar energy

C

NG

PVC

PiE

-PV NG

/TE

E Th

NG

Py

VC

E

Py

Multi-effects coupled nanogenerator (A)

(D)

Nylon FEP AgNWs/PDMS ITO PZT Ag TE module Acrylic (B)

Fig. 18 Design of the multi-effect nanogenerators. (A) Schematic diagram of the coupled technology for multi-energy harvesting. (B) Schematic diagram of the one-structured multi-effect coupled nanogenerators. (C) A photograph of the device. (D) SEM image of the freshly obtained AgNWs. AgNWs/PDMS, silver nanowires/polydimethylsiloxane; FEP, fluorinated ethylene propylene; ITO, indium tin oxide; PZT, lead zircornate titanate; SEM, scanning electron microscope. Reproduced with permission from Zhang KW, Wang SH, Yang Y. A one-structure-based piezo-tribo-pyrophotoelectric effects coupled nanogenerator for simultaneously scavenging mechanical, thermal, and solar energies. Adv Energy Mater 2017;7 (6):1601852, Copyright 2017, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

piezoelectric, pyroelectric, and photoelectric effects. Polyamide (nylon) was employed as a flexible vibrating film to generate triboelectrification with fluorinated ethylene propylene (FEP) film and applied strain to the neighboring PZT during vibration. Ag thin film under the PZT was served as the bottom electrode, while an ITO thin film incorporated with composite film of Ag nanowires and polydimethylsiloxane (AgNWs/PDMS) was used as the top electrode. A TE module supported by a radiator provided the temperature adjunction. Fig. 18(C) shows a photograph of the one-structured nanogenerator. Fig. 18(D) illustrates a SEM image of the AgNWs, which possessed lengths of 5–15 mm, with an aspect ratio of about 100. Fig. 19(A) shows a schematic diagram, demonstrating cross-sectional configuration of the one-structured multi-effect nanogenerator, combining with working principles, i.e., pyroelectric effect, photoelectric effect, and triboelectric–piezoelectric effect. Charge generation process of the PyENG unit is shown in Fig. 19(B). At equilibrium state (dT/dt¼ 0), electric dipoles in the PZT component were distributed in a given format at room temperature, so there were no electrons to flow, because no variation in spontaneous polarization took place. As the PZT component was heated, i.e., dT/dt40, the electric dipoles were disturbed to deviate with respect to their alignment axes, thus leading to a decrease in spontaneous polarization. Accordingly, there were electrons that flew from the ITO electrode to the Ag electrode. However, if the PZT component was cooled, i.e., dT/dto0, the spontaneous polarization was enhanced, since the degree of deviation of the electric dipoles was reduced. As a result, a negative pyroelectric potential was generated across the electrodes, i.e., a reverse signal was created. At the same time, ferroelectric PZT has been demonstrate to have visible-light photoelectric effect, so as to be able to have applications as solar cells [107–109]. As the PZT component was irradiated with light with proper energy, electron–hole pairs as

754

Pyroelectric Materials

A

Nylon

FEP

PyENG

B

AgNWs/PDMS

ITO

C

PZT

Ag

TE module

D

PVC

(i)

(i)

− +

TPiENG (i)

− +

Ps

− +

e− dT/dt = 0

Acrylic

(ii)

Polarization

e−

(ii)

(ii) − + Ps

(v)

e− − +

e−

e−

− +

dT/dt > 0

(iii) (iii)

e−

(iv) − +

Ps

dT/dt < 0

− + e−

Fig. 19 Device architecture and working mechanism of the one-structured piezo-tribo-pyro-photoelectric effect coupled nanogenerator. (A) Crosssectional diagram of the one-structured multi-effect nanogenerator. (B) Physical mechanism of pyroelectric effect in PyENG. (C) Light-induced charge separation and the corresponding distribution of charge density in the PVC with ITO/PZT/Ag sandwich structure. (D) Working principle of the vibrating TPiENG. FEP, fluorinated ethylene propylene; ITO, indium tin oxide; PVC, photovoltaic cell; PyENG, pyroelectric nanogenerator; PZT, lead zircornate titanate; TPiENG, triboelectric-piezoelectric nanogenerator. Reproduced with permission from Zhang KW, Wang SH, Yang Y. A onestructure-based piezo-tribo-pyro-photoelectric effects coupled nanogenerator for simultaneously scavenging mechanical, thermal, and solar energies. Adv Energy Mater 2017;7(6):1601852, Copyright 2017, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

charge carriers were generated, due to the absorbed photons, which were accumulated toward the cathode and anode driven by the polarization-induced depolarization electric field, as shown in Fig. 19(C,i). As a result, Schottky barriers at the interface and depolarization electric field in the PZT layer were established, as seen in Fig. 19(C,ii).

Pyroelectric Materials

755

Fig. 19(D) shows working principle of the TPiENG unit, with a combination of piezoelectric and triboelectric effects. A cycled compressive strain was applied to the FEP, generating air-driven vibration of the nylon film. In this case, nylon was much more triboelectrically positive as compared with FEP [110]. According to the triboelectric effect, the top nylon and the bottom FEP were in portion closely contacted, with equivalent positive and negative charges to be on the nylon and FEP, respectively, at the initial stage, as illustrated in Fig. 19(D,i). Because the device was in an electrostatic equilibrium state, no flow of electrons occurred. However, the polarized PZT layer was compressed, a positive piezoelectric potential was present between the ITO and Ag electrodes, with electrons flowing from Ag to ITO. As the contacted nylon and FEP was slightly separated, induced charges would be present on ITO, owing to the triboelectric charges on FEP, so that the electric potential would drive the electrons to flow from ITO to Ag apart, as observed in Fig. 19(D,ii). Meanwhile, the compressing strain on the PZT layer was decreased, because of the separation of the two layers in triboelectric component, thus leading to a decrease in polarization density, so that the electron flew from the Ag electrode to ITO electrode, in order to balance the extra polarization induced by the remove of the strain. When the nylon and FEP layers were entirely separated, no compressing strain was applied to the PZT layer. As a consequence, the piezoelectric signals disappeared, because constant spontaneous polarization was recovered, whereas negative signals were produced due to the triboelectric effect. Because the FEP layer was insulator, the produced negative triboelectric charges on its surface were stationary [111]. Once the nylon and FEP were

PyENG

PVC

600

TPiENG

1200 4 Current (nA)

0

−300 −600

600

Current (nA)

Current (nA)

300

0

2 0

−2

0

8

(A)

16 24 32 Time (min)

40

48

−600

0

8

(C)

16 24 32 Time (min)

40

−4

48

0

10

20 30 Time (s)

40

50

10

20 30 Time (s)

40

50

(E) 3

120

0

16 24 32 Time (min)

40

48

600

15

400

10 5

200

0 (D)

0

16 24 32 Time (min)

40

0 (F)

6

750

4

500

2

250

0

(H)

4

40

3

32

2 20 1 8

0

0 101 102 103 104 105 106 Load resistance (Ω)

1

48

8

1000

0 (G)

8

Output power (μW)

8

Output current (μA)

0

2

0

−30

Output current (μA)

(B)

30

101 102 103 104 105 106 Load resistance (Ω)

(I)

Output power (nW)

15

Voltage (V)

0

−120

Output current (μA)

Voltage (V)

60

Output power (μW)

Voltage (V)

60

101 102 103 104 105 106 Load resistance (Ω)

Fig. 20 Output characteristics of the individual PyENG, PVC, and TPiENG units. (A,C,E) Short-circuit currents. (B,D,F) Voltages measured with 1 GO in parallel. (G–I) Output currents and output powers as a function of external loading resistance. PyENG, pyroelectric nanogenerator; PVC, Polyvinyl chloride; TPiENG, triboelectric-piezoelectric nanogenerator. Reproduced with permission from Zhang KW, Wang SH, Yang Y. A one-structure-based piezo-tribo-pyro-photoelectric effects coupled nanogenerator for simultaneously scavenging mechanical, thermal, and solar energies. Adv Energy Mater 2017;7(6):1601852, Copyright 2017, WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim.

Pyroelectric Materials

756

completely separated, the system returned to equilibrium state, so that both the triboelectric and piezoelectric potential were absent, as demonstrated in Fig. 19(D,iii). After that, due to the vibration, the nylon film moved back to contact with the FEP layer. As a result, reversed current signals were generated, because of the triboelectric effect, as depicted in Fig. 19(D,iv). Once the PZT was fully compressed and the surfaces of nylon and FEP were closely contact, the polarization in the PZT layer was enhanced, thus resulting in the presence of reversed piezoelectric signals. The PyENG unit was characterized by periodically varying the temperature between 231C (room temperature) and 381C, with a TE module. Fig. 20(A) shows short-circuit currents of the PyENG unit, with the peak value to be 480 nA. The voltage oscillated with a peak value of 100 V, as illustrated in Fig. 20(B). Obviously, both the current and voltage responses well followed the variation period of the temperature. Fig. 20(G) shows current peak as a function loading resistance, which led to a maximum output power to be 13 mW at the loading resistance of 200 MO. The PVC was evaluated by using a full spectrum solar simulator. An incomplete recovery was observed for both the current and the voltage, as demonstrated in Fig. 20(C) and (D). A current peak of 890 nA and a voltage peak of 60 V were achieved, which could be attributed to the photoelectric effect coupled with the illumination-induced heat effect. The illumination resulted in an increase in temperature from 24 to 391C, well close to the temperature range for the PyENG unit characterization. Once approaching the equilibrium state, stable platforms of 170 nA and 48 V were achieved, which were closely related to the photoelectric effect without the inclusion of the heat effect. The PVC had a maximum output power of 7 mW, with a loading resistance of 100 MO, as seen in Fig. 20(H). Electrical performance of the TPiENG unit was examined at an air flow speed of about 15 m s 1. The current was an AC output with a peak value of 3.8 mA, whereas the output voltage could not be clearly identified, as illustrated in Fig. 20(E) and (F), respectively. This was because the internal resistance of the TPiENG unit was much lower than those of the PyENG and the PVC ones. As seen in Fig. 20(I), the TPiENG exhibited a maximum output power of 44 nW, with a loading resistance of 8 kO. Coupled nanogenerators, with various combinations of the individual units, such as “PyENG þ PVC,” “PyENG þ TPiENG,” “PVC þ TPiENG,” and “PyENG þ PVC þ TPiENG” were characterized. The output current of the PyENG unit measured had a highest value of 480 nA without the presence of platform, while the output current of the PVC unit displayed a peak value of 890 nA with a platform of 170 nA. In comparison, the combined component of “PyENG þ PVC” exhibited a peak output current of 1 mA, with a platform of 200 nA. Generally, the output voltage a nanogenerator made of two nanogenerators is determined by the one with a lower internal resistance [112]. The “PyENG þ PVC” unit had a voltage peak of 91 V was in between those of the PyENG (100 V) and the PVC (60 V), with the voltage slightly increasing to 50 V. Therefore, it was concluded that the PyENG and PVC units could be utilized either simultaneously or individually. This conclusion was also applicable to other combinations.

2.23.9

Future Directions

Researches of pyroelectric effect have demonstrated its potential in harvesting waste thermal energies. In contrast to TE energy harvesting that needs spatial gradients of temperatures, pyroelectric energy harvesting works when there are temporal temperature variations, i.e., time gradients of temperature. Although creation of spatial gradients rather than time gradients is more probable with wasted heat, pyroelectric energy harvesting produces a much-improved conversion efficiency – the ratio of net harvested energy to the heat drawn from the hot reservoir. Theoretically, it can be comparable to the conversion ratio of the Carnot cycle, and is independent of materials properties. Conversely, the conversion efficiency of TE effect remains largely constrained by the materials properties, as per earlier discussion. In accordance to the laws of thermodynamics, Carnot cycle is least realistic even though it is the most efficient. Comparatively, practical thermal energy harvesting through Olsen cycle has measured the largest energy density amongst other cycles. In consideration of theoretical prediction, an increase in energy density may be probable via by extending the electric field span, thus necessitating pyroelectric materials to possess even higher dielectric breakdown strength. On this point, bulk ceramics and single crystals encounter their own limitation, due to their relatively low breakdown strengths. Therefore, thin films and composites could be more suitable candidates for such special applications. Also, incorporation of pyroelectric effect with other effects, such as piezoelectric, triboelectric, solar cell, supercapacitors and batteries, to develop multiple functional devices with energy generation, storage and self-powering capabilities, could be another direction of research in this field, as demonstrated by the examples discussed at the end of the chapter. Moreover, flexible devices with applications in robots and wearable electronics could be of equivalent significances in the near future. Both topics have started to draw attention from all around the world.

References [1] [2] [3] [4] [5]

Arunachalam VS, Fleischer EL. Harnessing materials for energy – preface. MRS Bull 2008;33:261–3. Arunachalam VS, Fleischer EL. The global energy landscape and materials innovation. MRS Bull 2008;33:264–76. Haertling GH. Ferroelectric ceramics: history and technology. J Am Ceram Soc 1999;82:797–818. Damjanovic D. Ferroelectric, dielectric and piezoelectric properties of ferroelectric thin films and ceramics. Rep Prog Phys 1998;61:1267–324. Kong LB, Zhang TS, Ma J, Boey F. Progress in synthesis of ferroelectric ceramic materials via high-energy mechanochemical technique. Prog Mater Sci 2008;53:207–322.

Pyroelectric Materials

[6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71]

757

Whatmore RW. Pyroelectric devices and materials. Rep Prog Phys 1986;49:1335–86. Sebald G, Lefeuvre E, Guyomar D. Pyroelectric energy conversion: optimization principles. IEEE Trans Ultrason Ferroelectr Freq Control 2008;55:538–51. Bowen CR, Taylor J, LeBoulbar E, et al. Pyroelectric materials and devices for energy harvesting applications. Energy Environ Sci 2014;7:3836–56. Lubomirsky I, Stafsudd O. Invited review article: practical guide for pyroelectric measurements. Rev Sci Instrum 2012;83:051101. Wang SJ, Lu L, Lai MO. Pyroelectric materials for dielectric bolometers. Sci Adv Mater 2011;3:794–810. Kong LB, Li T, Hng HH, et al. Waste energy harvesting: mechanical and thermal energies. Heidelberg: Springer; 2014. Joshi JC, Dawar AL. Pyroelectric materials, their properties and applications. Phys Status Solidi A – Appl Res 1982;70:353–69. Zhong WL. Physics of ferroelectricity. Beijing: Science Press; 1996. Liu ST, Long D. Pyroelectric detectors and materials. Proc IEEE 1978;66:14–26. Kosorotov VF, Kremenchugskij LS, Levash LV, Shchedrina LV. Tertiary pyroelectric effect in lithium-niobate and lithium tantalate crystals. Ferroelectrics 1986;70:27–37. Poprawski R. Investigation of phase-transitions in NH4HSeO4 crystals by pyroelectric method. Ferroelectrics 1981;33:23–4. Poprawski R, Mroz J, Czapla Z. Investigation of ferroelectric phase-transition in NH4HSeO4 (AHSE) crystals by the pyroelectric method. Acta Phys Pol A 1980;57:429–32. Devonshire AF. Theory of ferroelectrics. Adv Phys 1954;3:85–130. Porter SG. A brief guide to pyroelectric detectors. Ferroelectrics 1981;33:193–206. White EAD, Wood JDC, Wood VM. Growth of large area, uniformly doped TGS crystals. J Cryst Growth 1976;32:149–56. Itoh K, Mitsui T. Studies of crystal-structure of triblycine sulfate in connection with its ferroelectric phase-transition. Ferroelectrics 1973;5:235–51. Loiacono GM, Osborne WN, Delfino M, Kostecky G. Single-crystal growth and properties of deuterated triblycine fluoroberyllate. J Crys Growth 1979;46:105–11. Keve ET, Bye KL, Annis AD, Whipps PW. Structural inhibition of ferroelectric switching in triglycine sulfate. 1. Additives. Ferroelectrics 1971;3:39–48. Bhalla AS, Fang CS, Cross LE. Pyroelectric properies of alanine and deuterium substituted TGSP and TGSAS single-crystals. Mater Lett 1985;3:475–7. Bhalla AS, Fang CS, Xi Y, Cross LE. Pyroelectric properties of the alanine and arsenic-doped triblycine sulfae single-crystals. Appl Phys Lett 1983;43:932–4. Felix P, Gamot P, Lacheau P, Raverdy Y. Pyroelectric, dielectric and thermal-properties of TGS, DTGS and TGFB. Ferroelectrics 1978;17:543–51. Shaulov A. Improved figure of merit in obliquely cut pyroelectric crystals. Appl Phys Lett 1981;39:180–1. Shaulov A, Smith WA. Optimum cuts of monoclinic M-crystals for pyroelectric detectors. Ferroelectrics 1983;49:223–8. Yamada T, Kitayama T. Ferroelectric properties of vinylindene fluoride-trifluoroethylene copolymers. J Appl Phys 1981;52:6859–63. Yamazaki H, Ohwaki J, Yamada T, Kitayama T. Temperature-dependence of the pyroelectric response of vinylidene fluoride trifluoroethylene co-polymer and the effect of its poling conditions. Appl Phys Lett 1981;39:772–3. Chung KT, Newman BA, Scheinbeim JI, Pae KD. The pressure and temperature-dependence of piezoelectric and pyroelectric response of poled unoriented phase-I poly(vinylidene fluoride). J Appl Phys 1982;53:6557–62. Fukada E, Furukawa T. Piezoelectricity and ferroelectricity in polyvinylidene fluoride. Ultrasonics 1981;19:31–9. Kumaragurubaran S, Takekawa S, Nakamura M, Kitamura K. Growth of 4-in diameter near-stoichiometric lithium tantalate single crystals. J Cryst Growth 2005;285:88–95. Byer NE, Vanderjagt A. Monolithic pyroelectric arrays. Ferroelectrics 1980;27:11. Jamieson PB, Abrahams SC, Bernstei Jl. Ferroelectric tungsten bronze-type crystal structures. I. Barium strontium niobate Ba0.27Sr0.75Nb2O5.78. J Chem Phys 1968;48:5048–57. Liu ST, Maciolek RB. Rare-earth-modified Sr0.5Ba0.5Nb2O6 ferroelectric-crystals and their applications as infrared detectors. J Electr Mater 1975;4:91–100. Liu ST, Bhalla AS. Some interesting properties of dislocation-free and La-modified Sr0.5Ba0.5Nb2O6. Ferroelectrics 1983;51:47–51. Luff D, Lane R, Brown KR, Marshall Hj. Ferroelectric ceramics with high pyroelectric properties. Trans J Br Ceram Soc 1974;73:251–64. Henson RM, Zeyfang RR, Linhart E. Pyroelectric properties of Pb(Ti, Zr, Fe1/2Ta1/2)O3 polycrystalline solid-solutions. Phys Status Solidi A – Appl Res 1978;46:511–5. Clarke R, Glazer AM, Ainger FW, et al. Phase-transition in lead zironate-titanate and their applications in thermal detectors. Ferroelectrics 1976;11:359–64. Whatmore RW, Bell AJ. Pyroelectric ceramics in the lead zirconate-lead titanate-lead iron niobate system. Ferroelectrics 1981;35:155–60. Whatmore RW. High-performance, conducting pyroelectric ceramics. Ferroelectrics 1983;49:201–10. Grabmaier BC. PbTiO3 grown from melt. Ferroelectrics 1976;13:501–3. Takeuchi H, Jyomura S, Ito Y, Nagatsuma K. Rare-earth subsituted piezoelectric PbTiO3 ceramcis for acustic-wave applications. Ferroelectrics 1983;51:71–80. Ichinose N. Electronic ceramics for sensors. Am Ceram Soc Bull 1985;64:1581–5. Whatmore RW, Osbond PC, Shorrocks NM. Ferroelectric materials for thermal IR detectors. Ferroelectrics 1987;76:351–67. Osbond PC, Whatmore RW. Improvements to pyroelectric ceramics via strontium doping of the lead zirconate-lead iron niobate-lead titanate system. Ferroelectrics 1991;118:93–101. Patel S, Chauhan A, Vaish R. Large pyroelectric figure of merits for Sr-modified Ba0.85Ca0.15Zr0.1Ti0.9O3 ceramics. Solid State Sci 2016;52:10–8. Liu X, Wu D, Chen Z, et al. Ferroelectric, dielectric and pyroelectric properties of Sr and Sn codoped BCZT lead free ceramics. Adv Appl Ceram 2015;114:436–41. Fadnavis SA, Katpatal AG. Nd3 þ doped lead germanate – a good candidate for pyroelectric detector. Ferroelectrics 1998;211:79–87. Ghulghule JR, Katpatal AG. Pyroelectric behaviour of pure and lanthanum-doped lead germanate single crystal. J Phys Chem Solids 1999;60:425–8. Wazalwar AV, Katpatal AG. Structural and pyroelectric properties of Nd3 þ þ K þ -doped ferroelectric lead germanate single crystals. Mater Lett 2002;55:221–9. Yue XF, Mendricks S, Nikolajsen T, et al. Transient enhancement of photorefractive gratings in lead germanate by homogeneous pyroelectric fields. J Opt Soc Am B – Opt Phys 1999;16:389–94. Houlton MR, Jones GR, Robertson DS. Study of growth defects in lead germanate crystals. J Phys D – Appl Phys 1975;8:219. Zwicker WK, Dougherty JP, Delfino M, Ladell J. Growth of high quality lead germanate crystals fo pyroelectric applications. Ferroelectrics 1976;11:347–50. Olsen RB, Bruno DA, Briscoe JM. Pyroelectric conversion cycle of vinylidene fluoride trifluoroethylene copolymer. J Appl Phys 1985;57:5036–42. Ikura M. Conversion of low-grade heat to electricity using pyroelectric copolymer. Ferroelectrics 2002;267:403–8. Kouchachvili L, Ikura M. Improving the efficiency of pyroelectric conversion. Int J Energy Res 2008;32:328–35. Kouchachvili L, Ikura M. Pyroelectric conversion – effects of P(VDF-TrFE) preconditioning on power conversion. J Electrost 2007;65:182–8. Sencadas V, Lanceros-Mendez S, Mano JF. Characterization of poled and non-poled b-PVDF films using thermal analysis techniques. Thermochim Acta 2004;424:201–7. Chu BJ, Zhou X, Ren KL, et al. A dielectric polymer with high electric energy density and fast discharge speed. Science 2006;313:334–6. Olsen RB, Briscoe JM, Bruno DA, Butler WF. A pyroelectric energy converter which employs regeneration. Ferroelectrics 1981;38:975–8. Olsen RB. Ferroelectric conversion of heat to electrical energy – a demonstration. J Energy 1982;6:91–5. Olsen RB, Brown DD. High efficiency direct conversion of heat to electrical energy related pyrorlectric measurements. Ferroelectrics 1982;40:17–27. Olsen RB, Bruno DA, Briscoe JM, Dullea J. Cascaded pyroelectric energy converter. Ferroelectrics 1984;59:205–19. Olsen RB, Bruno DA, Briscoe JM. Pyroelectric conversion cycles. J Appl Phys 1985;58:4709–16. Nguyen H, Navid A, Pilon L. Pyroelectric energy converter using co-polymer P(VDF-TrFE) and Olsen cycle for waste heat energy harvesting. Appl Therm Eng 2010;30:2127–37. Vanderpool D, Yoon JH, Pilon L. Simulations of a prototypical device using pyroelectric materials for harvesting waste heat. Int J Heat Mass Transf 2008;51:5052–62. Navid A, Vanderpool D, Bah A, Pilon L. Towards optimization of a pyroelectric energy converter for harvesting waste heat. Int J Heat Mass Transf 2010;53:4060–70. Carpinlioglu MO, Gundogdu MY. A critical review on pulsatile pipe flow studies directing towards future research topics. Flow Meas Instrum 2001;12:163–74. Ozawa M, Shinoki M, Nagoshi K, Serizawa E. Scaling of heat transfer characteristics in an oscillating flow. J Enhanc Heat Transf 2003;10:275–85.

758

Pyroelectric Materials

[72] Lee FY, Navid A, Pilon L. Pyroelectric waste heat energy harvesting using heat conduction. Appl Therm Eng 2012;37:30–7. [73] Navid A, Pilon L. Pyroelectric energy harvesting using Olsen cycles in purified and porous poly(vinylidene fluoride-trifluoroethylene) P(VDF-TrFE) thin films. Smart Mater Struct 2011;20:025012. [74] Navid A, Lynch CS, Pilon L. Purified and porous poly(vinylidene fluoride-trifluoroethylene) thin films for pyroelectric infrared sensing and energy harvesting. Smart Mater Struct 2010;19:055006. [75] Forster EO, Yamashita H, Mazzetti C, et al. The effect of the electrode-gap on breakdown in liquid dielectrics. IEEE Trans Dielectr Electr Insul 1994;1:440–6. [76] Chen FY, Fang YK, Hsu CY, Chen JR. Time response analysis of a pyroelectric detector. Ferroelectrics 1997;200:257–68. [77] Kao MC, Wang CM, Chen HZ, Lee MS, Chen YC. Thickness-dependent leakage current of (polyvinylidene fluoride/lead titanate) pyroelectric detectors. IEEE Trans Ultrason Ferroelectr Freq Control 2003;50:958–64. [78] Lee FY, Jo HR, Lynch CS, Pilon L. Pyroelectric energy conversion using PLZT ceramics and the ferroelectric-ergodic relaxor phase transition. Smart Mater Struct 2013;22(2):025038. [79] Dausch DE, Haertling GH. The domain switching and structural characteristics of PLZT bulk ceramics and thin films chemically prepared from the same acetate precursor solutions. J Mater Sci 1996;31:3409–17. [80] Vodopivec B, Filipic C, Levstik A, Holc J, Kutnjak Z. E-T phase diagram of the 6.5/65/35 PLZT incipient ferroelectric. J Eur Ceram Soc 2004;24:1561–4. [81] Bobnar V, Kutnjak Z, Pirc R, Levstik A. Electric-field-temperature phase diagram of the relaxor ferroelectric lanthanum-modified lead zirconate titanate. Phys Rev B 1999;60:6420–7. [82] Kajokas A, Matulis A, Banys J, et al. Dielectric dispersion and distribution of the relaxation times of the relaxor PLZT ceramics. Ferroelectrics 2001;257:69–74. [83] Viehland D, Dai XH, Li JF, Xu Z. Effects of quenched disorder on La-modified lead zirconate titanate: long- and short-range ordered structurally incommensurate phases, and glassy polar clusters. J Appl Phys 1998;84:458–71. [84] Bokov AA, Ye ZG. Recent progress in relaxor ferroelectrics with perovskite structure. J Mater Sci 2006;41:31–52. [85] Vodopivec B, Filipic C, Levstik A, et al. Dielectric properties of partially disordered lanthanum-modified lead zirconate titanate relaxor ferroelectrics. Phys Rev B 2004;69:224208. [86] Divya AS, Kumar V. A novel mechanism for relaxor-ferroelectric transition in PLZT(8/65/35). J Am Ceram Soc 2009;92:2029–32. [87] Shah S, Rao MSR. Preparation and dielectric study of high-quality PLZT x/65/35 (x ¼ 6,7,8) ferroelectric ceramics. Appl Phys A – Mater Sci Process 2000;71:65–9. [88] Wang D, Fotinich Y, Carman GP. Influence of temperature on the electromechanical and fatigue behavior of piezoelectric ceramics. J Appl Phys 1998;83:5342–50. [89] Pavel M, Rychetsky I, Petzelt J. Polar clusters in relaxor (Pb,La)(Zr,Ti)O3 revealed by second harmonic generation. J Appl Phys 2001;89:5036–9. [90] Lee FY, Goljahi S, McKinley IM, Lynch CS, Pilon L. Pyroelectric waste heat energy harvesting using relaxor ferroelectric 8/65/35 PLZT and the Olsen cycle. Smart Mater Struct 2012;21:025021. [91] Sebald G, Seveyrat L, Guyomar D, et al. Electrocaloric and pyroelectric properties of 0.75Pb(Mg1/3Nb2/3)O3-0.25PbTiO3 single crystals. J Appl Phys 2006;100. [92] Guyomar D, Pruvost S, Sebald G. Energy harvesting based on FE-FE transition in ferroelectric single crystals. IEEE Trans Ultrason Ferroelectr Freq Control 2008;55:279–85. [93] Khodayari A, Pruvost S, Sebald G, Guyomar D, Mohammadi S. Nonlinear pyroelectric energy harvesting from relaxor single crystals. IEEE Trans Ultrason Ferroelectr Freq Control 2009;56:693–9. [94] Kandilian R, Navid A, Pilon L. The pyroelectric energy harvesting capabilities of PMN-PT near the morphotropic phase boundary. Smart Mater Struct 2011;20. [95] McKinley IM, Kandilian R, Pilon L. Waste heat energy harvesting using the Olsen cycle on 0.945Pb(Zn1/3Nb2/3)O3-0.055PbTiO3 single crystals. Smart Mater Struct 2012;21. [96] Sebald G, Pruvost S, Guyomar D. Energy harvesting based on Ericsson pyroelectric cycles in a relaxor ferroelectric ceramic. Smart Mater Struct 2008;17:015012. [97] Yang Y, Wang SH, Zhang Y, Wang ZL. Pyroelectric nanogenerators for driving wireless sensors. Nano Lett 2012;12:6408–13. [98] Cuadras A, Gasulla M, Ferrari V. Thermal energy harvesting through pyroelectricity. Sens Actuators A – Phys 2010;158:132–9. [99] Zi YL, Lin L, Wang J, et al. Triboelectric-pyroelectric-piezoelectric hybrid cell for high-efficiency energy-harvesting and self-powered sensing. Adv Mater 2015;27:2340–7. [100] Zhang HL, Xie YH, Li XM, et al. Flexible pyroelectric generators for scavenging ambient thermal energy and as self-powered thermosensors. Energy 2016;101:202–10. [101] Gao FX, Li WW, Wang XQ, Fang XD, Ma MM. A self-sustaining pyroelectric nanogenerator driven by water vapor. Nano Energy 2016;22:19–26. [102] Vats G, Kumar A, Ortega N, Bowen CR, Katiyar RS. Giant pyroelectric energy harvesting and a negative electrocaloric effect in multilayered nanostructures. Energy Environ Sci 2016;9:1335–45. [103] Jiang SL, Liu P, Zhang XS, et al. Enhanced pyroelectric properties of porous Ba0.67Sr0.33TiO3 ceramics fabricated with carbon nanotubes. J Alloys Compd 2015;636:93–6. [104] Xie MY, Zabek D, Bowen C, Abdelmageed M, Arafa M. Wind-driven pyroelectric energy harvesting device. Smart Mater Struct 2016;25:125023. [105] Erturun U, Green C, Richeson ML, Mossi K. Experimental analysis of radiation heat-based energy harvesting through pyroelectricity. J Intell Mater Syst Struct 2014;25:1838–49. [106] Zhang KW, Wang SH, Yang Y. A one-structure-based piezo-tribo-pyro-photoelectric effects coupled nanogenerator for simultaneously scavenging mechanical, thermal, and solar energies. Adv Energy Mater 2017;7(6):1601852. [107] Kholkin A, Boiarkine O, Setter N. Transient photocurrents in lead zirconate titanate thin films. Appl Phys Lett 1998;72:130–2. [108] Poosanaas P, Dogan A, Thakoor S, Uchino K. Influence of sample thickness on the performance of photostrictive ceramics. J Appl Phys 1998;84:1508–12. [109] Qin M, Yao K, Liang YC. High efficient photovoltaics in nanoscaled ferroelectric thin films. Appl Phys Lett 2008;93:122904. [110] Zhu G, Zhou YS, Bai P, et al. A shape-adaptive thin-film-based approach for 50% high-efficiency energy generation through micro-grating sliding electrification. Adv Mater 2014;26:3788–96. [111] Yang Y, Zhang HL, Chen J, et al. Single-electrode-based sliding triboelectric nanogenerator for self-powered displacement vector sensor system. ACS Nano 2013;7:7342–51. [112] Jung WS, Kang MG, Moon HG, et al. High output piezo/triboelectric hybrid generator. Sci Rep 2015;5:9309.

Further Reading Batra AK, Aggarwal MD. Pyroelectric materials: infrared detectors, particle accelerators, and energy harvesters. Washington: SPIE; 2013. Gioanola L. Pyroelectric and Amr sensors for intelligent transportation systems. Saarbrücken: Lambert Academic Publishing; 2010. Khodayari A, Mohammadi S, Guyomar D. Pyroelectric energy harvesting: fundamentals and applications. Saarbrücken: VDM Verlag Dr. Müller; 2011. Kong LB, Li T, Hung HH, et al. Waste energy harvesting: mechanical and thermal energy. Heidelberg: Springer; 2014. Lang SB. Sourcebook of pyroelectricity. Boca Raton, FL: CRC Press; 1974.

Relevant Websites https://www.doitpoms.ac.uk/tlplib/pyroelectricity/printall.php DoITPoMS.

Pyroelectric Materials

https://en.wikipedia.org/wiki/Pyroelectricity Wikipedia. https://en.wikipedia.org/wiki/Pyroelectric_crystal Wikipedia. https://www.youtube.com/watch?v=aHp17E_4v_c Youtube. https://www.youtube.com/watch?v=qXLStQQxHzU Youtube. https://www.youtube.com/watch?v=qZuyFUrVJcA Youtube.

759

2.24 Insulation Materials Brock Conley, Cynthia A Cruickshank, and Christopher Baldwin, Carleton University, Ottawa, ON, Canada r 2018 Elsevier Inc. All rights reserved.

2.24.1 2.24.1.1 2.24.2 2.24.2.1 2.24.2.1.1 2.24.2.1.2 2.24.2.1.3 2.24.2.2 2.24.2.3 2.24.2.4 2.24.2.5 2.24.2.6 2.24.3 2.24.3.1 2.24.3.1.1 2.24.3.1.2 2.24.3.1.3 2.24.3.1.4 2.24.3.2 2.24.3.3 2.24.3.4 2.24.3.5 2.24.3.6 2.24.4 2.24.4.1 2.24.4.1.1 2.24.4.1.2 2.24.4.1.3 2.24.4.2 2.24.4.2.1 2.24.4.2.2 2.24.4.2.3 2.24.4.2.4 2.24.4.3 2.24.5 2.24.5.1 2.24.6 2.24.6.1

Introduction Purpose and Motivation Background and Fundamentals Modes of Heat Transfer Conduction Convection Radiation Thermal Bridging Building Envelopes in Building and Voluntary Codes Thermal Performance of Materials and Assemblies Moisture Performance of Materials and Assemblies Airtightness Systems and Applications Common Building Insulation Materials Blanket insulation Foam board based Spray insulation Insulated structural materials Above-Grade Walls Roofs and Attics Floors and Below-Grade Walls Pipes Thermal Storage Analysis and Assessment Material Property Testing Thermal conductivity Heat capacity Moisture permeability Building Envelope Assembly Testing Hot-box Guarded hot-box Evaluation under environmental conditions Computer simulation Exergy for an Insulated Wall at Steady State Results and Discussion Lifecycle Energy of Insulating Materials Case Studies and Examples Designing a High Thermal Resistance Wall With Vacuum Insulation Panels – ECHO – Team Ontario’s Entry to the 2013 Solar Decathlon Test Results Building Envelope Construction Lessons Learned Building Enclosure Simulation Future Directions Aerogels Vacuum Insulation Gas-Filled Panels Closing Remarks

2.24.6.2 2.24.6.3 2.24.6.4 2.24.6.5 2.24.7 2.24.7.1 2.24.7.2 2.24.7.3 2.24.8 References Further Reading Relevant Websites

760

Comprehensive Energy Systems, Volume 2

761 761 762 762 762 765 765 767 768 769 770 771 772 772 772 774 774 774 775 775 775 776 776 777 777 777 777 778 778 778 779 779 780 780 781 781 785 785 786 788 788 788 792 792 792 793 793 794 794 795

doi:10.1016/B978-0-12-809597-3.00252-2

Insulation Materials

Nomenclature

q q00 r R

Heat flow, W Heat flux though a point, W m2 Radius, m Thermal resistance of a material or assembly, m2 K W 1 Temperature, 1C or K Time, h Temperature difference, 1C or K Thermal transmittance, W m 2 K 1 Time rate of work, W Thickness of material, m Expected lifespan of building, years

HDD k L

Area, m2 Cooling degree days, days Coefficient of performance for heating or cooling equipment Emissions, kg Energy, Wh or J Emission intensity, gCO2 kWh 1 Time rate of exergy, W Heating degree days, days Thermal conductivity, m2 K W 1 Thickness of specimen, m

T t DT U _ W Dx Y

Subscripts CO2 cool d eff

Carbon dioxide Cooling equipment Destruction Effective value

embodied Encompassing any consumption of resources outside of operation heat Heating equipment j Surface or evaluation level o Reference level

A CDD COP e E ei _ Ex

2.24.1 2.24.1.1

761

Introduction Purpose and Motivation

Since 1973, the total annual global energy consumption has more than doubled, with annual greenhouse gas emissions increasing 110% in this period [1]. Globally, residential and commercial buildings account for over 20% of the annual global energy consumption, or 25 TWh annually [2]. This is expected to increase at a rate greater than 1.5% per annum between 2010 and 2040 [2]. When looking at individual developed regions, buildings account for almost 40% of energy consumption in the United States, 29% in Canada, and 38% in Europe [2–4]. When examining annual energy consumption in the residential sector, it can be seen that within all regions, space conditioning accounts for the greatest percentage of energy consumption ranging between 49 and 64% of annual consumption within buildings. This value is greatest in cold climates (northern countries, such as Canada), where space heating dominates [2,3]. Based on these statistics, the reduction in required energy for space conditioning within buildings could significantly reduce the total global energy consumption. Over the past couple of decades, the focus on reducing this energy consumption has been on the improved performance of mechanical equipment. This focus has seen the average effectiveness of new natural gas furnaces and boilers increase from below 70% to now approaching 98%, and heat pumps and air conditioning systems obtaining coefficient of performance (COP) in excess of 5 [5]. These devices are approaching the theoretical limits of their performance and as such, little room remains for decreasing energy consumption to meet space heating and cooling demands through improved efficiency of mechanical devices. As a result, building designers and engineers are required to search for new methods to improve the energy performance of buildings. At the same time, many building owners are looking to reduce their energy footprint and reduce energy costs. One of the most efficient and cost-effective methods for reducing energy consumption is improving the building envelope. An occupied building consumes energy to provide a conditioned space that provides indoor comfort for people to live or work. As the mechanical equipment heats or cools the space, the building is constantly exchanging energy with the outdoor environment. Space conditioning loads are predominantly the result of heat loss, or gain through the building envelope, including the exterior walls, roofs, basements, door, and windows among other penetrations, and the transfer of air between the conditioned and exterior conditions. As such, a building envelope with a higher thermal resistance and a reduction of air infiltration into the building will reduce the energy required for space conditioning. Increasing the performance of the building envelope has the potential to have the greatest impact on reducing the energy consumption on the building. That being said, insulation cannot just be continuously added due to a limit imposed on the envelope thickness and a diminished return on investment as more insulation is added. These factors should be weighted and consider during the design of the building envelope. Building insulation can be found in many different forms, including rigid boards, batt or blankets, or loose foam applied to or in cavities. In addition to form, they have many different installation methods, including mechanical fastening, loosely filling a cavity, or spray/blown-in. The different variety of building materials offers different thermal conductivities, and air and moisture permeance. This chapter introduces the fundamentals required when designing a building envelope including fundamental heat and mass transfer. A description of currently available insulating materials is presented, including the best applications for each type with sample assemblies. The evaluation methods for material properties and assembly performance is presented.

762

Insulation Materials

Emerging, high performance insulating materials, which could form the basis of future walls and their potential applications within building envelopes, are also discussed. Finally, a detailed case study is presented for the design process required when a highly insulated wall assembly is integrated into the Team Ontario Solar Decathlon house.

2.24.2

Background and Fundamentals

While the envelope can significantly improve the energy efficiency and indoor comfort of a building, there are factors that need to be considered before integration. Thermal bridges, moisture accumulation, air leakage, and long-term structural problems can occur if poorly planned or designed. These properties related to overall performance and together contribute to a great or poor building envelope. During the design of the building envelope, the thermal and moisture performance and the airtightness requirements have to be balanced and integrated into the remaining parts of the building design, including the architectural properties and mechanical equipment within the building.

2.24.2.1

Modes of Heat Transfer

The purpose of insulation is to impede the movement of heat in order to maintain the enclosed conditions. The modes of heat transfer that exist are conduction, convection, and radiation. These modes may act alone but more commonly are acting concurrently. Conduction is defined as the heat flow from one point to another through a medium (solid, liquid, or gas). Convection is defined as heat flow by means of a fluid medium flowing past the heat source (medium is in motion, liquid, or gas). Radiation is heat exchange by electromagnetic waves through a gaseous or vacuum medium between two surfaces. These are the heat transfer modes and are constantly present in buildings.

2.24.2.1.1

Conduction

Heat conduction is the transfer of thermal energy between “neighbors” due to a temperature gradient. In solids, it is due to the combination of vibrations of the molecules in a lattice and the energy transport by free electrons. In gases and liquids, conduction is due to the collisions and diffusion of the molecules during their random motion. Conduction occurs in a building when heat moves from the warm inside surface of the building enclosure to the cold outside surface. Heat can travel through a single material, such as concrete, or perhaps multiple materials in contact, like the sheathing faced with batt insulation. A temperature gradient is formed due to conduction across the building enclosure. The gradient is dependent on the conductivity, k, of the material. A lower conductivity yields a steeper temperature gradient, and a higher conductivity yields a shallower gradient. Therefore, a higher conductivity material provides heat to a simpler path to flow. The amount of heat transferred through the material, q, is equal to the temperature difference across the two surfaces, DT, multiplied by the conductivity, k, and surface area, A, and divided by the material thickness, Dx, as shown in Eq. (1), and illustrated in Fig. 1. In a building, this heat energy lost through the enclosure needs to be replenished by heat addition to maintain the interior conditions. Another way to denote a material’s resistance to heat flow is its thermal resistance, R. q ¼ ðk
A
DTÞ=Dx

ð1Þ

In many real systems, the conduction path flows through more than one material, with varying thermal conductivities and thicknesses. The best method to determine the effective thermal resistance of the insulating system is by creating a resistance network, illustrated in Fig. 2. The figure shows three materials that vary in thermal conductivity and thickness. Also shown are the temperature gradients through the assembly. The resistance network uses the thermal resistance value of each material, which is the Δx

Temperature gradient

A q

k

Thot

Tcold

Fig. 1 One-dimensional steady state conduction and temperature gradient through a solid material.

Insulation Materials

Δx1

Δx2

763

Δx3

A Tgradient

q

k1

k2

k3

Tcold

Thot

Tcold

Thot R1

R2

R3

Fig. 2 One-dimensional steady state conduction and temperature gradient through multiple layers of varying conductivity.

Δx1

Δx2

A

Δx3

k3

q

k1

k2

k4

Tcold

Thot

R3

Thot R1

Tcold R4

R2 Fig. 3 Two-dimensional conduction through multiple layers of varying conductivity.

inverse of the thermal conductivity per unit thickness, and analyzes the system as a steady state heat flow when no heat generation is present between the wall boundaries. The effective thermal resistance can be calculated using Eq. (2) when the material changes in the direction parallel to heat flow.


 Dx Dx Dx þ þ RSIeff ¼ k 1 k 2 k 3 RSIeff ¼ R1 þ R2 þ R3

ð2Þ When materials vary in the plane perpendicular to the heat flow, as in Fig. 3, the effective resistance is the inverse of the effective thermal resistance and is equal to the sum of the inverse of the thermal resistances, shown in Eq. (3). The resistance network shows the thermal resistances R2 and R3 in parallel, while others are in series. An example of parallel materials can be a thermal bridge, a variation in geometry or a change in material.

764

Insulation Materials

Wood stud Sheathing Exterior insulation

Cavity insulation Gypsum board

A q

Tinterior

Texterior

Fig. 4 Plan view of a typical constructed building envelope in cold dry climates.

Building envelopes contain multiple different materials to address the structural and thermal requirements, represented in Fig. 4. To augment the performance of the walls, insulation can be added between the structural members of the envelope thereby creating a parallel resistance network. The structural members, typically wood or steel, have a much high thermal conductivity value when compared to the insulation, which is commonly batt or fill insulation. For example, the RSI-value of the batt insulation is 2.5 m2 K W 1 (14 h 1F ft2 BTU 1) and a 2  4 wood stud has a RSI-value of three times lower and covers 6% of the wall area. The effective thermal resistance can be calculated using a modified version of Eq. (3), where only the parallel resistance network is considered. The effective RSI-value for that section would have an RSI-value of 1.9 m2 K W 1. It shows that a small area of low thermal resistance or high thermal conductivity in a system can have a large impact on the effective performance. In contrast, there may be a continuous layer of exterior insulation outboard of the sheathing. The continuous layer provides an RSI-value to the envelope that can be added to the total value, such as a series connection. Continuing the example from before, if a foam board with an RSI-value of 2.0 m2 K W 1 is added continuously along the outside, thermal resistance at the wood stud improves as much as the thermal resistance between the studs. This would provide an effective RSI-value of 3.9 m2 K W 1 for the assembly, excluding the other material layers that do not provide significant insulation. In conduction, the addition of insulators to the boundary of the system slows the rate heat flows. For the previous example, additional insulation to the outside or within the cavity will reduce the amount of heat transferred. The building envelope example above shows that in order to maintain the interior conditions of the building, additional insulation is added to the outside or within the cavity. By slowing the conduction through the envelope, the interior conditions require less energy input from mechanical equipment. Energy savings through proper and sufficient insulating materials within the building envelope can be achieved. Another example would be the use of a refrigerated container, such as a freezer or shipping container. The refrigerated interior needs to maintain an air temperature below 01C to avoid contamination, but the surrounding air temperature can differ by upwards of 251C. If the boundary or refrigerator wall insulation is insufficient, the unit consumes a large amount of energy. When the insulation is increased, energy use is reduced proportionally. It highlights the importance of proper insulation and its effectiveness in limiting conduction. RSIeff ¼ R1 þ

R2 R3 þ R4 R2 þ R3

ð3Þ

Insulating materials are not limited to covering flat and square surfaces. They may also be used to cover a cylindrical wall in some cases or more commonly to insulate cylindrical objects including pipes, tubes, or domestic hot water tanks. These systems typically contain a fluid that has a significantly different temperature compared to the surroundings. Additionally, cold pipes are typically insulated to prevent condensation from forming, by reducing the air temperature, and consequently the dew point, at the surface of the pipe. Radial coordinates are better at analyzing these situations, and cause a variation in calculating the RSI-value. By beginning with the steady axisymmetric conduction expression for cylindrical coordinates, the RSI-value of the cylindrical wall per unit length is represented by Eq. (4),
 1 r2 ln RSIcylindrical wall ¼ ð4Þ r1 2pk where r2 and r1 are the outer and inner radius in meter. For compounding layers of insulation or material, the principles from the series layers wall apply. The cylindrical wall diagram and resistance network are shown in Fig. 5. The performance of the cylindrical insulation has unique effects as it pertains to convection. Unlike square walls, as insulation is added to the cylindrical walls, the diameter and surface area exposed to the environment increases. Since heat exchange

Insulation Materials

765

Thot

r1 r2

Tcold

Thot

Tcold Ri

Rwall

Ro

Fig. 5 Temperature profile with resistance diagram of an insulated cylindrical wall.

by convection is a function of the surface area, a situation may occur where insulation is added and heat exchange is increased. This situation occurs when the thickness of insulation is below the critical radius of insulation. The critical radius is defined as the point where the energy exchange through conduction and convection is no longer increasing. After initially insulting the cylindrical wall, there is an increase in the rate of heat transfer because the increase in convection is larger than the reduction in conduction caused by the insulation. From the series resistance network, the rate of heat transfer is represented in Eq. (5), where the radius r1 and r2 are the inner and outer radiuses, k is the thermal conductivity of the insulation, h is the convective coefficient, and Ti and T1 are the temperatures at the interior surface of the wall and the environment, respectively. qcylindrical wall ¼

Ti T1   1 þ hð2pr 2Þ

r2 1 2pk ln r1

ð5Þ

To find the critical radius, the radius where the rate of heat transfer is maximized needs to be evaluated. This is achieved by differentiating Eq. (5) with respect to the exterior radius and the relationship between thermal conductivity and convection coefficient, shown in Eq. (6). rcritical ¼

k h

ð6Þ

This critical radius is important in many energy systems since reducing the energy exchange is paramount with the addition of insulation. In these applications, if the improper thickness of insulation is used, this can reduce the performance or efficiency as a whole.

2.24.2.1.2

Convection

Convection is the heat exchange though the bulk motion of fluid when a temperature difference exists. In buildings, the heat loss from the interior air to the interior wall surface would occur through convection. Similar to the thermal conductivity property, k, convection contains a convection coefficient, hc, that represents the amount of heat exchange between the surface or fluid and surrounding. When conduction and convection are present, the system can be analyzed with a resistance network. In Fig. 6, a temperature difference between the air temperature, TAir, and the surface temperature, THot, exists. An RSI-value can be created for the convection component by taking the inverse of the convection coefficient. The resistance network can be analyzed with Ohms Law to find the overall RSI-value. There are two distinct types of convective heat: natural and forced convection. Which heat transfer type is applicable is dependent on the bulk fluid motion. Natural convection is caused by naturally occurring fluid motion because of the difference in temperature of a surface and the fluid, and results in a lower convective coefficient. An example would be inside a dwelling heated with a radiator where the airflow is not aided by a fan. Another example is air losing heat to the cooler exterior walls, shown schematically in Fig. 7. Forced convection has a flow rate that is generated by an external source, such as a pump or fan, in order to increase the amount of heat transfer. An example of forced convection is high ventilation rates used in air conditioning equipment present in a dwelling or strong winds that interact with the exterior of the building envelope. Convection is occurring on both sides of the building envelope and it is an effective mechanism of exchanging energy. While convective loops and air leakage within the cavities are detrimental to thermal and overall performance, convection is a useful means to limit any overheating the home may encounter. For a building that is constantly in contact with sunlight, the exterior wall is exposed to a large amount of solar radiation. The solar radiation causes a significant increase to the temperature profile within the envelope, potentially reaching 601C. The hot exterior wall rejects its energy to the cooler outdoor air by forced convection. This is also an example of two heat transfer modes that are occurring simultaneously.

2.24.2.1.3

Radiation

Radiation, previously mentioned as solar radiation, is a heat transfer mechanism through electromagnetic waves. To radiate heat, there needs to be a temperature difference and a line of sight connection between the surfaces. While radiation is present between any surfaces that agree with those two requirements, solar radiation is the most significant factor with respect to buildings. Between the hot sky during sunny days and cold sky during winter nights, a large temperature difference between the building enclosure

766

Insulation Materials

Δx

hc

A

q Tair

k

Tcold

Thot Thot

Tair

Tcold R1

Rc

Fig. 6 Resistance network incorporating conductive and convective heat transfer.

hc A

Tinterior

Texterior

Fig. 7 Building envelope with natural convection at the interior surface.

and surrounding is present. The surroundings and building envelope are constantly exchanging energy through radiation. Radiation is evaluated with the temperature difference of the surface and surroundings to the fourth order, multiplied by the Stefan–Boltzmann constant, s, and the surface area, shown in Eq. (7).   4 4 ð7Þ Tsurrounding q ¼ s  A Tsurface A building envelope is constantly exchanging energy with the interior and exterior environment, and in some cases is exposed to all three heat transfer modes, shown in Fig. 8. In this case, heat transfer by conduction occurs through all insulating and structural materials of the envelope, heat transfer by convection occurs at the interior and exterior surfaces through air movement and heat transfer by radiation occurs between the outside surface and surroundings. R2–R6 represent the conduction through the structural and insulating materials. R1 and R7 are the convective effects at the surface of the building envelope interacting with the interior and exterior air temperatures. Finally, Rrad represents the radiation exchange between the exterior surface and the outdoors. It was mentioned in the previous section that radiation can overheat the building and temperatures in the envelope. While the solar radiation is a large heat source, it also provides natural lighting for the building and its occupants. Occupants commonly feel more comfortable in areas with a larger amount of natural lighting. Windows are the most common penetration in the building envelope used to provide the interior space with natural sunlight. However, the sunlight provides substantial heat directly to the

Insulation Materials

767

Tsurround

h1 h2

A

Tair,ext

Tair,int

Tinterior

Texterior

Tsurr Tair,int

R3 R1

Rrad R5

R2

R6

R4

Tair,ext R7

Fig. 8 Building envelope resistance network including the three heat transfer modes with the environment.

space. Excess glazing can provide the occupant more visual comfort but could raise the potential for overheating and thermal comfort. These factors need to be addressed during the design of the insulating and envelope system. The introduction of radiant barriers to the market has provided designers a way to control the radiation. These barriers are transparent metallic coatings that improve the emissivity of a material from around 90% (typical building materials) to as low as 3%. These materials are adhered to any insulation, roof membrane, or glazing to increase the effective thermal performance. The three heat transfer modes (conduction, convection, and radiation) are constantly interacting with the building envelope. Conduction can be reduced by adding insulating materials, either a higher performing material or a greater material thickness. Convection can be limited through reducing the airflow rates at the interior setting or the addition of baffles or siding at the exterior. The previous examples were used to highlight how resistance networks can be used to evaluate the effective RSI-value of insulating systems. While convection and radiation are present during many heat transfer problems, insulating materials are best used in reducing the amount of conduction the system experiences.

2.24.2.2

Thermal Bridging

A thermal bridge is an undesired path of low thermal resistance, through a material or layer of high thermal resistance. Typically, these thermal bridges provide an easy path for heat transfer through the enclosure of a building, increasing the amount of heat loss to the exterior and decreasing the effective thermal resistance of the building envelope. Thermal bridges within buildings are predominantly the results of two modes. The first is through structural members within the building envelope and the second is through the joints between individual pieces of insulation. When designing a building, the first step is to design and develop the structure, ensuring the building is structurally sound and will stand-up to the loads. These maybe environmental loads (wind, seismic, snow etc.) or the loads placed on the structure through the building application (live loads). These loads are predominately met using structural members and assemblies constructed using wood, concrete, or metal. Each of these materials has a thermal conductivity significantly greater than the insulating materials. When integrated into the insulating layers of a building, the high conductivity materials create a thermal bridge that reduce the effective thermal resistance of that layer. In addition to structural members, other common thermal bridges in buildings are the joint in which two insulating materials are connected or butted up to each other. In the perfect, ideal situation, the joint between the two materials would provide a continuous, homogenous insulating material, however, in building applications, the perfect joint is rarely possible, and the small air gap between materials provides an easier path for heat transfer. The thinner the insulating material, the greater the impact of the thermal bridges between each piece of insulation on the overall performance of the layer. This is particularly important with emerging insulating materials, such as vacuum insulation panels (VIPs), as each panel is relatively small when compared to a typical sheet in board insulation, and the difference in thermal resistances between the joint and the center of the insulating panel is significant. Thermal bridges are present in many building components and designs, although in almost every occurrence, good design can reduce or minimize thermal bridges. In residential applications, the most common, and in many cases, most significant thermal bridge is through the wood structure, whether it be at the studs in the walls, headers, footers, or the space between floors.

768

Insulation Materials

Most common construction practices utilize the space between the structural studs to insulate the house by installing batt insulation within the cavities. Although this is an effective use of the space and does not require additional wall thickness for insulation, it also means that the wood studs act as part of the insulating layer. As a result, if a wall constructed with 2  6 lumber and the studs spaced at 400 mm spacing, a continuous fiberglass batt insulating layer would provide an effective thermal resistance of 3.5 K m2 W 1, while this decreases to 3.25 K m2 W 1, or a 7% decrease in effective insulating values. This decrease is much more substantial if steel structural members are used. This becomes significant, as many new building codes require building envelopes to meet an effective thermal resistance. There are a number of methods that can be employed to reduce or eliminate this thermal bridge. The most common when constructing residential structures is to include a continuous layer of insulation on the exterior of the structural layer, while maintaining the insulation within the cavities between studs. The most common materials used for this exterior insulation are board insulations, including expanded and extruded polystyrene (XPS), polyisocyanurate, or high-density rock wool. This provides a continuous layer of insulation over-top of the structural members, reducing the percent decrease in thermal performance of the wall. Another method for reducing thermal bridging because of the structural members is to use alternative framing methods and layouts. Instead of using one thick stud, two smaller studs can be utilized, with each being offset from each other and with a thermal break between the two layers. This technique is commonly referred to as a double wall construction, and can increase the effective thermal resistance of the wall when employing similar thicknesses. Although this method provides greater thermal resistances through reduction in thermal bridges, it is far more complex to construct, and as a result can significantly increase construction costs, and has seen limited widespread implementation in practice. Although these methods have the potential to reduce thermal bridges, it does not eliminate the thermal bridge created by the structural member. To reduce any thermal bridging, one of two strategies must be employed. The first is to place all of the insulation on the outside of the structural members, removing the highly conductive material from the insulating layer. This strategy is more commonly utilized in commercial buildings, where a steel or concrete structure is erected, and then all of the insulation is attached to the outside of the building. The downside to this technique is thicker wall assemblies are required, as the insulating and structural layers are sequential as opposed to a combined single layer. The second method that can be employed is using structural materials that can also insulate. Currently, most structural materials provide little insulating value, as typically insulating materials are lightweight with little to no rigidity and cannot be used as a structural member. As such, one of the few materials currently available that provide the required building structure, while also insulating are structurally insulated panels (SIPs), which use an insulating foam core with structural facings (most commonly orientated strand board) affixed on each side of the foam. These panels are suitable for residential and low-rise applications, and can provide much higher effective thermal resistances for the same thickness, as a result, of eliminating the thermal bridge within the building envelope. In addition to thermal bridging through structural components and gaps in the insulation, many common building components also creates thermal bridges within the building envelope. Windows and doors produce some of the most significant thermal bridges within a building. Most window and doorframes are constructed from wood, aluminum, or polyvinyl chloride (PVC) plastics. These all create excellent conduits for heat to escape and are why window and doorframes are some of the most significant penetrations in any building. To reduce the thermal bridge created by these components, new materials are being integrated into these components. Window frames are being constructed out of insulating materials, including fiberglass and PVC plastic with embedded foam insulation. Aluminum spacers are being replaced with foam and rubber spacers, increasing the thermal performance of the window unit. Although significant advancements have been made in residential window units, almost all commercial and high-rise glazing units continue to use aluminum frames, and is one of the reasons high-rise wall assemblies have significantly lower effective thermal resistances when compared to low-rise residential units. Windows and doors are not the only architectural building features where significant thermal bridges exist. For example, balcony slabs on high-rise buildings are concrete cutouts that extend past the exterior of the building and act as a thermal bridge. The building itself is insulated, but the balcony slab is uninsulated. During cold windy days, the slab can become very cold, drawing heat out of the building through the opening required to attach the balcony to the main floor slab. Effectively, the balcony slabs become heat transfer fins that continuously removes heat from the building. Significant research is being conducted on different methods of creating a thermal break between the buildings. These include using a strip of dense insulation at the interface of the building and the balcony. Rebar and structural supports continue to pass through the dense insulation, but the thermal bridge is significantly reduced when compared to continuous concrete. This cannot only reduce the amount of energy loss through the thermal bridge, but also increase the occupant comfort by reducing cold spots within the building in close proximity that can occur near the balcony connection. When designing an insulating system for a building, for example, it is important to not only consider the material and the quantity of insulation. The design of the insulating system is of equal importance, as it is critical that a continuous layer of insulation is present. Insulation is only as good as its weakest point, and even a building envelope that has a very high nominal insulating value can actually perform very poorly if significant thermal bridging is present.

2.24.2.3

Building Envelopes in Building and Voluntary Codes

When designing a building envelope, it is necessary to examine the local building codes. In recent years, during the development of new codes, a greater emphasis has been placed on the energy efficiency of buildings through the reduction of energy consumption

Insulation Materials

769

in the building. The changes in code requirements are happening on a global scale, with many governing agencies implementing some type of energy clause within their building codes. Although each jurisdiction is responsible for implementing their own codes, many use global codes as the basis for their own codes. Two such codes are the International Energy Conservation Code (IECC), developed by the International Code Council, and ASHRAE 90.1 [6,7]. Both of these codes prescribe a minimum effective thermal resistance based on the climate zone of the building. A climate zone is a region with similar meteorological conditions, and in the case of prescribe insulating levels, is predominantly based on the heating and or cooling degree days (CDD) for a given location. Based on these zones, the IECC 2012, the minimum required thermal resistance range from 2.8 m2 K W 1 (15.8 h 1F ft2 BTU 1) for warm temperate climates to 4.9 m2 K W 1 (27.7 h 1F ft2 BTU 1) for cold, heating dominated climates. ASHRAE defines similar climate zones, and prescribes between 2.3 m2 K W 1 (13 h 1F ft2 BTU 1) and 5.0 m2 K W 1 (28.4 h 1F ft2 BTU 1) for the same conditions. Individual countries have developed or adopted their own energy codes. A worldwide leader in energy efficiency, Sweden was one of the first countries that implanted a strict energy code in the early 1970s, and currently require exterior walls to have a minimum thermal resistance of between 7 m2 K W 1 (37.7 h 1F ft2 BTU 1) and 8 m2 K W 1 (45.3 h 1F ft2 BTU 1) depending on the orientation of the wall [8]. Another northern country with its own energy code is Canada, which requires exterior wall insulating values of between 2.8 m2 K W 1 (15.9 h 1F ft2 BTU 1) and 3.1 m2 K W 1 (17.6 h 1F ft2 BTU 1) [9]. In Canada, each province is responsible for its own energy code, some of which are more stringent than the Canadian Energy Code. One of these is Ontario, which has increased the minimum required thermal resistance of exterior walls for new buildings in Ontario, requiring a minimum effective thermal resistance of 4.2 m2 K W 1 (23.8 h 1F ft2 BTU 1) and 4.8 m2 K W 1 (27.1 h 1F ft2 BTU 1) depending on the other installed energy saving features in the house. Japan has also implemented its own code with a wide range of prescribed values, as there is a wide range of climatic conditions within Japan. These range from 0.6 m2 K W 1 (3.4 h 1F ft2 BTU 1) in the southern areas to 2.6 m2 K W 1 (14.7 h 1F ft2 BTU 1) in the northern areas for residential buildings. For commercial buildings, an effective thermal resistance is not prescribed, but rather a maximum energy use intensity of the building. The building envelope must be designed to meet the required level [8]. The codes prescribe the minimum required performance of building envelopes. However, they do not exclude the use of additional insulation. Worldwide some national and international voluntary standards exist that can certify a building to a higher energy performance standard. To meet many of these voluntary standards, the thermal resistance of the building envelope must exceed, and in some cases greatly exceed the prescribed local codes. Most of these standards are performance-based codes, meaning a building must meet a certain energy consumption target, but how that target is met is up to the designer. As such, a wide range of thermal resistances for building envelopes can be seen within a building and achieve the same voluntary code. In addition to the thermal resistance of the building envelope, other factors must be considered within many of these codes, including the materials selected (limiting the amount of embodied energy), the airtightness of the envelope, the window area to wall area ratio and the airtightness of the building. These voluntary standards that may be selected by the designer or owner of the building may include Passive House [10], Leadership in Energy and Environmental Design (LEED) [11], Green Globes [12], among many others. In addition to these programs run by different organizations and governments, other initiatives are available including designing netzero or net-zero ready buildings [13]. Through both local building codes and voluntary standards, an overall effective thermal resistance of energy performance of the building is provided, but how that value code is met is up to the designer. The decisions include the type of insulation selected, the thickness and placement of each layer of insulation and the detailing that is done to prevent air and moisture infiltration. Within many of these building codes, provisions for air and vapor barriers are included, and it is of vital importance to understand the interaction between insulating layers and these barriers when designing a complete wall assembly. Although thermal performance is important, it is not the only factor in designing the envelope. These items and how they impact the wall performance are further discussed in the coming sections.

2.24.2.4

Thermal Performance of Materials and Assemblies

Typically, to reduce the energy transfer through an enclosure, increasing the effective thermal resistance or effective RSI-value of the boundary will achieve those goals. In building envelopes or enclosures, a combination of materials with varying thermal conductivity properties can cause problems in evaluating the overall performance. The existence of thermal bridges in building envelopes create easier paths for heat to escape. Thermal bridges are locations or points where the thermal conductivity is higher, allowing the heat transfer to become two-dimensional (2D) as opposed to one-dimensional (1D). In building envelopes, these may be as small as mechanical fasteners used to build and construct the assembly or as large as the wood or metal frame that have a significantly higher thermal conductivity when compared to the insulation used inside the cavity. When comparing building envelope designs, there are two separate methods used to evaluate the thermal performance, the nominal RSI-value, and the effective RSI-value. The nominal RSI-value is the RSI-value of the insulation at the center of the wall. For example, if a nominal RSI-3.5 m2 K W 1 (R-19.8 h 1F ft2 BTU 1) wall is required, the equivalent value of insulation is required in the cavity of the wall. The second method is to consider the effective RSI-value of the assembly. The effective RSI-value encompasses not only the cavity but also the entire wall assembly components together. For example, the stud and cavity layer RSI-value is averaged using the framing factor, which is the amount of frame in the wall area, or coverage areas for a single RSIvalue lower than the standard cavity insulation. Then, RSI-value of the remaining components are added to the averaged value to obtain the effective value. For example, if the insulation in the cavity were RSI-5 m2 K W 1 with 23% of the wall area being 38 mm

770

Insulation Materials

Roofs and attics

Building walls

Doors and windows Basement and foundation

Fig. 9 Illustrated heat losses in a residential building.

by 140 mm (2  6 lumber) studs at 406 mm (16 in.) on-center (OC) framing, the averaged value would be approximately RSI-3.4 m2 K W 1, significantly lower than the nominal value. The building codes have begun to require the effective RSI-value recently as opposed to the nominal value. The effective value encompasses a value that is more likely to be reached since it encompasses the thermal bridging that will take place in standard construction. Heat loss in a building is tied directly to the effective thermal resistance of the envelope. The heat losses or gains are caused when the outdoor temperature is different from the indoor temperature. In colder climates, an increased amount of heat will be lost to the outdoor environment and a larger load to the heating systems. However, increasing the effective thermal resistance of the building envelope will reduce the heat loss. In essence, the heat loss of a building is contributed to by three different components: heat through enclosure components (such as walls, doors, and windows), heat through ventilation, and heat through air infiltration of building seals and openings. The summation of these three transfers will be providing a net heat loss or gain through the boundary. The result will vary based on the indoor and outdoor temperatures, the effective thermal resistance of the building, and the air ventilation and infiltration rate. A building will experience heat losses through all the surfaces in contact with the environment, such as the walls, roofs, and basements, as illustrated in Fig. 9. By increasing the amount or changing the type of insulating material used for the envelope, the thermal performance can be improved and heat losses to the outside reduced. Improving the insulation of the building envelope includes benefits of lower energy consumption, reduced capacity requirements for heating and cooling equipment, and improved indoor comfort. As previously described, the total heat loss of the building is based on the envelope, ventilation, and infiltration. By reducing the losses through the envelope or infiltration, the required capacity of the heating and cooling equipment is reducing because the building will need less to offset any losses the building experiences. Through the lower heating, ventilation, and air conditioning (HVAC) requirements and lower losses, the conditions in the spaces will remain more consistent and reduced airflow rates provide greater indoor comfort for the occupants.

2.24.2.5

Moisture Performance of Materials and Assemblies

A common concern when designing building envelopes is how they able to wet and dry throughout the year. When using woodframed constructions, moisture trapped within the wall cavity could cause health, indoor comfort problems, and component deterioration. Moisture can be introduced to the building envelope forced through the environment, by rainfall, for example, or by the humid air diffusing through the envelope toward a dry air environment, which may be interior or exterior. The inevitable introduction of moisture to the building envelope has spurred the development and analysis into where, when, and how vapor barriers, retarders, and permeable materials should be installed [14]. The installation of these vapor-specific components are dependent and sensitive to many items, such as climatic conditions, cladding types, and structural components, which offer a wide variety of possibilities. Suggestions and recommendations are based on only using vapor barriers, retarders, or permeable materials when necessary and not introducing two barriers to the building enclosure to ensure the envelope will have drying potential inward and outward. The vapor permeability of building materials uses the unit “perm,” where a perm is 1 ng of water vapor per second, per square meter, per Pascal of pressure difference across the material (1 perm ¼ ng (s m2 Pa) 1). While there is a difference in definitions, a commonality between the American Society of Heating, Refrigeration, and Air-Conditioning Engineers (ASHRAE), the Canadian General Standards Board, and the International Building Code are to denote vapor barriers as materials with 0.1 perm or less, vapor retarders or vapor semi-impermeable materials are 1.0 perm or less, and vapor permeable materials have greater than 10 perms [15]. When materials have 1.0 and 10 perm is where the definition of vapor permeable materials will vary, with some

Insulation Materials

771

denoting the materials as either Class 2 or Class 3 vapor retarders depending on their permeability, while others call materials vapor semipermeable. The vapor barriers were introduced to retard the migration of moisture through the assembly [15]. In addition to vapor barriers and retarders, air barriers and rain screens exist in building enclosures to limit the movement of air and moisture. While vapor barriers are not intended to restrict the movement of air and air barriers are not intended to restrict to movement of moisture, there are materials that do perform both in the building enclosure. Air barriers will restrict the movement of the convective moisture transport through the building envelope. Air barriers are built to be continuous, durable, and robust such that they are not compromised during construction and are added to eliminate the flow of air through the assembly. This, in turn, will improve energy efficiency and indoor comfort levels of the building [16]. Rain screens are installed with the enclosure in order to control the rain penetration on the exterior wall. They also reduce and minimize the moisture load seen behind the cladding [17]. This has caused the building designers to integrate overhangs, windowsills, and impermeable surfaces, such as glass or vinyl, as the exterior cladding of buildings, unlike materials that absorb water like brick, wood, or stucco. Secondly, behind the cladding, a cavity or secondary boundary with greater water resistance is introduced to ventilate or dissipate the water back to the exterior [17]. Air barriers and rain screens are not considered vapor barriers or retarders, but are effective and common practices in reducing the moisture introduced into or transport through the enclosure. The climate regions are an important factor in deciding the cladding, barriers, and structural components used for the building enclosure. The hygrothermal loads and effects will be very different between a hot and humid climate and a cold or very cold climate. For example, in the cold climates wetting from the interior will occur during the heating season by air movement, or vapor diffusion caused by the higher levels of humidity on the inside compared to the outside. Control measures that can be implemented is to build an airtight enclosures and add vapor retarders or barriers to restrict the air and vapor movement through the envelope, and design the envelope to promote drying to the exterior [15]. In contrast to cold climates, hot and humid areas are prone to high levels of interior humidity during cooling periods as well as higher humidity levels on the exterior. The buildings must facilitate dehumidification of the indoor air, unlike cold climates [15]. Another difference is vapor retarders added toward the exterior to limit the vapor diffusion from the outside and to promote drying toward the interior. In new construction, initial moisture problems are nearly nonexistent since the building materials have been pretreated or have not been exposed to hazardous conditions. Initially, the moisture concerns during new construction are based on the long-term environmental conditions, properly sealing the building and how to prevent or retard moisture after it is introduced [18]. One method to alleviate these concerns is to add vapor barriers to the inside of the sheathing, such that moisture cannot permeate through the conditioned interior space. In existing buildings, however, moisture problems may have developed since initial construction. These issues may be visible through water stains, condensation or damaged finishes, or possibly condensation in cavities not visible without deconstructing layers of the envelope. If the problem sources are not found and resolved during renovations, it is possible that the deterioration and damage will continue to occur. Apart from existing issues, when updating an existing envelope it is possible that introducing new vapor barriers or added insulation on the exterior can move the dew point inward, and cause unwanted condensation or water entrapment. Since the envelope is exposed to the environment prior to renovation, there may be an accumulation or existence of moisture before the additional insulation or barriers are installed.

2.24.2.6

Airtightness

While heat may be lost through the walls by conduction, convection, and radiation, which is based the thermal properties of the construction materials, there may also be air leakage through the envelope. The insulating materials in the envelope can be impermeable to air, however, during construction, a number of joints, seams, and penetrations, including doors and windows, are introduced to the envelope [19]. Air barrier systems are composed of multiple air impermeable materials to reduce or restrict the movement of air through the building envelope. The components that comprise the system will be sealed with gaskets, weather-stripping, or sealant meant to create a continuous seal around the building, however, in reality, many unsealed sections exist in a building. The airtightness evaluation will differ for residential and commercial buildings. A blower door test is commonly performed during an energy audit of a building either to determine retrofit designs or to prove that the construction complies with voluntary standards. A blower door test uses a large fan unit, mounted on the exterior doorframe, to extract the air from the building or building zone and creates a lower air pressure compared to the exterior. This forces the higher-pressure outdoor air to leak through the building cracks and openings and the natural air infiltration rate can be determined by measuring the airflow for an extended period [20]. At the same time, tracer gasses can be used to find small cracks or leaks that may exist in the home, in a similar fashion to how infrared images are used to find localized thermal bridges in envelope components. Typically, tracer gasses previously included common gasses, such as helium, hydrogen, oxygen, and carbon monoxide but recently, professionals have been using nitrous oxide, sulfur hexafluoride, or various hydrocarbons [20]. In energy retrofits or energy-conscious homes, the air changes per hour (ACH) are targeted as a means to reduce the heat loss and ventilation requirements, since it prevents the warm or conditioned air inside the dwelling from being lost to the outside. According to the Canadian Mortgage and Housing Corporation, homes built between the 1950s and 1980s average an airtightness above 6.0 ACH at 50 Pa [21] and during retrofitted applications, and they require a minimum of 30% improvement on airtightness during a building envelope retrofit.

772

Insulation Materials

While airtightness can improve the overall thermal efficiency of your building, it may also cause unwanted indoor comfort effects. Increasing the building tightness can cause occupants to sense a lack of fresh air in the indoor environment due to the reduced air changes and ventilation. These effects can be mitigated through the addition of heat recovery ventilators (HRVs) or other equipment to bring fresh air into the building. HRVs provide the home with fresh air from the outside, and precondition the incoming air with the exhaust air during heating and cooling season. As homes become more airtight, pollutants may not be able to exit and fresh air may not be able to enter, and HRVs are able to solve both of these problems [22]. Some of the HRV systems are capable of transferring moisture and moderating the indoor humidity, and thereby improving the resistance against excess moisture through the building envelope.

2.24.3

Systems and Applications

As the building maintains the interior conditioned space at regulated temperature, it will be constantly losing or gaining energy from the outside depending on the heating or cooling season. During cooling season, the heat gains are typically found through the roofs, windows, and wall of the dwelling and during heating season, heat losses are found through the floors and basements in addition to the roofs, walls, and windows. In an effort to save on the space heating required maintaining the comfort levels, additional insulation could be added to these sections of the building.

2.24.3.1

Common Building Insulation Materials

Many factors contribute to the selection, which are insulation material to use in the buildings. Buildings are insulated with a variety of materials that use different types of fastening methods, air and vapor permeability, but above all else their different thermal insulating properties, specifically RSI-values (R-values) or thermal conductivities. The conventional building insulations can be found in loose-fill wool, a blanket roll, rigid board, sprayed, or built into structural materials. The benefits and advantages can include the way they are installed, their RSI-value per unit thickness, where they can be added, and their effects on thermal bridging, moisture transfer and air leakage. A summary of each insulation material is included in Table 1. The table lists, for each type of insulation, the range of RSI-values, common areas in the building where they would be found, methods to install the material, and the advantages of insulating the building with that material.

2.24.3.1.1

Blanket insulation

Blanket insulation is composed of flexible fibers based on fiberglass, mineral wool, plastic, or natural fibers. According to the U.S. Department of Energy [23], blanket insulation is the most common insulation in buildings and is widely available in the form of batts or blankets. They are available in widths and thicknesses that are suitable for standard spacing of stud and joists present in buildings. The insulation is typically installed within unfinished walls, and floor and ceiling joists where they can be fitted inside very easily since the insulations are made to the standard spacing outlined in building codes. However, if the spacing and insulation width do not align, the batt or roll can be simply hand-cut or altered on-site without consequence. Fiberglass blanket insulation is composed of fine glass fibers and provides an RSI-value that will vary based on the density of the material. For example, the RSI-value of the lower density fiberglass batt is RSI-1.94 versus RSI-2.64 for a high-density fiberglass batt for a 102 mm (400 ) deep cavity [24]. Mineral wool insulation is composed of either basalt and diabase, commonly named rock wool, or slag from a blast furnace, commonly referred to as slag wool. The materials that make up mineral wool are typically recycled materials from industrial processes. Another benefit to mineral wool is that it does not require additional materials or chemicals to make it fire resistant [25]. Similar to mineral wool, recycled materials are common in manufacturing blanket and wool insulations. Cellulose insulation is composed of recycled paper products that are fiberized to small pieces and tightly packed into building cavities [25]. The insulation also does not typically require a moisture barrier in the cavity and during manufacturing chemicals may be added to the composition to ensure the material is fire or insect retardant. Plastic fiber insulation is mainly composed of recycled plastic products, as opposed to recycled paper used in cellulose insulation. After being made into insulation batts, the plastic fibers need to be treated with a fire retardant such that it does not burn quickly, however, the insulation is prone to melting when exposed to high temperatures or flames [25]. While some insulation may be difficult to handle because of the irritations, they may cause on your skin, plastic fiber insulations avoid these issues, but they may be difficult to cut with standard tools compared to other batt insulations. Finally, these batts can be made of recycled natural materials, such as cotton, sheep wool, and straw among others to create the insulation fibers [25]. While they may provide an RSI-value nearly equivalent or below that of the previously mentioned fiberglass batts, they do offer some nonthermal benefits. Batts that use cotton insulation can be installed without any respiratory or skin personal protective equipment and can be manufactured using the waste from clothing factories and therefore require minimal energy to produce. Sheep wool insulation provides a comparable RSI-value and has the ability to absorb large amounts of moisture, however, it may degrade due to the chemicals used to resist fire, insects, and mold. Straw bale insulation offers effective sound-absorbing properties, but the expected RSI-value per unit thickness is much smaller compared to other available batt insulations [25].

Table 1

Summary of insulation materials and typical found properties, installation techniques, and advantages

Type

RSI-value per meter m K W 1 (R-value per inch) BTU h 1 1F 1 ft 2 in 1 [25,28]

Commonly installed locations in the building

Blanket: batts and rolls

21–32 (2.9–4.7)

Standard available sizes fitted for between studs, joists, and other cavities

• • •

Easy to install Suitable for standard stud and joists Cost effective

21–26 (2.8–3.7)

Floors Ceilings Unfinished walls Foundation walls Existing walls Unfinished walls Unfinished attics Difficult to reach cavities Floors Ceilings Unfinished walls Foundation walls

•

Loose-fill and blown-in

• • • • • • • • • • • •

•

Blown-in using specialized equipment Performed by professional

•

Adding insulation to finished, irregular shapes or around obstructions Retrofits or renovations

• • • • • • •

Existing walls Unfinished walls Unfinished attics Difficult to reach cavities Unfinished walls Ceilings Roofs

•

Rigid foam board

Spray foam

Structural insulated panels

25–46 (3.6–6.7)

21–42 (3.0–6.0)

25–34.7 (3.6–5)

Installation or fastening method

• • • •

• • • •

3.5 Effective (20 effective)

• • •

Foundation walls Exterior walls Flooring

• • •

•

Indoor and outdoor applications Covered by gypsum board for fire code on interior applications Covered by weatherproof facing for exterior installation Pressurized spray-foamed product Performed by professional

•

Relatively high RSI-value per unit thickness

•

Adding insulation to finished, irregular shapes, or around obstructions

Installed as a load bearing and structural component Pieces connect together for rapid construction Possible need of crane or heaving lifting equipment Poured during construction Performed by professionals Reinforced with steel rods

•

Insulation built into structural component High thermal resistance

•

• •

Insulation built into structural component High thermal resistance

Insulation Materials

Insulated concrete forms

Advantages

773

774

Insulation Materials

The overall thermal performance of the blanket insulation will vary depending on the individual insulation chosen and the depth of the cavity that they are fitted. However, on a per unit thickness basis, a standard blanket will provide between RSI 20–26 m2 K W 1 per meter (R2.9-3.8 h 1F ft2 BTU 1 in. 1), while a high performance blanket, involving a higher density material, may provide an RSI-value 32 m2 K W 1 m 1 (R-4.7 h 1F ft2 BTU 1 in. 1) [24]. Another valuable way to install many of the blanket insulations in through loose-fill or blown-in by professionals. Unlike the batts, loose-fill insulation does not require any handling or cutting of the material, while being of similar composition. The fibers, and in some cases foams as well, are used to fill wall cavities or enclosed spaces, such as attics and headers. They have the ability to conform to the space without disturbing the structural elements or finishes in the building. Due to their ease of installation and ability to conform around spaces, they are used significantly during retrofit or renovation activities.

2.24.3.1.2

Foam board based

Foam board insulation is provided in a rigid panel of insulation and offers an ability to easily fasten extra insulation to many facets of the building. The product is typically composed of either polystyrene or polyurethane, and is capable of providing an RSI-value of 27.6–45.0 m2 K W 1 m 1 (R4-6.5 h 1F ft2 BTU 1 in. 1) [24]. The insulation can be added all over a building, from the subfloor, exterior wall, interior wall, foundation wall, attic, and roof. The types of foam board that exist in expanded polystyrene (EPS), XPS, dense glass, and polyisocyanurate. These rigid boards are comprised of polystyrene foam or closed-cell plastic beads as the main material component. There are different manufacturing processes and final products available when polystyrene is used. XPS is manufactured through a continuous extrusion process and creates the homogenous closed-cell rigid board [25]. EPS is manufactured by using heat and pressure to fuse spherical beads of polystyrene together [11]. The performance differences between the two products are limited to the moisture permeability and RSIvalues. The condensed-beaded structure of the EPS allows moisture to be entrapped in the material and can degrade the performance and the structural integrity of the foam. This is unlike XPS, which is more comparable to a closed-cell systems since it does not allow moisture to be absorbed or condense inside the foam. Since there is a large number of companies manufacturing these products, each product should be evaluated individually, but RSI-values can range from RSI-26.3 to RSI-34.7 m2 K W 1 m 1 (R-3.8 to R-5.0 h 1F ft2 BTU 1 in. 1) [11]. Polyisocyanurate insulation is a type of closed-cell plastic that contains low conductivity gases within its cells. This type of insulation is able to provide a higher thermal performance compared to EPS and XPS but due to the low conductivity gas escaping the pores over time, it will have a lower RSI-value at the end of its life [25]. However, it remains possible that the end of life RSI-value remains above the other foam board insulations, around RSI-48 m2 K W 1 m 1 (R-7 h 1F ft2 BTU 1 in. 1). Manufacturers have used different metallic and plastic foils in order to reduce the degradation that the polyisocyanurate experiences with promising results to an improved RSI-value of 60 m2 K W 1 m 1 (R-8.7 h 1F ft2 BTU 1 in. 1) [25].

2.24.3.1.3

Spray insulation

Spray insulations are liquid foam or fibrous materials and provide a versatile method of insulating difficult shapes, across obstructions, while providing an above average RSI-value. They are blown-in by specialists or contractors typically in cavities, such as an attic, wall cavity, or a location with an abundance of obstructions. Since they are liquid foam prior to curing, they typically create an effective air barrier since they seal all the holes and seams. Sprayed foam insulation can also offer around two times the RSI-value per meter thickness compared to traditional batt and blanket insulations used for the same applications [24]. The moisture transport through the insulations can differ based on their foam-in-place type. Closed-cell foams offer a highdensity of closed cells filled with a gas that enables the foam to expand and seal all the seams in the cavity. The closed-cell foam also offers a strong resistance to heat, providing a high RSI-value per meter, and is stronger against moisture and air leakage compared to the second foam type, open-cell foams. Open-cell foams contain a lower density of cells and are filled with air as oppose to another expanding agent. The lack of expanding agent causes the foam to have a lower RSI-value per meter, and introduces the risk of the foam absorbing moisture. Therefore, open-cell foams are not recommended for below ground applications [25].

2.24.3.1.4

Insulated structural materials

Materials that offer structural and insulation properties required for the building enclosures are labeled as insulated structural materials can be found in the foundation and exterior walls of a building. Homes and buildings are commonly built with concrete and recently, the introduction of concrete with built-in insulation. Known as concrete block insulation and insulating concrete forms (ICFs), they offer a significant thermal improvement for basement and foundation walls compared to standard construction. ICFs have two added layers of foam insulation to provide the benefits of plain concrete as well as improved thermal performance and insulation values, while maintaining the appearance of a standard building [25]. Polystyrene foam interlocked and tied together are commonly used for the insulating core material offer an effective RSI-value of 3.5 m2 K W 1 (R-20 h 1F ft2 BTU 1) and will also be reinforced with steel rods [26]. While the ICFs are more costly than pouring concrete into standard forms, they do provide an insulated wall with a vapor barrier in a single construction step and also keeps the foundation from quickly freezing or thawing, reducing the risks of cracks, or leaks forming [26]. Another type of insulated structural material is structural insulated panels (SIP). The panels are prefabricated and beneficial when designing rapidly deployable buildings that only require connecting the panels together to build the home enclosure. Using these, prefabricated materials allow the home to be built much quicker than normal saving builder time and money [27]. The panels are usually built with foam-based insulation inside, like the previously discussed materials, for example, however novel

Insulation Materials

775

materials, like VIPs, have been implemented as well. The SIPs are offered in various sizes and may be so large that they require a crane in order to assemble the structure. They have been designed with residential buildings and commercial building kits. The panels must undergo precise and quality manufacturing to maintain their integrity and ensure their performance over the lifetime [26,27]. They need to have smooth surfaces and edges such that they interconnect and prevent gaps on the construction site. The connections create an airtight seal along the seams. The thin-profile insulation and airtightness created by the SIPs provide an excellent enclosure to maintain indoor comfort in an energy efficient manner [26]. The enclosure becomes so airtight when using SIPs that they require tightly controlled mechanical ventilation, such as HRVs, to ensure sufficient fresh air is being introduced to the indoor environment. Other areas of concern with the SIPs are the infiltration of animals and insects into the panels themselves, through small gaps at the connections or penetrating the panels themselves.

2.24.3.2

Above-Grade Walls

When adding insulation to the wall of a building, it can be installed on the interior or exterior of the building frame. In some cases, installing insulation on the inside or outside are better based on the exterior finish, moisture prevention benefits or other construction, and renovation-related questions. The walls of the building offer flexibility in the type of insulation used since they have both cavities and structural components that allow fastening. Inside the cavities, it is possible to add nearly any batt, blanket, blown-in, sprayed, and rigid foam board insulations. Building walls may be built using different types of construction, such as solid walls, concrete blocks, and framed walls. The solid walls are usually built of brick, stone, or concrete and do not contain a cavity, but could still contain a small drainage hole or plane to removed water from the wall [29]. Concrete block walls will have hollow cores, many thermal bridges and attention needs to be taken to ensure they are air-sealed between the interior finish and the block wall. Finally, frame walls can be built with using wood or steel framing pieces and contain cavities that can be insulated. In addition to the cavities, framed walls can be insulated from the interior or the exterior. When renovating and reinsulating walls of building, it may be performed through the interior by rebuilding existing wall, to build an additional interior wall or to renovate the exterior wall [29]. Rebuilding the existing wall will result in completely removing the wallboard, insulation in the cavity and any barriers prior to adding extra insulation. The cavity can be refilled with batt insulation or spray-foamed by a professional prior to reapplying the wallboard. Alternatively, the rigid foam board can be fastened to the exposed frame studs after filling the cavity with insulation to improve the effective RSI-value. Otherwise, building a framed wall on the interior can be done on any wall construction. However, it will cost valuable interior floor space. The newframed wall can be built at any distance from the old wall, as long as the cavities are completely filled and required barriers remain intact [29]. Proper air sealing along all the edges are required for both types of renovation methods. Unlike the interior, when adding insulation to the exterior of the walls during renovations, it is suggested to remove the cladding prior to adding insulation [29]. Although it is possible to apply additional insulation, such as foam board or blanket insulation, to the existing cladding if the existing moisture conditions allow the enclosure to properly dry after installation. Renovating walls is possible and can be a means to improve the energy efficiency of the building. Adding insulation to the building exterior for either new construction or renovation creates a moisture barrier from water infiltration into the envelope. Installing additional exterior insulation during renovations is less intrusive to the occupants when compared to adding insulation to the interior [30]. However, in some situation-specific instances where there are existing moisture concerns or other renovations inside the home, adding insulation to the interior is a better method to increase the thermal performance of the building.

2.24.3.3

Roofs and Attics

Roofs and attics are typically spaces at the top of the building and are exposed to the same conditions as the exterior walls with increased radiation or sun exposure. These spaces are typically uninhabited, and therefore, are a convenient space to add insulation, during either initial construction or renovation. Common materials used in the attic are loose-fill or batt insulation materials since their thickness does not cause many problems. While the recommended amount of insulation varies by climate and exterior conditions, the U.S. Department of Energy suggest that if the existing insulation in the attic is less than RSI-5.3 (R-30), the equivalent of 279 mm (1100 ) of fiberglass or 203 mm (800 ) of cellulose approximately, additional insulation could yield energy efficiency benefits [30]. In order to properly insulate the attic and maintain the benefits of the increased RSI-value, ensure the space does not contain any unwanted infiltration or leaks between the conditioned interior and the attic. Commonly used insulation in attics is batt, blanket, or loose-fill, but they are not limited to those types. It is also possible to use foam board or spray foam and obtain comparable energy efficiency, however, in most instances, it is simpler to handle and install in what is typically a confined space [31]. It is also important to maintain the ventilation in the attic as it reduces heat buildup in summer months, prolongs the lifetime of the roof, acts as the second line of defense against moisture penetration and reduces the chance of ice dams forming [31].

2.24.3.4

Floors and Below-Grade Walls

Floors may act as a barrier between conditioned and unconditioned spaces in commercial and residential buildings. A common example of this situation in a residential building is an unconditioned garage below a conditioned space inside the

776

Insulation Materials

dwelling. During the winter months, the temperature inside the garage will approach exterior conditions and without the insulation or proper air sealing, there will be heat loss as well as potential pollutants that may diffuse into the building interior. Basements are another source of heat loss during the wintertime due to the cooling of the surrounding ground. Approximately 20% of a standard home’s total heat loss can be attributed to the basement consisting of a largely uninsulated surface area below grade [30]. Some insulation may only be performed during the initial construction of the building through the use of ICFs, however, insulation can be still be added through renovation. The foundation walls that make up the basement will be made up of either poured concrete, concrete block or brick none of which offer much in the way of insulating the building. The below-grade rooms can be made more comfortable and the risk of moisture problems can be limited through the use of insulation on the basement or foundation walls. The basement has the ability to be insulated on the interior or exterior, each offering advantages and disadvantages. In most situations, insulating on the interior is the most practical and economical approach and can be achieved by adding rigid foam board or batt insulation with a wood-framed wall and gypsum board. This represents the easiest way to incorporate insulation in a renovation project, however, must not be performed when moisture issues exist. Adding the insulation to the foundation wall will cause the moisture to condense on the surface and rot the insulated timber wall [32]. In order to insulate the exterior of the foundation, an excavation must occur around the building prior to waterproofing and insulating the foundation. The benefits include moisture problems in the foundation are more visually apparent from the outside after excavating around the foundation; there are no disruptions inside the building and freeze-thaw stresses on the foundation are eliminated after the insulation is completed. However, some issues, in addition to the potentially high costs of excavating around the building, are the potential difficulties associated with excavating during certain periods of the year [32].

2.24.3.5

Pipes

Pipes are used to transport fluids, such as refrigerant, steam, or conditioned water at various temperatures and to inject or reject energy to a system. In some cases, such as district heating or building loops, the supply fluid may travel long distances before it reaches its application. Due to heat loss from the pipe to its environment, the supply temperature will decrease as the length of the pipe increases. Therefore, to meet a target temperature downstream, it is necessary to increase the level of insulation on the pipe or increase the supply temperature. The latter can be achieved using a boiler. Pipe insulation also reduces condensation that may form along the outer surface of a cold pipe by reducing the heat transfer between the surface of the pipe and its surroundings. For example, for indoor conditions of 211C and 40% relative humidity, condensation will form on the pipe when the surface temperature reaches around 71C. The heat exchanged at the pipe surface due to condensation will be transferred to the pipe fluid and increase its temperature. Pipes are common place in buildings to transport energy to meet the heating and cooling demands of the building. They may also be used in district heating and cooling systems, where the length of pipes is usually larger. Insulating these pipes will reduce the load on the heating or cooling system, as well as mitigate the unwanted energy exchange during transmission.

2.24.3.6

Thermal Storage

Thermal storage is a means to store energy that is not needed during the time period that it is produced. For example, solar energy can be stored during the day when it is available but used later in the evening when it is needed by occupants. Common thermal storage mediums include liquid, solid, air, or a mixture. An application of thermal storage can involve the use of a heat pump and cold and hot thermal storage, on the source and load side, respectively, to reduce peak demand on the electrical grid during the day. In this case, the thermal storages would be charged overnight and used during the day when electricity is most expensive (e.g., if electricity is priced using time-of-use billing). Insulating these systems is as important as efficiently harvesting the energy. Throughout the day or season, while the thermal storage is not either in use or charging, it will be exchanging energy with the environment. If the storage does not have sufficient continuous insulation or if a thermal bridge exists, the thermal energy could be lost faster than it is gained. Diurnal storage tanks used for domestic hot water systems, for example, contain water or a liquid solution. Since these storage tanks are charged and discharged daily, the effect of the overall heat transfer coefficient will be less due to the short time period in comparison to larger (more surface area) tanks and longer periods. Heat transfer characteristics are dependent on the surface and environment temperatures as well as the surface area in contact with the environment. The storage temperature will vary in temperature depending on the type of storage that exists, its application, temperature set-point and its location. Storages, for example, can be installed in a conditioned space (e.g., inside a building), a nonconditioned space (e.g., exposed to the outdoor weather) or buried in the ground. The importance of insulating storage is obvious when considering a seasonal thermal storage system, which has an annual charging and discharging cycle and is required to remain at the set-point longer than a diurnal thermal storage intended for daily consumption. If the rate of heat transfers are the same between the daily and seasonal storages, the seasonal storage would exchange more total energy over time that the daily storage.

Insulation Materials

2.24.4

777

Analysis and Assessment

The main purpose of building envelopes and insulating materials are to maintain the interior conditions related by their heat and mass transfer properties. Before the building materials can be implemented into building assemblies, it is important to understand their material properties, including the thermal conductivity, moisture permeability, air penetration, and heat capacity. These properties are important to determine the best use of each material, as well as to determine the interaction between materials. In addition, material properties are vitally important in the development of heat and mass transfer computer models of proposed wall assemblies. Once the material properties for each component of a wall assembly are determined, the complete wall assembly must be assessed as a unit. This is important to better understand the interaction between materials and how they operate as an assembly. It is important to measure the total effective thermal resistance of the assembly, as this will take into account any thermal weaknesses, including thermal bridges (usually the result of structural members) and penetrations. The thermal assessment of insulating materials and their components in the envelope are RSI-value for insulation or U-value, typically found in for windows. These properties can be found experimentally by using many different types of apparatuses, evaluating the materials at steady state or environmental conditions. This section will outline the different testing methods to determine both the fundamental properties of different insulations and structural materials, as well as measuring the heat and mass transfer within complete assemblies.

2.24.4.1

Material Property Testing

In order to understand how heat and mass transfer through the building envelope, it is necessary to understand the heat and mass transfer characteristics of each material. The main fundamental properties of the materials required are the thermal conductivity, heat capacitance, and moisture permeability. Other properties could be required for specific applications, including absorptivity and transitivity, however, are not usually applicable to opaque wall sections.

2.24.4.1.1

Thermal conductivity

Thermal conductivity (k) measures the materials resistance to heat transfer through it and is inversely proportional to the thermal resistance of the material. A common apparatus used to determine the thermal conductivity is a guarded hot plate, shown schematically in Fig. 10. A guarded hot plate is composed of at least two plates, one with a cold temperature and one with a hot temperature. Thermal conductivity is typically measured using a hot surface and cold surface at constant temperatures, and the material sample between the two surfaces. The total heat transfer is measured, either using a heat flux meter or by measuring the total heat input into the hot plate. Once the test reaches steady state conditions, which is noted as when the heat flux through the test material stops changing and remains constant for a prolonged period of time, readings can begin. Measurements over an extended period of time should be taken and averaged, reducing the error caused by minor fluctuations in the measurements. An uncertainty analysis should be conducted to determine the accuracy of the measurements. American Society for Testing Materials (ASTM) Standard C177-13 provides a detailed design guideline as to how to construct a guarded hot plate [33]. By knowing the temperature of each plate and the total heat flux, the thermal conductivity can be calculated using Eq. (8): k¼

ðqLÞ ADT

ð8Þ

where q is the total heat transfer through the specimen, L is the thickness of the specimen (i.e., the distance between the hot and cold plates), A is the area in which the heat flux is being measured, and DT is the temperature difference between the hot and cold plates.

2.24.4.1.2

Heat capacity

When designing a building envelope, the heat capacity of a material relates the amount of energy that is available to be stored as the wall heats up and cools down. The heat capacity also allows the amount of energy required for the wall to reach the steady state to be determined. This property is typically measured one of two ways. The first is with a sample at a constant temperature; it is placed against a hot surface, while the remaining sides are well insulated. Using temperature sensors on each surface, the amount of energy it takes for the sample to reach the temperature of the plate can be measured. When the heat capacity is coupled with the known rise in temperature of the sample, the heat capacitance can be determined. Cold plate Specimen Guard plate Specimen Cold plate Fig. 10 Schematic of a general guarded hot plate used to measure the thermal conductivity.

Metering plate Guard plate Edge insulation

778

Insulation Materials

The second method uses a calorimeter. In this method, a sample is heated up to a uniform temperature. It is then placed in a sealed, well-insulated container with a fluid (typically water) of a known volume, temperature, and heat capacitance. The material is left within the fluid until the sample and fluid reach a constant temperature. Using the final temperature of the sample and fluid, the heat capacitance of the sample can be calculated.

2.24.4.1.3

Moisture permeability

Moisture permeability is the material’s resistance to the water vapor diffusion through a unit of surface area. This can be experimentally determined using one of two methods, as described in ASTM standard E96M-15 [34]. The two methods are the desiccant method and the water method. In both methods, samples can be up to 32 mm (1.25 in.) in thickness. In the desiccant method, a desiccant material is first placed within a watertight container. The sample is then placed across the top of the container and sealed to the container, which is typically done using wax or a similar product. The sealed unit is then weighed using an analytical balance and then placed into a test chamber at a constant temperature and high relative humidity. Over time, the unit is periodically weighed, to determine the amount of moisture that has passed through the sample and accumulated in the desiccant material. Using the weight of the moisture that has passed through the sample and the time that has passed, the moisture permeability can be calculated, and is typically quantified as g h 1 m2. The wetted method uses a similar process; however, water is placed in the container instead of a desiccant. The container is then sealed and instead of placing it into a chamber with high humidity, it is placed into a very dry chamber. Over time, the container is weighed to determine the amount of water that has exited the container and using these data, the moisture transmission can be determined.

2.24.4.2

Building Envelope Assembly Testing

After measuring the thermal or moisture performance of individual building envelope components, the enclosure assemblies should be evaluated to measure how the components perform together. This usually entails building a representative sample of the wall assembly and testing it within a laboratory or test facility or installing instrumentation within an existing wall. Better results are typically obtained in a laboratory or test facility, and as such is the preferred method for new wall assemblies, however, evaluating existing walls may be necessary, for example, historical or irreproducible assemblies. The different test methods are outlined in detailed as follows.

2.24.4.2.1

Hot-box

The main function of a hot-box apparatus is to introduce a temperature difference on either side of an assembly to measure the heat flow and effective thermal resistance. The hot-box will regulate the air conditions on either side of the wall to replicate indoor and outdoor conditions. The apparatus tightly controls the temperature, limits the air velocity and sometimes the relative humidity on the indoor and outdoor side in order to reach steady state conditions and obtain a consistent heat flow through the assembly. In Fig. 11, a general schematic of a hot-box apparatus is shown, where the indoor and exterior chambers are labeled with the assembly specimen located between the chambers. The interior heating unit energy consumption will be measured and will be representative of the energy lost through the assembly. However, the hot-box will lose energy to the surrounding environment when there is an air temperature difference between the environment and interior chamber. Due to the losses that may accrue over a test period, the hot-box must be calibrated and using an assembly or material with a known RSI-value to quantify the heat through the chamber walls over the test period [35].

Specimen Baffle

Baffle Fans

Heating unit

Cooling unit

Interior chamber Exterior chamber Fans

Fig. 11 Schematic of a generic hot-box.

Surround panel

Insulation Materials 2.24.4.2.2

779

Guarded hot-box

A guarded hot-box is similar in functionality to the hot-box. However, an additional chamber is built to eliminate the heat losses through the interior chamber walls. A third section, called the guard chamber shown in Fig. 12, is added and surrounds the interior hot chamber. The guarded air temperature is tightly controlled to mimic the air temperature in the interior chamber. This helps to control the heat through the chamber walls. This heat loss can be neglected if the chamber walls are sufficiently insulated and the temperature difference between the chambers remain minimal, typically 0.11C. The heat flux of the metering chamber walls is measured during the evaluation period, but the purpose of the guarded chamber is to have an insignificant amount of heat through the metering chamber walls. By using the guarded hot-box apparatus, the heat flow is forced from the interior surface to the exterior surface of the specimen and the surrounding environment does not affect the metering chamber and heat flow if the apparatus is properly designed. In both the guarded hot-box and hot-box apparatuses, the effective insulating value of an assembly or individual material can be determined from steady state conditions. ASTM has developed a standard for operation, conditions, and requirements for these apparatuses [35]. In order to ascertain that the wall specimen is experiencing steady state conditions, multiple test conditions need to be met for five consecutive test periods. These conditions include:

• • • •

the the the the

average average average average

specimen surface temperature in the metering chamber does not vary by more than 70.251C; specimen surface temperature in the climate chamber does not vary by more than 70.251C; temperature within the air curtain in the metering chamber does not vary by greater than 70.251C; and energy input to the metering chamber does not vary by more than 71%.

Afterward, the effective RSI-value, R, of the assembly can be calculated using Eq. (9), R¼

DT ðtAÞ E

ð9Þ

where DT is the temperature difference between the interior and exterior surface of the assembly in 1C, E in the heat input to the metering chamber in Wh, t is the test period in hours, and A is the metering area or the area of the wall assembly specimen in contact with the metering chamber in m2. In addition to the effective RSI-value of an assembly, the steady state conditions, and thermal bridge effects can be determined by the hot-box and guarded hot-box apparatuses. In a laboratory setting, by using embedded instrumentation, such as temperature and heat flux sensors, temperature profiles and gradients within the assembly can be experimentally determined. The gradients for conventional building insulation materials will be through the thickness of the wall, however, using the novel materials mentioned later in the chapter, there can be significant temperature gradients at the interface of certain layers. The embedded sensors are also able to evaluate the RSI-value or thermal resistance at a point in the assembly (e.g., thermal bridge, center of stud, etc.). By measuring the heat flux and temperature difference across an interface or material, the RSI-value, R, can be calculated using Eq. (10), where q00 is the measured heat flux in W m 2 and DT is the temperature difference in 1C. R ¼ DT=q00

2.24.4.2.3

ð10Þ

Evaluation under environmental conditions

Unlike the hot-box and guarded hot-box apparatuses, in this test method the exterior of the building envelope is evaluated, while being exposed to the outdoor environmental conditions. This is known as in situ testing. The objective of in situ testing is to expose the wall sections to the full seasonal cycle using the unregulated outdoor conditions to evaluate the heat and moisture transfer through the assembly. The exposure to all weather and environmental conditions cause water infiltration and changes in

Interior chamber Baffle Heating unit (metered) Heating unit

Specimen Baffle Fans Cooling unit

Guarded chamber Exterior chamber Fans

Fig. 12 Schematic of a generic guarded hot-box.

Surround panel

780

Insulation Materials

temperature and humidity that may not be accurately replicated in a laboratory setting. The in situ specimens are typically installed for a test period of at least a year, in order to monitor the wetting and drying of the building envelope and associated insulating materials. During the testing period, the temperature and heat flux is measured at the interface of every layer for each cross-section to determine the nominal thermal resistance at that specific location. After collecting sufficient data determine by regression formulas outlined through ASTM or ISO testing standards, the weighted average effective thermal resistance of the building envelope can be calculated based on the coverage area of each cross-section [36]. Multiple specimens can be installed in the facilities and be independently tested and evaluated accurately since, unlike the hot-boxes, the heat addition inside the facility does not need to be monitored or is not a variable in the thermal evaluation. In addition to testing for moisture transfer within the wall assemblies, in situ testing can also be used to measure the thermal performance of the wall assembly. To acquire the necessary information, a temperature sensor is installed on the interior and exterior surfaces along with heat flux sensors. Measurements are taken at regular periods, for an extended period of time, and the average temperature difference between the interior and exterior surfaces as well as the average heat flux are used to calculate the thermal resistance at the point of measurement. The difference between this method and using a guarded hot-box is that only the thermal resistance at the point of measurement is calculated, as opposed to the overall effective thermal resistance of the entire assembly. As such, the thermal resistance at each unique cross-section must be measured. Using the calculated value of each cross-section, and taking an area weighted average of the complete wall assembly, the overall effective thermal resistance can be calculated. For most in situ building envelope apparatuses, there are multiple openings to install specimens since their test period are at least year. These opening may be oriented in different directions (e.g., North, South, East, and West, etc.) or perhaps all along the same face of a building in order to properly compare and evaluate the envelope assemblies. Since the apparatus will be set to a single locale for long testing periods, the calibration of computer models and simulations for an envelope design in one locale can aid in evaluating the same design in a different climate by using readily available weather data.

2.24.4.2.4

Computer simulation

Heat and moisture transfer software have been developed to study building envelope assemblies at steady state conditions or transient yearlong conditions. Certain programs are adept at performing hygrothermal simulations of the moisture transport and drying potential through the building envelope over a yearlong period, which when performed in situ would require a year worth of acquisition. Other programs offering the ability to simulate 2D, or three-dimensional (3D) heat transfer conditions are able to provide the effective RSI-value of the assembly and the isothermal profiles through the cross-section in order to compare different designs prior to the laboratory or in situ testing, which require more resources in terms of time, equipment, and materials. Two-dimensional heat transfer programs will simulate the heat transfer through the many building envelope components, typically using a finite element heat transfer analysis. The finite element method continuously solves the heat transfer conduction equation through each element and evaluates the total heat flow through the 2D cross-section. The program is capable of determining the effective RSI-value of the assemblies as well as the temperature gradients along the interfaces, which identify thermal bridge effects, and potential moisture, and condensation concerns. The hygrothermal simulations are performed at dynamic conditions, while coupling heat and moisture transfer through the building components. The simulation utilizes the exterior climate conditions from the chosen locale as the exterior boundary conditions and illustrates the moisture movement that is relevant in multilayered components, such as roofs, walls, or balconies, for example. After the simulation is completed, the interior building conditions and comfort levels can be determined, as well as whether the individual building envelope component interfaces can withstand the moisture damage accrued over time.

2.24.4.3

Exergy for an Insulated Wall at Steady State

Exergy (Ex) is a concept based on the Second Law of Thermodynamics and defined as the maximum available energy able to be extracted from a system before it reaches equilibrium with the environment (also known as the reference or dead state). It is the attempt to quantify the quality or work potential of energy. In contrast to energy, exergy can be destroyed through irreversible processes within the system. For a closed system at steady state, the exergy rate balance equation is expressed through Eq. (11),  X
dEx To _ _ _ d ¼0 Q W Ex ¼ 1 ð11Þ Tj dt   _ is the time rate of exergy transfer accompanying heat transfer in watts, To is the reference temperature in Kelvin, Tj where 1 TToj Q _ is the time rate of exergy accompanying work in watts, and Ex _ d is the time rate of exergy is the surface temperature in Kelvin, W destruction in watts. Let us consider the concept of exergy for an insulated wall at steady state where one surface of the wall is at 400K and the other surface is at 293K. An exergy analysis can be performed to determine the rate of exergy destruction through the wall. Assuming a reference temperature of 273K and a rate of exergy transfer accompanying heat transfer of 0.3 kW m 2 through the wall, the rate of

Insulation Materials

exergy destruction, in kW m

2

781

of wall surface, where no work is produced is:  X
To _ _ d =A ¼ Q=A 1 Ex Tj _ d¼ Ex




1

273K 400K




0:3

kW m2






1

273K 293K




0:3

kW m2



_ d =A ¼ 0:075 kW Ex m2 In this example, the exergy transferred into the wall is either destroyed within the wall (spontaneous heat transfer) or transferred out of the wall, where it is lost to the surroundings. Through this exercise, it becomes apparent that to minimize the exergy destroyed, the temperature gradient between both surfaces needs to minimized so that the heat flow is small. For example, _ d =A would be 0.037 kW m 2 if the heat flow were to be reduced to 0.15 kW m 2 by adding thicker or higher insulative material, Ex or effectively half. This analysis is relevant for many applications, including thermal energy storage systems where potential for work from the thermal storage can be significantly augmented by improving the insulation of the boundary.

2.24.5

Results and Discussion

While adding insulation is a simple and effective solution to conserving energy and improving energy efficiency, there is a point at which additional insulation will result in diminishing energy savings and no longer outweigh the monetary investment. This section analyses the embodied energy of common insulation materials for building envelopes in Canada.

2.24.5.1

Lifecycle Energy of Insulating Materials

When selecting the type and quantity of insulation to install within a building envelope, it is important to consider all energy aspects of the materials. In general practice today, a significant emphasis is placed on the amount of energy the building consumes as it is occupied. The amount of insulation is optimized to reduce the amount of heat transfer into or out of the building depending on the climate, while also commonly factoring in the cost of the material and the installation costs. What is often overlooked during the design phase is the lifecycle energy of the material. Commonly referred to as the cradle-to-grave energy, this value takes into account the amount of energy required for the production and installation of the insulation material, the amount of energy consumed, or saved during the life of the product and the amount of energy required to dispose of the material at the end of its useful life. A second value of importance is the lifecycle CO2 or CO2 equivalent, which measures the amount of CO2 or CO2 equivalent that is released into the atmosphere during the production, installation, and life of the unit. For both lifecycle calculations, the embodied energy or embodied CO2 of the material must be first calculated. The embodied values are the amount of energy or CO2 that must be considered before factoring in the energy and CO2 released during the operation of the building. What is included in calculating the embodied values varies depending on the source and the material being examined, but for most sources and materials, the following items are included when calculating the amount of energy consumed:

• • • • •

the the the the the

energy energy energy energy energy

required required required required required

to to to to to

extract the raw materials required to manufacture the product; manufacture the finished product; transport the materials (raw material to factory to building site); install the material within the building; and properly dispose of the material at the end of its useful life.

In addition to these values, some studies also include the following items:

• •

the energy used to maintain the product through its life; and the energy used when repairing or replacing the product during the lifespan of the total building.

Although there is a variety of embodied energy and CO2 equivalent values available, a comprehensive database of building materials has been assembled by Hammond and Jones [37]. Based on this database, the average embodied energy for the different insulations examined was 46.84 MJ kg 1. Table 2 below includes the embodied energy and embodied CO2 equivalent for some of the most common insulating materials. When looking at the building as a whole, the most common metric is to examine the lifecycle energy of the building. This takes into account the total-embodied energy within the building, and the amount of energy consumed to operate the building. Typically, only the energy to condition and operate the building (and not the total energy consumed within the building) are considered in the lifecycle energy and CO2 values. For example, the amount of energy required for space heating in the building would be included in the lifecycle energy, along with the amount of CO2 released to provide that heating, while the electricity to run the computers in the building would not be included in the energy and CO2 lifecycle calculations. To calculate the lifecycle energy, Elife, and CO2 values, eCO2 ;life , the embodied, heating, and cooling values must be calculated and summed. To do this,

782

Insulation Materials

Table 2 materials

Embodied energy and carbon for common insulating

Material

Embodied energy (MJ kg 1)

Embodied carbon (kg CO2eq kg 1)

Cellulose Fiberglass Mineral wool Rock wool Expanded polystyrene Polyurethane spray foam

0.94–3.3 28 16.6 16.8 88.6 101.5

0 1.35 1.28 1.12 3.29 4.26

Source: International Energy Agency. Key world energy statistics. Paris: International Energy Agency; 2016.

Table 3

Lifecycle energy for 1 m2 wall section using expanded polystyrene for 10-, 25-, and 50-year periods

Thickness (cm)

5 10 15 20 25 30 35 40 45 50 55 60

Annual space conditioning load (MJ)

124.4 62.2 41.5 31.1 24.9 20.7 17.8 15.6 13.8 12.4 11.3 10.4

Embodied energy (MJ)

111 222 332 443 554 665 775 886 997 1108 1218 1329

Lifecycle energy (MJ) 10-Year life

25-Year life

50-Year life

1355 844 747 754 803 872 953 1042 1135 1232 1331 1433

3222 1777 1369 1221 1176 1183 1220 1275 1342 1419 1501 1588

6332 3332 2406 1998 1798 1701 1664 1664 1688 1730 1784 1847

Eq. (12) must be utilized. Elife ¼

X

mi Eembodied;i þ ðHDD  COPheat þ CDD  COPcool Þ  Ueff Atotal 

24  3:6  Ylife 1000

ð12Þ

where mi is the mass of material i in kg, Eembodied,i is the embodied energy of material i in MJ kg 1, HDD is the heating degree days, CDD is cooling degree days, COPheat is the coefficient of performance of the heating system, COPcool is the coefficient of performance of the cooling system, Ueff is the effective heat transfer coefficient of the building in W m 2 K 1, Atotal is the overall area of the building envelope and Ylife is the expected lifespan of the building in years. This produces a final lifecycle energy value in MJ. Using this equation, the optimal level of insulation can be found. As additional insulation is added to the exterior of the building, its reduction in total energy consumption decreases. As such, as additional insulation is added, the total-embodied energy increases linearly; however, the resulting reduction in lifecycle energy is not linear and decreases as the thickness of the insulation increases. Consequently, there is an optimal point that can be found to have the lowest total lifecycle energy. When optimizing the thickness of the insulation installed within the building envelope, the location where the building will be located and its expected life are both critical in determining the optimal insulation level from a lifecycle energy perspective. Depending on the amount of insulation installed, the expected life of the building and the climate in which it is located, the embodied energy in the insulating materials can actually be greater than the operating energy required to heat and cool the building on an annual basis. Let us consider a building located in a climate with 2000 combined HDD and CDD. To simplify the calculation, a 1 m2 section of wall will be examined, the heating and cooling system has a COP of 1, and a single insulating material of increasing thickness will be assumed. Table 3 and Figs. 13 and 14 show the resulting total lifecycle energy for this small section using EPS and fiberglass insulation and assuming a 10-, 25-, and 50-year lifespan. From this example, we can observe that for buildings with a shorter expected life, the embodied energy becomes the dominant factor in the lifecycle energy calculation much quicker. For a 10-year life expectancy, the optimal insulation level is only 15 cm (5.9 in.). However, as the expected life increases, the operating energy required for space heating and/or cooling the building

Insulation Materials

783

Embodied energy compared to operating energy 3500

Embodied energy

3000

Energy (MJ)

2500

2000 Space conditioning load − 25 years

1500

1000

500

Space conditioning load − 50 years

0 0

10

20

30

40

50

60

70

Thickness (cm) Fig. 13 Comparison of embodied energy and operating energy for 25- and 50-year building life.

Comparison of lifecycle energy 7000

Total lifecycle energy (MJ)

6000

5000

4000 50 years 25 years 10 years

3000

2000

1000

0 0

10

20

30

40

50

60

70

Thickness (cm) Fig. 14 Lifecycle energy for a building exposed to 2000 degree days and an expected life of 10, 25, and 50 years.

becomes more dominant, with a building expected to last 50 years having an optimal insulation level for lifecycle energy of 35 cm (13.8 in.). As such, it is imperative that when designing a structure, the intended use of the building needs to be understood and that its expected life and the location where it is being installed will determine the impact of the embodied energy that will have on the optimization of the building envelope. To determine the impact on the optimal insulation levels, the number of degree days that the building is exposed to in the example was varied from the base 2000 degree days up to 6000 degree days. The impact of this change is shown in Fig. 15.

784

Insulation Materials

Comparison of lifecycle energy 20,000 18,000

Total lifecycle energy (MJ)

16,000

2000 Degree days

14,000 12,000

4000 Degree days

10,000 8000 6000

6000 Degree days

4000 2000 0 0

10

20

30

40

50

60

70

Thickness (cm) Fig. 15 Impact of degree days on the optimal level of insulation for the building envelope.

From this example, we can observe that as the number of degree days increases, the amount of insulation required to reach the optimal level for lifecycle energy increases. This is because when considering a building with a very high number of heating or CDD, it is more optimal from a pure energy perspective to continue to increase the insulation due to the disproportionately large amount of energy loss through the building envelope in comparison to the total-embodied energy within the insulating materials. For a 50-year period, if the building is exposed to 4000 degree days, an optimal insulation level of 55 cm (21.7 in.) is required, while at 6000 degree days, the optimal level is greater than 60 cm (23.6 in.). As such, buildings typically located in northern, heating dominated climates become an optimization exercise not in total lifecycle energy, but in cost of adding additional material in comparison to the energy being saved. In this example, a heating and cooling COP of 1 was used to illustrate the impact of location and expected life of the building on lifecycle energy calculations. In reality, the COP value can vary from as low as 0.9 using an on-site natural gas boiler to 4 when using a heat pump for heating and cooling. If the 25-year case is reexamined, varying the COP of the heating and cooling system, a new set of results are plotted in Fig. 16. From this example, we can observe that the higher the efficiency of the heating and cooling system, the greater the influence that the embodied energy within the insulating materials has on the lifecycle energy of the building. When a heating and cooling system provides a COP on an annual basis of 4, only 10 cm (4 in.) of insulation are required, compared to the 35 cm (13.8 in.) required when a COP of 1 is present for the space conditioning systems. While this previous example focused on examining the total lifecycle energy of a building, it is also important to look at the total lifecycle CO2 emissions of a building as well. The influence that the embedded CO2 has on the overall lifecycle CO2 is dependent not only on the heating and cooling load of the building, but also the fuel source being used to meet these loads. If natural gas is being used as the primary heating source in a heating dominated location, or the electrical supply within the jurisdiction where the building is located is predominately generated through the burning of fossil fuels (natural gas, coal, diesel, etc.), the amount of CO2 released for space heating and cooling will be many magnitudes greater than the amount of CO2 embodied within the material used to increase the thickness of the wall insulation. As such, if the design objective is to minimize the total lifecycle CO2, as much insulation should be used as is physically possible, while also considering cost and lifecycle energy optimization. If the building’s heating and cooling loads are met using electricity, and the building is located in a location that has a very high renewable energy penetration (most energy is generated using wind, solar, and hydro), at this point the embodied CO2 has a much larger impact on the total lifecycle CO2. To calculate the lifecycle CO2 of a building, Eq. (13) is used. eCO2 ;life ¼

X

mi Eembodied CO2 ;i þ ðHDD  COPheat þ CDD  COPcool Þ  Ueff Atotal  24  ei  Ylife

ð13Þ

where Eembodied CO2 ;i is the embodied CO2 of material i in kgCO2 kg 1, ei is the carbon emission intensity of the heating and cooling fuel in g CO2 kWh 1. This produces a final lifecycle CO2 emissions, eCO2 ;life , value in kg.

Insulation Materials

785

Comparison of varying coefficient of performances (COPs) on lifecycle energy consumption 3500

3000

Lifecycle energy (MJ)

2500

2000 COP1 COP2 COP3 COP4

1500

1000

500

0 0

10

20

30

40

50

60

70

Thickness (cm) Fig. 16 Impact of heating and cooling coefficient of performance (COP) on optimal insulation levels.

Using the equations for lifecycle energy and CO2 emissions provides an additional metric to examine the insulation being installed within a building. Depending on the overall objective of the building design, the insulation level can be optimized for either lifecycle energy or lifecycle CO2 emissions. These values can be used within the design constraints of the building and a cost optimization exercise can be used to develop an optimal building envelope for any given building, based on design goals, materials, expected life, and building location.

2.24.6

Case Studies and Examples

This section contains a thorough case study about the design, testing, and development of a high RSI-value building envelope for the Team Ontario’s entry into the 2013 solar decathlon to show the importance of insulation in residential buildings and its impact on energy consumption.

2.24.6.1

Designing a High Thermal Resistance Wall With Vacuum Insulation Panels – ECHO – Team Ontario’s Entry to the 2013 Solar Decathlon

The building envelope is critical in the design of high performance, energy efficient housing. As codes and voluntarily performance standards (LEED, Passive House, Net-Zero) [10,11], the ability to design and construct a high thermally insulating wall is pivotal to the success of these walls. Traditionally, the high R-values required for these projects have been obtained by simply increasing the thickness of the wall assembly. This has led to double wall construction, with or without exterior board insulation and interior insulation between wall constructions, leading to walls that commonly have a total thickness in excess of 300–400 mm (12–16 in.) and insulating values of approximately RSI-7 m2 K W 1 (R-40 h 1F ft2 BTU 1). Although this provides a significant energy improvement over a code built wall, these highly insulated walls are in excess of twice as thick as a traditional wall. In the current construction and real estate environment, where house are built to the allowable extent of the desired piece of land, this increase in wall thickness leads to a subsequent decrease in usable floor space. In a typical, single detached two-storey home, this lost floor space could be the equivalent of adding an extra bedroom. To help combat and start to reverse this trend of simply building thicker walls to increase thermal resistance, new, highly insulated building materials need to be integrated into the residential sector. As such, one of the design goals for Team Ontario as they designed and subsequently built their competition home for the 2013 U.S. Department of Energy Solar Decathlon was to develop a building envelope assembly that achieves a minimum total effective thermal resistance if RSI-8 m2 K W 1 (R-45 h 1F ft2 BTU 1), while having a total thickness equal to or less than 300 mm (12 in.). To achieve this effective thermal resistance goal, VIPs were selected due to their very high thermal resistance in comparison to their thin profile.

786

Insulation Materials

This project had a unique design statement, in which the house was being designed and built in the cold climate of Eastern Ontario, but also needed to be shipped 10,000 km (6000 mi) round trip to the competition site in Southern California. Additionally, the house had to be built in a modular fashion and be able to withstand the lifting and assembly process in both Canada before the competition and at the competition site. As such, a higher level of structural rigidity was required when compared to a site built, stick-framed house. This required that 38 mm by 152 mm (1.5 in. by 5.5 in.) lumber be used as the main structural members, as opposed to the preferred 38 mm by 89 mm (2  4) lumber, which would have allowed a thinner wall profile. Using this as the starting point, a wall was designed that incorporated VIPs to the outside of the structural members, allowing for an almost continuous layer of insulation and significantly reducing the thermal bridging within the assembly. Although VIPs have significant insulating benefits, they posed a substantial challenge in terms of how to successfully integrate the panels into a residential wall assembly. These panels are very fragile, and as a result, must be protected from threats, such as fasteners, improper handling, and other potential methods of puncture. As a result, the panels were placed in the center of the assembly, maximizing the distance between the interior and exterior surfaces, reducing the probability of accidently puncture caused by the homeowner making modifications to the house. A comprehensive installation plan was also developed, ensuring the panels could be installed without incident, including material handling and storage practices, as well as the installation process. Another design consideration was the impact of VIPs on the moisture transfer that occurs within a wall assembly. VIPs can significantly influence the moisture transfer within a wall assembly, as a VIP layer can cause a second vapor barrier and move the dew point of the wall. As the VIPs account for approximately 50% of the insulation, it was decided that the panels would be placed on the outside of the main structural members, ensuring the dew point remains outboard of the structural elements throughout the year, removing the potential for bulk moisture accumulation. This significantly reduced the possibility of rot or mold growth. However, it did not eliminate it, as there was still the potential for moisture to diffuse into the wall, and become entrapped between the VIPs and the traditional vapor barrier. To counter the possible moisture issues, no additional vapor barrier (typically polyethylene sheets) was installed on the inside of the wall, and instead, 50 mm (2 in.) of spray foam was applied within the stud cavities, allowing moisture a path out of the wall through the studs. As this is not traditionally the way walls dry, which typically sees moisture migrating to the exterior, two – 100 mm (4 in.) bands of moisture permeable EPS were installed within the wall. In addition to serving as drying paths, these bands also provided a location for mechanical and electrical penetrations, as well as a plane to fasten the exterior supports to the main structural components. Throughout the process, thermal modeling within THERM [37] was conducted to determine the overall effective thermal resistance of the wall. As different iterations were proposed, quick simulations were conducted to determine their effective thermal resistances, which were then inputted into a complete energy model in EnergyPlus [38] to determine the annual energy performance. Through these iterations, a final wall section was proposed (Fig. 17). During the initial design phase, a number of parameters were unknown and needed to be assumed. The biggest of these was the thermal resistance of the VIPs being used. A wide variety of quoted values were available within literature and product specifications, however, it was unknown the true value of the panels being proposed. As a result, a thermal conductivity of 0.002 W m 1 K 1 was assumed in the initial modeling, and an overall effective thermal resistance of the wall assembly of 14.7 m2 K W 1 (R-83 h 1F ft2 BTU 1) was obtained. To remove many of the assumptions, experimental validation of the thermal models was done, first through in situ testing, followed by laboratory testing in a guarded hot-box.

2.24.6.2

Test Results

In situ testing was first conducted on the proposed wall design, with a 2.4 m by 2.4 m (8 ft by 8 ft) exact replica built, instrumented, and installed within an outdoor test facility at Carleton University in Ottawa, Canada. Measurements were taken for Exterior Graphite EPS Vacuum insulation panels

Rainscreen airspace

2×3 Strapping

Wood siding

Plywood sheathing

2×6 Framing

Medium density spray foam

Ultility airspace Gypsum wall board

Interior Fig. 17 Cross-section of the final wall design, including 20 mm (13/16 in.) of vacuum insulation panels (VIPs) placed between the interior structural elements and an exterior 2×3 wall assembly for which allowed for cladding to be attached and provided a rain screen.

Insulation Materials

787

4 weeks in January and February, and measurements were taken at 1 min intervals. Each unique cross-section was instrumented to measure the temperature profile through the wall assembly, as shown in Fig. 18 for the test period, and to determine the average thermal resistance of the cross-section. Using the in situ data, and taking a weighted average of each cross-section, an overall effective thermal resistance of the wall assembly was determined to be 10.4 m2 K W 1 (R-59 h 1F ft2 BTU 1). This experimentally obtained thermal resistance was significantly lower than the originally predicted value, so the thermal resistance of each layer was determined using the measured temperature profile and heat flux through the wall assemblies. These were compared to the values used within the THERM modeling, and the actual thermal resistance of the VIPs for the 20 mm (13/16 in.) layer was experimentally found to be 5 m2 K W 1 (R-28 h 1F ft2 BTU 1), compared to the initially assumed value of 8.3 m2 K W 1 (R-47 h 1F ft2 BTU 1) for the VIP layer. When this experimentally determined thermal resistance was used in an updated model, the predicted effective thermal resistance was found to be 8.9 m2 K W 1 (R-50 h 1F ft2 BTU 1). To validate the in situ test and THERM modeling results, another replica wall was constructed and installed within a guarded hot-box at Carleton University, as shown in Fig. 19, allowing the wall to be experimentally evaluated at steady state conditions.

Team Ontario experimental wall data Jan. 11, 2013 − Feb. 8, 2013 20

60 50

10

40 Temperature (°C)

20

−10

10 −20

0 −10

−30

−20 −40 −30 Feb-08

Feb-06

Feb-04

Feb-02

Jan-31

Jan-29

Jan-27

Interior Sprayfoam-Plywood VIP-VIP Exterior surface

Jan-25

Jan-23

Jan-21

Jan-19

Jan-17

Jan-15

Jan-13

Jan-11

−40

−50

Interior surface Plywood-VIP VIP-EPS Exterior

Fig. 18 Temperature profile through the wall assembly with vacuum insulation panels (VIPs) over the 4 weeks testing period in Winter 2013.

Fig. 19 Vacuum insulation panel (VIP) layer of the test wall being installed within the guarded hot-box surround panel for laboratory testing.

Heat flux (W m−2)

0

30

788

Insulation Materials

30

Temperature (°C)

20 2SF PLY

10

2 PLY VIP 2 VIP VIP

0 0

5

10

−10

15

2 VIP EPS Climate side Metering box

−20 −30

Time (h)

Fig. 20 Steady state temperature profile for a 15 hour test in the guarded hot-box. EPS, expanded polystyrene; VIP, vacumm insulation panel.

This wall specimen was instrumented to provide a temperature profile at the same cross-sections as the in situ testing, with the temperature profile over the length of the test shown in Fig. 20. Instead of using a weighted average, the total heat transfer through the wall was measured. Coupled with the measured interior and exterior temperatures, the overall effective thermal resistance of the wall assembly was found to be 8.2 m2 K W 1 (R-46.5 h 1F ft2 BTU 1). This value was within the experimental uncertainty of the test, validating both the in situ test results and updated thermal model in THERM.

2.24.6.3

Building Envelope Construction

Following the detailed analysis of the proposed wall assembly, this design was selected as the complete house. Although the thermal performance was optimized and fit within the parameters, a number of construction challenges remained. A detailed panel layout was designed based on the proposed geometry of house, including the size and location of windows, doors, and required mechanical penetrations. This was required to maximize the VIP coverage in the wall assembly. In areas where the house dimensions did not match the VIP dimensions, changes were proposed to the architects to change the location or size of the feature, and in many cases, a compromise was found between the engineers and the architects. This demonstrated an important aspect of designing with new materials. After the panel layout had been developed, the few remaining areas that did not have VIP coverage were filled with EPS. The building envelope was then constructed as per the designed specifications. Special care was required in the storage and handling of the panels. This included keeping the panels in dry conditions, avoiding materials being stacked on top of them, and away from tools that could potentially damage the panels. When installing the panels, it was imperative that the panels were not dropped or left on the ground around the exterior of the building. To improve the building envelope performance, the seams between the panels were taped, reducing potential air infiltration into the building. The installed panels on the exterior of the structural members are shown in Fig. 21. Once the VIPs were installed on the exterior of the structural members, strapped EPS was installed on the exterior of the VIPs (Fig. 22), allowing the cladding to be installed on the exterior of the EPS (Fig. 23). A picture of the finished exterior is shown in Fig. 24.

2.24.6.4

Lessons Learned

This project successfully showed that VIPs could be utilized in a residential building. A building envelope was designed, tested, and constructed with an overall effective thermal resistance of 8.2 m2 K W 1 (R-46.5 h 1F ft2 BTU 1) in a total thickness of 300 mm (12 in.), which could be easily reduced if not for some of the unique design criteria required for a Solar Decathlon Project. With current design and building practices, the design process when using VIPs is far more complex and involved when compared to the design of a traditional house. As more demonstration projects are completed successfully, and the knowledge base is increased, this process will become more streamlined, and consequently, reduce the cost of implementing VIPs into a residential building envelope. Once successful, based on the outcomes of this project, the implementation of VIPs into residential buildings could allow for the energy consumption required for space heating and cooling to be reduced by half, while keeping the same overall wall thickness and therefore not compromising the interior floor space of the building.

2.24.6.5

Building Enclosure Simulation

Following the experimental analysis at steady state and in situ conditions, a 2D heat transfer computer simulation was conducted. In the Team Ontario Solar Decathlon House, incorporating the VIPs to the exterior wall insulation of the building introduced a nonhomogenous thermal resistance to the layer. The temperature deviations that are caused by the VIPs are difficult to visualize with experimental measurements due to a lack of granularity. However, computer simulations are a useful tool to determine the temperature gradients and effective RSI-value for multiple cross-sections and designs. For 2D simulations, when thermal resistance

Insulation Materials

789

Fig. 21 Vacuum insulation panel (VIP) layer being installed on the exterior of structural members.

Fig. 22 Strapped expanded polystyrene (EPS) exterior layer installed and ready for siding.

discontinuities exist in the vertical and horizontal directions of a single layer, multiple cross-sections need to be simulated, and a weighted averaging technique based on the height ratios of the cross-sections should be used. When integrating VIPs into building envelopes, installing them as a continuous layer, similar to foam board as an exterior insulation, is not the most efficient method due to cost and protection the VIP from puncture. Recently, the VIPs have been encased in rigid foam board prior to being fastened to the exterior of the building to simplify the installation method at the construction site. Fig. 25 shows a panel with VIPs encased in XPS with the top removed and distinct spaces created for a thin, high performance wall, and to avoid puncturing the VIP before completing construction of the building. However, this does create a nonhomogeneous layer and requires simulation of multiple cross-sections to determine the effective RSI. A drawing of an exterior wall assembly using VIP encased in XPS as the exterior insulation is shown in Fig. 26. The wall is built with 38 mm by 89 mm (2  4 lumber) wood studs on 406 mm (16 in.) center spacing with RSI-2.4 m2 K W 1 (R-14 h 1F ft2 BTU 1) batt insulation in the cavity between the studs.

790

Insulation Materials

Fig. 23 Pine siding installed to provide a rain screen with small gaps to allow for drying.

Fig. 24 Finished product at the competition site in California (photo credit: U.S. Department of Energy).

The simulation consists of modeling two separate cross-sections, one incorporating VIPs and one excluding VIPs. By using interior and exterior boundary conditions, and the thermal properties of the building materials given from by the manufacturers or found in the THERM database, simulation outputs valuable information regarding the assembly. The VIP and non-VIP crosssections were simulated to have RSI-values of 9.27 m2 K W 1 (52.6 h 1F ft2 BTU 1) and 4.48 m2 K W 1 (25.4 h 1F ft2 BTU 1), respectively. The effective thermal resistance value of the either assembly is found by using the coverage ratios, where 88% of the wall assembly is covered by VIPs and the remaining 12% is covered by 50 mm (2 in.) of XPS. Therefore, the effective thermal resistance of the entire wall assembly is simulated to be 8.87 m2 K W 1 (50.2 h 1F ft2 BTU 1). This shows that when the VIPs are encased in insulation as a protective measure, they can maintain a high level of thermal performance in the building envelope with a minimal loss in effective RSI when compared to a complete VIP layer. In addition to the RSI-value of a cross-section, a desirable output of a simulation is the isotherms of a model. Understanding how the heat will permeate through the wall assembly is also of interest when the insulating layers contain discontinuities that will act as thermal bridges. The isotherms present in the wall assembly from the exterior to the interior with the VIP-encased panel are

Insulation Materials

791

Fig. 25 A photo of a sheet of extruded polystyrene (XPS) with embedded vacuum insulation panels (VIPs) to augment the insulation. This was installed within a guarded hot-box test facility to measure the effective RSI-value.

Stud

Batt insulation

Drywall

Plywood

XPS

VIP

Fig. 26 Composition of simulated exterior high RSI-value wall assembly. VIP, vacuum insulation panel; XPS, extruded polystyrene.

Interior

Exterior Fig. 27 Schematic of a wall assembly evaluated with two-dimensional steady state without isotherms.

Interior

19.4 16.8 14.1 11.5 8.9

6.3

−17.3

Exterior Fig. 28 Isotherms through the wall assembly. Note the thermal bridge that exists at the extruded polystyrene (XPS) of the composite insulation panel.

shown in Fig. 27. Even without considering the edge effects caused by the lower thermal resistance along the perimeter of the VIPs, a large temperature difference along the wood stud–gypsum board interface is caused by the XPS aligned with the wood studs, two low thermal resistive materials compared to the VIP (Fig. 28).

792

Insulation Materials

2.24.7

Future Directions

With the increase in building efficient and high performance buildings, the use of high R-value materials has been growing. These novel materials offer a large thermal resistance with a small thickness in order to highly insulate the building without compromising the internal floor area of the home or the building’s footprint caused by increasing the thickness of the building envelope. The materials that will be discussed can offer as high as 10 times the thermal resistance per unit thickness when compared to the conventional insulation materials discussed previously [39].

2.24.7.1

Aerogels

Aerogels are very porous solid and lightweight materials that are composed of a silica dioxide gel and a liquid solvent. Through a chemical process and drying the gel, the liquid pores are replaced with gas, providing the low density solid. Aerogels are an effective insulator because the small gas-filled pores significantly limit the means of heat transfer. The internal structure limits the air movement through the material, which reduces convection [40]. Silica is known as having a low thermal conductivity, and the pores in the material cause many thermal breaks further improving the effective thermal conductivity. Aerogels are often provided on a roll, and cover the surface of the building envelope. Compared to conventional insulations, aerogels can provide 2–2.5 times the thermal resistance, which offer a significant reduction in insulation thickness required to create a high performance envelope. Aerogels have the flexibility and versatility to be mechanically fastened to many facets of the building, including the interior or exterior walls, windows, doors, attics, among many others. Due to their high compressive stress, they can also be used as underfloor insulation when the thickness is a limitation [41]. In buildings, the aerogels offered can be translucent or opaque, which further opens the possibility of applications, specifically large thermal improvements on windows and skylights [40]. Unfortunately, a high cost per unit area limits its widespread implementation.

2.24.7.2

Vacuum Insulation

VIPs are a highly insulating material that has been implemented into many appliances, such as refrigerators and freezers, and the transportation industry where refrigeration is required. VIPs consist of an evacuated, open porous core enclosure within a several metallized laminate layers that provide a high thermal resistance per unit thickness [39]. The thermal resistance benefits are provided by the lack of convective and conductive currents through the panel caused by the porous structure and vacuum induced inside. Some VIP manufacturers have claimed a thermal resistance as high as RSI-234.6 per meter (R-40 in. 1) through the center of the panel. However, the panel does not have a homogenous thermal resistance. Experimental studies have shown that a 32% drop in thermal resistance may exist along the perimeter of VIPs and this reduction would vary between each panel from different manufacturers [42]. The thermal resistance variations that exist along the perimeter are caused by the different sealing or folding processes of the metallic envelope to contain the vacuum. Another factor would be the type of metallic envelope and core material used to fabricate the VIPs. As such, the VIPs contain an inherent thermal bridge along the perimeter of the panels. The perimeter has a significant increase in thermal conductivity when compared to the center of the panel because of the manufacturing technique to seal the seams and the conductivity of the metallized foil envelope [43]. The thermal bridge can be seen through infrared thermal images taken during laboratory testing in Fig. 29. The panel edges are distinctly shown in the thermal images by the variation in colors related to the surface temperature. Even though there are substantial advantages of VIPs, many disadvantages and challenges exist before they can be implemented into the building envelope as an insulator. Since the VIP’s low conductivity is based on the vacuum contained within the enclosure, if they are punctured, or the enclosure is compromised, there is a large reduction in performance. In building construction, there are many hazards, such as mechanical fasteners (e.g., screws, nails, etc.) and the physical handling of the panels, could lead to puncture before the building envelope construction is complete [44]. Another effect of VIPs when installed inside the building envelope is their degradation in thermal performance over its lifespan. The panels will experience a drop in RSI-value after fabrication due to increasing internal pressure and moisture accumulation within the panel. While the VIPs create an impermeable vapor barrier within the building envelope, there can be vapor and air diffusion into the core of the panel either through the metallic foil or through imperfections created during manufacturing or installation [39]. Some studies have shown that moisture accumulation will have a greater effect on VIP degradation than increased air pressure, and a response has been the addition of desiccants to the core material. The core material’s conductivity is increased as moisture enters the foil and is trapped within the pores. This creates a thermal bridge for heat to travel. After installation into a building enclosure, the VIPs are introduced the environment and moisture passing through the envelope, either through diffusion or unintended leakage. The moisture is unavoidable after installation, and the effects should be taken into consideration during the analysis for the lifetime of the envelope, whether it is 10, 15, 30, or 50 years [39,43]. However, even though the panel will experience degradation over time, the VIPs will maintain an RSI-value greater than common building insulations [45]. The high thermal performance of VIPs offer the ability to create thin, high RSI-value building envelopes that comply with voluntary building standards without sacrificing the internal floor area or increasing the footprint of the building required by the

Insulation Materials

793

20.3 14.0

23.0

Fig. 29 Infrared thermal image of interior surface temperature of vacuum insulation panels (VIPs) installed in a building envelope. Reproduced from Conley B, Cruickshank CA. Evaluation of thermal bridges in vacuum insulation panels assemblies through steady-state testing in a guarded hot box. In: eSim 2016, Hamilton; 2016.

additional thickness when using common insulations. Even though drawbacks, such as performance degradation, thermal bridges along the perimeter, fragility and constructability exist, they offer a high-upside solution to thermally inefficient buildings without increasing wall thickness.

2.24.7.3

Gas-Filled Panels

Gas-filled panels (GFP) are similar to VIPs such that they have a metallic envelope. However, instead of being evacuated, they are filled with a low conductivity gas, and since the enclosed gas remains at ambient pressure, the porous structure is not required. The GFPs still contain a baffle structure inside in order to restrict the movement of gas and limit convection [46]. Since the filled gas has a strong effect on the effective thermal conductivity, varying the gas can change the thermal performance. Some studies have suggested that GFPs would be able to obtain an effective thermal conductivity of 0.035 W m 1 K 1 for a 25-mm panel thickness when using air as the main gas, and 0.0106 W m 1 K 1 with krypton [46]. Typically, gases with higher molecular weight and mono-atomic gases will offer a smaller thermal conductivity, and should be chosen as the fill gas. The envelope foil must act as an effective gas barrier in two directions, such that it effectively keeps the low conductivity gas inside and the air and moisture outside the panel. The GFP lifecycle strongly depends on the gas transmission rate through the envelope foil as the effective thermal conductivity of the panel [46]. Similar to VIPs, the lifespan of the panels are dictated by the diffusion through the foil.

2.24.8

Closing Remarks

In conclusion, the building enclosure significantly impacts the overall efficiency of a building, in terms of thermal performance and indoor comfort. The enclosure can be improved through addition insulation applied to various locations in the building either during the initial construction or during a renovation. Another method to improve the efficiency is to increase the airtightness of the enclosure, by sealing the seams and imperfections that may exist. Finally, improving the enclosure will affect the required capacity of the heating and cooling equipment in a building, further improving the amount of energy and costs required to maintain indoor comfort. The various types of insulation discussed in this chapter have a number of benefits, installation requirements, and challenges. The building designer is responsible for selecting a building enclosure that meets the required performance. The insulation materials or building assemblies may be experimentally tested or simulated to validate the thermal or moisture performance of the design at specific climate, since climatic zones play a large role in the enclosure requirements. As building designs are trending toward a lower energy consumption, the enclosure insulation becomes increasingly important. Adopting new and higher performing insulating materials with careful enclosure design will be paramount to the future of buildings.

794

Insulation Materials

References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46]

International Energy Agency. Key world energy statistics. Paris: International Energy Agency; 2016. Office of Energy Analysis. U.S. Energy Information Administration. International energy outlook 2016. Washington, DC: U.S. Department of Energy; 2016. Natural Resources Canada – Office of Energy Efficiency. Energy use data handbook 1990 to 2013. Ottawa, ON: Natural Resources Canada; 2015. EuroStat. Final energy consumption, EU-28, 2014, European Commission; 2016. U.S. Department of Energy. Furnaces and boilers. Available from: http://energy.gov/energysaver/furnaces-and-boilers. International Code Council. 2012 International energy conservation code. Illinois: International Code Council Inc.; 2011. ASHRAE. Energy standard for buildings except low-rise residential. Atlanta, GA: ASHRAE; 2013. International Energy Agency. Energy efficiency requirements in building codes, energy efficiency policies for new buildings. Paris: International Energy Agency; 2008. National Research Council of Canada. National energy code of canada for buildings 2011. Available from: http://www.nrc-cnrc.gc.ca/eng/publications/codes_centre/ 2011_national_energy_code_buildings.html; 2016. Passive House Institute US. Available from: http://www.phius.org; 2016. U.S. Green Buildiing Council. LEED. Available from: http://www.usgbc.org/leed; 2016. Green Building Initiative. Available from: http://www.thegbi.org/; 2016. U.S. Department of Energy. Zero energy ready home. Available from: http://energy.gov/eere/buildings/zero-energy-ready-home; 2016. Wilkinson J, Ueno K, De Rose D, Straube J, Fugler D. Understanding vapour permeance and condensation in wall assemblies. In: 11th Canadian conference on building science and technology, Banff, Alberta; 2007. Lstiburek J. Understanding vapour barriers. Westford, MA: Building Science Corporation; 2005. Straube J. The influence of low-permeance vapour barriers on roof and wall performance. In: Proceedings of performance of whole buildings; 2011. Brown W, Chown G, Poirier G, Rousseau M. Designing exterior walls according to the rainscreen principle. Ottawa, ON: National Research Council of Canada; 1999. Lstiburek J. Moisture control for buildings. ASHRAE J 2002;44(2):36–41. Canada Mortgage and Housing Corporation. Energy efficiency building envelope retrofits for your house. Canada: Canada Mortgage and Housing Corporation; 2015. US Department of Energy. Blower door tests. Available from: http://energy.gov/energysaver/blower-door-tests; 2016. Canada Mortgage and Housing Corporation. Before you start your energy efficiency retrofit – The building envelope. Canada: Canada Mortgage and Housing Corporation; 2015. Natrual Resources Canada. Heat/energy recovery ventilators; 2016. US Department of Energy. Insulation. Available from: http://energy.gov/energysaver/insulation; 2016. US Department of Energy. Insulation materials. Available from: http://energy.gov/energysaver/insulation-materials; 2016. US Department of Energy. Types of insulation. Available from: http://energy.gov/energysaver/types-insulation; 2016. Mullens M, Arif M. Structural insulated panels: impact on the residential construction process. Constr Eng Manag 2006;132(7):786–94. Medina MA, King JB, Zhang M. On the heat transfer rate reduction of structural insulated panels (SIPs) outfitted with phase change materials (PCMs). Energy 2008;33 (4):667–78. Canada Mortgage and Housing Corporation. Insulating your house. Canada: Canada Mortgage and Housing Corporation; 2009. Natural Resources Canada. Keeping the heat in – chapter 7: walls, Natural Resources Canada; 2012. US Department of Energy. Where to insulate in a home. Available from: http://energy.gov/energysaver/where-insulate-home; 2016. Natural Resources Canada. Keeping the heat in – chapter 5: roofs and attics, Natural Resources Canada; 2012. Natural Resources Canada. Keeping the heat in – chapter 6: basement insulation, Natural Resources Canada; 2012. ASTM International. Standard test method for steady sate heat flux measurements and thermal transmission properties by means of a guarded hot plate apparatus. West Conshohocken, PA: ASTM International; 2013. ASTM International. Standard test methods for water vapor transmission of materials. West Conshohocken, PA: ASTM International; 2015. ASTM International. Standard test method for thermal performance of building materials and envelope assesmblies by means of a hot box apparatus. West Conshohocken, PA: ASTM International; 2011. ASTM International. Standard practice for determining thermal resistance of building envelope components for in-situ data. West Conshohocken, PA: ASTM International; 2007. Hammond G, Jones C. Inventory of carbon & energy (ICE). Available from: http://www.circularecology.com/embodied-energy-and-carbon-footprint-database.html; 2011 [accessed 01.08.17]. THERM. Lawrence Berkeley National Laboratory (LBNL). Available from: http://windows.lbl.gov/software/therm/therm.html; 2016. National Renewable Energy Laboratory. Energyplus. Available from: https://energyplus.net/; 2016. Mukhopadhyaya P, Kumaran K, Ping F, Normandin N. Use of vacuum insulation panel in building envelope construction: advantages and challenges. In: 13th Canadian conference on building science and technology, Winnipeg, MB; 2011. Baetens R, Jelle B, Gustavsen A. Aerogel insulation for building applications: a state-of-the-art review. Energy Build 2011;43(4):761–9. Fickler S, Milow B, Ratke L, Schnellenbach-Held M, Welsh T. Development of high performance aerogell concrete. In: 6th International building physics conference; 2015. Conley B, Cruickshank CA. Evaluation of thermal bridges in vacuum insulation panels assemblies through steady-state testing in a guarded hot box. In: eSim 2016, Hamilton; 2016. Fricke J, Heinemann U, Ebert H. Vacuum insulation panels – from research to market. Vacuum 2008;82(7):680–90. Baetens R, Jelle BP, Thue JV, et al. Vacuum insulation panels for building applications: a review and beyond. Energy Build 2010;42):147–72. Kalnaes S, Jelle B. Vacuum Insulation panel products: a state-of-the-art review and future research pathways. Appl Energy 2014;116:355–75.

Further Reading ASHRAE. ASHRAE handbook of fundamentals. Atlanta, GA: ASHRAE; 2013. Jelle B. Traditional, state-of-the-art and future thermal building insulation materials and solutions – properties, requirements and possibilities. Energy Build 2011;43 (10):2549–63. Modera MP, Persily AK. Airflow performance of building envelope components. West Conshohocken, PA: ASTM International; 1995. Moran MJ, Shapiro HN, Boettner DD, Bailey MB. Fundamentals of engineering thermodynamics. vol. 7. Hoboken, NJ: John Wiley & Sons; 2010. Staube J. High performance building enclosures. Sommerville, MA: Building Science Press; 2012.

Insulation Materials

Relevant Websites https://ashrae.org/ American Society of Heating, Refrigeration and Air-Conditioning Engineers (ASHRAE). https://buildingscience.com/ Building Science Corporation. https://www.cmhc-schl.gc.ca Canadian Housing and Mortgage Corporation (CMHC). BuildingGreen.com Green Building Information. http://www.nrcan.gc.ca/home Natural Resources Canada (NRCan). http://www.nrcan.gc.ca/energy/offices-labs/office-energy-efficiency NRCan Office of Energy Efficiency. http://energy.gov/ United States Department of Energy (US DoE).

795

2.25 Hydrophobic Materials Bekir S Yilbas, Haider Ali, Muhammad R Yousaf, and Abdullah Al-Sharafi, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia r 2018 Elsevier Inc. All rights reserved.

2.25.1 Introduction 2.25.2 Basics of Wetting States 2.25.2.1 Wetting on Smooth Surfaces 2.25.2.2 Wetting on Rough Surfaces 2.25.2.3 Contact Angle Hysteresis 2.25.2.4 Lubricant Impregnated Surfaces 2.25.2.5 Particle Adhesion on Surfaces 2.25.2.5.1 Johnson–Kendall–Roberts model 2.25.2.5.2 Derjaguin, Muller, Toropov model 2.25.2.5.3 Hamaker’s model 2.25.2.5.4 Rumpf–Rabinovich model 2.25.2.5.5 Gravitational force 2.25.2.5.6 Lift force (inertial force and shear stress) 2.25.2.5.7 Drag force due to pressure and shear on the particle surface 2.25.2.5.8 Centrifugal force 2.25.2.5.9 Friction force 2.25.2.5.10 Coriolis force 2.25.3 Surface Treatment Methods for Hydrophobicity 2.25.3.1 Laser Texturing of Surfaces 2.25.3.2 Modified Silica Nanoparticle Deposition 2.25.3.3 Polydimethylsiloxane Replica Molding 2.25.3.4 Nanoparticle Film Forming 2.25.3.5 Lubricant Impregnated Surfaces 2.25.4 Case Study for Laser Texturing and Polydimethylsiloxane Replication of Textured Surface 2.25.4.1 Experimental 2.25.4.2 Findings and Discussions 2.25.5 Case Study for Solvent Induced Crystallization of Polycarbonate Surface and Environmental Dust Effects 2.25.5.1 Experimental 2.25.5.2 Findings and Discussions 2.25.6 Conclusions 2.25.6.1 Laser Texturing and Polydimethylsiloxane Replication of Textured Surface 2.25.6.2 Solution Crystallization of Polycarbonate Surface and Environmental Dusts 2.25.6.3 Future Directions Acknowledgment References Further Reading Relevant Websites

Nomenclature A f F Fg g m r rc R

796

Hamaker constant Surface fraction of interface Force Gravitational force Gravitational acceleration Mass of the particle Roughens ratio Distance of the particle from the rotational center Radius of the particle

Re* u* (n) y f g r m o mf

797 798 799 799 800 801 802 802 803 803 803 804 804 804 804 804 804 805 805 805 807 808 809 810 810 811 822 822 823 828 828 829 829 829 830 831 831

Reynolds number Friction velocity Kinetic viscosity Contact angle Interfacial area Surface energy between two surfaces Density of the dust particle Dynamic viscosity of fluid Angular velocity of the disk Friction coefficient

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00253-4

Hydrophobic Materials

Abbreviations CA CAH OTS PDMS

2.25.1

Contact angle Contact angle hysteresis Octadecyltrichlorosilane Polydimethylsiloxane

PFOTS TEOS UV–vis WCA

797

Trichloro (1H,1H,2H,2H-perfluorooctyl) Tetraethoxysilane Ultraviolet–visible Water contact angle

Introduction

Climate change increases the frequency of regular dust storms around the globe particularly in the Middle East and the Sahara region. The performance of solar energy harvesting devices is highly dependent on the amount of incident solar energy reaching at the active surface of the energy harvesting devices. Dust settlements on such surfaces degrade the device performance in terms of efficiency and output power and require additional efforts to remove dust from surfaces. Several methods have been purposed for dust removal from surfaces and some of these include sonic and mechanical excitation of dust particles, mechanical brushing, air jet blowing, and water cleaning. Most of these methods involved with sophisticated devices or external efforts, such as energy to operate. Scarcity of the clean water limits the practical applications of water jet and water film cleaning of surfaces and, hence, minimization of water consumption during cleaning becomes necessary. On the other hand, adopting self-cleaning surfaces can minimize external energy use in energy harvesting devices and can provide effective cleaning process toward removing the dust particles. In general, self-cleaning process utilizes hydrophobic characteristics of surfaces; in which case, low free energy of surface and reduced particle contact area at the surface, due to airgap within surface texture, provide weak adhesion of particles on surfaces. In addition, low contact angle hysteresis (CAH) of hydrophobic surfaces enables water droplet rolling on the inclined hydrophobic surface. This, in turn, facilitates picking up of weakly adhered dust particles from the hydrophobic surface by rolling water droplets. On the other hand, self-cleaning characteristics of optically transparent wafers are of interest in many applications and it becomes critical for efficient harvesting of solar energy in harsh environments [1]. Generating self-cleaning characteristics of optically transparent wafer surfaces are one of the current challenges because the task of achieving durable surfaces in the harsh environmental conditions is extremely difficult. As seen in nature, lotus leaves red rose petals, fish scales, etc. [2] inspire to design and fabricate artificial surfaces with self-cleaning characteristics. Several examples, as observed in nature, bring out the importance of surface hydrophobicity to achieve self-cleaning characteristics at the surface. Hydrophobic characteristic of surfaces mainly depends on surface texture and surface free energy of substrate materials. Surface texture composing of micro/nanopillars with low surface energy is desirable for achieving hydrophobic characteristic of surfaces [3]. Although several methods and strategies are introduced to create hydrophobic surfaces, some of these methods are related to multistep procedures and mostly involve with harsh conditions, specialized reagents, and high cost. These include phase separation [4], electrochemical deposition [5], plasma treatment [6], sol–gel processing [7], electrospinning [8], laser texturing [9], and solution immersion [10]. Challenges are faced during surface texturing and chemical/physical modification of surfaces toward reducing surface free energy because of high cost, long processing time, and equipment and skilled man power requirements. Consequently, development of new technologies for cost-effective processing toward improving surface hydrophobicity of optically transparent wafers becomes essential. Laser surface texturing through combination of controlled melting and ablation at the surface is one of the promising techniques to achieve superhydrophobic surfaces on ceramic and polycarbonate substrates [11,12]. Although laser textured surfaces under assisting gas environment provides superhydrophobic surface characteristics, the resulting surface properties, such as UV transmittance, do not meet requirements of practical applications for solar energy harvesting as part of solar selective surfaces or solar volumetric absorber [13]. In addition, laser texturing toward achieving surface superhydrophobicity involves with an energy intensive process [14], which is costly. However, laser textured superhydrophobic surface can be replicated using cost-effective forming process. Silicon-based organic polymers, such as polydimethylsiloxane (PDMS), can be one of the excellent candidates for effective forming process. PDMS is optically clear, inert, nontoxic, and it has nonflammable characteristics. In addition, PDMS in a liquid form has superior rheological properties, which is suitable for copying and reproducing fine-sized textured surfaces. Reconstruction of surface topology by using PDMS is reported in early studies [15,16]; however, the process of copying and replicating of surfaces with presence of micro/poles and the characteristics of the reconstructed surface require further study toward achieving exact reproduction of topology of the textured surface. This becomes necessary particularly for surface topologies with complicated geometric features, such as laser textured surfaces [11]. Considerable research studies were carried out to examine laser surface texturing and surface hydrophobicity. A study on laser texturing of alumina surface and wetting characteristics of treated surfaces was carried out by Shen et al. [17]. They indicated that laser surface wettability modification brought new applications for both laser and materials for industry. Laser gas-assisted texturing of alumina surface for improved hydrophobicity was investigated by Yilbas et al. [18]. They demonstrated that high pressure nitrogen assisting gas caused formation of AlN and AlON species at the surface, which modified the surface energy after the treatment process. The presence of AlN contributes to hydrophobicity enhancement at the surface. In addition, Wenzel and Cassie and Baxter states were present at the treated surface due to the variation in the surface texture. Effect of laser fluence on surface microstructure of alumina ceramic was examined by Harimkar and Dahotre [19]. They indicated that laser fluence influenced the microstructure in terms of changes in morphology and (1 1 0) crystallographic texture of the surface grains.

798

Hydrophobic Materials

Therefore, the microstructural observations could be used to establish the guidelines for optimizing the laser fluence to achieve the desired morphology of the surface grains and extent of texture in the surface modified alumina ceramic. Laser carbonitriding of alumina surface was carried out by Yilbas [20]. The findings revealed that high temperature gradient was developed in the irradiated region, which in turn, resulted in high residual stress levels in this region. X-ray photoelectron spectroscopy (XPS) and X-ray diffraction (XRD) data showed the presence of Al(C, N) and AlN compounds in the surface region. In addition, the microhardness in the surface region of the laser treated workpiece increased significantly. Laser gas-assisted nitriding and sol–gel coating of alumina surfaces was examined by Yilbas et al. [21]. They demonstrated that the laser treated and sol–gel coated alumina surfaces provided superior surface characteristics in the harsh environments because of weak adhesion between the mud formed from the dust particles and the coating surface. This was associated with the small texture height of the sol–gel coating, which lowered the area of the interfacial contact between the mud and the coated surface, and relatively lower surface energy of the sol–gel coating as compared to that of the laser treated surface. Laser gas-assisted nitriding of alumina surfaces was studied by Yilbas et al. [22] and they showed that two regions were formed in the laser irradiated zone. The first region below the surface was dense and composes of a-Al2O3 and AlN, while in the second region, which was below the first region, randomly stacked lamellae structure was observed. On the other hand, research into hydrophobic surface studies using PDMS gained considerable inertia in recent years. The preparation of a PDMS/CaCO3 for a superhydrophobic coating was presented by Yuan et al. [23]. They showed that contact and sliding angles of the superhydrophobic coating having the optimum ratio of PDMS and nano-CaCO3 particles could reach 160 and 3 degrees, respectively. A morphological study on direct polymer cast micro-textured hydrophobic surfaces using PDMS was carried out by Adithyavairavan and Subbiah [24]. They measured geometric texture parameters such as peak height, peak/ valley base width and interspacing between adjacent peak/valleys for the lotus leaf and its replicates, which displayed shorter peak heights and larger base widths. The replicates had contact angles (CAs) of 132.1 and 129.2 degrees, respectively, which were closest to the CA of the lotus leaf, 152.9 degrees. A thermal imprint technique for preparation of superhydrophobic polymer coatings were introduced by Lin et al. [25]. They demonstrated that all the samples using sandpaper as templates exhibited stable Cassie or metastable wetting characteristics. The rougher surfaces provided some large protrusions, which immersed into water droplets, leading to high water penetration depth and thus stickiness. A study on PDMS micro/nanohybrid surface for increasing hydrophobicity was carried out by Kim et al. [26]. They showed that the micro/nanohybrid PDMS surface had higher CA as compared to those of flat, nano and micropatterned PDMS surfaces. Robust hydrophobic surfaces with various micropillar arrays formed using PDMS was investigated by Yeo et al. [27]. They evaluated hydrophobicity of various PDMS surfaces and indicated that the apparent CA of robust hydrophobic surface monotonically increased as top surface area decreased for a given perimeter and height. Formation of superhydrophobic PDMS surfaces using ultrafast laser-induced surface modification was examined by Yoon et al. [28]. They demonstrated that negative replica of processed PDMS surface exhibited large CAH with a sliding angle of 90 degrees, while positive replica maintained superhydrophobicity. On the other hand, wetting characteristics of surfaces play an important role in many of the processes involving liquid/surface interactions [29]. Development of surfaces with enhanced wetting/non-wetting characteristics to meet required applications has received a great deal of attention in recent years both from industry and academia [30,31]. The chemical nature of a surface and its texture characteristics can both be adjusted to allow the surfaces to be used in self-cleaning [32], anti-biofouling [33], anti-smudge [34], anti-fog [35], and low-drag applications [36]. Self-cleaning surfaces with high optical transmittance are of special interest because of their wide range of applications, for example, windows for skyscrapers, protective covers for photovoltaic (PV)-panels, windshield of cars, etc. The required effort to remove the dust particles from such surfaces is considerably reduced. This, in turn, lowers the costs associated with the cleaning of conventional wetting surfaces. In order to explore the surface wetting phenomenon, it is important to highlight some of the properties that govern the wettability of surfaces. The development of superhydrophobic surfaces is a multistep process and usually involves with expensive and hazardous chemicals, longer processing times, and skilled man power. One of the facile and cost-effective techniques to obtain surfaces with excellent non-wetting properties is to replicate the texture of a superhydrophobic surface using PDMS. In this case, the textured surface can be copied and replicated several times by PDMS without any damage to the surface. In this chapter, firstly, some basic definitions and background for surface hydrophobicity are introduced. Later, two case studies are incorporated to account for the processes involved with surface hydrophobicity. In the first case study, alumina tiles which exhibit superhydrophobic characteristics after texturing by laser, and silicon micropost arrays are replicated to obtain textured PDMS surfaces. Non-wetting characteristics and transmittance of the resultant surfaces are to be presented. In the second case study, solvent induced texturing of polycarbonate surfaces are to be presented in the light of the previous study [37].

2.25.2

Basics of Wetting States

Wetting of surfaces is a complex phenomenon, influenced not only by the surface energies of the relevant phases, but by the surface texture of the solid substrate as well. In the coming sections, some theoretical background for droplet behavior on smooth and rough surfaces will be presented. For the case of droplets on rough surfaces, two different states, Wenzel and Cassie–Baxter will be discussed. It will be followed by a brief overview of lubricant impregnated surfaces.

Hydrophobic Materials

la sa

799

Air Liquid

 sl

Solid Fig. 1 Force balance of interfacial tensions at three-phase contact line.

2.25.2.1

Wetting on Smooth Surfaces

A surface’s wettability is most often expressed by the CA. It is defined as the angle formed between the solid/liquid interface and the liquid/vapor interface when a droplet is placed on a surface. This has been illustrated in Fig. 1. The CA is dictated by the equilibrium of the interfacial tensions at the three-phase contact line, i.e.: gla cosy þ gsl ¼ gsa

ð1Þ

where gla, gsl, and gsa are liquid–air, solid–liquid, and solid–air interfacial tensions, respectively. The above equation can also be expressed as: cosy ¼

gsa

gsl

ð2Þ

gla

The above equation is called as Young’s equation of CA and is applicable to smooth surfaces [38]. gsl in the above equation can be estimated as [39] pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð3Þ gsl ¼ gsa þ gla 2 gsa  gla

Surfaces with water contact angle (WCA) greater than 90 degrees are called as hydrophobic and those having WCA less than 90 degrees are called as hydrophilic. According to Eqs. (2) and (3), a surface can be made hydrophobic by decreasing the solid–air interfacial tension. The lowest reported solid–air interfacial tension is for surfaces with trifluoro methyl (–CF3) groups (6 mN m 1) [40] and the corresponding WCA observed on such surfaces is 120 degrees. A further increase in CA requires texturing of the surfaces.

2.25.2.2

Wetting on Rough Surfaces

There are two primary wetting regimes on a rough surface: 1. Wenzel regime or homogeneous regime. 2. Cassie–Baxter regime or nonhomogeneous/composite regime. Assuming that the liquid completely wets a surface, Wenzel developed a model predicting the relation between the CA of a droplet on a rough surface to that on the smooth surface [41]. cosyrough ¼ rcosyflat

ð4Þ

where yrough and yflat are the CAs of the droplet on rough and flat surfaces, respectively. r is the roughness ratio and is defined as the ratio of surface area to the flat projected area. Now if a surface is hydrophobic to start with, i.e., yflat is greater than 90 degrees, then introducing roughness will make it more hydrophobic (since r is always greater than 1). New CA yrough will be greater than yflat. If a surface is hydrophilic to start with, i.e., yflat is less than 90 degrees, then introducing roughness will make it more hydrophilic. New CA yrough will be less than yflat. This has been illustrated in Fig. 2. The Wenzel state for a droplet on a rough surface is shown in Fig. 3. The Wenzel state dictates that the liquid will completely fill the texture of a solid surface. However, complete submergence of texture with the liquid becomes less energetically favorable for a hydrophobic surface. This is because the system is relatively at a higher energy when the liquid is fully wetting the texture. More stable state is attained when air pockets are formed in between the texture. Cassie–Baxter extended Wenzel’s work and incorporated the effect of trapped air pockets resulting in a composite solid–liquid–air interface as opposed to homogeneous solid–liquid interface [42]. cosyrough ¼ rfs cosyflat þ fA cosyLA

ð5Þ

where fs and fA being the fractional liquid–solid and liquid–air interfacial areas (fs þ fA ¼ 1). Since cosyLA ¼ 180 degrees (CA of water with air), entrapment of air pockets will result in an increased CA. Putting cosyLA ¼ 180 degrees, the above equation reduces to: cosyrough ¼ rfs cosyflat

ð1

fs Þ

ð6Þ

Based on Eq. (6), a decrease in the value of fs can increase the value of yrough and a surface can be made superhydrophobic (CA greater than 150 degrees). The Cassie–Baxter state for a droplet placed on a rough surface is shown in Fig. 4.

800

Hydrophobic Materials

180 flat=150 degrees

160

flat=120 degrees

140

rough

120 100

flat=90 degrees

80 60

flat=60 degrees

40 20 0 1.0

flat=30 degrees 1.1

1.2

1.3 1.4 1.5 1.6 Roughness ratio ‘r ’

1.7

1.8

Fig. 2 Evolution of contact angle (CA) with roughness ratio “r” for a surface. Reproduced from Nosonovsky M, Bhushan B. Roughness-induced superhydrophobicity: a way to design non-adhesive surfaces. J Phys Condens 2008;20:225009.

Air Droplet

Rough solid Fig. 3 Wenzel state for a liquid droplet placed on a rough surface.

Air Liquid

Rough solid Fig. 4 Cassie–Baxter state for a liquid droplet placed on a rough surface.

Eq. (6) can also be expressed in terms of fA, where fA is the fractional liquid–air interfacial area: cosyrough ¼ rcosyflat

fA ðrcosyflat þ 1Þ

ð7Þ

The evolution of CA with roughness ratio r and the liquid–air fraction area fA on surfaces for which yflat is 120 and 150 degrees has been shown in Fig. 5. For both of the surfaces (yflat ¼120 degrees and yflat ¼ 150 degrees), an increase in CA is brought about by an increase in both, the roughness ratio r and the liquid–air fractional area fA.

2.25.2.3

Contact Angle Hysteresis

In general, two types of CA measurements are performed: static and dynamic. Static CA is measured by placing the droplet on a flat surface and then recording the value of the CA once an equilibrium is established. Dynamic CAs are measured during the droplet growth and shrinkage. These are nonequilibrium CAs. Dynamic CA measurement is also performed by tilting the surface and noting the values of the CAs at the front and the back of the droplet. These CAs are called as advancing and receding CAs and the difference between their values is called as CAH. CAH is highest just before the droplet starts sliding/rolling on the surface. CAH is a measure of how much energy is dissipated during droplet flow [43]. For surfaces with low CAH, water rolls-off at very small tilt angles and carries away dust particles along its path, giving rise to the so called self-cleaning effect [43].

Hydrophobic Materials

801

180

A=0.6 170

A=0.4 A=0.6

A=0.2

160

rough

A=0.8

A=0.2

A=0

150

A=0

140 130

flat=120 degrees

120 110

flat=150 degrees 1

1.2

1.4 1.6 Roughness ratio ‘r ’

1.8

2.0

Fig. 5 Evolution of contact angle (CA) with roughness ratio “r” and liquid–air fractional area fA for surfaces for which yflat is 120 and 150 degrees. Reproduced from Nosonovsky M, Bhushan B. Roughness-induced superhydrophobicity: a way to design non-adhesive surfaces. J Phys Condens 2008;20:225009.

Using Eq. (7), a relation for the CAH for a rough surface can be derived. The resultant equation is shown below. cosyrough adv

cosyrough rec ¼ rð1

fA Þðcosyflat adv

cosyflat rec Þ þ Hr

yflat adv and yflat rec represent the CAH for the smooth surface and Hr represents the surface roughness effect. The above equation can be simplified to obtain the following expression [44]: pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi cosyrec flat cosyadv flat yrough adv yrough rec ¼ r 1 fA pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2ðrcosyflat þ 1Þ

ð8Þ

ð9Þ

For a homogeneous or Wenzel interface, fA ¼ 0. Increasing roughness ratio r for such a case would increase CAH. For a composite or Cassie–Baxter interface, CAH is directly proportional to (1 fA), i.e., a decrease in liquid–air fractional area will increase the CAH. This brings about the importance of microair pockets in the texture which increase the CA and lower the CAH for a surface. Presence of a nanoscale texture on top of a microscale texture can also significantly increase the liquid–air fractional area and lower the CAH. A nanoscale texture thus prevents the pinning of the droplet and allows it to roll-off at small inclination angles, thereby yielding the self-cleaning effect.

2.25.2.4

Lubricant Impregnated Surfaces

Conventional superhydrophobic surfaces rely on air pockets trapped between the texture to maximize the droplet CA with the surface. Such surfaces continue to exhibit non-wetting behavior as long as stable air pockets are maintained beneath the droplet [45]. These tiny air pockets, however, are unstable and collapse under conditions involving large wetting pressure, high temperature or humidity [46,47], or when a damage occurs to the surface texture [48]. A low surface tension liquid can also sometimes displace these air pockets and penetrate the texture [49]. In all such cases, the droplet pins to the surface. The surface thus loses its self-cleaning ability. To overcome the limitations associated with lotus leaf inspired surfaces, a new type of pitcher-plant (Nepenthes) inspired surfaces called slippery lubricant infused porous surfaces (SLIPS) have been reported [50]. These surfaces do not rely on trapped air pockets inside the texture to repel liquids. The texture is instead filled with a lubricant, thus providing a surface with an overlying liquid interface that is ultrasmooth, chemically homogeneous, continuous and provides an extremely low CAH for a broad range of liquids [50]. Surfaces with high optical transmittance find their use in wide range of applications, such as in construction industry, automobile industry, optoelectronics, and energy conversion devices (PV panels). In all such applications, maintaining high optical transmittance is important for esthetic and performance reasons. Surfaces are prone to lose their transmittance due to accumulation of air borne dust particles at the surface. Maintaining optical clarity of the surfaces thus becomes challenging especially at large scales, such as skyscraper windows and protective covers for PV cells in solar farms. All such cases bring about the importance of optically transparent surfaces that are cost-effective to clean. One of the promising solutions is to alter the wetting characteristics of a surface and make it superhydrophobic. Natural superhydrophobic surfaces like lotus plant leaves can remain clean since water droplets roll away at small inclination angles and take away dust particles along its path. This gives rise to the so called lotus effect or self-cleaning effect (Fig. 6). A superhydrophobic surface, thus, can clean itself by using very little quantity of water in contrast to the conventional wetting surfaces. Glass is an excellent choice of material in applications requiring high transmittance. Due to its hydrophilic nature, water droplets stick to the glass surface. These water droplets catch the dust particles and form a mud solution at the surface which greatly

802

Hydrophobic Materials

Water droplet

Dust particles

Fig. 6 Lotus or self-cleaning effect whereby the dust particles are carried away by the rolling droplet on a superhydrophobic surface.

lowers its transmittance. This can be avoided by coating the glass surface with a superhydrophobic coating. Silica nanoparticle coating is an ideal choice because of its matching refractive index with glass which greatly improves the optical transmittance of the coating. Silica nanoparticle-based coatings also possess good scratch and wear resistant properties [51]. Another facile approach to obtain optically transparent superhydrophobic coatings is to replicate textured surfaces having superhydrophobic characteristics. PDMS, a silicon-based organic polymer, is an ideal choice because of its high transparency, hydrophobic nature, and the ability to copy submicron features during the replication process. Filling the texture of a superhydrophobic surface with some lubricant oil can overcome many limitations associated with conventional superhydrophobic surfaces. A lubricant oil provides a smooth and homogeneous interface that can repel variety of fluids give very low CAH and improve the optical transparency of the coatings [50,52].

2.25.2.5

Particle Adhesion on Surfaces

Dust settlement at the surfaces gives rise to adhesion of dust particles and they have detrimental effects on solar energy harvesting devices due to optical scattering and absorption of solar radiation at the surface. On the other hand, several methods are introduced to remove the dust particles from the surface. A simple and traditional cleaning method is done by using a mop and water supply; a method that is labor intensive. Pressurized water can be used to avoid moping the surfaces while minimizing mechanical scratches formed at the surface. This process increases the process time and lowers the cleaning efficiency; however, it enables to clean the spots those are difficult to be cleaned by a mop. The steam and compressed air can be used to remove the dust accumulation from the surfaces. In addition, there are methods where the need for the cost of labor is minimized. Some of these include robot device, automatic brushes, and sprinkle water. Introducing electric current to remove and repel the dust particles from the surface is an alternative method to be implemented for surface cleaning. Utilizing mechanical vibration or centrifugal forces are examples of incorporating the dynamic methods toward cleaning of surfaces from the contaminations, such as environmental dusts and small size waste particles. In order to assess the dust particles adhesion, understanding of the contact mechanics associated with the particles becomes essential. In general, the adhesion of dust particle is affected by many factors, such as van der Waals force, static electric charge, relative humidity, size of the contact area, roughness of the surface, chemistry of particle agglomeration, duration of contact, local temperature, and other factors. In the case of the influence of van der Waals forces, it is assumed that van der Waals forces are the most common forces that contributes to the particle adhesion on the surface. The models developed describing adhesion of particles on surfaces are briskly introduced below.

2.25.2.5.1

Johnson–Kendall–Roberts model

The fundamental understanding of adhesion of spherical particles on the flat surfaces is introduced incorporating van der Waals forces, which is known as Johnson–Kendall–Roberts (JKR) model for adhesion, i.e.: F¼

3 pRg 2

ð10Þ

where R is radius of the particle and g is surface energy between two surfaces. However, the above equation is applicable for large size soft bodies having the high surface energies.

Hydrophobic Materials 2.25.2.5.2

803

Derjaguin, Muller, Toropov model

Derjaguin, Muller, Toropov (DMT) model for adhesion is similar to that of JKR model; however, the resulting force equation uses the different constant, i.e.: F ¼ 2pRg

ð11Þ

The force equation introduced above is appropriate for small size hard solid particles with low surface energies.

2.25.2.5.3

Hamaker’s model

One of the important adhesion force equation incorporates the Hamakar’s model. The force balance relies on the vertical forces and it is schematically shown in Fig. 7. The resulting equation is: F¼

AR 12z2

ð12Þ

where A is Hamaker constant and zo is separation distance between particle and the flat surface, which is in the order of 0.3 or 0.4 nm [53]. Hamaker’s model did not consider the particle contact area to the surface.

2.25.2.5.4

Rumpf–Rabinovich model

This model utilizes Van der Waal’s forces acting on a particle when located on the rough surface. It includes the statistical distribution of surface asperities and the effect of root mean square (RMS) of the rough surface on the adhesion force. Fig. 8 shows the schematic view of the particle and the surface. The resulting adhesion force takes the following form: AR F¼ 12z2

1 1 þ 1:48RRMS

þ

1 2 1 þ 1:48z2RMS

!

ð13Þ

In general, the equations developed determining the adhesion force can be used according to the surface texture characteristics. However, equations based on JKR and DTM models are the only equations incorporating the free energy of the surfaces. One of the methods to remove the dust particles is to utilize the centrifugal acceleration at the surface via rotation of the surface. In this case, various forces should considered for the dust removal from the surface through rotating the surface. These forces are listed below:

R 1 2

z

Fig. 7 Force balance for hard solid particle. Here, g is the surface energy of mediums, and z is the minimum spacing of contacting particle from the flat surface.

R

z

Fig. 8 A schematic view of the particle and the surface.

804

Hydrophobic Materials

2.25.2.5.5

Gravitational force

One of the forces contributing to the particle adhesion onto the surface is gravitational force which can be written as: Fg ¼ mg ¼

4 3 pR rg 3

ð14Þ

where m is mass of the particle, r is density of the dust particle, and g is gravitational acceleration.

2.25.2.5.6

Lift force (inertial force and shear stress)

Lift force can be generated by inertial force during the rotation of the solid body, such as dust particles. The inertial lift force can cause torque acting on the particle at the point contact between particle and the center of rotation. In addition, other type of lift force is also generated because of the flow shear stress acting on the particle. The combination of these forces gives rise to lifting of the particle from the surface. The lift force can be derived from the Navier–Stoke equation for a spherical shape particle where particle is almost touching the surface for the case where the Reynolds number remains below unity. Therefore, the lift can be written as: FL ¼

9:22m2   3 Re r

ð15Þ

where m is dynamic viscosity of fluid and Re is Shear Reynolds number which is defined by: Ru v

ð16Þ

sffiffiffiffiffiffiffi 2to r

ð17Þ

Re ¼ where (n) is kinetic viscosity and u is friction velocity: u ¼

where to is the shear stress at the wall. The shear stress of rotating disk system incorporating the simplified form of the Navier–Stoke equation can be written as: 0 pffiffiffiffiffiffiffiffi ð18Þ to ¼ rrGο vo3 0

where to is shear stress at the wall (y¼ 0), r is the position of particle from the center of rotation, Go dimensionless constant at the 0 wall (Gz ¼ 0 ¼ 0:61592) and o is the angular velocity of the disk.

2.25.2.5.7

Drag force due to pressure and shear on the particle surface

In general drag force can be categorized into three groups. The first group is pure pressure drag force when the area normal to the flow is relatively large, the second group is involved with the pure shear friction drag force, which is resulted due to friction, and the last group is associated with the combination of pressure and shear drag forces. The drag force acting on the spherical particle can be considered as the third type, which involves with the combination of pressure and shear drag force. The drag force for the spherical solid body can be written as: FD ¼ 10:2pmRu

ð19Þ

where m is the flow velocity.

2.25.2.5.8

Centrifugal force

This is related to the rotational acceleration of the particle during rotation. For the spherical particle, it takes the form: F ¼ mo2 r ¼

2.25.2.5.9

4 3 2 pR ro r 3

Friction force

This force is associated with friction factor between the particle and the surface and it can be written as:   Ff ¼ mN ¼ m Fa þ Fg Fl

2.25.2.5.10

ð20Þ

ð21Þ

Coriolis force

Coriolis force is the force that acts on reverse direction of the main flow direction. It can be written as: F ¼ 2mo2 r ¼

8 3 2 pR ro r 3

here, o is the rotational speed, R is the particle diameter, and r is the distance of the particle from the rotational center.

ð22Þ

Hydrophobic Materials

2.25.3

805

Surface Treatment Methods for Hydrophobicity

There are several methods considered to create hydrophobic characteristics of the surfaces. In general, surface hydrophobicity is associated with low surface energy and surface texture composing of micro/nanopillars. The processes related to the surface treatment process are given below under the appropriate subsections.

2.25.3.1

Laser Texturing of Surfaces

Laser surface texturing through combination of controlled melting and ablation at the surface is one of the promising techniques to achieve superhydrophobic characteristics on ceramic and polycarbonate substrates [11]. Considerable research studies were carried out to examine laser surface texturing and surface hydrophobicity. Laser texturing of a zirconia surface under nitrogen environment in the presence of TiC and B4C was studied by Yilbas et al. [54]. TiC and B4C particles were placed in a thin carbon film at the surface. The carbon film distributed the particles evenly at the surface and enhanced the absorption of the incident laser beam. The carbide particles enhanced the surface roughness without forming any cracks due to the generation of thermal stress arising from a mismatch between thermal expansion coefficients. The laser textured surface comprised of micro/nanopoles and cavities and showed hydrophobic characteristics with an average CA of 130 degrees. Laser ablation of rare-earth oxide ceramics toward achieving superhydrophobic characteristics was carried out by Azimi et al. [55]. A hierarchal texture consisting of micro and nanostructures was obtained after laser texturing. The texture consisted of micron-sized mud-crack like features covered with nano-sized protrusions. The surface showed exceptional water repellency causing the impinging water droplets to bounce off. Triantafyllidis and Stott [56] studied surface treatment of alumina-based ceramics with the use of combined laser sources. They demonstrated that the use of combined laser sources altered the cooling rates thus avoiding the buildup of large thermal stresses. This resulted in a crack free laser treated surface. Surfaces treated with a single laser beam, on the other hand, showed solidification cracking. Laser gas-assisted nitriding of alumina surfaces was studied by Yilbas et al. [22]. They showed that the depth of laser treated zone extended almost 40 mm below the surface and composed of two regions. The first region was dense and composes of a-Al2O3 and AlN. In the second region, randomly stacked lamellae structure was observed. The laser treated surface showed uniform features, provided that some locally scattered submicron cracks were observed. In another study, laser texturing of alumina tiles under high pressure nitrogen environment was carried out by Yilbas et al. [18] to enhance the surface hydrophobicity. Laser texturing resulted in the formation of fine-sized micro and nanotextures at the surface. The surface energy was also modified due to the formation of AlN and AlON at the surface. Two different states, Wenzel state and Cassie–Baxter state, existed on the surface depending upon the surface roughness. Laser micromachining of thin alumina sheets was introduced as a one-step process to generate superhydrophobic characteristics by Jagdheesh [57]. The effect of laser processing parameter (pulses per unit area) on the change in surface morphology and chemistry was studied. It was shown that changes in morphology influence the wetting properties more than chemical changes at the surface. Laser surface texturing of 304S15 stainless steel to develop superhydrophobic characteristics was investigated by Ta et al. [9]. Instead of common pico/femtosecond laser systems employed for texturing, compact, and cost-effective nanosecond fiber laser system was used. Surfaces showed hydrophilic characteristics immediately after laser irradiation. When placed in ambient air for over a period of 13 days, surfaces began to demonstrate superhydrophobic characteristics. Optimized laser processing parameters gave a WCA of 154 degrees and a CAH of 4 degrees. Laser patterning of copper and brass surfaces using infrared nanosecond laser for enhancing surface hydrophobicity was studied by Ta et al. [58]. Wetting characteristics of laser irradiated surfaces changed from hydrophilic to superhydrophobic with time. This was associated with the partial deoxidation of the oxides formed as a result of laser irradiation. Surfaces showed superhydrophobic characteristics with small tilt angles after an extended period of time (around 11 days). Laser irradiated copper and brass surfaces also demonstrated self-cleaning ability. These surfaces also showed the potential to be used as sensors in chemical sensing applications.

2.25.3.2

Modified Silica Nanoparticle Deposition

The superhydrophobic surfaces can be generated through silica nanopartilces deposition at the surface, which in turn generates the surface texture composing of hierarchical structures [59]. In this case, the silica-PDMS film is deposited by using simple immersion method for 10 min [59]. Two kinds of silica particles are used which have 7 and 14 nm sizes. The concentration of silica particle is also varied to observe the effect to WCA. Fig. 9 shows water droplet CA for various silica nanoparticle sizes and concentrations. The hysteresis CA also shows similar trend that 14 nm particle sizes has 10 degrees and 7 nm particle size has 40 degrees while the combination of 14 and 7 nm results in 30 degrees. From Fig. 10, silica particles having 14 nm sizes result in higher CA than those of 7 nm particle size. This is due to the morphology and distribution of surface-structures (Fig. 9). Fig. 10 shows scanning electron microscope (SEM) micrographs of silica particles at the PDMS surfaces with different particles size. It is evident from SEM micrographs that silica particle-PDMS

806

Hydrophobic Materials

(A)

(B)

(D)

(C)

(E)

(F)

Fig. 9 Water contact angle (WCA) measurement of silica-polydimethylsiloxane (PDMS) surfaces with different particle size and concentration: (A) 7 nm and 0.5% w v 1: 93 degrees, (B) 7 nm and 1.0% w v 1: 99 degrees, (C) 7 nm and 2.0% w v 1: 130 degrees, (D) 14 nm and 0.5% w v 1: 95 degrees, (E) 14 nm and 1.0% w v 1: 103 degrees, and (F) 14 nm and 2.0% w v 1: 147 degrees. Reproduced from Gao N, Yan YY, Chen XY, Mee DJ. Superhydrophobic surfaces with hierarchical structure. Mater Lett 2011;65:2902–5.

AccV Spot Magn 10.0 kV 3.0 500×

Det WD SE 10.0

100 µm

(A)

AccV Spot Magn 10.0 kV 3.0 500×

(D)

AccV Spot Magn Det WD 10.0 kV 3.0 5000× SE 10.0

10 µm

AccV Spot Magn Det WD 10.0 kV 3.0 10000× SE 10.0

(B)

Det WD SE 10.1

100 µm

AccV Spot Magn Det WD 10.0 kV 3.0 5000× SE 9.9

5 µm

(C)

10 µm

AccV Spot Magn Det WD 10.0 kV 3.0 10000× SE 9.9

(E)

5 µm

(F) 1

Fig. 10 SEM images of silica-PDMS surfaces with (A) 14 nm silica particle and 2.0% w v silica concentration and (B) 7 nm silica particle and 2.0% w v 1 silica concentration. SEM, scanning electron microscope; PDMS, polydimethylsiloxane. Reproduced from Gao N, Yan YY, Chen XY, Mee DJ. Superhydrophobic surfaces with hierarchical structure. Mater Lett 2011;65:2902–5.

surfaces with 7 nm have more grove like structures and irregularity while minimizing the amount of air pocket to maintain the water droplet at high CA. In addition, coating method could be used to deposit silica particles to generate hydrophobic surfaces [60]. Fig. 11 shows the schematic view of the processes involved in drop coating technique. Firstly, octadecyltrichlorosilane (OTS) is grown onto silica

Hydrophobic Materials

807

500 nm M-SiO2 In ethanol Drop coating Dried

PDMS layer Dried Cured Drop coating

Drop coating

100 nm M-SiO2 In ethanol

PDMS in hexane

Fig. 11 Illustration of drop coating procedure of superhydrophobic PDMS-based surface. PDMS, polydimethylsiloxane. Reproduced from Ke Q, Fu W, Jin H, Zhang L, Tang T, Zhang J. Fabrication of mechanically robust superhydrophobic surfaces based on silica micro-nanoparticles and polydimethylsiloxane. Surf Coatings Technol 2011;205:4910–4.

142 degrees

WZU

(A)

155 degrees

147 degrees

SEI

10.0 kV ×10,000

1 µm

WD5.7 mm

WZU

(B)

SEI

10.0 kV ×10,000

1 µm

WD7.7 mm

WZU

SEI

10.0 kV ×10,000

1µm

WD5.7 mm

(C)

Fig. 12 Top side: SEM image PDMS/silica coating on glass substrate using (A) unmodified 500 nm silica particle (B) modified 500 nm silica particle; bottom side: SEM image and WCA of PDMS/M-silica at different curing temperature: (A) 601C, (B) 1001C, and (C) 1401C. SEM, scanning electron microscope; PDMS, polydimethylsiloxane; WCA, water contact angle. Reproduced from Ke Q, Fu W, Jin H, Zhang L, Tang T, Zhang J. Fabrication of mechanically robust superhydrophobic surfaces based on silica micro-nanoparticles and polydimethylsiloxane. Surf Coatings Technol 2011;205:4910–4.

particle surface. PDMS elastomer is prepared with the weight ratio of curing agent over PDMS 1:10. Later, solution of 500 nm modified silica particle in ethanol is dropped onto the glass surface. Once the surface is dried, additional 100 nm modified silica particles are dropped onto the coated glass surface. Upon drying of the surface it is, then, cured at 1001C for 24 h. Finally, PDMS layer is dropped onto the surface. Modified silica particles having OTS layer on the surface are shown in Fig. 12 for different curing temperatures. The hierarchical structures forming superhydrophobic characteristics at the surface can be achieved using PDMS casting and laser etching [61]. The procedure of this method is illustrated in Fig. 13. In this case, the micropillar is formed via PDMS casting and the sub-microroughness is achieved by laser etching technique.

2.25.3.3

Polydimethylsiloxane Replica Molding

Studies related to PDMS-based hydrophobic surfaces gained considerable attention in recent years. PDMS is optically clear, inert, nontoxic, and it has nonflammable characteristics. In addition, PDMS in liquid form has superior rheological properties, which is suitable for copying and reproducing fine-sized features. Research into replica molding using PDMS gained considerable attention in recent years [62–64]. Shao et al. [62] successfully replicated different micropatterns with very high aspect ratios using PDMS replica molding technique. The challenging part was the release of the PDMS from the mold. In order to facilitate the release of negative PDMS replica from the SU-8 master mold, the

808

Hydrophobic Materials

PC nanopillars

PDMS PDMS solution

(A)

PDMS

PC nanopillars

PDMS

PDMS nanopillars

(B)

(C) Fig. 13 (A) The schematic process of lase etching, (B) the schematic process of PDMS casting, and (C) PDMS nanopillars after etching. PDMS, polydimethylsiloxane. Reproduced from Jin M, Feng X, Zhai J, Cho K, Feng L, Jiang L. Super-hydrophobic PDMS surface with ultra-low adhesive force. Macromol Rapid Commun 2005;26:1805–9.

master mold was first treated with hexamethyldisilazane (HMDS) vapor. Similarly, the PDMS negative mold was treated with 1H,1H,2H,2H-perfluorodecyltrichlorosilane (PFDTS) vapor in order to ease the release of replicated PDMS from negative PDMS mold. Preventing PDMS–PDMS bonding is a major challenge in replica molding process. By coating the negative PDMS mold with hydroxypropylmethylcellulose, a hydrophilic polymer, Yang et al. [63] and Gitlin et al. [64] were successfully able to replicate different microstructures by PDMS replica molding process. Yeo et al. [65] used PDMS double casting technique to replicate micropatterned silicon SU-8 master molds. Micropatterns with circular, triangular, and cross shapes were successfully replicated and the wettability of the resultant surfaces was analyzed. The experimental WCA matched with the theoretical Wenzel and Cassie–Baxter CAs. It was also found out that surfaces with higher aspect ratio enhanced the hydrophobicity of the surfaces owing to the increased surfaces roughness and formation of large air pockets beneath the water droplet. Adithyavairavan and Subbiah [24] replicated the lotus leaf microstructures using two-step polymer casting technique. By using three different polymers, vinyl polysiloxane (VPS), PDMS and polymethylmethacrylate (PMMA), a total of six negative–positive mold combinations were studied. The lotus leaf replicate obtained from the VPS–PMMA mold, the texture geometric parameters, such as peak height, peak width, and the interspacing between the peaks were closest to that of the actual lotus leaf. The CA measured for VPS–PMMS replicate was 132.1 degrees while that of the actual lotus leaf was 152 degrees. The reason for the difference between the CAs was attributed to the absence of hydrophobic wax that not only covers the lotus leaf microstructure but gives it a nanotexture as well. A thermal imprint technique for preparation of superhydrophobic polymer coatings was introduced by Lin et al. [25]. A mold was first prepared by casting PDMS over the sand papers. The prepared PDMS molds were superhydrophobic and showed CAs as high as 170 degrees. The pattern of the PDMS molds was then transferred to the commercial polymer coatings using thermal imprint method. The non-wetting characteristics of the polymer coatings were enhanced and WCAs of around 150 degrees were observed. The water droplets, however, stuck to the surface due to the presence of large features protruding from the surface. Laser irradiation of PDMS to generate superhydrophobic characteristics was studied by Yoon et al. [28]. PDMS surfaces, irradiated by a femtosecond layer, showed a WCA of 170 degrees and a sliding angle of 2 degrees. Superior superhydrophobic characteristics of the laser irradiated surface were associated with the presence of micro and nanostructures, very similar to the ones observed in lotus leaf. The negative PDMS replica of the laser treated surface had a smaller CA with very large hysteresis. This was due to the absence of the micro and nanosize features that are emerging from the surface for laser treated surface. This caused the water droplets to be in contact in Wenzel state rather than in Cassie–Baxter state. The positive PDMS replica, on the other hand, had a high WCA (150 degrees) and small sliding angles, similar to that of laser treated surface.

2.25.3.4

Nanoparticle Film Forming

Silica nanoparticles have been extensively used to give wetting/non-wetting characteristics to a variety of substrates. By functionalizing the nanoparticles with appropriate functional groups, superhydrophilic/superhydrophobic or omniphilic/omniphobic surfaces could be realized. Stober et al. [66] pioneered the technique for the synthesis of spherical silica nanoparticles using tetraethyl orthosilicate (TEOS), ammonium hydroxide, ethanol, and water. The effect of concentration of TEOS, water, and ammonium hydroxide on the mass fraction and final particle size has been extensively studied [67–69]. One of the interesting properties of silica nanoparticles is that its surface can easily be functionalized with some desired functional group. Suratwala et al. [70] modified the surface of silica sols by reaction with HMDS and ethoxytrimethylsilane (ETMS) to produce hydrophobic trimethylsily functionalized silica

Hydrophobic Materials

809

nanoparticles. The percentage of functionalization varied between 5% and 33% and was strongly dependent upon the starting nanoparticles chemistry, HMDS concentration, and reaction time. Yong et al. [71] used one pot technique to prepare spherical monodispersed-functionalized silica nanoparticles directly from TEOS. The surfaces of silica nanoparticles were modified and made hydrophobic by reaction with organosilanes, such as 3-aminopropyltrimethoxysilane (APTMS), 3-isobutyltrimethoxysilane (ITMS), and octyltriethoxysilane (OTES). All the chemicals, including the modifier silane, were mixed in one pot to obtain octyl-, minopropyl-, and isobutyl-functionalized silica nanoparticles. By using a layer-by-layer (LBL) technique, Bravo et al. [72] fabricated a multilayered silica nanoparticle superhydrophobic film on glass. The glass slides were dipped repeatedly in different solutions to create an adhesion layer, followed by the body and the top layer. A hierarchal texture was created by using 50 and 20 nm silica nanoparticles in the body layer and 20 nm particles in the top layer. The number of layers were optimized for the highest optical transmittance (around 90%) and maximum WCA (161 degrees). The procedure adopted was very tedious and involved a lot of chemicals. Ling et al. [73] employed a simple dip coating technique to obtain a silica nanoparticle coating on the glass slide. Prior to dip coating, an amine terminated self-assembled monolayer was grown on the surface of glass. It enabled the silica nanoparticles to leave large gaps in between while forming a coating at the glass surface. Static WCA of 152 degrees was observed after treatment with trichloro (1H,1H,2H,2H-perfluorooctyl) (PFOTS). The CAH, however, was relatively higher, averaging around 25 degrees. The UV/Vis spectrum showed a slight decrease in transmittance at smaller wavelengths. Li et al. [74] fabricated a transparent superhydrophobic coating on glass by LbL deposition of poly allylamine hydrochloride (PAH)/SiO2 nanoparticle films on top of PDDA-silicate film. It was followed by calcination and chemical vapor deposition (CVD) of fluoroalkylsilane. A total of four cycles of PAH/SiO2 deposition yielded a static WCA of 157 degrees. By using spray coating technique, Ogihara et al. [75] formed a hydrophobic silica nanoparticle coating on the glass surface. Functionalized silica nanoparticles suspended in propanol were used for the coating purposes. The obtained coating consisted of very fine micro and nanostructures and yielded a high optical transparency. The effect of surface treatment of the glass on the hydrophobic characteristics of the resultant surfaces was also investigated. For the untreated glass surfaces, the silica nanoparticle films did not show superhydrophobic characteristics. This was due to the high surface energy of the glass surfaces which caused the water droplets to stick to it. For the glass surfaces treated with fluorine groups, formation of a uniform silica nanoparticle film was difficult because of the poor wettability of the glass. Glass surfaces treated with dodecyl groups were most suited for forming a transparent superhydrophobic silica nanoparticle film.

2.25.3.5

Lubricant Impregnated Surfaces

Filling the texture of a superhydrophobic surface with some lubricant oil is a relatively new concept and such surfaces have gained quite an attention in recent years [50,76,77]. Traditional superhydrophobic surfaces rely on the presence of air pockets to repel the droplets. The air pockets, however, are unstable and a water droplet can easily penetrate the texture and be in the Wenzel state. This situation is avoided in lubricant infused surfaces by selecting a lubricant oil that completely fills the texture and provides an interface that is smooth and chemically homogeneous. Lubricant infused surfaces, however, become more complex due to the introduction of an additional phase and such a system has far more possible thermodynamic states which may exist when a droplet is placed on such surfaces. Smith et al. [78] investigated these states and the criterion under which one thermodynamic state may be more favorable. It was shown that a total of 12 different thermodynamic states are possible in these complex four phase systems. It is desirable for lubricant impregnated surfaces to have a stable lubricant film beneath as well as outside the droplet. Such thermodynamic state is possible if the following two conditions are satisfied: yos(a) ¼ 0 and yos(w) ¼ 0, i.e., the CA of oil(o) on solid(s) in the presence of air(a) and water(w) is zero. In case of a nonzero CA, complete encapsulation of texture by the lubricant ceases to occur. A critical CA based on the texture parameters has been defined as yc ¼ cos

1

1 r

f f

ð23Þ

where “r” is the roughness ratio and f is the projected area of the surface occupied by the solid. A surface for which 0oyos(a)oyc, the fractional area of the surface, as defined by f, will not be covered by thin lubricant film. It will instead be exposed to air/vapor phase. Surfaces for which yos(a)4yc, the liquid will displace the lubricant and get impaled by the texture. Similar arguments can be made when 0oyos(w)oyc. The surfaces features, as defined by f, will not be covered by lubricant but rather by the liquid. When yos(w)4yc, the liquid displaces the lubricant and completely wets the texture as in Wenzel state. Kim et al. [79] investigated the effect of surface roughness on lubricant loss under high shear conditions and found out that nanostructured surfaces were best at retaining a thin layer of lubricant because of the capillary forces. It was also concluded that two-tiered roughness was not necessary and that a single level of roughness was sufficient enough to hold the lubricant in place. Smith et al. [78], however, reported that the presence of a nanostructure on top of a microstructure significantly reduced water droplet pinning for surfaces for which the emergent texture features were not covered by a lubricant film. Anand et al. [80] studied droplet condensation on lubricant impregnated surfaces and found out that the droplets stayed mobile and did not pin to the substrate. Surfaces for which the lubricant was not fully encapsulating the texture, a nanotexture allowed the droplets to move with significant velocities. This provided fresh nucleation sites for the vapors to condense, grow, and subsequently be removed from the

810

Hydrophobic Materials

surface. In case of conventional superhydrophobic surfaces, nucleation of droplet starts within the texture itself. The droplet is thus impaled by the texture and does not stay mobile. Kim et al. [81] demonstrated that lubricant infused surfaces can be used as ice-repellent coatings because of their enhanced condensate removal abilities. Aluminum was used for study because of its wide applicability in transportation and construction industry. After forming a texture on aluminum surface, it was infused with a lubricant. They demonstrated that the lubricant infused aluminum surfaces resisted ice-formation at subzero temperatures. In this case, the condensating water droplets stayed mobile and did not pin to the surface. Such surfaces also showed very low ice-adhesion and allowed its easy removal once formed at the surface.

2.25.4

Case Study for Laser Texturing and Polydimethylsiloxane Replication of Textured Surface

Laser treatment of alumina surfaces and surface hydrophobicity was studied earlier [18] and laser processing was found to be inefficient in terms of processing cost despite the achievement of scratch resistant and hard surface textures with hydrophobic characteristics. One of the cost-effective processes realizing a hydrophobic surface is to copy and replicate laser textured surfaces using PDMS. In this case, laser textured alumina surface can be copied and replicated several times by PDMS without losing surface texture characteristics. Although PDMS copied and deposition of the functionalized silica particles at the surface was studied earlier [82], the fundal texture characteristics of the laser treated and PDMS replicated surfaces are left for future study. Consequently, in the present study, laser texturing of alumina surface is carried out and surface texture features are analyzed using the analytical tools. Laser textured surfaces are copied and replicated by PDMS using the dip coating technique. The resulting textures on PDMS wafer surfaces are characterized incorporating the analytical tools including SEM and atomic force microscope (AFM), Fourier transform infrared spectroscopy (FTIR), XRD, optical transmittance, and CA measurements using sessile droplet method. Scratch tests are carried out on the copied and replicated PDMS surfaces using the micro/nanotribometer.

2.25.4.1

Experimental

Alumina (Al2O3) tiles with 3-mm thickness were used as workpieces. The CO2 laser (LC-ALPHAIII) was used to irradiate the alumina tile surfaces. The nominal output power of the laser was 2 kW and the irradiated spot diameter at the workpiece surface was about 200 mm. High pressure nitrogen gas jet emerging from the conical nozzle was utilized during the laser heating of the surfaces. The laser pulsing frequency was set at 1500 Hz, which in turn gave rise to about 68% overlapping ratio for the irradiated spots at the surface. The initial tests were conducted to select the laser treatment parameters so that the laser parameters resulting in crack free surfaces were selected. The results of the initial tests revealed that increasing laser power by 10% while keeping laser scanning speed constant gave rise to large cavity formation at the surface. On the other hand, reducing laser scanning speed by 10% while keeping the laser output power same resulted in cracks at the surface. Consequently, laser treated surface properties became highly dependent on the proper selection of the laser processing parameters. Therefore, through controlling the laser power settings, beam intensity distribution, pulse repetition rate, spot size, and the scanning speed, crack free surface texture could be realized. Laser treatment conditions are given in Table 1. Liquid PDMS, which belongs to a group of polymeric organosilicon compounds, was used to replicate laser textured surface. PDMS (Sylgard 184, Dow Corning) was prepared by mixing the elastomer base with hardening agent in 10:1 wt%. The mixture was deposited onto the laser textured surface and degasified in a vacuum chamber at 0.1 bar for 30 min. The deposited and degasified PDMS was left in an oven at 1501C for 30 min for curing purposes. Solidified PDMS was then removed from laser textured surface after the curing period was over. PDMS was prepared again in 10:1 wt% (elastomer base to hardening agent ratio), poured over the copied PDMS, degassed and then cured in an oven using same parameters. Fig. 14 shows the schematic view of laser texturing and PDMS copying and replication of the laser textured surfaces. Material and surface characterization of laser textured, and PDMS copied and replicated surfaces was carried out using XRD and SEM. Jeol 6460 electron microscopy was used for SEM examinations and Bruker D8 Advanced having CuKa radiation was used for XRD analysis. A typical setting of XRD was 40 kV and 30 mA and scanning angle (2y) was ranged 20–80 degrees. AFM/SPM microscope in contact mode was used to analyze the surface texture. The tip was made of silicon nitride probes (r¼ 20 60 nm) with a manufacturer specified force constant, k, of 0.12 Nm 1. The standard test method for Vickers indentation hardness of advanced ceramics (ASTM C1327-99) was adopted to measure the surface microhardness and microphotonics digital microhardness tester (MP-100TC) was used for this purpose. The measurements were repeated five times at each location for the consistency of the results. A linear microscratch tester (MCTX-S/N: 0104300) was used to determine the friction coefficient of the laser treated, and, PDMS copied and PDMS replicated surfaces. Table 1

Laser processing parameters

Feed rate (m s 1)

Power (W)

Frequency (Hz)

Nozzle gap (mm)

Nozzle diameter (mm)

Focus diameter (mm)

N2 pressure (kPa)

0.1

2000

1500

1.5

1.5

0.3

600

Hydrophobic Materials

811

.5 µm

Z=2 Y=

Alumina

m

Laser source

µ 1.5

 = 150 degrees

 = 65 degrees

X = 1.5

µm

Textured alumina

Alumina

Copying

.9nm

Z=940

S

Y=

PDM

15

PDMS

nm

 = 122 degrees

Copied PDMS

Textured alumina

Copying Z = 3.2 Y=

S

20

PDM

 = 128 degrees

nm

PDMS

Copied PDMS

µm

Replicate of textured alumina

Fig. 14 Schematic view of laser texturing, PDMS copying, and replicating laser textured alumina surface. PDMS, polydimethylsiloxane.

The equipment was set at the contact load of 0.03 N and end load of 5 N. The scanning speed was 5 mm min 1 and loading rate was 1 N s 1. The total length for the scratch tests was 1 mm. FTIR was carried out incorporating Nicolet Nexus 670 FT-IR Spectrometer. The wetting experiment was performed using Kyowa (model – DM 501) CA goniometer. A static sessile drop method was considered for the CA measurement. The WCA between the water droplet and the surface was measured with the fluid medium as deionized water. Droplet volume was controlled with an automatic dispensing system having a volume step resolution of 0.1 mL. Still images were captured, and CA measurements were performed after 1 s of deposition of water droplet on the surface. UV–vis spectrophotometer (Jenway 67 series) was used to measure the transmittance of PDMS copied and replicated workpieces.

2.25.4.2

Findings and Discussions

Laser texturing of alumina surface for improved hydrophobicity is considered and copying/replication of textured surfaces by PDMS is realized for a single-step innovative processing. Laser textured, and PDMS copied and replicated surfaces are characterized using the analytical tools. Fig. 15 shows SEM micrographs of laser textured surface. Since laser repetitive pulses with 1500 Hz frequency is used to irradiated alumina surfaces, overlapping of irradiated spots forms regular ablation/melting tracks at the surface along the laser scanning direction (Fig. 15(A)). Overlapping ratio of irradiated spots is in the order of 68%. Moreover, combination of ablation and melting at the laser irradiated surface gives rise to formation of small cavities and poles at the irradiated surface (Fig. 15(B)). Although the distribution of poles demonstrates a nonuniform behavior, some degrees of a hierarchical texture is formed at the surface (Fig. 15(C) and (D)). This situation is also seen from Fig. 16, in which three-dimensional (3-D) optical image of the laser textured surface is shown.

812

Hydrophobic Materials

(A)

(B)

(C)

(D)

Fig. 15 SEM micrographs of laser textured alumina surface: (A) overlapping of laser irradiated spots, (B) small size pillars and cavities, (C) hierarchical surface texture, and (D) nanosize pillars in hierarchical texture. SEM, scanning electron microscope.

0.31

0.63

0.94

Fig. 16 Three-dimensional optical image of laser textured surface, in which regular scanning tracks are visible.

1.26 mm

Hydrophobic Materials

813

In addition, the pole height varies within 2.3–0.02 mm, which can be observed from Fig. 17, in which AFM image is shown (Fig. 17(A)) together with the surface texture profile along a line scanned (Fig. 17(B)). The surface roughness of the laser treated surface is in the order of 0.82 mm. It is evident from AFM image that the surface consists of fine-sized poles and cavities. Since laser beam intensity distribution is Gaussian across the irradiated spot, laser beam intensity remains high at the irradiated spot center. This gives rise to a local evaporation at the surface while melting takes place toward the edge of the irradiated spot. The melt flows from the irradiated spot edge toward the cavity center modifying the cavity size and texture at the irradiated surface. The flow of melt from the neighboring irradiated spot toward the previously formed spot, due to consecutive pulses, further modifies the surface texture, which in turn gives rise to a hierarchical topography of the surface. The closed examination of the surface reveals that no microcrack and large size void is formed at the surface (Fig. 15(B)).

6 Z: 2.

µm

Y: .0

15 µm

.0 µm

5 X: 1. (A)

µm 15 2.0 10

1.5 1.0

5 0.5 0.0

0 µm

(B)

0

5

10

15

2.3

Displacement (µm)

Displacement (µm)

1.8

0.3

0.8

1.8

1.3 0

4

9

Position (µm)

0

5 10 Position (µm)

Fig. 17 Atomic force microscope (AFM) image and line scan at laser textured surface: (A) three-dimensional image of surface, and (B) surface texture along the rakes at the surface.

814

Hydrophobic Materials

Laser treated layer

(A)

(B)

(C)

(D)

Fig. 18 SEM micrographs of cross-section of laser textured surface: (A) laser treated layer, (B) dense zone at the surface, (C) small size columnar structure, and (D) increasing columnar structure with increasing depth. SEM, scanning electron microscope.

This is attributed to the self-annealing effect generated during the repetitive pulse heating of the surface. In this case, heat conduction from the recently formed irradiated spot toward the previously formed spot alters the cooling rates below the heated surface. This in turn modifies the cooling rates and reduces the thermal stress in the surface vicinity. Moreover, the molten over flow at the laser treated surface is not observed from SEM micrographs. This indicates that combination of controlled ablation and melting in the irradiated region lowers the presence of molten material at the surface; in which case, the molten flow, which possibly covers the resulting surface texture, is avoided. Fig. 18 shows SEM micrograph of the cross-section of laser treated layer. The depth of laser treated layer extends almost 40 mm below the surface and it is formed mainly from three zones. In the first zone, a dense layer consisting of fine size grains are formed. The formation of the dense layer is attributed to the high cooling rates at the surface and formation of nitride compound in the surface region because of the presence of high pressure nitrogen assisting gas. It should be noted that nitrogen at high pressure is used as an assisting gas during laser treatment process. The presence of nitrogen compound is also evident from the XRD of the surface, which is shown in Fig. 19. In this case, the peaks observed in the diffractogram are as follows: al2O3 (ICSD collection codes 010426 and 073076), AlN (ICSD collection codes 041358 and 082790), and AlN (ICSD collection codes 070032 and 070033). The formation of aluminum nitride is attributed to the following reactions: a single alumina oxide and carbon monoxide are formed at the treated surface through a reaction Al2O3 þ 2C-Al2O þ 2CO. As a second step, AlN is formed through the reaction Al2O þ CO þ N2-2AlN þ CO2 under high pressure nitrogen environment [83]. In the second zone of the laser treated layer, columnar structures with small sizes are formed just below the dense layer. The formation of small size columnar structure is related to the heat diffusion in alumina during the processing; in which case, temperature gradient reduces with increasing depth [84]. As depth below the surface increases, the size of the columnar structure increases. However, the heat affected zone is not observed below the melted layer, which is attributed to the low thermal diffusivity of alumina. Fig. 20 shows SEM micrographs of PDMS surface after removal from the laser textured alumina surface. Since liquid PDMS has excellent rheological properties, it wets the texture features of laser treated alumina surface prior to its solidification. Therefore, it copies almost exactly the surface texture of laser treated alumina when removed in a solid phase from the textured workpiece surface (Fig. 20(A)). These give rise to formation of micro/nano-sized voids and cavities at the solid PDMS surface (Fig. 20(B)).

Hydrophobic Materials

815

3000 -Al2O3 -Al2O3 AlN

Laser treated

Relative intensity

2500

2000

1500

1000

As received

500

0 30

40

50 60 2  (degress)

70

80

Fig. 19 X-ray diffraction (XRD) of laser treated and as received alumina.

(A)

(B)

(C)

(D)

Fig. 20 SEM micrographs of PDMS copied surface: (A) overlapping of laser irradiated spots, (B) fine size textures copied, (C) fine size copied texture with alumina residue (marked circle), and (D) submicron copied texture. SEM, scanning electron microscope; PDMS, polydimethylsiloxane.

816

Hydrophobic Materials

However, some whiskers like texture are not clearly visible at the solid PDMS surfaces (Fig. 20(C)). Some fine size textures are partially copied (Fig. 20(D)) from the laser treated surface. This behavior is related to strong adhesion between whiskers on the laser textured surface and PDMS, and odd orientation of whiskers on the laser textured surface, which causes broken whiskers during the removal of PDMS from the laser textured surface. In the case of complex geometric shapes, such as odd-shaped texture peaks on laser textured surface, some residues of solid phase PDMS remained at the laser textured surface because of the strong adhesion between solidified PDMS and the laser textured surface. The large size textures are copied by PDMS without loss of the texture features, which is associated with the low elastic modulus of solid phase of PDMS; in which case, it behaves like elastic body during the removal from the laser textured surface. In addition, no rapturing of PDMS surface is observed after removal from the laser treated surface. Fig. 21 shows SEM micrographs of replicated PDMS. Replication is done by recasting the PDMS on copied PDMS surface. The similar arguments can be applied for the PDMS replicated laser textured surface (Fig. 15(A)–(D)) to those for the PDMS copied surface (Fig. 20).

(A)

(B)

(C)

(D)

Fig. 21 SEM micrographs of PDMS replicated laser textured surface: (A) overlapping of replicated laser irradiated spots, (B) fine size replicated texture with alumina residue (marked circle), (C) fine size textures replicated, and (D) submicron replicated texture. SEM, scanning electron microscope; PDMS, polydimethylsiloxane.

Hydrophobic Materials

817

Fig. 22(A) and (B) shows the 3-D optical image of copied and replicated laser textured surfaces, respectively. It is evident that the replicated surface texture shows identical surface pattern to that of laser textured surface (Fig. 16). Figs. 23 and 24 show AFM images of the PDMS copied and replicated surfaces, respectively. The general features of the PDMS replicated surfaces (Fig. 24(A)) are similar to the laser textured surface (Fig. 17(A)). The averaged textures heights of PDMS replicated (Fig. 24(B)) and laser textured surfaces (Fig. 17(B)) are almost the same. Consequently, replicated surface has the texture morphology similar to that of the laser textured surface, provided that some fine texture details of the laser treated surface are not observed at the PDMS replicated surface. The surface free energy of laser treated workpiece is measured using the CA sessile droplet method for water, glycerol, and diiodomethane [85]. The surface energy formulation incorporating the CA measurement technique is presented in the early study [85]. Only the surface energy data based on the CA measurements are given herein and the details of the analysis can be found in the previous study [86]. The surface energy determined is in the order of 54.7 mJ m 2, which is slightly higher than that reported in the previous study (53 mJ m 2) for AlN [86]. This is attributed to the nonuniform distribution of nitrogen compounds formed at the surface, which is also observed from energy dispersive spectroscopy (EDS) data given in Table 2, i.e., at different locations of the laser treated surface, nitrogen concentration differs significantly. Consequently, it is most probable that nitrogen compounds are formed locally in the near region of the irradiated spot center, where surface temperature remains high. Nevertheless, the surface energy measured for laser textured surface is very close to that of presented in the open literature [87]. Fig. 25 shows FTIR data for PDMS copied and replicated surfaces. In general, two groups of data can be observed, namely SiMethyl (SiMe) and siloxane (SiOSi) groups. A doublet at 1100 cm 1 and 1020 cm 1 correspond to asymmetric and symmetric stretching vibration, respectively [88]. The absorption at 800 cm 1 is attributed to out-of-plane oscillation of the Si–CH3 bonding. The SiMe structure peak occurs at 1250 cm 1 [89]. Slight differences in absorption as compared to that reported in the literature [90] are associated with the solidification rates of liquid PDMS, which takes longer duration for solidification at the laser treated surface while slightly modifying the absorption characteristics. The presence of absorption peak at 2960 corresponds to –CH3 (asymmetric) bend stretching vibration [91]. In addition, the presence of peak at 670 cm 1 demonstrates that Al–N vibrations take place for the PDMS copied and replicated wafers. This is attributed to the residues of textured alumina surfaces, such as whiskers, which remain in the PDMS copied surface after removal from laser textured surface and later transferred to PDMS replicated surface.

0.31

0.63

0.94

1.26 mm

(A)

0.31

0.63

0.94

1.26 mm

(B) Fig. 22 Three-dimensional optical image of copied and replicated PDMS surfaces: (A) copied surface and (B) replicated surface.

818

Hydrophobic Materials

Z: 940.7 nm

.5 Y: 1 m .0 µ

X: 15.0 µm (A) 15 800

nm

600

10

400 5 200 0 µm

0

5

10

442 Displacement (µm)

Displacement (µm)

696

496

296

0 (B)

15

5

10

Position (µm)

242

0

5 10 Position (µm)

Fig. 23 Atomic force microscope (AFM) image of PDMS copied surface: (A) three-dimensional image of copied PDMS surface and (B) line scan at the copied surface. PDMS, polydimethylsiloxane.

Fig. 26 shows water droplet images obtained for laser textured (Fig. 26(A)), and PDMS copied (Fig. 26(B)), and PDMS replicated (Fig. 26(C)) surfaces. The CA remains slightly higher for PDMS duplicated surfaces (Fig. 26(C)), and then follows PDMS copied surface. Low surface energy and micro-/nano-textured alumina surface resulted in high CA of the droplets. The CA of the droplets varies within 10% over the laser treated surface. This behavior is attributed to the one or all of the followings: (1) locally varying surface energy due to nonuniform distribution of nitrogen compounds at the surface, and (2) nonhomogeneous distribution of the micro/nanosize poles and cavities at the surface. Nevertheless, the coverage area of the low CA region is small as compared to that of the high CA region at the surface, i.e., the low CA region is in the order of 7% of the total area of the laser treated surface. Some small differences in between CAs due to copied and replicated PDMS surfaces are associated with the surface texture differences between two surfaces. In this case, some of the nano and sub-nanosize features imbedded in the copied PDMS surface is protruded from the surface for the PDMS replicated surface.

Hydrophobic Materials

819

1 µm

Z: 2.

.0

15 X: µm

0 µm

. Y: 15 (A)

µm 15 1.5 10 1.0 5

0.5

0.0

0 (B)

µm

0

5

10

1229 Displacement (µm)

1.4 Displacement (µm)

15

0.9

1099

899

699

0.4 0

5

10

Position (µm)

0

5 10 Position (µm)

Fig. 24 Atomic force microscope (AFM) image of PDMS replicated surface: (A) three-dimensional image of replicated PDMS surface and (B) line scan at the replicated surface. PDMS, polydimethylsiloxane.

Table 3 gives the CA and hysteresis (yhysteresis ¼ yAdvancing yReceding, where yAdvancing is the advancing angle and yReceding is the receding angle) of laser textured, PDMS copied and replicated surfaces. The hysteresis corresponding to copied and replicated PDMS surfaces are slightly higher than that of that of the laser textured surfaces. This is because of the loss of some nanosize whiskers on PDMS copied and replicated surfaces, which lowers the lotus effect on the surface. Fig. 27 shows the optical transmittance of PDMS copied and replicated wafers. The transmittance data for as received flat PDMS are also included for the comparison reason. The loss of transmittance is low for the PDMS copied and PDMS replicated wafers. Fig. 28 shows the friction coefficients of the as received and laser treated alumina surface together with PDMS surface. The friction coefficient attains low values for the laser treated surfaces, which is related to the surface hardness improvement after the laser treatment process. In this case, surface hardness increased significantly (1650 ( þ 50/ 50) HV) as compared to as received

820

Hydrophobic Materials

Table 2

Energy dispersive spectroscopy (EDS) results in the surface region of the laser treated alumina

7 µm

Electron image 1

1.0 800

As received Copied

0.8

Replicated

Absorptance

1020 1100

0.6

0.4 670

1250

0.2 2960 0.0 600

1200

1800

2400

3000

3600

Wave number (cm−1) Fig. 25 FTIR plots of as received plane PDMS, PDMS copied surface, and PDMS replicated surface. FTIR, Fourier transform infrared spectroscopy; PDMS, polydimethylsiloxane.

 = 150 degree  = 65 degree

As received alumina surface

 = 122 degree

PDMS copied surface (B)

Laser textured alumina surface (A)

 = 128 degree

PDMS replicated surface (C)

Fig. 26 Contact angle of water droplet on different surfaces. As received surface. (A) Laser textured alumina surface. (B) PDMS copied surface. (C) PDMS replicated surface. PDMS, polydimethylsiloxane.

Hydrophobic Materials

821

Table 3 Contact angles (CAs) measurement results for as received, laser treated, polydimethylsiloxane (PDMS) copied and replicated surfaces

Untreated surface Laser treated surface PDMS copied PDMS replicated

CA (degrees)

Hysteresis (degrees)

65.3 ( þ 5/ 5) 150.2 ( þ 5/ 5) 122.4 ( þ 5/ 5) 128.2 ( þ 5/ 5)

42 18 32 24

100

Transmittance

80

60

40

As received PDMS 20

Copied PDMS Replicated PDMS

0 300

400

500

600 700 Wavelength (nm)

800

900

Fig. 27 UV transmittance of as received plane PDMS wafer, PDMS copied, and PDMS replicated surfaces. UV, ultraviolet; PDMS, polydimethylsiloxane.

1.2 As received alumina Laser treated alumina

1.0

Friction coefficient

As received PDMS Copied PDMS

0.8

Replicated PDMS 0.6

0.4

0.2

0 0.0

0.1

0.2 0.3 Distance (mm)

0.4

0.5

Fig. 28 Friction coefficient for as received, and laser textured alumina surface, and PDMS copied and replicated surfaces. PDMS, polydimethylsiloxane.

alumina surface (1100 ( þ 50/ 50)). The attainment of increased surface hardness is associated with the dense layer formed at the surface under the high cooling rates during laser treatment process. The friction coefficient for PDMS surface is higher than those corresponding to laser treated and as received alumina surfaces. This behavior is attributed to low hardness and small elastic modulus of PDMS.

822

Hydrophobic Materials

2.25.5

Case Study for Solvent Induced Crystallization of Polycarbonate Surface and Environmental Dust Effects

Polycarbonate wafers are used as protective cover for PV panels. Solvent crystallization of polycarbonate surface results in micro/ nanospherulites and fibrils at the surface, which improve significantly surface hydrophobicity and self-cleaning characteristics of crystalized surface. Dust accumulation and mud formation on crystalized polycarbonate surface are critical for the performance of PV panels and it needs to be examined thoroughly. Climate change results in regular sand storms taking place in the Middle East, particularly in Saudi Arabia [92], and desert dust is one of the major constitutes of the particles in storm. Small size dust particles can suspend in the air over many days after the storm is over and some of these particles settle on the exposed surfaces in environments. Although these particles are small in size, they cover large area over the time and modify optical, texture, and other characteristics of the surfaces. Mineral dust particles have tendency to absorb and scatter solar radiation while lowering solar power reaching surfaces and altering environmental temperatures. However, some of the incident solar power losses are partially compensated by long wave emission of radiation by dust particles, which may not contribute considerably to solar power harvesting in terms of useful energy generation through concentrated solar heating, electricity generation by PV, thermal volumetric solar absorption, etc. Minimization of dust settlement and accumulation on active surfaces of solar power applications is one of the recent challenges to be met. On the other hand, development of new and cost-effective self-cleaning surfaces is promising to minimize and avoid after dust effects on active solar energy harvesting surfaces. One of the methods to minimize the efforts required to remove the settled dust particles at the surface is to modify surface texture and surface energy toward improving hydrophobicity. Low surface energy and texture consisting of micro/nanopillars give rise to improved surface hydrophobicity and low adhesion between dust particles and the surface. Generating the cost-effective hydrophobic surfaces toward solar energy harvesting applications are challenging because of keeping optical properties of the surface after texturing is difficult to achieve. Hydrophobicity of a surface depends and interfacial energies of solid and liquid, surface texture, and Laplace pressure as demonstrated by Wenzel and other researchers. Mimicking the nature, such as lotus leaves surface, enables to create surfaces of superhydrophobic characteristics. A substantial increase in hydrophobicity can be achieved when a combination of chemical modification and surface roughness of the substrates is integrated. Many techniques and processes have been developed to enhance the hydrophobicity of surfaces using this strategy; however, some of these techniques involve multistep procedures and harsh conditions or required specialized reagents and equipment. In addition, micro/nanotexturing of surfaces by a laser controlled ablation provides improved surface hydrophobicity; however, thermal stress field developed in the laser textured substrates can limit the practical applications. Surface crystallization of polymeric materials, such as polycarbonate wafers, gives rise to hierarchical textures compose of micro/nanostructures. This is because of chain flexibility of polycarbonate molecules in the glass structure, which presents high crystallization ability when subjected to solvents such as acetone. One of the techniques to crystallized polymeric material surfaces is to use solvents, such as acetone, through the immersion technique. Although crystallization process is fast and generates hierarchical texture structures at surfaces, optical characteristics of the crystalized surfaces change because of the chain restrictions of large size molecules in crystal structures. On the other hand, polycarbonate wafer (p-hydroxypheyl) is one of the candidates to replace silicon base protective glasses for PV applications. This is because of its low density, high fracture toughness, and mechanical flexibility. Considerable research studies were carried out to examine crystallization of polycarbonate wafers. Adhesion between the dust particles and the solid surface is governed by the van der Waals forces; however, in the humid air conditions the formation of mud solution modifies and alters interfacial adhesion force because of the cohesive effect of the crystalized dried mud solution. In humid environments, the dust particles settled at surfaces and adsorb water from humid atmosphere forming a mud at the surface. Some of alkaline and earth alkaline compounds in the dust particles dissolve in water and the mud solution sediments on solid substrate surface while forming chemically active liquid layer at the interface between the mud and the surface. The mud solution dries with the mud at the surface with progressing time. The dry mud solution can result in covalent bounding at the solid substrate surface while increasing forced required removing the dry mud from the surface. In order to reduce the efforts required to remove the dry mud from the surfaces, investigation of the dry mud adhesion on hydrophobic surfaces become essential. In the present study, adhesion of dry mud on crystalized polycarbonate surface is examined. The surface of polycarbonate wafer is crystallized to achieve a surface texture composing of micro/nanostructures consisting of spherulites and fibrils. The solvent immersion technique incorporating acetone is used to crystalize polycarbonate surfaces. The characteristics of crystalized surface including morphological, optical, hydrophobic, and tribological features are assessed through analytical tools. Influence of the dry mud formed from the dust particles on the characteristics of the crystalized surface is presented in line with the previous study [37].

2.25.5.1

Experimental

Polycarbonate wafers of 3 mm thickness were used as workpieces. Polycarbonate wafer had excellent optical clarity with high toughness and it was derived from p-hydroxypheyl. Polycarbonate wafers were cleaned ultrasonically prior to immersion into a liquid acetone for 2 min duration. However, several tests were conducted to select the appropriate immersion duration for crystallization of polycarbonate surface. Therefore, the immersion duration resulting in crystal structures with hydrophobic characteristics was selected. An experiment was carried out to examine the effect of the actual mud formation, because of water vapor condensation onto the accumulated dust particles at the surface, on the characteristics of crystalized polycarbonate surface. A dust layer of 300 mm

Hydrophobic Materials

823

thickness was formed from the dust particles collected from the local environment on workpiece surface. The dust layer of 300 mm thickness resembles the actual dust accumulation on surfaces in open environment over 1 week period. The desalinated water having same volume of dust layer was dispensed gradually on to the dust layer at the workpiece surface. It should be noted that initially some tests were carried out to measure the amount of water vapor absorbed by the dust particles, due to condensation in the local humid environment, over 6 h period. The tests results revealed that the amount of condensate had almost same volume of dust over 6 h. In order to resemble the water condensation on the layer of dust particles in the humid air, dispensed water with the same amount of dust particles in the layer was left at the workpiece surface without mechanical mixing of the dust particles with water. This gave rise to natural formation of mud at the workpiece surfaces. The workpieces were, then, located in a local normal air ambient for 3 days to dry. The crystalized and as received polycarbonate surfaces with the presence of dry mud were tested for adhesion work and friction coefficient measurements. Once the adhesion and friction tests were completed, the dry mud layer was removed from the surface with a pressurized desalinated water jet having 2-mm diameter and 2 m s 1 velocity. The water jet-assisted cleaning process was continued for 20 min for each workpiece surface. Finally, the analytical tools were used to assess the morphology, microhardness, and hydrophobic characteristics of the cleaned surfaces.

2.25.5.2

Findings and Discussions

Surface crystallization of polycarbonate wafers is carried out to achieve a hierarchical texture at the surface toward generating the hydrophobic surface characteristics. Influence of dust and mud formed from the dust particles on texture, hydrophobicity, and friction coefficient of crystalized polycarbonate surface is presented in the light of the previous study [37]. Fig. 29 shows SEM micrographs of crystalized polycarbonate surface. Crystalized surface demonstrates hierarchical structure, which composes of microsize spherulites and nanosize fibrils (Fig. 29(A) and (B)). The formation of micro/nanosize spherulites and fibrils is associated with the acetone-induced crystallization. Spherulites aggregate at crystalized surface and surface area of spherulites is partially covered by fibrils (Fig. 29(C) and, in some cases, fibrils forms like a compact structure at spherulites surface (Fig. 29(D))). After immersion, crystals growth radially from the potential nucleation sites on polycarbonate surface and extension of some branches occurs from the nucleation sites during growth of crystals; in some cases, a premature formation of spherulites without fibrils are observed (Fig. 29(E)). In addition, connected spherules form from the intermittent branching, which appear locally at crystallized surface (Fig. 29(F)). Acetone has Hildbrand solubility parameters in the order of 20.1–20.3 J1/2 cm 3/2 and it possesses miscible characteristics. Initial stage of immersion process, acetone diffuses into the polymeric structure and gradually forms swollen film at the surface (gelated layer) during the dissolution. As the immersion time progress, diffusion increases and the glass transition temperature of polymeric film reduces behind diffusion front. This causes plasticizing of swollen polymer. The diffusion process is non-Fickian and the diffusion front in between swollen film and solid polycarbonate penetrates almost at a constant velocity into the solid phase of amorphous polycarbonate. When polycarbonate wafer is removed from the immersion bath, acetone residues evaporate at the surface and spherulites are initiated to form at the surface. The transition glass temperature reduces during acetone evaporation while giving rise to supercooling of the swollen film. This in turn results in crystalized surface composing of micro/nanosize spherulites and fibrils. Spherulites demonstrates scattered pattern at the surface, which is associated with the transient nature of evaporation of acetone from polycarbonate surface, which causes locally scattered nucleation centers at the surface. In general, three consecutive phases are associated with crystallization of polycarbonate surface. These include initiation of crystallization, primary crystallization, and secondary crystallization. A nucleus emerges in the early phase; in which case, polymer chains align in a parallel way gradually and the chains are added to the nucleus. Crystal growth becomes spontaneous as the nucleus size reaches to the critical size. As crystallization progresses, bundle-like or lamellar crystallization takes place; however, types of crystallization depend on the size of primary nucleus and free energy of the surface normal to the chain direction per unit area. In the case of mud formed at crystalized polycarbonate surface, mud cross-section composes of porous structures. This can be seen from Fig. 30(A), in which SEM micrograph of mud cross-section on crystalized polycarbonate surface is shown. Mud consists of the dust particles with varying sizes within 0.01–20 mm (Fig. 30(B)). The EDS data for the dust particles are given in Table 4. The dust particles possess alkali and alkali earth metals (Na, Ca) and they dissolve in water during the mud formation while forming a mud solution. The mud solution penetrates through the mud cross-section along the pores and reaches at crystalized polycarbonate surface. The mud solution accumulates at the interface of the mud and crystalized surface while forming a dense layer upon drying the solution (Fig. 30(C)). However, some of the mud solution could not reach at crystalized surface and captured in a large size pore in the mud. It dries by time and forms a dense structure in this region (Fig. 30(D)). On the other hand, once the dry mud is removed from crystalized surface by pressurized desalinated water jet, residues of the mud particulates remain at crystalized surface indicating the strong adhesion between mud residues and crystalized surface. The total area covered by mud residues is estimated as almost 7%. In order to assess the elemental distribution of the dry mud solution, further experiments are carried out. Fig. 31 shows SEM micrographs of mud solution on crystalized polycarbonate surface after drying. It is evident that the crystals are formed with varying sizes (Fig. 31) and EDS data for the dry mud solution demonstrate that alkaline (Na, K) and alkaline earth metallic (Ca) compounds are formed on crystalized polycarbonate surface. In addition, some attachments of sub-microsize loose dust particles are observed on crystal structures formed (Fig. 31). In addition, inductively coupled plasma mass spectrometry analysis of mud solution (Table 5) reveals same conclusion of the EDS data. The pH of mud solution increases to reach pH ¼ 7.4 after 24 h.

824

Hydrophobic Materials

(A)

(B)

(C)

(D)

(E)

(F)

Fig. 29 SEM micrographs of crystalized polycarbonate surface: (A) crystalized surface formed from micro/nanospherulites and fibrils, (B) fibrils on spherulite surface, (C) spherulites with large area coverage of fibrils, (D) compact fibrils on spherulites surface, (E) initiation of radial growth of spherulites, and (F) radially extended spherulites. SEM, scanning electron microscope. Reproduced from Yilbas BS, Ali H, Al-Aqeeli N, Abu-Dheir N, Khaled M. Influence of mud residues on solvent induced crystalized polycarbonate surface used as PV protective cover. Sol Energy 2016;125:282–93.

Therefore, OH ions in the mud solution are responsible for high basic pH level, which is associated with the dissolution of the alkaline and alkaline earth metallic compounds in the water. Moreover, NaOH attacks at crystalized polycarbonate surface and it gives rise to local damages on the surface, which can be observed from SEM micrograph (Fig. 32(A) and (B)). Therefore, influence of the mud on crystalized polycarbonate surface is detrimental prior to its drying.

Hydrophobic Materials

825

Void

× 150

15 KV

100 µm 0000

13

39

× 1,500

15 KV

SEI

(A)

10 µm

23

41

SEI

(B)

Dry mud solution

Dry mud solution

× 500

5 KV

50 µm 0000

13

× 600

5 KV

39 SEI

(C)

20 µm 0000

13

39 SEI

(D)

Fig. 30 SEM micrograph of dry mud cross-section on polycarbonate surface: (A) large view of cross-section, (B) voids in the cross-section, (C) dried mud solution at the interface of the dry mud and polycarbonate surface, and (D) local accumulation of mud solution. SEM, scanning electron microscope. Reproduced from Yilbas BS, Ali H, Al-Aqeeli N, Abu-Dheir N, Khaled M. Influence of mud residues on solvent induced crystalized polycarbonate surface used as PV protective cover. Sol Energy 2016;125:282–93.

Table 4

Elemental composition of the dust (wt%)

Si

Ca

Na

S

Mg

K

Fe

Cl

O

11.4

7.8

3.4

2.1

2.9

1.3

1.3

1.1

Balance

In order to assess the texture characteristics of crystallized polycarbonate surface prior to the mud formation, AFM is carried out. AFM microimages of crystallized polycarbonate surfaces and line scans are shown in Fig. 33. AFM image shows nonspherical textures at the crystalized polycarbonate surface (Fig. 33(A)). This reveals that spherulites have nonspherical morphology and their sizes vary at the surface. From the line scan (Fig. 33(B)), crystalized surface has a texture composing of micro/nanosizes of spherulites and fibrils. The presence of fibrils results in ripples/wavy like behavior of texture profile along the line scan (Fig. 33(B)). The roughness of crystalized polycarbonate surface varies within 3.6–4.3 mm. In the case of crystalized polycarbonate surface after dry mud removal, surface texture is modified significantly after the dry mud removal because of the residues of dry mud at the surface (Fig. 33(C)); in which case, texture height increases and ripples/waviness in texture profile almost disappears (Fig. 33(D)). This indicates that the dry mud residues cover the fibrils and modifies spherulites morphology. In addition, small cavity with deep penetration is also observed on crystallized polycarbonate surface (Fig. 33(D)) because of hydroxyl attach (NaOH) at the surface prior to drying of mud solution. The parameters affecting the liquid drop CA are the surface free energy of the substrate material and the surface texture. The CA of a liquid droplet on a perfectly smooth and chemically homogenous surface can be formulated by Young’s equation; however,

Hydrophobic Materials

826

Fig. 31 SEM micrographs of crystalized mud solution. SEM, scanning electron microscope. Reproduced from Yilbas BS, Ali H, Al-Aqeeli N, AbuDheir N, Khaled M. Influence of mud residues on solvent induced crystalized polycarbonate surface used as PV protective cover. Sol Energy 2016;125:282–93.

Table 5

Internal cathodic protection data for the mud solution after 6 h dissolution time of dust particles in desalinated water

Ca

Na

Mg

K

Fe

Cl

309,300

44,200

69,820

32,600

1724

38,200

Damage site

Dry mud residues 5 KV (A)

× 7000

2 µm 0000

16

35

15 KV

SE I

× 7000

2 µm

0000

16

35 SE I

(B)

Fig. 32 SEM micrographs of crystalized polycarbonate surface after dry mud removal by pressurized desalinated water jet: (A) localized mud residues at crystalized surface and (B) local damage site at crystalized surface [37]. SEM, scanning electron microscope. Reproduced from Yilbas BS, Ali H, Al-Aqeeli N, Abu-Dheir N, Khaled M. Influence of mud residues on solvent induced crystalized polycarbonate surface used as PV protective cover. Sol Energy 2016;125:282–93.

the formulation is limited to exceptionally smooth and homogenous surfaces, which may not be applicable for the textured surfaces. Wenzel and Cassie–Baxter relations for apparent CA include surface roughness and can possibly provide realistic data. On the other hand, liquid droplet has liquid–solid and liquid–vapor interfaces, which are included in the CA formulation. Therefore,

Hydrophobic Materials

Z: 7.5

.0 µm

827

µm

Z: 16

Y: 15 µm .0

9 Y: 0.0 µm

.0 µm

.0 X: 90

X: 15

µm (B)

(A)

5 4

3

Z (µm)

Z (µm)

4

2 1

2 1

0

18 µm

3

0 5 10 15 20 25 30 35 40 X (µm)

0

10 µm

Line scan

0 2 4 6 8 10 12 14 X (µm) Line scan

Fig. 33 AFM images of crystalized and dry mud removed polycarbonate surface: (A) crystalized polycarbonate surface and line scan for texture and (B) dry mud removed from crystalized polycarbonate surface by a pressurized desalinated water jet and line scan for texture. AFM, atomic force microscope; SEM, scanning electron microscope. Reproduced from Yilbas BS, Ali H, Al-Aqeeli N, Abu-Dheir N, Khaled M. Influence of mud residues on solvent induced crystalized polycarbonate surface used as PV protective cover. Sol Energy 2016;125:282–93.

the relation developed for the CA is: cosyc ¼ f1 cosy1 þ f2 cosy2

ð24Þ

where yc is the apparent CA, f1 is the surface fraction of liquid–solid interface, f2 is the surface fraction of liquid–vapor interface, y1 is the CA for liquid–solid interface, and y1 is the CA for liquid–vapor interface. For the air–liquid interface, f1 can be represented as f, which is the solid fraction, and air fraction (f2) becomes (1 f). In the formulation, f varies within 0–1, such that f¼0 corresponds to the case where the liquid droplet is not in contact with the surface, while f ¼1 is case of the surface, which is completely wetted. However, the contact mode changes from Cassie–Baxter state to Wenzel state when the surface texture becomes sparse or when the droplets impact onto the surface with a high velocity. Crystallized polycarbonate surface has the micro/nanosize texture feature, which in turn gives rise to the Cassie–Baxter state at the surface. Consequently, surface texture with combination of micro/nanospherulites and fibrils causes air pockets to be trapped within the texture, while resulting in Cassie–Baxter state at crystallized polycarbonate surface. In addition, some small variations in CAs at crystallized polycarbonate are associated with nonuniform distribution of spherulites and fibrils at crystalized surface. Fig. 34 shows droplet images and CAs corresponding to as received, crystalized, and dry mud removed surfaces. However, water droplet CA corresponding to mud removed crystalized polycarbonate surface is considerably less than that of crystalized surface, which is attributed to the mud residues modifying surface texture completely and altering surface energy. The hysteresis angle corresponds to crystalized surface is found to be less than that corresponding to mud removed crystalized polycarbonate surface. This is because of almost total coverage of nano and sub-nanosize fibrils by mud residues on crystalized surface. Fig. 35 shows the UV–visible transmittance data for untreated, crystallized, and crystalized mud removed polycarbonate wafers. Crystallization of polycarbonate surface reduces the transmittance because of scattering of incident radiation at the textured surface. However, transmittance reduces significantly for the crystalized and mud removed surface. The reduction in the transmittance is associated with the mud residues, which are scattered at crystalized surface while blocking the incident radiation.

828

Hydrophobic Materials

 = 102 degree

 = 68 degree

 = 155 degree

Fig. 34 Contact angle (CA) for as received, crystalized, and crystalized and mud removed by a pressurized desalinated water jet polycarbonate surfaces. Reproduced from Yilbas BS, Ali H, Al-Aqeeli N, Abu-Dheir N, Khaled M. Influence of mud residues on solvent induced crystalized polycarbonate surface used as PV protective cover. Sol Energy 2016;125:282–93.

100

Transmittance

80

60

Crystallized As received

40

Crystalized and dry mud removed

20

0 400

500

600 Wavelength (nm)

700

800

Fig. 35 Transmittance for as received, crystalized, and crystalized and mud removed by a pressurized desalinated water jet polycarbonate wafers. Reproduced from Yilbas BS, Ali H, Al-Aqeeli N, Abu-Dheir N, Khaled M. Influence of mud residues on solvent induced crystalized polycarbonate surface used as PV protective cover. Sol Energy 2016;125:282–93.

2.25.6

Conclusions

Surface hydrophobic characteristics of substrates are important in terms of self-cleaning applications of solar energy harvesting devices. Because of air pollution and environmental dust accumulation on surfaces of such devices, the performance of devices reduces gradually and efforts required to regain device performance becomes expensive. This is because of the detrimental chemical effect of the dust particles settled at the surface. Consequently, improvement of surface characteristics toward achieving self-cleaning at the surface becomes essential. One of the methods to achieve self-cleaning characteristics at the surface is to generate hydrophobic characteristics at the surface. In this case, dust particles adhesion becomes minimal and the effort required to remove such particles becomes small. In this chapter, the basic understanding of surface hydrophobicity is introduced and assessments methods are provided. In order to enhance the understanding of the processes and techniques involved with hydrophobicity, two case studies are introduced, namely laser texturing of alumina surface and PDMS replication of the textured surface, and solution crystallization of polycarbonate surface. The findings related to each case study are presented under the appropriate subsections:

2.25.6.1

Laser Texturing and Polydimethylsiloxane Replication of Textured Surface

Laser treatment of alumina surface toward improving surface hydrophobicity is carried out and surface characteristics are analyzed using the analytical tools. Liquid PDMS is used to copy and replicate laser textured surfaces. Surface texture of PDMS copied and replicated wafers are examined in details using scanning electron and AFM. PDMS copied and replicated surfaces are further characterized incorporating FTIR and UV transmittance tests. Hydrophobic characteristics of laser textured, and PDMS copied and replicated surfaces are assessed using the sessile droplet method. The friction coefficient of laser textured and PDMS copied and replicated surfaces are measured using the microtribometer. It is found that laser textured surfaces demonstrate hydrophobic characteristics, which is associated with micro/nanocavities and poles formed at the surface. Since repetitive laser pulses are introduced during laser scanning of the surface, self-annealing effect is created through heat conduction from the recently formed irradiated spot toward the early formed laser spot in between the two consecutive pulses. This, in turn, eliminates the crack

Hydrophobic Materials

829

formation in the laser treated layer. A dense layer consisting of fine size grains are formed in the surface region because of the high cooling rates. Use of high pressure nitrogen assisting gas gives rise to formation of AlN at the laser treated surface, which lowers the surface energy of the laser treated surface. PDMS copied and replicated surfaces give rise to a similar texture distribution over the surface as the laser treated surface. However, some of the detailed features, such as fine size odd-shaped whiskers, are not copied properly from the laser textured surface. PDMS copied and replicated surfaces result in hydrophobic characteristics; however, CAH appears to be slightly higher than that of laser textured surface. This behavior is attributed to incomplete copying and replicating laser treated surface due to the presence of the fine size whiskers at the laser treated surface. FTIR data reveal that Al–N peak occurs for the PDMS copied and replicated surfaces, which is related to the residues of the broken whiskers from laser textured surface. UV transmittance of PDMS copied and replicated wafers reduce slightly as compared to that of the as received flat PDMS wafer. The friction coefficient of the laser treated surface is lower than that of the as received alumina surface, which is associated with the microhardness enhancement after the laser treated process. The friction coefficient of PDMS copied and replicated surfaces are higher than laser treated surface, which is related to the small hardness and high elastic modulus of PDMS.

2.25.6.2

Solution Crystallization of Polycarbonate Surface and Environmental Dusts

Polycarbonate wafers can be used as a protective cover for PV panels to protect the surface from environmental hazardous including dust, humidity, rain, etc. Self-cleaning of polycarbonate wafer surfaces is one of the methods to minimize dust and mud effects in a humid environment. Crystallization of polycarbonate surfaces by using a solvent immersion method provides hierarchical texture improving hydrophobic and self-cleaning characteristics of the surface. In the present chapter, dust accumulation and mud formation on crystalized polycarbonate surfaces are simulated experimentally to resemble the dust settlement at the surface in the humid environment. Influence of dry mud residues and mud solution at crystalized surface is examined using analytical tools including optical, electron scanning, and AFM, EDS, XRD, FTIR, UV-transmission, and scratch tester. Hydrophobicity of crystalized polycarbonate surface prior and after dry mud removal is assessed incorporating sessile drop method. Friction coefficient of crystalized surface prior and after dry mud removal is measured by using a microtribometer. It is found that crystalized surface composes of micro/nanosize spherulites and fibrils forming a hierarchal texture, which increases hydrophobicity of the surface considerably as observed from high CAs of water droplets. The mud formed from the dust particles at crystalized surface gives rise to a strong bonding at the surface. In this case, alkaline and alkaline earth metal compounds, which are present in the dust, dissolve in water forming a chemically active solution, which settles at the interface between the mud and crystalized surface because of gravity. Adhesion between the mud and crystalized polycarbonate surface increases considerably once the dust solution dries. Dry mud residues are observed at crystalized surface despite cleaning the surface with pressurized desalinated water. Fibrils are almost covered by the dry mud residues after cleaning crystalized surface. This lowers surface hydrophobicity considerably and enhances the friction coefficient at the surface. Scratch hardness increases after surface crystallization and it increases further after the dry mud removal from the surface. Surface crystallization lowers transmittance of polycarbonate wafer because of scattering of incident visible radiation by the surface texture. The mud residues further lower the transmittance of crystalized polycarbonate surface.

2.25.6.3

Future Directions

The maximum solar energy harvesting is one of the challenges in clean energy research. Improvement of surface characteristics toward enhanced solar absorption is the main focus in solar energy harvesting studies. However, frequent environmental dust storms give rise to excessive dust accumulation on the selective solar absorbing surfaces and the PV protective glass surfaces. Cleaning of such surfaces from the environmental dusts is expensive and involves with energy intensive processes. In addition, the use of large amount of clean water, for cleaning purposes, brings difficulties in the areas where the clean water scarcity is high. Consequently, self-cleaning of such surfaces becomes necessary for cost-effective operation of energy harvesting devices with high performances. The essential component of self-cleaning is to generate surface hydrophobicity. One of the important current challenges is to enhance the hydrophobicity of the transparent substances while keeping the optical transmittance high. Since surface texturing is essential toward enhancing hydrophobicity, the textured surface modifies the optical scattering at the surface while lowering the optical transmittance. Therefore, a trade exists between the texture parameters and the optical properties of the surface. Therefore, optimization study incorporating the artificial techniques becomes essential to solve such coupling problem. In addition, green processes for generating surface hydrophobicity has a critical importance to minimize the hazardous environmental effects. Consequently, the chemicals involved with surface texturing and lowering free energy of surfaces should be selected carefully in the process. Although many techniques have been introduced for cost-effective processes, one-step method for surface texturing and lowering surface energy has yet to be developed. Hence, the challenge is not only to development of cost-effective one-step method, but resulting surface should balance between the hydrophobicity and the optical transmittance of the textured surface. Therefore, the future research directions should fulfill the current challenges while meeting the requirements of cost-effective green processing.

Acknowledgment The authors acknowledge the financial support of King Fahd University of Petroleum and Minerals (KFUPM) through Project# MIT11111-11112 to accomplish this work.

830

Hydrophobic Materials

References [1] Strauss JA, Soave PA, Ribeiro RS, Horowitz F. Absorber and self-cleaning surfaces on modified polymer plates for solar harvesting in the humid (sub)tropics. Sol Energy 2015;122:579–86. [2] Bhushan B, Jung YC, Koch K. Self-cleaning efficiency of artificial superhydrophobic surfaces. Langmuir 2009;25:3240–8. [3] Azimi G, Dhiman R, Kwon H-M, Paxson AT, Varanasi KK. Hydrophobicity of rare-earth oxide ceramics. Nat Mater 2013;12:315–20. [4] Han JT, Xu X, Cho K. Diverse access to artificial superhydrophobic surfaces using block copolymers. Langmuir 2005;21:6662–6. [5] Shirtcliffe NJ, McHale G, Newton MI, Chabrol G, Perry CC. Dual-scale roughness produces unusually water-repellent surfaces. Adv Mater 2004;16:1929–32. [6] Kinoshita H, Ogasahara A, Fukuda Y, Ohmae N. Superhydrophobic/superhydrophilic micropatterning on a carbon nanotube film using a laser plasma-type hyperthermal atom beam facility. Carbon N.Y. 2010;48:4403–8. [7] Latthe SS, Imai H, Ganesan V, Rao AV. Superhydrophobic silica films by sol–gel co-precursor method. Appl Surf Sci 2009;256:217–22. [8] Ma M, Mao Y, Gupta M, Gleason KK, Rutledge GC. Superhydrophobic fabrics produced by electrospinning and chemical vapor deposition. Macromolecules 2005;38:9742–8. [9] Ta VD, Dunn A, Wasley TJ, et al. Laser textured superhydrophobic surfaces and their applications for homogeneous spot deposition. Appl Surf Sci 2016;365:153–9. [10] Zhang X, Guo Y, Zhang P, Wu Z, Zhang Z. Superhydrophobic CuO@Cu2S nanoplate vertical arrays on copper surfaces. Mater Lett 2010;64:1. [11] Yilbas BS, Matthews A, Karatas C, et al. Laser texturing of plasma electrolytically oxidized aluminum 6061 surfaces for improved hydrophobicity. J Manuf Sci Eng Trans ASME 2014;136:54501. [12] Yilbas BS, Khaled M, Abu-Dheir N, et al. Wetting and other physical characteristics of polycarbonate surface textured using laser ablation. Appl Surf Sci 2014;320:21–9. [13] Liu Y, Chen J, Guo D, Cao M, Jiang L. Floatable, self-cleaning, and carbon-black-based superhydrophobic gauze for the solar evaporation enhancement at the air–water interface. ACS Appl Mater Interfaces 2015;7:13645–52. [14] Yilbas BS, Ali H. Laser texturing of Hastelloy C276 alloy surface for improved hydrophobicity and friction coefficient. Opt Lasers Eng 2016;78:40–147. [15] Tavares MTS, Santos ASF, Santos IMG, et al. TiO2 PDMS nanocomposites for use on self-cleaning surfaces. Surf Coatings Technol 2014;239:16–9. [16] Yilbas BS, Ali H, Al-Aqeeli N, Khaled M, Abu-Dheir N, Varanasi KK. Solvent-induced crystallization of a polycarbonate surface and texture copying by polydimethylsiloxane for improved surface hydrophobicity. J Appl Polym Sci 2016;133:43467. [17] Shen Y, Liu D, Zhang W, Dearden G, Watkins K. Ultrafast laser surface wettability modification on alumina surface. Chinese Opt Lett 2013;11:21403. [18] Yilbas BS, Khaled M, Abu-Dheir N, Aqeeli N, Furquan SZ. Laser texturing of alumina surface for improved hydrophobicity. Appl Surf Sci 2013;286:161–70. [19] Harimkar SP, Dahotre NB. Effect of laser fluence on surface microstructure of alumina ceramic. Adv Appl Ceram 2006;105:304–8. [20] Yilbas BS, Akhtar SS, Karatas C. Laser carbonitriding of alumina surface. Opt Lasers Eng 2011;49:341–50. [21] Yilbas BS, Ali H, Al-Aqeeli N, Oubaha MM, Khaled M, Abu-Dheir N. Laser gas assisted nitriding and sol–gel coating of alumina surfaces: effect of environmental dust on surfaces. Surf Coatings Technol 2016;289:11–22. [22] Yilbas BS, Karatas C, Arif AFM, Abdul Aleem BJ. Laser gas assisted nitriding of alumina surfaces. Surf Eng 2009;25:235–40. [23] Yuan Z, Bin J, Wang X, et al. Preparation of a polydimethylsiloxane (PDMS)/CaCO3 based superhydrophobic coating. Surf Coatings Technol 2014;254:97–103. [24] Adithyavairavan M, Subbiah S. A morphological study on direct polymer cast micro-textured hydrophobic surfaces. Surf Coatings Technol 2011;205:4764–70. [25] Lin Y-C, Hsu S-H, Chung Y-C. Thermal imprint techniques for preparation of superhydrophobic polymer coatings. Surf Coatings Technol 2013;231:501–6. [26] Kim DS, Lee B-K, Yeo J, Choi MJ, Yang W, Kwon TH. Fabrication of PDMS micro/nano hybrid surface for increasing hydrophobicity. Microelectron Eng 2009;86:1375–8. [27] Yeo J, Choi MJ, Kim DS. Robust hydrophobic surfaces with various micropillar arrays. J Micromech Microeng 2010;20:25028. [28] Yoon TO, Shin HJ, Jeoung SC, Park Y-I. Formation of superhydrophobic poly(dimethysiloxane) by ultrafast laser-induced surface modification. Opt Express 2008;16 (17):12715. [29] Adamson AW, Gast AP. Physical chemistry of surfaces. New York, NY: Wiley; 1997. [30] Parkin IP, Palgrave RG. Self-cleaning coatings. J Mater Chem 2005;15:1689–95. [31] Callies M, Quéré D. On water repellency. Soft Matter 2005;1:55–61. [32] Fürstner R, Wilhelm B, Christoph N, Peter W. Wetting and self-cleaning properties of artificial superhydrophobic surfaces. Langmuir 2005;21:956–61. [33] Fusetani N. Biofouling and antifouling. Nat Prod Rep 2004;21:94. [34] Muthiah P, Bhushan B, Yun K, Kondo H. Dual-layered-coated mechanically-durable superomniphobic surfaces with anti-smudge properties. J Colloid Interface 2013;409:227–36. [35] Howarter J, Youngblood J. Self-cleaning and next generation anti-fog surfaces and coatings. Macromol Rapid Commun 2008;29:455–66. [36] Bixler G, Bhushan B. Fluid drag reduction and efficient self-cleaning with rice leaf and butterfly wing bioinspired surfaces. Nanoscale 2013;5:7685. [37] Yilbas BS, Ali H, Al-Aqeeli N, Abu-Dheir N, Khaled M. Influence of mud residues on solvent induced crystalized polycarbonate surface used as PV protective cover. Sol Energy 2016;125:282–93. [38] Young T. An essay on the cohesion of fluids. Philos Trans R Soc London 1805;95:65–87. [39] Israelachvili JN. Intermolecular and surface forces. 3rd ed. Cambridge, MA: Elsevier Academic Press; 2011. [40] Shibuichi S, Onda T, Satoh N, Tsujii K. Super water-repellent surfaces resulting from fractal structure. J Phys Chem 1996;100:19512–7. [41] Wenzel RN. Resistance of solid surfaces to wetting by water. Ind Eng Chem 1936;28:988–94. [42] Cassie ABD, Baxter S. Wettability of porous surfaces. Trans Faraday Soc 1944;40:546. [43] Bhushan B, Jung YC, Koch K. Micro-, nano- and hierarchical structures for superhydrophobicity, self-cleaning and low adhesion. Philos Trans R Soc London A Math Phys Eng Sci 2009;367:1631–72. [44] Michael N, Bhushan B. Hierarchical roughness makes superhydrophobic states stable. Microelectron Eng 2007;84:382–6. [45] Lafuma A, Quéré D. Superhydrophobic states. Nat Mater 2003;2:457–60. [46] Deng T, Varanasi KK, Hsu M, et al. Nonwetting of impinging droplets on textured surfaces. Appl Phys Lett 2009;94:133109. [47] Reyssat M, Yeomans JM, Quéré D. Impalement of fakir drops. Europhys Lett 2008;81:26006. [48] Bocquet L, Lauga E. A smooth future? Nat Mater 2011;10:334–7. [49] Tuteja A, Choi W, Ma M, et al. Designing superoleophobic surfaces. Science 2007;318:1618–22. [50] Wong T-S, Kang SH, Tang SKY, et al. Bioinspired self-repairing slippery surfaces with pressure-stable omniphobicity. Nature 2011;477:443–7. [51] Ebert D, Bhushan B. Transparent, superhydrophobic, and wear-resistant coatings on glass and polymer substrates using SiO2, ZnO, and ITO nanoparticles. Langmuir 2012;28:11391–9. [52] Yao X, Hu Y, Grinthal A, Wong T-S, Mahadevan L, Aizenberg J. Adaptive fluid-infused porous films with tunable transparency and wettability. Nat Mater 2013;12:529–34. [53] Li Q, Rudolph V, Peukert W. London-van der Waals adhesiveness of rough particles. Powder Technol 2006;61:248–55. [54] Yilbas BS. Laser texturing of zirconia surface with presence of TiC and B4C: surface hydrophobicity, metallurgical, and mechanical characteristics. Ceram Int 2014;40:16159–67. [55] Azimi G, Kwon H-M, Varanasi KK. Superhydrophobic surfaces by laser ablation of rare-earth oxide ceramics. MRS Commun 2014;4:1–5. [56] Triantafyllidis D, Li L, Stott FH. Surface treatment of alumina-based ceramics using combined laser sources. Appl Surf Sci 2002;186:140–4. [57] Jagdheesh R. Fabrication of a superhydrophobic Al2O3 surface using picosecond laser pulses. Langmuir 2014;30:12067–73.

Hydrophobic Materials

831

[58] Ta DV, Dunn A, Wasley TJ, et al. Nanosecond laser textured superhydrophobic metallic surfaces and their chemical sensing applications. Appl Surf Sci 2015;357:248–54. [59] Pan L, Yu G, Zhai D, et al. Hierarchical nanostructured conducting polymer hydrogel with high electrochemical activity. Proc Natl Acad Sci USA 2012;109:9287–92. [60] Ke Q, Fu W, Jin H, Zhang L, Tang T, Zhang J. Fabrication of mechanically robust superhydrophobic surfaces based on silica micro-nanoparticles and polydimethylsiloxane. Surf Coatings Technol 2011;205:4910–4. [61] Jin M, Feng X, Zhai J, Cho K, Feng L, Jiang L. Super-hydrophobic PDMS surface with ultra-low adhesive force. Macromol Rapid Commun 2005;26:1805–9. [62] Shao G, Wu J, Cai Z, Wang W. Fabrication of elastomeric high-aspect-ratio microstructures using polydimethylsiloxane (PDMS) double casting technique. Sensors Actuators A Phys 2012;178:230–6. [63] Yang L, Hao X, Wang C, Zhang B, Wang W. Rapid and low cost replication of complex microfluidic structures with PDMS double casting technology. Microsyst Technol 2014;20:1933–40. [64] Gitlin L, Schulze P, Belder D. Rapid replication of master structures by double casting with PDMS. Lab Chip 2009;9:3000. [65] Yeo J, Kim DS. The effect of the aspect ratio on the hydrophobicity of microstructured polydimethylsiloxane (PDMS) robust surfaces. Microsyst Technol 2010;16:1457–63. [66] Stöber W, Fink A, Bohn E. Controlled growth of monodisperse silica spheres in the micron size range. J Colloid Interface Sci 1968;26:62–9. [67] Bogush GH, Tracy MA, Zukoski CF. Preparation of monodisperse silica particles: control of size and mass fraction. J Non Cryst Solids 1988;104:95–106. [68] Ibrahim IAM, Zikry AAF, Sharaf MA, Zikry A. Preparation of spherical silica nanoparticles: stober silica. J Am Sci 2010;6:985–9. [69] Wang X-D, Shen Z-X, Sang T, et al. Preparation of spherical silica particles by Stöber process with high concentration of tetra-ethyl-orthosilicate. J Colloid Interface Sci 2010;341:23–9. [70] Suratwala TI, Hanna ML, Miller EL, et al. Surface chemistry and trimethylsilyl functionalization of Stöber silica sols. J Non Cryst Solids 2003;316:349–63. [71] Yong WYD, Zhang Z, Cristobal G, Chin WS. One-pot synthesis of surface functionalized spherical silica particles. Colloids Surfaces A Physicochem Eng Asp 2014;460:151–7. [72] Bravo J, Zhai L, Wu Z, Cohen RE, Rubner MF. Transparent superhydrophobic films based on silica nanoparticles. Langmuir 2007;23:7293–8. [73] Ling XY, Phang IY, Vancso GJ, Huskens J, Reinhoudt DN. Stable and transparent superhydrophobic nanoparticle films. Langmuir 2009;25:3260–3. [74] Li Y, Liu F, Sun J. A facile layer-by-layer deposition process for the fabrication of highly transparent superhydrophobic coatings. Chem Commun 2009;14:2730–2. [75] Ogihara H, Xie J, Saji T. Controlling surface energy of glass substrates to prepare superhydrophobic and transparent films from silica nanoparticle suspensions. J Colloid Interface Sci 2015;437:24–7. [76] Epstein AK, Wong T-S, Belisle RA, Boggs EM, Aizenberg J. Liquid-infused structured surfaces with exceptional anti-biofouling performance. Proc Natl Acad Sci USA 2012;109:13182–7. [77] Huang X, Chrisman JD, Zacharia NS. Omniphobic slippery coatings based on lubricant-infused porous polyelectrolyte multilayers. ACS Macro Lett 2013;2:826–9. [78] Smith JD, Dhiman R, Anand S, et al. Droplet mobility on lubricant-impregnated surfaces. Soft Matter 2013;9:1772–80. [79] Kim P, Kreder MJ, Alvarenga J, Aizenberg J. Hierarchical or not? Effect of the length scale and hierarchy of the surface roughness on omniphobicity of lubricant-infused substrates. Nano Lett 2013;13:1793–9. [80] Anand S, Paxson AT, Dhiman R, Smith JD, Varanasi KK. Enhanced condensation on lubricant-impregnated nanotextured surfaces. ACS Nano 2012;6:10122–9. [81] Kim P, Wong T-S, Alvarenga J, Kreder MJ, Adorno-Martinez WE, Aizenberg J. Liquid-infused nanostructured surfaces with extreme anti-ice and anti-frost performance. ACS Nano 2012;6:6569–77. [82] Yilbas BS, Yousaf MR, Ali H, Al-Aqeeli N. Replication of laser-textured alumina surfaces by polydimethylsiloxane: improvement of surface hydrophobicity. J Appl Polym Sci 2016;133:44015. [83] Huang Z, Cho S-J, Jiang D, Tan S. Surface nitridation of Al2O3 based composite by N2-HIP post-treatment. J Mater Sci 1999;34:2023–7. [84] Yilbas BS, Akhtar SS, Karatas C. Laser cutting of alumina tiles: heating and stress analysis. J Manuf Process 2013;15:14–24. [85] van Oss CJ, Good RJ, Busscher RJ. Estimation of the polar surface tension parameters of glycerol and formamide, for use in contact angle measurements on polar solids. J Dispers Sci Technol 1990;11:75–81. [86] Sun C-C, Lee S-C, Hwang W-C, Hwang J-S, Tang I-T, Fu Y-S. Surface free energy of alloy nitride coatings deposited using closed field unbalanced magnetron sputter ion plating. Mater Trans 2006;47:2533–9. [87] Ji AL, Ma LB, Liu C, Li CR, Cao ZX. Synthesis and characterization of superhard aluminum carbonitride thin films. Diam Relat Mater 2005;14:1348–52. [88] Aguiar KR, Santos VG, Eberlin MN, et al. Efficient green synthesis of bis(cyclic carbonate)poly(dimethylsiloxane) derivative using CO2 addition: a novel precursor for synthesis of urethanes. RSC Adv 2014;4:24334–43. [89] Smith AL. The analytical chemistry of silicones. New York, NY: Wiley; 1991. [90] Gaboury SR, Urban MW. Spectroscopic evidence for Si–H formation during microwave plasma modification of poly(dimethylsiloxane)elastomer surfaces. Polym Commun 1991;32:390–2. [91] Kim SH, Dugger MT, Mittal KL. Adhesion aspects in MEMS-NEMS. Boca Raton, FL: CRC Press; 2011. [92] Kalderon-Asael B, Erel Y, Sandler A, Dayan U. Mineralogical and chemical characterization of suspended atmospheric particles over the east Mediterranean based on synoptic-scale circulation patterns. Atmos Environ 2009;43:3963–70.

Further Reading Bushan B. Biomimetics, bioinspired hierarchical-structured surfaces for green science and technology. Berlin; New York, NY: Springer; 2016. ISBN: 978-3-319-28284-8. Daoud WA. Self-cleaning materials and surfaces: a nanotechnology approach. Chichester: Wiley; 2013. ISBN-13: 978-1119991779. Junhui He. Self-cleaning coatings: structure, fabrication and application. Cambridge: Royal Society of Chemistry; 2016. ISBN: 978-1-78262-286-4. Xiangyu Yin, Bo Yu. Antifouling self-cleaning surfaces. Berlin; Heidelberg: Springer; 2014. ISBN: 978-3-662-45203-5.

Relevant Websites http://www.blogionik.org/self-cleaning-surfaces/. Blogionik. http://jncc.defra.gov.uk/page-5592. Joint Nature Conservation Committee. http://www.explainthatstuff.com/how-self-cleaning-windows-work.html. Explain That Stuff.

2.26 Dust Repellent Materials Bekir S Yilbas, Ghassan Hassan, and Abdullah Al-Sharafi, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia r 2018 Elsevier Inc. All rights reserved.

2.26.1 2.26.2 2.26.2.1 2.26.2.2 2.26.2.3 2.26.3 2.26.4 2.26.5 2.26.5.1 2.26.6 2.26.6.1 2.26.6.1.1 2.26.6.1.2 2.26.6.1.3 2.26.6.1.4 2.26.6.1.5 2.26.6.2 2.26.6.2.1 2.26.6.2.1.1 2.26.6.2.1.2 2.26.7 2.26.7.1 2.26.7.1.1 2.26.7.1.2 2.26.7.1.3 2.26.7.1.4 2.26.7.1.5 2.26.7.2 2.26.8 2.26.8.1 2.26.8.2 2.26.8.3 2.26.9 2.26.9.1 2.26.9.1.1 2.26.9.1.2 2.26.9.1.3 2.26.10 2.26.10.1 2.26.10.2 2.26.10.3 2.26.10.4 2.26.10.5 2.26.10.6 2.26.10.7 2.26.10.8 2.26.10.9 2.26.11 2.26.11.1 2.26.11.2 2.26.11.3 2.26.11.4

832

Introduction Some Methods of Dust Removal Dust Migration (Flow) Around the Globe Dust and Mud Adhesion Dust Mitigation Techniques Surface Characteristic Surface Characterization Tools Dust Effect on Selective Surfaces for Solar Energy Harvesting Effect of Dust on PV Modules Output Power Analysis of Dust Adhesion on Hydrophilic and Hydrophobic Surfaces Particles Detachment and Contact Analysis Johnson–Kendall–Roberts analysis Derjaguin–Muller–Toporov analysis The Maugis and Pollock model Hamaker model Rumpf–Rabinovich model Free Surface Energy and Adhesion Surface free energy calculations Wenzel's model Cassie–Baxter model Dynamics of Dust Removal and Surface Cleaning External Dust-Removal Forces Gravitational force Shear stress and inertial force (lifting force) Drag forces (pressure and shear stress) Centrifugal force Frictional force The Mechanism of Particle Removal Surface Characteristics (Hydrophobic and Super-Hydrophobic Coating) Mimicking Nature Hierarchical Structure in Plant Leaves Plant Leaves With Unitary Structure Surface Texturing (Materials, Mechanism, and Processes) to Produce Hydrophobic and Super-Hydrophobic Surfaces Roughening of Low Surface Energy Material Silicones polydimethylsiloxane Organic and inorganic materials Fluorocarbons Producing and Modifying a Rough Surface Using Material of Low Surface Energy Hydrothermal Reaction and Wet Chemical Reaction Electrochemical Deposition Layer-by-Layer and Self-Assembly Lithography Etching and Chemical Vapor Deposition Electrospinning Technique Sol–Gel Method and Polymerization Reaction Other Techniques Impact Dynamics of Water Droplets on Super-Hydrophobic Surfaces Dust Characterization Particle Sizing Analysis Particle Chemical Analysis Dust Particle Chemistry Dust Particle Adhesion

Comprehensive Energy Systems, Volume 2

833 834 834 834 835 835 836 836 837 838 838 839 839 840 841 841 841 841 842 842 843 843 843 843 844 844 844 844 845 845 845 845 845 845 845 848 848 848 848 848 849 849 849 850 850 852 852 852 853 854 854 857

doi:10.1016/B978-0-12-809597-3.00254-6

Dust Repellent Materials 2.26.12 Electrostatic Dust-Removal Methods 2.26.13 Characterization and Analysis of Environmental Dust 2.26.13.1 Dust Particles Characterization 2.26.13.2 Analysis of the Mud Formed From the Dust Particles 2.26.13.3 Analysis of Mud Residues 2.26.14 Environmental Dust Effect on Solar Thermal Selective Surfaces 2.26.14.1 Laser Texturing of Zirconia Surface and Environmental Dust Effect 2.26.14.1.1 Laser texturing process and parameters 2.26.14.1.2 Results of laser texturing and dust adhesion on zirconium surface 2.26.14.2 Laser Surface Texturing of Tungsten and Environmental Dust Effect on Textured Surfaces 2.26.14.2.1 Laser texturing parameters and surface characteristics 2.26.14.2.2 Results and discussion 2.26.14.3 Laser Surface Texturing of Alumina Surface and Environmental Dust Effects on Textured Surfaces 2.26.14.3.1 Experimental 2.26.14.3.2 Results and Discussion 2.26.15 Conclusions 2.26.15.1 Future Directions Acknowledgment References Relevant Websites

Nomenclature A CA CAH f F Fg g m OTS PDMS PFOTS r rc

2.26.1

Hamaker constant Contact angle Contact angle hysteresis Surface fraction of interface Force Gravitational force Gravitational acceleration Mass of the particle Octadecyltrichlorosilane Polydimethylsiloxane Trichloro(1H,1H,2H,2H-perfluorooctyl) Roughens ratio Distance of the particle from the rotational center

R Re TEOS u v UV–vis WCA y f g r m o mf

833 858 859 860 861 863 863 863 864 864 867 868 868 871 872 872 875 877 877 877 880

Radius of the particle Reynolds number Tetraethoxysilane Friction velocity Kinetic viscosity Ultraviolet-visible Water contact angle Contact angle Interfacial area Surface energy between two surfaces Density of the dust particle Dynamic viscosity of fluid Angular velocity of the disk Friction coefficient

Introduction

Solar energy harvesting has a crucial role for widening renewable energy applications around the Globe. Utilization of high temperature resistant ceramic components becomes unavoidable for solar thermal applications. Zirconium nitride can serve as a selective surface for solar energy harvesting in thermal systems, which is because of excellent optical properties in terms of absorption and emission [1]. On the other hand, climate change has imitated dust storms around the Middle East [2]. The dust settlement on solar energy harvesting surfaces lowers device performance in terms of output power and efficiency [3]. The environmental dust particles compose of various elements and compounds including alkaline and alkaline earth metals [4]. In humid air ambient, water condensates on the dust particles and some compounds of the dust particles dissolve into water condensate. This forms chemically active liquid solution, which accumulates at the interface of the device surface and the dust particles under the gravity [4]. Chemically active liquid solution causes some asperities on the device surface, such as pin holes and pit sites, while causing permeant damages [5]. In addition, once the liquid solution dries at the device surface, adhesion between the dried liquid and the surface becomes significantly strong and the efforts required removing the dried solution and the mud from the surface increase significantly. One of solutions to create self-cleaning characteristics at the surface is to improve surface hydrophobicity. In general, surface hydrophobicity is associated with the micro/nanoscale surface texture and low surface free energy. However, plane yttria stabilized zirconia surface has high surface free energy and demonstrates hydrophilic characteristics. Consequently, altering characteristics of zirconia surface through surface processing becomes essential to generate surface hydrophobicity. Several methods have been suggested and many techniques were reported for generating hydrophobic characteristics at the surfaces [6–11]. The techniques reported were involved with multi-steps processes and harsh conditions. Some of

834

Dust Repellent Materials

these processes include phase separation [6], electrochemical deposition [7], plasma treatment [8], sol–gel processing [9], electrospinning [10], and solution immersion [11]. Laser gas assisted texturing offers considerable advantages over the multi-step processes for hydrophobizing surfaces, which is particularly true for ceramic surfaces [12]. Some of these advantages include high speed processing, precision of operation, and low cost. In laser texturing process, combination of melting and evaporation of ceramic surface takes place via ablation [12]. This in turn generates surface texture consisting of micro/nano poles and cavities. The use of high pressure assisting gas, such as nitrogen, generates nitride compounds on the laser textured surface [12]; in which case, nitride compounds have low surface energy than the oxides [12]. In addition, arrays of micro/nano size pole/pillar improve the water-repelling capacity of the laser textured surface. However, dust settlement on the laser textured surface changes the surface characteristics, such as, surface hydrophobicity. Hence, investigation of environmental dust particle adhesion and influence of dried mud solution on the characteristics of laser textured zirconia surface becomes essential.

2.26.2

Some Methods of Dust Removal

Several means of dust-removal mechanisms are present, such as natural means, electrostatic means, mechanical means, and selfcleaning by means of surface nano-films. The natural powers such as gravitation and wind are not efficient in dust-removal process. The mechanical technique that used to clean the surfaces mainly includes blowing, brushing, and ultrasonic driving and vibrating. There are many problems associated with the cleaning process by using mechanical methods. Firstly, the brushing cleaning method is inefficient because of the strong adhesion of the dust particles on surfaces. Secondly, the surfaces were damaged by the brush during wiping. However, air blowing is ineffective due to unsatisfactory maintainability of the equipment used and high energy consumption during the process. Consequently, self-cleaning methods becomes a promising technique for cleaning the dusty surfaces. The self-cleaning Nano-film method required the surface to be covered by a pellucid nano-film that made the surface superhydrophobic.

2.26.2.1

Dust Migration (Flow) Around the Globe

According to the World Meteorological Organisation (WMO) protocol, the events of dust flow has been classified into different categories based on the dust visibility. The first category includes the dust-in-suspension that reduces the visibility up to 10 km. The second category is involved with the blowing dust that raised sand and dust at the observation time and reduces the visibility from 1 to 10 km. The third is the dust storm with strong winds, which lift of soil and dust particles and visibility reduces to 200–1000 m. Finally, severe dust Storm with strong winds that lifting large dust quantities and decreasing the visibility to 200 m [13]. It was reported that dust emission varied with space and time and greatly depended on land surface and atmospheric conditions. Different technologies and measurement methods is established to track the flow of dust around the globe. Satellite has been used extensively to track the dust migration and identifying the physical quantities. Estimation of dust physical quantities are difficult due to the mixture of signals (from the dust, clouds, land surface, and other aerosol) that detected by the satellite. Therefore, satellite data can be integrated with other radiation measurements to insure accurate estimations [13]. Dust storms originate in semi-arid and arid regions (e.g., Mongolia, the Middle East, and the Sahara). Dust storms throughout these places were placed more than 5000 million tons of dust into the atmosphere per year. North Africa region is considered as the world's largest source of dust to the atmosphere which emitted around 1012 g of dust annually. 250  103 g of dust is transported over the tropical North Atlantic Ocean [13]. Dust storms that originate from the Sahara (North Africa) migrate and influence the air quality and climate of the Middle East, Africa, Asia, Europe, the Americas, and the Caribbean. Middle East is considered as the most affected region by dust because of the frequent dust event around the year [14]. The wind force passing over wide deserts motivate the loosely sand particles to vibrate and leap. Then, the small dust particle starts to travel in suspension. The regional sand-dust storms have a wide extension that covers different countries. Removing of vegetation cover is a direct cause of dust storms generation that result in loosening the top soil cover that prevent the existence of loosely sand particles [15]. The map shown in the Fig. 1 demonstrates the dust potential rating classes around the globe [3]. It has been observed that the Middle East North Africa (MENA) region and Australia exhibiting the largest dust potential.

2.26.2.2

Dust and Mud Adhesion

Dust settlement onto the surfaces located to open environments causes irrecoverable damages on surfaces and lowers the system performance, such as those associated with solar thermal and solar photovoltaic (PV) applications. There are many surface treatment methods being reported in the literature for minimizing the dust and mud effect on the surface characteristics and surface performance [16]. However, self-cleaning or cost-effective removal of dust particles from such surfaces remains challenging. The dissolved ions (Na þ , K þ , Ca þ 2) from dust particles attract mud and clay minerals while forming different structure. These ions penetrate the mud layers and hold them together. The process of dust–water interaction and mud formation can take place according to the following steps: soil and dust particles surfaces are composed of hydroxyls and oxygen layers; therefore, hydroxyls attracted the negative corners and oxygen with positive corners which easily generated hydrogen bonding between water molecules. Dust particles reactions take place with molecules of polarized water which separate single ions (Na þ , K þ , Ca þ 2)

Dust Repellent Materials

180°

130°W

80°W

30°W

20°E

70°E

120°E

170°E 70′N

70′N

50′N

60′N

30′N

30′N

Dust potential rating class

10′N

835

10′N

None Very low Low

10′S

10°S

Moderate High 30′S

80°S

Very high

60°S 50′S

70′S 0

1000 2000 3000 4000 5000

kilometers

180°

130°W

80°W

30°W

20°E

70°E

120°E

170°E

70°S Projection: World Robinson, WGS 1984

Fig. 1 Map of global dust potential (integrated desert terrain forecasting for military operations). Reproduced from Sissakian VK, Al-Ansari N, Knutsson S. Sand and dust storm events in Iraq. Nat Sci 2013;5(10):1084–94.

surrounded by hydration sheaths and dissolved in the water molecules forming oxygen and hydroxyls (KOH, NaOH, etc) stretch. Solid undissolved dust particles have pendant hydroxyl (–OH) groups which can form bonds or strong polar attractions to mud solution while forming inorganic surfaces. During drying, the dissolved ions (Na þ , K þ , Ca þ 2, Cl , SiO ) attract mud structure due to electrostatic and ionic bonding force. These ions dissolved in the mud solution while particles holding them together and forming crystals in between. It stays in the mud structure during water evaporation via increasing the adhesion force.

2.26.2.3

Dust Mitigation Techniques

In reviewing the various dust-impact mitigation approaches, the focus is on the techniques that deal with the effect of dust as it is one of the most significant environmental factors affecting PV module performance in the regions where solar power may be a particularly viable alternative to fossil fuels as a result of the high level of sunshine. To recover the performance of a PV module that is covered with dust, it is important to carry out periodic cleaning. However, the required frequency of cleaning depends on environmental conditions. There are a variety of methods that have been used or developed to mitigate the dust effect on PV module performance. Some also contribute to mitigating other climate factors as well, such as temperature and humidity. The reported dust mitigation methods in the literature can be divided into four categories: spontaneous, mechanical and electromechanical, electrostatic shields, and micro- and nano-scale surface functionalization. The purpose of micro- and nano-scale surface fabrication is to develop self-cleaning surfaces with optimal optical properties. This method enables the creation of a superhydrophobic surface that has low wettability and high water droplet mobility. Such a surface can enhance the cleaning efficiency and thereby reduce the necessary cleaning frequency. Super-hydrophobic surfaces are comprised of a micro- or nano-structure surface coated by a thin film of low surface energy material or vice versa [17].

2.26.3

Surface Characteristic

Most of today’s technologies, such as self-cleaning technologies, have been derived from nature. Several surfaces in nature have self-cleaning features, for example, butterfly’s wings and plant leaves. During 20th century, this technology has received great attention due to the wide range of applications (PV cleaning, cement to textiles, etc.). In addition, this technology provides different solutions in reducing maintenance cost, reducing cleaning time, and eliminating manpower efforts [18]. Coatings used in self-cleaning applications are classified due to the action of water into hydrophobic and hydrophilic categories. For hydrophobic technique, the droplet rolls, and slides thereby cleaned the surface, while in the hydrophilic technique the cleaning takes place by water spreading on the surface [18]. The self-cleaning property is mainly affected by surface contact angle (CA). The surface contact angle is the angle formed between the solid surface and liquid droplet at the three phase boundaries. The surfaces are classified based on water contact angle (WCA) as: hydrophilic surface for CAo90 degrees, hydrophobic surface for CA490 degrees, super (Ultra) hydrophobic for CA4150 degrees, and super (Ultra) hydrophilic for CA close to zero degree (see Fig. 2).

836

Dust Repellent Materials

  < 90° - Hydrophilic surface

  > 90° - Hydrophobic surface

  > 150° - Ultra-hydrophobic surface Fig. 2 The contact angle representation for hydrophobic, hydrophilic, and super-hydrophobic surfaces. Reproduced from Ganesh VA, Raut HK, Nair AS, Ramakrishna S. A review on self-cleaning coatings. J Mater Chem 2011;21(41):16304–22.

Super-hydrophobic surfaces with water-repelling characteristics are of great importance among different areas (waterrepellent automotive parts, water-proofing of textiles, biofouling prevention, drag reduction in micro-channels, and self-cleaning windows) [19]. The common method for studying super-hydrophobic materials is based on static and dynamic CAs. Dynamic CAs are measured by evaluating the receding and advancing CAs for the sliding water drops [20].

2.26.4

Surface Characterization Tools

The study of dust and dry mud on selective surfaces undergoes several experimental techniques (mechanically and chemically) using different characterization tools. The characterization tools that used for dust particles examinations can be divided into three main groups: mechanical testing tools, microstructure characterization tools, and chemical investigation instruments. The mechanical equipment includes tensile testing machine and scratching testing machine that used to measure the interfacial forces (adhesion and cohesion force) between the dry mud and substrate. In addition, the microstructure characterization methods which included scanning electron microscopy (SEM), energy dispersive spectroscopy (EDS), atomic force microscopy (AFM), X-ray diffraction (XRD), and 3D imaging. The SEM is a multipurpose device available for the microstructure identification, chemical composition, and morphology exploration of dust particles, while EDS is an elemental analysis technique which permits to recognize what are the dust particles elements and the percentage of their quantities. The AFM provides a 3D profile of the dust particle and dry mud layer in nanoscale. The AFM measures the forces between the surface and sharp probe (usually made of silicon nitride) at short distance. The XRD known as compound analysis technique is used for phase identification and unknown solids determination of dust particles contents. The 3D optical microscope is a light microscope based on white light interferometry that uses the light photons beam to form detailed three-dimensional images from the specimen surface. The advantage of the image taken by the light electron microscope was its accuracy and high magnification which showed more information about surface topography. The 3D optical microscopy can provide detailed information about the shape and size of dust particles as well as measuring surface topography of dry mud layer. Dust particles and dry mud solution undergo several chemical investigations using Fourier transform infrared spectroscopy (FTIR) and PH measurements. FTIR technique can be conducted to obtain the emission or absorption infrared spectrum of dust solid particles. FTIR analysis provides different information such as identifying of unknown dust materials, determining the quality or consistency of a sample and the amount of dust components in a mixture. The PH is an abbreviation of “pondus hydrogenii” and was proposed by S.P.L. Sorensen in 1909. The PH measurements provide an indication of the mud solution weather it is basic or acidic solution. The definition based on hydrogen ion activity given as PH ¼ log10 aHþ .

2.26.5

Dust Effect on Selective Surfaces for Solar Energy Harvesting

Effective solar energy harvesting is of tremendous importance, as a solution for the increasing demand of power consumption that is coupled with the depletion of fossil fuels resources. The sector of renewable energy needs to provide a considerable share of future development in order to take advantage of incident solar radiation as a source of clean energy without involving any

Dust Repellent Materials

837

emission product. The fields of solar energy applications have grown steadily especially in electric power generation sector. Several factors motivated the use of renewable energy such as global warming reduction, low maintenance requirement, and cost effectiveness for remote areas with no electric grid connections. PV solar energy technologies and their applications have received global research attention and becoming an interesting area for financial investments [21,22]. The power output from solar energy systems depends on several environmental factors such as wind velocity, rainfall, ambient temperature, and dust accumulation on the system protective surfaces. The nature of dust accumulation and its effect on the PV module performance needs current attention. The literature review indicated the importance of the influence of dust deposition on protective surfaces in the energy production [23]. The effect of dust and aerosol minerals on anthropogenic still not obvious, because of uncertainties of product microscopic levels such as particles distribution, shape, and size [24]. On the other hand, the effect of dust soiling on solar PV modules has a great impact on light transmission, through protective covers of suggested and many (SAM). This soiling process on solar panels – due to airborne dust – happens due to dry dust deposition because of gravitational settling and wet particles deposition. Such soiling absorbs and reflects incident solar radiation leading to a reduction of light transmittance. Various problems are associated with dust deposition on solar energy protective surfaces such as a reduction in the output energy yield, difficulty in predicting the power output due to dust deposition losses, time and cost associated with cleaning solar protective surfaces. Self-cleaning or cost-effective removal of dust particles is still remaining challenging. The dust particles absorb water vapor in the humid air environments and forms mud at the surfaces and become hard to remove. Therefore, reducing the amount of accumulated dust on the solar energy protective surfaces is of immense importance.

2.26.5.1

Effect of Dust on PV Modules Output Power

The degradation in the performance of PV system is due to environmental and climate conditions. Significant advances have been made to improve PV systems performance within the last few decades, but the impacts of climate conditions (such as dust accumulation and high ambient temperature) on PV system performance still remain as a challenge. The contrast between environmental conditions from location to another around the world has a corresponding different level of impact on the performance of PV module system in different zones [25]. Several studies have been carried out on the effects of different environmental conditions on the performance of PV system. Mekhilef et al. [26] reviewed the effects of humidity level, dust accumulation and wind velocity surrounding the PV system. They demonstrated that each condition influences the other condition, and thus, all conditions should be considered together. The effect of dust accumulation on PV module output power during time periods, 1940–90 and 1990–2010 has been reviewed by Mani and Pillai [27]. They used a suitable maintenance/ cleaning cycle for their PV systems that considered the prevalent environmental and climatic conditions. However, in this survey section, the focus is on the effect of dust deposition on solar panels performance and the impact of other environmental conditions on the level of dust accumulation. Moreover, the impact of adhesion and cohesion forces on dust deposition as well as dust particles interactions is also considered. The effect of dry mud spots and the dust particles on the output power of PV Modules have been studied by numerous researchers. Kurokawat [28] studied the effect of small dirt spot between 5 and 100 mm in diameter on the power output of module surfaces. The power output for 3% uncleaned area was reduced by 50%. Elminir et al. [29] studied the effect of dust sediment on 100 samples of glass tested at various azimuth and tilt angles for more than 6 months exposure to the environment. The sediment of dust particles varied from 15.84 to 4.48 g/m2 for 0 to 90-degrees tilt angle with transmittance ranging from 52.5% to 12.4%, respectively. In addition, the output power of the device is decreased by 17.4% at 45-degrees tilt angle for 1 month exposure. Jiang et al. [30] investigated the deposition effect of airborne dust on PV panel surfaces. The main focus was to examine performance analysis of polycrystalline and monocrystalline silicon cells experimentally. Based on the experimental setup, the relative humidity was controlled to 60% and the PV module temperature was 251C using controlled air environments while a small fan was used to simulate the air velocity. Based on these conditions, it was observed that deposition of dust grew from 0 to 22 g/m2 the output efficiency reduction increased from 0 to 26%. Also, it was found that module surface material affected the rate of dust accumulation, for example, for the same concentration of dust, the glass surface modules have slow degradation rate than polycrystalline silicon modules. Brown et al. [31] used an artificial dust (o70 mm) for soiling test to predict the PV modules performance in Arizona. The PV modules using anti-soiling hydrophilic coating resulted in 5% performance improvement. Coating and other mitigation techniques has been reported later on in this review. Touati et al. [32] observed the effect of dust accumulation on different PV technologies and its sensitivity to humidity and temperature. The obtained results indicated that monocrystalline modules the efficiency reduced by 10% due to dust accumulation of 100 days. The authors recommended more durable panels for dust accumulation such as semi-flexible PV modules. Rajput et al. [33] studied the influence of dust deposition on the electrical efficiency of monocrystalline PV modules. The findings revealed that a maximum efficiency of 6.38% for clean modules and a maximum efficiency of 0.64% for dusty modules. Also, this result indicated a 92.11% of the reduction in modules power output and the reduction of modules efficiency was 89%. Ghazi et al. [34] presented a review on the effect of dust on flat surfaces in the Middle East and North Africa (MENA) region. The MENA region exhibited the worst dust deposition districts in the world. Sudan has the worse deposition of dust 9 times compared to the United Kingdom. Moreover, dust storms in Baghdad had a significant impact on the intensity of solar radiation [35]. Boyle et al. [36] did investigate the influence of transmissivity due to normal soiling on the glass covers of PV panels in the United States. They reported a 2 g/m2 of deposit dust after 5 weeks of exposure. In addition, every g/m2 accumulated on the surface reduced the transmissivity of light by 4.1%. The solar radiation transmissivity was not influenced by the incidence angle of irradiance which changed linearly with dust accumulation. The authors compared their results with other published results, and they found a large difference in the rate of dust

838

Dust Repellent Materials

deposition and its effect on the value of transmissivity. They concluded that the effect of dust accumulation on PV modules glass covers transmissivity is location dependent.

2.26.6

Analysis of Dust Adhesion on Hydrophilic and Hydrophobic Surfaces

The cleaning and detection of dust particles from surfaces is of great importance in many fields such as PV applications. A wide range of parameters significantly affect the particle–surface interactions. These parameters include particle shape, particle size, surface roughness, and adhesion bond between particles and surfaces [37]. Generally, adhesion forces between small particles and surfaces result from capillary forces, electrostatic forces, and van der Waals forces. Van der Waals forces basically originate from intermolecular attraction between the dust particles molecules and the surfaces. This force is an attractive force that prevents small particles resuspension. Furthermore, the capillary forces and electrostatic forces strongly influence the adhesion of particles. Strong capillary forces generated under humid environment due to the formation of water meniscus around the particles. This capillary force is much higher than van der Waals forces. In addition, the electrostatic forces are important during the presence of electric field which lead to particle charging. The Coulomb force magnitude depend on the electric field strength and the magnitude of particle charge. These forces (capillary and electrostatic forces) are attractive forces that add to van der Waals force. The environmental conditions and physicochemical properties during experiments will decide the predominance force among others [37]. The adhesion force is the sum of these different forces. Fad ¼ Fc þ Fes þ Fvdw

ð1Þ

where is the total adhesion force, Fc is the capillary force, Fes is the electrostatic force, and Fvdw is the van der Waals force. The adhesion force between particles and among particles is van der Waals interactions during the absence of external electrical field and without chemical bonding. This attraction force (van der Waals interactions) produces due to the random electrons movement that generates dipoles (concentrated charge areas). The van der Waals force is 10 times greater than the electrostatic force. Moreover, the van der Waals force can increase 5 times under compression force which increases the real area of contact between the particles and the surfaces. Generally, the electrostatic forces are small to influence adhesion forces when external forces supplied.

2.26.6.1

Particles Detachment and Contact Analysis

As indicated earlier, there are different adhesion forces which prevent the removal of small particle such as electrostatic forces, capillary force and van der Waals forces. Note that van der Waals force is always exist in detachment and adhesion of micro/ submicron particle size. Particle–surface interaction between two elastic spheres was firstly studied by Hertz [37]. A circular contact area is formed during applying of normal force as shown in the Fig. 3. Based on Hertz model the radius of contact circle aH is:
 Pdp 1=3 aH ¼ ð2Þ 2K where dp is the spherical particle diameter and K is Young’s modulus that is given as: ! 1 4 1 ðnÞ21 1 ðnÞ22 K¼ þ E1 E2 3

ð3Þ

where (n) and E are the Poisson’s ratio and Young’s modulus, respectively.

P

dp

Surface O Fad



Fig. 3 Attachment of plastic sphere particle with solid surface. Reproduced from Kohli R, Mittal KL. Developments in surface contamination and cleaning, Volume 4: detection, characterization, and analysis of contaminants. Norwich, NY: William Andrew; 2011.

Dust Repellent Materials

839

An attractive molecular force exists between the two surfaces in contact is due to the decrease in the surface energy. For elastic sphere in contact, an energy balance is produced between the stored elastic energy and released surface energy. The free energy loss Es and the force Fs can be represented as following expressions: pa2 Dg

Es ¼

and

ð4Þ

Fs ¼ dEs =dδ

ð5Þ

Es ¼

ð6Þ

Combining these equations with Hertz analysis gives: and




a¼

pRDg  3Fs R 1=3 4E

ð7Þ

The above theory is valid only when the Hertzian assumptions exist. Furthermore, several modifications have been added to the Hertz analysis considering the effect of adhesion and high elastic modulus spheres. Particles–substrate interactions have been evaluated theoretically using several adhesion measurements and analytical models.

2.26.6.1.1

Johnson–Kendall–Roberts analysis

Johnson, Kendall, and Roberts (JKR) formulated this theory in order to incorporate the adhesion effect that missing in the Hertzian contact by using of a balance between surface energy loss and the stored elastic energy [38]. JKR analysis considers only the adhesion and contact pressure effect inside the contact area. Based on the modified Hertz analysis, referred to as JKR analysis, expressions for attractive force and contact radius at no externally applied force are: Fs ¼

3 pRDg 2

ð8Þ

and a¼




9pDgs R2 2E

1=3

ð9Þ

where Fs is the force required to pull the surfaces apart and R is radius of particle and g is surface energy [39]. However, the contact radius of a particle based on JKR model in the presence of externally applied force and the force required to remove the particle JKR ) can be represented as: (Fad 2 3 sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 dP 4 3 3pdP WA 2 5 3 ð10Þ P þ pdP WA þ 3pdP WA P þ aJKR ¼ 2K 2 2 and

JKR Fad ¼

3 pdP WA 4

ð11Þ

where P is the externally applied force and WA is the thermodynamic work. The JKR model is valid only for relatively large surfaces with high surface energy. The JKR analysis predicts that the attractive force between the two surfaces in contact and the contact area should be finite at zero applied load [37].

2.26.6.1.2

Derjaguin–Muller–Toporov analysis

Derjaguin et al. [40] developed a model for a sphere with high elastic modulus whose profile does not change outside the contact area. They found the relation of contact radius and the corresponding pull of force to be: a3DMT ¼

dP ½P þ pdP WA Š 2K

ð12Þ

and JKR ¼ pdP WA FDMT

ð13Þ

For negligible elastic deformation of the sphere on a rigid surface in the absence of externally applied force, the Derjaguin–Muller–Toporov (DMT) model is given as: Es ¼ 2pRDg

ð14Þ

DMT equation is valid for both hard and small particles which has low surface energy. There are different assumptions in JKR and DMT models lead to different predictions. The JKR model predicts the particle-surface contact area to be finite at the moment of separation while the DMT predicts a zero contact radius. Another difference is that the surface force according to JKR analysis is assumed to act inside the contact region of particle and substrate. However, the DMT analysis considered the force outside contact

840

Dust Repellent Materials

area. From these differences in the assumptions, it can conclude that in terms of predicted pull-off force, the DMT model is 4/3rd greater than the JKR model. Muller et al. [41] reported that JKR model is appropriate for large particle with high surface energy and low Young’s modulus. However, the DMT model is more applicable for hard and small size particles with low surface energy and high Young’s modulus. The transition between the two models can be represented in terms of mT as: mT ¼

2.26.6.1.3



8dP WA2 9K 2 z30

1=3

ð15Þ

The Maugis and Pollock model

Maugis and Pollock (MP) [42] introduced the MP model for particle-surface adhesion considering plastic deformation. The model was developed by extending the JKR model and the DMT model. The contact radius can be predicted using MP model as: a2MP ¼

½P þ pdP WA Š pH

ð16Þ

where H is the material harness. It was observed that the JKR model and DMT model predicted the contact radius and the pull-off force for elastic contact between particle and surface interaction while the MP model considered the plastic deformation on the contact radius and pull-off force. Fig. 4 shows the variation between the non-dimensional resistance moment of (M ad ) adhesion with the external applied force  P predicted from different models. The maximum resistance moment for JKR model is just before the separation point at P ¼0.47 while for DMT model and MP model occur much before separation point at P ¼ 0.5 and P ¼ 0.44, respectively. Table 1 listed the maximum adhesion moment calculated from JKR, DMT, and MP models. 0.5

JKR 0.4

0.3 M *ad

DMT 0.2 MP 0.1

0

0

0.2

0.4 P*

0.6

0.8

Fig. 4 Variation of resistance moment of adhesion for JKR, DMT, and MP models. Reproduced from Kohli R, Mittal KL. Developments in surface contamination and cleaning, Volume 4: detection, characterization, and analysis of contaminants. Norwich, NY: William Andrew; 2011.

Table 1 The maximum resistance moment of adhesion for JKR, DMT, and MP models Adhesion model JKR model DMT model MP model

Maximum resistance moment of adhesion h 4 5 i1=3 WA dP JKR Mad; max ¼ 2:707 h K4 5 i1=3 WA dP DMT Mad; max ¼ 1:725 K MP Mad; max ¼

2pðdP WA Þ3=2 pffiffiffiffiffi 3 3H

Abbreviations: DMT, Derjaguin–Muller–Toporov; JKR, Johnson–Kendall–Roberts; MP, Maugis and Pollock.

Dust Repellent Materials 2.26.6.1.4

841

Hamaker model

This model is defined for a van der Waals body to body interaction, which is [26]: F¼

AR 12Z 2

ð17Þ

where Z is separation distance between surface and particles and A (A ¼ p2  C  r1  r2 ) is Haymaker constant, which ignores the influence of an intervening medium between the two particles of interaction.

2.26.6.1.5

Rumpf–Rabinovich model

Rumpf developed the Hamaker model by considering the effect of roughness height on the adhesion forces. In the Rumpf model, the surface is assumed to contain semi spherically shaped asperities with centers on the surface [43] " # AH rR R ad FRump ¼ þ ð18Þ 6Z02 r þ R ð1 þ r=Z0 Þ2 Rabinovich et al. [44] further developed the Rumpf model by incorporating the effects of both the height and width of surface asperities in their model [43] " # AH rR R ad FRabinovich ¼ þ ð19Þ 6Z02 r þ R ð1 þ ymax =Z0 Þ2

2.26.6.2

Free Surface Energy and Adhesion

The energy that is normally associated with bonding to other atoms is available at the surface. This energy required to create a new surface, is referred to as free surface energy. When a bond is formed between two materials with surface energies per unit area g1 and g2, interface with an interfacial energy per unit area, say, g12, is formed. The work of adhesion is the energy required to separate the two surfaces per unit area and it can be written as Dupré equation [45]: Wad ¼ Dg ¼ g1 þ g2

ð20Þ

g12

and g12 ¼ jg1

g2 j

ð21Þ 2

where Dg representing the reduction in the surface energy of the system per unit area (negative value in mJ/m , dynes/cm). For identical materials in contact, Dg represents the cohesion work that equal to 2g. It was observed from surface free energy theory that higher surface energy lead to stronger bonds between the solid surfaces in contact. Therefore, the theory of adhesion suggests materials with low surface energy (g) and low Dg. The total energy change during separation of two surfaces in contact is demonstrated in Fig. 5.

2.26.6.2.1

Surface free energy calculations

Surface energy can be defined also as the excess energy at the surface. The surface energy of solid–liquid in contact is ranged in between 0.05 and 0.2 J/m2 while for solids to vapor in contact is less than 2 J/m2. The surface energy will vary with the deformation of solid surfaces due to stretching of the interatomic bonds of the solid surface. This deformation can be written as:
 ∂G GS ¼ g þ ð22Þ ∂A

Solid 1, 1 Solid 1, 1

Contact Separation

Interface, 12

−12

WA

Solid 2, 2 Solid 2, 2

Fig. 5 The work of adhesion during contact interface separation. Reproduced from Petean P, Aguiar M. Determining the adhesion force between particles and rough surfaces. Pow Technol 2015;274:67–76.

842

Dust Repellent Materials

A critical situation when different phases meet together (gas, liquid, and solid). The material’s surface energy can be derived from Young’s equation which relates the surface tensions between the three phases: solid, liquid, and gas: or it can be written as

gSV ¼ gSL þ gLV cosyy cosyy ¼

gSV

ð23Þ

gSL

ð24Þ

gLV

where gSV is surface energy of solid against vapor, gSL is surface energy of solid against liquid and gLV is surface energy of liquid against vapor and yY Young’s CA between the liquid and the surface as shown in the Fig. 6. Young’s equation is proving that surface energy is greatly affecting by CA. The adhesion work of solid to liquid (WSL) can be expressed by combing Young's equation with Dupre’s equation. This combination indicates that the strength of adhesion is directly related to the CA between solid and liquid. It can be observed that the adhesion energy can be measured using the CA. WSL ¼ gSV þ gLV

gSL ¼ gLV ð1 þ cosyÞ

ð25Þ

Rough surfaces have different effects on the CAs of wetting liquids. Therefore, a modified form of Young’s equation is important to satisfy the practical working conditions. Cassie–Baxter and Wenzel are the two main models that attempt to describe the wetting of textured surfaces. 2.26.6.2.1.1 Wenzel's model Describes the homogeneous wetting regime, as shown in Fig. 7, and is defined by the following equation for the CA on a rough surface: cosy ¼ rcosy

where r is the roughness parameter and y is the apparent CA. Combine this formula with Young’s equation: r




gSV

gSL gLV



¼ cosy

ð26Þ ð27Þ

The apparent CA (y) corresponds to the stable equilibrium. The roughness ratio, r is a measure of how surface roughness affects a homogeneous surface. y is the Young CA for an ideal surface. It was important to mention that Wenzel's equation illustrates the CA of a rough surface is varies from the intrinsic CA, but it does not describe contact angle hysteresis (CAH). 2.26.6.2.1.2 Cassie–Baxter model When dealing with a heterogeneous surface, the Wenzel model is not sufficient. A more complex model is needed to measure how the apparent contact angle changes when various materials are involved: cosy ¼ rf f cosyY þ f

ð28Þ

1

LG

C

SG

SL

Fig. 6 Digram of surface energy on water contact angle. Reproduced from Whyman G, Bormashenko E, Stein T. The rigorous derivation of Young, Cassie–Baxter and Wenzel equations and the analysis of the contact angle hysteresis phenomenon. Chem Phys Lett 2008;450(4):355–9.

Fig. 7 Wenzel’s model. Reproduced from Whyman G, Bormashenko E, Stein T. The rigorous derivation of Young, Cassie–Baxter and Wenzel equations and the analysis of the contact angle hysteresis phenomenon. Chem Phys Lett 2008;450(4):355–9.

Dust Repellent Materials

843

where the rf is the roughness ratio of the wet surface area and f is the contact fraction of solid surface area wet by the liquid. A summation of all forces around the three-phase contact line results in a pinning force. Cassie–Baxter model shown in Fig. 8 can also be recast in the following equation: gcosy ¼

2.26.7

N X  fi gi;sv

gi;sl

n¼1



ð29Þ

Dynamics of Dust Removal and Surface Cleaning

2.26.7.1

External Dust-Removal Forces

Particle removal mechanism required an external force. This external forces mechanism is generated from different sources such as substrate vibration and fluid flow (hydrodynamic forces) around small particles [37].

2.26.7.1.1

Gravitational force

Gravitational force is the force that acting on the particle–surface adhesion and it can be written as: Fg ¼ mg ¼

4 3 pR rg 3

ð30Þ

where g is the gravitational force, r is the dust particle density, and m is the mass of the particle.

2.26.7.1.2

Shear stress and inertial force (lifting force)

In general, inertial force causes the lift force. The inertial-lift forces can create a torque that can act as the center of rotation at the contact point between the particle and wall surface. The flow shear stress creates additional lift force that causes torque force on the particle as shown in Fig. 9. Based on Navier–Stoke equation, Cherukat et al. [46] derived the lift force formula for spherical shape particle for Reynolds number below unity and the particle touch the surface. After that Zhang et al. [47] has simplified the model to

Fig. 8 Cassie–Baxter model. Reproduced from Whyman G, Bormashenko E, Stein T. The rigorous derivation of Young, Cassie–Baxter and Wenzel equations and the analysis of the contact angle hysteresis phenomenon. Chem Phys Lett 2008;450(4):355–9.

FL Mb dp

FD

Fg Ff Surface O Fad



Fig. 9 The force for particle–surface interaction. Reproduced from Kohli R, Mittal KL. Developments in surface contamination and cleaning, Volume 4: detection, characterization, and analysis of contaminants. Norwich, NY: William Andrew; 2011.

844

Dust Repellent Materials

be written as: FL ¼

9:22m2 ðRe Þ3 r

ð31Þ

where Ru v

ð32Þ

sffiffiffiffiffi 2t r

ð33Þ

Re ¼ and u ¼

Here Re is the shear Reynold number, m is the dynamic viscosity, v is the kinetic viscosity, u is the friction velocity, and t is the shear stress.

2.26.7.1.3

Drag forces (pressure and shear stress)

Drag force can be divided into pressure force and shear force. Firstly, when the area normal to the flow direction is wide it is called pure pressure drag force. The force caused by friction force to parallel area of the particle is called pure friction drag force. Another case is the combination between pressure and friction forces. Goldman et al. [48] derived the drag force formula from Navier–Stoke equation as: FD ¼ 10:2pmRu

ð34Þ

while the hydrodynamic drag force according is given as: FD ¼

2.26.7.1.4

prfD CD V 2 d2P Cc

ð35Þ

Centrifugal force

The dust adhesion on rotating disk system is highly affected by centrifugal force that provide the force required to move the particle from the center to the edge of rotation. The centrifugal force is function of disk radius (r), mass (m), and rotational velocity (w). F ¼ mw2 r ¼

2.26.7.1.5

4 3 2 pR rw r 3

ð36Þ

Frictional force

Friction force is the force acting in reverse direction of particle movement. Friction force is function of normal force (N) and coefficient of friction (m), i.e.: Ff ¼ mN ¼ mFg

2.26.7.2

ð37Þ

The Mechanism of Particle Removal

The particle can be removed from a substrate when the net external force overcomes the magnitude of adhesion forces. There are three removal mechanisms of particle from the substrate: sliding, rolling, and lifting. The force balance of spherical particle is shown in the Fig. 9. The moment and force balance for each removal mechanism is listed in Table 2. The particle is separated by rolling mechanism if the moment of external forces overcomes the adhesion resistance moment around the point O as shown in the Fig. 9. It can be assumed that after rolling starts, the particle removal mechanism is due to the turbulence fluctuations and lift force. Soltani and Ahmadi [49] studied the different removal mechanisms of spherical particles from solid surfaces. The results revealed that the spherical solid particles are easily removed by rolling than sliding or lifting mechanisms. This result can be attributed to the low flow velocity required to roll the spherical particles compared with the sliding and lifting flow velocity. Table 2

Removal mechanisms of solid particle

Removal mechanism Sliding Rolling Lifting

Moment and force balance   FD  Ks F ad þ

Fg   Mhyd ¼ FD d2P þ FL a þ Mh  Fad þ Fg a ¼ Mad FL  Fad þ Fg

Dust Repellent Materials

2.26.8

845

Surface Characteristics (Hydrophobic and Super-Hydrophobic Coating)

2.26.8.1

Mimicking Nature

Nature is the inspiration source for researchers to develop functional systems contained self-cleaning characteristics. Ward et al. [50] observed that louts leaf raised from muddy water remained clean and untouched by pollutants. The secret behind this phenomenon was demonstrated in 1960s using SEM technique. The surface has different microscopic roughness scale with epicuticular wax crystalloids which lead to super-hydrophobic leaves [51]. Guo et al. [52] reported that super-hydrophobic leaves could be divided into hierarchical micro/nano structure and unitary microstructure. This revelation improves the development toward mimicking the natural super-hydrophobic behavior.

2.26.8.2

Hierarchical Structure in Plant Leaves

The textured surface providing the plant leafs with super-hydrophobic property. Fig. 10 shows the hierarchical structures for different plant leaves surfaces. For example, the SEM for lotus leaf shown in Fig. 10(A,B) has 162 degrees WCA with uniform texture with 10 mm flanged. Nanorod like structure was observed in Fig. 2(B) with 50 nm diameter which gives rise to behave as super-hydrophobic materials [53]. The plant surface might possess papillae structure which increases the WCA. The SEM image shown in Fig. 10(C) represents the surface of rice leaf with 5–8 mm papillae diameter. The surface hydro-hydrophobicity will enhance also due to presence of air trapping mechanism. This mechanism generated in surfaces sublayers having innumerable nano-pins. Fig. 10(E,F) shows SEM images for rice and lotus leaves with innumerable nano-pins (159 degrees WCA) [54]. Jiet et al. [55] reproduced a super-hydrophobic surface using casting technique to replicate a template of lotus leaf structure. The observed surface was textured by intricate nanostructures to obtain the space between micro-structured protrusions and valleys with a papillae diameter of 6 mm as shown in Fig. 11. However, obtaining lotus leaf template using artificial surface is still remains challenging for research.

2.26.8.3

Plant Leaves With Unitary Structure

The uniformly slick fibers distribution will form unitary structure which differ from hierarchical structures. Fig. 12(A) shows a Ramee leaf face exhibiting unitary structure with fiber diameter of 2 mm. Surprisingly, unitary structure was obtained also in Chinese watermelon surface as shown in Fig. 12(C,D) both plant leaves with unitary structure exhibit a 159 degrees WCA and considered as super-hydrophobic [56]. Ding et al. [57] fabricated unitary structure like surface using electrostatic layer deposition. The unitary structure was made by depositing cellulose acetate fibers that revealing a smooth super-hydrophobic surface with 140 degrees WCA.

2.26.9

Surface Texturing (Materials, Mechanism, and Processes) to Produce Hydrophobic and SuperHydrophobic Surfaces

Researchers focused on developing methods in order to fabricate materials exhibiting very low surface energy as well as controlling its morphology in micro/nano sacle. In general, methods of fabricating super-hydrophobic substrates can be divided into two parts: (1) producing a rough substrate using material exhibiting low surface energy and (2) lowering surface energy of rough surfaces [18].

2.26.9.1 2.26.9.1.1

Roughening of Low Surface Energy Material Silicones polydimethylsiloxane

Modified silica nanoparticles and polydimethylsiloxane (PDMS) are organosilicon compounds used to produce super-hydrophobic surfaces due to its hydrophobic and intrinsic deformability properties. Different methods were introduced in the literature in order to fabricate super-hydrophobic materials using PDMS. Khorasani et al. [58] used CO2 pulsed laser to initiate peroxide groups on PDMS. This peroxide groups introduced 2-hydroxyethyl methacrylate (HEMA) graft polymerization on PDMS surface. The observed WCA was 175 degrees due to the chain ordering and porosity on the PDMS surface. Another method was reported by Ma et al. [59] to fabricate super-hydrophobic membrane using electrospinning technique. PDMS block together with polystyrene homo-polymer was used to make the electrospun fibers. It was reported that the surface roughness with PDMS component, resulted in WCA of 163 degrees. Ke et al. [60] fabricated a super-hydrophobic surface form PDMS and silica nanoparticles using drop coating method. The drop coating technique process that used in the experiment is demonstrated in the Fig. 13. The octadecyltrichlorosilane (OTS) was developed onto silica nanoparticle surface. The drop coating procedure can be summarized in four steps. Firstly, PDMS elastomer was prepared with 1:10 weight ratio between curing agent to PDMS. Secondly, a 500 nm of ethanol – modified silica nanoparticle solution (M-SO2) dropped onto glass surface. After drying of the solution, a 100 nm of M-SO2 was added to the coated glass substrate. Thirdly, the dried M-SO2 solution was cured for 24 h at

846

Dust Repellent Materials

(A)

(B)

(C)

(D)

(E)

(F)

Fig. 10 Scanning electron microscopy (SEM) images of natural super-hydrophobic surfaces with hierarchical structures, where (A,B) lotus leaf structure, (C,D) rice leaf structure, and (E,F) taro leaf structure. Reproduced from Otten A, Herminghaus S. How plants keep dry: a physicist's point of view. Langmuir 2004;20(6):2405–8.

1001C. Finally, PDMS film was dropped onto the substrate before final curing. The results revealed that the super-hydrophobic surface produced using this technique has 15572 degrees CA. Another method used to texture the surface is called laser etching technique. Jin et al. [61] produced a PDMS super-hydrophobic surface (hierarchical–structured) using casting and laser etching. The submicro roughness was made by laser etching while the micropillar was made by PDMS solution casting. The produced surface texture had a micro-, submicro-, nano-structure that generated from submicro-/nano groves and micropillars. Fig. 14 shows the hierarchical–structured surface using both casting and

Dust Repellent Materials

(A)

847

(B)

Fig. 11 Biomimetic super-hydrophobic surfaces replicated from lotus leaf surface structure template. Reproduced from Sun M, Luo C, Xu L, et al. Artificial lotus leaf by nanocasting. Langmuir 2005;21(19):8978–81.

(A)

(B)

(C)

(D)

Fig. 12 Scanning electron microscopy (SEM) images of natural super-hydrophobic surfaces with a unitary structure: (A,B) Ramee leaf with WCA of 164; (C,D) Chinese watermelon surface with WCA of 159 degrees. Reproduced from Ma M, Gupta M, Li Z, et al. Decorated electrospun fibers exhibiting superhydrophobicity. Adv Mater 2007;19(2):255–9.

848

Dust Repellent Materials

500 nm M-SiO2 In ethanol Drop coating Dried

PDMS layer Dried Cured Drop coating

Drop coating

PDMS in hexane

100 nm M-SiO2 In ethanol

Fig. 13 The drop coating method of preparing super-hydrophobic polydimethylsiloxane (PDMS) based surface. Reproduced from Ke Q, Fu W, Jin H, Zhang L, Tang T, Zhang J. Fabrication of mechanically robust super-hydrophobic surfaces based on silica micro-nanoparticles and polydimethylsiloxane, Surf Coat Technol 2011;205:4910–4.

laser etching technique. The developed hierarchical structure using casting and laser etching technique has static and dynamic WCA from this surface were 162 and o5 degrees, respectively.

2.26.9.1.2

Organic and inorganic materials

Recently, paraffinic hydrocarbons (organic materials) have been used to produce super-hydrophobic instead of fluorocarbons and silicones compounds. Lu et al. [62] used low-density polyethylene method to form super-hydrophobic coating. Super-hydrophobic surface with high porosity was produced by controlling the nucleation rate and crystallization time. The obtained WCA using this method was 173 degrees. On the other hand, inorganic compounds such as TiO2 and ZnO were used to produce reversibly switchable wettability films. Feng et al. [63] synthesized nanorods of ZnO with very low surface energy exhibiting super-hydrophobic properties. It was observed that, ZnO super-hydrophobicity changed to super-hydrophilicity when the surface exposed to UV radiation. The reported explanation for this phenomenon was due to the production of electron hole pair which adsorbs the hydroxyl group from the film surface.

2.26.9.1.3

Fluorocarbons

Fluorocarbons are polymers that have very low surface energy. Super-hydrophobic properties can be obtained from these polymers by roughening techniques. Shiu et al. [64] used oxygen plasma to treat Teflon in order to produce rough surface. It was revealed that the WCA was 168 degrees.

2.26.10 2.26.10.1

Producing and Modifying a Rough Surface Using Material of Low Surface Energy Hydrothermal Reaction and Wet Chemical Reaction

Chemical and hydrothermal reactions are direct techniques used to control morphology and dimensionality of nanostructured materials. Jiang et al. [65] produced a super-hydrophobic layer using chemical composition technique on copper substrate. It was observed that tetradecanoic acid solution modified the copper surface and result in super-hydrophobic layer. Recently, the hydrothermal technique was developed to fabricate functional materials with variable morphologies [66]. Hou et al. [67] oxidized Zn metal combined with n-octadecyl thiol to produce ZnO nanorod with super-hydrophobic characteristics. The array of ZnO nanorod was synthesized using catalysts, short-wavelength lasing, piezoelectric materials, and gas sensors. Fig. 15 shows the biomimetic super-hydrophobic surfaces fabricated using chemical and hydrothermal reaction.

2.26.10.2

Electrochemical Deposition

This technique is a widely used to produce micro/nanoscale super-hydrophobic surface structure [68]. Bell et al. [69] formed a super-hydrophobic surface using galvanic deposition technique. The formed surface can easily float on water and the measured CA was 173 degrees. Jiang et al. [70] produced copper mesh with micro/nanoscale hierarchical structured.

Dust Repellent Materials

Convex width

849

Microconvex

(A)

(B)

(C)

(D)

Fig. 14 Scanning electron microscopy (SEM) micrograph for: (A) laser etched PDMS surface, (B) and (C) surface with higher magnification, and (D) comparison between flat PDMS and laser etched PDMS. Reproduced from Jin M, Feng X, Xi J, et al. Super-hydrophobic PDMS surface ultralow adhesive force. Macromol Rapid Commun 2005;26:1805–9.

2.26.10.3

Layer-by-Layer and Self-Assembly

Layer-by-layer (LBL) technique is demonstrated as a substrate adsorption in compounds with opposite charge to form multilayer films. Zhai et al. [71] developed a super-hydrophobic film using LBL method based on lotus leaves template. A polyelectrolyte multilayer surface was achieved by developing a honeycomb-like structure. It was reported that, the multilayer substrate has superhydrophobic properties due to the treatment with semi-fluorinated silane.

2.26.10.4

Lithography

This method is used to develop micro/nano patterns. The lithographic techniques involves X-ray lithography, photolithography, nanosphere lithography, soft Lithography, and electron beam lithography [72,73]. For example, Notsu et al. [74] fabricated superhydrophilic/hydrophobic patterns on gold surface using photocatalytic lithography. The surface exhibiting super-hydrophobic characteristics with 164 degrees WCA.

2.26.10.5

Etching and Chemical Vapor Deposition

Mainly, this technique is used for polymers to develop different morphologies. Qian and Shen [75] fabricated super-hydrophobic by roughening polycrystalline metals using chemical etching method. This technique was based on dissolving of etchant in grain dislocations. It was reported that treatment of the substrate with fluoroalkylsilane resulted in super-hydrophobic film (153 degrees WCA). On the other hand, chemical vapor deposition is also used to develop micro/nanostructured surfaces. For example, Ci et al.

Dust Repellent Materials

850

(A)

(B)

(C)

(D)

Fig. 15 Biomimetic super-hydrophobic surfaces fabricated by (A,B) wet chemical reaction and (C,D) hydrothermal reactions. Reproduced from Meng Q, Hu J. A review of shape memory polymer composites and blends. Compos Part A Appl Sci Manuf 2009;40(11):1661–72.

[76] fabricated a super-hydrophobic surface with 170 degrees WCA. This substrate was constructed based on aligned double walled carbon nanotube.

2.26.10.6

Electrospinning Technique

It is the dominant method in fine nanofibers synthetization. The researchers use Electrospinning technique to modify surface roughness in order to achieve super-hydrophobic features. Ma et al. [10] produced super-hydrophobic surface using electrospinning with chemical vapor deposition. The process was carried out through two steps: (1) electrospun of poly-caprolactone and (2) coating with polymerized perfluoroalkyl ethyl methacrylate. It was observed that, a super-hydrophobicity was introduced to the surface with 175 degrees WCA (see Fig. 16).

2.26.10.7

Sol–Gel Method and Polymerization Reaction

The Sol–gel technique is a general method being used in all solid substrates [77]. For instance, Huang et al. [78] used ethylene glycol and hexamethylenetetramine to produce super-hydrophobic surface. It was reported that the super-hydrophobicity of the film was controlled by adjusting the film pore size. Furthermore, Rao et al. [79] produced a super-hydrophobic silica aerogel using sol–gel method on MTMS precursor. The silica aerogel solution was made from MTMS, ethanol and water with different molar ratio. The catalyst (ammonium hydroxide) was used to speed up the gelation time. Fig. 17 shows the SEM micrographs of the developed super-hydrophobic surface. The result revealed that the highest CA was 173 degrees. Polymerization is successful technique in fabricating super-hydrophobic materials. Choi et al. [80] designed a surface exhibiting super-hydrophobic properties

Dust Repellent Materials

851

(B)

(A)

Fig. 16 Electrospun nanofiber surface with 175 degrees WCA: (A) before coating and (B) after coating. Reproduced from Ma M, Mao Y, Gupta M, Gleason KK, Rutledge GC. Super-hydrophobic fabrics produced by electrospinning and chemical vapor deposition. Macromolecules 2005;38 (23):9742–8.

Acc.V Spot Magn Det WD Exp 3.00 kV 3.0 5000x TLD 1.8 1

5 μm

Acc.V Spot Magn Det WD Exp 3.00 kV 3.0 5000x TLD 2.2 1

5 μm

(A)

(B) Fig. 17 Scanning electron microscopy (SEM) micrographs of (A) hexamethyldisilazane (HMDS) modified and (B) the trimethylsilyl chloride (TMCS) modified silica super-hydrophobic films. Reproduced from Rao AV, Latthe SS, Nadargi DY, Hirashima H, Ganesan V. Preparation of MTMS based transparent super-hydrophobic silica films by sol–gel method. J Colloid Interface Sci 2009;332:484–90.

Dust Repellent Materials

852

Materials

Process

Silicones (PDMS)

Etching/casting

Fluorocarbons (Teflon)

Stretching/casting

Organic materials (PS+DMF)

Electrospinning

Inorganic materials (ZnO)

Two step method

Polycrystalline metals (Al/Cu)

Etching

Composites (Au) Polyelectrolyte + Silica nano particles Alkyoxysilane

Roughening the surface of low surface energy material

Hydrophobic/ superhydrophobic coatings Lithography LBL and colloidal assembly

Making rough surface and modifying the surface with material of low surface energy

Self-cleaning coatings

Sol-gel processing

Titania

Titania + dopants Zinc oxide

CVD/PVD/ Sol-gel/ spraying

Hydrophilic photocatalytic coating

Tungsten oxide

Fig. 18 Summary of various materials and fabrication procedures. Reproduced from Ganesh VA, Raut HK, Nair AS, Ramakrishna S. A review on self-cleaning coatings. J Mater Chem 2011;21(41):16304–22.

using polymerization technique. They used an oxide base surface coated with random copolymer. The measured WCA for the substrate was 163 degrees.

2.26.10.8

Other Techniques

Many methods were reported in the literature in the past few years to produce rough surfaces. For example: electro-spraying, sandblasting, and texturing are techniques being used to develop self-cleaning surfaces. Fig. 18 shows the various methods and fabrication techniques for self-cleaning surfaces.

2.26.10.9

Impact Dynamics of Water Droplets on Super-Hydrophobic Surfaces

Wang et al. analyzed the water droplet rebound and impact dynamics on super-hydrophobic surfaces. The study was conducted using carbon nanotube (CNT) arrays grown on Si substrates (using thermaly assisted chemical vapor deposition) having differet static CAs of 140 and 163 degrees. It was found that for 163 degrees static CA, the water droplets were bounced off many times (Fig. 19(A)) [81]. On the other hand, for reduced static CA of 140 degrees, the water droplet remaind pinned without rebouncing (Fig. 19(B)). This behavior of droplet stucking was attributed to the droplet momentum at first impact which enabled it pentrating and displacing the super-hydrophobic air pockets under static condition. Consequently, the contact area between the water droplet and CNT array increased significantly which increase the adhesion force (Van der Waals) of the water droplet to the substrate. The outcome of this study shows the challenges to maintain the nanostructured super-hydrophbic surfaces under impact conditions [81].

2.26.11

Dust Characterization

Dust particles are characterized to better understand the cause of the effect of dust deposition on PV module performance and to come up with effective mitigation techniques. Dust particles characterization methods include particle sizing, morphology analysis, and chemical composition analysis.

Dust Repellent Materials

0s

9 ms

2 ms

5 ms

12 ms

14 ms

16 ms

1 mm

MWNT array

3.7 ms

0s

(A)

853

5.7 ms

31.3 ms

21.3 ms

1 mm

MWNT array (B)

Fig. 19 Water droplet snapshots impacting carbon nanotube arry surface with: (A) static contact angle of 163 degrees and (B) static contact angle of 140 degrees. Reproduced from Wang Z, Lopez C, Hirsa A, Koratkar N. Impact dynamics and rebound of water droplets on superhydrophobic carbon nanotube arrays. Appl Phys Lett 2007;91:23105.

25

% Number

% Area

% Volume

20

%

15 10 5 0 0.1

1

10 Particle size (μm)

100

1000

Fig. 20 The fractions of the number, area, and volume of the particles distribution. Reproduced from Said SAM, Walwil HM. Fundamental studies on dust fouling effects on PV module performance. Sol Energy 2014;107:328–37.

2.26.11.1

Particle Sizing Analysis

The particle sizes and morphology have been determined using both optical techniques (including commercial particle analyzers), SEM, and recently, scanning probe microscopies. These important studies involved evaluating particle size distributions (Fig. 20) using various microscopy techniques. Table 3 shows the dust particle sizing and distribution morphology characteristics for dust collected from different cities in different countries. The literature indicates that the deposition of finer particles has a more significant effect on PV module performance than that of coarser particles. El-Shobokshy and Hussein [88] studied carbon, cement, and three types of limestone particles having median diameters of 5, 10, 50, 60, and 80 m. The dust particles were deposited on a PV surface at a controlled surface-mass density, and the power output was measured. Depositing equal densities (surface-mass density of 25 g/m2) of limestone particles of different sizes on a PV module surface showed that the finer particles caused a greater power reduction. Their results also showed that normalized power output in the case of cement and carbon particles dropped by 40% and 90%, respectively. This was attributed to the greater uniformity of distribution of the fine particles on the surface causing more scattering losses [88]. Moreover, high wind speeds remove coarser particles more effectively than fine ones [89]. On the other hand, gravity influenced significantly the dust deposition rate. The deposition rate due to gravity for small dust particles (Dpo5 mm) was 5% and increased to 75% for large dust particles (Dp45 mm) [90].

854

Dust Repellent Materials

Table 3

Observed dust particle sizing and distribution morphology characteristics

Study

Area of dust collection

Particle sizing (mm)

Additional

Qasem et al. [82]

Kuwait

4–8

Said and Walwil [83]

Dhahran, KSA

0.5–176

Bouaouadja et al. [84]

Algeria

95–780

Mastekbayeva and Kumar [85] Kazem and Chaichan [86]

Bangkok, Thailand Oman

53–75 2–63

Wang and Gong [87]

Qatar

Average 2 mm

The major grain size was silt. The small silt grains were of slate, whereas the bigger grains were quartz Different and irregular shapes, but generally, tend to be spherical shape The shape of the grains is irregular but approximately spherical Soft Bangkok clay used to prepare the artificial dust Dust deposition on PV was found to vary from one location to another Some non-uniform particles of few tens of micron

Table 4

The dust chemical composition collected from different locations

Study

Area of dust collection

Dust composition

Qasem et al. [82,92] Said and Walwil [83] Elminir et al. [29] Adinoyi and Said [93]

Kuwait Dhahran, KSA Helwan, Cairo, Egypt Dhahran, KSA

Clarke et al. [94]

24 sites across western Libya

The dominating component of dust was quartz followed by calcite and albite 60% of dust particles were calcite and quartz Quartz and calcite predominated with smaller amounts of dolomite and clay minerals The dust was composed of oxygen (58%), calcium (13%), carbon (10%) and sulfur (6%) Quartz with lesser amounts of calcite, illite, and halite

2.26.11.2

Particle Chemical Analysis

The nature of the dust particles, including chemical composition and color, plays a major role in the degree of reduction in glass cover transmittance and hence PV performance. Kaldellis et al. [91] have deposited various specific red soils, limestone’s, and ash’s deposition densities on the surface of PV-panels implying the deterioration of their performance when dust particles are deposited on their surface. The mean power reduction between the clean and the polluted PV pair, vary from approximately 3 to 5 W for red soil particles’ accumulation ranging from 0.12 to 0.35 g/m2, 4 to 7 W for red soil particles’ accumulation ranging from 0.28 to 1.51 g/m2, and 1 to 8 W for red soil particles’ accumulation ranging from 0.63 to 3.71 g/m2, respectively. These results indicated that red soil deposition on PVs’ surfaces causes the most considerable impact on PVs’ performance and thus the highest reduction in generated energy, followed by the limestone and secondly by the carbon-based ash. An amount of 0.35 g/m2 of red soil deposition on PV-panels’ surfaces may reduce the generated energy by almost 7.5% while approximately the same deposition density for limestone (0.33 g/m2) causes almost 4% energy reduction. However, in another study conducted by [88] 28 g/m2 of carbon deposit caused a greater reduction in PV module power output than 73 g/m2 of cement or 250 g/m2 of limestone. They attributed the reason for effective solar radiation absorption by carbon and hence adversely affects PV performance. Elminir et al. [29] conducted extensive mineralogical analysis using XRD to identify the chemical composition of the deposited layers of dust particles on PV module in Egypt. The dust particles were mostly composed of quartz and calcite, with smaller amounts of dolomite and clay minerals. Table 4 shows the observed dust chemical composition collected from different locations. Fig. 21 presents the results of the XRF analysis. The major constituents were silicon from desert sand (quartz, or silicon dioxide, SiO2) calcium from the mineral calcite (calcium carbonate, CaCO3). The minor constituents consisted of aluminum, iron, magnesium, potassium, and sodium. Similarly, Said and Walwil [83] reported that oxygen had the highest chemical concentration followed by calcium, silicon, sulfur, and iron as shown in Fig. 22. Also, it is found that quartz and calcite compounds occupied more than 60% of dust particle content (Fig. 23). Figs. 24 and 25 shows some SEM micrographs of dust particles. The Figures indicate that dust particles are composed of different shapes and sizes. Large dust particles attach to the small dust particles, because of the electrostatic charges of small dust particles [95]. Small dust particles exposed to the solar irradiation for long durations and attaching of ionic compounds to the dust particles caused static charging of the particles [96].

2.26.11.3

Dust Particle Chemistry

Dust particles absorb water vapor in humid air environments and form mud at the surfaces. Once the mud becomes dried at high temperature conditions under the solar radiation, it becomes difficult to remove from the glass cover surfaces [97]. Water interaction with dust particles took place due to effective adsorption of water molecules by dust particles surfaces [98]. The interaction between water molecules, dissolved ions, and soil particles occurred because of the unbalanced force field which

Dust Repellent Materials

20

Dec. Jan. Feb. Mar. Apr. May. Jun.

18 Relative concentration (%)

855

16 14 12 10 8 6 4 2 0 0

Na

Ca

Zn

Si

S Cl Al K Mineralogical composition

Ti

Fe

Ce

Mg

Fig. 21 Organic components of the polluting material using XRF analysis. Reproduced from Elminir HK, Ghitas AE, Hamid R, El-Hussainy F, Beheary M, Abdel-Moneim KM. Effect of dust on the transparent cover of solar collectors. Energy Conv Manag 2006;47(18):3192–203.

60 EDS

XRF

50

% Weight

40 30 20 10 0 Ca

Si

Fe

Mg

Al

K

Ti

Zn

Sr

S

Cl

C

O

Na

Chemical element Fig. 22 X-ray fluorescence (XRF) and energy dispersive spectroscopy (EDS) chemical element analysis. Reproduced from Said SAM, Walwil HM. Fundamental studies on dust fouling effects on PV module performance. Sol Energy 2014;107:328–37.

40 35

% Weight

30 25 20 15

34 26

32

30

10 11

5 0 Calcite, magnesium, syn

Quartz

Dolomite Chemical compound

Gypsum

Calcite

Fig. 23 X-ray diffraction (XRD) qualitative and quantitative analysis. Reproduced from Said SAM, Walwil HM. Fundamental studies on dust fouling effects on PV module performance. Sol Energy 2014;107:328–37.

856

Dust Repellent Materials

Fig. 24 Scanning electron microscopy (SEM) micrograph of dust. Reproduced from Adinoyi MJ, Said SAM. Effect of dust accumulation on the power outputs of solar photovoltaic modules. Renew Energy 2013;60:633–6.

Fig. 25 Scanning electron microscopy (SEM) micrograph small dust particles adhering to the surface of large dust particles. Reproduced from Hassan G, Yilbas BS, Said SAM, Matin A. Chemo-mechanical characteristics of mud formed from environmental dust particles in humid ambient air. Sci Rep 2016;6.

depended on the particle size. It should be noted that the formed mud solution had chemically active characteristics [5,99]. The dissolution of dust particles, such as calcite (CaCO3) and halite (NaCl), are expressed by Eqs. (38)–(40) shown below. In the first example, halite reactions take place with molecules of polarized water which separate chloride and sodium into single ions which become surrounded by hydration sheaths and dissolved in the water molecules. The carbonation process takes place in calcite (CaCO3) dissolution, where the carbon dioxide attracted and dissolved in water producing carbonic acid [100]. NaCl þ H2 O-Naþ1 þ Cl

1

þ H2 O

ð38Þ

H2 O þ CO2 -H2 CO3

þ

H2 CO3 þ CaCO3 -Caþ2 þ 2ðHCO3 Þ þ

þ2

ð39Þ 1

ð40Þ

During drying, the dissolved ions (Na , K , Ca , Cl , SiO ) attract mud structure due to electrostatic and ionic bonding force. These ions become dissolved in the mud solution particles holding them together and forming crystals in between. The ions in the mud structure during water evaporation increasing the adhesion force. Yilbas et al. [101,102] demonstrated that dust particles consist of neutral and ionic compounds. The alkaline and alkaline earth metallic compounds of dust particles dissolve in the water condensate on surfaces in humid environments, which gives rise to the formation of a chemically active mud solution that flows around dust particles under the effect of gravity and reaches the solid surface where the dust particles have settled. This chemically active mud solution layer has a major effect on increasing the adhesion force between the formed mud and surface.

Dust Repellent Materials

Table 5

857

Effect of cover surface on photovoltaic performance

Study and location

Surface

Testing

KSA [83]

Coated and uncoated

KSA [106]

Anti-reflecting coating

NREL [107]

Grooved, pyramid structured, lightly textured, and flat glasses

Belgium [23]

Three different coated glass samples

KSA [106] Málaga [108]

Texturing and anti-reflecting coating Three different types of glasses (textured with little pyramids follow an angle pattern and pyramids show an orthogonal pattern)

The transmittance reduces by 30% for coated glass and after 37% for uncoated glass after 40 days of exposure The power output enhanced by 8% using anti-reflective coatings on PV glass cover surface Module Voc was higher by 50, 80, and 10 mV for grooved, pyramid structured, and lightly textured glass module as compared to the reference module Transmittance decreased by (%): multilayer (ML) (0.85), self-cleaning (SC) (1.30), anti-reflection (AR) (1.75) and regular glass (2.63) Texturing and anti-reflecting coating reduced the dust accumulation by 5% Energy loss due to soiling for coated modules was 10% and 12% for uncoated modules during summer months

Qasem et al. [92] investigated the air pollution impact on PVs’ performance for several amounts of carbon-based ash deposited on PV panel surfaces. They compared the power output decrease of the artificially polluted panel with the clean one control for the different quantity of ash deposited on the PV panel surface. The results showed that for ash mass depositions of 0.63, 1.89, 3.15, and 3.78 g/m2; the panel power output dropped by 2.3%, 7.5%, 17%, and 27%, respectively.

2.26.11.4

Dust Particle Adhesion

Another issue is the impact of humidity in dust fouling. Vapor condensation on the PV module surface forms capillary bridges in gaps between the particles and the surface, generating large meniscus forces that enhance adhesion between the particle and surface, which encourages dust buildup [103–105]. Table 5 reports the effect of surface on the dust deposition amount. Generally, an increase in absolute humidity causes an increase in dust accumulation [109,110]. Mekhilef et al. [26] reported that adhesion force between dust particles and surfaces was highly influenced by the atmosphere humidity. As the relative humidity decreased, the efficiency of solar panel increased due to less dust particles adhered to the surface. In relation with particles size, Corn [111] studied the adhesion force between solid particles and demonstrated that their adhesion force increased significantly with the particle size. Furthermore, the contact area between a particle and rough surface was found to have a major role in the adhesion between the particles and surface. The adhesion of dust and contact potentials were investigated by Penney [112]. It was reported that the adhesive forces of electrostatically deposited dust were much greater than a similar dust deposited mechanically. Each particle in an electric field takes specific orientation by the dipole moment which is produced by the contact potential differences. The coulomb forces between the particle layers producing high adhesive forces by the dipoles orient electrostatically. In general, dust particle charge should be influenced by the adhesion force between the dust particles and PV glass cover and decrease the PV voltage output [96]. Mclean [113] presented the cohesion of dust layer and the cohesive force in electrostatic precipitators on the sediment layers of dust in particular. An electrostatic precipitator has a significant cohesive force that influences the sediment layers because of the electric field of particles in air influenced by the corona current across the layers. It was found that the electric field which flows through the layer had a linear proportional to the cohesive force approximately. Somasundaran et al. [114] reported using a cohesive force apparatus. It was improved to measure the various shapes, various sizes, and nature of the particles chemical structure under various conditions for very low cohesive force of around 1 nN. The cohesive force between glass surfaces and dust particles increases with the decrease of PH and an additional amount of salt resulting in a significant increase of the cohesive force. Also, the cohesion between dust particles and surfaces was reduced by the interaction of anionic surfactant with polyethylene oxide layer. Various models for adhesion force measurements are reported [45,115]. These models are the JRK model, Rabinovich model and Derjaguin, Muller, and Toporov (DMT) models, all of which are used to characterize the adhesion force for surfaces and adhering dust particles. For similar particles, the adhesion force varied due to different values of substrate roughness and the actual contact area which has an important role influencing particles adhesion. Kazmerski et al. [116] evaluated the basic interactions of dust particles adhesion to the PV module surfaces. Their main interest was to investigate morphology, and chemical mechanisms of dust particles soiling. The measured adhesion forces of dust particles with surface chemistry were correlated using Cuddihy principles. The results revealed that relatively high adhesive forces are due to dust particles chemical bonding to the glass surface. One of the chemical solutions was suggested by Brown et al. [117] in 2012. They applied an anti-soiling hydrophilic coating to the glass cover which reduced the dust soiling amount on the surface. However, it is indicated that beads with large size do experience large adhesion force due to increase in the contact area between the bead and the surface [83]. In relation with adhesion forces measurements, Kazmerski et al. [116,118] reported that the inter-particle adhesion force is higher than particle-module glass cover adhesion. On the other hand, Hassan et al. [95] measured the adhesion force of the dry mud formed from environmental dust

858

Dust Repellent Materials

Adhesion force (μN)

12.0 10.0 8.0 6.0 4.0 2.0 0.0 0

5

10

15

20

25

30

35

40

45

50

Bead size (μm) Fig. 26 Effect of particle size on adhesion force from different studies. Reproduced from Hassan G, Yilbas BS, Said SAM, Matin A. Chemomechanical characteristics of mud formed from environmental dust particles in humid ambient air. Sci Rep 2016;6; Mohandes BMA, Lamont LA. Application study of 500 W photovoltaic (PV) system in the UAE. Appl Sol Energy 2009;45:242–7; and Brown K, Narum T, Jing N, Soiling test methods and their use in predicting performance of photovoltaic modules in soiling environments. In: 38th Photovoltaic specialists conference (PVSC), IEEE; 2012. p. 1881–5.

+

Dust particle

1 2 3 (A)

VM (B)

Fig. 27 Electric curtain: (A) Three-phase electric curtain and (B) the electrical field distribution between the electrodes of the shield. Reproduced from Calle CI, Buhler CR, McFall JL, Snyder SJ. Particle removal by electrostatic and dielectrophoretic forces for dust control during lunar exploration missions. J Electrostat 2009;67:89–92 and Sims RA, Biris AS, Wilson JD, et al. Development of a transparent self-cleaning dust shield for solar panels. In: Proceedings of ESA-IEEE joint meeting on electrostatics; 2003. p. 814.

particles on the PV glass cover. It was reported that the adhesion force increased due to formation of dry mud solution film at the interface of the dry mud-glass surface. Fig. 26 shows variation of adhesion force due to dust particles size.

2.26.12

Electrostatic Dust-Removal Methods

The electrostatic approaches that used for mitigating the negative effects of dust on lunar solar panels were firstly proposed by NASA. The PV module surface can be attached with an electrodynamic screen which is made of transparent plastic sheets. This transparent plastic sheet is a UV-radiation resistant, such as, polyethylene terephthalate (PET). In addition, conducting electrodes that are made of transparent indium tin oxide sheets were attached to the PV surface in spiral or parallel configuration. These conducting electrodes are embedded beneath a thin transparent film. AC voltage supply (single phase or multiple phases) connect the electrodes and produce an electromagnetic field which repels the dust particles from the surface. Fig. 27 shows the electric curtain method proposed by NASA. Gaofa et al. [119] reviewed the method of dust removal from PV panel surfaces. These methods are natural means, self-cleaning, mechanical means electrostatic means. It was reported that electric curtain method is best strategy for dust removal. Mazumder et al. [120] reported that the dust removal efficiency from a PET surface using electrodynamics screen was 80% which lead to increase in solar cell performance of up to 90%. The electrostatic buildup surfaces (sticky surface) are more likely to accumulate dust than the smoother surfaces. Dust promotes dust which means initial dust quantity tends to promote or attract further dust settlement and the surface becomes amenable to more dust deposition. The efficiency of electrodynamics screen depends on several parameters including type of dust particle, dust deposition rate, the applied voltage, and method of operation. The type of accumulated dust particles is significantly affects the dust-clearing ability

Dust Repellent Materials

859

Arduino controller

Solar panel

Electrode +

Electrode −

Weight sensor (load cell)

Power supply unit

Battery

Power convertor

− +

Fig. 28 Arduino interface to electrostatic precipitator. Reproduced from Hudedmani MG, Joshi G, Umayal RM, Revankar A. A comparative study of dust cleaning methods for the solar PV panels. Adv J Grad Res 2017;1:24–9.

using such an electric curtain. In this regards, Sims et al. [121] developed a transparent electrodynamic shield to protect the PV panel surfaces from accumulated dust particles. They tested three different powders (lactose, acrylic polymer powder, and mars dust simulant). A Mars dust simulant was observed to be the easiest particles to be removed from the electrodynamic shields, with more than 90% being cleaned using high voltage, while relatively lower values were reported for lactose and acrylic polymer powder. The behavior observed from Mars dust simulant was due to the small size of the particles. Atten et al. [122] reported that the electric curtain is directly proportional particle size and it has to overcome the interaction van der Waal forces between the dust particles and the surface. However, there is possibility of intermittent operation of electrodynamic screens under normal operating conditions. A comparative study was conducted by Sharma et al. [123] to analyze the efficiency of electrodynamic screens under both intermittent and continuous operations with different dust deposition rates. The results revealed that the average dust-removal efficiency during intermittent operation approximately 90%, whereas it was when the screen was activated continuously was over 95%. Moreover, Sharma et al. [123] studied the power needed to operate the electrodynamic screen. It was observed that the power required to operate the electrodynamic screen is highly dependent on dust deposition. The average power consumption of the screen was approximately 10 W/m2 for a dust load of 0.6 mg/cm2, that is,5 W/m2 for removing the dust and 5 W/m2 for the supply operation. The power consumption varied between 1.0 and 2.9 mW when the when dust accumulation varied between 0.4 and 0.6 mg/cm2. It was recommended that, a low-power microcontroller system can be used instead of a digital signal controller system in order to decrease the power required for the supply operation [124]. The dust-clearing ability of a transparent electrodynamic shield is sensitive to changes in the pulse (wave shape) frequency and amplitude of the applied voltage. Low frequencies lead to better removal efficiency by increasing the velocity across the shield electrodynamic surface. Higher voltages improve also the dust removal efficiency, as do pulsed and triangular signals [125]. Recently, Hudedmani et al. [126] conducted a comparative study of various dust cleaning techniques (manual cleaning automatic wiper based cleaning, vacuum suction cleaning, and electrostatic precipitator). The results revealed that using of Arduino-controller electrostatic precipitator resulted in utilizing the solar energy effectively and enhancing the efficiency of the solar panel. Fig. 28 shows the configuration of Arduino-controller electrostatic precipitator used in the study. Most electrostatic experimental investigation of dust accumulation deposition has been conducted for artificial dust. Therefore, it is recommended that the electrostatic deposition behavior and the electrochemical properties of natural dust needs be investigated for different environmental conditions. Furthermore, the electrostatic attraction impact on dust settlement behavior can be affected by the electrostatic environments which influence their deposition degree [27].

2.26.13

Characterization and Analysis of Environmental Dust

As the dust particles accumulates in humid air ambient, mud forms because of the accumulation of the dust particles and the condensation of water vapor on the surface of the dust particles. The dust particles are composed of soluble particles, such as alkaline metals (e.g., Na, K), and non-soluble compounds such as silica and calcite (CaCO3). The soluble compounds alter the base of the solution and increase adhesion of mud onto the surface by forming covalent bonds between the mud and the solid surface. Although adhesion between the dry dust particles and the substrate surface is governed by van der Waals forces, the cohesive effect due to the crystallized solution at the interface increases the mud adhesion at the surface. Consequently, the mud residues that remain at the substrate surface modify the chemical and physical characteristics of the surface including its surface

860

Dust Repellent Materials

texture, optical and tribological properties, stress levels, and surface hydrophobicity. Such changes can consequently reduce the performance of the solid substrate in a given application. To address this issue, the dust accumulation at the surface can be minimized creating surface self-cleaning effects, this procedure fails in the humid air environments because of the excessive interfacial forces generated between the solid surface and the wetted dust particles.

2.26.13.1

Dust Particles Characterization

Characterization of atmospheric airborne dust during the wet seasons in East Africa was investigated by Mkoma et al. [127]. They demonstrated that common crystal and sea-salt elements, including Na, Mg, Al, Si, Cl, Ca, Ti, Mn, Fe, Sr, NO3 , and P (and to a lesser extent Cu and Zn) tended to be coarse particles. In addition, the aerosol chemical mass content of the aerosol was determined to consist of 48% organic matter, 44% crustal matter, 4% sea salt, and 2% elemental carbon was observed. A characterization of atmospheric aerosols was also carried out by Maenhaut et al. [128]. They showed that most of the Ca was water soluble; the mineral dust Ca was presumably mostly present as calcite, and perhaps also in part as gypsum. In contrast, only half of the K content was water soluble, indicating that it was to a large extent associated with insoluble mineral dust. Patterns of dust retention on urban trees in oasis cities were examined by Baidourela and Zhayimuj [129]. The findings revealed that, dust that had accumulated on tree leaves was mainly of local urban origin, and the heavy metal concentrations at different sites varied significantly. The morphology of atmospheric particles in a semi-arid region of India was studied by Mishra et al. [130]. They demonstrated that the influence of the dust aspect ratio on dust scattering was significant for dust with a high hematite content. Since the characteristics of the dust particles vary around the globe, consideration of revisiting the characterization becomes essential. The characterization process and methods adopted are given in line with the previous study [4]. The dust was collected from the solar energy laboratory of research institute at King Fahd University of Petroleum and Minerals (KFUPM), which is located close to the city of Dammam in Saudi Arabia. The dust accumulated on the surface of the protective glass of PV panels was collected every 2 weeks, which was repeated for over the period of 12 months. The dust particles collected were analyzed in terms of weight, size, shape, and elemental composition using the analytical tools. The findings revealed that the dust particles collected over 2 weeks periods within 12 months have similar characteristics in terms of elemental composition, size distribution, and shape. However, the amount of dust particles accumulated at the surface of the PV protective layer 5 g/m2 within 2 weeks; however, it varied within 15% (by weight) over 12 months. This was attributed to wind speed and its direction. Although the wind speed and direction changed over the time, the average wind speed remained about 4 m/s over a year. Dust accumulation and associated mud formation were simulated in laboratory environments that mimicked environmental humid air conditions, in which the air pressure was atmospheric, temperature was kept at 361C, and relative humidity was 80%. In the real environment, the condensation of water vapor onto dust particles triggers the formation of mud on the substrate surface. To simulate actual mud formation on the polycarbonate (PC) surface due to the condensation of water vapor on the accumulated dust particles, the following experiment was conducted. A 300-mm-thick layer of dust particles, which were collected from the local environment, was formed on the PC surfaces. Mud formation on the PC surfaces from the dust particles was then allowed to proceed in a humidity chamber, which simulated humid air conditions. Initial condensation tests were performed in local humid air for a period of 4 h to estimate the amount of condensate that accumulated prior to dust formation. This step resulted in the natural formation of mud on the PC surface in the chamber. The PC samples with the mud were kept in local ambient air for 3 days to dry. The adhesion work of the sample surfaces with the dry mud was measured. Then, the dry mud was removed from the surface using a desalinated pressurized water jet 2 mm diameter at a velocity of 2 m/s. The cleaning process was undertaken for 20 min for each sample surface tested. Fig. 29 shows SEM images of the dust particles accumulated on a PV protective glass test surface in Dammam, Saudi Arabia. The size of the dust particles varied from the nanometer range to 30 mm, exhibiting an average size of 1.2 mm. Some of the smaller dust particles in the submicrometer range were attached to the surfaces of larger dust particles (Fig. 29(A,B)). The bright areas typically observed for the smaller particles are indicative of the occurrence of electron charging in the SEM chamber during imaging; this showed that the small particles were charged, which created forces for attachment to the surfaces of the large particles. In general, the dust particles exhibited various shapes with round corners, sharp edges, or flake-like structures. P2 The geometric features of dust particles can be classified by the shape factor, RShape ¼ 4pA , where P is the perimeter of the dust pðLproj Þ2 particle, and the aspect ratio, AAspect ¼ 4A , where A is the cross-sectional area, and Lproj is the longest projection length of the dust particle. The aspect ratio is related to the approximate particle roundness and represents the ratio of the major-to-minor axes of an ellipsoid that is best fit to the particle. The shape factor is the inverse of the particle circularity, which is associated with the complexity of the particle (i.e., a shape factor of unity corresponds to a perfect circle). The particle diameter and area can be obtained from the measurements, where the diameter of a circle with an equivalent area is considered for circular dust particles, and an ellipse model is used when the longest projection is assumed to be the major axis. The particle cross-sectional area is preserved for non-circular dust particles. The relationship between the particle size and the aspect ratio or the shape factor is not simple. An inverse relationship is observed between the particle size and the aspect ratio, whereas a direct relationship is observed between the particle size and the shape factor. In this case, the particle aspect ratio reduces as the shape factor increases with increasing particle sizes. The cross-sectional area for a typical dust particle of size 1.8 mm is in the order of 2.5 mm2 and it gives rise to the shape factor of 1.05. However, the cross-sectional area of a typical dust particle of size 15 mm is in the order of 175 mm2 and the corresponding shape factor is about 3.18. The shape factor approached unity for the small particles (r2 mm), whereas for the large particles (Z10 mm), the median shape factor approached 3.

Dust Repellent Materials

(A)

861

(B)

Fig. 29 Scanning electron microscopy (SEM) images of the dust particles: (A) small particles attached to large particles and (B) various sizes of small particles.

Table 6 Si 12.3

Elemental composition of dust (wt%) determined by energy dispersive spectroscopy (EDS) Ca 8.2

Na 3.6

S 2.4

Mg 2.6

K 1.2

Fe 1.1

Cl 0.9

O Balance

Table 6 shows the elemental composition of the dust particles. In general, Si, Ca, Mg, Na, K, S, O, and Fe were the most common elements detected in the dust particles regardless of the size; however, the concentrations of Na, K, Ca, and O were found to increase, and chlorine was also present in the small particles (r2 mm). The changes in the elemental composition of the small particles (r2 mm) can be attributed to the prolonged time the particles spend in the atmosphere and the longer time periods of the particles interaction with solar radiation when compared with the larger particles. Therefore, the prolonged exposure of smallsized particles to the atmosphere allows for the attachment of ionic compounds in regions near the sea. The EDS data showed that the elemental contents of the particles were non-uniformly distributed regardless of the particle geometrical features, which may be attributed to the geological distribution of the local desert. The concentration of chlorine varied among the different dust particles, and the EDS data were not consistent with the molar ratio of NaCl shown in Table 6. This discrepancy suggests that the dust particles do not contain salt crystals, but rather dissolved chlorine in the compounds. Additionally, the high Si content measured in the flake-like particles indicates that Si was primarily present as silica particles. Fig. 30 shows the X-ray diffractogram of the dust particles and reveals the presence of K, Na, Ca, S, Cl, and Fe peaks. The iron peak coincided with the aluminum and silicon peaks, and the sodium and potassium peaks are likely associated with sea salt because the dust particles were collected from a region near the Arabian Gulf. The presence of sulfur may be related to the presence of calcium, such as anhydrite or gypsum components (CaSO4), in the dust particles. Iron is likely associated with the clayaggregated hematite (Fe2O3).

2.26.13.2

Analysis of the Mud Formed From the Dust Particles

Fig. 31 shows the SEM images of the top surface and cross-section of the dry mud formed on the PC surface, respectively. The dry mud surface was composed of closely packed dust particles, micro voids, and adhered large dust particles. Furthermore, the morphology of the dry mud surface exhibited an irregular topology with an average surface roughness on the order of 2.6 mm. Close examination of the SEM image revealed the presence of locally scattered white dense regions on the surface, which is likely related to the dried mud solution that is primarily composed of Na, K, and Ca (Table 2). The alkaline materials (e.g., Na and K) and the alkaline earth metallic compounds dissolved in the water, thus creating a mud solution. Some of the mud solution sediments at the mud–PC interface were obtained upon gravitational settling; however, a small amount remained at the surface where the dust particles were small, and adhered strongly and formed a white residue after drying. To determine the pH of the mud solution, the dust particles were mixed with ultra-clean desalinated water at a ratio of 1:4 and placed in an ultrasonic shaker for 15 min. The pH of the mud solution was recorded over a given time period. The pH increased with time, and the solution remained basic at pH ¼8.4, which is associated with the presence of OH ions in the solution. In this case, the dissolution of the alkaline and alkaline earth metallic compounds in the water is responsible for the formation of OH ions. The data obtained from the quadrupole inductively coupled plasma mass spectrometry analysis revealed that the mud

862

Dust Repellent Materials

200 Calcite (CaCO3) Quarz (SiO2) Gypsum (CaSO4.2H2O) Arhdride (CaSO4) Calcium sulfate hydrate (CaSO4.0.15H2O) Magnetite (Fe3O4) FeO KCl MgO NaCl Dolamid

Relative intensity (a.u)

160

120

80

40

0 20

30

40

50

60

70

80

2 Fig. 30 X-ray diffractogram of the dust particles.

5 kV

×430

50 μm 0000 13 39 SE I

(A)

15 kV

×400

50 μm 0000 13 39 SE I

(B)

Fig. 31 Scanning electron microscopy (SEM) images of the dry mud cross-section: (A) small dust particles forming a dense structure and dried solution trapped in the cavity as indicated by the circle and (B) large particles as indicated by the circle.

Table 7 Elemental composition of the dried mud solution on the polycarbonate surface assessed by energy dispersive spectroscopy. Each spectrum corresponds to a different location on the surface as indicated in the scanning electron microscopy image Ca 7.2

Na 0.4

S 2.2

Mg 1.7

K 0.2

Fe 0.4

Cl 0.1

O Balance

solution contained alkaline and alkaline earth metals. Consequently, the sediment mud solution contained alkaline and alkaline earth metals after drying, as evidenced in the EDS data (Table 7). The surface microhardness data were obtained at different locations on the dry mud surface. The microhardness increased in regions where the small dust particles were in close proximity. This result can be attributed to the strong cohesive forces among the small particles and the presence of the dry solution in this region, which acts as a binding agent among the particles. In contrast, the microhardness was lower in regions containing large particles, which is related to the weak bonding among the large particles due to the voids present among them. The undissolved oddly shaped, large dust particles were responsible for the nearby fine void formations in the dry mud; however, these voids appeared to be randomly scattered on the mud surface. The microhardness values of individual dust particles that were Z10 mm in size were also measured. The findings showed that the microhardness values varied significantly within the range of 6.7–26.3 HV. In the mud cross-section images (Fig. 31(A,B)), porous structures were found to cover a large area of the cross-section surface that were likely due to the naturally formed dust layer from the irregularly shaped

Dust Repellent Materials

15 kV

X8, 000

10 μm 0000 11 40 SE I

(A)

10 kV

X8, 000

2 μm

863

0000 21 40 SE I

(B)

Fig. 32 Mud residues and cavities formed at the PC surface after mud removal: (A) large and small mud residues – the large residues are indicated by the circle and (B) dried solution crystals.

dust particles. It should be noted that mechanical compaction was not applied to the dust layer on the PC surface prior to mud formation. The liquid solution flowed within the porous structures and hardened at the interface, forming a thin layer between the mud and the PC surface. However, some of the liquid solution remained in the cavities formed between the undissolved irregularly shaped dust particles in the mud. The film formed from the dried solution at the interface contained alkaline elements (e.g., Na, K), alkaline earth metals (e.g., Ca), and Fe, Mg, and Cl, as also shown in the EDS data in Table 2.

2.26.13.3

Analysis of Mud Residues

To assess the mud that remained on the surface, the PC surface interface with the dry mud was cleaned using a jet of water. A pressurized water jet that was 2 mm in diameter was sprayed at a velocity of 2 m/s and directed normally to the mud surface during the cleaning process for 20 min. After drying the surface naturally in an air, the PC surface was examined. Fig. 32(A,B) shows SEM images of the PC surface after mud removal. A few locally scattered mud residues were observed on the PC surface (Fig. 32 (A)). Additionally, crystal structures were found on the surface (Fig. 32(B)). The presence of mud residues after cleaning with pressurized water indicates the presence of strong adhesion between the dry mud and the PC surface. Owing to the mud solution sediment at the interface of the mud and PC, which contains alkaline hydroxides (e.g., NaOH and KOH) due to the high pH (8.4), nucleophile (OH ) attacks take place at the PC surface depending on the local concentration of NaOH and KOH in the mud solution. Thus, micromolecules with carbon bonded to three oxygen molecules of the PC (i.e., a carbonate link) at the surface are attacked by OH due to differences in electronegativity. An anion returns to the aqueous phase, and the reaction continues until complete depolymerization of the PC at the surface. Similar scenarios apply to KOH. The degradation of carbonate groups leads to a series of reactions due to the elimination of CO2 and CO in the absence of free radical reactions (i.e., first-order reaction). Consequently, local degradation creates small cavities at the surface, particularly in the region near the mud residues where the mud solution sediments locally at the surface. The dried mud solution at the interface of the dry mud and the PC surface increases adhesion between the mud residues and the PC surface despite the use of the high pressure water jet during surface cleaning. Furthermore, the presence of Cl in the dried mud solution at the dry mud–PC interface (Table 7) suggests that the simultaneous presence of bisphenate anions and chlorine could lead to chlorination of the aromatic rings with the formation of polychlorinated bisphenates. However, under basic conditions (e.g., pH ¼ 8.4), polychlorinated bisphenates are known to oxidize to form radicals, in which quinones form depending on the oxidative process that occurs [131]. This process contributes to the degradation of the PC surface, thereby enhancing the formation of small cavities on the PC surface.

2.26.14

Environmental Dust Effect on Solar Thermal Selective Surfaces

Environmental dust has significant effect on the solar thermal selective surfaces, which are widely used in solar volumetric receivers. Some of the case studies relevant to dust effect on the laser textured surfaces in relation to solar absorption applications are given in line with the previous studies.

2.26.14.1

Laser Texturing of Zirconia Surface and Environmental Dust Effect

In order generate texture and nitride compounds at the surface in relation to solar energy absorption, controlled laser melting/ablation method is accommodated. The experimental study for texturing of zirconia surface is presented in line with the previous study [132].

864

Table 8

Dust Repellent Materials

Laser processing parameters

Feed rate (m/s)

Power (W)

Frequency (Hz)

Nozzle gap (mm)

Nozzle diameter (mm)

Focus diameter (mm)

N2 pressure (kPa)

0.1

1800

1500

1.5

1.5

0.2

600

2.26.14.1.1

Laser texturing process and parameters

A CO2 laser (LC-ALPHAIII) delivering nominal output power of 2 kW was used to irradiate the workpiece surface. The nominal focal length of the focusing lens was 127 mm. The laser beam diameter focused at the workpiece surface was 0.2 mm. Nitrogen assisting gas emerging from the conical nozzle and coaxially with the laser beam was used. The laser melting parameters are given in Table 8. The zirconia tiles 25 mm  15 mm  3 mm were used in the experiments. JEOL JDX-3530 scanning electron microscope was used to obtain photomicrographs of the cross-section and surface of the workpieces after the tests. The Bruker D8 Advance having Cu–Ka radiation was used for XRD analysis. A typical setting of XRD was 40 kV and 30 mA. The wetting experiment was performed using Kyowa (model-DM 501) CA goniometer. A static sessile drop method was considered for the CA measurement. The WCA between the water droplet and the laser treated surface was measured with the fluid medium as de-ionized water. The droplet volume was controlled with an automatic dispensing system having a volume step resolution of 0.1 mL. Still images were captured, and the CA measurements were performed after 1 s of deposition of the water droplet on the surface. The experiments were repeated three times at different locations at the laser treated and as received surfaces. The error estimated is in the order of 7%. In order to determine the surface free energy of the laser treated surface CA measurements are extended to include Glycerol and Diiodomethane. The surface energy of solids and liquids can be divided into components. The surface energy determined is in order of 49.33 mJ/m2, which is slightly less than those reported in the previous study (52.6 mJ/m2) for ZrN [133]. An experiment was carried out to examine the effect of the actual mud formation, because of water vapor condensation onto the accumulated dust particles at the surface, on the characteristics of the laser textured zirconia surface. A dust layer of 300 mm thickness was formed from the dust particles collected from the local environment on workpiece surface. The dust layer of 300 mm thickness resembles the actual dust accumulation on surfaces in open environment over 1-week period. The desalinated water having same volume of dust layer was dispensed gradually on to the dust layer at the workpiece surface. It should be noted that initially some tests were carried out to measure the amount of water vapor absorbed by the dust particles due to condensation in the local humid environment, over 6-h period. The tests results revealed that the amount of condensate had almost same volume of dust over 6 h. In order to resemble the water condensation on the layer of dust particles in the humid air, dispensed water with the same amount of dust particles in the layer was left at the workpiece surface without mechanical mixing of the dust particles with water. This gave rise to natural formation of mud at the workpiece surfaces. Afterwards, the workpieces were located in a local normal air ambient for 3 days to dry. The laser textured surface with the presence of dry mud was tested for adhesion work and friction coefficient measurements. The linear micro-scratch tester (MCTX-S/N: 01–04300) was used for the tangential force and adhesion work measurements. During the experiments, the equipment was set at the contact load of 0.03 N and end load of 5 N. The total length for the scratch tests was 5 mm and the scanning speed was kept at 5 mm/min with the loading rate of 5 N/s. Once the adhesion and friction tests were completed, the dry mud layer was removed from the surface with a pressurized desalinated water jet having 2 mm diameter and 2 m/s velocity. The water jet assisted cleaning process was continued for 20 min for each workpiece surface. Finally, the analytical tools were used to assess the morphology and hydrophobic characteristics of the cleaned surfaces.

2.26.14.1.2

Results of laser texturing and dust adhesion on zirconium surface

The findings are discussed in the line of the previous study [132]. Fig. 33 shows SEM and AFM images of the laser treated zirconia surface. Due to the high repetition rate of laser pulses (1500 Hz of pulse repetition rate), the regular patterns are formed at the surface along the scanning direction of the laser irradiation (Fig. 33(A)). The close examination of the surface reveals that molten flow across these patterns is not observed. This indicates that excessive heating and melting at the surface did not occur during the laser texturing of the surface. In general, the irradiated surface is free from asperities such as large size open cavities and cracks. The crack free surface is associated with the self-annealing effect of the irradiated spots along the laser scanning tracks. In this case, the heat transfer from the lately formed irradiated spots toward the early formed spots alters the cooling rates below the surface. This creates a self-annealing effect minimizing the high thermal strain levels and thermal stresses in this region. The close examination of the surface shows that textured surface composes of micro/nano pillars and cavities. Although the surface melting and evaporation takes place at the irradiated spot, the evaporated region is limited to small area at the surface. This is because of the laser power intensity distribution at the surface, which follows the Gaussian distribution. Therefore, laser pulse intensity remains high at the irradiated spot center and reduces toward the edge of the irradiated spot. Consequently, evaporation takes place at the close region of the irradiated pot center due to the high intensity and melting occurs in the close region of the irradiated spot edges. Moreover, melt flow from the irradiated spot edges toward the cavity modifies the cavity size and depth at the irradiated spot center. This in turn results in micro/nano size pillars and shallow depth cavities at the surface (Fig. 33(B)). The average roughness of the laser textured surface is in the order of 0.85 mm. Some small cavities are formed at the surface and the cavity depth appears to be shallow, which is in the order of 60 nm. Elemental composition of as received and laser treated zirconia surfaces is given in Table 9. The elemental composition remains almost uniform at the surface; in which case, laser

Dust Repellent Materials

15 kV

×100

100 μm 0000 11 40 SE I

(A)

2 μm

×7,500

15 kV

865

0000 11 40 SE I

(B)

Fig. 33 SEM micrographs and AFM images of laser textured zirconia surface: (A) regular laser scanning tracks and (B) micro-size cavity (marked in circle).

Table 9

EDS data for elemental composition of laser treated workpiece surface (wt%)

Spectrum

Y

N

O

Zr

As received Laser textured

5.1 4.9

0 5

65.8 58

Balance Balance

14,000

t-(111)

6000

2000

ZrN

t-(113) t-(222)

t-(202)

As received surface

t-(200)

4000

t-(113) t-(222)

ZrN

8000

t-(202)

t-(200)

10,000 Relative intensity

Laser textured surface

t-(111)

12,000

0 20

30

40

50

60

70

80

2 Fig. 34 X-ray diffractogram of laser textured and as received surface.

treatment does not significantly change the elemental composition of the substrate material. The presence of nitrogen is associated with the nitride compound formation at the surface under the high temperature surface treatment process. The nitride compound such as ZrN, can be formed through two steps reactions including formation of the tetragonal structure of zirconia (t-ZrO2) and later initiation of zirconium nitride (ZrN) formation under the surface in high pressure nitrogen ambient. The formation of ZrN is also evident from Fig. 34, in which X-ray diffractogram of the laser treated and as received zirconia surfaces are shown. Tetragonal ZrO2 (t-ZrO2) peaks are dominant for the surface of as-received zirconia while ZrN peaks are present for the laser treated substrate surface. Since water condensates on the dust particles in humid air environments and the dust particles possess alkaline (Na, K) and alkaline earth (Ca) metallic compounds, the dissolution of these compounds in water alter the chemical state of condensate water. In order to assess this situation, water is mixed with dust particles by 1:10 ratio (10 being water) and the liquid solution from this

866

Dust Repellent Materials

15 kV

×800

20 μm 0000

23

40 SE I

Fig. 35 Scanning electron microscopy (SEM) micrographs of dried liquid solution and crystals with varying sizes are formed at the surface.

Fig. 36 Scanning electron microscopy (SEM) micrographs of dry mud surface. Small porous formed on dry mud surface due to liquid solution sediments under gravity (marked in circles).

mixture is extracted. The solution extracted is dispensed on the laser textured zirconia surface. In addition, pH of the liquid solution is measured after 6 h of mixing and it is observed that the pH of the liquid solution increases to þ 7.6 showing the basic state. The increase of pH of the liquid solution is because of the presence of OH ions in the solution, which are due to the dissolution of alkaline and alkaline earth metals in water condensate. In order further assess the characteristics of the liquid solution, the dispensed solution on the laser textured zirconia surface is left for drying in a control chamber at room temperature. Fig. (35) shows SEM micrographs of the dried solution. The dried solution forms crystals at the zirconia surface with varying sizes. Formation of crystals is associated with the presence of OH compounds in the liquid solution. Hence, the dried liquid solution adheres to zirconia surface via crystal formation upon its drying. In order to investigate the mud formation on zirconia surface due to water condensate on the dust particles, actual humid air conditions are resembled and mud is formed on zirconia surface accordingly. The mud formed on zirconia surface is dried, similar to that of the liquid solution, and the dry mud characteristics are analyzed. Fig. 36 shows SEM micrographs of the dry mud surface. The dust particles at the dry mud surface attach with the fine size bonding layers in between them. However, some small porous like structures are observed at the dry mud surface. The porous like structures are related to the liquid solution, which flows toward the laser textured zirconia surface under the gravity. Fig. 37 shows SEM micrographs of the dry mud removed laser textured surface. Some dust residues are observed on the laser textured surface (Fig. 37). This indicates the strong adhesion of the dry mud at the surface. However, no chemical attack, resulting in pit sites or pin holes, is observed at the laser textured surface after the dry mud removal. However, some closely attached structures at the surface are observed which are associated with the mud solution formed after desalinated water jet cleaning of the surface. The tangential force required to remove the dry mud from the laser textured surface is measured by using the micro-tribometer. The tangential forces measured for as received zirconia and the laser textured surfaces are also shown in Fig. 38 for comparison.

Dust Repellent Materials

10 kV

5 μm 0000

×4,300

867

21 40 SE I

Fig. 37 Scanning electron microscopy (SEM) micrographs of laser textured surface after dry mud removal. Attachment of fine size mud residues on surface is evident.

0.8 Dry mud Laser textured As received

Tangential force (N)

0.6

0.4

0.2

0 0.0

0.2

0.4

0.6

0.8

1.0

Distance (mm) Fig. 38 Tangential force along the scratch length for dry mud removal, laser textured surface, and as received surface.

Tangential force variation along the scanning length remains low, which is attributed to the low friction coefficient of as received surface. Tangential force for the laser textured surface is also relatively lower than that corresponding to the forced required removing the dry mud from the laser textured surface. However, the tangential force for the laser textured surface remains slightly lower than as received surface. The attainment of slightly low tangential force is related to the low surface energy of the laser treated surface because of nitride compound formation at the surface, and surface hardness increases after the laser treatment process, i.e., it increases from 1600740 HV (as received surface) to 1960740 HV (laser treated surface). Moreover, some small variations of tangential force occur along the scanning direction. This variation is attributed to the micro/nano textures of the laser treated surface. The attainment of high values of the tangential force for the dry mud removal from the laser textured surface indicates the strong adhesion between the dry mud and the laser textured surface. The strong adhesion force is because of the interlayer of dried mud solution between the dry mud and the laser textured surface, which acts like an adhesion layer at the textured surface. The adhesion work is determined from the integration of the tangential force along the scanning distance. However, the area under the tangential force for the laser textured surface is considered to be the frictional work. Therefore, subtraction of the area under the tangential force due to the dry mud removal and the frictional work results in adhesion work of the dry mud on the laser textured surface, which is estimated as 0.145 mJ.

2.26.14.2

Laser Surface Texturing of Tungsten and Environmental Dust Effect on Textured Surfaces

Tungsten oxide has good absorption characteristics and it is used as one of the selective surface materials for solar energy harvesting in volumetric receivers. In order to form tungsten oxide and texture the surface with micro/nano pillars, the method of

868

Dust Repellent Materials

laser control melting/ablation can be used, which is similar to that of zirconia surface. Since the method adopted is similar to that used for zirconium texturing, it is briskly introduced in the line of the previous study. Considerable research studies were carried out to examine solar selective surfaces in relation to energy harvesting applications of tungsten. Spectrally selective absorber coating of tungsten alloy was studied by Dan et al. [134] for solar thermal applications. They demonstrated that the tungsten based selective absorbing coating with excellent thermal stability was possible and the coating could be used for photo-thermal conversion at temperatures of up to 5001C. The microstructure, chromaticity, and thermal stability of tungsten alloy toward possible application of spectrally selective solar absorber were examined by Gao et al. [135]. They showed that coating consisted of tungsten carbide and aluminum oxide anti-reflectance layers exhibited excellent selective absorbing properties with a high solar absorptance and a low thermal emittance. Tungsten based solar absorber coatings for combined heat and power systems were investigated by Wackelgard et al. [136]. The findings revealed that the resulting coatings adhered well to the substrate of stainless steel and showed stable response to oxidation at 3501C. Investigation of solar selective absorbers based alumina-tungsten cermet was carried out by Rebouta et al. [137]. They indicated that the coatings exhibited a solar absorptance of 93%–95% and an emissivity of 7%–10% (at 4001C) and they had excellent thermal stability coating during heattreatments at 4001C in air for 2500 h and at 5801C in vacuum for 850 h. Tungsten black absorber for solar energy harvesting was studied by Fan et al. [138]. They introduced a structured tungsten slab with subwavelength periodicity displaying near-complete absorptivity throughout the entire solar spectrum over a wide angular range. A study for enhancement of absorptance of surfacetextured tungsten thin film in relation to solar applications was carried out by Gao et al. [139]. They showed that the surfacetextured single layer tungsten thin film exhibited a high absorptance of 0.74 and a low emittance of 0.05 due to the high aspect ratio subwavelength structures. Thermal stability of W-cermet-based spectrally selective solar absorber was examined by Cao et al. [140]. They demonstrated that tungsten improved thermal stability and reduced the infrared emittance, which could be a good candidate for the infrared reflector layer. Tungsten nanowire metamaterial as selective solar thermal absorbers was studied by Chang et al. [141]. They showed that the nanowire-based selective solar absorber with base geometric parameters could reach 83.6% of thermal conversion efficiency with low independence of incident angle. The viability of micro/nano textured tungsten as an efficient solar absorber was investigated by Ungaro et al. [142]. They indicated that micro/nano textured tungsten had extremely high absorption across the solar spectrum along with relaxed requirements for manufacturing, allowing them to be applied for power generation. Investigation of tungsten carbide nanostructures for solar applications was carried out by Vijayakumar et al. [143]. They showed that the PV performance was strongly affected by the sintering temperature of the tungsten carbide composite material and the tungsten carbide nanorods sintered at 8001C demonstrated improved performances.

2.26.14.2.1

Laser texturing parameters and surface characteristics

The laser treatment process and resulting surface characteristics are given briskly. Tungsten tiles with 3 mm thickness were used as workpieces. The CO2 laser (LC-ALPHAIII) was incorporated to irradiate the tungsten surfaces. The nominal output power of the laser was 2 kW and the irradiated spot diameter at the workpiece surface was in the order of 200 mm. Oxygen at high pressure (600 kPa) was used as an assisting gas during the texturing. The laser pulsing frequency was set at 1500 Hz, which in turn gave rise to 70% overlapping ratio for the irradiated spots at the surface. In order to avoid several repeats during the experiments, the initial tests were conducted to select the laser texturing parameters in such a way that the laser parameters resulting in free surface asperities, such as micro-cracks, voids, and large size cavities, were selected. The results of the initial tests revealed that reducing laser power by 10% while keeping laser scanning speed same gave rise to small size texture height without micro/nano pillars at the surface. On the other hand, reducing laser scanning speed by 10% while keeping the laser output power same resulted in cracks and cavities at the surface. Consequently, through controlling the laser power settings, beam intensity distribution, pulse repetition rate, spot size, and the scanning speed, crack free surface texture could be realized. Laser surface texturing conditions are given in Table 8. Material characterization of the laser textured surfaces was carried out using Jeol 6460 electron microscope. Bruker D8 Advanced having Cu–Ka radiation was used for XRD analysis. A typical setting of XRD was 40 kV and 30 mA and scanning angle (2y) was ranged 20–80 degrees. A linear micro-tribometer (MCTX-S/N: 01–04300) was used to determine the tangential force required to remove the dry mud formed the laser textured surface. The equipment was set at the contact load of 0.03 N and end load of 5 N. The scanning speed was 5 mm/min and loading rate was 1 N/s.

2.26.14.2.2

Results and discussion

Fig. 39 shows SEM micrographs of laser textured surface. Laser textured surface has regular laser scanning tracks with almost equal spacing (Fig. 39(A)). The laser scanning tracks are formed through overlapping of laser irradiated spots at the surface. Since laser pulses have high repetition rate, they overlap during scanning. The laser pulse intensity decays exponentially across the irradiated spot due to Gaussian beam intensity distribution. In this case, the peak power intensity occurs at the irradiated spot center and the power intensity reduces at the irradiated spot edge. This, in turn, gives rise to surface evaporation in the central region of the irradiated spot and melting in the close region of the irradiated spot edges. However, melt flow from the cavity edges toward the cavity center modifies the surface texture. Consequently, micro/nano size pillars ((Fig. 39(B)) and small size cavities (Fig. 39(C)) are formed at the surface. The close examination of SEM micrographs reveals that no micro-cracks and no other asperities, such as large size voids and porous like structures, are formed at the surface. In addition, sharp peaks with various heights and some small valleys are formed; in which case, the pillar heights changes at the surface. The average roughness of the surface is in the order of 0.57 mm. Table 9 gives EDS data at the surface prior and after the laser texturing. Elemental composition does not change

Dust Repellent Materials

20 kV

50 μm 0000

×370

11 46 SE I

20 kV

(A)

1 μm

×16,000

0000

869

11 40 SE I

(B)

Fig. 39 Scanning electron microscopy micrographs of laser textured surface: (A) regular scanning patterns and (B) micro/nano size textures.

1600 W WO2 WO3 WO3.H2O

Intensity (a.u.)

1200

800

400

0 20

30

40

50

60

70

80

90

2 Fig. 40 X-ray diffractogram of laser textured surface.

considerably after the laser texturing. In addition, oxygen is observed in the EDS data after the laser texturing. The presence of oxygen on the treated surface indicates that tungsten oxide compound is formed at the surface during the laser treatment process because of high pressure oxygen. Fig. 40 shows X-ray diffractogram of laser treated surface. The presence of oxide peaks reveals that WO3 is formed on the surface during the laser gas assisted texturing. It should be noted that high pressure oxygen is used as an assisting gas during the laser processing. This, in turn, gives rise to formation of oxide species at the surface. Fig. 41 shows SEM micrographs of cross-section of the laser treated layer. The depth of laser treated layer extends almost 25 mm below the surface (Fig. 41(A)). A dense layer is formed in the surface vicinity of the laser treated layer (Fig. 41(B)). The formation of dense layer is associated with high cooling rates at the surface due to the impingement of the assisting gas and formation of oxide compounds, which give rise to volume shrinkage at the surface. The cross-section of the laser treated layer is free from asperities such as voids. In addition, no micro-cracks are observed in the dense layer despite the high cooling rates occurring at the surface. Some small size grains are observed in the region close to the dense layer. The formation of small size grains are attributed to the high cooling rates at the surface. The demarcation line is observed at a depth below the surface, where the laser treated layer ends (Fig. 41(A)). The average microhardness is in the order of 430720 HV, which is almost 1.5 times of the untreated base material hardness (280715 HV). The microhardness enhancement is associated with the dense layer formed at the surface under the high cooling rates because of the high pressure assisting gas. It should be noted that the hardness measurements are repeated ten times at different locations on the surface and the average microhardness values are reported. The error estimated for the hardness measurements based on the repeats is in the order of 5%.

870

Dust Repellent Materials

Dense layer

Laser treated layer

(A)

(B)

Fig. 41 Scanning electron microscopy (SEM) micrographs of laser treated layer cross-section: (A) depth of treated layer and (B) dense layer at the surface.

0.03 0.03

O-W-O O-H

Absorbance

0.02 O-W-O

-OH

0.02 0.01 0.01 0.00 600

1000

1400

1800 2200 Wavenumber (cm−1)

2600

3000

Fig. 42 Fourier transform infrared spectroscopy (FTIR) data for laser treated surface.

Fig. 42 shows FTIR data obtained for the laser treated workpiece surface. The absorption peaks for WO3 are located at 730 cm 1 and 890 cm 1, which are associated with the stretching vibrations of O–W–O bonds. The weak absorption peak is observed at 945 cm 1, which is related to the stretching vibration of W ¼ Ot, here Ot corresponds to therminal oxygen [144–146]. It should be noted that after the laser treatment, the fine structures are formed in the surface region (Fig. 41(A)), which alters the lattice symmetry to monoclinic [147]. This, in turn, gives rise to various polar vibrations of cubic WO3. Moreover, the peak at 1590 cm 1 corresponds to the (O–H) bending mode while the small peak at 2932 cm 1 demonstrate ( OH) group of WO3 [148,149]. This attributed to the stretching vibration of (OH), which is structurally bonded to WO3 matrix and indicates the possibility of Coulombic attractive forces to be operative between the hydroxyl groups on WO3 [150]. In the humid air ambient, the water condensates on the dust particles and forms the mud-like structures on the hydrophobic surface. Upon drying at high temperatures, the soft mud structure turns into the solid dry mud adhering to the surface. The tangential force required to remove the dry mud from the textured surface is measured using the micro-tribometer. This provides the force required to remove the dry mud from the laser textured surface. The tangential force variation along the hydrophobic surface is shown in Fig. 43. Moreover, the friction force on the laser textured surface with absence of the dry mud is also provided for comparison. The tangential force remains almost three times of the frictional force along the surface. This demonstrates that the adhesion of the dry mud on the laser textured surface is significant and force required to remove the dry mud from the textured surface is considerable. Moreover, some oscillations are observed in the tangential force. This behavior is attributed to the locally strong adhesion of the dry mud on the surface. In this case, the dried liquid solution forms an interlayer between the dry mud and the textured surface. This gives rise to increased adhesion of the dry mud at the textured surface. The wetting state of the liquid

Dust Repellent Materials

871

1.5 Laser treated As received

Tangential force (N)

1.2

0.9

0.6

0.3

0.0 0.0

0.2

0.4 0.6 Distance (mm)

0.8

1.0

Fig. 43 Tangential force required to remove dry mud from laser textured surface and frictional force for as received surface.

solution on the textured surface depends on the spreading coefficient of the solution [151]. In line with the previous study [151], the spreading coefficient remains negative while indicating formation of locally scattered small liquid solution ponds rather than forming a liquid film on the textured surface. Consequently, upon drying of the liquid solution, locally scattered crystallized structures are formed in between the dry mud and the textured surface, where the ponds of the liquid solution are formed initially. This enhances the tangential force locally along the textured surface.

2.26.14.3

Laser Surface Texturing of Alumina Surface and Environmental Dust Effects on Textured Surfaces

Laser texturing of alumina surfaces and environmental dust effect on the textured surface is presented in line with the previous study [152]. Aluminum oxynitride (AlON) is one of the ceramics that has high hardness and superior optical characteristics such as high optical transmittance in the near-ultraviolet, visible, and mid-wave-infrared radiations [153]. It finds applications in various industries, particularly for those involved with blast-resistant transparent windows and infrared optics. Aluminum oxynitride has a cubic spinel structure, and it can be fabricated by the conventional methods incorporating ceramic powder processing technologies [154]. However, producing a thin layer of aluminum oxynitride coating on the ceramic surfaces via the method of conventional powder processing is challenging and may involve a high cost. One of the methods to produce aluminum oxynitride film on the ceramic surfaces is to incorporate laser gas assisted surface processing [155]. Laser surface treatment of alumina tiles with the presence of assisting gas, composing of a mixture of nitrogen and oxygen, in the treated region can produce a thin layer of aluminum nitride and aluminum oxynitride film at the surface while improving the tribological properties of the treated surface significantly [155]. Since the laser controlled melting/ablation is involved with high temperature processing, care must be taken to reduce the surface defects such as micro-cracks, deep cavities, and voids [156]. Forming a thin film of coating on ceramic surface prior to laser treatment improves surface properties in terms of microhardness and fracture toughness significantly [157]. In addition, the selection of laser processing parameters becomes critical for securing the desired surface topology and microstructure [158]. Laser texturing can also be used to improve surface hydrophobic characteristic, which in turn lowers the friction coefficient and minimizes the adhesion of the particles on the surface [152]. Consequently, investigation of laser oxynitriding of alumina surface toward improving the surface hardness and tribological properties becomes essential. Considerable research studies were carried out to examine laser treatment of alumina surfaces. A study on aluminum oxynitride thin films grown by ion-beam-assisted pulsed laser deposition was carried out by Zabinski et al. [159]. They showed that the films were nearly stoichiometric except for depositions in a vacuum and the nitrogen concentration could be controlled through selection of gas pressure and ion energy. In addition, the crystalline Al–O–N films showed improved hardness in comparison to amorphous films. Laser texturing of alumina surface for improved hydrophobicity was studied by Yilbas et al. [160]. The findings revealed that laser controlled ablation resulted in micro/nano texturing of the surface. Although the surface texture did not exactly follow the regular pattern, pillars, and dimples like structures were formed, and the averaged surface roughness was within the submicro scale. In general, laser texturing improved the surface hydrophobicity; however, Wenzel and Cassie and Baxter states were present at the treated surface due to the variation in the surface texture. Tribology and super-hydrophobicity of the lasercontrolled-melted alumina surfaces with the presence of hard particles at the surface were examined by Yilbas et al. [161]. They demonstrated that laser treatment produced micropoles, nanopoles, and small size cavities at the surface, which enhanced surface hydrophobicity. The microhardness of the laser-treated surface increased almost 50% because of the dense layer formed on the surface, and the residual stress was in the order of 2 GPa, which was compressive. The scratch resistance and friction coefficient of the laser-treated surface were superior. Laser micro machining of epoxy/alumina nanoparticle composite generating the micropatterns was investigated by Psarski et al. [162]. They showed that the textured surface had hierarchical topography, which

872

Dust Repellent Materials

consisted of hexagonally spaced microcavities and nanoparticle agglomerates. Laser short-pulse fabrication of a super-hydrophobic Al2O3 surface was studied by Jagdheesh [163]. The findings revealed that the geometry of the laser-machined pattern played a major role in changing the wetting properties rather than the chemical changes induced on the surface. The micropillars exhibited a consistent super-hydrophobic surface with a static CA in the order of 15073 degrees.

2.26.14.3.1

Experimental

Alumina (Al2O3) tiles (Ceram Tec-ETEC, 2010) with 4 mm thickness were used as workpieces. The CO2 laser (LC-ALPHAIII) was used to irradiate the alumina tile surfaces. The nominal output power of the laser was 2 kW and the irradiated spot diameter at the workpiece surface was about 200 mm. The mixture of nitrogen and oxygen at high pressure (600 kPa) was used as an assisting gas P during the texturing. The partial pressure of nitrogen was PNT2 ¼ 23, where PN2 is the nitrogen pressure in the mixture and PT is the P total pressure of mixture and the particle pressure of oxygen was POT2 ¼ 13, where PO2 is the oxygen pressure in the mixture. The laser pulsing frequency was set at 1500 Hz, which in turn gave rise to 70% overlapping ratio for the irradiated spots at the surface. The initial tests were conducted to select the laser texturing parameters in such a way that the laser parameters resulting in surface defects, such as micro-cracks, voids, and large size cavities, were avoided. The results of the initial tests revealed that reducing laser power by 10% while keeping laser scanning speed same gave rise to small size texture height without micro/nano pillars at the surface. On the other hand, reducing laser scanning speed by 10% while keeping the laser output power same resulted in cracks and cavities at the surface. Consequently, through controlling the laser power settings, beam intensity distribution, pulse repetition rate, spot size, and the scanning speed, crack free surface texture could be realized. Laser surface texturing conditions are given in Table 8.

2.26.14.3.2

Results and Discussion

The findings of laser texturing of alumina and environmental dust effect on the laser textured surface are presented in line with the previous study [155]. Fig. 44 shows SEM micrographs of the laser textured surface. The textured surface contains regular scanning patterns (Fig. 44 (A)), which are formed due to overlapping of the laser repetitive pulses on the workpiece surface during the scanning. Since the scanning speed of the laser beam is kept constant and the frequency of the repetition rate of the laser pulses is 1500 Hz, the overlapping ratio of the irradiated spots is in the order of 70% on the workpiece surface. The laser pulse intensity follows the Gaussian distribution across the irradiated spot on the surface, i.e., the maximum peak occurs at the irradiated spot center. This

Laser scanning tracks

20 kV

×120 100 μm 0000 09 48 SE I

(A)

20 kV

×2,300

10 μm

07 40 SE I

(B)

Dense structure Treated layer

15 kV

×6,500

2 μm 0000 11 38 SE I

(C) Fig. 44 Scanning electron microscopy (SEM) micrographs of laser textured surface: (A) regular laser scanning tracks, (B) textured surface, and (C) cross-section of laser treated layer and dense structure at the surface.

Dust Repellent Materials

873

gives rise to high intensity laser beam interaction of the surface in the region of the irradiated spot center. Therefore, evaporation takes place in this region while surface melting occurs in the near region of the irradiated spot edges. The flow of molten material from the region close to the irradiated spot edge toward the irradiated spot center modifies the size of the cavity created through the surface evaporation. Consequently, combination of evaporation and melting at the irradiated surface modifies the surface texture. This, in turn, gives rise to the formation of shallow cavity and micro/nano size poles and pillars at the surface (Fig. 44(B)). Although melting of the substrate material occurs in the close region of the irradiated spot edge, no overflow of the molten material along the scanning tracks are observed. In addition, the proper selection of the laser treatment parameters, through evaluating the initial tests, provides controlled melting and ablation at the surface. In general, laser treated surface is free from defects sites such as pores, large size cavities, and micro-cracks despite the fact that the mixture of oxygen and nitrogen at high pressure is used as an assisting gas during the texturing, which increases the cooling rates at the surface. In this case, the assisting gas enhances the convection heat transfer from the irradiated surface toward its surroundings while increasing temperature gradients at the irradiated surface. However, the presence of the crack free laser treated surface indicates that the heat conduction from the recently formed laser scanning tracks toward the previously formed tracks generate a self-annealing effect below the surface. This, in turn, lowers the temperature gradient below surface vicinity and suppresses the high thermal strain in this region. Consequently, micro-crack formation during the laser texturing of the surface is avoided by the self-annealing effect. The dense layer is formed in the surface vicinity of the laser treated layer. This situation can be observed from Fig. 44(C) in which the crosssection of the laser treated layer is shown. The laser treatment results in a uniform thickness of the treated layer, which extends almost 20 mm below the surface. The formation of dense layer is associated with the high cooling rates at the surface and formation nitride compounds such as AlN and AlON in the surface region of the laser treated layer, which are less than the density of alumina. This in turn results in volume shrinkage in the surface region. The formation of AlN and AlON compounds is evident from Fig. 45, in which X-ray diffractogram of the laser treated surface is shown. The formation of aluminum oxynitride (AlON) is related to the high temperature processing at the surface during the laser treatment. Since the assisting gas pressure is high at the surface because of the stagnation flow of the impinging nitrogen assisting gas, 2AlO þ N2-2AlON reaction takes place in the surface region while forming AlON compound. During the laser heating of the substrate surface, the transformation of g-Al2O3 phase into thermodynamically stable a-Al2O3 can also take place, which can be observed after comparing diffractograms of the laser treated and as received workpiece surfaces. Table 9 gives the EDS data for as received and the laser textured surfaces. Elemental composition before and after the laser treatment process changes slightly; however, nitrogen is observed from the EDS data despite the fact that quantification of nitrogen is involved with errors because of being the light element. Nevertheless, the presence of nitrogen in EDS data reveals that nitride compounds are formed on the laser textured surface. The microhardness of the laser textured surface increases notably (1750730 HV) as compared to that of the as received alumina (1130730 HV). This is attributed to the formation of AlON and AlN compounds at the laser treated surface, particularly AlON, which has considerably high hardness value (1750730 HV), and high cooling rates from the surface due to convection cooling effect of the assisting gas. Fig. 46(A) shows SEM micrographs of the dry mud surface on the laser textured surface. The dry mud surface consists of undissolved dust particles and some small cavities in between the particles. The cavities are formed due to the dissolution of alkaline (Na, K) and alkaline earth (Ca) metals in water; in which case, liquid solution flows toward the interface between the mud and the textured surface under the gravitational force while leaving some small cavity like structures on the mud surface. To assess the dissolution of the alkaline and alkaline earth dust compounds in water, the dust particles are mixed with desalinated water for few hours and the liquid solution formed due to this mixing is analyzed in terms of pH and its composition via using inductively 6000 CaCO3 NaCl MgO AlO(OH)

Laser treated and mud removed

Relative intensity

5000

4000

3000

 − Al2O3  − Al2O3 AIN AION

Laser treated

2000

1000 As received 0 20

30

40

50

60

70

2  (degrees) Fig. 45 X-ray diffractogram of as received alumina, laser textured, and dry mud removed laser textured surface.

80

874

Dust Repellent Materials

Textured surface

Small voids

Dry mud

Dried solution

5 kV

×7,000

4 kV

2 μm 0000 16 35 SE I

(A)

×1,500

10 μm 0000 27 35 SE I

(B)

Fig. 46 Scanning electron microscopy (SEM) micrographs of dry mud: (A) surface of dry mud and presence of voids at the surface and (B) cross-section of dry mud and presence of dried liquid at the interface of textured surface and dry mud.

coupled plasma (ICP) spectrometer. The findings revealed that pH of the solution is in the order of 7.6 and elements of Na, K, Ca, and Mg are observed from the ICP data. Consequently, the dissolution of alkaline and alkaline earth metal compounds in water is responsible for the increase of solution pH and formation of cavity like structure on the dry mud surface. Fig. 46(B) shows SEM micrographs of the mud cross-section on the laser textured surface. However, the wetting state of the laser textured surface depends on the the spreading coefficient of the liquid solution (SslðaÞ ). In this case, the spreading coefficient of the liquid solution yields SslðaÞ ¼ gla

gsa

gsl

where gla is the surface tension of liquid solution in air, and gsa is the surface free energy of laser textured surface in air, and gsl is the interfacial tension between the liquid solution and the laser textured surface. The surface tension of the liquid solution is measured using the capillarity tube method and it is found to be in the order of 0.085 mJ/m2, which is close to that of the water. However, the interfacial tension between the liquid solution and the laser textured surface could not be measured with accuracy with our current laboratory facilities. Therefore, we consider that the interfacial tension between the liquid solution and textured surface is in the same order as desalinated water and the laser textured surface. This assumption yields the interfacial tension in the order of 86.1 mJ/m2164[]. It should be noted that this assumption can be justifiable because of the surface tension of the liquid solution, which is very close to the surface tension of the water. Therefore, the spreading coefficient of the liquid solution on the laser textured surface (SslðaÞ ) becomes in the order of 124.1 mJ/m2, which is SslðaÞ o0. Therefore, the liquid solution partially wets the textured surface while forming separated small liquid islands in between the dust particles and the laser textured surface. Consequently, the dried liquid solution forms small islands of dried film at the interface of the dust particles and textured surface. This situation can be seen from Fig. 46(B); in which case, small bright regions at the interface corresponds to the dried liquid solution. Fig. 47 shows the tangential force required to remove the dry mud from the laser textured surface and the tangential forces corresponding to the laser textured and as received surfaces. The tangential force demonstrates oscillatory behavior along the scratch length. This is attributed to the strong adhesion of the dry mud on the laser textured surface because of the presence of dried mud solution at the interface. Since the dried mud solution does not extend uniformly forming a film at the interface, the locally scattered dried mud solution at the interface is responsible for the strong adhesion of the dry mud at the surface. The adhesion work can be determined from the integration of the the force difference between the tangential and friction along the scratch length. In this case, the adhesion work is estimated in the order of 0.275 mJ. The tangential force on the laser textured and as received surfaces is associated with the frictional force. Consequently, the frictional force increases for the laser textured surface because of the surface texture, which increases the adhesion of the indenting tip to the workpiece surface. Fig. 48 shows SEM micrographs of the laser textured surface after the dry mud removed by a water jet. Few mud residues are evident on the laser textured surface (Fig. 48(A)) after the dry mud removal. The mud residues strongly adhere at the textured surface and the elemental analysis of these residues are similar to that of normal dust particles. However, the elemental concentration of alkaline earth metal (Ca) remains slightly higher in the mud residues, which can be associated with the dried liquid solution at the interface between the mud and the laser textured surface. The close examination of the micrographs (Fig. 48(B)) reveals that some crystallized structures are present on the laser textured surface after the dry mud is removed by a water jet. The crystal structures are attributed to the residues of the dried liquid solution on the surface, which is consistent with the previous findings [95]. Nevertheless, the surface is free from mud induced defects such as local pit sites or corrosion induced pin holes unlike that of PC surface reported in the previous study [95]. The dry mud residues lower the water droplet CA, which can be seen from the Fig. 31. This indicates that the hydrophobic characteristics of the laser textured surface are influenced by the dry mud residues. The microhardness of the surface after the dry mud removal remains almost same as the laser textured surface, which is in the order of 1700 HV The FTIR is carried out on the laser textured surface after the mud removed by the pressurized water jet.

Dust Repellent Materials

875

0.8 Laser treated with presence of dry mud 0.7

Laser textured surface As received

Tangential force (N)

0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0

0.2

0.4

0.6

0.8

1.0

Distance (mm) Fig. 47 Tangential force along the scratch length for as received, laser textured, and laser textured with presence of dry mud surfaces.

15 kV (A)

×37 500 μm 0000 25 40 SE I

15 kV

×130 100 μm 0000 23 40 SE I

(B)

Fig. 48 Scanning electron microscopy (SEM) micrographs of mud residues at the surface: (A) mud residues at the surface, and (B) some crystalized dried mud solution at the surface after water jet removal of dry mud.

2.26.15

Conclusions

Environmental dust has severe effects on the thermal and optical performances of the solar energy harvesting devices. This is due to the fact that the dust accumulation forms a layer at the surface while scattering and reflecting the incident solar radiation. In order to sustain the solar energy harvesting device efficiency, dust-removal from the surface becomes necessary. One of the alternatives for dust-removal process is to generate self-cleaning surfaces. This involves surface texture with micro/nano pillars. One of the methods to achieve such texture arrangement is to use laser beam while generating controlled melting/ablation at the surface. On the other hand, dust particles have various shapes and elemental composition. Small dust particles attach to large size particles due to electrostatic forces. In the humid ambient, some dust particle compounds dissolve in condensed water while forming the chemically active fluid, which in turn damage the energy harvesting device surface. The conclusion derived from the present study is given below. Environmental dust particles varied in size; the average particle size was 1.2 mm. Their elemental composition was non-uniform within the variously sized dust particles. Small dust particles attached to the large particles due to the forces exerted by charges present on the different materials. The aspect ratio and shape factor of the dust particles changed with particle size; however, no simple correlation was found among the dust particles, shape factor, and aspect ratio. The shape factor approached unity for small particles (r2 mm), whereas for large particles (Z10 mm), the median shape factor approached 3. The liquid solution extracted from the dust particles contained ionic compounds (OH ), which increased the solution pH (8.4). When mud is formed from the dust particles on the PC sheet, the dissolved alkaline and alkaline earth metal hydroxyls (OH ) flow at the interface of the mud and PC surface while form a layer in this region after drying. Alkaline hydroxyls attacked the PC surface, thereby altering the

876

Dust Repellent Materials

vibrational state of the macromolecules at the surface region, and increasing the microhardness and lowering the UltravioletVisible (UV–visible) transmittance of the resulting PC surface. In addition, the bonding of calcite and the formation of hydroxyls compounds at the PC surface increased the adhesion of the dry mud on the surface. The adhesion work required to remove the dry mud from the PC surface increased significantly because of the presence of the dried mud solution at the interface of the dry mud and the PC sheet surface. The influence of chemo-mechanical behavior of the mud formed from the dust particles on PC surfaces is novel and significantly alters the properties of the PC. The findings provide broad insight into the performance of the PC for solar energy systems when subjected to the environmental dust and mud. The combination of surface evaporation and melting results in micro/nano size pillars at the textured surface. Since the repetitive laser irradiated spots have a high frequency (1500 Hz), they give rise to the irradiated spot overlapping ratio of 72% at the surface. This in turn gives rise to regular laser scanning tracks at the treated surface. The condensation of water onto the environmental dust particles in the humid air ambient is examined and its influence on the laser textured zirconia surface is analyzed. The dust particles are collected from the local area (Dhahran) of Saudi Arabia and humid air ambient is simulated incorporating the actual local environmental conditions. Adhesion force between the dry mud, which is formed from the dust particles mimicking water condensate onto the dust particles, and laser textured zirconia surface is measured using the micro-tribometer. The adhesion work required to remove the dry mud from the laser textured zirconia surface is determined from the tangential force data. In order to assess the influence of the dry mud on the laser textured surfaces, a desalinated water jet is used to clean the dry mud from the surface. The dust particles and the mud residues on the dry mud removed surface are characterized using the analytical tools. In addition, the morphological and elemental changes of laser treated surface prior and after removing the dry mud are investigated. In general, the laser treated surface is free from large size asperities such as large size cracks and open voids. The nitride compounds are formed at the laser textured surfaces after laser texturing under the ambient of high pressure nitrogen assisting gas. The laser textured surfaces demonstrate hydrophobic characteristics because of the arrays of micro/nano pillars and nitride compounds formed at the surface. The dust particles possess various elements including silicon, alkaline (Na, K) and earth alkaline (Ca) metals, sulfate, oxide, and chloride compounds. The dust particles have various shapes and sizes and the average dust particles are in the order of 1.2 mm. Small dust particles have electrostatic changes and they attach to the large size particles. The alkaline and earth alkaline metallic compounds in the dust particles dissolve in water condensate while forming a chemically active liquid solution, which sediments on the laser texture surface under the gravity. The crystals with various sizes are formed on the laser textured surface upon drying of the liquid solution. The tangential force required for removal of the dry mud from the textured surface is significantly higher than the frictional force. This behavior is related to the strong adhesion between the dry mud and the textured surface. In this case, dried liquid solution in between the dry mud and the laser textured surface plays an important role toward increasing adhesion. This study provides useful information on the zirconium nitride formation through laser surface processing, which can be used for solar absorption as a selective surface. In addition, it also gives insight into the characteristics of laser textured yttria-stabilized zirconia surface when subjected to the dusty environments in a humid air condition. The use of nitrogen and oxygen mixture, as an assisting gas during laser texturing, facilitates the formation of AlN and AlON compounds at the surface. In addition, laser texturing results in uniform treated layer below the surface with dense structures in the surface vicinity. The formation of the dense structures is associated with the high cooling rates at the surface and the volume shrinkage because of the formation of AlN and AlON compounds in the surface region. The dense layer formed in the surface vicinity increases the microhardness of the surface significantly. The laser textured surface composes of micro/nano size pillars, which demonstrate the hierarchical structures at the surface. The hydrophobicity of the textured surface improves significantly. This is associated with the surface texture pattern and the nitride and oxynitride compounds formed at the surface, which gives rise to lower surface energy than that of alumina. The dust particles have different sizes and shapes, and compose of various elements. The averaged dust particles are in the order of 1.2 mm. Water condensate on the dust particles in humid air ambient dissolves alkaline and alkaline earth compounds in the dust while forming a liquid solution. The liquid solution flows toward the surface and forms an interfacial layer between the dust particles and the laser textured surface. Since the spreading coefficient of the liquid solution on the laser textured surface is negative (Sr0), the liquid solution partially wets the surface rather than forming a continuous film at the dust and the textured surface interface. Once the liquid solution dries, the adhesion between the dry mud and the laser textured surface increases significantly. This, in turn, gives rise oscillatory behavior of the tangential force required to remove the dry mud from the textured surface. In this case, the presence of the dried solution in between the dry mud and the textured surface increases the tangential force. The dry mud residues appear at the textured surface once it is removed by the pressurized water jet. The mud residues contain alkaline (Na) and alkaline earth metals (Ca) while indicating that the dried liquid solution is responsible for the attachment of the dry mud residues on the textured surface. In addition, the dry mud residues alter the hydrophobic state of the surface while lowering the water droplet CA. However, microhardness of the surface is not affected by the dry mud residues at the surface. The present work gives insight details of the dust and mud effects on the laser textured surface of alumina tiles and provides useful information on the laser textured characteristics including texture morphology, hydrophobicity, and microhardness. In the case of laser texturing of tungsten surface, the laser beam intensity remains higher at the irradiated spot center and the surface evaporation takes place at the irradiated spot center while surface melting occurs in the close region of the irradiated spot edges. The laser textured surface is free from asperities such as cracks and large size voids. The presence of oxygen as an assisting gas results in formation of WO3 at the laser textured surface. Laser treatment improves the hardness of the surface; in which case, the surface microhardness increases almost 1.5 times of the base material hardness after the laser treatment. The laser texturing alters the hydrophobic state of the surface. In this case, water droplet CA increases from 64 to almost 94 degrees after the laser treatment process,

Dust Repellent Materials

877

which is associated with the surface texture formed after the laser treatment process. The liquid solution with pH¼ 8.6 is formed once the water condensate on the dust particles in a humid air ambient. The dried liquid solution composes of various sizes of crystal structures on the textured surface. In addition, the mud-like structure is formed on the laser textured surface due to the partial dissolution of the dust particles in the water condensate. Once the mud-like structures are dried, the tangential force required to remove the dry mud from the textured surface becomes almost three times of the frictional force at the surface. This is more pronounced when the dried liquid solution is present at the interface of the dry mud and the textured surface. UV visible absorption characteristic of the laser textured surface is influenced by the dry mud and dried liquid residues at the surface. In this case, absorption of incident radiation reduces by almost 20% because of the scattering of the incident UV radiation by the dry mud and dried liquid residues at the surface. The present study gives insight into the laser gas assisted texturing of W and formation of WO3 and provides useful information for improvement of the absorption of solar radiation as well as the effect of environmental dust on the absorption.

2.26.15.1

Future Directions

Energy harvesting is the great importance in clean energy research; however, environmental dust effect elevates this importance because of its detrimental effect energy harvesting device performance. Recent changes in climate causes dust storms, particularly in the Middle East, while influencing device sustainable energy harvesting and operation. In order to improve the device performance and minimize dust accumulation effect investigation of the dust accumulation after effect becomes essential. One of the costeffective method for environmental dust-removal is to create the self-cleaning structures at the energy harvesting devices. Developing sustainable self-cleaning surfaces in harsh environments (high temperature and humidity) is challenging because of the organic materials involved on the textured surfaces. Such as Polystyrene/OTS or -Perfluorooctyltriethoxysilane decomposes and losses its hydrophobic characteristics at high temperature and humid ambient. Consequently, forming and development of hydrophobic self-cleaning surface is necessary. The important element of the self-cleaning surface is optical transparency. When such surfaces are used in PV applications, the textured surface does not block, scatter, or reflect the incoming radiation onto the device active surface. Although oil impregnation can improve the optical transmittance of textured surfaces, the dust accumulation deteriorates the quality of the oil film due to cloaking process. In addition, some of the transparent oils are expensive and requires regular replication on the textured surface. Therefore, maintaining the oil film is the challenges for the transparent self-cleaning surfaces. In addition, when texturing surface, environmental friendly texturing methods becomes vital, since the process is involved with some chemicals, which may damage the environment during the processing. Therefore, green processes for generating selfcleaning surfaces remains prime importance. Therefore, the future research directions should fulfill the current challenges while meeting the requirements of cost-effective green processing. There is almost no control over the environmental dust settlement on the surfaces I open environments and dust migration around the Globe. Consequently, living in the dust free environments is unlikely in the near future. This, in turn, elevates the importance of dust research toward safety, protection, and removal of environmental dust from surfaces, which have great importance for energy harvesting.

Acknowledgment The authors acknowledge the financial support of King Fahd University of Petroleum and Minerals (KFUPM) through Project# MIT11111–11112 to accomplish this work.

References [1] Meng J-p, Liu X-p, Fu Z-q, Zhang K. Optical design of Cu/Zr 0.2AlN0.8/ZrN/AlN/ZrN/AlN/Al34O62N4 solar selective absorbing coatings. Sol Energy 2017;146:430–5. [2] Parolari AJ, Li D, Bou-Zeid E, Katul GG, Assouline S. Climate, not conflict, explains extreme Middle East dust storm. Environ Res Lett 2016;11(11):114013. [3] Beattie NS, Moir RS, Chacko C, Buffoni G, Roberts SH, Pearsall NM. Understanding the effects of sand and dust accumulation on photovoltaic modules. Renew Energy 2012;48:448–52. [4] Yilbas BS, Ali H, Al-Aqeeli N, et al. Characterization of environmental dust in the Dammam area and mud after-effects on bisphenol-A polycarbonate sheets. Sci Rep 2016;6:24308. [5] Yilbas BS, Hassan G, Ali H, Al-Aqeeli N. Environmental dust effects on aluminum surfaces in humid air ambient. Sci Rep 2017;7. [6] Han JT, Xu X, Cho K. Diverse access to artificial superhydrophobic surfaces using block copolymers. Langmuir 2005;21(15):6662–5. [7] Shirtcliffe NJ, McHale G, Newton MI, Chabrol G, Perry CC. Dual‐Scale Roughness Produces Unusually Water‐Repellent Surfaces. Adv Mater 2004;16(21):1929–32. [8] Kinoshita H, Ogasahara A, Fukuda Y, Ohmae N. Superhydrophobic/superhydrophilic micropatterning on a carbon nanotube film using a laser plasma-type hyperthermal atom beam facility. Carbon 2010;48(15):4403–8. [9] Latthe SS, Imai H, Ganesan V, Rao AV. Superhydrophobic silica films by sol–gel co-precursor method. Appl Surf Sci 2009;256(1):217–22. [10] Ma M, Mao Y, Gupta M, Gleason KK, Rutledge GC. Superhydrophobic fabrics produced by electrospinning and chemical vapor deposition. Macromolecules 2005;38 (23):9742–8. [11] Zhang X, Guo Y, Zhang P, Wu Z, Zhang Z. Superhydrophobic CuO@ Cu 2 S nanoplate vertical arrays on copper surfaces. Mater Lett 2010;64(10):1200–3. [12] Yilbas B, Ali H, Al-Aqeeli N, Oubaha M, Khaled M, Abu-Dheir N. Laser gas assisted nitriding and sol–gel coating of alumina surfaces: effect of environmental dust on surfaces. Surf Coat Technol 2016;289:11–22. [13] Shao Y, Dong C. A review on East Asian dust storm climate, modelling and monitoring. Global Planet Change 2006;52(1):1–22. [14] McGee D, Winckler G, Stuut J, Bradtmiller L. The magnitude, timing and abruptness of changes in North African dust deposition over the last 20,000 yr. Earth Planet Sci Lett 2013;371:163–76. [15] Sissakian VK, Al-Ansari N, Knutsson S. Sand and dust storm events in Iraq. Nat Sci 2013;5(10):1084–94.

878

Dust Repellent Materials

[16] Ramli MA, Prasetyono E, Wicaksana RW, Windarko NA, Sedraoui K, Al-Turki YA. On the investigation of photovoltaic output power reduction due to dust accumulation and weather conditions. Renew Energy 2016;99:836–44. [17] Cao L. Superhydrophobic surface: Design, fabrication, and applications, University of Pittsburgh, Pittsburgh; 2010. [18] Ganesh VA, Raut HK, Nair AS, Ramakrishna S. A review on self-cleaning coatings. J Mater Chem 2011;21(41):16304–22. [19] Zimmermann J, Reifler FA, Fortunato G, Gerhardt LC, Seeger S. A Simple, One‐Step Approach to Durable and Robust Superhydrophobic Textiles. Adv Func Mater 2008;18(22):3662–9. [20] Miwa M, Nakajima A, Fujishima A, Hashimoto K, Watanabe T. Effects of the surface roughness on sliding angles of water droplets on superhydrophobic surfaces. Langmuir 2000;16(13):5754–60. [21] Assi A, El Chaar L, Effect of wind blown sand and dust on photovoltaic arrays. In: The 23rd European photovoltaic solar energy conference; 2008. [22] Katkar A, Shinde N, Patil P. Performance & evaluation of industrial solar cell wrt temperature and humidity. Int J Res Mech Eng Technol 2011;1(1):69–73. [23] Appels R, Lefevre B, Herteleer B, et al. Effect of soiling on photovoltaic modules. Sol energy 2013;96:283–91. [24] McCarthy JJ, Canziani OF, Leary NA, Dokken DJ, White KS, editors. Climate change 2001: impacts, adaptation, vulnerability. Cambridge, UK: Cambridge University Press; 2001. [25] Jordan DC, Wohlgemuth JH, Kurtz, SR. Technology and climate trends in PV module degradation. In: 27th European photovoltaic solar energy conference and exhibition; 2012. p. 3118–24. [26] Mekhilef S, Saidur R, Kamalisarvestani M. Effect of dust, humidity and air velocity on efficiency of photovoltaic cells. Renew Sustain Energy Rev 2012;16(5):2920–5. [27] Mani M, Pillai R. Impact of dust on solar photovoltaic (PV) performance: Research status, challenges and recommendations. Renew Sustain Energy Rev 2010;14 (9):3124–31. [28] Kobayashi S-i, Iino T. Kobayashi H, Yamada K, Yachi T, Degradation of output characteristics of a small photovoltaic module due to dirt spots. In: 27th international conference on telecommunications (INTELEC'05); 2005, p. 435–9. [29] Elminir HK, Ghitas AE, Hamid R, El-Hussainy F, Beheary M, Abdel-Moneim KM. Effect of dust on the transparent cover of solar collectors. Energy Conv Manag 2006;47 (18):3192–203. [30] Jiang H, Lu L, Sun K. Experimental investigation of the impact of airborne dust deposition on the performance of solar photovoltaic (PV) modules. Atmos Environ 2011;45(25):4299–304. [31] Brown K, Narum T, Jing N. Soiling test methods and their use in predicting performance of photovoltaic modules in soiling environments. In: 38th photovoltaic specialists conference (PVSC); 2012. p. 001881–5. [32] Touati F, Al-Hitmi M, Bouchech H. Towards understanding the effects of climatic and environmental factors on solar PV performance in arid desert regions (Qatar) for various PV technologies. In: 2012 first international conference on renewable energies and vehicular technology (REVET), IEEE; 2012, p. 78–83. [33] Rajput DS, Sudhakar K. Effect of dust on the performance of solar PV panel. Int J ChemTech Res 2013;5(2):1083–6. [34] Ghazi S, Sayigh A, Ip K. Dust effect on flat surfaces – a review paper. Renew Sustain Energy Rev 2014;33:742–51. [35] Kazem AA, Chaichan MT, Kazem HA. Dust effect on photovoltaic utilization in Iraq. Renew Sustain Energy Rev 2014;37:734–49. [36] Boyle L, Flinchpaugh H, Hannigan M. Natural soiling of photovoltaic cover plates and the impact on transmission. Renew Energy 2015;77:166–73. [37] Kohli R, Mittal KL. Developments in surface contamination and cleaning, Volume 4: detection, characterization, and analysis of contaminants. Norwich, NY: William Andrew; 2011. [38] Johnson K, Kendall K, Roberts A. Surface energy and the contact of elastic solids. In: Proceedings of the royal society of London A: mathematical, physical and engineering sciences, the royal society; 1971. p. 301–13. [39] Li Q, Rudolph V, Peukert W. London-van der Waals adhesiveness of rough particles. Pow Technol 2006;161(3):248–55. [40] Derjaguin BV, Muller VM, Toporov YP. Effect of contact deformations on the adhesion of particles. J Colloid Interface Sci 1975;53(2):314–26. [41] Muller V, Yushchenko V, Derjaguin B. On the influence of molecular forces on the deformation of an elastic sphere and its sticking to a rigid plane. J Coll Interface Sci 1980;77(1):91–101. [42] Maugis D, Pollock H. Surface forces, deformation and adherence at metal microcontacts. Acta Metal 1984;32(9):1323–34. [43] Korayem MH, Hezaveh HB, Taheri M. Dynamic modeling and simulation of rough cylindrical micro/nanoparticle manipulation with atomic force microscopy. Microsc Microanal 2014;20(6):1692–707. [44] Rabinovich YI, Adler JJ, Ata A, Singh RK, Moudgil BM. Adhesion between nanoscale rough surfaces: I. Role of asperity geometry. J Coll Interface Sci 2000;232(1):10–6. [45] Petean P, Aguiar M. Determining the adhesion force between particles and rough surfaces. Pow Technol 2015;274:67–76. [46] Cherukat P, Mclaughlin JB. The inertial lift on a rigid sphere in a linear shear flow field near a flat wall. J Fluid Mech 1994;263:1–18. [47] Zhang F, Busnaina AA, Fury MA, Wang S-Q. The removal of deformed submicron particles from silicon wafers by spin rinse and megasonics. J Electron Mater 2000;29 (2):199–204. [48] Goldman AJ, Cox RG, Brenner H. Slow viscous motion of a sphere parallel to a plane wall – I Motion through a quiescent fluid. Chem Eng Sci 1967;22(4):637–51. [49] Soltani M, Ahmadi G. On particle adhesion and removal mechanisms in turbulent flows. J Adhes Sci Technol 1994;8(7):763–85. [50] Dorrer C, Rühe J. Some thoughts on superhydrophobic wetting. Soft Matter 2009;5(1):51–61. [51] Parker AR, Lawrence CR. Water capture by a desert beetle. Nature 2001;414(6859):33–4. [52] Luo Z, Zhang Z, Hu L, et al. Stable bionic superhydrophobic coating surface fabricated by a conventional curing process. Adv Mater 2008;20(5):970–4. [53] Mock U, Michel T, Tropea C, Roisman I, Rühe J. Drop impact on chemically structured arrays. J Phys: Condens Matter 2005;17(9):S595. [54] Otten A, Herminghaus S. How plants keep dry: a physicist's point of view. Langmuir 2004;20(6):2405–8. [55] Sun M, Luo C, Xu L, et al. Artificial lotus leaf by nanocasting. Langmuir 2005;21(19):8978–81. [56] Ma M, Gupta M, Li Z, et al. Decorated electrospun fibers exhibiting superhydrophobicity. Adv Mater 2007;19(2):255–9. [57] Ogawa T, Ding B, Sone Y, Shiratori S. Super-hydrophobic surfaces of layer-by-layer structured film-coated electrospun nanofibrous membranes. Nanotechnology 2007;18:165607. [58] Khorasani MT, Mirzadeh H, Kermani Z. Wettability of porous polydimethylsiloxane surface: morphology study. Appl Surf Sci 2005;242:339–45. [59] Ma M, Hill RM, Lowery JL, Fridrikh SV, Rutledge GC. Electrospun poly (styrene-block-dimethylsiloxane) block copolymer fibers exhibiting superhydrophobicity. Langmuir 2005;21:5549–54. [60] Ke Q, Fu W, Jin H, Zhang L, Tang T, Zhang J. Fabrication of mechanically robust superhydrophobic surfaces based on silica micro-nanoparticles and polydimethylsiloxane, Surf. Coat Technol 2011;205:4910–4. [61] Jin M, Feng X, Xi J, et al. Super-hydrophobic PDMS surface ultra-low adhesive force. Macromol Rapid Commun 2005;26:1805–9. [62] Lu X, Zhang C, Han Y. Low-density polyethylene superhydrophobic surface by control of its crystallization behavior. Macromol Rapid Commun 2004;25:1606–10. [63] Feng X, Feng L, Jin M, Zhai J, Jiang L, Zhu D. Reversible super-hydrophobicity to super-hydrophilicity transition of aligned ZnO nanorod films. J Am Chem Soc 2004;126:62–3. [64] Shiu J-Y, Kuo C-W, Chen P, Mou C-Y. Fabrication of tunable superhydrophobic surfaces by nanosphere lithography. Chem Mater 2004;16:561–4. [65] Wang S, Feng L, Jiang L. One-step solution-immersion process for the fabrication of stable bionic superhydrophobic surfaces. Adv Mater 2006;18:767–70. [66] Srinivasan S, Praveen VK, Philip R, Ajayaghosh A. Bioinspired superhydrophobic coatings of carbon nanotubes and linear systems based on the ‘bottom-up’ selfassembly approach. Angew Chemie 2008;120(31):5834–8. [67] Meng Q, Hu J. A review of shape memory polymer composites and blends. Compos Part A Appl Sci Manuf 2009;40(11):1661–72.

Dust Repellent Materials

879

[68] Safaee A, Sarkar DK, Farzaneh M. Superhydrophobic properties of silver-coated films on copper surface by galvanic exchange reaction. Appl Surf Sci 2008;254:2493–8. [69] Larmour IA, Bell SEJ, Saunders GC. Remarkably simple fabrication of superhydrophobic surfaces using electroless galvanic deposition. Angew Chemie 2007;119:1740–2. [70] Wang S, Song Y, Jiang L. Microscale and nanoscale hierarchical structured mesh films with superhydrophobic and superoleophilic properties induced by long-chain fatty acids. Nanotechnology 2006;18:15103. [71] Zhai L, Cebeci FC, Cohen RE, Rubner MF. Stable superhydrophobic coatings from polyelectrolyte multilayers. Nano Lett 2004;4:1349–53. [72] Li L, Li Y, Gao S, Koshizaki N. Ordered Co3 O4 hierarchical nanorod arrays: tunable superhydrophilicity without UV irradiation and transition to superhydrophobicity. J Mater Chem 2009;19:8366–71. [73] Cai W, Li Y, Duan G. Wettability and superhydrophobicity of 2-D ordered nano-structured arrays based on colloidal monolayers. J Adhes Sci Technol 2008;22:1949–65. [74] Notsu H, Kubo W, Shitanda I, Tatsuma T. Super-hydrophobic/super-hydrophilic patterning of gold surfaces by photocatalytic lithography. J Mater Chem 2005;15:1523–7. [75] Qian B, Shen Z. Fabrication of superhydrophobic surfaces by dislocation-selective chemical etching on aluminum, copper, and zinc substrates. Langmuir 2005;21:9007–9. [76] Ci L, Vajtai R, Ajayan PM. Vertically aligned large-diameter double-walled carbon nanotube arrays having ultralow density. J Phys Chem C 2007;111:9077–80. [77] Zhang X, Kono H, Liu Z, et al. A transparent and photo-patternable superhydrophobic film. Chem Commun 2007;46:4949–51. [78] Huang Z, Zhu Y, Zhang J, Yin G. Stable biomimetic superhydrophobicity and magnetization film with Cu-ferrite nanorods. J Phys Chem C 2007;111:6821–5. [79] Rao AV, Latthe SS, Nadargi DY, Hirashima H, Ganesan V. Preparation of MTMS based transparent superhydrophobic silica films by sol–gel method. J Colloid Interface Sci 2009;332:484–90. [80] Cho WK, Park S, Jon S, Choi IS. Water-repellent coating: formation of polymeric self-assembled monolayers on nanostructured surfaces. Nanotechnology 2007;18:395602. [81] Wang Z, Lopez C, Hirsa A, Koratkar N. Impact dynamics and rebound of water droplets on superhydrophobic carbon nanotube arrays. Appl Phys Lett 2007;91:23105. [82] Qasem H, Betts TR, Mullejans H, AlBusairi H, Gottschalg R. Effect of dust shading on photovoltaic modules; 2011. [83] Said SAM, Walwil HM. Fundamental studies on dust fouling effects on PV module performance. Sol Energy 2014;107:328–37. [84] Bouaouadja N, Bouzid S, Hamidouche M, Bousbaa C, Madjoubi M. Effects of sandblasting on the efficiencies of solar panels. Appl Energy 2000;65:99–105. [85] Mastekbayeva GA, Kumar S. Effect of dust on the transmittance of low density polyethylene glazing in a tropical climate. Sol Energy 2000;68:135–41. [86] Kazem HA, Chaichan MT. Experimental analysis of the effect of dust’s physical properties on photovoltaic modules in Northern Oman. Sol Energy 2016;139:68–80. [87] Jingshu Wang ZZ, Gong H. Modeling of dust deposition affecting transmittance of PV modules. Jocet 2017;5(3):217–21. [88] El-Shobokshy MS, Hussein FM. Degradation of photovoltaic cell performance due to dust deposition on to its surface. Renew Energy 1993;3:585–90. [89] El-Shobokshy MSS, Mujahid A, Zakzouk aKMKM. Effects of dust on the performance of concentrator photovoltaic cells. IEE Proc I Sol State Electron Dev 1985;132:5–8. [90] Lu H, Lu L, Wang Y. Numerical investigation of dust pollution on a solar photovoltaic (PV) system mounted on an isolated building. Appl Energy 2016;180:27–36. [91] Kaldellis JK, Fragos P, Kapsali M. Systematic experimental study of the pollution deposition impact on the energy yield of photovoltaic installations. Renew Energy 2011;36:2717–24. [92] Qasem H, Betts TR, Müllejans H, AlBusairi H, Gottschalg R. Dust-induced shading on photovoltaic modules. Prog Photovolt Res Appl 2014;22:218–26. [93] Adinoyi MJ, Said SAM. Effect of dust accumulation on the power outputs of solar photovoltaic modules. Renew Energy 2013;60:633–6. [94] O’Hara SL, Clarke ML, Elatrash MS. Field measurements of desert dust deposition in Libya. Atmos Environ 2006;40:3881–97. [95] Hassan G, Yilbas BS, Said SAM, Matin A. Chemo-mechanical characteristics of mud formed from environmental dust particles in humid ambient air. Sci Rep 2016;6. [96] Dastoori K, Al-Shabaan G, Kolhe M, Thompson D, Makin B. Impact of accumulated dust particles’ charge on the photovoltaic module performance. J Electrostat 2016;79:20–4. [97] Yilbas BS, Ali H, Al-Aqeeli N, et al. Characterization of environmental dust in the dammam area and mud after-effects on bisphenol – a polycarbonate sheets. Sci Rep 2016;6:24308. [98] FitzPatrick EA, Fitzpatrick EA. Soil microscopy and micromorphology. Chichester: John Wiley and Sons; 1993. [99] Zhang H, Li X, Du C, Qi H. Corrosion behavior and mechanism of the automotive hot-dip galvanized steel with alkaline mud adhesion. Int J Miner Metall Mater 2009;16:414–21. [100] Giovanoli R, Schnoor JL, Sigg L, Stumm W, Zobrist J. Chemical weathering of crystalline rocks in the catchment area of acidic Ticino lakes, Switzerland. Clays Clay Miner 1988;36:521–9. [101] Yilbas BS. Effect of mud drying temperature on surface characteristics of a polycarbonate PV protective cover. Sol Energy 2017;143:63–72. [102] Yilbas BS, Ali H, Khaled MM, Al-Aqeeli N, Abu-Dheir N, Varanasi KK. Influence of dust and mud on the optical, chemical, and mechanical properties of a PV protective glass. Sci Rep 2015;5:15833. [103] Zhang X, Shi F, Niu J, Jiang Y, Wang Z. Superhydrophobic surfaces: from structural control to functional application. J Mater Chem 2008;18:621–33. [104] Verma LK, Sakhuja M, Son J, et al. Self-cleaning and antireflective packaging glass for solar modules. Renew Energy 2011;36:2489–93. [105] Jing L, Zhi-Jun Z, Ji-Lin Y, Yi-Long B. A thin liquid film and its effects in an atomic force microscopy measurement. Chinese Phys Lett 2009;26:86802. [106] Said SAM, Al-Aqeeli N, Walwil HM. The potential of using textured and anti-reflective coated glasses in minimizing dust fouling. Sol Energy 2015;113:295–302. [107] Duell M, Ebert M, Muller M, et al. Impact of structured glass on light transmission, temperature and power of PV modules. In: 25th European Photovoltaic Solar Energy Conference Exhibition/5th World Conference Photovoltaic Energy Conversion, Valencia, Spain, September, 6–10, 2010. p. 3867–72. [108] Piliougine M, Cañete C, Moreno R, et al. Comparative analysis of energy produced by photovoltaic modules with anti-soiling coated surface in arid climates. Appl Energy 2013;112:626–34. [109] Ming C, Khai B, Chen E, Peng K. Microelectronics reliability humidity study of a-Si PV cell. Microelectron Reliab 2010;50:1871–4. [110] Mohandes BMA, Lamont LA. Application study of 500 W photovoltaic (PV) system in the UAE. Appl Sol Energy 2009;45:242–7. [111] Corn M. The adhesion of solid particles to solid surfaces, I. a review. J Air Pollut Control Assoc 1961;11:523–8. [112] Penney GW, Klinger EH. Contact potentials and the adhesion of dust, ransactions of the american institute of electrical engineers, Part I. Commun Electron 1962;81:200–4. [113] McLean KJ. Cohesion of precipitated dust layer in electrostatic precipitators. J Air Pollut Control Assoc 1977;27:1100–3. [114] Somasundaran P, Lee HK, Shchukin ED, Wang J. Cohesive force apparatus for interactions between particles in surfactant and polymer solutions. Colloids Surf A Physicochem Eng Asp 2005;266:32–7. [115] Kumar A, Staedler T, Jiang X. Journal of colloid and interface science role of relative size of asperities and adhering particles on the adhesion force. J Colloid Interface Sci 2013;409:211–8. [116] Kazmerski LL, Diniz ASAC, Maia CB, et al. Fundamental studies of adhesion of dust to PV module surfaces: chemical and physical relationships at the microscale. IEEE J Photovoltaics 2016;6(3):719–29. [117] Brown K, Narum T, Jing N., Soiling test methods and their use in predicting performance of photovoltaic modules in soiling environments. In: 38th Photovoltaic specialists conference (PVSC), IEEE; 2012. p. 1881–5. [118] Kazmerski LL, Sonia A, Diniz AC, et al. Soiling particle interactions on PV modules: surface and inter–particle adhesion and chemistry effects. Pvsc 2016;2–4. [119] He G, Zhou C, Li Z. Review of self-cleaning method for solar cell array. Procedia Eng 2011;16:640–5. [120] Mazumder MK, Sharma R, Biris AS, Zhang J, Calle C, Zahn M. Self-cleaning transparent dust shields for protecting solar panels and other devices. Sci Technol 2007;25:5–20. [121] Sims RA, Biris AS, Wilson JD, et al. Development of a transparent self-cleaning dust shield for solar panels. In: Proceedings of ESA-IEEE joint meeting on electrostatics; 2003. p. 814. [122] Atten P, Pang HL, Reboud JL. Study of dust removal by standing-wave electric curtain for application to solar cells on Mars. IEEE Trans Ind Appl 2009;45:75–86.

880

Dust Repellent Materials

[123] Sharma R, Wyatt CA, Zhang J, Calle CI, Mardesich N, Mazumder MK. Experimental evaluation and analysis of electrodynamic screen as dust mitigation technology for future Mars missions. IEEE Trans Ind Appl 2009;45:591–6. [124] Bock JP, Robinson JR, Sharma R, and others. An efficient power management approach for self-cleaning solar panels with integrated electrodynamics screens. In: Proceedings of ESA annual meeting on electrostatics; 2008. [125] Biris AS, Saini D, Srirama PK, et al. Electrodynamic removal of contaminant particles and its applications. In: Conference record, 2004 IEEE industry applications conference 2004; 39th IAS Annual Meeting, vol. 2; 2004. p. 1283–6. [126] Hudedmani MG, Joshi G, Umayal RM, Revankar A. A comparative study of dust cleaning methods for the solar PV panels. Adv J Grad Res 2017;1:24–9. [127] Mkoma SL, Maenhaut W, Chi X, Wang W, Raes N. Characterisation of PM 10 atmospheric aerosols for the wet season 2005 at two sites in East Africa. Atmos Environ 2009;43:631–9. [128] Maenhaut W, Raes N, Wang W. Analysis of atmospheric aerosols by particle-induced X-ray emission, instrumental neutron activation analysis, and ion chromatography. Nucl Instrum Meth Phys Res Sect B Beam Interact Mater Atoms 2011;269:2693–8. [129] Baidourela A, Zhayimu K. Patterns of dust retention by urban trees in Oasis cities. Nature 2015;14:53. [130] Mishra SK, Agnihotri R, Yadav PK, et al. Morphology of atmospheric particles over semi-arid region (Jaipur, Rajasthan) of India: Implications for optical properties. Aerosol Air Qual Res 2015;15:974–84. [131] Mill T, Hendry DG. Kinetics and mechanisms of free radical oxidation of alkenes and olefins in the liquid phase. Compr Chem Kinet 1980;1–87. [132] Yilbas BS. Laser treatment of zirconia surface for improved surface hydrophobicity. J Alloys Compd 2015;625:208–15. [133] Sun C-C, Lee S-C, Hwang W-C, Hwang J-S, Tang I-T, Fu Y-S. Surface free energy of alloy nitride coatings deposited using closed field unbalanced magnetron sputter ion plating. Mater Trans 2006;47:2533–9. [134] Dan A, Jyothi J, Chattopadhyay K, Barshilia HC, Basu B. Spectrally selective absorber coating of WAlN/WAlON/Al2O3 for solar thermal applications. Sol Energy Mater Sol Cells 2016;157:716–26. [135] Gao XH, Guo ZM, Geng QF, Ma PJ, Wang AQ, Liu G. Microstructure, chromaticity and thermal stability of SS/TiC–WC/Al2O3 spectrally selective solar absorbers. Sol Energy Mater Sol Cells 2017;164:63–9. [136] Wäckelgard E, Mattsson A, Bartali R, et al. Development of W-SiO2 and Nb-TiO2 solar absorber coatings for combined heat and power systems at intermediate operation temperatures. Sol Energy Mater Sol Cells 2015;133:180–93. [137] Rebouta L, Sousa A, Capela P, et al. Solar selective absorbers based on Al2O3:W cermets and AlSiN/AlSiON layers. Sol Energy Mater Sol Cells 2015;137:93–100. [138] Rephaeli E, Fan S. Tungsten black absorber for solar light with wide angular operation range. Appl Phys Lett 2008;92:211107. [139] Gao X-H, Guo Z-M, Geng Q-F, Ma P-J, Liu G. Enhanced absorptance of surface-textured tungsten thin film for solar absorber. Surf Eng 2016;32:840–5. [140] Cao F, Kraemer D, Sun T, Lan Y, Chen G, Ren Z. Enhanced thermal stability of W-Ni-Al2O3 cermet-based spectrally selective solar absorbers with tungsten infrared reflectors. Adv Energy Mater 2015;5. [141] Chang J-Y, Wang H, Wang L. Tungsten nanowire metamaterials as selective solar thermal absorbers by excitation of magnetic polaritons. J Heat Transfer 2017;139:52401. [142] Ungaro C, Gray SK, Gupta MC. Black tungsten for solar power generation. Appl Phys Lett 2013;103:71105. [143] Vijayakumar P, Pandian MS, Lim SP, et al. Investigations of tungsten carbide nanostructures treated with different temperatures as counter electrodes for dye sensitized solar cells (DSSC) applications. J Mater Sci Mater Electron 2015;26:7977–86. [144] Yamanaka K, Oakamoto H, Kidou H, Kudo T. Peroxotungstic acid coated films for electrochromic display devices. Jpn J Appl Phys 1986;25:1420. [145] Varshney PK, Ramachandran R, Agnihotry SA, et al. Colloidal tungstic oxide films for electrochromic applications: preliminary studies; 1999. [146] Dengel AC, Griffith WP, Powell RD, Skapski AC. Studies on transition-metal peroxo complexes. Part 7. Molybdenum (VI) and tungsten (VI) carboxylato peroxo complexes, and the X-ray crystal structure of K2[MoO(O2)2 (glyc)]2H2O. J Chem Soc Dalt Trans 1987;5:991–5. [147] Borstel G, Krumins A, Millers D. Defects and surface-induced effects in advanced perovskites, vol. 77. Berlin: Springer Science & Business Media; 2012. [148] Deepa M, Srivastava AK, Sood KN, Murugan AV. Nanostructured tungsten oxide-poly (3,4-ethylenedioxythiophene): poly (styrenesulfonate) hybrid films: synthesis, electrochromic response, and durability characteristics. J Electrochem Soc 2008;155:D703–10. [149] Chaudhary V, Kaur A. Enhanced room temperature sulfur dioxide sensing behaviour of in situ polymerized polyaniline–tungsten oxide nanocomposite possessing honeycomb morphology. RSC Adv 2015;5:73535–44. [150] Sharma N, Deepa M, Varshney P, Agnihotry SA. FTIR and absorption edge studies on tungsten oxide based precursor materials synthesized by sol–gel technique. J Non Cryst Solids 2002;306:129–37. [151] Yilbas BS, Yousaf MR, Al-Sharafi A, et al. Silicone oil impregnated nano silica modified glass surface and influence of environmental dust particles on optical transmittance. RSC Adv 2017;7:29762–71. [152] Yilbas BS, Ali H, Khaled M, Al-Aqeeli N, Al-Sulaiman F. Laser gas assisted texturing of alumina surfaces and effects of environmental dry mud solution on surface characteristics. Ceram Int 2016;42:396–404. [153] Wahl JM, Hartnett TM, Goldman LM, Twedt R, Warner C. Recent advances in ALON (TM) optical ceramic. In: Proceedings of SPIE, vol. 5786; 2005. p. 71–82. [154] Wang J, Zhang F, Chen F, et al. Fabrication of aluminum oxynitride (AlON) transparent ceramics with modified gelcasting. J Am Ceram Soc 2014;97:1353–5. [155] Yilbas BS, Matthews A, Karatas C, et al. Laser texturing of plasma electrolytically oxidized aluminum 6061 surfaces for improved hydrophobicity. J Manuf Sci Eng 2014;136:54501. [156] Oosterbeek RN, Ward T, Ashforth S, Bodley O, Rodda AE, Simpson MC. Fast femtosecond laser ablation for efficient cutting of sintered alumina substrates. Opt Lasers Eng 2016;84:105–10. [157] Yilbas BS, Akhtar SS, Karatas C. Laser carbonitriding of alumina surface. Opt Lasers Eng 2011;49:341–50. [158] Cannillo V, Sola A, Barletta M, Gisario A. Surface modification of Al–Al 2 O 3 composites by laser treatment. Opt Lasers Eng 2010;48:1266–77. [159] Zabinski JS, Hu JJ, Bultman JE, Pierce NA, Voevodin AA. Stoichiometry and characterization of aluminum oxynitride thin films grown by ion-beam-assisted pulsed laser deposition. Thin Solid Films 2008;516:6215–9. [160] Yilbas BS, Khaled M, Abu-Dheir N, Aqeeli N, Furquan SZ. Laser texturing of alumina surface for improved hydrophobicity. Appl Surf Sci 2013;286:161–70. [161] Yilbas BS, Bhushan B, Aleem BJA, Gaseem Z. Tribology and superhydrophobicity of laser-controlled-melted alumina surfaces with hard particles. JOM 2014;66:1068–79. [162] Psarski M, Marczak J, Grobelny J, Celichowski G. Superhydrophobic surface by replication of laser micromachined pattern in epoxy/alumina nanoparticle composite. J Nanomater 2014;2014:41. [163] Jagdheesh R. Fabrication of a superhydrophobic Al2O3 surface using picosecond laser pulses. Langmuir 2014;30:12067–73. [164] Zan H-W, Chou C-W, Wang C-H, Yen K-H, Hwang J-C. Pentacene patterning on Aluminum Nitride by water dipping. J Electrochem Soc 2008;155:J321–5.

Relevant Websites http://www.springer.com/gp/book/9781461409250 Springer – Biomimetics in Materials Science. http://eu.wiley.com/WileyCDA/WileyTitle/productCd-1119991773.html; https://www.amazon.com/Self-Cleaning-Materials-Surfaces-Nanotechnology-Approach/dp/1119991773 Wiley – Self-Cleaning Materials and Surfaces: A Nanotechnology Approach.

2.27 CO2 Capturing Materials Tugba Davran-Candan, Yeditepe University, Istanbul, Turkey r 2018 Elsevier Inc. All rights reserved.

2.27.1 Introduction 2.27.2 CO2 Absorption 2.27.2.1 Basic Chemistry of CO2 Capture 2.27.2.2 Conventional Absorption Materials and Their Challenges 2.27.2.3 Emerging Absorption Materials 2.27.2.3.1 Amine blends 2.27.2.3.1.1 N-methyldiethanolamine blends 2.27.2.3.1.2 2-Amino-2-methyl-1-propanol blends 2.27.2.3.1.3 Piperazine blends 2.27.2.3.2 Alkyl amines containing multiple amino groups 2.27.2.3.3 Ionic liquids 2.27.3 Adsorption 2.27.3.1 Physical Adsorbents 2.27.3.2 Chemical Adsorbents 2.27.3.2.1 Support selection and amine functionalization 2.27.3.2.2 Amine-functionalized ordered mesoporous silica sorbents 2.27.3.2.2.1 Mobile Composition of Matter No.41 2.27.3.2.2.2 Santa Barbara amorphous-15 2.27.3.2.2.3 Thermal stability of functionalized mesoporous silica 2.27.3.2.2.4 Effect of moisture on the adsorption capacities of mesoporous silica sorbents 2.27.3.2.3 Amine-functionalized carbonaceous sorbents 2.27.3.2.3.1 Activated carbons 2.27.3.2.3.2 Carbon nanotubes 2.27.4 Illustrative Example: What to and Not to Have for a Feasible Adsorbent? 2.27.4.1 mesoporous silica support-41 or Santa Barbara amorphous-15? 2.27.4.2 Amine Type and Preparation Method 2.27.4.3 Physical Properties of Support 2.27.4.4 Calcination 2.27.4.5 Operational Variables 2.27.5 Future Directions 2.27.6 Closing Remarks Appendix References Relevant Websites

Abbreviations AEAPMS AEAPS AEEA AEPD AHA AHPD AMP AMPD APTES APTMS [bmim] þ CCS CO2-BOL DBU [DCA]

N-(2-aminoethyl)-3aminopropylmethyldiethoxysilane N-[3-(Trimethoxysilyl)propyl] ethylenediamine 2-(2-Aminoethylamino)ethanol 2-Amino-2-ethyl-1,3-propanediol Aprotic heterocyclic anions 2-Amino-2-(hydoxymethyl)-1,3-propanediol 2-Amino-2-methyl-1-propanol 2-Amino-2-methyl-1,3-propanediol (3-Aminopropyl)-triethoxy-silane (3-Aminopropyl)-trimethoxy-silane 1-n-Butyl-3-methylimidazolium Carbon capture and sequestration CO2-binding organic liquids 1,8-Diazabicyclo-[5.4.0]-undec-7-ene Dicyanamide

Comprehensive Energy Systems, Volume 2

DEA DEAPTMS DETA DMAPS EDA ENRTL IGMWNT IPCC MAPS MCM-41 MDEA MEA Met MOF P66614 PCC

doi:10.1016/B978-0-12-809597-3.00255-8

906 883 883 883 885 885 885 886 887 888 889 890 891 891 892 892 892 895 897 899 899 899 900 901 901 901 904 904 904 905 905 906 910 912

Diethanolamine [3-(Diethylamino)propyl]trimethoxysilane Diethylenetriamine (N,N-Dimethylaminopropyl)-trimethoxysilane Ethylenediamine Electrolyte non-random two liquid Industrial grade multiwalled carbon nanotubes Intergovernmental Panel on Climate Change (N-methylaminopropyl)-trimethoxy-silane Mobile Composition of Matter No.41 N-methyldiethanolamine Monoethanolamine Methioninate Metal organic framework Trihexyl(tetradecyl)phosphonium Post-combustion carbon capture

881

882

CO2 Capturing Materials

PEHA PEI PEI-EC PE-MCM-41 PILs Pro PZ RILs

2.27.1

Pentaethylenehexamine Polyethylenimine Ethylenediamine end-capped PEI Pore-expanded MCM-41 Protic ionic liquids Prolinate Piperazine Reversible ionic liquids

SBA-15 TEA TEPA TETA TRI TSIL

Santa Barbara Amorphous type material No.15 Triethanolamine Tetraethylenepentamine Triethylenetetramine 3-[2-(2-Aminoethylamino)ethylamino] propyl trimethoxysilane Task-specific ionic liquids

Introduction

Anthropogenic greenhouse gas emissions have increased dramatically since the beginning of the industrial revolution parallel to the growth observed in population and economy. These unprecedented levels in the atmosphere were accused of being one of the prevailing causes of the global warming experienced since the mid-20th century. Carbon dioxide (CO2) is the main anthropogenic greenhouse gas and its current (July 2016) concentration is measured as B400 ppm in the atmosphere by the Mauna Loa Observatory. According to the fifth assessment report of Intergovernmental Panel on Climate Change (IPCC), continued emissions will cause further warming and long-lasting changes in the climate system, increasing the risk of droughts, floods, heat waves as well as cyclones and wildfires [1]. In the report it was stated that even according to the most stringent mitigation scenarios (which are characterized by 40–70% reduction in the CO2 emissions by 2050 compared to 2010 and almost zero emissions by 2100), the projected level reaches approximately 450 ppm by the year 2100 which corresponds to a 21C increase in the global mean surface temperature relative to preindustrial levels. Hence, it is evident that serious action should be taken to cut off CO2 emissions in order to keep the temperature increase in the allowable limits. According to the statistics provided by the International Energy Agency [2], use of energy through the combustion of fossil fuels constitutes the largest source of anthropogenic CO2 emissions. In 2014, electricity and heat generation were shown to be responsible for 42% of the global emissions associated with fossil fuel combustion, followed by the sectors of transport (23%) and industry (19%). Although employment of non-emitting energy sources such as nuclear, hydropower, or other renewable sources is expanding every day, the share of fossil fuels within the global total primary energy supply is still above 80% and this trend seems to continue in the near future as well. In this manner, it becomes evident that development of technologies that would reduce the CO2 emissions is crucial also for a sustainable energy future, making this field of study closely related to the energy sector. CO2 capture from large point sources such as coal-based power plants and its subsequent sequestration (carbon capture and sequestration – CCS) in depleted oil and gas fields or deep saline aquifer formations have been considered as being a promising way of reducing anthropogenic emissions. CCS process involves three steps, namely, the separation of CO2 from the flue gas, compression, and transportation to the storage site and sequestration. Among these steps, the capture step is the most expensive one making the CCS process still far from being feasible, so development of cost-effective capture methods receives a great deal of attention from the scientific community. There are three different approaches that are in use for the capture of CO2, i.e., precombustion, post-combustion, and oxy-fuel combustion. In the precombustion capture, the primary fuel (such as coal) is partially oxidized in steam and air (or oxygen) mixture to give synthesis gas (syngas), which is primarily made of hydrogen (H2), carbon monoxide (CO), and CO2. The syngas subsequently undergoes the water-gas shift reaction converting CO and water (H2O) to H2 and CO2. Carbon dioxide is then separated from the H2 stream and stored. In the oxy-combustion capture, oxygen stream is used for combustion rather than an air stream. In this way the resulting flue gas mainly consists of CO2 and H2O, making the separation of CO2 easy. Finally, in the postcombustion capture, CO2 is separated from the flue gas that is obtained through the combustion of fuel with an air stream. Although flue gas streams contain higher concentration of CO2 in the pre- and oxy-combustion methods making separation easier, post-combustion method is the most widely used technology in the field due to the very high investment costs of gasification and air/oxygen separation involved in the pre- and oxy-combustion [3]. The most conventional and mature technology that is employed in the post-combustion capture is the chemical absorption of CO2 by using aqueous solutions of amines. Among the amine structures considered, monoethanolamine (MEA) is the most extensively handled structure, being used as the benchmark to compare the performance of newly developed materials. However, despite the high reactivity and kinetics, MEA-containing absorption systems are still far away from being feasible for daily use due to high regeneration costs and there is an extensive on-going research for the development of alternative absorption systems. Another approach that would eliminate the drawbacks associated with the aqueous absorption is the adsorption of CO2 from the flue gas. Development of high-capacity and highly stable adsorbents that would operate under the flue gas conditions is the main objective of the investigations carried out in the field. This chapter aims to provide a deep understanding of the CO2 capturing materials, which is critical for the fulfillment of 2100 emission targets so as to minimize the adverse effects of climate change through the mitigation of CO2 emissions while ensuring a sustainable energy future. An extensive review of the literature about the materials developed for absorption and adsorption of CO2 will be given in Sections 2.27.2 and 2.27.3, respectively; followed by the illustrative example (Section 2.27.4) involving the

CO2 Capturing Materials

883

comparative assessment of various amine-functionalized ordered mesoporous silica sorbents. Finally, future directions will be discussed in Section 2.27.5 and a brief conclusion of the key points will be provided in Section 2.27.6.

2.27.2

CO2 Absorption

Absorption is the intake of a material into the bulk phase of another, which may be physical or chemical. Physical absorption follows Henry’s law and it is temperature and pressure dependent. In chemical absorption, on the other hand, there is a chemical reaction that is taking place between the absorbent and absorber. Among the separation methods regarded for post-combustion CO2 capture, absorption is the most mature and prevalent technology. In a typical post-combustion application (e.g., coal power plant) the flue gas contains 10–15% CO2, 70–75% N2, 8–10% H2O, 3–4% O2, and trace amounts of SOx and NOx [4], which implies that the materials developed for absorption should have strong interactions with CO2 but no interactions with the other components of the flue gas. Accordingly, chemical absorption has attracted more attention than physical absorption in postcombustion carbon capture (PCC) and amino group-containing materials with their high affinity toward CO2 have extensively been under consideration for this purpose. This section contains a detailed review of the amine structures used for this purpose.

2.27.2.1

Basic Chemistry of CO2 Capture

Use of alkyl amines in the removal of acidic gases such as H2S from natural gas and other fuel gases (gas sweetening) is a very common process that has been in use for several decades. Alkyl amines, with their basic character, chemically absorb acidic gases through an acid–base reaction, which means that CO2 – being acidic – can selectively be absorbed from the flue gas by liquid amine solutions. Since the interaction mechanism between CO2 and amines play a critical role in the overall capacity of the absorbents, a good knowledge of the reaction pathway is always required. According to the well-accepted zwitterion mechanism, CO2 is captured in the form of carbamate in the case of primary (R1–NH2) and secondary (R1–R2–NH) amines (Eqs. (1) and (2)) [5,6]. The lone pair on the N atom first attacks the CO2, leading to the formation of the zwitterion. This zwitterionic intermediate is subsequently deprotonated by a base (generally another amine molecule but may also be H2O or OH ) producing the carbamate ion as well as the protonated base molecule. RNH2 þCO2 2 RNHþ 2 COO þ RNHþ 2 COO þ B2 RNHCOO þ BH

ð1Þ ð2Þ

When the amino group is attached to a tertiary carbon atom in the primary, and to a secondary or a tertiary carbon in the secondary amine, the amine structure is said to be sterically hindered and for these structures carbamate formation was shown to be unstable leading to the formation of bicarbonate as the final product [7]. Despite the controversies regarding the bicarbonate formation, CO2 reacts with the amine molecule in the presence of H2O in order to form bicarbonate ion and the protonated base molecule, according to the most prevailing base-catalyzed hydration mechanism (Eq. 3). This reaction pathway was originally proposed for tertiary amines (R1–R2–R3–N), which are known to be incapable of producing carbamates. However, as da Silva and Svendsen [8] later claimed, as long as the base functionality is of appropriate strength and accessible to H2O molecules, basecatalyzed hydration mechanism can explain the bicarbonate formation in aqueous sterically hindered primary/secondary amine solutions as well [9]. RNH2 þ CO2 þ H2 O2 HCO3 þ RNHþ 3

ð3Þ

As the reaction stoichiometry implies, 0.5 and 1 mol of CO2 is captured for each mole of amine group; if the reaction proceeds through carbamate and bicarbonate formation pathways, respectively. Accordingly, bicarbonate formation seems to be more feasible in CO2 capture in terms of equilibrium capacities, yet much slower kinetics of the base-catalyzed CO2 hydration mechanism restricts the use of these amines. Finally, all the chemical reactions of CO2 with amines are reversible permitting the regeneration of amine solutions by heating the CO2-rich solution.

2.27.2.2

Conventional Absorption Materials and Their Challenges

Use of aqueous MEA (see Fig. 1 for the structure) is an effective and economical way of acidic gas removal from natural gas streams, which has been in service for more than 60 years. Although this technology is quite mature and effective for gas treating, improvements are required for the adaptation of the process to large-scale CO2 removal from the flue gas streams. First of all, different from the high-pressure gas treatment process, flue gas from a power plant is available at pressures close to atmospheric values resulting in very low CO2 partial pressures. Consequently, absorption materials with very high affinities are required for the selective absorption of CO2 under low partial pressures. However, the heat of absorption is usually very high for such strong interactions requiring too much energy for the regeneration step, which in turn determines the feasibility of the overall process. Next, the presence of O2 in the flue gas generally causes severe corrosion and solvent degradation problems, whereas SOx reacts irreversibly with MEA leading to the formation of corrosive salts. Finally, compared to the conventional use in gas treatment processes, CO2 capture systems should deal with much larger amounts of gas [10]. Keeping all these challenges in mind, it can be

884

CO2 Capturing Materials

H N

HO

NH2 OH

HO

NH2

HO

DEA

MEA

AMP H2N

N

OH

HO OH

HO MDEA

AMPD OH HN

HO

OH

OH

HO

NH NH2 AHPD

NH2 AEPD

PZ

H N

H N H2N

NH2

HO

NH2 DETA

AEEA

H N

NH2 N H

H2N TETA

H N H2N

H N N H TEPA

NH2

Fig. 1 Molecular structures of the commonly used amines.

concluded that the absorbents developed for CO2 capture should have high absorption capacity at low total and partial pressures, should be tolerant to the presence of O2 and SOx and should not require too much energy for the regeneration step. MEA, being able to remove B90% of the CO2 in the flue gas and producing a B99.95% pure CO2 product (on a dry basis), is regarded as the state-of-the-art solvent in the field. Usually a 30 wt% MEA solution is employed for the absorption of CO2. Inhibitors and other additives are added to the solution in order to prevent corrosion and solvent degradation caused by oxygen, respectively. For maximum efficiency, the absorption column is placed after the desulfurization step and operated at around 401C and ambient pressure, which requires the cooling of the flue gas before the inlet of the absorption column. This cooling is achieved either in the SOx stripper (if it is present) or in a cooler. After the absorption step, the CO2-rich solution is heated in the regenerator where CO2 is stripped from the solution at 100–1201C [10]. These operational conditions were determined through an optimization of the parameters regarding the challenges associated with the capacity, kinetics, regeneration, corrosion, and degradation. Despite the promising nature of the current technology, MEA-based systems are still far away from being used in large-scale CO2 capture due to their high-energy consumption in the regeneration step. To overcome these problems, much of the research is devoted to finding alternative solvents with high capture capacity and stability, fast reaction kinetics, low regeneration energy, and corrosivity. Diethanolamine (DEA) and N-methyldiethanolamine (MDEA) are the two structures (Fig. 1) that have been studied for this purpose. When compared to MEA, DEA and MDEA were found to be less corrosive and less volatile [11,12], making both structures potentially better than MEA in terms of corrosion and solvent loss. DEA – being a secondary amine – is expected to follow the zwitterion mechanism as MEA (primary amine) does, whereas MDEA – being a tertiary amine – is supposed to capture CO2 in the form of bicarbonate (base-catalyzed hydration mechanism). Hence, from the equilibrium CO2 loading point of view, reaction stoichiometry implies that MDEA should absorb higher amounts of CO2 than MEA and DEA can do. However,

CO2 Capturing Materials

885

experimental studies demonstrated that absorption capacity followed the order: MEA–DEA 4 MDEA, stemming from the very slow reaction kinetics of tertiary amines [11–13]. On the other hand, energy required for the regeneration of the solvent plays a very critical role in the determination of the solvents’ feasibility and this energy is the sum of the energies required to heat CO2-rich solution to regeneration temperature (sensible heat), vaporize the solvent (heat of vaporization), and desorb CO2 (heat of reaction). In this manner, the stability of the reaction product (heat of reaction) is one of the main factors contributing to the regeneration costs and the energy requirements for desorption of CO2 of the three solvents were reported to be in the order MEA 4 DEA 4 MDEA [14]. Correspondingly, it was obvious that among the structures considered no single amine (DEA or MDEA) was superior to MEA, both having their own advantages as well as disadvantages. Sterically hindered amines came into view at this point as potential alternatives to the existing solvent systems. Sartori and Savage [7], in their pioneering work, defined the steric hindrance in primary amines as the presence of an amino group attached to a tertiary carbon and showed that 2-amino-2-methyl-1-propanol (AMP), being the sterically hindered version of MEA (see Fig. 1), had superior CO2 solubility with respect to MEA, which was later confirmed by other experiments as well [15,16]. AMP is one of the most extensively studied hindered amines in the literature and the increase observed in the equilibrium capacity of this structure was associated with the reduced carbamate stability due to the bulk substitutions at the a-carbon [7]. When carbamate loses stability, bicarbonate dominates the reaction product as in the case of tertiary amines, resulting in higher equilibrium solubility due to the reaction stoichiometry and lower regeneration energies due to the reduced product stability. Since reaction kinetics plays a vital role in the realization of high equilibrium loadings, absorption kinetics of CO2 into aqueous AMP solutions were investigated broadly. Even though the second-order rate constants reported by early kinetic studies (under the assumption that bicarbonate formation in AMP systems took place through the zwitterionic intermediate) were much higher than those of MDEA, they were still much lower than the values estimated for MEA [17,18]. Other sterically hindered amines such as 2-amino-2-methyl-1,3-propanediol (AMPD), 2-amino-2-ethyl-1,3-propanediol (AEPD), and 2-amino-2-(hydoxymethyl)-1,3-propanediol (AHPD) were also studied (see Fig. 1 for the structures). These structures were derived from AMP by substituting OH groups, the steric hindrance increasing in the order AMP o AMPD o AHPD o AEPD. At 401C, for 10 wt% solutions, all three structures had lower CO2 solubilities compared to MEA at lower partial pressures, whereas the situation is completely reversed above 50 kPa, typical to that of physical absorbents [19]. Their CO2 absorption rates were further measured and modeled at various temperatures close to 401C under the assumption that these sterically hindered structures reacted with CO2 also through the zwitterionic intermediate similar to AMP. Measurements demonstrated that their reaction kinetics were much slower with respect to MEA kinetics, following the order AEPD o AHPD o AMPD o AMP [18,20–22]. This ordering was indeed just the opposite of that of the steric hindrance given above, indicating that reduced hindrance led to a more pronounced reaction rate [22].

2.27.2.3 2.27.2.3.1

Emerging Absorption Materials Amine blends

Early investigations manifested the truth that primary and secondary amines were advantageous with respect to reaction kinetics but they had low equilibrium loading capacities and very high regeneration energies designated by the reaction stoichiometry and the product (carbamate) stability, respectively. Tertiary and sterically hindered amines, on the other hand, had higher loadings and lower product stabilities but slower reaction kinetics. At this point, the idea of combining the advantageous properties of the two groups by mixing primary/secondary amines with tertiary/sterically hindered ones has emerged as a promising solution and various amine blends have been under inquiry for this purpose since 1990s. 2.27.2.3.1.1 N-methyldiethanolamine blends Shen and Li examined the solubility of CO2 in aqueous mixtures of MEA (primary) and MDEA (tertiary) at various temperatures and CO2 partial pressures between 1 and 2000 kPa [23]. At 401C, equilibrium loading of the 30 wt% MDEA solution was reported to be the highest above 17 kPa, followed by that of the 12 wt% MEA þ 18 wt% MDEA, 24 wt% MEA þ 6 wt% MDEA, and 30 wt% MEA, respectively; indicating a decrease in the solubility with an increase in the MEA concentration. Dawodu and Meisen, however, observed that the equilibrium loadings rose with increasing MEA concentrations from 0.8 to 2.1 M in the solution at 701C and low pressures (below B200 kPa) for solutions of equal molarity and argued that CO2 absorption was restricted by the reaction stoichiometry of the carbamate formation reaction at higher pressures [24]. The authors also examined DEA þ MDEA blends and showed that at 701C the equilibrium solubility is unaffected by the choice of MEA or DEA, while MEA þ MDEA solution had higher solubility compared to that of DEA þ MDEA at higher temperatures, which may be indicative of a lower regeneration energy for the latter solution [24]. The kinetics of CO2 absorption in MEA/DEA þ MDEA blends have also been studied extensively, much more extensively indeed, due to the importance of kinetics (together with the mass transfer coefficient) in the determination of the column height. Early studies investigating the absorption of CO2 reported good agreement between experimental and predicted results by making use of the shuttle mechanism [25], parallel-reversible-chemical-reaction model according to both the film and penetration theories under the assumption that tertiary alkanolamine completely deprotonates the primary alkanolamine [26], modified pseudo-firstorder model based on the film theory with the assumption of shuttle mechanism [27] and Higbie’s penetration theory with the assumption of reversible reactions [28]. Experiments performed with aqueous solutions containing 1.5 wt% MEA þ 28.5 wt% MDEA, 3 wt% MEA þ 27 wt% MDEA, and 4.5 wt% MEA þ 25.5 wt% MDEA at atmospheric pressure and 401C pointed out that

886

CO2 Capturing Materials

addition of small amounts MEA to an aqueous MDEA solution enhanced the rate of absorption significantly, the effect being much pronounced for shorter contact times and higher MEA concentrations [28]. Similar results were obtained for solutions containing various concentrations of MEA (0.1, 0.2, 0.3, 0.4, and 0.5 kmol/m3) mixed with aqueous MDEA solutions of 1 and 1.5 kmol/m3 concentrations at 30, 35, and 401C by using a hybrid reaction rate model, a zwitterion mechanism for MEA and a pseudo-first-order reaction model for MDEA [29]. Ramachandran et al. later used the same hybrid reaction model to study the kinetics of CO2 absorption into aqueous blends of MEA þ MDEA at various concentrations at temperatures 25, 30, 40, 50 and 601C but reported that zwitterion and termolecular mechanisms could not predict the rate constants in their original forms [30]. Recently, Sema et al. performed a comprehensive mass transfer and reaction kinetics study of CO2 absorption into aqueous blends of 3 wt% MEA þ 27 wt% MDEA, 5 wt% MEA þ 25 wt% MDEA, and 7 wt% MEA þ 23 wt% MDEA at atmospheric pressure and over a temperature range 25–601C [31]. By using a second-order zwitterion mechanism for MEA and the base-catalyzed hydration (termolecular) mechanism for MDEA, they reported an increase in the enhancement factor (absorption rate) with an increase in MEA concentration. Lin et al. carried out a similar kinetic examination to that of Liao et al. [29] by using aqueous blends of DEA (0.1, 0.2, 0.3 and 0.4 kmol/m3) þ MDEA (1 and 1.5 kmol/m3) and reported significant enhancements in the CO2 absorption rate with the addition of small amounts of DEA [32]. The performance of the aqueous MEA þ MDEA and DEA þ MDEA blends has also been evaluated with respect to their regeneration energy requirements. Idem et al. examined the CO2 capture performance of single amine (MEA) and mixed-amine (MEA/MDEA ¼ 4:1) solvents of 5 kmol/m3 total amine concentration in terms of various criteria including the heat requirement for solvent regeneration under realistic feed conditions and argued that the heat penalty for solvent regeneration was reduced significantly when an amine blend was used instead of a single amine [33]. Experiments implemented in a bench-scale gas stripping and solvent regeneration system with aqueous single (MEA, DEA, and MDEA) and blended (MEA þ MDEA and DEA þ MDEA) alkanolamine solutions at 4.0 kmol/m3 total amine concentration, on the other hand, exhibited that the reboiler heat duty of the single amine solutions followed the order MEA 4 DEA 4 MDEA, while the heat duties of the blended amines were in between those of the parent alkanolamines [14]. The magnitude of the reduction in regeneration heats was found to be proportional to the concentration of the tertiary amine in the solution but the relation was found to be nonlinear, stemming from the fact that heat of vaporization was more dominant in the determination of heat duty rather than the heats of reaction. 2.27.2.3.1.2 2-Amino-2-methyl-1-propanol blends As discussed in the previous section, even though AMP was shown to capture CO2 much faster than MDEA, rate of absorption was still much slower in aqueous AMP solutions compared to that in MEA solutions [17,18]. This has led to an extensive investigation of AMP blends with primary or secondary amines that have much faster reaction kinetics in the literature. Equilibrium solubility of CO2 was estimated by making use of the modified Kent–Eisenberg model in aqueous MEA þ AMP and DEA þ AMP solutions. Li and Chang [34] performed experiments with various MEA þ AMP blends of total amine concentration of 30 wt% at CO2 partial pressures ranging from 1.0 to 200 kPa and demonstrated that equilibrium loading increased with increasing AMP concentrations in the solution at 601C, above B7 kPa partial pressure. The trend was exactly the opposite (loading decreased with increasing AMP concentrations) below that partial pressure, resulting in a crossing of the equilibrium solubility curves of AMP and MEA. Similar observations were later also reported by Park et al. [16] at 40, 60, and 801C and at partial pressures of CO2 ranging from 0.1 to 50 psia. The intersection of the two solubility curves (AMP and MEA) was shown to appear at a higher partial pressure as the temperature increased from 40 to 801C. For DEA þ AMP blends with 30 wt% of total amine concentrations, on the other hand, the same trend that the CO2 solubility raised with increasing AMP concentrations in the solution was discovered, except for the fact that DEA was not as reactive as MEA at lower partial pressures. In MEA þ AMP blends, although MEA reacted much better than AMP at lower partial pressures, the stoichiometry limited the amount of CO2 absorbed in the solution and AMP began to take an active role in the absorption as partial pressure increased. In the case of DEA, however, CO2 solubility at low partial pressures was quite small, similar to that in AMP and as a result, addition of DEA into aqueous AMP solutions did not lead to any enhancements in the CO2 solubility at lower partial pressures [16]. Rate of CO2 absorption into aqueous blends of MEA þ AMP [35] and DEA þ AMP [36] was measured at 30, 35, and 401C using a laboratory wetted wall column. In both studies, aqueous solutions of eight different concentrations were studied by using a hybrid reaction rate model. This model consisted of a first-order reaction mechanism for MEA and the zwitterion mechanism for AMP for the MEA þ AMP systems, whereas reaction rate of DEA and AMP were modeled by making use of the zwitterion and a second-order mechanism, respectively, in DEA þ AMP systems. Both measurements indicated that addition of small amounts of MEA or DEA into aqueous solutions of AMP led to significant improvements in the absorption rates. Mandal et al. reported comparable results that were obtained by using equilibrium-mass transfer-reaction kinetics-based combined model, which was developed according to Higbie’s penetration theory, for both blends [37,38]. A comparative investigation performed by the same group with MEA þ AMP þ H2O and MEA þ MDEA þ H2O solutions of 30 wt% total amine concentration exhibited that addition of MEA to both solutions raised the CO2 absorption rate but the enhancement observed in MEA þ AMP þ H2O blend was more significant than that of MEA þ MDEA þ H2O [39]. Reboiler heat duties calculated for the MEA þ AMP þ H2O solutions of 4.0 kmol/m3 total amine concentration were shown to be between those of the parent amines, as it was in the case of MDEA blends [14]. The decline observed in the heat requirement with increasing AMP concentrations in the solution mixture was quite significant up to an AMP concentration of 2.7 kmol/m3. After this point, the change in the reboiler duty became negligible implying the fact that heat of reaction was not the only component determining the energy requirement for regeneration.

CO2 Capturing Materials

887

2.27.2.3.1.3 Piperazine blends Piperazine (PZ) – initially used as an effective activator for the other solvents – is a cyclic diamine (Fig. 1) that has attracted growing attention for CO2 capture due to its higher reaction kinetics, better thermal and oxidative degradation properties, slightly lower volatility, higher loading capacity, significantly lower corrosiveness to stainless steel, and better regeneration performance compared to 30 wt% aqueous MEA solutions [13,40]. However, limited water solubility of PZ resulted in very narrow operating conditions for CO2 absorption, which in turn popularized its use as a component of an amine blend rather than a single amine solution [40]. Since tertiary and sterically hindered amines were shown to have very slow reaction kinetics, PZ is usually added to aqueous MDEA or AMP solutions in order to eliminate the disadvantages associated with kinetics. Early VLE data corresponding to aqueous MDEA þ PZ solutions with 4.28 kmol/m3 MDEA and 0–0.515 kmol/m3 PZ at 343K revealed that increasing concentrations of PZ enhanced the CO2 solubility in the solution [41]. Bishnoi and Rochelle later claimed that these results were obtained under conditions where the ratio of total CO2 to PZ is close to or greater than 1 (corresponding to high loading region) and the equilibrium partial pressure of CO2 over MDEA þ PZ blend did not differ significantly from that in MDEA alone under these conditions [42]. They conducted experiments with 4 M MDEA þ 0.6 M PZ solutions at 40 and 701C in the low loading region and developed a model that would successfully predict their own data together with those reported before in the high loading region (those in Ref. [41]). Their results expressed that PZ had a serious effect on the CO2 solubility only when the ratio of total CO2 to PZ is less than 1. In this region, PZ-activated MDEA solutions were found to possess higher CO2 solubility with respect to DEA þ MDEA but lower solubility compared to MEA þ MDEA blends. Absorption studies performed more recently with various amine blends demonstrated that 20 wt% PZ þ 10 wt% MDEA blend had a much higher absorption capacity than 40 wt% DEA þ 10 wt% MDEA and 40 wt% MEA þ 10 wt% MDEA at 401C and 12 kPa CO2 pressure [43]. Xu et al. explored the kinetics of CO2 absorption into aqueous blends of MDEA (1.75–4.21 kmol/m3) þ PZ (0.041–0.21 kmol/ 3 m ) at a temperature range 30–701C using a model that was based on a rapid pseudo-first-order reversible reaction mechanism between CO2 and PZ parallel to the reaction between CO2 and MDEA [44]. By making a comparison of the absorption rate coefficients of MDEA solutions with and without PZ, they concluded that PZ accelerated the CO2 absorption rate in the solution. However, it was later declared that due to the very small concentrations of PZ involved in Xu et al.’s experiments, the enhancement factor calculated for MDEA þ PZ blends did not indeed differ much from that of 4.21 M MDEA solution [45]. Besides, since the absorption amount was small, their data were argued to be prone to large errors. To overcome these shortcomings, Bishnoi and Rochelle performed experiments with 4 M MDEA þ 0.6 M PZ blends between 22 and 701C and simulated their data by making use of a model based on chemical reactions and transport effects with the eddy diffusivity theory [45]. Their results showed that addition of PZ into aqueous solutions of MDEA improved the absorption kinetics significantly even at high loadings. Comparison made to MEA þ MDEA and DEA þ MDEA blends, on the other hand, proved that PZ was much more effective than MEA and DEA, especially at low loadings. Heat of absorption of MDEA þ PZ blends were investigated at low loadings and temperatures between 35 and 651C, by using the reaction calorimetry. For solutions of various concentrations it was realized that addition of PZ boosted the heat of absorption, the effect being much more pronounced at lower loadings [46]. For the aqueous PZ þ AMP blends, on the other hand, Yang et al. were the first to report equilibrium solubility data [47]. They performed experiments in a vapor-recirculation equilibrium cell at temperatures 40, 60, and 801C along with pressures up to 152 kPa with solutions of six different concentrations. Their experimental data was modeled using the modified Kent–Eisenberg model without taking the presence of all piperazine species in the liquid phase (such as piperazine carbamate, piperazine dicarbamate, and protonated piperazine carbamate) into consideration and good agreement was obtained between the calculated and measured values. Furthermore, equilibrium CO2 loading was displayed to be higher for the blends in which the concentration of PZ was greater. However, this enhancement provided by PZ was valid only in the low to moderate loading region, the situation becoming just the opposite above a CO2 loading of approximately 0.75 mol/mol amine at 601C. Following this work, Dash et al. studied the VLE of aqueous AMP þ PZ blends at a temperature and pressure range 25–551C and 0.1–1450 kPa, respectively, in 28 wt% AMP þ 2 wt% PZ, 25 wt% AMP þ 5 wt% PZ, and 22 wt% AMP þ 8 wt% PZ solutions [48]. Different from Yang et al., their model was developed based on the electrolyte nonrandom two liquid (ENRTL) theory and involved the interactions of the piperazine species in the liquid phase. From their plot of equilibrium CO2 partial pressures against CO2 loading for different compositions at 551C, it was obvious that increasing concentrations of PZ improved the solubility up to a loading of 0.75 mol/mol amine. Subsequently, the same group executed a similar VLE study by using AMP þ PZ solutions of 40 and 50 wt% total amine concentrations and observed comparable results with the previous one along with the solid precipitation in 50 wt% blend below 451C [49]. For similar total amine concentrations of 30 wt%, AMP þ PZ blends were shown to possess a slightly higher absorption capacity compared to MDEA þ PZ blends for PZ concentrations of 20 and 10 wt% at 401C and 12 kPa CO2 partial pressure [43]. One of the earliest studies inspecting the effect of PZ on the kinetics of CO2 absorption in aqueous PZ þ AMP blends was performed by Seo and Hong [50]. Absorption rates in 0.55–3.35 kmol/m3 AMP þ 0.058, 0.115, and 0.233 kmol/m3 PZ were measured at a temperature range 30–401C by employing a wetted-sphere absorption apparatus. Their results claimed that addition of even small amounts of PZ into aqueous solutions of AMP promoted the apparent reaction rate due to PZ’s contribution to the zwitterion deprotonation and its direct reaction with CO2, the effect being less obvious at higher AMP concentrations. However, it was later claimed that these results were prone to large measurement errors since the experiments were performed at relatively higher partial pressures of CO2, resulting in almost total depletion of PZ at the gas–liquid interface [51]. Similar measurements were subsequently performed over a temperature and CO2 partial pressure range 30–401C and 2.63–4.55 kPa, respectively, in

888

CO2 Capturing Materials

solutions containing 1.0–1.5 kmol/m3 AMP and 0.1–0.4 kmol/m3 PZ. Identical conclusions were arrived with those of Seo and Hong by making use of a hybrid reaction rate model, a second-order reaction for CO2-PZ and a zwitterion mechanism for CO2AMP [52]. Finally, Samanta and Bandyopadhyay modeled the absorption rate data obtained at 25–401C and 2–14 kPa CO2 partial pressure with solutions of 30 wt% total amine concentrations (PZ concentration ranging from 2 to 8 wt%), by making use of a comprehensive coupled mass transfer-reaction kinetics-equilibrium model [51]. Likewise, they observed that addition of small amounts of PZ enhanced the kinetics of absorption significantly under the conditions employed. Comparison of the CO2 absorption rate in 10 wt% MDEA þ 20 wt% PZ to that of 10 wt% AMP þ 20 wt% PZ, on the other hand, demonstrated that the two values were close to each other [43]. Dash et al. reported the average heat of absorption of CO2 into aqueous blends of 22 wt% AMP þ 8 wt% PZ over 40–701C as 70 kJ/mol, claiming a 15% decrease with respect to that in aqueous PZ solutions [48]. Despite the blends of AMP þ PZ and MDEA þ PZ have more frequently been investigated, aqueous mixtures of MEA þ PZ have also attracted attention in the field. Dang and Rochelle studied the CO2 solubility and absorption rate in aqueous MEA þ PZ blends with total amine concentrations ranging from 1.0 to 5.0 M, at temperatures 40 and 601C [53]. For blends containing 0.6–1.2 M PZ, equilibrium partial pressures of CO2 was found to be 2–5 times smaller than that of single MEA solutions at high loadings (40.4–0.5 mol CO2/mol amine) indicating an increase in the equilibrium loading with rising PZ concentration, whereas PZ had no significant effect on equilibrium CO2 partial pressures at low loadings (o0.2–0.3 mol CO2/mol amine). Later, experiments carried out with various blends of total amine concentrations of 30 and 50 wt% at 401C temperature and 100 kPa total pressure (CO2 partial pressure ¼ 12 kPa) evidenced that equilibrium absorption capacity of aqueous MEA þ PZ solutions were in between the equilibrium capacities of the parent solvents (i.e., 0.50 mol CO2/mol amine for MEA and 1.06 mol CO2/mol amine) and the loading ascended with a rise in the concentration of PZ in the solution [43]. CO2 absorption rates in MEA þ PZ blends with 0.6–1.2 M PZ in the solution were reported to be 1.5–2.5 times greater compared to that of single MEA solutions, the enhancement factor being much larger in lean solutions for both aqueous MEA and MEA þ PZ, which was attributed to the depletion of MEA and PZ at higher CO2 loadings [53]. For loadings less than 0.4 mol CO2/mol amine, PZ species were argued to contribute more than 60% of the total absorption rate, while that of MEA was limited to 20–30% for the whole range of loading. Similar findings that the addition of PZ increased the absorption rate of CO2 in aqueous MEA solutions were also declared by others [43].

2.27.2.3.2

Alkyl amines containing multiple amino groups

Since the reactive part of all the amines that have been considered for CO2 absorption is the amino group, use of structures with multiple amino groups has come out as a promising way of increasing the capacity more recently which might in turn lead to the development of carbon capture materials for commercial use. 2-(2-Aminoethylamino)ethanol (AEEA) is an alkanolamine with two amino groups (Fig. 1), one being primary and the other secondary. In a comparative study involving various amine solutions of 5 wt% concentration, absorption and desorption of CO2 was explored at 231C and the boiling temperatures of each solution, respectively [54]. Among the solutions inspected, AEEA demonstrated the highest CO2 loading (1.348 mol/mol amine) after the polyamine, the amount being much higher than that of MEA (0.813 mol/mol amine) under the same conditions, and a much better regeneration capacity compared to the MEA solution. The same group later conducted experiments with AEEA þ MDEA blends as well and concluded that while the absorption capacity was slightly enhanced in the presence of MDEA, its desorption was significantly improved [55]. Subsequently, Ma’mun et al. performed a screening study of various single and mixed amine solutions at 401C, based on their CO2 absorption rates as a function of loading [56]. Their results implied that 30 wt% AEEA solution had a much faster absorption rate than MEA solution of the same concentration above a CO2 loading of B0.20 mol/mol amine along with higher absorption capacity, lower vapor pressure and higher maximum net cyclic capacity. When it was used in a MDEA blend (2.6 M MDEA þ 0.62 M activator), its performance was observed to be inferior to that of PZ but superior to that of MEA at all CO2 loadings. Lepaumier et al. measured the degradation rates of 4 mol/kg AEEA solutions in the presence and absence of O2 and CO2, separately at 2 MPa and 1401C [57,58]. They estimated that its thermal and oxidative degradation rates were much lower than those of MEA. However, AEEA was reported as being the least stable amine structure in the presence of CO2 under the conditions employed. Finally, heats of absorption of CO2 in MEA and AEEA solutions were studied in a reaction calorimeter at a temperature range 40–1201C and the two values were concluded to be similar [59]. All these findings, together with the high corrosiveness of AEEA [56], state that in spite of its better absorption and regeneration properties, its use may still be limited due to its high heat of absorption and corrosiveness. Diethylenetriamine (DETA) is a triamine containing two primary and one secondary amino group (Fig. 1). Absorption studies performed with aqueous DETA solutions of 30 wt% concentration at 40, 60, and 801C and 112.78 kPa demonstrated that at all temperatures, capture capacity of DETA was much higher with respect to the loadings obtained in aqueous MEA and DEA solutions of the same concentration, stemming from the fact that there were more reactive sites (amino groups) in the structure of DETA [60]. When the concentration of DETA solution increased from 1.0 to 2.9 kmol/m3 at a temperature range 25–601C; however, there was a decline in the CO2 loading at all temperatures studied [61]. Choi et al. employed DETA as an activator in aqueous MDEA blends of 30 wt% total amine concentration, at temperatures of 40, 60, and 801C [62]. They observed that the presence of 10 wt% DETA in the solution enhanced the CO2 loading significantly, at all the temperatures studied. Kinetics of the DETA solutions has also been investigated extensively along with the absorption studies. Hartono et al. used the pseudo-first-order assumption that was validated through the penetration theory in order to model their experimental data obtained in aqueous

CO2 Capturing Materials

889

solutions of DETA with 1.0, 1.5, 2.0, 2.5, and 2.9 kmol/m3 concentrations at a temperature range 25–601C [61]. They concluded that the reaction was much faster in solutions with higher concentrations and both the termolecular and the zwitterion mechanisms could successfully explain the kinetic data. In a complementary study of the same group, the relative effects of the primary and secondary amino groups on the kinetics of absorption were explored [63]. Since the main contribution to the CO2 absorption at low loadings belonged to the primary amino groups in the DETA structure [61], the reaction rates of the primary groups were measured at low loadings, while those of the secondary groups were obtained in the presence of H2SO4 that was employed to neutralize the primary groups. Their measurements indicated that the primary amino groups absorbed CO2 much faster than the secondary ones. When used as an activator in 30 wt% MDEA solutions, on the other hand, DETA could successfully promote the absorption flux at 401C [62]. To complete the evaluation of DETA as a potential capture material, heat duty of regeneration (or heat of absorption) was often studied, as well. At 251C, the heat of CO2 absorption was estimated to be quite similar in 30 wt% DETA and MEA solutions [60]. When used in MDEA solutions, however, although the calculated heat of absorption was much higher than that of the single MDEA solution, MDEA þ DETA solution had considerably lower heat of absorption with respect to the 30 wt% MEA solution [62]. Triethylenetetramine (TETA) and tetraethylenepentamine (TEPA) are the two structes that are also considered for CO2 capture. TETA has four (2 primary, 2 secondary), TEPA has 5 (2 primary, 3 secondary) amino groups (see Fig. 1). Comparative studies performed with aqueous DETA, TETA, and TEPA solutions of 30 wt% total concentration at 40, 60, and 801C demonstrated that CO2 loading increased in the order of increasing amino groups, all being much larger than that of the 30 wt% MEA solution [60]. When the heats of absorption were taken into consideration, the trend was observed to be just the opposite. As the number of secondary amino groups escalated in the structure, secondary carbamates were observed in the solution along with the primary ones and this led to a decrease in the heat of absorption [60]. Similar findings that absorption capacity was improved with the number of amino groups in the activators was also reported for MDEA þ DETA/TETA/TEPA solutions of 30 wt% total amine concentration, MDEA þ TEPA possessing the highest loading amount [62]. Despite the heats of absorption of all three blends were significantly higher than that of single MDEA, they seemed to be much more feasible compared to 30 wt% MEA solution. Finally, all MDEA blends showed higher absorption fluxes with respect to single MDEA but there was a slight decrease in the overall mass transfer coefficient in the order DETA 4 TETA 4 TEPA. Even though the molecules discussed above were found to be advantageous in terms of reaction kinetics and absorption capacity, high degradation rates of these structures still limits their use for CO2 capture in single amine solutions but their adoption as activators in tertiary amine solutions attracts great attention in the community [64].

2.27.2.3.3

Ionic liquids

Ionic liquids (molten salts) are defined as being salts which have melting points less than 1001C and consist exclusively of ions. Due to their negligible vapor pressures, wide range of solubilities and miscibilities along with generally nonflammable and thermally stable character, they are regarded as being novel solvents for green chemistry. Apart from these, ionic liquids – due to their structure consisting of ions – have dual functionalities, which can be tailored to a great extent, by varying the anionic or the cationic group [65]. After the discovery that significant amounts of CO2 dissolved in imidazolium-based ionic liquids to facilitate the extraction of dissolved product [66], these structures have attracted increasing attention as potential CO2 capture materials. Although there is a vast number of possibilities of structures (anions and cations) that can be used to form the ionic liquid, imidazolium, pyrollidinium, pyridinium, guanidinium, phosphonium, morpholinium, piperidinium, sulfonium, and ammonium are the most commonly employed cations (see Fig. 2), while hexafluoophosphate, tetrafluoroborate, alkylsulphate triflate, dicyanamide, bis (trifluoromethylsulfonyl)imide are the most familiar anionic groups (see Fig. 2) encountered in ionic liquids for CO2 capture [67]. Early studies concentrated on the investigation of materials that can only physically absorb (without a chemical reaction) CO2. Despite the initial claims that anionic groups dominated the dissolution process, it was later concluded that solubility was indeed highly affected by the entropic effects rather than the solute–solvent interactions and it was improved with increasing molecular weight, molar volume as well as the free volume [67]. Nonetheless, solute–solvent interactions should not be regarded as being completely ineffective because fluor-containing ionic liquids had much higher CO2 solubilities compared to the non-fluorinated ones. N2 solubility, on the other hand, was shown to be much smaller than that of CO2 leading to high CO2/N2 selectivities of the ionic liquid structures considered [67]. Notwithstanding their promising solubilities and selectivities, many ionic liquids were reported to be highly viscous; viscosity getting higher for the structures involving longer alkyl chains. This may, though, be eliminated or at least be lessened with a proper design of the ionic liquid structure [67]. Although the physical ionic liquids could absorb and separate CO2 to a certain level, their performance was not high enough to employ them in large-scale capture processes. This led to the idea of functionalization of the cationic or anionic groups in the ionic liquids with amines so as to enhance CO2 absorption through chemical interactions, resulting in the structures known as task-specific ionic liquids (TSIL). Bates et al. were the first to tether a primary amino moiety covalently to an imidazolium cation [68]. It was observed that the molar uptake of CO2 by the amine-incorporated TSIL at 221C approached to 0.5 mol/mol amine, which was the theoretical maximum achieved by the primary amines through the zwitterion mechanism. Along with this ratio, NMR and FTIR spectra that were obtained in the presence and absence of CO2 also verified that absorption was chemical rather than physical. Similar to the conventional amines, the process was reversible and the ionic liquid could be regenerated upon heating to 80–1001C. Following this, Sanchez et al. studied CO2 absorption in primary and tertiary-amine-functionalized 1-n-butyl-3-methylimidazolium tetrafluoroborate ([bmim] þ [BF4] ) and 1-n-butyl-3-methylimidazolium dicyanamide ([bmim] þ [DCA] ) [69]. Significant enhancement in CO2 absorption was reported for both ionic liquids upon amine functionalization at 301C and a pressure up to 1 MPa. The effect of amine incorporation was much pronounced for [bmim] þ [BF4] , although the absorption capacities of

890

CO2 Capturing Materials

R4

R2

N

N

+

N

R1

+

Pyrollidinium

R6

O

C

R1

N

R2

R3 Guanidium

R2

F B

−

Morpholinium

F

Tetrafluoroborate

−

O

O

R1

Alkylsulphate

Sulfonium F

F C

S

C S

O O

O

F

N

R1

R2

F −

+

F

O− O

F

O Triflate

Bis(trifluoromethylsulfonyl)imide F

C N

R1

S

Piperidinium

S O

R1

+

R2

C F

Ammonium −

O

N

R3 R2

F

R3 R2

N S

F

+

R1

O

R1

F

N

+

R3

+

Phosphonium

R4

N

R4 N

+

Pyridinium

R1 +

F

N

R2

Imidazolium

R5 N

P

R1

F

−

F

P

C N

Dicyanamide

F F Hexafluorophosphate

−

Br

−

Cl

F

Bromide

Chloride

Fig. 2 Molecular structures of the commonly used anions and cations in the formation of ionic liquids.

nonfunctionalized forms of both ionic liquids were almost the same. Besides, tethering of primary amines were found to be more effective than the tertiary ones in boosting the absorption capacities, in line with the literature data claiming that tertiary amines were less reactive than the primary amines. Theoretical calculations later demonstrated that a more favorable reaction stoichiometry (approaching to 1 mol CO2/mol amine) could be possible by functionalizing the cationic group rather than the anion since tethering the amine to the cation favored the carbamic acid production more readily compared to the carbamate formation [70], which was also verified by experiments performed with phosphonium-based amino acid ionic liquids [71]. Gurkan et al. measured the heats of CO2 absorption in trihexyl(tetradecyl)phosphonium prolinate ([P66614][Pro]) and trihexyl (tetradecyl)phosphonium methioninate ([P66614][Met]) at 251C and 2–3 bar as 80 and 64 kJ/mol, respectively [71]. These numbers were close to the values calculated for MEA ( 85 kJ/mol) [67] and indicated that regeneration energies were high. At this point, exploitation of aprotic heterocyclic anion (AHAs) has emerged as a viable approach due to their 1:1 reaction stoichiometry with CO2 in addition to their highly tunable character. Gurkan et al. discovered that 3-substituted cyanopyrrolide had 39 kJ/mol less absorption enthalpy compared to the unsubstituted pyrrolide, while substitution in position 2 decreased the reaction enthalpy further [72]. With the experimental results complementing these findings, AHA-based ionic liquids have appeared as a new class of ionic liquids, the properties of which may be tuned to obtain promising CO2 capture materials [72,73]. Subsequently, Wang et al. synthesized basic ionic liquids by neutralizing weak proton donors (trizole, terazole, imidazole, pyrazole, oxazolidinone, phenol, indole, and bentrizole) with phosphonium hydroxide [74]. They reported that the absorption capacity and the reaction enthalpy were significantly affected by the basicity of the anionic groups, both increasing with higher pKa values. Another class of ionic liquids, super base-derived protic ionic liquids (PILs), was also demonstrated to reversibly capture CO2 with high capacities at atmospheric pressure and 231C [75]. These novel structures were obtained by the combination of super bases (MTBD or P2-Et) with partially fluorinated alcohols (TFE, TFPA, or HFPD), imidazole, pyrrolidone, and phenol. In addition to their high CO2 capacity, their gravimetric capacities were also much better compared to the conventional ionic liquids. Another class of ionic liquids is the reversible ionic liquids (RILs), which are formed through the reaction of CO2 with a neutral organic molecule (alcohol and amine) and revert back to their original nonionic form upon exposure to N2 or Ar [76]. In the pioneering work of Jessop et al., it was demonstrated that an ionic liquid was reported when a 1:1 mixture of 1,8-diazabicyclo[5.4.0]-undec-7-ene (DBU) and 1-haxanol was exposed to CO2 at atmospheric pressure and room temperature [76]. Following this, Heldebrant et al. introduced the concept of “CO2-binding organic liquids” (CO2-BOL) as a version of the RIL [77]. These ionic liquids, because they are liquids before and after CO2 exposure, do not involve excessive inert solvents that reduce the weight and volumetric capacity of the capture material. Accordingly, DBU:1-haxanol CO2-BOL was demonstrated to have much better gravimetric capacity (19%) compared to that of the conventional aqueous MEA solutions (7%). Furthermore, CO2 was reported to be absorbed weakly in the form of alkylcarbonate salts and this – together with the lower heat capacities of the CO2-BOL – led these systems attract growing attention in the field [77]. Even though various ionic liquids have extensively been studied for their potential use in CO2 capture, they still do not provide viable solutions to the existing problems due to their high viscosities and costs [67].

2.27.3

Adsorption

As it was discussed thoroughly in the previous section, amine-based CO2 scrubbing technologies are still much far away from realizing the objected levels of atmospheric CO2 declared by the mitigation policies due to problems associated with their

CO2 Capturing Materials

891

high-energy requirements in the regeneration step, corrosive nature, and formation of environmentally hazardous products as a result of the amine degradation. Correspondingly, parallel to the on-going search for better absorbents; other technologies that would eliminate the challenges inherent in absorption are also taken into consideration in the community. One of the proposed solutions is to utilize solid adsorbents for the capture of CO2 rather than aqueous solutions of absorbents. Adsorption is the buildup of a species on a solid surface due to the interactions between the surface and the species. If this interplay is physical in nature (i.e., van der Waals), the species is physisorbed at the surface without requiring an activation barrier. The binding energies are much weaker in physical adsorption compared to chemisorption that takes place through a chemical reaction. When CO2 is captured through adsorption rather than absorption, the energy that is required for the heating of the CO2-rich solution to the regeneration temperature and vaporization of the solvent will be eliminated to a great deal along with the corrosion problems essential to the wet scrubbing technologies. However, dealing with low CO2 partial pressures at a temperature range B100–1201C still imposes some challenges. It was claimed that an ideal adsorbent should have high adsorption capacity with high CO2 uptakes at lower partial pressures, fast adsorption kinetics, high CO2 selectivity and stability, mild conditions for regeneration, tolerance to the presence of impurities, and low cost [78]. Gray et al. proposed that a potential candidate should have a working capacity of at least 3 mmol CO2/g sorbent for a considerable energy reduction with respect to the conventional MEA systems without sacrificing too much from its capacity [79]. Both physical and chemical adsorbents have attracted substantial attention in this search for better capture materials. Although this section provides an extensive review of the adsorbents developed, the emphasis will be on the chemical sorbents involving amino groups. Further information about physical adsorbents may be found in the very useful reviews published before [78,80,81].

2.27.3.1

Physical Adsorbents

Porous carbonaceous materials (i.e., activated carbon, carbon nanotubes, and graphene/graphite-based materials) constitute one of the most frequently investigated physical sorbents in the field of CO2 capture due to their low cost, high surface area, and thermal stability, easy regeneration (usually with pressure swing adsorption), as well as tunable pore size and structure. Early studies performed with activated carbons that were obtained from various raw materials demonstrated that the equilibrium CO2 capacity was highly dependent on temperature and pressure, i.e., capacities were quite low under flue gas conditions, resulting from the low CO2 affinity of the sorbents [78]. Furthermore, their selectivity and tolerance to the presence of water vapor were also much below the required levels [78]. Following studies concentrated on succeeding an improvement in the surface area and pore structure of these materials in order to obtain higher capacity and selectivity. Accordingly, carbon nanotubes have emerged as potential capture materials due to their well-defined pore structures. However, investigations carried out on carbon nanotubes did not report any better results; both their capacities and selectivities were similar to the previously explored carbonaceous materials (activated carbons) at low partial pressures [78]. More recently, graphene and graphite-based adsorbents have also been under consideration. When unmodified, the performance of graphene was similar to other carbonaceous materials, i.e., low capture capacity at higher temperatures and lower pressures, and still more work is required for the assessment of graphene-based materials as potential CO2 capture materials [81,82]. Zeolites are highly ordered, microporous aluminosilicates that have attracted huge attention as a sorbent due to their tunable pore structure and adsorption characteristics. Early research on zeolites has concentrated on the effect of Si/Al ratio as well as the number and nature of the extraframework cation on adsorption characteristics and established that zeolites with low Si/Al ratios together with high content of extraframework cations were more favorable in terms of CO2 capture [78]. They were, on the other hand, reported to have very fast reaction kinetics, high stability in the absence of water, easy regeneration plus high capacity and selectivity for CO2 at high pressures and low temperatures [78,80]. However, their capacity and selectivity decreased extremely under conditions close to that of flue gas treatments (low pressure, higher temperature and presence of water vapor), restricting their use severely as potential adsorbents [78,80]. Metal organic frameworks (MOFs) form another class of sorbents that have extensively been investigated much recently. MOFs are porous, crystalline compounds that are formed by metal ions or clusters coordinated to organic ligands. Their pore structure and adsorption behaviors may easily be tailored by changing the metallic species or the organic linkers, due to which they were thought of having high potential for CO2 adsorption. However, most of the MOFs were shown to have high capacity and selectivity only at high pressures and low temperatures unless they were modified with chemisorption-inducing groups, similar to other physisorbents considered before [78,81]. Furthermore, despite their high adsorption kinetics, hydrothermal stability of MOFs may be problematic for some kinds [78].

2.27.3.2

Chemical Adsorbents

Notwithstanding the high surface area and easily tunable characteristics of the physical adsorbents discussed above, CO2 capture capacities as well as selectivities obtained for those materials are still much below the required levels under the conditions of flue gas treatment. This has led to the idea of combining high affinity of amine scrubbing technologies with lower regeneration heats of solid adsorbents stemming from the absence of water and resulted in the introduction of amine-functionalized solid sorbents into the CO2 capture literature. In this section, a detailed analysis of the amine-functionalized sorbents will be presented with the emphasis on ordered mesoporous silica and carbon. Discussion about functionalized zeolites and MOFs can be found elsewhere (Ref. [81]).

892

CO2 Capturing Materials

2.27.3.2.1

Support selection and amine functionalization

Amine-functionalized solid sorbents constitute of support materials, the surfaces of which are modified by the addition of aminocontaining species. Usually, materials with high surface area, proper porosity, good mechanical strength, fast kinetics, high hydrothermal stability, and amine affinity are chosen as supports [83]. Accordingly, in addition to carbonaceous sorbents; silicabased, porous adsorbents have also frequently been under consideration in the synthesis of amine-modified solid sorbents. This group of materials are known as ordered mesoporous silica and are famous of their very large specific surface areas, ordered pore systems, and well-defined pore size distributions with much larger-pore sizes than that of zeolites [84]. They are synthesized through the condensation of the silica precursors under basic conditions in the presence of ionic structure-directing agents, which are subsequently removed by extraction or calcination [84]. Mobile Composition of Matter No.41 (MCM-41) and SBA-15 (Santa Barbara Amorphous type material) are the most widely investigated members of this class of materials. Amino groups can be loaded onto various supports through impregnation, grafting and co-condensation. Impregnation involves the contact of the support material with a liquid solution of amines to be loaded and the deposition is achieved through the weak van der Waals interactions between the support and the amine-containing species. Post-synthetic grafting refers to the subsequent modification of the surface of supports with amine-containing organic groups. This modification is attained through the chemical reaction of the support (silanol groups at the pore surfaces) with amino-containing moieties (organosilanes), usually retaining the mesoporous structure of the support [84]. Co-condensation, finally, includes the simultaneous synthesis and functionalization of the support in a reaction mixture of the aminosilane, structure-directing agent and the silica precursor. Since the amino groups are incorporated into the matrix of the functionalized material, it may be possible that some of them are inaccessible to CO2 molecules [83]. Based on the preparation method involved, amine-modified solid sorbents are grouped into three classes, namely, Class 1, Class 2, and Class 3 sorbents [85]. In Class 1 sorbents, amino-containing species is physically attached (impregnation) onto the support, whereas a covalent bond is formed between the support and the amine (grafting) in Class 2 adsorbents. Finally, Class 3 sorbents are synthesized through the in situ polymerization of aminopolymers, starting from an amine-containing monomer [84]. Fig. 3 presents the most commonly used amine structures for functionalization.

2.27.3.2.2

Amine-functionalized ordered mesoporous silica sorbents

2.27.3.2.2.1 Mobile Composition of Matter No.41 The term “molecular basket” was first introduced by Xu et al. for the amine-functionalized CO2 sorbents, where the mesoporous silica support (MCM-41) acted as the basket and the polyethylenimine (PEI) groups (branched, MwB600), by providing numerous adsorption sites, ensured the basket to be a molecular one [86]. In their pioneering work, capture capacity was shown to improve with increasing PEI (branched) loadings, from 0.2 mmol/g sorbent at 0 wt% to 3.02 mmol/g sorbent at 75 wt% PEI, under pure CO2 flow at 751C. When the adsorption capacity was based on the PEI amount rather than the sorbent, it reached a maximum corresponding to 4.88 mmol/g PEI at 50 wt% loading, which was almost twice the capacity achieved by pure PEI (without support), illustrating the synergistic effect of MCM-41 on the adsorption capacity. Further increase of PEI loading, however, resulted in depreciation in the adsorbed quantities. Based on this finding, the authors concluded that the channels of MCM-41 were fully filled with PEI at 50 wt% loading and beyond this amount all the amine structures were coated on the external surface of the support and it was indeed the amino groups impregnated in the channels of MCM-41, which contributed the molecular basket, i.e., increased the CO2 capture capacity. Apart from the PEI amounts, the effect of adsorption temperature on the capacity was also investigated under pure CO2 atmosphere. For 50 wt% PEI-impregnated samples, temperature was displayed to have an enhancing effect on the captured amounts despite the exothermic nature of adsorption. This was suggested to be originating from the fact that adsorption of CO2 was kinetically – rather than thermodynamically – controlled under these conditions. Subsequently, Heydari-Gorji et al. synthesized two different types of MCM-41 and impregnated them with PEI (linear, average Mn B423) in order to determine the effect of surface groups on CO2 adsorption capacity [87]. In the first sample, the surfactant template and the pore-expanding agent were removed by calcination carried out at 5501C, while in the other, ethanol extraction was employed to discard the swelling agent. Both of the supports were loaded with 55 wt% PEI and tested for their CO2 adsorption capacities under a pure CO2 flow at various temperatures. Despite the second sample had a lower pore size and pore volume, its adsorption capacity was superior to that of the first sample at all the temperatures investigated. Based on the rising behavior of the capacity with increasing temperature, it was argued that the adsorption process was diffusion controlled and PEI imposed a lower diffusion resistance on the ethanol extracted sample compared to the calcined one, the superiority being originated from the long-chain alkyltrimethylammonium cations present on the surface of the solvent-extracted MCM-41. The observation that adsorption capacity decreased with increasing amount of amine beyond the optimum loading suggested that pore volume may be a crucial parameter on the determination of the adsorption capacity, indicating that higher pore volumes may lead to higher amine loadings. Accordingly, Sayari group introduced the pore-expanded MCM-41 (PE-MCM-41) – the postsynthesis hydrothermally treated version of the corresponding mesoporous silica – for which the pore volume varied from typically 0.8 up to 3.6 cm3/g [88]. When both conventional MCM-41 and PE-MCM-41 were impregnated with DEA, the latter exhibited a better performance (2.36 mmol/g) at 251C under a 5% CO2 flow, due to the increased amount of amine content originating from the larger pore volume of the PE-MCM-41. However, estimated CO2/DEA ratio (C/N) was lower (0.37) than the theoretical amount (0.5) and the authors argued that this resulted at least partly from the deactivation of DEA through the interactions with the support surface.

CO2 Capturing Materials

OH N H

NH2 H2N

N

N

N H

N

N N

H2N

HO

NH2 H N

893

NH2

NH2

n

OH Branched PEI

TEA H N

H N

NH2

N H

H2N

N H PEHA

O

H N Si

O

H2N

NH2

N H

NH2

O

EDA TRI O O

NH2

Si O

NH2

Si

O

O O APTMS

APTES O

O

O O

H N

N

Si

Si O O

DMAPS

MAPS

O

O O

H N

Si

N

O

Si

NH2

O

AEAPMS

DEAPTMS O O

H N

Si

H2N

O AEAPS

Fig. 3 Molecular structures of the commonly used amines and amine-containing organosilanes.

In order to see the effect of molecular weight on the adsorption performance of amine functionalized MCM-41 sorbents, a comparative study of MCM-41 impregnated with various amine structures was executed [89]. Ethylenediamine (EDA), TEPA, and PEI with two different average molecular weights (MwB600 and MwB1800) were employed as the amine structures. Similar to the previous results reported for PEI-impregnated MCM-41 [86], adsorption capacities were demonstrated to improve with increasing

894

CO2 Capturing Materials

amine loadings up to a point (40 wt%) for all the amines involved. At 40 wt% amine loading, the capacities decreased in the order TEPA 4 PEI (MwB600)4PEI (Mw B1800)4EDA at 351C under a 10% CO2 atmosphere. Surprisingly, EDA-MCM-41 displayed a very poor performance (B0.6 mmol/g) that was almost identical to that of unmodified MCM-41, which was attributed to the EDA losses encountered during the impregnation procedure stemming from the low molecular weight of EDA. In contrast, TEPA, PEI (MwB600) and PEI (MwB1800) were shown to be much successfully loaded onto the mesoporous silica support and the adsorption capacities declined with increasing molecular weight, whereas the thermal stabilities were much higher in the case of larger molecules. Performance of various amine structures were further investigated in a comparative study conducted with 40 wt% EDA, DETA, TEPA, and pentaethylenehexamine (PEHA)-impregnated MCM-41 at 351C and 0.1 bar CO2 partial pressure [90]. CO2 adsorption capacity of the adsorbents followed the order EDA o DETA o TEPA o PEHA, in line with the amine content of the sorbents. If the C/N ratio was taken into consideration, however, all numbers were close to one another (between 0.22 and 0.29) but much lower than the theoretical value of 0.5. Furthermore, CO2 adsorption over TEPA and PEHA-impregnated MCM-41 were reported to be kinetically controlled, while the lower molecular weight EDA and DETA-functionalized mesoporous silica were declared to be under thermodynamic control below 901C. To unveil the effect of the presence of the second amine, mixed amine-impregnated MCM-41 sorbents were synthesized with various loadings of TEPA-MCM-41, AMP-MCM-41, and TEPA-AMP-MCM-41 [91]. Among the sorbents prepared, 30 wt% TEPA-30 wt% AMP-MCM-41 achieved the highest adsorption capacity (3.01 mmol/g), which was much above the level (2.45 mmol/g) accomplished by 60 wt% TEPA-MCM-41 at 701C and 0.15 atm CO2 partial pressure, suggesting that the presence of AMP provided a synergistic effect on the capture capacity. Apart from the amine-impregnated MCM-41 materials, adsorbents manufactured through amine grafting have also been under consideration. One of the early studies performed with 3-[2-(2-aminoethylamino)ethylamino]propyl trimethoxysilane (TRI)grafted MCM-41 and PE-MCM-41 implied that the amount loaded onto the PE-MCM-41 was slightly greater than that of the conventional MCM-41 for the same amount of silane added under identical conditions of preparation [92]. Although the amine content was only slightly favored in the case of PE-MCM-41, at 251C and 0.05 atm CO2 partial pressure there was a significant boosting in the equilibrium adsorption capacity (from 0.97 to 1.41 mmol/g) and the apparent adsorption rate of TRI-PE-MCM-41 sorbent, which was partly attributed to the larger pore diameter and pore volume of the support. In this study, amine-containing moieties were loaded onto the surface of the MCM-41 support at 1101C using toluene as the solvent. In order to resolve the potential effect of grafting conditions on the adsorption capacity, TRI-MCM-41 samples were prepared through water-aided grafting at various temperatures by the same group and then were tested in terms of their adsorption performances at 251C and 0.05 atm CO2 partial pressure [93]. Their findings revealed that the use of PE-MCM-41 instead of conventional MCM-41 resulted in much higher amine loadings and for all the conditions explored, the highest amount of loading was obtained at 851C. Besides, use of water during grafting led to a serious increment in the amount of amine tethered onto the support at all temperatures, which in turn brought about a profound improvement in the adsorption capacity from 1.41 mmol/g reported in the previous study [92] to 2.65 mmol/g, together with the employment of lower grafting temperatures (851C). After the optimization of the synthesis conditions of TRI-PE-MCM-41, the adsorption performance of the sorbent was examined under various operational conditions [94]. As expected, TRI-PE-MCM-41 outperformed the corresponding PE-MCM-41 and MCM-41 in terms of adsorption capacity at low to moderate pressures under pure CO2 flow due to the chemical interaction between the amino groups and CO2. However, there was a notable reduction in the slope of its isotherm at moderate to high pressures suggesting that the strength of this interaction weakened at higher pressures. Nevertheless, adsorption capacity further increased with pressure verifying the contribution of physical adsorption at this pressure range. The decline observed in the adsorption capacity with increasing temperature at all pressures, on the other hand, showed that the adsorption process was thermodynamically controlled for TRI-PE-MCM-41 under the conditions involved in the study. (3-Aminopropyl)-trimethoxy-silane (APTMS) and (3-Aminopropyl)-triethoxy-silane (APTES) are other aminosilane structures that have been frequently employed for the functionalization of MCM-41 support. Adsorption capacity of APTMS-grafted PE-MCM-41 was examined under a gas flow containing 5% CO2 at 251C and was reported as 2.05 mmol/g [95]. The corresponding CO2/N ratio was estimated as 0.49 that was compatible with the carbamate formation reaction but the values greater than 0.5 at higher CO2 concentrations were indicative of the presence of physical adsorption along with the chemical one. In a comparative study of APTMS, (N-methylaminopropyl)-trimethoxy-silane (MAPS) and (N,N-dimethylaminopropyl)trimethoxy-silane (DMAPS)-grafted PE-MCM-41, the adsorption performances of primary (APTMS), secondary (MAPS) and tertiary (DMAPS) amino groups (see Fig. 2 for structures) were examined and compared to that of TRI-PE-MCM-41, which contains both primary and secondary groups [96]. At 251C and 0.05 atm CO2 partial pressure, capture capacities followed the order TRI 4 primary 4 secondary 4 tertiary; while the CO2/N ratio declined in the sequence of primary 4 secondary B TRI 4 tertiary. The more steep slope observed in the isotherm of primary amines with respect to the secondary ones claimed that chemical interactions between CO2 and the amino groups at lower concentrations were more evident in the case of APTMS that was consistent with the estimated isosteric heats of adsorption for the APTMS and MAPS-modified adsorbents. Much recently, CO2 adsorption behavior of APTES and TRI-grafted PE-MCM-41 sorbents were investigated at various temperatures and partial pressures of CO2 [97]. Similar to the previous study, although the adsorption capacity of TRI-PE-MCM-41 was reported to be higher than that of APTMS-PE-MCM-41 under flue gas conditions (at 751C and 0.2 bar pressure), CO2/N ratio was more favorable in the case of the latter. In order to determine the effects of double functionalization on the adsorption performance of amine-modified PE-MCM-41, three different groups of sorbents were prepared [98]. Sorbents in the first group were synthesized through the impregnation of

CO2 Capturing Materials

895

PE-MCM-41 with PEI (branched, MwB800), TEPA, and PEHA, while APTMS, N-[3-(trimethoxysilyl)propyl]ethylenediamine (AEAPS), and TRI-grafted samples constitute the second group. For the third-group materials, on the other hand, APTMS and TRIgrafted materials were further impregnated with PEI, TEPA, and PEHA. At 451C and 1 bar of CO2 partial pressure, both the amine content and the adsorption capacity were shown to increase in the order of APTMS o AEAPS o TRI but the amino efficiencies (CO2/N) of the three materials were almost identical. For impregnated samples under the same conditions, there occurred a significant increment in the adsorption capacities of all the samples with increasing % loadings. At 50 wt% loading, CO2 amounts adsorbed by PEI, TEPA, and PEHA-impregnated sorbents were much higher than the amounts captured by grafted samples, while the amino efficiencies were much favorable in the case of grafted materials. When APTMS-grafted PE-MCM-41 was further impregnated with various loadings of PEI, PEHA, and TETA; the adsorption capacity was found to be superior to the grafted material at all loadings. However, amino efficiency declined with increasing impregnated amounts, indicating the presence of pore blockage and/or support saturation. In the case of double-functionalized TRI-PE-MCM-41, this blockage and/or saturation was more profound stemming from the longer organic chains of the TRI structure leading to amino efficiencies as low as 0.09. A similar survey was conducted by Wang et al. using APTMS-grafted, TEPA-impregnated and double-functionalized MCM-41 sorbents at 701C and 0.15 bar CO2 partial pressure [99]. For 20, 30, 40, 50, and 60 wt% TEPA-impregnated MCM-41 samples, CO2 uptake was shown to escalate with increasing loadings. For the APTMS-grafted samples, on the other hand, adsorption capacity was improved with amine loadings up to 30 wt% but decreased beyond that point due to the restriction of diffusion by the aggregated APTMS in the channels or on the external surface of MCM-41. To investigate the effect of double functionalization 20, 30, 40 and 50 wt% APTMS-grafted MCM-41 samples were further impregnated with TEPA with various loadings. Equilibrium adsorption capacities of double-functionalized sorbents were usually higher compared to APTMS-grafted samples, the capacity improving with increasing amounts of TEPA up to a point. Under the conditions studied, 30 wt% APTMS-40 wt% TEPA-MCM-41 was shown to have the best adsorption performance with an uptake of 3.45 mmol/g. 2.27.3.2.2.2 Santa Barbara amorphous-15 SBA-15 is another ordered mesoporous silica structure that has attracted very high attention for its application in the manufacture of amine-modified sorbents for CO2 capture. Yue et al. functionalized two different types of SBA-15, namely, SBA(P) (waterwashed support without calcination) and SBA(C) (calcined support at 5501C) by impregnation of TEPA and DEA onto both supports and tested the samples for their adsorption capacities at 751C under pure CO2 atmosphere [100]. Capture capacities of DEA-impregnated SBA(P) and SBA(C) were found to be much lower than their TEPA-impregnated counterparts, stemming from the higher amine group density of the TEPA structure. When a mixture of the two amines with a total wt% of 50 was loaded onto the two supports, an adsorption capacity of 3.70 mmol/g was obtained for the 30 wt% TEPA-20 wt% DEA-SBA(P)-15 sorbents at 751C under a pure CO2 environment, which was indicative of a synergistic effect between TEPA and DEA. This effect was attributed to the presence of –OH groups in the structure of DETA and was found to be present for samples prepared with SBA(C) as well. Sanz et al. investigated pure CO2 adsorption isotherms of PEI (MwB800)-impregnated SBA-15 with various loadings at 25, 45, and 751C [101]. At 451C and atmospheric pressure, the amount of CO2 adsorbed onto the sorbent was shown to improve with increasing amounts of PEI, while the amino efficiency declined from 0.26 to 0.17 at 10 and 70 wt% PEI, respectively. The reason why the observed amino efficiencies were much below the theoretical amount of 0.5 was stated to be the branched structure of the employed PEI, for which 25% of all the amino groups involved were tertiary. As for the decrease in the amino efficiency, limited accessibility of the internal amino groups due to the increasing packing effect of the PEI molecules were held responsible. Similar to the findings reported for MCM-41 [86], both the adsorption capacity and the amino efficiency were shown to be more favorable at higher temperatures originating either from the kinetic effects or the expansion of PEI structure with temperature. Comparable effects of temperature on adsorption amount were reported by subsequent studies performed with 50 wt% PEI (linear, MwB423)-impregnated various SBA-15 at 0.15 bar CO2 partial pressure [102]. The samples differed from one another in their pore structures and the total pore volume was claimed to be an important parameter in the determination of the samples’ adsorption performances, the highest capacity (2.39 mmol/g) being recorded for the adsorbent prepared with the highestpore-volume support at 751C. A comparison of the adsorption behaviors of PEI-impregnated SBA-15 and MCM-41 structures was reported in a couple of studies. Son et al. synthesized 50 wt% PEI (linear, MwB600)-functionalized SBA-15 and MCM-41 samples and claimed that SBA15 performed much better than MCM-41 in terms of adsorption capacity and kinetics at 751C under a pure CO2 flow [103]. They attributed this to the larger pore diameters of SBA-15 and concluded that adsorption performance was highly determined by the average pore diameter of the support. Effect of pore length on CO2 adsorption was also investigated by using PEI (MnB423)impregnated PE-MCM-41 and SBA-15 with various pore diameters (7.2 and 10.5 nm) and morphology (conventional and platelet) [104]. Their results revealed that at 1 bar of CO2 partial pressures and a temperature range 25–1001C, SBA-15 (platelet) outperformed all the other sorbents, capacities decreasing in the order: SBA-15 (platelet)4SBA-15 (Dp ¼ 10.5 nm)BSBA-15 (Dp ¼ 7.2 nm)4PE-MCM-41. The effect of pore diameter on the adsorption capacity was ruled out under the conditions involved in the study, based on the almost identical performances of SBA-15 (Dp ¼10.5 nm) and SBA-15 (Dp ¼ 7.2 nm). However, pore length was seen to be quite determining on the adsorption properties (capacity and kinetics), samples with shorter pore lengths performing better due to the diminished diffusion pathways associated with them. Finally, Yan et al. carried out a comparative examination of MCM-41 and SBA-15 with various microporosities impregnated with 20 wt% PEI (MnB600) [105]. Their adsorption isotherms at 0, 5 and 101C displayed that PEI-MCM-41 had an inferior capacity with respect to the others at all the

896

CO2 Capturing Materials

temperatures considered, stemming from the smaller pore size of the MCM-41 support. Unlike the previously reported temperature effects [101], adsorption capacities were found to diminish with increasing temperature proposing that the process was under thermodynamic control. Furthermore, although microporosity had a pronounced effect on the adsorption capacities of unmodified supports, their role subsided significantly with the introduction of PEI and the CO2 uptake of the amine-impregnated SBA-15 samples correlated with their surface areas. SBA-15 has attracted more attention for grafting of amines rather than impregnation. One of the early studies in the field analyzed the CO2 capture amounts of AEAPS-grafted SBA-15 at 251C and 0.15 bar CO2 partial pressure [106]. Their estimated capacity was around 0.45 mmol/g, which was much below the required levels. Linfang et al. prepared APTES-grafted SBA-15 samples by using post-synthesis silylation and one-step silylation methods using ethanol as the solvent [107]. The first sample was synthesized with calcined SBA-15; while for the second one as prepared SBA-15 (no calcination) was employed. Their characterization results demonstrated that one-step silylation resulted in higher amine loadings, which in turn led to more favorable adsorption capacities. This was attributed to the translation of free silanol into hydrogen-bonded silanol upon calcination that was claimed to be inactive during surface functionalization. Further study of grafting conditions (temperature and reflux time) affirmed that samples refluxed at 501C for 20 h achieved the highest amount of adsorption (B0.95 mmol/g) at 0.02 MPa CO2 partial pressure and 251C temperature inline with its highest amine content. When temperature was raised to 651C, there was a decrease in the capture capacity of the one-step-grafted sample, indicative of a thermodynamically controlled adsorption under the conditions considered. Effect of calcination on CO2 adsorption properties of APTMS-grafted SBA-15 samples was further investigated by Wang et al. [108]. For this, two different methods were employed for the template removal, namely, ethanol extraction and calcination. The silanol density of the extracted samples was reported to be significantly higher than that of the calcined one, giving rise to higher amine contents in the ethanol extracted samples. However, despite the improvement of the silanol density was very pronounced, the amine content of the uncalcined sorbents did not differ drastically from the other samples; which was indicative of inaccessible silanol groups. Nevertheless, APTMS-SBA-15 prepared through ethanol extraction was shown to adsorb more CO2 at 251C and the estimated isosteric heats of adsorption demonstrated that chemisorption was still effective for those adsorbents even above a CO2 loading of 1.1 mmol/g. In a comparative study involving APTES, AEAPS, TRI, MAPS, DMAPS, and N-(2-aminoethyl)-3-aminopropylmethyldiethoxysilane (AEAPMS)-grafted SBA-15, three different grafting conditions were involved [109]. Adsorption capacities measured at 601C and 0.15 kPa CO2 partial pressure were superior for samples prepared by boiling SBA-15 in water for 2 h, independent of the amine (APTES, AEAPS, or TRI) grafted onto the sample. Both the amine content and the adsorbed CO2 amount increased in the order of APTES o AEAPS o TRI (0.66, 1.36, and 1.58 mmol/g, respectively). Samples involving MAPS, DMAPS, and AEAPMS, on the other hand, were prepared without boiling the SBA-15 support and when compared to their counterparts (APTES, AEAPS, and TRI grafted under the same conditions), the adsorption capacity was shown to escalate in the sequence of DMAPS o MAPS o APTES o AEAPS o AEAPMS o TRI. The very small amounts of capture achieved by DMAPS (0.05 mmol/g) suggested that tertiary amine groups were not reactive in the absence of water. Adsorption performances of primary, secondary, and tertiary amino groups were studied by synthesizing APTMS, MAPS, and [3-(diethylamino)propyl]trimethoxysilane (DEAPTMS)-grafted SBA-15 sorbents [110]. At 251C and atmospheric pressure, primary amine-containing APTMS-SBA-15 was found to be the most favorable material in terms of adsorption capacity and amino efficiency, which was followed by the secondary (MAPS) and the tertiary (DEAPTMS) amine-grafted SBA-15, respectively. To inspect the effect of preparation method on the adsorption behavior; impregnation, grafting, and co-condensation methods were utilized in order to functionalize SBA-15 by APTMS, AEAPS, TRI, and PEI (MwB800) [111]. Among the samples synthesized, co-condensed sorbents possessed the lowest amount of amine content (in the range 1.3–2.7 wt%) compared to the impregnated and grafted materials. This, as expected, brought about the lowest adsorption capacities and amino efficiencies for APTMS, AEAPS, and TRI-co-condensed SBA-15 samples at 451C and 0.15 bar CO2 partial pressure, which was attributed to the inaccessibility of some amino groups due to their locations (inside the silica walls). Adsorption capacities of APTMS, AEAPS, and TRI-grafted SBA15 were much greater and ranged between 0.75 and 1.38 mmol CO2/g sorbent depending on the nitrogen content of the sorbent and increased in the order of APTMS o AEAPS o TRI, whereas the highest amino efficiency was estimated for the mono-amine (APTMS) containing adsorbent. Moreover, CO2 adsorption amounts were improved with increasing organic loading of PEIimpregnated samples and reached a value of 1.62 mmol/g for 50 wt% PEI-SBA-15. The amino efficiencies of the samples, however, declined with increasing loading implying the diffusional limitations imposed by the PEI structures deposited on the external surface of the support. Following this, the effect of preparation method was further explored by the same group by using APTMS and TRI-grafted sorbents along with PEI (branched, MwB800) and TEPA-impregnated SBA-15 [112]. At 251C under a pure CO2 flow, adsorption capacities of the grafted materials were reported as 1.40 and 1.74 mmol/g for APTMS and TRI-grafted SBA-15, respectively; the latter possessing higher content of nitrogen. Similarly, CO2 uptake of the 50 wt% TEPA-imgregnated SBA-15 (2.22 mmol/g) was found to be superior to that of the 50 wt% PEI-impregnated sample (1.72 mmol/g) due to the higher content of amine in the TEPA-containing samples. Although the nitrogen content of the PEI-impregnated samples was much higher compared to the grafted sorbents, its performance did not differ notably from them resulting from the low amino efficiencies of the PEI-containing samples, as discussed in the previous study [111]. When temperature was increased to 1101C, a decline was observed in the capture capacities of all the samples but this was more pronounced for the impregnated sorbents. Next, the effect of CO2 concentration on the adsorption performance of the sorbents was investigated at 451C by using pure and 15% (by volume) CO2 flows. Although lower uptakes were recorded by all the samples under lower CO2 partial pressures (except for 70 wt%

CO2 Capturing Materials

897

TEPA-SBA-15), they could still maintain most of their capacities under dilute streams (especially impregnated samples with high loadings), stemming from the chemical interaction between CO2 and the amino groups. Much recently, Vilarrasa-Garcia et al. synthesized various SBA-15 supports with modified pore widths and lengths through the addition of different chemicals and functionalized them with APTES grafting or PEI (branched, MnB600) impregnation [113]. Based on the characterization of the manufactured support structures, they claimed that P6 mm hexagonal symmetry of the conventional SBA-15 structures was lost and a transition from a hexagonal arrangement of parallel channels to a mesoocellular foam structure took place upon addition of a swelling agent (heptane) or ammonium fluoride. It was furthermore proposed that addition of ammonium fluoride did not only gave rise to a mesocellular foam structure but also ended up with a lower silanol density, which in turn led to restricted amounts of nitrogen grafted onto the surface. As a result, adsorption performances of APTES-grafted SBA-15 prepared through the addition of ammonium fluoride were found to be inferior with respect to the samples prepared with conventional support at 251C and atmospheric pressure. On the other hand, although both the amino efficiencies and the amounts of CO2 adsorbed by 30 wt% PEI-impregnated samples were much lower than the values observed for APTESgrafted ones, addition of fluoride seemed to have a boosting effect on the relative contribution of chemisorption with respect to physisorption for PEI-containing samples, which was related to the decreased micropore volumes of the fluoride-containing sorbents. Besides, the CO2 uptake of the 50 wt% PEI-impregnated fluoride-containing SBA-15 was shown to increase much more than that of the conventional support, indicating that CO2 experienced less diffusional resistances in the case of mesocellular foam structure due to its more open porous structure. Similar to MCM-41, double functionalization of SBA-15 was also considered. For this purpose, Sanz et al. prepared APTMS and TRI-grafted PE-SBA-15 samples in addition to PEI (branched, MnB800) and TEPA-impregnated ones and then compared the adsorption performances of these with those of the double-functionalized samples that were obtained with the impregnation of TEPA and PEI onto APTMS and TRI-grafted sorbents [114]. At 451C and 1 bar pressure, CO2 uptakes of APTMS-PE-SBA-15 and TRI-PE-SBA-15 were measured to be 1.49 and 1.62 mmol/g, respectively; whereas their corresponding amino efficiencies were 0.48 and 0.34. These results did not differ much from those reported previously for conventional SBA-15 grafted with the same structures, implying that use of a pore-expanded support did not provide any advantages due to the fact that a significant fraction of the pores were still empty even after the functionalization. For the 50 wt% PEI and TEPA-modified sorbents, in contrast, eventhough the amine content of the impregnated SBA-15 and PE-SBA-15 were almost identical; both the adsorption performances and the amino efficiencies of PEI and TEPA-impregnated PE-SBA-15 surpassed those of the functionalized SBA-15 samples, indicative of the enhancement of CO2 diffusion in the presence of larger pores. In this way, adsorption capacity of TEPA-PE-SBA15 reached a value of 3.73 mmol/g at 451C and atmospheric pressure. When APTMS-grafted PE-SBA-15 was impregnated with 30 wt% PEI and TEPA, CO2 amount adsorbed by the resulting sorbents were reported as 2.52 and 3.16 mmol/g, respectively, both being much larger than those obtained with the singly modified samples. The corresponding amino efficiencies were estimated as 0.33 and 0.40, respectively; implying a much efficient use of impregnated amino groups. Next, TRI-grafted PE-SBA-15 was further impregnated with 30 wt% TEPA, resulting in a CO2 uptake of 3.89 mmol/g under the conditions employed. The CO2/N ratio for this sample was recorded as 0.42, which was much higher than those estimated for TRI-PE-SBA-15 and 30 wt% TEPA-PE-SBA-15. Despite the improvement observed in the adsorption performances of the synthesized samples at first might seem to be originating from the higher amine contents of the double-functionalized samples, this probability was ruled out by considering the lower or similar amino efficiencies of the 50 wt% PEI and TEPA-PE-SBA-15 in spite of their higher nitrogen contents compared to the double-functionalized sorbents. Instead, it was proposed that a synergistic effect between grafting and impregnation occurred, originating mainly from the boosting of the zwitterion deprotonation by the highly mobile impregnated amino groups. Finally, an adsorption capacity as high as 4.88 mmol/g was reported for 50 wt% TEPA-impregnated APTMS-PE-SBA-15. Following this work, adsorption performances of the same sorbents were studied in terms of kinetics by the same group and it was claimed that CO2 adsorption rate was much faster for the impregnated samples compared to grafted and double-functionalized ones, and the rate was highly dependent on the amine type [115]. As it was for the impregnated samples, adsorption capacities of MCM-41 and SBA-15 grafted with the same aminosilane (APTES) were also investigated comparatively at 251C and 0.1 bar of CO2 partial pressure [116]. Among the samples studied, APTES-SBA-15 was found to be more favorable in terms of adsorption capacity with an uptake amount of 1.54 mmol/g. The superiority of these samples were attributed to the larger pore size of the SBA-15 support together with its higher amine surface density which was argued to assure the presence of two nitrogen atoms in close proximity for the deprotonation of the zwitterionic intermediate. Subsequently, Chang et al. analyzed the adsorption performances of APTES-grafted MCM-41, SBA-15, and PE-SBA15; and reported that the adsorbents prepared with SBA-15 could achieve much higher amounts of CO2 adsorbed compared to the others [117]. To further investigate the adsorption characteristics of modified SBA-15; APTES, AEAPS, and TRI-grafted samples were prepared by using SBA-15 that was synthesized by two different methods. Among them, TRI-grafted vacuum dehydrated SBA-15 demonstrated the highest capture capacity (2.41 mmol/g) and amino efficiency (0.66) at 601C and 0.15 bar CO2 partial pressure. Furthermore, CO2 uptake of the TRI-grafted SBA-15 was shown to rise with increasing CO2 concentrations but decrease with increasing temperature. 2.27.3.2.2.3 Thermal stability of functionalized mesoporous silica Long-term stability is something indispensable for the development of feasible sorbents for CO2 capture. In addition to cyclic stability, potential adsorbents should also possess high thermal stability since the regeneration of the functionalized sorbents are usually achieved through desorption of CO2 at elevated temperatures owing to the exothermic nature of adsorption. For this

898

CO2 Capturing Materials

purpose, most of the materials developed for CO2 adsorption were evaluated in terms of their cyclic and thermal stabilities along with their capture capacities. Early investigation of cyclic adsorption (751C) and desorption (751C) of 50 wt% PEI-impregnated MCM-41 demonstrated that the adsorption capacity was constant and desorption was complete after seven cycles of adsorption and desorption [86]. Similar findings were reported for 50 wt% PEI-impregnated SBA-15, as well [102]. When PEI was loaded onto SBA-15 supports with various pore structures and pore volumes, both the adsorption capacity and cyclic stability were found to be dependent on the pore diameter and pore volume. Cyclic performances of the larger-pore samples, for which higher CO2 uptakes were recorded, were shown to be fairly stable after 12 cycles of adsorption (751C) and desorption (1001C), whereas adsorption capacity of the sorbent with the smallest pore size and pore volume decreased quickly with each cycle, losing 13.5% of its initial capacity, which was attributed to the loss of some PEI accumulated on the external surface of the support. In a comparative examination of cyclic performances (adsorption 451C, desorption 451C (30 min) þ 1101C (2 h)) of 50 wt% PEI and TEPA-impregnated-SBA-15 sorbents, adsorption capacity of the PEI-impregnated samples were shown to remain almost constant (94.5% of the initial capacity) after 10 cycles [112]. For the TEPA-impregnated samples, however, CO2 uptake was subjected to a more significant decline in the second cycle, but the sample could still capture 86.3% of the initial amount after 10 cycles. This stability loss was explained through partial evaporation or degradation of TEPA that might be taking place under the conditions employed due to the higher volatility of the TEPA molecule. A more systematic inspection of the effect of molecular weight on the thermal stabilities of impregnated sorbents was performed with 40 wt% EDA, TEPA, and PEI with two different average molecular weights (MwB600 and MwB1800)-functionalized MCM-41 [89]. The smaller weight loss of EDA-impregnated material above 1001C suggested that some of the EDA that was intended to be loaded onto the support had already volatilized during the preparation process due to its high volatility. In addition, the temperature at which the weight loss reached a maximum point increased in the order of EDA-MCM-14oTEPA-MCM-41oPEI (MwB600)-MCM-41oPEI (MwB800)MCM-41, implying that the thermal stability of the samples were enhanced with increasing molecular weights of the amines. When the TEPA, PEI (600) and PEI (800)-impregnated samples were exposed to 10 adsorption (351C)–desorption (1001C) cycles, on the other hand, adsorption capacities of the PEI-modified sorbents were seen to be almost constant, while a 7.4% drop was recorded in the final capacities of TEPA-impregnated samples with respect to the initial ones, further confirming the good thermal and cyclic stability of the PEI-modified sorbents. Cyclic capacity for the 30 wt% TEPA-20 wt% DEA-SBA-15, in contrast, was reported to be still 95.8% of the initial amount after the sixth cycle, illustrating the enhancement provided by DETA [100]. Improvements of the cyclic capacity of TEPA-containing sorbents in the presence of the second amine structure were also demonstrated by Wang et al. for 30 wt% TEPA–30 wt% AMP-MCM-41 [91]. After 15 cycles of adsorption (701C) and desorption (1001C), mixed-amine sorbent could still maintain 96% of its initial capacity, which was much higher compared to only TEPAimpregnated sorbents. DEA-PE-MCM-41 was also investigated in terms of thermal and cyclic stability [88]. Thermogravimetric analysis of the sample demonstrated that a higher portion of the DEA was lost below 2801C, while the remaining loss took place above 3001C. When the sample was heated to 2501C, held at that temperature for 15 min and then cooled to 251C, it could adsorb only 4.15% of the initial value, which was indicative of a large extent of deactivation due to the decomposition observed below 2801C. Cyclic capacity of the DEA-PE-MCM-41 was shown to be fairly stable, however, after seven cycles of adsorption (251C) and desorption (751C), losing only 3.3% of the initial capacity. When compared to impregnated samples, thermal and cyclic stabilities of grafted samples were declared to be much better. Wang et al. studied thermal regeneration of APTMS-grafted SBA-15, for which the template removal was achieved by ethanol extraction [108]. After eight cycles of continuous adsorption (251C) and desorption (901C), adsorption capacity of the sample was shown to be almost constant. However, if regeneration was performed at 1501C; APTMS-SBA-15 could achieve 18% less adsorption with respect to the sample regenerated at 901C. Based on this, the authors concluded that temperatures higher than 901C should be avoided for degassing. Similarly, APTES-SBA-15 [107] and TRI-PE-MCM-41 [94] were demonstrated to sustain their adsorption capacities after three and seven adsorption–desorption cycles, respectively. To investigate the origin of unstability, Sayari et al. performed a comparative study by using APTMS and MAPS-grafted PE-MCM-41 [96]. Among the sorbents considered, MAPS-PE-MCM-41 had the highest cyclic stability, preserving its working capacity even after 60 adsorption (551C) and desorption (1201C) cycles, whereas APTMS-PE-MCM-41 was the least stable one, losing 21% of its initial capacity under the same conditions. This significant degree of deactivation observed for the primary amine was attributed to the formation of urea under anhydrous conditions through the isocyanate intermediate and carbamate dehydration. When APTMS-grafted ordered mesoporous silica was further impregnated with TEPA, adsorption capacity of the MCM-41 supported sorbent was detected to lose only 3.43% of its initial activity after 10 cycles of adsorption (701C) and desorption (1001C) [99], while the PESBA-15 supported one was subjected to a slight but sustained decrease, maintaining only 87% of its initial capacity at the end of five cycles [115]. In summary, although impregnated mesoporous silica sorbents could achieve higher amine loadings compared to the grafted ones, they may suffer from low thermal and cyclic capacities stemming from the weak interactions between the amine structures and the mesoporous silica surfaces. However, in most of the studies investigating the long-term stability of the adsorbents, the number of adsorption and desorption cycles that was involved were at most 10–15, which was still far below the realistic numbers encountered in industry. More realistic conditions should be employed for the investigation of cyclic capacity before reaching exact conclusions.

CO2 Capturing Materials

899

2.27.3.2.2.4 Effect of moisture on the adsorption capacities of mesoporous silica sorbents Since flue gas streams contain 8–10% H2O, moisture tolerance is crucial for the development of feasible sorbents. Accordingly, effect of moisture on the adsorption capacities of amine functionalized mesoporous silica sorbents has been quite frequently under consideration in the CO2 adsorption community. However, although mechanistic studies have demonstrated that CO2 may be captured in terms of bicarbonate rather than carbamate in the presence of water, depending on the type of the amine structure involved [7], there was no general consensus reached about the effect of water in the literature. Early experiments performed with 50 wt% PEI-MCM-41 at 751C under various flow compositions claimed that moisture had a promoting effect on the adsorption capacity and further stated that this promoting effect is more pronounced when the concentration of moisture is lower than that of the CO2 [118]. In order to explain the origin of this enhancement, adsorption runs were conducted for unmodified MCM-41 in the presence and absence of moisture. It was observed that addition of H2O hardly affected the results; ruling out the possibility that the promoting effect provided by the moisture originated from the dissolution of CO2 in water and confirming the role of bicarbonate formation in this promotion. Further studies conducted with various impregnated mesoporous silica sorbents, however, reported differing findings. When DEA-PE-MCM-41 was investigated at 251C under 5% CO2 flow with 28% relative humidity, adsorption capacity was found to be almost identical (2.85 mmol/g) to the value recorded under anhydrous conditions [88]. In contrast, at 451C and 0.15 bar CO2 partial pressure, adsorption capacities of 50 wt% PEI and TEPA-impregnated SBA-15 were measured to be 1.96 and 3.67 mmol/g under a flow of 5% H2O, respectively [112]. When compared to the capture capacities obtained under anhydrous conditions (1.70 mmol/g for PEI and 2.58 mmol/g for TEPA), the boosting effect of moisture on the adsorption capacities was evident. Water tolerance of grafted sorbents was also under consideration. Wang et al. reported CO2 uptakes of two different APTMSgrafted SBA-15 samples at 251C under a flow of 3% H2O in balance with CO2 [108]. According to their results, APTMS-SBA-15 (calcined) achieved weight gains of 61 and 121 mg/g, respectively, under anhydrous and moist conditions, whereas the corresponding value for the ethanol extracted sample increased from 76 to 143 mg/g, in the presence and absence of water, respectively. However, the authors claimed that they did not have a mass spectrometer to differentiate between the adsorbed amounts of CO2 and water but still argued that the capacities were improved in the presence of moisture. The positive effect of H2O on the adsorption capacity of APTMS-PE-MCM-41 was indeed demonstrated by using a mass spectrometer much earlier than this study [95]. At 251C under a flow of 5% CO2, adsorption capacities of the samples increased with increasing relative humidity, all values being much higher with respect to the dry capacity. In contrast, in a comparative study conducted much formerly, adsorption performance of APTMS-SBA-15 was illustrated to remain constant in the presence of humidity (601C, 15 kPa CO2 and 12 kPa H2O), similar to MAPS, DMAPS, and AEAPMS-grafted samples [109]. AEAPS and TRI-grafted sorbents, on the other hand, were declared to possess slightly improved capacities under moist conditions. Similar findings that TRI-grafted mesoporous silica sorbents could achieve slightly higher amounts of CO2 adsorption when water was involved in the reactant stream were reported by other groups as well [92–94,117].

2.27.3.2.3

Amine-functionalized carbonaceous sorbents

2.27.3.2.3.1 Activated carbons Although amine modification of carbon-based materials has not attracted as high attention as functionalized mesoporous silicas, they are still among the most extensively investigated materials in the CO2 capture literature. Early experiments performed with DETA, PEHA, and PEI-impregnated commercial mesoporous carbon demonstrated that capture capacity of unmodified samples were much higher than those of the impregnated samples at 251C [119]. It was claimed that the dominant interaction mechanism between CO2 and the sorbents was physical in nature at room temperature and upon impregnation a significant amount of the pores were blocked by the amine moieties, leading to a decrease in the amount of CO2 physisorbed in the case of amineimpregnated materials. As temperature got higher, there was a significant drop in the amount captured by all samples, which was consistent with physical adsorption. However, the decline observed for the bare carbon support was more notable than that of the others and the capture capacity of DETA-impregnated samples exceeded the others above B601C suggesting the contribution of chemical adsorption at medium temperatures. Nevertheless, adsorption capacities reported in the study were much below the required amounts. Subsequently, Bezerra et al. explored the adsorption performances of triethanolamine (TEA) and MEA-impregnated microporous carbons at a temperature and pressure range 25–751C and 0.1–10 bar, respectively [120]. At 1 bar, the adsorption capacity of the unmodified sample was reported to be 83 mg/g at 251C and 31 mg/g at 751C, the decrease with temperature being indicative of physisorption. For MEA-impregnated samples, however, the capacities were 45 and 75 mg/g, respectively, at 25 and 751C. This enhancement at higher temperatures indicated that chemisorption played an active role in the capture of CO2 at 751C when the support was modified with MEA. TEA impregnation, on the other hand, did not result in promising capture capacities at both temperatures, which was attributed to the poor textural properties of the samples. More recent experiments carried out with PEI-modified mesoporous carbon investigated the effect of PEI loading, CO2 concentration, and temperature on the adsorption capacities of the sorbents [121]. At 751C and atmospheric pressure, capture capacity of the unmodified carbon support was measured as 0.05 mmol CO2/g sorbent under a gas flow involving 15% CO2, due to the limited physisorption. When PEI was impregnated on the mesoporous carbon, sorption capacity was observed to rise with increasing PEI amounts up to a loading of 75 wt% under the same conditions. Further increase of PEI to 80% resulted in a significant decline in the sorption capacity, stemming from the inaccessibility of the PEI moieties due to the

900

CO2 Capturing Materials

blockage of the pores. When the amine utilization ratio was taken into account; adsorbents with 65 wt% PEI were found to be superior to all other adsorbents with a utilization ratio of 63%. Supplemantary measurements taken with increasing CO2 concentrations from 5 to 80% showed that capture capacity increased with rising concentrations and 65 wt% PEI-impregnated mesoporous carbon was capable of completely removing the CO2 even at the highest concentration. On the other hand, experiments performed at various temperatures (between 30 and 1001C) indicated that temperature increase up to 801C improved the sorption capacity due to the dominance of the kinetic factors in this region. Finally, the presence of moisture (80% relative humidity) had a positive effect on the adsorption capacity rather than deteriorating the results as in the case of physical adsorbents. Gibson et al. studied the effect of pore structure on the CO2 adsorption behaviors of various polyamine-impregnated porous carbons by employing microporous and mesoporous carbons [122]. Pore volume and surface area of both carbon supports were proven to decline with increasing loadings of TETA, the decrease being more distinguished in the case of microporous carbon due to the partial or complete blockage of the pores depending on the amount of loading. In contrast, mesoporous carbons seemed to experience only a small level of pore blocking through the whole range of loading suggesting that mesoporous supports should be involved in adsorption studies in order to obtain a more homogenous distribution of the amine groups. To verify the superiority of mesoporous carbon, adsorption experiments under conditions similar to flue gas treatment (0.1 bar CO2 partial pressure, 751C temperature) were conducted by using TETA and PEI-impregnated carbons. For both of the supports, activation with amino groups resulted in significant enhancements in the adsorption capacities. However, this enhancement was restricted by the amount of amine loading in the case of microporous carbons, i.e., there was an optimum loading amount beyond which adsorption capacity decreased with increasing amine loadings due to the limited access to active sites in the inner pores. Furthermore, when the adsorption performances of TETA and PEI-impregnated mesoporous carbons were compared, lower molecular weight amines (TETA) were found to be superior to higher molecular weight structures (PEI) in terms of adsorption capacity but inferior in terms of stability. More recently, Kongnoo et al. investigated the CO2 adsorption characteristics of MEA and DEA-impregnated mesoporous commercial carbons [123]. Both samples demonstrated better performances with respect to the unmodified support at a temperature and pressure range 40–701C and 20–500 kPa, respectively, but the improvements observed upon DEA impregnation was more pronounced compared to MEA. This was in complete agreement with the surface characterization of both samples indicating that MEA molecules brought about more partial and full blockages of the inner pores of the support resulting in more hindered mass transfer. 2.27.3.2.3.2 Carbon nanotubes In addition to activated carbons, amine functionalized carbon nanotubes have also been investigated frequently as potential capture materials. Different from the porous carbon supports, both impregnation and grafting were used for the loading of amino groups. Su et al. investigated the CO2 capture performances of APTES, AEAPS, and PEI-modified multiwalled carbon nanotubes at various temperatures under 15% CO2 concentrations [124]. At low temperatures, all functionalized samples achieved higher capacities compared to the unmodified nanotubes. This enhancement was much greater for APTES and AEAPS-grafted sorbents at low temperatures. As temperature increased, adsorption capacities of all the samples (including the unmodified nanotubes) were found to decline resulting in low capacities under flue gas conditions and indicating that the interaction between the sorbents and CO2 was physical. Average isosteric heat of adsorption for APTES-grafted multiwalled carbon nanotubes was calculated as 11.8 kJ/mol in a following study, verifying the physical nature of this interplay [125]. Due to the high costs of commercially purified multiwalled carbon nanotubes, their cheaper versions – industrial grade multiwalled carbon nanotubes (IGMWNT) – have also been considered for their applications in CO2 capture. Liu et al. explored the adsorption capacities of TEPA-impregnated IGMWNTs and demonstrated that they have comparable capacities to those of TEPA-functionalized commercially purified multiwalled carbon nanotubes at amine loadings between 0 and 60% [126]. Capture capacities of IGMWNTs improved with increasing TEPA loadings at 401C and 0.1 bar CO2 partial pressure reaching a value of 2.593 mmol/g at 50 wt% TEPA loadings. Further increase in the amount of amine to 60 wt%, however, resulted in a slight decrease in the adsorption capacity. This was in complete agreement with the previous findings reported for amine-impregnated activated carbons [121,122] and attributed to the blockage of the pores that was evident from the sharp decline observed in the pore volume of 60 wt% TEPA-impregnated IGMWNTs. Temperature effect on adsorption capacities of 50 wt% TEPA-impregnated IGMWNTs as well as the isosteric heats of adsorption were also explored. Although adsorption capacity increased from 2.145 mmol/g at 201C to 3.088 mmol/g at 701C, the trend was the opposite beyond this point indicating partial desorption of CO2 above this temperature. Notwithstanding this increasing behavior of capture capacity with temperature (up to a point), there was a continuous decline in the isosteric heats of adsorption with increasing adsorption capacity pointing out a combination of physical and chemical adsorption. When IGMWNTs were impregnated with PEI and ethylenediamine endcapped PEI (PEI-EC), PEI-EC-impregnated nanotubes were found to be superior in terms of adsorption capacity at 651C and a CO2 partial pressure of 0.1 bar, originating from the amino groups present at the end of PEI molecules in the PEI-EC structure [127]. Similar to the TEPA-impregnated IGMWNTs [126], adsorption capacity was improved with increasing temperature up to 701C which indicated that kinetic limitations were dominant at the low temperature region. In contrast, estimated isosteric heat of adsorption (73.7 kJ/mol) proved that the interaction between CO2 and the PEI-EC-impregnated IGMWNT was mainly chemical.

CO2 Capturing Materials

2.27.4

901

Illustrative Example: What to and Not to Have for a Feasible Adsorbent?

An extensive review of the literature corresponding to CO2 adsorbents has been discussed above. As it was clear from the given discussion, potential sorbents have widely been investigated in terms of their adsorption capacities and kinetics in addition to thermal and cyclic stabilities. In doing so, various mesoporous or microporous sorbents were modified with different amine structures and the ones acquiring high potentials of application were further investigated for optimization of a couple of parameters, like the operation conditions, amine type and loading, physical properties of the modified and unmodified support (pore size, pore volume, surface are), preparation methods and conditions, pretreatment conditions, etc. Taking the huge variety of these conditions together with the conflicting findings reported in the literature (e.g., no general consensus on the effect of water) into account, the necessity to develop a solid understanding of the relation between these parameters and the adsorption performance of the sorbents is evident for the development of a feasible material. For this purpose, literature data corresponding to the adsorption performances of various amine-modified mesoporous silica are tabulated and analyzed comparatively in this section, in order to obtain a rough correlation about how these parameters affect the performance of materials. For the evaluation of performances, adsorption capacities along with amino efficiencies (CO2/N ratio) will be used. Although there is a vast majority of data available in the literature at various temperatures and pressure ranges, only the data corresponding to certain pressure (0.10–0.15 bar and around 1 bar) and temperature (45–1001C) ranges were considered here, to be able to perform a rough comparative analysis of such a complicated system without having to use complex modeling tools. Table 1 presents the data used for comparative analysis, whereas the whole data extracted from literature (in the mentioned pressure and temperature range) are provided in the Appendix.

2.27.4.1

mesoporous silica support-41 or Santa Barbara amorphous-15?

MCM-41 and SBA-15 are the most frequently considered mesoporous silica structures for the amine functionalization for CO2 capture purposes. Based on the literature review presented above, silica type is believed to affect the adsorption performances of the sorbents through the physical properties of the support only, meaning that as long as the pore size and pore diameter are identical, adsorption properties of two similar adsorbents should also be comparable under complementary conditions independent of the support type. When TRI was grafted onto MCM-41 (data 36), whose pore size, pore volume, surface area, and amine content were reported to be 5.1 nm, 1.28 cm3/g, 894 m2/g and 4.86 mmol/g, respectively; adsorption capacity was measured to be 1.27 mmol/g, while the amino efficiency was estimated as 0.26 at 451C and 0.15 bar CO2 pressure [98]. If TRI was used to modify SBA-15 with 8.9 nm pore size, 1.1 cm3/g pore volume and 775 m2/g surface area, on the other hand, it could adsorb 1.38 and 1.03 mmol/g CO2 for amine contents of 5.21 (data 80) and 4.5 mmol/g (data 78), respectively [111]. The corresponding amino efficiencies were calculated as 0.27 and 0.23, respectively. At higher pressures of CO2 (B1 bar), when APTMS (Data 34 and 85) was loaded onto the MCM-41 and SBA-15 supports, the properties of which were mentioned above, CO2 uptakes of the two sorbents were recorded as 0.87 mmol/g for MCM-41 and 0.97 mmol/g for SBA-15 at 451C [98,111]. Similar results were arrived for AEAPS-modified sorbents when data 35 and 87 were evaluated together. Eventhough the physical properties of the two supports were not absolutely identical for the cases considered, they might be regarded as being close enough to suggest that employment of MCM-41 or SBA-15 did not lead to significant differences on the adsorption performances of the two resulting sorbents, as long as the other parameters are comparable.

2.27.4.2

Amine Type and Preparation Method

Effect of amine type on the adsorption performances of various sorbents has indeed been investigated by various researches. As it is clear from the comparison of data 12 and 20, adsorption performances of APTMS and MAPS differ significantly from each other at 551C and 0.13 bar CO2 partial pressure, when they were grafted onto identical supports with comparable amine contents (4.6 and 3.67 mmol N/g, respectively, for APTMS and MAPS), which may be attributed to the higher reactivity of primary amino groups, as it was claimed in the corresponding study [96]. Accordingly, the amine structures investigated in the literature will be evaluated here based on their molecular structures. When APTMS and TRI were loaded onto the same SBA-15 support with similar amine loadings (1.9 mmol/g for APTMS and 1.8 mmol/g for TRI) through co-condensation, adsorption capacities of the two sorbents were found to be identical at 451C and B0.12 bar CO2 partial pressure, although APTMS involved only primary amines, while TRI had one primary and two secondary amino groups (data 90 and 91). However, both the adsorption capacities and the estimated amino efficiencies were very small, stemming from the inaccessibility of the amino groups due to the incorporation of them into the walls of the silica support [111]. Since it is not known exactly which amino groups are unattainable, comparison of these two adsorption data does not really provide any useful insights about the role of amino moieties on the adsorption capacity. For APTES and TRI-grafted SBA-15 samples with similar nitrogen contents, on the other hand, amounts of CO2 captured by the two sorbents differed notably at 601C and 0.15 bar CO2 partial pressure (data 68 and 72), APTES-SBA-15 achieving almost twice the CO2 uptake of TRI-SBA-15. This may be due to the higher reactivity of the primary groups but may also originate

902 CO2 Capturing Materials

Table 1 Ref.

86 86 86 94 94 96 96 96 96 97 97 97 97 97 97 98 98 98 98 98 98 101 108 108 109 109 109

Literature data used for comparative assessment of various sorbents

#

1 2 3 4 5 8 12 16 20 26 27 28 24 31 33 34 35 36 38 40 42 57 61 64 68 71 72

Support

MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15

Amine

PEI PEI PEI TRI TRI APTMS APTMS APTMS MAPS APTES APTES APTES MAPS TRI TRI APTMS AEAPS TRI PEI PEHA TEPA PEI APTMS APTMS APTES AEAPS TRI

N content (mmol/g)

11.67 11.67 11.67 7.9 7.9 4.6 4.6 4.6 3.67 3.16 3.16 3.16 3.67 6.39 6.39 2.43 4.00 4.86 6.23 6.07 2.57 5.86 2.2 3.2 2.61 4.61 2.75

Molar amounts of amino groupsa Pb

3.49 3.49 3.49 1 1 1 1 1 0 1 1 1 0 1 1 1 1 1 4.65 2 2 4.65 1 1 1 1 1

Sc

6.98 6.98 6.98 2 2 0 0 0 1 0 0 0 1 2 2 0 1 2 9.3 4 3 9.3 0 0 0 1 2

Td

3.49 3.49 3.49 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4.65 0 0 4.65 0 0 0 0 0

Support properties

Peraparation conditions

Dp (nm)

Vp (cm3/g)

SBET (m2/g)

Calcination

Prep. meth.e

2.75 2.75 2.75 11.7 11.7 11 11 11 11 30 30 30 11 30 30 5.1 5.1 5.1 5.1 5.1 5.1 8.9 7.4 7.6 6 6 6

1 1 1 3.09 3.09 2.81 2.81 2.81 2.81 2.6 2.6 2.6 2.81 2.6 2.6 1.28 1.28 1.28 1.28 1.28 1.28 1.1 0.92 1.04 1.07 1.07 1.11

1480 1480 1480 1230 1230 1153 1153 1153 1153 1045 1045 1045 1153 1045 1045 894 894 894 894 894 894 775 824 786 820 820 910

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No Yes Yes Yes

I I I G G G G G G G G G G G G G G G I I I I G G G G G

Operational variables Tads (1C)

pCO2 (bar)

50 75 100 45 55 40 55 70 55 45 60 75 70 45 60 45 45 45 45 45 45 45 50 50 60 60 60

1.00 1.00 1.00 1.02 1.01 0.13 0.13 0.13 0.13 0.12 0.14 0.14 0.13 1.00 1.00 1.00 1.00 0.15 1.00 1.00 1.00 0.13 0.15 0.15 0.15 0.15 0.15

Adsorption performance CO2 capacity (mmol/g) 1.00 2.55 2.50 2.48 2.29 1.98 1.85 1.53 0.75 1.38 1.34 1.21 0.51 2.16 2.34 0.87 1.51 1.27 1.50 1.35 0.55 1.11 0.84 1.09 0.66 1.36 0.35

CO2/N 0.09 0.22 0.21 0.31 0.29 0.43 0.40 0.33 0.20 0.44 0.42 0.38 0.14 0.34 0.37 0.36 0.38 0.26 0.24 0.22 0.21 0.19 0.38 0.34 0.25 0.30 0.13

109 111 111 111 111 111 111 111 111 111 111 112 112 112 112 112 112 112 112 112 112 113 113 113 113 114 114 117 117 117 117 117

73 78 80 82 83 84 85 86 87 90 91 94 96 98 102 103 104 105 106 107 108 113 114 115 117 121 123 135 138 139 140 141

SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15

TRI TRI TRI TRI TRI APTMS APTMS AEAPS AEAPS APTMS TRI PEI PEI TEPA TRI APTMS PEI PEI PEI PEI TEPA APTES APTES PEI PEI PEI TEPA AEAPS TRI TRI TRI TRI

4.85 4.50 5.21 5.71 5.71 2.57 2.57 4.07 4.07 1.9 1.8 2.43 5.93 2.57 3.79 2.64 2.43 4.14 5.93 11.86 2.57 3.55 3.55 6.09 6.09 6.43 6.71 3.25 3.2 3.68 3.68 3.68

1 1 1 1 1 1 1 1 1 1 1 4.65 4.65 2 1 1 4.65 4.65 4.65 4.65 2 1 1 3.49 3.49 4.65 2 1 1 1 1 1

2 2 2 2 2 0 0 1 1 0 2 9.3 9.3 3 2 0 9.3 9.3 9.3 9.3 3 0 0 6.98 6.98 9.3 3 1 2 2 2 2

0 0 0 0 0 0 0 0 0 0 0 4.65 4.65 0 0 0 4.65 4.65 4.65 4.65 0 0 0 3.49 3.49 4.65 0 0 0 0 0 0

6 8.9 8.9 8.9 8.9 8.9 8.9 8.9 8.9 8.9 8.9 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 7.95 7.95 7.95 7.95 15.2 15.2 6.3 6.3 6.3 6.3 6.3

1.11 1.1 1.1 1.1 1.1 1.1 1.1 1.1 1.1 1.1 1.1 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 1.7 1.7 1.7 1.7 1.18 1.18 0.61 0.61 0.61 0.61 0.61

910 775 775 775 775 775 775 775 775 775 775 587 587 587 587 587 587 587 587 587 587 908 908 908 908 428 428 737 737 737 737 737

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No Yes Yes Yes Yes Yes

G G G G G G G G G C C I I I G G I I I I I G G I I I I G G G G G

60 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 60 45 60 45 45 60 60 40 60 80

0.15 0.15 0.15 0.15 1.00 0.15 1.00 0.15 1.00 0.12 0.11 1.00 1.00 1.00 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.13 0.13 0.10 0.10 1.00 1.00 0.15 0.15 0.10 0.10 0.10

1.1 1.03 1.38 1.36 1.82 0.75 0.97 1.13 1.34 0.04 0.04 0.47 1.70 0.45 1.10 0.86 0.24 0.44 1.27 2.05 0.21 1.39 1.25 1.01 0.97 1.83 2.16 1.73 1.5 1.65 1.62 1.58

0.23 0.23 0.27 0.24 0.32 0.29 0.38 0.28 0.33 0.02 0.02 0.19 0.29 0.18 0.29 0.33 0.10 0.11 0.21 0.17 0.08 0.39 0.35 0.17 0.16 0.28 0.32 0.53 0.47 0.45 0.44 0.43

a

Calculated for 1 mol of amine structure, total molar amount of amino groups in PEI was calculated by dividing molecular weight by the weight of the repeating unit, branched PEI involves 25% primary, 50% secondary, 25% tertiary groups. Primary. c Secondary. d Tertiary. e Preparation method, I for impregnation, C for co-condensation, G for grafting. b

CO2 Capturing Materials 903

904

CO2 Capturing Materials

from the steric hindrance or diffusion resistance imposed by the longer chain of the TRI molecule, as claimed in the data source [109]. Consideration of data 71 and 73 along with 135 and 138 revealed that for similar amine contents at 601C and 0.15 bar CO2 partial pressure, both the adsorption capacity and the amino efficiency were slightly better for AEAPS-SBA-15 with respect to those of TRI grafted SBA-15 sample; while both sorbents achieved comparable amounts of CO2 adsorption at 451C and the same CO2 partial pressure (data 86 and 102). This variation in the relative performances of the two samples with respect to temperature may be an indication of changing attitudes of the sorbents corresponding to temperature. It was indeed previously shown that CO2 adsorption on TRI-SBA-15 was thermodynamically controlled, i.e., capacity declined with increasing temperature [117], yet there were no records of CO2 uptake of AEAPS-SBA-15 with changing temperature. Hence no exact conclusions could be drawn regarding the different relative capacities of the two sorbents at two distinct temperature values. When the adsorption performances of PEI-impregnated samples were compared to those of TEPA and PEHA-impregnated ones (data 38 and 40, 94 and 98, 107 and 110, 121 and 123), it was realized that for similar amounts of nitrogen content both sorbents achieved almost identical amounts of CO2 adsorption at 451C. If the distribution of amino groups within each structure was taken into consideration (TEPA has the highest fraction of primary amines), it may be claimed that primary amino groups were not the only structures playing active roles in the adsorption of CO2. Thus, despite the much frequently pronounced superiority of the primary amino groups, comparative analysis of the adsorption data extracted from literature did not provide very determining conclusions about the effect of molecular structure on the adsorption performance. Additional variables like the viscosity or the chain length of the amine structures may also be involved to model this complicated data set in order to arrive at more precise results. Apart from various amine structures, effect of preparation method on the achieved adsorption amounts has also been investigated extensively. Generally, as stated in the previous section, impregnated samples resulted in higher amine contents but suffer from low thermal stability due to the weak interactions between the support and the amine groups. When similar amounts of amine containing samples were evaluated together (data 84 and 108, 84 and 104, 103 and 108, 86 and 105, 78 and 105, 57 and 82, 34 and 42, 83 and 96), grafted samples were almost always superior to the impregnated ones in terms of adsorption capacity and amino efficiency.

2.27.4.3

Physical Properties of Support

Previous studies affirmed that pore size and pore volume influenced the adsorption capacity of the sorbents significantly due to an increase in the amount of amine loaded onto the support [88,93]. Indeed it may be speculated that larger pore diameters and pore volumes enhanced the adsorption performances of the sorbents only because they improved the amine content of the samples. Accordingly, it may be expected that adsorption capacities of two samples should not differ significantly from each other as long as their amine contents are identical but no data pair to confirm this idea could be found. Besides, experiments performed at 251C under 5% CO2 atmosphere exhibited that use of PE-MCM-41 rather than conventional MCM-41 led to considerable improvements in the adsorption capacity eventhough the amine contents of the two resulting samples were similar, suggesting that larger pores might indeed provide advances in terms of diffusion as well [92].

2.27.4.4

Calcination

Experiments performed to reveal the effect of calcination on adsorption performances of various samples asserted that silanol density is much higher in the absence of calcination, which brought about higher amine contents leading to better adsorption performance of the sample [106,107]. At 501C and 0.15 bar CO2 partial pressure, APTMS-grafted onto calcined SBA-15 with an amine content of 2.2 mmol N/g could adsorb 0.84 mmol CO2 /g, while APTMS-tethered onto ethanol extracted support with an amine content of 3.2 mmol N/g bound 1.09 mmol CO2/g (data 61 and 64). Almost equal amino efficiencies of the two sorbents offered that the calcination affected the performance of adsorbents through the amine content and as long as the nitrogen amounts involved in the sorbents are comparable, their adsorption performances should also not differ from each other significantly. To further verify this, calcination effect should be evaluated for impregnated samples but no such data pair was available in the set to verify this proposal.

2.27.4.5

Operational Variables

Temperature is one of the variables, the effect of which has been investigated the most. For APTMS, APTES, and MAPS-grafted samples adsorption was shown to be under thermodynamic control (data 26, 27 and 28; 113 and 114; 8, 12 and 16; 20 and 24), i.e., adsorption capacity decreased with increasing temperature as it was demonstrated by previous experiments [96,97,113]. Consideration of data 139, 140 and 141 together with 4 and 5, on the other hand, illustrated that adsorption capacity of TRI functionalized samples diminished at elevated temperature [94,117], while data 31 and 33 indicated just the opposite, i.e., adsorption was diffusion limited [97]. Adsorption behavior of PEI-grafted samples with increasing temperature varied as well. Data points 1, 2 and 3 illustrated that adsorption capacity was favored at higher temperatures [86], whereas data 115 and 117

CO2 Capturing Materials

905

suggested that CO2 uptake diminished when temperature rose from 45 to 601C but the change was not so noticeable [113]. Based on these observations, it may be argued that for smaller pore sizes and larger amine structures diffusion of CO2 could be restricted and as a result an increment might be observed in the adsorbed amounts upon a temperature increase. However, more data are required for the verification of this argument. Although only certain regions of CO2 partial pressure range were considered, it may be claimed that adsorption capacity was improved with rising pressure, the enhancement being more pronounced in the low-pressure regions that was typical of chemisorption. Finally, the effect of moisture on the adsorption capacity could not be evaluated here since there were no moist data available at the selected regions of temperature and pressure. Finally, despite the comparative analysis of the literature data performed here provided some basic insights into the CO2 adsorption by amine-functionalized mesoporous silica sorbents, more complicated modeling tools are required for the establishment of more accurate correlations.

2.27.5

Future Directions

Extensive research carried out in the field of CO2 capture displayed that employment of amines in the adsorption or absorption of the gas from high-emission point sources is one of the most feasible ways of capturing CO2 due to the high interaction of the amino groups with CO2. Although amine scrubbing is still the state-of-the-art technology in the field, it seems that its industrial use in the near future is improbable due to the challenges associated with the high regeneration costs. Use of amine functionalized adsorbents emerge as an alternative for aqueous amine systems but development of solid sorbents is in the early phases of research that concentrates mostly on the evaluation of capture performances of various materials in terms of capacity and kinetics. However, there is a huge variety in the available amine structures and testing the capture characteristics of all those molecules requires tedious and expensive experiments. Rational design of solid sorbents followed by fast screening methods to evaluate their performances may eliminate these costly experiments to a great extent but a deep understanding of the structure and activity relationship of the amines together with the complex relation between the stated parameters and the adsorption performances are required for the design of potential capture materials and at this point, even the way how temperature and moisture affect the CO2 uptakes of various sorbents is not explicit yet. If the factors determining the adsorption performance of various structures are resolved through molecular modeling used in combination with experiments and complex modeling tools, then the design of sorbents with the desired properties may be achieved accordingly. Once potential capture materials are developed, they should be tested under realistic conditions, not only in terms of capacity and kinetics, but also in terms of thermal and cyclic stability.

2.27.6

Closing Remarks

CO2 is one of the anthropogenic greenhouse gases, the increasing concentration of which is held responsible for the climate change observed for the last years. These emissions are mostly originated from the global energy supply sector and it seems that concentration of CO2 will go on increasing in the atmosphere without the implementation of supportive and effective environmental policies that regulates the greenhouse gas emissions by governments since the dominance of the fossil fuels in the energy market is likely to continue in the near-future. In this manner, development of highly efficient technologies that would effectively capture CO2 without bringing too much extra production costs in the energy market becomes inevitable and capture from large point sources like coal-fired power stations by using amine scrubbing and its subsequent sequestration is one of the potential ways of achieving this. The benchmark technology in the field of CO2 scrubbing is the absorption by aqueous MEA solutions. However, its large-scale use is restricted by the high regeneration costs stemming from the high-energy requirement of the heating of CO2-rich solution to regeneration temperature, vaporization of the solvent and regeneration of the amine. Since MEA binds CO2 in the form of very stable carbamates, alternative amine structures that would lead to less stable reaction products were considered and tertiary along with the sterically hindered amines were found to be promising in terms of reduced product stability and more favorable reaction stoichiometry. However, both structures reacted very slowly compared to MEA, which restricted their use as single amine solutions but promoted their adoption in blends with other amines with much faster reaction kinetics. In addition to amine blends, ionic liquids and structures containing more than one amino group have also been investigated much recently but no single material that would overcome all those problems associated with the existing technology could be identified, so the search for better capture materials is still going on. Employment of amine-functionalized solid sorbents has emerged as a potential solution, combining high reactivity of amine structures with reduced regeneration energy due to the absence of water. For that purpose various supports like ordered mesoporous silica were modified with diverse amine structures through grafting, co-condensation or impregnation and tested for their adsorption performances. Eventhough the adsorption capacities were significantly enhanced upon amine functionalization, this method is still under development and realistic evaluation of adsorption technology including the economics is still absent.

906

Appendix

Ref.

#

Adsorption data extracted from literature Support

Amine

N content (mmol/g)

Molar amounts of amino groupsa Pb

86 86 86 94 94 96 96 96 96 96 96 96 96 96 96 96 96 96 96 96 96 96 96 96 96 97 97 97 97 97 97 97 97 98 98 98 98

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37

MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41

PEI PEI PEI TRI TRI APTMS APTMS APTMS APTMS APTMS APTMS APTMS APTMS APTMS APTMS APTMS APTMS MAPS MAPS MAPS MAPS MAPS MAPS MAPS MAPS APTES APTES APTES APTES TRI TRI TRI TRI APTMS AEAPS TRI PEI

11.67 11.67 11.67 7.9 7.9 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 4.6 3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.67 3.16 3.16 3.16 3.16 6.39 6.39 6.39 6.39 2.43 4.00 4.86 2.25

3.49 3.49 3.49 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 4.65

Sc

6.98 6.98 6.98 2 2 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 0 0 0 0 2 2 2 2 0 1 2 9.3

Td

3.49 3.49 3.49 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4.65

Support properties Dp (nm)

Vp (cm3/g)

SBET (m2/g)

2.75 2.75 2.75 11.7 11.7 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 11 30 30 30 30 30 30 30 30 5.1 5.1 5.1 5.1

1 1 1 3.09 3.09 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.81 2.6 2.6 2.6 2.6 2.6 2.6 2.6 2.6 1.28 1.28 1.28 1.28

1480 1480 1480 1230 1230 1153 1153 1153 1153 1153 1153 1153 1153 1153 1153 1153 1153 1153 1153 1153 1153 1153 1153 1153 1153 1045 1045 1045 1045 1045 1045 1045 1045 894 894 894 894

Peraparation conditions Calcination

Prep. meth.e

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

I I I G G G G G G G G G G G G G G G G G G G G G G G G G G G G G G G G G I

Operational variables Tads (1C) 50 75 100 45 55 40 40 40 40 55 55 55 55 70 70 70 70 55 55 55 55 70 70 70 70 45 60 75 75 45 45 60 60 45 45 45 45

pCO2 (bar) 1.00 1.00 1.00 1.02 1.01 0.10 0.12 0.13 0.98 0.10 0.12 0.13 0.98 0.10 0.12 0.13 0.98 0.10 0.12 0.13 0.98 0.10 0.11 0.13 0.98 0.12 0.14 0.14 1.00 0.14 1.00 0.15 1.00 1.00 1.00 0.15 1.00

Adsorption performance CO2 capacity (mmol/g) 1.00 2.55 2.50 2.48 2.29 1.95 1.97 1.98 2.33 1.82 1.84 1.85 2.13 1.49 1.52 1.53 1.91 0.69 0.72 0.75 1.21 0.47 0.49 0.51 0.96 1.38 1.34 1.21 1.49 1.84 2.16 1.96 2.34 0.87 1.51 1.27 0.56

CO2/N 0.09 0.22 0.21 0.31 0.29 0.42 0.43 0.43 0.51 0.40 0.40 0.40 0.46 0.32 0.33 0.33 0.42 0.19 0.20 0.20 0.33 0.13 0.13 0.14 0.26 0.44 0.42 0.38 0.47 0.29 0.34 0.31 0.37 0.36 0.38 0.26 0.25

CO2 Capturing Materials

Table A1

38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83

MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 MCM-41 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15

PEI PEI PEHA PEHA TEPA TEPA APTMS þ PEI APTMS þ PEI APTMS þ PEI APTMS þ PEHA APTMS þ PEHA APTMS þ PEHA TRI þ PEI TRI þ PEI TRI þ PEI TRI þ PEHA TRI þ PEHA TRI þ PEHA TRI þ TEPA PEI APTES APTES APTES APTMS APTMS APTMS APTMS APTMS APTES APTES APTES AEAPS AEAPS AEAPS TRI TRI TRI MAPS DMAPS AEAPMS TRI TRI TRI TRI TRI TRI

6.23 10.01 6.07 9.36 2.57 7.00 3.93 7.50 10.86 4.07 10.86 11.29 6.07 8.57 12.00 6.71 9.00 11.64 12.86 5.86 2.56 2.56 2.56 2.2 2.2 3.2 3.2 3.2 1.11 2.57 2.61 2.26 3.76 4.61 2.75 4.85 5.8 1.88 1.79 3.52 4.50 4.50 5.21 5.21 5.71 5.71

4.65 4.65 2 2 2 2 5.65 5.65 5.65 3 3 3 5.65 5.65 5.65 3 3 3 3 4.65 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 0 0 1 1 1 1 1 1 1

9.3 9.3 4 4 3 3 9.3 9.3 9.3 4 4 4 11.3 11.3 11.3 6 6 6 5 9.3 0 0 0 0 0 0 0 0 0 0 0 1 1 1 2 2 2 1 0 1 2 2 2 2 2 2

4.65 4.65 0 0 0 0 4.65 4.65 4.65 0 0 0 4.65 4.65 4.65 0 0 0 0 4.65 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0

5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 5.1 8.9 9.2 9.2 9.2 7.4 7.4 7.6 7.6 7.6 6 6 6 6 6 6 6 6 6 6 6 6 8.9 8.9 8.9 8.9 8.9 8.9

1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.28 1.1 0.95 0.95 0.95 0.92 0.92 1.04 1.04 1.04 1.11 1.11 1.07 1.11 1.11 1.07 1.11 1.11 1.07 1.11 1.11 1.11 1.1 1.1 1.1 1.1 1.1 1.1

894 894 894 894 894 894 894 894 894 894 894 894 894 894 894 894 894 894 894 775 582 582 582 824 824 786 786 786 910 910 820 910 910 820 910 910 820 910 910 910 775 775 775 775 775 775

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No Yes Yes No No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

I I I I I I GþI GþI GþI GþI GþI GþI GþI GþI GþI GþI GþI GþI GþI I G G G G G G G G G G G G G G G G G G G G G G G G G G

45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 65 65 65 50 50 50 50 50 60 60 60 60 60 60 60 60 60 60 60 60 45 45 45 45 45 45

1.00 0.98 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.13 0.10 0.15 1.00 0.15 1.00 0.11 0.15 1.00 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 1.00 0.15 1.00 0.15 1.00

1.50 2.20 1.35 2.00 0.55 1.59 1.39 1.45 1.14 1.50 2.05 2.37 1.79 1.15 1.12 1.56 1.44 2.06 2.13 1.11 0.45 0.61 0.99 0.84 1.24 1.04 1.09 1.31 0.15 0.52 0.66 0.26 0.87 1.36 0.35 1.1 1.58 0.25 0.05 0.91 1.03 1.34 1.38 1.73 1.36 1.82

907

0.24 0.22 0.22 0.21 0.21 0.23 0.35 0.19 0.10 0.37 0.19 0.21 0.30 0.13 0.09 0.23 0.16 0.18 0.17 0.19 0.17 0.24 0.39 0.38 0.56 0.33 0.34 0.41 0.14 0.20 0.25 0.12 0.23 0.30 0.13 0.23 0.27 0.13 0.03 0.26 0.23 0.30 0.27 0.33 0.24 0.32 (Continued )

CO2 Capturing Materials

98 98 98 98 98 98 98 98 98 98 98 98 98 98 98 98 98 98 98 101 107 107 107 108 108 108 108 108 109 109 109 109 109 109 109 109 109 109 109 109 111 111 111 111 111 111

Ref.

#

908

Table A1

Continued Support

Amine

Molar amounts of amino groupsa Pb

111 111 111 111 111 111 111 111 112 112 112 112 112 112 112 112 112 112 112 112 112 112 112 112 112 112 112 112 113 113 113 113 113 113 113 114 114 114 114

84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122

SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15

APTMS APTMS AEAPS AEAPS APTMS APTMS APTMS TRI TRI APTMS PEI PEI PEI PEI TEPA TEPA TEPA TEPA TRI APTMS PEI PEI PEI PEI TEPA TEPA TEPA TEPA APTES APTES APTES PEI PEI PEI PEI APTMS TRI PEI PEI

2.57 2.57 4.07 4.07 0.9 0.9 1.9 1.8 3.79 2.64 2.43 4.14 5.93 11.86 2.57 6.36 11.00 14.86 3.79 2.64 2.43 4.14 5.93 11.86 2.57 6.36 11.00 14.86 3.55 3.55 3.55 6.09 6.09 6.09 6.09 3.14 4.86 6.43 9.43

1 1 1 1 1 1 1 1 1 1 4.65 4.65 4.65 4.65 2 2 2 2 1 1 4.65 4.65 4.65 4.65 2 2 2 2 1 1 1 3.49 3.49 3.49 3.49 1 1 4.65 4.65

Sc

0 0 1 1 0 0 0 2 2 0 9.3 9.3 9.3 9.3 3 3 3 3 2 0 9.3 9.3 9.3 9.3 3 3 3 3 0 0 0 6.98 6.98 6.98 6.98 0 2 9.3 9.3

Td

0 0 0 0 0 0 0 0 0 0 4.65 4.65 4.65 4.65 0 0 0 0 0 0 4.65 4.65 4.65 4.65 0 0 0 0 0 0 0 3.49 3.49 3.49 3.49 0 0 4.65 4.65

Support properties Dp (nm)

Vp (cm3/g)

8.9 8.9 8.9 8.9 8.9 8.9 8.9 8.9 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 8.5 7.95 7.95 7.95 7.95 7.95 7.95 7.95 15.2 15.2 15.2 15.2

1.1 1.1 1.1 1.1 1.1 1.1 1.1 1.1 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 0.96 1.7 1.7 1.7 1.7 1.7 1.7 1.7 1.18 1.18 1.18 1.18

SBET (m2/g)

775 775 775 775 775 775 775 775 587 587 587 587 587 587 587 587 587 587 587 587 587 587 587 587 587 587 587 587 908 908 908 908 908 908 908 428 428 428 428

Peraparation conditions Calcination

Prep. meth.e

Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes No No No No

G G G G C C C C G G I I I I I I I I G G I I I I I I I I G G G I I I I G G I I

Operational variables Tads (1C) 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 45 60 45 45 60 60 45 45 45 45

pCO2 (bar) 0.15 1.00 0.15 1.00 0.12 1.02 0.12 0.11 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.15 0.10 0.13 0.13 0.10 0.13 0.10 0.13 1.00 1.00 1.00 1.00

Adsorption performance CO2 capacity (mmol/g) 0.75 0.97 1.13 1.34 0.09 0.21 0.04 0.04 1.52 1.06 0.47 0.64 1.70 2.23 0.45 1.60 2.58 2.51 1.10 0.86 0.24 0.44 1.27 2.05 0.21 1.26 2.28 2.61 1.36 1.39 1.25 1.01 1.05 0.97 1.01 1.49 1.62 1.83 3.14

CO2/N 0.29 0.38 0.28 0.33 0.10 0.23 0.02 0.02 0.40 0.40 0.19 0.15 0.29 0.19 0.18 0.25 0.23 0.17 0.29 0.33 0.10 0.11 0.21 0.17 0.08 0.20 0.21 0.18 0.38 0.39 0.35 0.17 0.17 0.16 0.17 0.48 0.33 0.28 0.33

CO2 Capturing Materials

N content (mmol/g)

114 114 114 114 114 114 115 115 115 115 117 117 117 117 117 117 117 117 117 117 117

123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143

SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15 SBA-15

TEPA TEPA APTMS þ PEI APTMS þ TEPA APTMS þ TEPA TRI þ TEPA APTMS APTMS þ PEI APTMS þ TEPA APTMS þ PEI APTES APTES AEAPS AEAPS TRI TRI TRI TRI TRI TRI TRI

6.71 10.07 7.64 7.93 10.93 9.29 2.00 7.00 7.57 10.71 1.68 1.54 3.25 2.81 3.68 3.2 3.68 3.68 3.68 3.68 3.68

2 2 5.65 3 3 3 1 5.65 3 5.65 1 1 1 1 1 1 1 1 1 1 1

3 3 9.3 3 3 5 0 9.3 3 9.3 0 0 1 1 2 2 2 2 2 2 2

0 0 4.65 0 0 0 0 4.65 0 4.65 0 0 0 0 0 0 0 0 0 0 0

15.2 15.2 15.2 15.2 15.2 15.2 15.2 15.2 15.2 15.2 6.3 6.3 6.3 6.3 6.3 6.3 6.3 6.3 6.3 6.3 6.3

1.18 1.18 1.18 1.18 1.18 1.18 1.18 1.18 1.18 1.18 0.61 0.61 0.61 0.61 0.61 0.61 0.61 0.61 0.61 0.61 0.61

428 428 428 428 428 428 428 428 428 428 737 737 737 737 737 737 737 737 737 737 737

No No No No No No No No No No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

I I GþI GþI GþI GþI G GþI GþI GþI G G G G G G G G G G G

45 45 45 45 45 45 45 45 45 45 60 60 60 60 60 60 40 60 80 40 80

1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.15 0.15 0.15 0.15 0.15 0.15 0.10 0.10 0.10 0.15 0.15

2.16 3.73 2.53 3.16 4.89 3.89 0.72 2.35 2.93 2.80 1.06 0.43 1.73 1.22 2.41 1.5 1.65 1.62 1.58 2.56 2.33

0.32 0.37 0.33 0.40 0.45 0.42 0.36 0.34 0.39 0.26 0.63 0.28 0.53 0.43 0.65 0.47 0.45 0.44 0.43 0.70 0.63

a

Calculated for 1 mol of amine structure, total molar amount of amino groups in PEI was calculated by dividing molecular weight by the weight of the repeating unit, branched PEI involves 25% primary, 50% secondary, 25% tertiary groups. Primary. c Secondary. d Tertiary. e Preparation Method, I for impregnation, C for co-condensation, G for grafting. b

CO2 Capturing Materials 909

910

CO2 Capturing Materials

References [1] Pachauri RK, Meyer LA, editors. IPCC: Climate Change 2014: synthesis report. Contribution of working groups I, II and III to the fifth assessment report of the Intergovernmental Panel on Climate Change [Core writing Team]. Geneva: IPCC; 2014. 151 pp. [2] IEA. CO2 emissions from fuel combustion. Paris: IEA; 2016. [3] Metz B, Davidson O, de Coninck HC, Loos M, Meyer LA, editors. IPCC IPCC special report on carbon dioxide capture and storage, prepared by working group III of the intergovernmental panel on climate change. Cambridge: Cambridge University Press; 2005. 442 pp. [4] Favre E. Membrane processes and postcombustion carbon dioxide capture: challenges and prospects. Chem Eng J 2011;171:782–93. [5] Caplow M. Kinetics of carbamate formation and breakdown. J Am Chem Soc 1968;90:6795–803. [6] Danckwertz PV. The reaction of CO2 with ethanolamines. Chem Eng Sci 1976;34:443–6. [7] Sartori G, Savage DW. Sterically hindered amines for CO2 removal from gases. Ind Eng Chem Fundam 1983;22:239–49. [8] da Silva EF, Svendsen HF. Computational chemistry study of reactions, equilibrium, and kinetics of chemical CO2 absorption. Int J Greenhouse Gas Control 2007;1:151–7. [9] Davran-Candan T. DFT Modeling of CO2 interaction with various aqueous amine structures. J Phys Chem A 2014;118:4582–90. [10] Chapel DG, Mariz CL. Recovery of CO2 from flue gases: commercial trends. In: Presented at the canadian society of chemical engineers annual meeting, saskatoon, saskatchewan, canada, october 4–6; 1999. [11] Veawab A, Tontiwachwuthikul P, Chakma A. Corrosion behavior of carbon steel in the CO2 absorption process using aqueous amine solutions. Ind Eng Chem Res 1999;38:3917–24. [12] Kohl AL, Nielsen R. Gas purification. Houston, TX: Gulf Professional Publishing; 1997. [13] Khan AA, Halder GN, Saha AK. Comparing CO2 removal characteristics of aqueous solutions of monoethanolamine, 2-amino-2-methyl-1-propanol, methyldiethanolamine and piperazine through absorption process. Int J Greenhouse Gas Control 2016;50:179–89. [14] Sakwattanapong R, Aroonwilas A, Veawab A. Behavior of reboiler heat duty for CO2 capture plants using regenerable single and blended alkanolamines. Ind Eng Chem Res 2005;44:4465–73. [15] Tontiwachwuthikul P, Meisen A, Lim CJ. Solubility of CO2 in 2-amino-2-methyl-1-propanol solutions. J Chem Eng Data 1991;36:130–3. [16] Park SH, Lee KB, Hyun JC, Kim SH. Correlation and prediction of the solubility of carbon dioxide in aqueous alkanolamine and mixed alkanolamine solutions. Ind Eng Chem Res 2002;41:1658–65. [17] Yih SM, Shen KP. Kinetics of carbon dioxide reaction with sterically hindered 2-amino-2-methyl-1-propanol aqueous solutions. Ind Eng Chem Res 1988;27:2237–41. [18] Alper E. Reaction mechanism and kinetics of aqueous solutions of 2-amino-2-methyl-1-propanol and carbon dioxide. Ind Eng Chem Res 1990;29:1725–8. [19] Park J-Y, Yoon SJ, Lee H, et al. Density, viscosity, and solubility of CO2 in aqueous solutions of 2-amino-2-hydroxymethyl-1,3-propanediol. J Chem Eng Data 2002;47:970–3. [20] Yoon SJ, Lee H, Yoon J-H, et al. Kinetics of absorption of carbon dioxide into aqueous 2-amino-2-ethyl-1,3-propanediol solutions. Ind Eng Chem Res 2002;41:3651–6. [21] Bougie F, Iliuta MC. Kinetics of absorption of carbon dioxide into aqueous solutions of 2-amino-2-hydroxymethyl-1,3-propanediol. Chem Eng Sci 2009;64:153–62. [22] Yoon J-H, Baek J-I, Yamamoto Y, Komai T, Kawamura T. Kinetics of removal of carbon dioxide by aqueous 2-amino-2-methyl-1,3-propanediol. Chem Eng Sci 2003;58:5229–37. [23] Shen KP, Li MH. Solubility of carbon dioxide on aqueous mixtures of monoethanolamine with methyldiethanolamine. J Chem Eng Data 1992;37:96–100. [24] Dawodu OF, Meisen A. Solubility of carbon dioxide in aqueous mixtures of alkanolamines. J Chem Eng Data 1994;39:548–52. [25] Critchfield J, Rochelle GT. CO2 absorption into aqueous MDEA and MDEA/MEA solutions. In: AIChE National Meeting, March 30, paper No.43e. Houston, TX: American Institute of Chemical Engineers; 1987. [26] Versteeg GF, Kuipers JAM, Van Beckum FPH, Van Swaaij WPM. Mass transfer with complex reversible chemical reactions-II. Parallel reversible chemical reactions. Chem Eng Sci 1990;45:183–97. [27] Rangwala HA, Morrell BR, Mather AE, Otto FD. Absorption of CO2 into aqueous tertiary amine/mea solutions. Can J Chem Eng 1992;70:482–90. [28] Hagewiesche DP, Ashour SS, Al-Ghawas HA, Sandall OC. Absorption of carbon dioxide into aqueous blends of monoethanolamine and N-methyldiethanolamine. Chem Eng Sci 1995;50:1071–9. [29] Liao C-H, Li M-H. Kinetics of absorption of carbon dioxide into aqueous solutions of monoethanolamine þ N-methyldiethanolamine. Chem Eng Sci 2002;57:4569–82. [30] Ramachandran N, Aboudheir A, Idem R, Tontiwachwuthikul P. Kinetics of the absorption of CO2 into mixed aqueous loaded solutions of monoethanolamine and methyldiethanolamine. Ind Eng Chem Res 2006;45:2608–16. [31] Sema T, Naami A, Fu K, et al. Comprehensive mass transfer and reaction kinetics studies of CO2 absorption into aqueous solutions of blended MDEA-MEA. Chem Eng J 2012;209:501–12. [32] Lin C-Y, Soriano AN, Liu M-H. Kinetics study of carbon dioxide absorption into aqueous N-methyldiethanolamine þ diethanolamine. J Taiwan Inst Chem Eng 2009;40:403–12. [33] Idem R, Wilson M, Tontiwachwuthikul P, et al. Pilot plant studies of the CO2 capture performance of aqueous MEA and mixed MEA/MDEA solvents at the University of Regina CO2 capture technology development plant and the Boundary Dam CO2 capture demonstration plant.Ind Eng Chem Res 2006;45:2414–20. [34] Li M-H, Chang B-C. Solubilities of carbon dioxide in water þ monoethanolamine þ 2-amino-2-methyl-1-propanol. J Chem Eng Data 1994;39:448–52. [35] Xiao J, Li C-W, Li M-H. Kinetics of absorption of carbon dioxide into aqueous solutions of 2-amino-2-methyl-1-propanol þ monoethanolamine. Chem Eng Sci 2000;55:161–75. [36] Wang H-M, Li M-H. Kinetics of absorption of carbon dioxide into aqueous solutions of 2-amino-2-methyl-1-propanol þ diethanolamine. J Chem Eng Jpn 2004;37:267–78. [37] Mandal BP, Bandyopadhyay SS. Absorption of carbon dioxide into aqueous blends of 2-amino-2-methyl-1-propanol and monoethanolamine. Chem Eng Sci 2006;61:5440–7. [38] Mandal BP, Biswas AK, Bandyopadhyay SS. Absorption of carbon dioxide into aqueous blends of 2-amino-2-methyl-1-propanol and diethanolamine. Chem Eng Sci 2003;58:4137–44. [39] Mandal BP, Guha M, Biswas AK, Bandyopadhyay SS. Removal of carbon dioxide by absorption in mixed amines: modeling of absorption in aqueous MDEA/MEA and AMP/MEA solutions. Chem Eng Sci 2001;56:6217–24. [40] Freeman SA, Dugas R, van Wagener DH, Nguyen T, Rochelle GT. Carbon dioxide capture with concentrated, aqueous piperazine. Int J Greenhouse Gas Control 2010;4:119–24. [41] Xu G-W, Zhang C-F, Qin S-J, Gao W-H, Liu H-B. Gas-liquid equilibrium in a CO2–MDEA–H2O system and the effect of piperazine on it. Ind Eng Chem Res 1998;37:1473–7. [42] Bishnoi S, Rochelle GT. Thermodynamic of piperazine/methyldiethanolamine/water/carbon dioxide. Ind Eng Chem Res 2002;41:604–12. [43] Adeosun A, El Hadri N, Goetheer E, Abu-Zahra MRM. Absorption of CO2 by amine blens solution: an experimental evaluation. Int J Eng Sci 2013;3:12–23. [44] Xu G, Zhang C, Qin S, Wang Y. Kinetics study on absorption of carbon dioxide into solutions of activated methyldiethanolamine. Ind Eng Chem Res 1992;31:921–7. [45] Bishnoi S, Rochelle GT. Absorption of carbon dioxide in aqueous piperazine/methyldiethanolamine. AIChE J 2002;48:2788–99.

CO2 Capturing Materials

911

[46] Svensson H, Hulteberg C, Karlsson HT. Heat of absorption of CO2 in aqueous solutions of N-methyldiethanolamine and piperazine. Int J Greenhouse Gas Control 2013; 17:89–98. [47] Yang Z-Y, Soriano AN, Caparanga AR, Li M-H. Equilibrium solubility of carbon dioxide in (2-amino-2-methyl-1-propanol þ piperazine þ water). J Chem Thermodyn 2010;42:659–65. [48] Dash SK, Samanta AN, Bandyopadhyay SS. Solubility of carbon dioxide in aqueous solution of 2-amino-2-methyl-1-propanol and piperazine. Fluid Phase Equilibr. 2011;307:166–74. [49] Dash SK, Samanta AN, Bandyopadhyay SS. Experimental and theoretical investigation of solubility of carbon dioxide in concentrated aqueous solution of 2-amino-2methyl-1-propanol and piperazine. J Chem Tehrmodyn 2012;51:120–5. [50] Seo DJ, Hong WH. Effect of piperazine on the kinetics of carbon dioxide with aqueous solutions of 2-amino-2-methyl-1-propanol. Ind Eng Chem Res 2000;39:2062–7. [51] Samanta A, Bandyopadhyay SS. Absorption of carbon dioxide into aqueous solutions of piperazine activated 2-amino-2-methyl-1-propanol. Chem Eng Sci 2009;64:1185–94. [52] Sun W-C, Yong C-B, Li M-H. Kinetics of the absorption of carbon dioxide into mixed aqueous solutions of 2-amino-2-methyl-1-propanol and piperazine. Chem Eng Sci 2005;60:503–16. [53] Dang H, Rochelle GT. CO2 absorption rate and solubility in monoethanolamine/piperazine/water. Sep Sci Technol 2003;38:337–57. [54] Bonenfant D, Mimeault M, Hausler R. Determination of the structural features of distinct amines important for the absorption of CO2 and regeneration in aqueous solution. Ind Eng Chem Res 2003;42:3179–84. [55] Bonenfant D, Mimeault M, Hausler R. Comparative analysis of carbon dioxide absorption and recuperation capacities in aqueous 2-(2-aminoethylamino)ethanol (AEE) and blends of aqueous AEE and N-methyldiethanolamine solutions. Ind Eng Chem Res 2005;44:3720–5. [56] Ma’mun S, Svendsen HF, Hoff KA, Julliussen O. Selection of new absorbents for carbon dioxide capture. Energy Convers Manage 2007;48:251–8. [57] Lepaumier H, Picq D, Carrette P-L. New amines for CO2 capture. I. Mechanisms of amine degradation in the presence of CO2. Ind Eng Chem Res 2009;48:9061–7. [58] Lepaumier H, Picq D, Carrette P-L. New amines for CO2 capture. II. Oxidative degradation mechanisms. Ind Eng Chem Res 2009;48:9068–75. [59] Kim I, Svendsen HF. Heat of absorption of carbon dioxide (CO2) in monoethanolamine (MEA) and 2-(aminoethyl)ethanolamine (AEEA) solutions. Ind Eng Chem Res 2007;46:5803–9. [60] Kim YE, Moon SJ, Yoon YI, et al. Heat of absorption and absorption capacity of CO2 in aqueous solutions of amine containing multiple amino groups. Sep Purif Technol 2014;122:112–8. [61] Hartono A, da Silva EF, Svendsen HF. Kinetics of carbon dioxide absorption in aqueous solution of diethylenetriamine (DETA). Chem Eng Sci 2009;64:3205–13. [62] Choi SY, Nam SC, Yoon YI, Park KT, Park S-J. Carbon dioxide absorption into aqueous blends of methyldiethanolamine (MDEA) and alkyl amines containing multiple amino groups. Ind Eng Chem Res 2014;53:14451–61. [63] Hartono A, Svendsen HF. Kinetics reaction of primary and secondary amine group in aqueous solution of diethylenetriamine (DETA) with carbon dioxide. Energy Procedia 2009;1:853–9. [64] Gao H, Xu B, Liu H, Liang Z. Effects of amine activators on aqueous N,N-diethylethanolamine solution for postcombustion CO2 capture. Energy Fuels 2016;30:7481–8. [65] Freemantle M. An Introduction to ionic liquids. Cambridge, UK: RSC Publishing; 2009. [66] Blanchard LA, Hancu D, Beckman EJ, Brennecke JF. Green processing using ionic liquids and CO2. Nature 1999;399:28–9. [67] Ramdin M, de Loos TW, Vlugt TJH. State-of-the-art of CO2 capture with ionic liquids. Ind Eng Chem Res 2012;51:8149–77. [68] Bates ED, Mayton RD, Ntai I, Davis JH. CO2 capture by a task-specific ionic liquid. J Am Chem Soc 2002;124:926–7. [69] Sanchez LMG, Meindersma GW, de Haan AB. Solvent properties of functionalized ionic liquids for CO2 absorption. Chem Eng Res Des 2007;85:31–9. [70] Mindrup EM, Schneider WF. Computational comparison of tethering strategies for amine functionalized ionic liquids. In: Seddon K, Rogers R, Plechkova N, editors. ACS symposium series. Washington, DC: American Chemical Society; 2009. p. 419–30. [71] Gurkan BE, de la Fuente JC, Mindrup EM, et al. Equimolar CO2 absorption by anion-functionalized ionic liquids. J Am Chem Soc 2010;132:2116–7. [72] Gurkan B, Goodrich BF, Mindrup EM, et al. Molecular design of high capacity, low viscosity, chemically tunable ionic liquids for CO2 capture. J Phys Chem Lett 2010;1:3494–9. [73] Seo S, Quiroz-Guzman M, deSilva MA, et al. Chemically tunable ionic liquids with aprotic heterocyclic anion (AHA) for CO2 capture. J Phys Chem B 2014;118:5740–51. [74] Wang C, Luo X, Luo H, et al. Tuning the basicity of ionic liquids for equimolar CO2 capture. Angew Chem Int Ed 2011;50:4918–22. [75] Wang C, Luo H, Jiang D-E, Li H, Dai S. Carbon dioxide capture by superbase-derived protonic ionic liquids. Angew Chem Int Ed 2010;49:5978–81. [76] Jessop PG, Heldebrant DJ, Li X, Eckert CA, Liotta CL. Reversible nonpolar-to-polar solvent. Nature 2005;436:1102. [77] Heldebrant DJ, Yonker CR, Jessop PG, Phan L. Organic liquid CO2 capture agents with gravimetric CO2 capacity. Energy Environ Sci 2008;487:487–93. [78] Sayari A, Belmabkhout Y, Serna-Guerrero R. Flue gas treatment via CO2 adsorption. Chem Eng J 2011;171:760–74. [79] Gray ML, Champagne KJ, Fauth D, Baltrus JP, Pennline H. Performance of immobilized tertiary amine solid sorbents for the capture of carbon dioxide. Int J Greenhouse Gas Control 2008;2:3–8. [80] Samanta A, Zhao A, Shimizu GKH, Sarkar P, Gupta R. Post-combustion CO2 capture using solid sorbents: A review. Ind Eng Chem Res 2012;51:1438–63. [81] Wang J, Huang L, Yang R, et al. Recent advances in solid sorbents for CO2 capture and new development trends. Energy Environ Sci 2014;7:3478–518. [82] Balasubramanian R, Chowdhury S. Recent advances and progress in the developments of graphene-based adsorbents for CO2 capture. J Mater Chem A 2015;3:21968–89. [83] Wang Q, Luo J, Zhong Z, Borgna A. CO2 capture by solid adsorbents and their applications: current status and new trends. Energy Environ Sci 2011;4:42–55. [84] Hoffman F, Cornelius M, Morell J, Froeba M. Silica-based mesoporous organic-inorganic hybrid materials. Angew Chem Int Ed 2006;45:3216–51. [85] Choi S, Drese JH, Jones CW. Adsorbent materials for carbon dioxide capture from large anthropogenic point sources. ChemSusChem 2009;2:796–854. [86] Xu X, Song C, Andersen JM, Miller BG, Scaroni AW. Novel polyethylenimine–modified mesoporous molecular sieve of MCM-41 type as high-capacity adsorbent for CO2 capture. Energy Fuels 2002;16:1463–9. [87] Heydari-Gorji A, Belmabkhout Y, Sayari A. Polyethlenimine-impregnated mesoporous silica: Effect of amine loading and surface alkyl chains on CO2 adsorption. Langmuir 2011;27:12411–6. [88] Franchi RS, Harlick PJE, Sayari A. Applications of pore-expanded mesoporous silica. 2. Development of a high-capacity, water-tolerant adsorbent for CO2. Ind Eng Chem Res 2005;44:8007–13. [89] Liu Z-I, Teng Y, Zhang K, Cao Y, Pang W-P. CO2 adsorption properties and thermal stability of different amine-impregnated MCM-41 materials. J Fuel Chem Technol 2013;41:469–76. [90] Liu Z, Teng Y, Zahang K, Chen H, Yang Y. CO2 adsorption performance of different amine-based siliceous MCM-41 materials. J Energy Chem 2015;24:322–30. [91] Wang X, Guo Q, Zhao J, Chen L. Mixed amine-modified MCM-41 sorbents for CO2 capture. Int J Greenhouse Gas Control 2015;37:90–8. [92] Harlick PJE, Sayari A. Applications of pore-expanded mesoporous silicas. 3. Triamine silane grafting for enhanced CO2 adsorption. Ind Eng Chem Res 2006;45:3248–55. [93] Harlick PJE, Sayari A. Applications of pore-expanded mesoporous silica. 5. Triamine grafted material with exceptional CO2 dynamic and equilibrium adsorption performance. Ind Eng Chem Res 2007;46:446–58. [94] Belmabkhout Y, Sayari A. Effect of pore expansion and amine functionalization of mesoporous silica on CO2 adsorption over a wide range of conditions. Adsorption 2009;15:318–28. [95] Serna-Guerrero R, Da’na E, Sayari A. New insights into the interactions of CO2 with amine-functionalized silica. Ind Eng Chem Res 2008;27:9406–12. [96] Sayari A, Belmabkhout Y, Da’na E. CO2 deactivation of supported amines: does the nature of amine matter? Lngmuir 2012;28:4241–7.

912

CO2 Capturing Materials

[97] Loganathan S, Tikmani M, Mishra A, Ghoshal AK. Amine tethered pore-expanded MCM-41 for CO2 capture: Experimental, isotherm and kinetic modeling studies. Chem Eng J 2016;303:89–99. [98] Sanz R, Callja G, Arencibia A, Sanz-Perez ES. CO2 capture with pore-expanded MCM-41 silica modified with amino groups by double functionalization. Micropor Mesopor Mater 2015;209:165–71. [99] Wang X, Chen L, Guo Q. Development of hybrid amine-functionalized MCM-41 sorbents for CO2 capture. Chem Eng J 2015;260:573–81. [100] Yue MB, Sun LB, Cao Y, et al. Promoting the CO2 adsorption in the amine-containing SBA-15 by hydroxyl group. Micropor Mesopor Mater 2008;114:74–81. [101] Sanz R, Calleja G, Arecibia A, Sanz-Perez ES. CO2 adsorption on branched polyethylenimine-impregnated mesoporous silica SBA-15. Appl Surf Sci 2010;256:5323–8. [102] Yan X, Zhang L, Zhang Y, Yang G, Yan Z. Amine-modified SBA-15: effect of pore structure on the performance for CO2 capture. Ind Eng Chem Res 2011;50:3220–6. [103] Son W-J, Choi J-S, Ahn W-S. Adsorptive removal of carbon dioxide using polyethylenimine-loaded mesoporous silica materials. Micropor Mesopor Mater 2008;113:31–40. [104] Heydari-Gorji A, Yang Y, Sayari A. Effect of pore length on CO2 adsorption over amine-modified mesoporous silicas. Energy Fuels 2011;25:4206–10. [105] Yan X, Komarneni A, Yan Z. CO2 adsorption on Santa Barbara Amorphous-15 (SBA-15) and amine-modified Santa Barbara Amorphous-15 (SBA-15) with and without controlled microporosity. J Colloid Interface Sci 2013;390:217–24. [106] Zheng F, Tran DN, Busche BJ, et al. Ethylenediamine-modified SBA-15 as regenerable CO2 sorbent. Ind Eng Chem Res 2005;44:3099–105. [107] Wang L, Ma L, Wang A, Liu Q, Zhang T. CO2 adsorption on SBA-15 modified by aminosilane. Chin J Catal 2007;28:805–10. [108] Wang L, Yang RT. Increasing selective CO2 adsorption on amine-grafted SBA-15 by increasing silanol density. J Phys Chem C 2011;115:21264–72. [109] Hiyoshi N, Yogo K, Yashima Y. Adsorption characteristics of carbon dioxide on organically functionalized SBA-15. Micropor Mesopor Mater 2005;84:357–65. [110] Ko YG, Shin SS, Choi US. Primary, secondary, and tertiary amines for CO2 capture: Designing for mesoporous CO2 adsorbents. J Colloid Interface Sci 2011;361:594–602. [111] Sanz R, Calleja G, Arencibia A, Sanz-Perez ES. Amino functionalized mesostructured SBA-15 silica for CO2 capture: Exploring the relation between adsorption capacity and the distribution of amino groups by TEM. Micropor Mesopor Mater 2012;158:309–17. [112] Sanz-Perez ES, Olivares-Marin M, Arencibia A, et al. CO2 adsorption performance of amino-functionalized SBA-15 under post-combustion conditions. Int J Greenhouse Gas Control 2013;17:366–75. [113] Vilarrasa-Garcia E, Moya EMO, Cecilia JA, et al. CO2 adsorption on amine modified mesoporous silicas: Effect of the progressive disorder of the honeycomb arrangement. Micropor Mesopor Mater 2015;209:172–83. [114] Sanz R, Calleja G, Arencibia A, Sanz-Perez ES. Development of high efficiency adsorbents for CO2 capture based on a double-functionalization method of grafting and impregnation. J Mater Chem A 2013;1:1956–62. [115] Sanz R, Calleja G, Arencibia A, Sanz-Perez ES. CO2 uptake and adsorption kinetics of pore-expanded SBA-15 double-functionalized with amino groups. Energy Fuels 2013;27:7637–44. [116] Zelenak V, Badanicova M, Cejka J, et al. Amine-modified ordered mesoporous silica: Effect of pore size. Chem Eng J 2008;144:336. [117] Chang F-Y, Chao K-J, Cheng H-H, Tan C-S. Adsorption of CO2 onto amine-grafted mesoporous silicas. Sep Purif Technol 2009;70:87–95. [118] Xu X, Song C, Miller BG, Scaroni AW. Influence of moisture on CO2 separation from gas mixture by a nanoporous adsorbent based on polyethylenimine-modified molecular sieve MCM-41. Ind Eng Chem Res 2005;44:8113–9. [119] Plaza MG, Pevida C, Arenillas A, Rubiera F, Pis JJ. CO2 capture by adsorption with nitrogen enriched carbons. Fuel 2007;86:2204–12. [120] Bazerra DP, Oliveira RS, Vieira RS, Cavalcante Jr. CL, Azevedo DCS. Adsorption of CO2 on nitrogen-enriched activated carbon and zeolite 13X. Adsorption 2011;17:235–46. [121] Wang J, Chen H, Zhou H. Carbon dioxide capture using polyethylenimine-loaded mesoporous carbons. J Environ Sci 2015;25:124–32. [122] Gibson JAA, Gromov AV, Brandani S, Campbell EEB. The effect of pore structure on the CO2 adsorption efficiency of polyamine impregnated porous carbons. Micropor Mesopor Mater 2015;208:129–39. [123] Kongnoo A, Intharapat P, Worathanakul P, Phalakornkule C. Diethanolamine impregnated palm shell activated carbon for CO2 adsorption at elevated temperatures. J Environ Chem Eng 2016;4:73–81. [124] Su F, Lu C, Cnen W, Bai H, Hwang JF. Capture of CO2 from flue gas via multiwalled carbon nanotubes. Sci Total Environ 2009;407:3017–23. [125] Hsu S-C, Lu C, Su F, Zeng W, Chen W. Thermodynamics and regeneration studies of CO2 adsorption on multiwalled carbon nanotubes. Chem Eng Sci 2010;65:1354–61. [126] Liu Q, Shi Y, Zheng S, et al. Amine-functionalized low-cost industrial grade multi-walled carbon nanotubes for the capture of carbon dioxide. J Energy Chem 2014;23:111–8. [127] Liu Q, Shi Y, Wang Q, et al. Carbon dioxide capture with polyethlenimine-functionalized industrial-grade multiwalled carbon nanotubes. Ind Eng Chem Res 2014;53:17468–75.

Relevant Websites https://sequestration.mit.edu Carbon Capture and Sequestration Technologies at MIT. http://www.ccsassociation.org/what-is-ccs/capture/ Carbon Capture and Storage Association. https://www3.epa.gov/climatechange/ccs/ Environmental Protection Agency. https://ec.europa.eu/clima/policies/lowcarbon_en European Commission. https://www.iea.org/topics/ccs/ International Energy Agency. https://www.netl.doe.gov/research/coal/carbon-capture U.S. Department of Energy.

2.28 Anti-Corrosive Materials Bekir S Yilbas, Ihsan-ul-Haq Toor, and Abdullah Al-Sharafi, King Fahd University of Petroleum and Minerals, Dhahran, Saudi Arabia r 2018 Elsevier Inc. All rights reserved.

2.28.1 Introduction 2.28.2 Some Aspects of Corrosion 2.28.2.1 Corrosion Measurement Techniques 2.28.2.1.1 Tafel extrapolation method 2.28.2.1.2 Linear polarization resistance 2.28.2.1.3 Electrochemical noise measurement 2.28.2.2 Electrochemical Impedance Spectroscopy 2.28.2.3 Corrosion Behavior of Stainless Steels and Alloying Elements 2.28.3 Results and Discussion 2.28.3.1 Corrosion Response of Laser Treated Alumina 2.28.3.2 Corrosion Response of Laser Treated Ti-6Al-4Al Alloy 2.28.3.3 Corrosion Response of Laser Treated Aluminum–Silicon Alloy 2.28.3.4 Corrosion Response of Laser Treated High Strength Low Alloy Steel 2.28.3.5 Corrosion Response of Laser Treated Copper Alloy (Bronze) 2.28.3.6 Corrosion Response of Laser Treated Hastelloy Alloy 2.28.4 Conclusions 2.28.5 Future Directions 2.28.6 Closing Remarks Acknowledgment References Further Reading Relevant Websites

Nomenclature ba bc E

2.28.1

Anodic Tafel constant (V) Cathodic Tafel constant (V) Polarization potential (V)

Ecorr Icorr Ie Rp

913 916 917 918 918 919 922 922 923 923 924 928 931 932 935 939 940 940 941 941 943 943

Corrosion potential (V) Corrosion current (A) Polarization current (A) Polarization resistance (V/A)

Introduction

Various metallic or composite materials are used for improved thermal energy storage and energy harvesting devices, in particular solar energy harvesting systems. Some of these materials include metallic nitrides or oxides. The metallic nitrides, such as titanium nitride, are mainly used as selective surfaces to harvest solar irradiation at high temperatures. However, when these surfaces are exposed to humid and hot environments with the presence of conductive ions, such as chlorine compounds, they suffer from electrochemical reactions leading to pitting corrosion. The surface, then, degrades its desired physical properties including absorption and emission characteristics. This, in turn, lowers the performance characteristics of the selective surface during energy harvesting. In addition, the efforts toward removal of the pit size and regaining the surface characteristics become extensive and very costly. Consequently, the replacement of the selective surface becomes unavoidable in such cases. In this case, in the humid air environments, the dissolution of dust particles in condensed water on selective surfaces triggers the electrochemical reactions. This, in turn, results in permanent damages on the surface. Although theses damages are localized in terms of the area and do not extend over the entire surface, these surfaces need to be replaced in order to sustain the efficient operation of the solar energy harvesting device. On the other hand, there are several processes that can be listed for nitriding the metallic materials such as titanium and its alloys. Some of these methods include gas nitriding, ion implantation, nitride coating via physical and/or chemical vapor deposition, laser gas assisted treatment, and similar. Laser gas assisted treatment has several advantages over the other methods. Some of these advantages include high processing speed, localized treatment, precision of operation, and low cost. Laser surface nitriding involves with high-temperature processing, which requires high power laser processing on the metallic surface. However, thermal stresses are developed due to the high cooling rates in the laser treated region during the high speed processing. This lowers the corrosion resistance of the laser gas nitride surfaces. Nevertheless, with control of laser pulses and assisting gas pressure,

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00256-X

913

914

Anti-Corrosive Materials

crack formation can be avoided. Since the nitride layer acts like a passive layer at the surface, it prevents the formation if pitting corrosion sites at the surface. In addition, alumina is also used in photovoltaic panel frames and in some other construction sites of the solar volumetric receivers. Alumina tiles have resistance to high temperature, corrosion, and wear. Alumina tiles are generally produced by alumina powders through sintering, which results in structural inhomogeneity in the tiles. In this case, irregular orientation of powders with varying sizes result in small cavities and pores in the structure, which limits the practical applications of the resulting tiles. One of the methods to improve such a structure in the surface region is to apply a control melting in this region using a laser beam. Due to the attainment of high temperature in the laser irradiated region, structural changes are unavoidable in this region. In laser controlled melting process, nitrogen is used as an assisting gas; in which case, aluminum nitride can be formed in the laser irradiated region. The corrosion response of the surface changes because of the structural modification and nitride species formation in the surface region. Considerable research studies were carried out to examine the corrosion resistance of alumina. The corrosion properties of plasma-sprayed alumina surfaces were examined by Harju et al. [1]. They showed that the dissolution-reprecipitation and surface site redistributions occurred after the corrosion tests; however, overall, only minor changes in surface properties resulted from this restructuring process. The laser cladding of Al2O3 on Mg alloy and wear and corrosion properties of the resulting surface was examined by Ya-Li et al. [2]. They showed that the hardness, wear, and corrosion resistance of the laser remelted coating were much higher than those of the plasma-sprayed coatings due to formation of dense column-like crystal structures after the laser treatment process. The laser surface treatment of Al2O3based refractory and surface chemical resistance characteristics were studied by Lawrence and Li [3]. They indicated that the use of the different lasers resulted in dissimilar microstructures in the surface region because of the different rate of solidification. The corrosion of alumina ceramics in acidic aqueous solution at high temperatures and pressures was investigated by Schacht et al. [4]. They showed that at high temperatures, intergranular corrosion was observed, provided that the dissolution of the grains did not occur. The laser treatment of aluminum alloys for pitting corrosion protection was studied by Chong et al. [5]. They showed that for the large area of the laser beam overlapping, the laser surface treatment had no detrimental effect on pitting potential, but the cellular layers that were formed on the surface were prone to pitting. Aluminum alloys are also widely used in photovoltaics as the frame material and the constructional support. This is because of their low densities and good resistance to corrosion. Although aluminum alloys are resistance to aqueous corrosion, in some cases, the pitting corrosion at the surface is unavoidable. Surface treatment toward generating a electrochemically passive layer has been a wide interest; however, the treatments may involve with expensive and timely processes. The laser controlled melting offers considerable advantages over the conventional surface treatment methods, such as plasma arc melting and induction heating, to generate the passive layer on the substrate surface. In this case, the formation of passive layer through altering the surface microstructure by using a laser gas assisted processing is possible, which, in turn, prevents the formation of pit nucleation sites at the surface. However, laser heating process involves with high cooling rates and, in some cases, excessive stress sites can be formed at the surface. This, in turn, causes formation of the microcracks while lowering the corrosion resistance due to breakdown of the passive layer at the treated surface. Laser surface treatment of aluminum alloys was carried out by Xu et al. [6]. They showed that selection of laser parameters were very important to obtain smooth single treated tracks and minimize the surface roughness due to overlapping of the tracks. Laser induced crystallization of amorphous silicon and aluminum thin films were investigated by Paduru et al. [7]. They indicated that the polycrystalline Al–Si layer was formed after laser processing and the crystallization rate increased both with laser power density and exposure time. Surface properties of Al–Si alloy irradiated by picosecond laser were studied by Zhu et al. [8]. They demonstrated that the laser scanning speed had significant effect on the uniform distribution of elemental composition in the treated region. Laser treatment of the Al–Si alloy surface for wear applications was carried out by McCay et al. [9]. Their findings revealed that the hardness of the treated surface increased with increasing silicon content; however, the segregation of the silicon, as pockets of high silicon content, had an adverse effect on the wear resistance of the treated surface. Laser treatment of aluminum alloys and structural changes in the treated layer was examined by Pokhmurs'ka et al. [10]. They showed that the influence of structural heterogeneity of the laser treated surface on the variation of corrosion potentials decreased as the corrosion activity of the medium increased. Investigation of the aluminum oxide and silicon nitride thin films for anticorrosion protection of surfaces was carried out by Qu et al. [11]. They demonstrated that the forming of aluminum oxide and silicon nitride films at the metallic surface provided good corrosion resistance of the surface. The corrosion behavior of aluminum alloy/silicon carbide composites was examined by Pardo et al. [12]. They showed that the presence of silicon carbide at the surface enhanced the corrosion resistance of the surface. Localized corrosion of silicon-reinforced aluminum composites in aqueous solution was studied by Ding and Hihara [13]. They indicated that the presence of silicon served as cathodic sites and partially exacerbated the corrosion at the surface. The corrosion resistance of silicon nitride bonded silicon carbide refractory blocks was studied by Etzion and Metson [14]. They demonstrated that the corrosion rate was enhanced as the pore sites at the surface increased. The influence of microstructure and composition on electrochemical behavior of Al–Si coatings on steel was examined by Schoukensa et al. [15]. The findings revealed that preferential dissolution of aluminum resulted in a shift in open current potential and it affected the subsequent anodic behavior of the surface. Localized corrosion of Al–Si carbide composite in chloride containing environment was investigated by Ahmad et al. [16]. They observed that pit sites could be correlated with the secondary phase particles, which contributed to the pitting behavior of the surface. Bronze is a copper alloy and it is used in solar thermal receivers due to its combination of high electrical and thermal conductivities. The corrosion and erosion can take place at the fluid–metal interface and this adverse effect accelerates for the surfaces where the nonuniform microstructures are present. In general grain refining, with uniform microstructures at the surface vicinity of the alloy, improves the corrosion and erosion resistance at the surface. One of the methods to improve the alloy

Anti-Corrosive Materials

915

microstructure in the surface vicinity is to introduce control melting incorporating the high power laser beam. In addition, laser treatment of the surfaces is involved with precision of operation, short treatment duration, shallow heat affected zone, and low cost. However, the proper selection of the laser parameters is vital in the surface treatment process to avoid the surface asperities because of the excessive heating. The surface asperities such as cavities and microcracks act like a defect site for accelerated corrosion at the surface. Some studies were carried out to examine laser treatment and corrosion resistance of copper alloys. Laser treatment of bronze surface and corrosion resistance was examined by Tang et al. [17]. They showed that the galvanic effect between the laser treated and as-received samples were small, which justified the use of laser surface alloying as a feasible method in the local surface treatment of bronze. The cavitation erosion resistance of laser treated bronze was studies by Kwok et al. [18]. Their findings revealed that the cavitation erosion resistance of the laser treated surface was improved by at most 7.5 times that of as-received bronze surface. In addition, laser treatment enhanced the corrosion resistance of the surface considerably. Electrochemical response of the laser treated bronze surface was examined by Klassen et al. [19]. They illustrated that the oxide formed at the surface during laser treatment behaved as the passivation film improving the corrosion resistance of the surface. The corrosion resistance of laser treated bronze surface was studied by Mazurkevich et al. [20]. They indicated that the laser treatment of the surface improved the corrosion resistance notably. The influence of laser treatment on the surface properties of copper alloys was investigated by Garbacz et al. [21]. They incorporated Raman Spectroscopy to examine the phase composition of the corroded layers at the laser treated surface. The corrosion characteristics of laser treated Ni–Al–bronze surface were examined by Kawazoe et al. [22]. Their findings revealed that Ni–Al–bronze had quenching characteristics closely related to that of steel and the corrosion resistance of the surface improved after the laser treatment process. Titanium alloys are widely used in energy and aerospace sectors due to their high strength to density ratio is high. Although alloys have low density, poor tribological properties of the surface limit the practical applications of the alloy in the wearing environments. The tribological properties of the alloy surface can be improved through surface treatment processes. There are many techniques for the treatment of the alloy surface such as gas nitriding, plasma assisted processes, and laser processing. Laser surface treatment has many advantages over the other techniques and some of these include high speed of operational, high precision, local treatment, and low cost. However, laser surface treatment is involved with high cooling rates, which in turn results in high-temperature gradients in the irradiated region of the workpiece. This results in attainment of high strain and stress levels in the irradiated region and influences the fracture toughness of the treated surface. This is particularly true for the cladded surfaces by a laser beam with the presence of hard particles. However, the stress levels can be reduced through multiple scanning of the surface via creating a self-annealing effect in the laser treated layer. In addition, the presence of hard particles modifies the surface texture of the treated layer, which is expected to cause an adverse effect on the electrochemical response of the laser treated surface when subjected to the electrolytic solution. Some studies related to electrochemical response of titatnium alloys were carried out previously [23–34]. The corrosion behavior of the titanium alloy treated by a laser shock processing was studied by Hua et al. [23]. They demonstrated that the average corrosion rate of TC11 titanium alloy surface, treated by a laser shock processing, was more than 50% lower than that of the untreated alloy surface. The corrosion resistance of laser machined titanium alloy sheet was investigated by Shanjin and Yang [24]. They indicated that the surface morphology and the corrosion resistance of the cut sections were changed for different assisting gases used. Electrochemical behavior of nano- and femtosecond laser textured titanium alloy was examined by Jeong et al. [25]. They showed that laser treatment altered the potentiodynamic and AC impedance response of the alloy surface because of the microstructural changes in the surface region. Laser surface alloying of titanium with nickel and palladium for increased corrosion resistance was carried out by Blanco-Pinzon et al. [26]. The findings revealed that the corrosion rates of the optimum palladium-alloyed surfaces were about two orders of magnitude lower than that of untreated titanium. The corrosion and nickel release of laser surface melted titanium alloy behavior in a salt solution were investigated by Cui et al. [27]. They indicated that the calcium-phosphate layer was observed on the surface of the samples after immersion in a salt solution for 15 days. A comprehensive review for corrosion behavior of laser nitrided titanium alloy was presented by Razavi et al. [28]. They included various aspects of corrosion response of the laser treated titanium alloy surface. Wear and corrosion properties of laser gas nitrided titanium alloy surface were studied by Zhang et al. [29]. They suggested that the performance and composition of the surface of the titanium alloy were significantly improved by laser gas assisted nitriding; in which case, improved corrosion and wear resistances were resulted as compared to the untreated alloy surface. The corrosion behavior of laser-alloyed copper with titanium fabricated under high power diode laser was examined by Wong et al. [30]. They demonstrated that improvement of corrosion resistance of the surface was attributed to the presence of titanium in the intermetallic and metallic phases to form the protective oxide layer at the surface. Laser surface modification to improve the corrosion resistance of titanium and titanium alloys was presented by Yu [31]. He covered the recent investigations of laser surface modification to improve the corrosion resistant of titanium and its alloys and included the microstructures and the properties for corrosion resistance. The corrosion and wear behavior of laser nitrided biomedical titanium and its alloys were studied by Sathish et al. [32]. They reported that the laser high speed processing of the alloy enabled smooth and crack free surfaces. Interfacial reactions in a titanium alloy with laser-embedded SiC particles and intergranular corrosion susceptibility of the alloy surface were investigated by Kooi and De Hosson [33]. They showed that the interfacial reaction between titanium and SiC largely improved wear resistance of the alloy and it did not have an adverse effect on the corrosion properties. Microstructural and corrosion evaluation of laser treated titanium alloy surface was studied by Geetha et al. [34]. They indicated that the laser treatment improved significantly the corrosion resistance of the surface due to the formation of fine grains at the surface. Some research studies reported earlier deal with the corrosion response of some metallic and nonmetallic alloys used in energy harvesting systems, the fundamental understanding of corrosion mechanisms and pit

916

Anti-Corrosive Materials

formation is necessary to further explore the topic. The following section provides the fundamental understanding of the corrosion process.

2.28.2

Some Aspects of Corrosion

The degradation of material properties due to its interaction with the environment is called corrosion. It is a natural process, as materials tend to go back to their most thermodynamically stable state (oxidized states, ores etc.) by reacting with the environment. However, the rate of corrosion or oxidation varies among metals. The metals can be free of oxidation or corrosion in inert atmosphere and/or vacuum. There are four important components of a corrosion cell and if any one of them is missing, corrosion will not take place. These four components/processes include anodic half-cell reaction, cathodic half cell reaction, electrical path, and the presence of electrolyte. This also provides a fundamental rule to protect the metals from corrosion, as by removing any one of the process, reaction will not take place. Let us consider a metal (M) which is immersed in its own solution (Fe in FeSO4 solution). After some time an equilibrium will be established (no net current flow) between metal and its ions which can be written as M ¼ Mzþ þ ze Here M is for the metal atoms, Mz þ is the metal ion z is the valence and electron is represented by e . The above reaction can be divided into two other reactions, namely, anodic and cathodi half-cell reactions, respectively. In the anodic/oxidation half-cell reaction (given below), metal is oxidized to give metal ions (cations) and electrons as below. M-Mzþ þ ze ðanodic half

cell reactionÞ

So the metal is oxidized or corrdoes diuring an anodic reaction. The electrons produced during the anodic eaction must be consumded by cathodic reaction/reduction reaction as Mzþ þ ze -M ðcathodic half

cell reactionÞ

There should not be any accumulation of changes during the corrosion reaction, i.e., electrons produced by anodic reaction, must be consumed by cathodic reaction. There can be many different cathodic reactions and few of them are given as

•

Hydrogen ions reduction in an acid (without oxygen): 2Hþ þ 2e -H2

•

Oxygen reduction in an acid (with oxygen): O2 þ 4Hþ þ 4e -2H2 O

•

Oxygen reduction in neutral solution. O2 þ 2H2 O þ 4e -4ðOH Þ

These cathodic/reduction reactions occur at the surface of cathode (cathodic site) and cathode itself does not participate in the reaction, i.e., it does not corrode during the process, rather is cathodically protected. Cathodic site or electrode just facilitates the reduction reaction by allowing the electron transfer for reduction reaction. Therefore, these cathodic materials must be capable of conducting the electrons and can be non-metallic as well. Hence, both the half-cell reactions have their own significance and are important for the corrosion process. Oxidation reaction causes the corrosion to occur by ionizing the metal to its metal ions. On the other hand, reduction reaction can be used for electroplating applications by depositing the ions in solution as metal on cathodic surface. Another important component of corrosion cell/process is the presence of electrolyte for ionic current conduction such as water, seawater and acidic/basic solutions. The conductivity of the solution will affect the corrosion rate, such as seawater is more corrosive than distilled water owing to its high conductivity. The fourth and last component of corrosion cell is the presence of electrical path for electron movement between anodic and cathode sites. If the electrons produced during oxidation reaction are unable to reach the cathode site, no corrosion will take place. So therefore, one of the corrosion protection strategies is to stop the electron transferring from anode to cathode. These anodic and cathodic reactions can occur on the surface of the same metal as well and it can be best explained using rust formation on steel surface when it is exposed to atmospheric conditions. Some part of the steel becomes anode and rest behaves as cathode. Electron produced by anodic reactions traveled within the same metal to its surface where they are consumed by one of many cathodic reactions discussed above such as reduction of dissolved oxygen or water to produce hydrogen gas and this type of electrochemical cell is best explained in Fig. 1. There are many reasons why some part of the same surface becomes anode and others cathode. Some of these factors are: (1) there can be some inherent defects/inclusions, different phases, microstructural variations and so on which can make some parts more active than the rest of the surface, (2) the location of these sites is not well defined and they can be very near and/or away from each other, (3) there can be some concentration variations over the surface, such as low to high oxygen at different locations over the surface, which can give rise to different anode and cathode site son the same surface.

Anti-Corrosive Materials

917

Aqueous solution (electrolyte) M(OH)2

Anodic reaction M→

M2+

+

2e−

Cathodic reaction 2OH−

M2+

1/2O2 + H2O + 2e− → 2OH−

e− e−

Metal

Fig. 1 Corrosion of a metal in an aerated environment (electrochemical cell is established between anodic and cathodic sites). Reproduced from Bardal E. Corrosion and protection. Berlin: Springer Science & Business Media; 2007.

1.4 1.2 e

1.0 0.8 0.6 Eo (V)

0.4

d

0.2 0.0

f

Passivity

−0.2

Corrosion

c

−0.4 −0.6

Corrosion

a

b

10−2

−0.8

10−4

−1.0

Immunity

−6

10

−1.2 1

2

3

4

5

6

7 8 pH

9 10 11 12 13 14

Fig. 2 The potential–pH (Pourbaix) equilibrium diagram of iron in water at 251C. Reproduced from Bardal E. Corrosion and protection. Berlin: Springer Science & Business Media; 2007.

It is important to note that the electron flow between anode and cathode sites is measured in terms of corrosion current, i.e., rate of electron production and consumption. This rate should be equal otherwise rather than charge conservation; there will be buildup of the charge, which will hinder reaction kinetics. In order to move the electrons from anode to cathode, there must be enough driving force or potential difference available; otherwise, electron flow will not occur. Thermodynamic potential can tell about the tendency of a reaction, i.e., reaction can occur spontaneously, non-spontaneously or in equilibrium/no reaction. Thermodynamic data in terms of “Pourbaix diagrams” can provide information about the possibility of different reactions for pure metals when immersed in water. Pourbaix diagrams are basically potential–pH diagrams, which can provide an information about possibility of corrosion, passivation and immunity of the metal in water (PB diagram of iron is shown in Fig. 2). Therefore, such a diagram based on potential-pH measurements can predict, if (a) metal will remain in immune region, or (b) in passive region, where the tendency to corrode decreases significantly owing to passive film formation, or (c) in corrosion region, where no more oxide film is stable. The dashed lines (a) and (b) in Fig. 2 represent water stability regions at different E-pH values. Below line (a), hydrogen will be stable and above line, (b) oxygen will be stable. However in between (a) and (b) water will be stable. Among some of the limitations of Pourbaix diagrams, one is its inability to measure the rate of reaction. There might be a thermodynamic tendency to corrode; however, corrosion rate might be very low, so in actual corrosion might not be a big issue.

2.28.2.1

Corrosion Measurement Techniques

In the laboratory, corrosion rates can be measured by accelerated corrosion tests, which can be destructive and nondestructive. Polarization methods are common among nondestructive corrosion testing. Corrosion rates can be measured by applying a current to the metal under investigation in order to generate a polarization diagram. This polarization curve will give the change in potential as a function of applied current. There are two kinds of polarization, i.e., anodic polarization is developed when the metal surface is polarized in the positive direction by the application of current. On the other hand, if the metal surface is polarized

918

Anti-Corrosive Materials

in the negative direction, the metal/electrode surface is said to cathodically polarize. The concentration of metal ions in solution, dissolved oxygen, film formation and other related factors can affect the degree of polarization. If the reaction rate is disturbed by environmental factors such as concentration of the species, the polarization is called concentration polarization. On the other hand, if the process is affected by adsorption/film formation, then it is called activation polarization.

2.28.2.1.1

Tafel extrapolation method

In Tafel analysis, data obtained from the polarization experiment (anodic or cathodic) are used to measure the corrosion rate of the material. In many cases, it is preferred to use cathodic polarization data due to its ease of measurement during experiments. In Fig. 3, dashed lines, which represent total anodic and cathodic reactions corresponding to metal dissolution and hydrogen evolution respectively, are superimposed. At high applied current densities, the current density applied and the corresponding hydrogen evolution current density is almost similar. In order to calculate the corrosion rate from such a polarization curve, Tafel (liner region) is extrapolated to the Ecorr (corrosion potential) as shown in Fig. 3. Corrosion potential is the point where both the hydrogen evolution and metal dissolution rates are equal and it represents the corrosion rate of the system in terms of corrosion current density. The Tafel constants (V/decade), which can be calculated by anodic and cathodic portions of the curve can be affected by either concentration or activation polarizations. These polarizations will affect the Tafel curves/region; however, if any of the two curves, either anodic or cathodic, has the linear portion, then that can be sued to extrapolated to corrosion potential to calculate the corrosion rate. Any of the two polarization curves can be used to calculate corrosion rate, as shown in Fig. 4. Tafel analysis can be effectively used to measure even very small corrosion rates with high accuracy. It can give direct corrosion current values, which can be used to measure corrosion rate. It is a robust corrosion measurement technique, so it is a very useful method to analyze the corrosion behavior of alloys and to evaluate the performance of corrosion inhibitors.

2.28.2.1.2

Linear polarization resistance

Polarization resistance of a material is defined as the DE/Di (slope of a potential–current density curve at the free corrosion potential, yielding the polarization resistance Rp. Rp is related to the corrosion current (icorr) with the help of the Stern–Geary approximation (under activation controlled polarization). dE ba  bc ¼ 2:3ðba þ bc ÞIcorr dIe In this equation, anodic and cathodic Tafel constants are represented by ba and bc, respectively. Polarization response of about 710 mV around corrosion potential can be sued to measure the slope DE/DI and corrosion current can be calculated using the above equation provided Tafel constants are known. This method can also be used to measure the corrosion rate in metals −0.40 Noble

−0.42 Experimental curves −0.44 −0.46

Anodic curve calculated from cathodic data

Potential, E (V) vs. SCE

−0.48 Anodic −0.50 −0.52 −0.54 −0.56

Cathodic

−0.58

Active

−0.60 −0.62 −0.64 −0.66 −0.68 102

103 104 Applied current density (μA/cm2)

105

Fig. 3 Cathodic polarization data showing Tafel behavior in room temperature deaerated (A) 1 N H2S04 for carbon steel. Reproduced from Jones DA. Principles and prevention of corrosion. New York: Macmillan; 1992.

Anti-Corrosive Materials

E

919

E

Ecorr

Ecorr

Icorr Icorr = Ilim log I

(A)

log I

(B)

E

E

Ecorr Ecorr Icorr = Ip Icorr

log I

(C)

(D)

log I

Fig. 4 Corrosion current density calculation by Tafel extrapolations: (A) Under activation controlled anodic and cathodic reactions, (B) anodic under activation control and diffusion-controlled cathodic reaction, (C) cathodic reaction under activation control and an irregular anodic response, and (D) a passive response by metal and corrosion current is equal to passive current.

showing passive behavior (ba-1) and under the diffusion controlled cathodic process. dE bc ¼ ðPassive stateÞ 2:3Icorr dIe dE ba ðdiffusion‐controlled cathode reactionÞ ¼ 2:3Icorr dIe In order to measure the LPR, one has to

• • • • •

Draw a polarization curve Measure Icorr within 10–20mV, i.e., in the linear region around corrosion potential, in either noble or active direction Plot a graph between overvoltage (Eapp Ecorr) versus (i) Rp will be the slope of the potential–current density curve near Ecorr Rp ¼ ∆E/∆I as ∆E-0

Among the benefits of Linear polarization resistance (LPR), one is to be able to measure the corrosion rate without much disturbance or polarization under natural conditions. LPR can give instantaneous corrosion rates of the materials in just few minutes of time without disturbing the system much (small polarization). Like Tafel analysis, this technique can also be used to measure very low corrosion rates.

2.28.2.1.3

Electrochemical noise measurement

The phenomenon of metastable pitting means that such pits initiate and grow for some time before being passivated again. Usually such pits appear far below the stable pitting potential (Epit), which is one of the very important parameter to compare the

920

Anti-Corrosive Materials

0.3 Potential (V)

0.2 0.1 0 −0.1

0

5

200

400

600

800 Time (min)

1000

1200

1400

200

400

600

800 Time (min)

1000

1200

1400

× 10−6

Current (A)

4 3 2 1 0

0

Fig. 5 Potential and current noise transients for 304 stainless steel in 0.5 M NaCl þ borate buffer with a bias potential of 200 mV.

pitting tendency of different stainless steel alloys. Traditionally, these metastable pits are of um size and their lifetime is on the order of seconds or less. The metastable pitting events can be characterized by potential transients in the active direction at open circuit or by current transient under an applied anodic potential (Fig. 5). If the pits are stable, they will survive and continue to grow; otherwise metastable pits will stop growing and will be passivated [35]. During an electrochemical process, there are fluctuations in current or potential and these fluctuations are termed as “electrochemical noise”. This technique has been in use for last many years for successful interpretation of corrosion phenomenon in corrosion monitoring industry, due to its obvious advantages over other monitoring techniques such as equipment simplicity and easy data interpretation. This technique can be applied to real structures and no external perturbation of the corroding system is required. Localized corrosion problems such pitting, which are difficult to detect and interpret with other monitoring techniques can easily be detected and interpreted with this technique [36]. This technique has been used to study various types of corrosion systems, such as uniform corrosion [37], pitting corrosion [38,39], and crevice corrosion [40,41] and to evaluate corrosion inhibitors and organic coatings [42]. A correlation has been found between the detected EN (potential noise and current noise) and the nature of localized corrosion processes [43]. Every time domain function has a counterpart in the frequency domain. In the case of an autocorrelation function (ACF), which is a function of the shift variable, the counterpart is called the power spectral density (PSD). A plot of PSD against frequency is called the Power spectrum. It indicates how the sequence’s power or energy is distributed in the frequency domain, and is a widely used measure of random sequences. Formally, the ACF and power spectrum of a digital sequence are related as a Fourier transform pair. When the power contained in the frequency interval between f and f þ δf is calculated, one does not distinguish between positive and negative f, but rather regards f as varying from 0 to þ 1. In such cases, we define the one-sided PSD as jHðf Þj2 þ jHð f Þj2 for 0rf r1 PSD is a real function of frequency, with no information about the relative phases of the various frequency components present. Hence PSD relates to power, rather than amplitude or phase. Two major problems associated with spectral estimation techniques are aliasing and leakage. Aliasing is an error introduced due to the sampling rate being too slow. It results in the representation of a high frequency component by a lower frequency component. The rule governing proper sampling, referred to as the Nyquist sampling theorem, states that the sampling rate must be at least twice the frequency of the highest frequency component in the waveform being sampled. In other words, there must be at least two samples per cycle for any frequency component we wish to define. If there are fewer – if the sampling rate is less than twice the highest frequency component – then aliasing occurs. For the one sample per second sampling rate the Nyquist frequency is 0.5 Hz. Leakage results from the basic assumption of the Fourier transform that a finite time record is assumed to be periodic. If, for example, a sinusoidal process is sampled and a whole number of cycles are recorded then the Fourier transform will infer correctly that, the process is an infinitely long sine wave and will resolve its characteristic frequency in the frequency domain. However, if a whole number of sine waves is not recorded, then the Fourier transform assumes a waveform distorted from the original. The above methods of spectral estimation do not always give acceptable spectral resolution. Various more modern approaches have been developed, in which the random sequence is modeled as the output of a processor driven by noise. By making

Anti-Corrosive Materials

921

reasonable assumptions about the structure and order of the system it is often claimed that it is possible to extrapolate the ACF outside the observation interval. Such model-based methods include the so-called moving average, auto regressive moving average and maximum entropy techniques. Spectral density estimation is a second order statistical manipulation and the available techniques could be broadly classified into two categories: linear or nonadaptive methods and nonlinear or adaptive methods. Techniques using Fourier based transformation of the ACF are classified as linear methods as they only involve the use of linear operations on the available time series, as illustrated in Fig. 6. The maximum entropy method (MEM) and a similar technique called the maximum likelihood method are nonlinear or adaptive technique as their design is data independent. The MEM was originally developed by Burg and has since been the subject of further work by Burg and others [43]. Whereas the Fourier transform computes the coefficients of a series of sine waves that sum to the observed time record, the MEM effectively computes the coefficients of a series of sine waves that sum to the observed time record, the MEM effectively computes the coefficients of a particular class of digital filter that would give the observed time record when applied to a white noise input signal. Briefly, the MEM mathematically ensures that the fewest possible assumptions are made about unmeasured data by choosing the spectrum that is the most random or has the maximum entropy for the process under investigation, and is consistent with all known data. Fig. 7 shows the comparison between fast-Fourier transformation (FFT) and MEM. Compared with FFT, the proponents of the MEM claimed that it had some advantages.

PSD

in

Low

PSD (”10”A”/Hz)

Frequency 1000

100

Cut-off frequency

Roll-off slope

MEM FFT

10

0.01 Frequency (Hz)

1E−3

0.1

Fig. 6 Suggested parameters from the power spectral density (PSD) plot to estimate the pitting corrosion tendency of an alloy.

10−8 10−9

PSD (A2 Hz−1)

10−10

MEM FFT

10−11

10−10 10−11

MEM FFT

10−12 10−13 10−2

10−1

100

101

102 10−1 Frequency (Hz)

100

101

102

Fig. 7 The comparison of the power spectral density (PSD) data by fast-Fourier transformation (FFT) and maximum entropy method (MEM).

922 2.28.2.2

Anti-Corrosive Materials Electrochemical Impedance Spectroscopy

Impedance techniques are nowadays widely applied for investigating the mechanism and rate of corroding systems. Their use is generally restricted to homogeneous corrosion situations in which, at least statistically, the whole sample surface undergoes the same processes. Some trends to extend impedance techniques to localized corrosion are reported in the recent literature. They deal merely with pitting, abrasion and other types of corrosion. Essential electrochemical impedance techniques provide more or less time resolved and surface averaged information on the interface phenomenon. It must be a priori a difficult task to deduce from this kind of experiment an accurate interpretation of localized corrosion. Various approaches have been proposed for improving the applicability of a.c. techniques to this problem. They can be classified as (1) experiments specially designed in order to localize as far as possible the electrochemical measurement in the immediate vicinity of the corrosion location, and (2) experiments tentatively interpreted in terms of an electrical network which must represent at the same time the frequency dependence of the faradaic processes and their spatial distribution. Both approaches will be discussed on the basis of a number of recent studies reported in the literature or performed in our laboratories. The electrochemical impedance spectroscopy (EIS) is an important tool for examination of passive layers and pitting corrosion. Mansfeld et al. [44–46] have demonstrated that pitting corrosion of aluminum alloys can be detected by characteristic changes of impedance spectra. The phase angle occurred to be a very sensitive indicator of the initiation of pitting corrosion. An estimation of pit growth rate has been made by Mansfeld et al. [47,48]. The obtained relationships were similar to those of Hounkeler and Boehni [49] being pits growth rates as a function of exposure time. The paper of Scully et al. [50] indicates that impedance spectroscopy can be used to investigate the passivity of aluminum as well as its pitting behavior. EIS method has been employed to detect and monitor pitting corrosion of other materials than aluminum and its alloys [51]. Oltra and Keddam [52,53] have compared impedance results with the theoretical model of pitting corrosion. Wenger et al. [54] have shown that the characteristic features of the impedance spectra of stainless steel do not only depend on electrochemical state of the pits and of the remaining part of electrode surface, but also on the geometrical changes related to the development of pits. The change of roughness effect induced by the development of pits on steel was very seldom considered [55]. According to the opinion of Wenger et al. [54] the application of constant phase element instead of capacity should not be explained a priori by the influence of diffusion and possible porosity effect. However, Park and Pyun [56] elucidate the occurrence of constant phase element by roughening of the electrode surface resulting from the formation and growth of pits. Recently, Darowicki et al. [57,58] derived a new impedance method, which allows obtaining the changes of impedance spectra as a function of time. This method was successfully applied to investigation of the stress corrosion cracking process [59], measurement of electrode surface changes [60] and impedance investigation of kinetics and mechanism of electrode reaction during linear changes of potential [61]. The application of this new method to investigation of pits initiation process on stainless steel is very interesting due to existence of some controversial opinions about the mechanism of this stage of pitting corrosion. The penetration mechanism has been discussed by Hoar et al. [62], Marcus and Herbelin, [63] and Yang et al. [64]. The defect point model was described by Lin et al. [65], Chao et al. [66] is developed from of the mentioned above. The adsorption model has been described by Kolotyrkin [67] and Strehblow [68]. However, the film-breaking model has been suggested by Vetter and Strehblow [69] and Sato et al. [70].

2.28.2.3

Corrosion Behavior of Stainless Steels and Alloying Elements

Stainless steel alloy 304 is traditionally used as a material for hot water storage tank applications and many other different industrial applications because of its good formability, corrosion resistance and mass production; however in the past some incidents were reported in which hot water storage tanks made by SS304 fractured by stress corrosion cracking (SCC) even after few months in operation. Also in those industrial applications where bit higher chloride content is present, SS304 is not recommended due to possible problems of localized corrosion. Steel making companies are developing new materials which can fulfill industrial demands considering mechanical and corrosion properties. In spite of the high thermodynamic reactivity of metals, what makes our metals-based civilization possible is the phenomenon of passivity [71]. Passivity infers kinetic stability of reactive metals in contact with oxidizing aqueous environments, when thermodynamics indicates a large driving force (negative change in the Gibbs free energy) for the reaction of the metal with oxygen (either from O2 or H2O). The observed kinetic stability is due to the formation of a “passive” reaction product film on the surface that effectively isolates the reactive metal from the corrosive environment, as originally postulated by Faraday (the brief history of passivity is given by Uhlig [72]). It is generally agreed that the quality of the “passive” reaction product film (i.e., “passive film”) in terms of the ion- and electron-transport properties, or in terms of its structure and chemistry is responsible for the corrosion resistance of metal and alloy [73,74]. The co-relation between passivity and corrosion resistance can be understood by the fact that almost 30% cases of corrosion failure are caused by localized corrosion (pitting, crevice, stress corrosion cracking, etc.). Alloying elements in stainless steels are added to achieve certain changes either to the corrosion resistance or to the microstructure (which in turn influences the mechanical and fabrication properties). A common evaluation of corrosion resistance of stainless steel grades is the pitting resistance equivalent (PRE), which is usually evaluated as: PRE ¼ %Cr þ 3.3  %Mo þ 16  %N. This is a neat equation, but unfortunately only a guide to rank different stainless steel grades, and is not a predictor of resistance to any particular corrosive environment. What is apparent is that pitting corrosion resistance can be increased by molybdenum, but also by chromium or by nitrogen additions, which are much cheaper than molybdenum. Si is usually used as a

Anti-Corrosive Materials

923

deoxidizer in stainless steels in order to improve their oxidation resistance and reported to increase hardness and reduces ductility [75–77]. According to M. A. Streicher [75], alloying steels with small concentrations of silicon (up to 2.5 wt%) increases the resistance of the alloy to pitting corrosion, due to the changes produced at the grain boundaries while Rhodin [76] said that the positive effect produced by silicon is due to the increased stability of the passive state, resulting from the increase in silicon content of the protective film. Thomasov et al. [77] studied the effect of various alloying elements on the variations in the intensity and nature of pitting corrosion as well as the mechanisms involved in rendering stainless steels less susceptible to pitting corrosion by the specific effect of the alloying elements. Thomasov et al. [77] also measured the effect of silicon on the susceptibility to pitting corrosion by determining the breakdown potentials in 0.1 N NaCl solution and found that when (18Cr–14Ni) steel was alloyed with 5 wt% Si or vanadium, breakdown potential was increased up to 1.5V (relative to the normal hydrogen electrode). Their work concluded that steel containing about 2.5 and 5 wt% silicon are most resistant to pitting corrosion. When analyzing the pitting initiation sites, they found that alloying of chromium-nickel steel with Mo, V or Si simultaneously increases its resistance to pitting corrosion and shifts the zone of probable pit initiation from the peripheral areas of the grain to the grain boundary material. This may be regarded as an increase in stability of the passive state of the austenite grain proper. When alloyed with Mo, V or Si, only a narrow grain boundary zone remains susceptible to pitting corrosion under very aggressive test conditions. The enhanced resistance to pitting corrosion when steel was alloyed with Si, Mo or V, can be explained by the increased stability of the passive film due to the presence of higher concentration of these elements (more than in the alloy) in the protective film [78].

2.28.3

Results and Discussion

In the present section, the electrochemical response of some of metallic and nonmetallic materials are introduced. These materials include laser textured alumina surfaces, aluminum silica alloy, titanium alloy, and laser treated steel. The findings are presented in line with the previous studies [79–84].

2.28.3.1

Corrosion Response of Laser Treated Alumina

To determine the corrosion resistance of the laser treated surface, electrochemical tests are carried out in line with the previous study [79]. In this case, the laser treated (one side untreated and the other side laser treated) and as-received samples of Al2O3 is used in the corrosion experiments. The specimens with dimension of 30 mm in nominal diameter and 5 mm thickness were ultrasonically degreased in distilled water. An aqueous solution containing 5% NaCl was prepared using deionized water. The specimens were, then, immersed in the sodium chloride solution and continuously stirred using a magnetic stirrer for seven days at room temperature. The findings of the corrosion tests are given below in line with the previous study [79]. Fig. 8 shows the SEM micrographs of laser treated surface. It is evident that regular patterns are formed at the surface, which is due to overlapping of the laser irradiated spot during the laser scanning. Moreover, the surface is free from the cracks at low laser output power and some locally scattered micro-cracks are observed at the surface for high laser output power. The cracks are mainly situated around the overlapping patterns. This is because of the high thermal stress levels developed in these regions. It should be noted that AlN has low density than alumina; therefore, volume shrinkage in the AlN region results in stress levels [80]. This, in turn, contributes to the crack formation in the surface region. Laser melting modifies the microstructure in the surface region, which is evident when comparing the micrographs of laser treated and untreated surfaces, i.e., microsize alumina powders appear at the untreated surface while the continuous condensed structure is observed at the laser treated surface. The laser treated region extends 40 mm below the surface. There is no micro-crack observed across the laser treated sample cross-section. However, some small microcracks are seen in the surface region for high laser power output, provided that the microcracks do not extend into the sample cross-section. Fig. 9 shows the SEM micrographs of laser treated and untreated Al2O3 surfaces after the corrosion tests. SEM micrographs reveal that, in general, the local pitting of laser treated surface occurs at the crack tips for high power laser output. This indicates that pitting initiates at local stress region. Moreover, as the pitting size increases, it grows along the crack length through dissolution of the treated surface. This may occur because of the non-homogenized structures developed in this region. Therefore, formation of AlN at the surface after the laser treatment process could results in locally non-homogeneous distribution of nitrogen concentration at the surface. In this case, potential difference in nitride species and alumina results in active sites at the surface. It should be noted that AlN and a–Al2O3 peaks are observed from the XRD diffractograms, which demonstrates the co-existence of Al2O3 and AlN compounds in the surface region. The volume shrinkage along the AlN region causes micro-level stress formation in these regions. Moreover, partially dissolved neighboring materials at the surface results in delamination and the formation of the pit sites. However, this situation occurs randomly at the surface and no regular pits pattern is observed. The pits are shallow, which, in turn, indicates that no secondary pit sites are formed. In this case, passivation in the pit sites prevents the secondary pitting by the corrosion product. In the case of alumina as-received surfaces, the scattered few pit sites are observed. The small voids between the alumina powders are responsible for the formation of the pit sites; in which case, the voids act as the active crevice corrosion sites. Fig. 10 shows surface roughness of the laser treated and untreated workpieces. It is evident that the roughness of the laser treated surface is slightly higher than the untreated surface. This is because of the laser induced melting tracks, which are developed at the surface after the laser treatment process (Fig. 8). Nevertheless, the surface roughness of the laser treated surface is in the order of 7 mm.

924

Anti-Corrosive Materials

Micro-cracks

Laser power = 120 W Scanning speed = 5 cm/s

Laser power = 180 W Scanning speed = 5 cm/s

Microvoids

Laser tracks

Laser power = 180 W Scanning speed = 5 cm/s

Untreated surface

Fig. 8 SEM micrographs of top surface of laser treated and untreated workpieces. Reproduced from Yilbas B, Khaled M, Karatas C. Laser remelting of alumina tile surfaces: corrosion testing in aqueous solution. Corrosion Eng Sci Technol 2011;46(4):477–80.

Consequently, the textural irregularities generated at the surface increases the surface area subjected to the electrolytic solution. However, this does not cause significant increase in the pitting sites at the surface provided that only few scattered pit sites are observed.

2.28.3.2

Corrosion Response of Laser Treated Ti-6Al-4Al Alloy

To assess the electrochemical response of the laser treated alloy surface, the corrosion tests (Potentiodynamic polarization and Tafel analysis) are carried out in line with the previous study [81] while incorporating a three electrodes cell, which composed of a specimen as a working electrode, a Pt wire as a counter electrode, and a saturated calomel reference electrode (SCE). The laser treated specimens are degreased in benzene, cleaned ultrasonically, and subsequently washed with distilled water prior to electrochemical tests. The investigations were carried out with an exposed working electrode area of 0.3 cm2 in 0.5 M NaCl solution at room temperature. The experiments were performed using PCI4/750 Gamry potentiostat and repeated several times to ensure the reproducibility of the data. DC105 corrosion software was used to analyze the Tafel region, while potentiodynamic polarization experiments were performed at a scan rate of 0.5 mV/s. Fig. 11 shows the SEM micrographs of laser treated surface. The laser scanning tracks are visible and close examination of the surface reveals that laser treated surface is free from large size defects. Since laser beam intensity distribution at the workpiece surface is Gaussian, the peak intensity occurs at the irradiated spot center. This, in turn, causes partial evaporation of the treated surface, due to the attainment of high temperature at the irradiated spot center, and forms shallow cavities in the region of the irradiated spot center. Moreover, workpiece surface is scanned with high frequency laser repetitive pulses (1500 Hz) and initially formed small cavity around the irradiated spot center is filled with the molten flow from the second spot located adjacent to the initially evaporated region. This in turn results in continuous laser scanning tracks (Fig. 11(A)). The depth of the cavity formed during surface evaporation becomes shallow because of the molten flow from its neighborhood irradiated spot (Fig. 11(B)). The presence of continuous laser scanning tracks gives rise to a regular surface texture. This, in turn, results in low surface roughness, which is in the order of 1.37 mm. Although laser melting and solidification takes place at high cooling rates, no thermally induced cracks are observed at the surface of the laser treated layer. This is associated with the self-annealing effect of the finally formed laser scanning tracks on the initially formed laser tracks. In this case, conduction heat transfer from the finally formed laser tracks toward the regions, where the laser tracks are formed initially, is responsible for creating the self-annealing effect in the treated layer. Moreover, the close examination of the surface reveals the presence of partially imbedded B4C particles at the surface (Fig. 11(C)). This is

Anti-Corrosive Materials

925

Pit sites

Laser output power = 180 W

Laser output power = 120 W

Pit site

Fig. 9 SEM micrographs of workpiece surfaces after the corrosion tests. Reproduced from Yilbas B, Khaled M, Karatas C. Laser remelting of alumina tile surfaces: corrosion testing in aqueous solution. Corrosion Eng Sci Technol 2011;46(4):477–80.

because of the higher melting temperature of B4C particles as compared to that corresponding to the base material. The mismatch between the thermal expansion coefficient of B4C and the base material results in micro-stresses in the near region of the B4C particles, which contributes to the stress levels in the treated layer. However, the self-annealing effect reduces the stress levels in the surface region. High cooling rates causes the formation of fine grains in the surface region (Fig. 11(D)). Since the self-annealing effect reduces the stress levels, no micro-crack or crack-network is observed in the region of the fine grains. Fig. 12 shows the results obtained using Potentiodynamic polarization tests, which are conducted in 0.5 M NaCl solution at room temperature. The findings revealed that the presence of nitride compounds increase the corrosion resistance of the alloy surface when the properties were measured in terms of pitting potential (Epit), corrosion potential (Ecorr) and passive current density (ip) values, as compared to that of untreated surface. The laser treated and untreated surfaces exhibit active behavior and a stable passive range is observed; however, as soon as the pitting potential is reached, anodic current increased abruptly. The pitting potential, which is an important parameter to investigate the localized corrosion resistance of passive materials, is found to be higher for nitrided sample (0.44 VSCE) as compared to that of untreated alloy, which is found to be 0.3 VSCE. Consequently, the effect of laser treatment is more significant in improving the localized corrosion resistance of the alloy surface. Corrosion potential

926

Anti-Corrosive Materials

5 μm

Fig. 10 Surface roughness of laser treated surface. Reproduced from Yilbas B, Khaled M, Karatas C. Laser remelting of alumina tile surfaces: corrosion testing in aqueous solution. Corrosion Eng Sci Technol 2011;46(4):477–80.

Laser scanning tracks

Laser treated surface

Laser scanning tracks

(A)

(B)

Fine grains

B4C particles

(C)

(D)

Fig. 11 SEM micrographs of the top surface of laser treated titanium alloy surface with presence of B4C particles: (A) Regular laser scanning tracks, (B) laser scanning tracks with fine and shallow cavity, (C) partially imbedded B4C particles, and (D) fine sized grains at the dense layer surface.

(Ecorr) of laser treated surface is slightly noble and found to be 0.3 VSCE, as compared to 0.43 VSCE, for untreated sample surface in 0.5 M NaCl solution at room temperature. After the corrosion potential measurements, a stable passive range is observed in both surfaces, which is particularly true for the laser treated surface. This behavior is attributed to the formation of nitride and other carbide compounds at the surface, which improve the corrosion resistance. In the case of untreated surface, passive range is not very consistent and anodic current increases until it reaches the pitting potential. A comparison of the passive current density (ip) of the surfaces shows that the laser treated surface has less value as compared to untreated surface, again suggesting a stable and protective passive film formation by laser treatment of the alloy surface. Passive current density of the laser treated surface is of the order of 1.11  10 6 A/cm2 as compared to 3.9  10 6 A/cm2 for the untreated surface, which is consistent with the early work [82]; in this case, it was reported that the presence of nitride compounds improves the pitting corrosion resistance of austenitic stainless steel and

Anti-Corrosive Materials

927

0.8

Potential (VSCE)

0.4

0.0

Untreated

−0.4

Laser treated

−0.8

−1.2

1E−10 1E−9

1E−8

1E−7

1E−6

1E−5

1E−4

1E−3

0.01

0.1

1

Current density (A/cm2) Fig. 12 Potentiodynamic polarization response of the laser treated and untreated alloy in 0.5 M NaCl solution at room temperature. Reproduced from Yilbas BS, Toor I-u-H, Malik J. Corrosion resistance of laser treated titanium alloy with B4C particles at the surface. Int J Mater Res 2014;105 (10):975–82.

their weldments [83]. In addition, in titanium alloys, nitrogen and carbon stabilize the alpha phase [82]. Number of possible mechanisms are suggested to explore the improvement in localized corrosion resistance due to presence of nitride compounds. The two most comprehensive mechanisms state that, i.e., (1) Nitrogen enriches at the metal/film interface by anodic segregation and so passivity is improved by subsequent retardation of base metal dissolution and [84] and (2) by forming a dense oxynitride passive layer on the surface of these materials [85]. However, carburizing environments and phase equilibrium of Ti–C differs from those for Ti–N, as there is a very low solid solubility for carbon in alpha titanium [86]. Therefore, the effect of TiN and TiC compounds on the corrosion resistance of the laser treated surface are different; in which case, nitride compounds have considerable effect on the corrosion resistance of the laser treated surface. On the other hand, the EIS gives information on the polarization resistance (Rp), which is used in corrosion rate calculations and is inversely proportional to corrosion rate of that particular material. It is obtained by measuring the impedance of the electrode/electrolyte system over a wide range of frequencies. In the present study, EIS was carried out at open circuit potential (OCP), by applying a sinusoidal potential perturbation of 10 mV with frequency sweep from 100 kHz to 0.01 Hz. Polarization resistance value (Rp) of laser treated alloy is found to be much higher than that of untreated alloy surface, confirming the results discussed for the polarization curves. Polarization resistance value (Rp) value for the laser treated surface is 2.8  103 O cm2 as compared to 7.2  106 O cm2 for that of untreated surface. These results show the positive effect of the laser treatment process on the corrosion resistance properties of the surface. Polarization resistance (Rp) is widely used to characterize the material’s resistance to polarization and, thus, can determine the protectiveness of film on the material surface [87]. The higher value of Rp (7.2  106 O cm2 ) for laser treated surface means that the alloy surface becomes less active owing to more homogeneous and uniform surface structure after the laser treatment. These results are in good agreement with those discussed in the previous section of potentiodynamic polarization, where laser the treated sample shows higher pitting potential values. It is also evident from Fig. 13 that the pit sizes are small and shallow on the laser treated surface indicating the presence of locally scattered corrosion sites (Fig. 13(A)). The corrosion products, such as oxides, are formed at the pit sites, they are not notable in the surface ergion. This is associated with presence of few small areas of pit sites rather than a large area oxidation at the laser treated surface. The presence of B4C particles at the laser treated surface does influence the pitting size and formation at the surface (Fig. 13(B)). It acts as a passive region at the surface suppressing the extension of pit sites while contributing to the improvement of the corrosion resistance of the laser treated alloy surface. Consequently, B4C particles do not behave like active sites at the surface during the electrochemical tests, despite their contribution to the micro-stress stress levels at the surface. Nevertheless, the total coverage area of the surface by B4C particles is small because of their concentration in the laser treated layer, which is 5%. Therefore, their influence on the micro-stress levels and the pit sites is not considerable at the surface. Microsize holes are observed at the surface of the laser treated surface (Fig. 13(C)); however, they are few in numbers and locally scattered.

Anti-Corrosive Materials

928

Pit site Fine grains

B4C particles (A)

(B)

Sub-micro holes

(C) Fig. 13 SEM micrographs of pit sites of laser treated surface: (A) electrochemically tested surface and presence of fine grains at the surface, (B) pit site around B4C particle, (C) close view of electrochemically tested surface and two sub-micro holes. Reproduced from Yilbas BS, Toor I-u-H, Malik J. Corrosion resistance of laser treated titanium alloy with B4C particles at the surface. Int J Mater Res 2014;105(10):975–82.

2.28.3.3

Corrosion Response of Laser Treated Aluminum–Silicon Alloy

In line with the previous study, the corrosion tests (potentiodynamic polarization) were carried out in a three electrodes cell, which composed of a specimen as a working electrode, a Pt wire as a counter electrode, and a SCE. The workpieces were degreased in benzene, cleaned ultrasonically, and subsequently washed with distilled water prior to potentiodynamic tests. The tests were carried out according to ASTM G5 and G61, which were widely used for conducing different potentidoynamic polarization studies. 3M electrochemical tape was used to select a surface area of 0.07 cm2 for subsequent electrochemical tests. An alligator clip with a copper wire was used to immerse the required area under investigation in to the electrolyte and for working electrode connections. The investigations were carried out in deaerated 0.2M NaCl solution at room temperature using PCI4/750 Gamry potentiostat and repeated several times to ensure the reproducibility of the data. The workpieces were cathodically cleaned for 3 min at 1.0 VSCE and allowed to stabilize for 30 min under the open circuit conditions prior to the potentiodynamic polarization tests were performed. The use of 0.2 M NaCl solution provides slow reaction rate, which could be sufficiently observed with PCI4/750 Gamry potentiostat. However, increasing concentration causes high reaction rate and early passivation of the surface is resulted, which unable to assess the corrosion characteristics of the laser treated surface. DC105 corrosion software was used to analyze the Tafel region, while Potentiodynamic polarization experiments were performed at a scan rate of 0.5mV/s. During the polarization experiments, potential was scanned from 1.2 to 0.4 VSCE. Tafel extrapolation could be carried out in a separate experiment or on the existing potentiodynamic polarization curve and later was used in this work. A region within 7100 mVSCE around Ecorr was selected for Tafel fit and subsequently corrosion rate data were obtained. The findings of the electrochemical tests for the aluminum–silicon alloy are presented in line with the previous findings. Fig. 14 shows the SEM micrographs of laser treated surface. The surface exhibits regular laser scanning tracks, which compose of irradiated laser spots. Since the frequency of the laser repetitive pulses is high (1500 Hz), the overlapping ratio of the laser spots is on the order of 72%. It should be noted that the overlapping ratio is determined from the coverage area of previously formed irradiated spot by the lately formed irradiated spot at the workpiece surface. Therefore, the overlapping ratio corresponds to the coverage area over the irradiated spot area. The high overlapping ratio results in large coverage area of the previously irradiated spot and causes continues melting at the surface during the laser scanning of the surface by consecutive pulses. Since the laser treatment process is

Anti-Corrosive Materials

929

Laser scanning tracks

Overlapping of spots

Fig. 14 SEM micrographs of laser treated surface.

involved with the controlled melting, no over flow of the melted material across the laser scanning tracks is observed. In addition, the treated surface is free from large scale asperities and the cavities. It should be noted that laser output power is controlled to avoid the surface evaporation during the laser processing. However, locally scattered few small cavities are formed at the surface, which are associated with the surface evaporation. This occurs because of the high-temperature oxidation reactions despite the fact that high pressure nitrogen is used as an assisting gas in the treated section. The presence of small amount of oxygen causes exothermic reactions to take place in the irradiated region [88]. Therefore, excessive heating at the surface, due to exothermic reactions, gives rise to the evaporation and thermal erosion in these regions. Moreover, microcrak networks are not observed despite the occurrence of high cooling rates at the surface. It should be noted that melted surface solidifies quickly causing hightemperature gradient formation in the surface region. This, in turn, triggers thermal stress development in this region. In addition, the assisting gas contributes to the cooling rates, since it impinges onto the surface while generating a convective cooling effect at the surface. However, lately formed laser scanning tracks act like heat sources for initially formed laser scanning tracks because of the conduction heating. This, in turn, modifies the cooling rates and generates self-annealing effect on the laser treated surface. Consequently, thermal stress formed at the surface region is not high enough to cause crack and crack network formations at the surface. The close examination of the SEM micrographs reveals that fine grains are formed at the surface due to high cooling rates at the surface. Fig. 15 shows the results of Potentiodynamic polarization response of the laser treated sample surfaces in 0.2M NaCl solution at room temperature. In general, laser treated workpiece surfaces improve the corrosion resistance slightly as compared to that of the bare (non-treated) surface, except for workpiece no. S4, which has lower corrosion resistance than the bare surface. The treated surface undergoes fast solidification at high cooling rates and demonstrates the advantages of refinement and reduced microsegregation; therefore, the treated surface is expected to have the improved corrosion resistance than that of the bare surface. The

930

Anti-Corrosive Materials

0.2

Potential (VSCE)

0.0

AISi-S1 AISi-S2 AISi-S3 Bare surface In 0.2M NaCl solution at room temp.

AISi-S4

−0.2 − −0.4

−0.6

−0.8

−1.0 1E−10

1E−8

1E−4

1E−6

0.01

2

Current density (A/cm ) Fig. 15 Potentiodynamic polarization response of laser treated and bare workpieces in 0.2M NaCl solution. AlSi – S1, S2, S3, and S4 represent the laser treated workpiece number.

Table 1

Electrochemical test results for bare and laser treated workpieces in 0.2 M NaCl aqueous solution Ecorr (mVSCE)

Bare surface Laser treated Laser treated Laser treated Laser treated

surface surface surface surface

#1 #2 #3 #4

0.72 0.70 0.70 0.72 0.70

icorr  10

9

5.80 4.50 3.92 0.0039 28

(A)

ip  10 5.10 4.80 4.00 3.90 5.00

3

(A)

Corrosion rate (mpy) 0.53 0.50 0.45 0.30 1.12

attainment of low corrosion resistance for workpiece no. S4 can be attributed to the surface texture of the treated workpiece, since the treatment is carried out at higher laser output power as compared to the other workpieces, i.e., high surface roughness, due to the local evaporation, causes accelerated corrosion at the surface. In this case, electrolytic solution trapped in between the closely spaced surface roughness peaks triggers the crevice corrosion in this region while contributing to the corrosion current. In addition, microstructural changes at the treated surface contribute to the electrochemical response of the surface [89], i.e., the deficiency of Al-rich (a) phase and Si particles in refined microstructures contributed to the low corrosion resistance at the surface. It should be noted that during the solidification of the laser treated layer, the grain boundaries can be imperfectly conformed due to the deformation in the atomic level, mainly on the Al-rich phase side of the interface. Since silicon grows from the melt in a faceted manner (smooth growth interface) and a-phase solidifies with surfaces that are rough, localized deformation can take place in the microstructure, which influences the corrosion resistance at the treated surface in consistent with the previous studies [89,90]. On the other hand, aluminum oxide and aluminum nitride formation enhances the corrosion resistance at the surface in line with the previous findings [79]. Therefore, the possible explanation for the improvement of corrosion resistance of the laser treated surface, as compared to bare workpiece, is due to the passive oxide and nitride film formation and the microstructural refinement at the surface during the treatment process. In addition, the coarser dendritic structures yield higher corrosion resistance than finer dendritic structures which is associated with the morphology of the interdendritic eutectic mixture [91,92]. Moreover, the corrosion potential (Ecorr) of all specimens is found to be in the range of 0.7 to 0.72 VSCE, while the corrosion current density (icorr) is found to be the least for specimen S3, i.e., it is 6.3  10–11 A. the same trend is observed for the corrosion rate calculated by Tafel analysis and it is found to be the least for specimen S3, which is 0.3 mils per year (mpy). However, the corrosion rate is the highest for specimen S4 as given in Table 1. Fig. 16 shows the SEM micrographs of laser treated and bare workpiece surfaces after the electrochemical tests. The pit sites are visible at the surfaces of all workpieces, which is particularly more pronounced for the sample no. S4. In general, the pits are shallow and small for the laser treated surfaces. This indicates the presence of the passive layer due to oxide and nitride compounds formed at the surface during the laser treatment process, since aluminum oxide is extremely resistance to the corrosion attack [93]. Moreover, the close examination of the pit sites indicates that intergranular attack is governing process for the pitting, particularly for sample number S4. In this case, the initial pit sites are formed through generation of cracks around the grain

Anti-Corrosive Materials

931

Corrosion products

(A)

(B) Corrosion products

Corrosion products

(C)

(D) Crevice like pitting site

(E) Fig. 16 SEM micrographs of pit sites at laser treated and bare surfaces: (A) Laser treated surface, (B) Bare surface, (C) Laser treated at high power intensity (100 W), (D) Initiation of pit site and cracking, (E) Crevice like pitting site, where the pit depth is large.

boundaries because of the volume expansion as shown in Fig. 16(D)). The pit sizes are relatively smaller for laser treated surfaces as compared to that of the bare surface. However, no secondary pitting is observed despite the fact that the corrosion product is present in some pit sites. This indicates the passivation of the pit site once it is generated; therefore, the shallow pits are resulted at the surface. Moreover, few deep pit sites are observed, which are associated with the local surface roughness peaks triggering the crevice corrosion in this region. It should be noted that the pit depth extends toward the laser treated layer when the crevice like corrosion takes place locally. Nevertheless, this effect is only observed locally and the pit site sizes are small, i.e., no large scale elongated pits are observed due to the secondary pitting at the surface.

2.28.3.4

Corrosion Response of Laser Treated High Strength Low Alloy Steel

High-strength low-alloy steel (HSLA) has improved mechanical properties and the low carbon content of the alloy enables to retain formability and weldability, which enable the alloy to be used in energy harvesting devices. Although high-strength lowalloy steel has high resistance to wear and corrosion, surface properties, such as corrosion resistance and microhardness, can be further improved through forming a fine grain dense layer at the surface through a control melting. One of the techniques in line with the surface properties improvement is to use a high intensity laser beam, since laser controlled melting of the alloy surface has many advantages over the conventional melting methods. Some of these advantages include precision of operation, local treatment, fast processing, and low cost. However, laser heating causes high rate of melting and consequence solidification in the treated layer. This, in turn, results in formation of high-temperature gradients in the treated region while limiting the practical application of the treated surface. In addition, high intensity laser beam, more than the evaporation threshold, causes attainment

932

Anti-Corrosive Materials

of high temperature at the surface and results in formation of evaporation induced cavities at the surface. However, proper selection of laser treatment parameters minimizes the evaporation induced cavity formation and other surface asperities during the laser treatment process. Although laser treatment of the alloy surface improves the surface properties, development of high thermal stress field in the treated layer can exceed the yielding limit of the alloy and causes failure of the treated surface through multiple cracking or a crack network formation. This particularly lowers the corrosion resistance of the resulting surface. To examine the corrosion resistance of high strength low alloy steel, the electrochemical tests are carried out. The findings are discussed herein in the light of the previous study [94]. The corrosion tests (potentiodynamic polarization and Tafel behavior) were carried out in a three electrodes cell, which composed of a specimen as a working electrode, a Pt wire as a counter electrode, and a SCE. The specimens were degreased in benzene, cleaned ultrasonically, and subsequently washed with distilled water prior to electrochemical tests. The investigations were carried out with an exposed working electrode area of 0.07 cm2 in 0.2 M NaCl solution at room temperature in PCI4/750 Gamry potentiostat and repeated several times to ensure the reproducibility of the data. DC105 corrosion software was used to analyze the Tafel region, while Potentiodynamic polarization experiments were performed at a scan rate of 0.5mV/s. Fig. 17 shows the optical micro-photograph and SEM micrographs of laser treated surface prior to the corrosion tests. Laser treated surface composes of regular laser scanning tracks, which are formed due to the overlapping of the laser irradiated spots. The laser pulsing frequency is 1500 Hz and it results in overlapping ratio of 72% for two consecutive spots at the surface. This, turn, causes continuous melting of the substrate along the laser scanning tracks. Laser treated surface is free from large scale asperities including voids and cracks. Although few shallow cavities are observed at the surface, they are locally scattered and do not conform a regular pattern. The formation of shallow cavities is associated with the local evaporation at the surface during the laser heating process. In this case, laser intensity distribution is Gaussian and the peak intensity corresponds to the center of the irradiated spot. In addition, the presence of small oxygen in the irradiated region triggers the exothermic oxidation reactions at the surface, which in turn, provides extra energy in the laser interaction zone. Consequently, local evaporation of the surface becomes unavoidable, despite the use of high pressure nitrogen assisting gas in the irradiated region. Nevertheless, the size of evaporated surface is small and its depth is shallow. The surface roughness is in the order of 1.8 mm, which is small for laser treated surface [95]. However, some textures in the surface roughness chart represent the presence of a cavity having shallow depth. Moreover, high heating rates cause thermal expansion of the substrate material in the surface region during the heating cycle and thermal contraction causes the formation of high stress levels in the surface region during the cooling cycle. Although the cooling rate is high at the surface, no thermally induced cracks are observed in the surface region. This is attributed to the self-annealing effect of the lately formed laser scanning tracks, which act like a heat source influencing the cooling rates in the region of the initially formed tracks. Therefore, heat conduction from the recently formed laser tracks toward the initially form tracks modifies the cooling rate in the laser treated layer. The closed examination of the micrograph reveals the presence of fine grains at the surface. This is attributed to the high cooling rates at the surface; in which case, assisting gas contributes to the high cooling rates at the surface through a convection heat transfer. Fig. 18 shows the potentiodynamic polarization response of laser treated and untreated sample surfaces, which are tested in 0.2M NaCl solution at room temperature. It is clear from Fig. 18 that laser treated (sample 1) exhibited better corrosion resistance than untreated sample (sample 2) in terms of pitting potential (Epit) and passive current density (ip). Corrosion potential (Ecorr) is found to be, i.e., 0.63 VSCE and 0.57 VSCE for laser treated sample (sample 1) and untreated sample (sample 2), respectively. Passive current density (iP) as well as corrosion current (icorr) density of laser treated sample is much less than that corresponding to untreated sample. All these results suggest a stable and more protective film is formed on laser treated sample surface, which means that laser treatment has a positive effect on the corrosion properties of the high strength low alloy steel. Therefore, formation of nitride compounds and fine grains at the surface act like as a self-protective layer at the surface. This, in turn, improves the corrosion resistance of the laser treated surface, which is in line with the previous study [96]. Moreover, low roughness of the laser treated surface prevents the initiation of the crevice corrosion at the surface. It should be noted that large surface texture allows electrolytic solution to fill the gabs between the texture peaks at the surface. This, causes, serve corrosion attacked while generating deep and large pit sites at the surface. However, this situation does not take place at the treated surface due to low surface roughness. Fig. 19 shows SEM micrographs of laser treated and untreated surfaces after the potentiodynamic tests. In general, locally scattered pits are formed at the surface after the electrochemical tests. Small and shallow pit sites are visible at the laser treated surface; however, the pit size increases for the untreated surface. The localized pitting results in low pitting potential during the electrochemical tests, which can be seen from Fig. 18. The close examination of the pit sites reveals that no secondary pitting takes place at the surface of the laser treated sample. In this case, formation of nitride species and fine grains at the surface are responsible for preventing the secondary pitting, i.e., they act like a protective layer while lowering the pit size at the treated surface. The self-annealing effect of the lately formed laser scanning tracks results in the reverse transformation of some δ-ferrite into g-austenite in the treated layer, which in turn improve the localized corrosion resistance of the laser treated surface.

2.28.3.5

Corrosion Response of Laser Treated Copper Alloy (Bronze)

In order to investigate the electrochemical response of bronze, electrochemical tests are carried out with the association of the previous study [97]. The findings are discussed herein in line with the previous study [97]. The corrosion tests (potentiodynamic polarization, Tafel behavior and EIS) were carried out in a three electrodes cell, which composed of a specimen as a working

Anti-Corrosive Materials

933

Laser scanning tracks (A) 400 μm

Overlapping of laser spots Laser scanning tracks (B)

(C)

Fine grains (D) Fig. 17 Optical and SEM images of laser treated surface prior to electrochemical tests: (A) Optical image, (B) SEM micrograph showing regular laser scanning tracks, (C) overlapping of laser irradiated spots, and (D) fine size grains formed at the laser treated surface. Reproduced from Yilbas B, Toor I-u-H, Malik J, Patel F. Laser treatment of high strength low alloy steel and electrochemical response of the surface. Ind Lubrication Tribol 2015;67(2):166–171.

electrode, a Pt wire as a counter electrode, and a SCE. The specimens were degreased in benzene, cleaned ultrasonically, and subsequently washed with distilled water prior to electrochemical tests. The investigations were carried out with an exposed working electrode area of 0.07 cm2 in 0.1M NaCl solution at room temperature in PCI4/750 Gamry potentiostat and repeated several times to ensure the reproducibility of the data. DC105 corrosion software was used to analyze the Tafel region, while potentiodynamic polarization experiments were performed at a scan rate of 0.5 mV/s. EIS measurements were carried out at OCP, by applying a sinusoidal potential perturbation of 10 mV with frequency sweep from 100 kHz to 0.01 Hz. The impedance data were analyzed and fitted to appropriate equivalent electrical circuit using the GAMRY framework software. Fig. 20 shows the results of potentiodynamic polarization response of laser treated and untreated surfaces in 0.1M NaCl solution at room temperature. It is clear from Fig. 20 that laser treated sample exhibits better corrosion resistance than the untreated specimen in terms of pitting potential (Epit) and passive current density (ip). Corrosion potential (Ecorr) is found to be

934

Anti-Corrosive Materials

−0.4 Sample 1 Sample 2 In 0.2M NaCl solution at room temp.

Potential (VSCE)

−0.5

−0.6

−0.7

−0.8

1E−8

1E−7

1E−6 1E−5 1E−4 Current density (A/cm2)

1E−3

Fig. 18 Potentiodynamic polarization response of laser treated (Sample 1) and untreated (Sample 2) specimens in 0.2M NaCl solution. Reproduced from Yilbas B, Toor I-u-H, Malik J, Patel F. Laser treatment of high strength low alloy steel and electrochemical response of the surface. Ind Lubrication Tribol 2015;67(2):166–71.

Pit site initiation Pit site

(B)

(A)

Pit site

Corrosion induced cracks

(C)

(D)

Fig. 19 SEM micrographs of laser treated and untreated surfaces after electrochemical tests: (A) pit site initiation at laser treated surface, (B) pit site formed at laser treated surface, (C) pit site formed at untreated surface, and (D) corrosion induced cracking at untreated surface. Reproduced from Yilbas B, Toor I-u-H, Malik J, Patel F. Laser treatment of high strength low alloy steel and electrochemical response of the surface. Ind Lubrication Tribol 2015;67(2):166–171.

Anti-Corrosive Materials

935

1.000 V

Potential (VSCE)

500.0 mV

Laser treated

0.000 V

Untreated −500.0 V

−1.000 V 1.000 nA 10.00 nA 100.0 nA 1.000 μA 10.00 μA 100.0 μA 1.000 mA 10.00 mA Current density (A/cm2) Fig. 20 Potentiodynamic polarization response of laser treated and untreated specimens in 0.1 M NaCl solution. Reproduced from Yilbas B, Malik J, Patel F, Karatas C. Electrochemical testing of laser treated bronze surface. J Alloys Compounds 2013;563:180–5.

Table 2

Results of electrochemical tests for laser treated and untreated alloy in 0.1M NaCl solution [97] Ecorr (mVSCE)

Laser treated Untreated

400 120

icorr

iP

EPit (mVSCE)

10 (nA) 100 (nA)

10 (uA) 300 (uA)

330 70

Source: Yilbas B, Malik J, Patel F, Karatas C. Electrochemical testing of laser treated bronze surface. J Alloys Compounds 2013;563:180–5.

400 mV4 120 mV for laser treated and untreated specimens, respectively. Laser treated specimen shows a pitting potential of 300 mVSCE as compared to 70 mVSCE for untreated specimen. Passive current density (iP) as well as corrosion current (icorr) density of laser treated specimen are much less than that of untreated specimen. All these results suggest that a stable and more protective film is formed on the laser treated specimen surface; therefore, laser treatment has a positive effect on the corrosion properties of bronze surface. Table 2 summarized the results of potentiodynamic polarization. To confirm the potentiodynamic polarization data, the values of corrosion rate for two specimens are calculated using Tafel analysis. It is found that corrosion rate of laser treated specimen (0.00037mpy) is much lower than that of untreated specimen (0.0083mpy), which is in agreement the results shown in Fig. 20. Furthermore, EIS measurements are carried out at OCP, by applying a sinusoidal potential perturbation of 10 mV with frequency sweep from 100 kHz to 0.01 Hz. Moreover, it can be observed that polarization resistance value (Rp) of laser treated specimen is much higher than that of untreated specimen, suggesting that it has higher corrosion resistance. The large semicircle of laser treated specimen corresponds to higher corrosion resistance behavior than untreated specimen. These results show the positive effect of laser treatment on the corrosion properties of bronze. Fig. 21 shows the SEM micrographs of pit sites at the laser treated and as-received workpiece surfaces. It is evident from the SEM micrographs that pit site is smaller and shallow for laser treated surface as compared to corresponding to as-received surface. Consequently, laser treated layer acts as a passive layer lowering the surface pitting during the electrochemical testing. The pits formed at the surface do not have a regular pattern and the secondary pitting in the pit sites is not observed for both laser treated and as-received surfaces.

2.28.3.6

Corrosion Response of Laser Treated Hastelloy Alloy

Hastelloy has good sulfidation resistance and high metallurgical stability, which makes it preferable material for high-temperature applications of thermal energy storing and electricity generating devices such as gas turbines. In addition, it has low cycle fatigue resistance superior to that of most solid solution-strengthened alloys and it has a very good resistance to hot corrosion. The alloy is

936

Anti-Corrosive Materials

Fig. 21 SEM micrographs of laser pit sites after electrochemical tests: (A) Laser treated surface, and (B) As-received surface. Reproduced from Yilbas B, Malik J, Patel F, Karatas C. Electrochemical testing of laser treated bronze surface. J Alloys Compounds 2013;563:180–5.

Table 3

Results of electrochemical tests for laser treated and untreated samples in 0.5M NaCl solution Ecorr (mVSCE)

Laser treated surface Untreated surface

320 338

EPit (mVSCE) 837.3 345

used to manufacture high-temperature gas path components such as turbine combustors, flame holders, liners, pressure vessels of some nuclear reactors, chemical reactors, and pipes/valves in the chemical industry. Several surface treatment techniques are available to improve the surface properties of metallic alloys; however, laser treatment is considered to be one of the preferable techniques with improved performance as compared to conventional methods in terms of chemical cleanliness, thermal penetration, and surface profile. To investigate the corrosion resistance of the laser treated Hastelloy alloy, potentiodynamic tests are carried out in line with the previous study [87]. The test results in terms of findings are presented herein relevant to previous study [97]. Table 3 summarized the results of potentiodynamic polarization. Corrosion potential (Ecorr) was found to be 320 mV for laser treated surface. Pitting potential of the laser treated sample was found to be 837.3 mVSCE and that of untreated was 345 mVSCE. Passive current density (iP) as well as corrosion current density of laser treated surface was less than the untreated surface, which suggests a stable passive film (lower corrosion rate) on the former than the as-received surface. Passive current density on untreated surface showed a continuous increase along with localized breakdown/repassivation phenomenon, which is related to unstable films (metastable pitting). Table 4 summarized the results of Tafel analysis and it is clear that corrosion rate of laser treated surface was decreased significantly as compared to untreated surface. The results presented showed that laser treated specimen has exhibited the greatest corrosion resistance in terms of high Epit, lower icorr, and lower iP as compared with the untreated sample surfaces. These improvements in pitting potential and passive current density, along with more noble corrosion potential can be related to

Anti-Corrosive Materials

Table 4

937

Result from Tafel slopes and polarization resiatnce by LPR method

ba (V/decade) Βc (V/decade) Icorr (nA) Ecorr (V) Corrosion rate (mpy) Polarization resistance (Rp) ohms

Laser treated

Laser treated

Untreated

0.0504 0.0167 0.0624 0.0998 3.904E 4 8.46

0.1024 0.0557 0.512 0.192 3.202E 3 4.7

Untreated

In 0.5M NaCl solution at room temp.

Current density (A/cm2)

1E−7

1E−8 Untreated

1E−9 Laser treated

1E−10 0

500

1000

1500

Time (s) Fig. 22 Current decay curves at an applied potential of 200mVSCE in 0.5M NaCl solution. The constant potential experiments were repeated three times and based on the repeatability of the experimental data, the estimated error was within 6%.

the dissolution of carbides, inclusions, and precipitates during the laser treatment, which were initially present in the untreated material. In addition, fine and homogenous structures formed at the surface after laser treatment process decreased the possible pitting sites substantially which ultimately increased the pitting potential, decreased the passive current density as well as decreased the overall corrosion rate of laser treated surface. The presence of nitrogen also improved the corrosion properties of laser treated surface. It was discussed previously that nitrogen in stainless steels and other alloys had a significant effect on the improvement of the overall corrosion resistance. This improvement was attributed to the modifications in the properties of the passive films caused by nitrogen. It was also reported that nitrogen could improve the resistance to pitting corrosion of stainless steels by preferential enrichment at the film/metal interface through anodic segregation, so that the passivity was improved with subsequent retardation of metal dissolution [98]. The findings are further complemented by experiments such as linear polarization resistance and current decay curves. Linear polarization resistance (Rp) is widely used to characterize the material’s resistance to polarize and thus can determine the protectiveness of film on material surface. The higher the polarization resistance the lower is the corrosion current density, i.e., corrosion rate. This method can be used to check the polarization resistance, i.e., corrosion rate of coating, or any other surface treatment. Polarization resistance of laser treated surface is found to be 8.46 Ω as compared to 4.76 Ω for untreated surface. This trend is similar to the one observed based on corrosion current density calculation by Tafel analysis, which shows that corrosion current of 0.0624 nA for laser treated surface and 0.512 nA for untreated surface. This suggests that laser treated surface has the highest corrosion resistance properties as compared to untreated surface. A higher value of Rp (8.46 Ω) for laser treated surface means that after laser treatment, the alloy surface becomes less active owing to more homogeneous and uniform surface structure. These results are in good agreement with those discussed in the previous section of Potentiodynamic polarization, where laser treated specimen shows higher pitting potential values. In addition, in order to confirm the positive electrochemical effect on the laser treated surface, current decay curves are measured for one hour in 0.5 M NaCl solution at room temperature. A constant potential of 200 mVSCE is applied in these measurement to examine the passive film stability. The potential of 200 mV is selected because at this potential both the samples are in a stable passive region. Results presented in Fig. 22 shows significant decrease in

938

Anti-Corrosive Materials

Untreated surface

Pitted region

Laser treated surface

Laser treated surface

Small shallow pits Pit initiation

Untreated surface

Large pit site

Corrosion product Large pit site Untreated surface

Fig. 23 SEM micrographs of pit sites at laser treated and untreated surfaces.

current density for laser treated surface as compared to untreated surface. This finding, again, complements the results obtained by linear polarization resistance as well as by Potentiodynamic experiments while suggesting that passive film on laser treated surface is more stable as compared to untreated sample. After the corrosion experiments (Potentiodynamic polarization), scanning electron microscope is conducted over the large area at the surface to investigate and compare the surface degradation of treated surfaces and correlate those with electrochemical tests. It is clear from the Fig. 23 that laser treated surface has less pit sites as compared to untreated surface. Since polarization scans are stopped as soon as there is significant increase in current, so no deep pit sites are expected to form at the surface. The initiation of pit sites is evident from SEM micrographs and pits are randomly distributed at the surface. The elongated large pit sites are clearly observed for untreated surfaces. This is in agreement with the Tafel results; in which higher corrosion rate is observed for untreated surface. However, few small pit sites are observed on the laser treated surface. The pits are shallow and no micro-cracks emanate from the pit sites. This is attributed to the formation of nitride species and fine grains formed at the surface region, which act as a self-protective layer at the surface and, thus, improving the corrosion resistance of the laser treated surface. The Surface roughness measurement is carried out to compare the roughness pattern of the laser treated and untreated surfaces and to correlate the surface roughness data with the corrosion behavior of the surfaces. It is well documented that corrosion properties of the materials are seriously affected by their surface roughness profile, because surface roughness affects the ability of

Anti-Corrosive Materials

939

pits or corrosion initiation sites to form on the metal surface. Various researchers [99–101] examined this affect for stainless steel alloys and found that a fine surface had higher corrosion resistance, as the ability of pitting to occur/corrosion initiation sites to form was reduced on the smooth surfaces. In other words, an increase in surface roughness increased the possibility of pitting corrosion. It is clear that surface defects contribute to the formation of corrosion initiation sites. In addition, for rough surfaces, the time of contact (diffusion) of corrosion causing species, such as chloride ions, increases. This, in turn increases the corrosion rate by decreasing the repassivation ability of the surface and so there are more pit sites forming at the surface. In the present study, surface roughness measurement is carried out using Mitutoyo Surftest (SJ-301) equipment and the average roughness (Ra) of the laser treated surface is found to be 0.63 mm, as compared to that of untreated surface (0.755 mm). It is evident from these values that laser treated surface is relatively smoother than the untreated surface, which ultimately reduces the corrosion initiation sites, i.e., lead to improved corrosion resistance of laser treated surface, which is discussed in the former paragraph.

2.28.4

Conclusions

Energy materials, particularly for those expose to harsh environments, suffer from electrochemical attacks because of the presence of ionic solutions in the environmental sites. The care is needed to minimize the effects of the electrochemical attacks on the materials used in energy harvesting devices because of minimization of maintaining cost and sustainable operation of devices with high performance. Although several precautions are introduced for preventive maintenance of such devices, minimization of the defects due to the electrochemical attacks is still one of the current challenging problems. In this section, the conclusions derived from studies associated with the electrochemical tests are presented. In the case of laser treated alumina surfaces, it is found that the laser treated surface is free from pores and voids; however, locally formed microcracks are observed at the surface at high laser output power. A few pit sites are formed at the laser irradiated surface. This is because of: (1) locally scattered non-homogeneous formation of AlN, and (2) changes in the surface texture with a high surface roughness. Nevertheless, the laser treatment provides a reasonably good corrosion resistance at the surface. Surface roughness measurements prior and after the laser treatment process reveals that the surface roughness is in the order of 7 mm after the laser treatment process. Although the surface texture increases the surface area in contact with the electrolytic solution, only few scattered shallow pit sites are observed. Pits are formed mainly around the crack tips at the surface and, in some case; the pit size extends along the crack length. However, the voids between the alumina particles for the untreated workpieces act like the corrosion centers initiating the crevice corrosion. This results in pit formation in these regions. The titanium alloy with presence of B4C particles in the laser treated surface demonstrates that laser treated layer extends almost 45 mm below the depth with a uniform thickness along the treated layer surface. A dense layer consisting of fine grains, nitride and carbide compounds, and B4C particles is formed in the surface region of the laser treated layer. The treated layer is free from voids and micro-cracks despite the fact the differences between the thermal expansion coefficients of B4C and the titanium alloy. This is attributed to the self-annealing effect of the lately formed laser scanning tracks on the initially formed scanning tracks. The feathery like structures are formed in the near region of the surface because of the nitride diffusion through the grain boundaries. The corrosion resistance of the laser treated layer improves significantly as compared to that corresponding to the untreated surface. This behavior is evidenced through potentiodynamic and AC impedance test results. The improvement of the corrosion resistance is associated with the formation of passive layer at the surface because of nitride and carbide compounds, such as TiN and TiC. The formation of TiC is attributed to the presence of carbon film prior to laser treatment process and high pressure nitrogen assisting gas is responsible for TiN formation at the surface. The presence of B4C particles at the surface does not have an adverse effect of the corrosion resistance of the laser treated surface. This is because of the contribution of B4C particles to the passive layer formed at the workpiece surface during the electrochemical tests. Aluminum–silicon alloy after laser surface treatment demonstrates improved corrosion resistance, which is mainly, because of the passive layer formed at the surface. High-temperature oxidation reactions taking place during laser treatment increase the thermal erosion and trigger the evaporation at the surface due to the excessive heating. This, in turn, modifies the surface texture while increasing the surface roughness. However, at high laser power intensities, the electrolytic solution occupying closely spaced surface roughness peaks triggers the crevice corrosion in this region. In addition, microstructural defects contribute to the attainment of low corrosion resistance of the surface, which is observed for samples. The occurrence of low corrosion resistance for laser treated surfaces is associated with the development of passive layer at the surface through forming nitride and oxide compounds. The pits formed at the surface are shallow and small in size and no secondary pitting is observed in the pit sites for the entire laser treated surface. The microstructural defects cause grain boundary cracking at the surface after the corrosion tests. This, in turn, triggers the formation of deep pit sites at the surface. High strength low alloy steel is one of the good candidates for possible usage in the construction of the thermal energy storage devices. The corrosion tests revealed that the pit sites formed at the laser treated surface is also examined in line with the electrochemical response of the surface. It is found that the laser treated surface is free from large scale surface asperities including cracks and voids. However, some shallow locally distributed few cavities are observed at the surface, which are associated with the surface evaporation during the laser scanning of the surface. Laser treated layer consists of the dense layer compose of fine grains and nitride species, then follows feathery like and dendritic structures below the dense layer. The cellular structures are formed toward the heat affected zone. The demarcation line is clearly observed between the laser treated layer and the base material. The roughness of the laser treated surface is in the order of 1.8 mm, which does not result closely spaced tall surface textures.

940

Anti-Corrosive Materials

Consequently, low surface roughness suppresses the initiation of crevice corrosion between the tall textures at the surface. The formation of nitride compounds and fine grains act like as a self-protective layer at the surface; therefore, a stable and protective film on the laser treated workpiece surface lowers the corrosion current. In this case, laser surface treatment has a positive effect on the corrosion properties of the treated workpiece. The pits at the surface do not form a regular pattern and they are shallow. The secondary pitting is prevented by the protective layer formed at the laser treated surface. Copper alloys, particularly bronze, has high thermal and electrical conductivities, which makes it attractive for construction of thermal energy devices. Laser treatment provides the dendritic structures, which are formed below the surface due to relatively slower cooling rates as compared to that at the surface. Large grains are observed in the heat affected zone region. The corrosion current density for the laser treated surface is much less than that of the as-received surface indicating the laser treatment provides protective layer at the surface. This finding is also supported by the EIS results. The pits sites on the laser treated surface appear to be shallower and smaller in size than that corresponding to the as-received surface. The close examination of the pit sites revealed that no secondary pitting takes place at the laser treated and untreated surfaces. The corrosion testing of laser treated Hastelloy alloy demonstrates that laser treated layer extends uniformly below the surface. The fine grains are formed at the surface because of the high cooling rates and nitride compounds. Although high cooling rates generate high thermal stress at the surface region, no crack or crack network is observed. The closed examination of the laser scanning tracks reveals that the overlapping ratio of the laser spots is on the order of 65%, which produces a continuous melting at the surface. The dense structures developed at the surface consist of fine grains and as the depth below surface increases in the laser treated layer, the cellular structure is formed. The groove between the cells takes place due to the segregations at the grain boundaries and the cell size varies because of nonuniform cooling in this region. Electrochemical test results revealed that laser treatment improves the corrosion resistance of the alloy significantly. The pit sites formed at the surface is shallow and randomly scattered. Few small pit sites are observed at the surface of the treated samples; however, elongated large pits are formed at the surface. The formation of fine grains and low surface roughness contributes to the improved corrosion resistance of the laser treated surfaces.

2.28.5

Future Directions

Selection of proper materials for energy harvesting and storage devices is critical to main the efficient operation of the thermal system particularly in harsh environments because of the corrosion attack. In open environments, the presence of ionic compounds becomes unavoidable due to the gas emission from industrial and domestic sites, which in turn triggers the corrosion effect on the energy materials. In addition, the suspended dust particles in open air contributes to the environmentally presence of the ionic compounds. The ionic compounds, such as compounds of alkaline and alkaline earth metals (NaCl , KCl , etc.), dissolves in the water condensate while resulting in chemically active solution on the surfaces. In some cases, acids are also formed during dissolution of ionic compounds in water condensate, which in turn accelerates the corrosion and degradation of the materials. This becomes unavoidable failure of surfaces and scientific research to challenge this problem toward minimizing this effect requires further efforts particularly in the humid environments. Therefore, to meet with the challenges minimizing greenhouse gas emissions becomes essential part of the future research directions. On the other hand, the surfaces used in energy harvesting or storing devices remain critical to avoid or minimize electrochemical reactions taking place due to chemically active ionic solutions in open environment. In this case, one-step method surface processing generating the passive layer against the electrochemical reactions becomes demanding. This is because of avoiding multiple processes requirement for creating the passive layer on the surfaces. There several methods are introduced to improve corrosion resistance of the surface, such as coating, electrochemical deposition of passive layers, ion implantation, etc. Currently, research into coatings through the processes of physical/ chemical deposition, electroplating, high velocity spraying, painting some others gain acceleration toward improving the coating quality and reducing the processing cost. Research into thermal processing of surfaces for generating chemically passive layer also gains inertia in recent years. Laser processing of such surfaces is one of the favorable processes among the others such as induction heating, quenching, plasma treatment and others. Although laser surface processing involves with high temperatures and high cooling rates in the treated region, control of laser parameters minimizes the adverse effects (micro-cracks, cavity formation, etc.) during the processing. However, the process optimization through proper selection of the process parameters is one of the future challenges when the new materials are introduced. Therefore, the future challenge is not only to development of cost effective one step surface processing for generating the passive layer, but resulting surface should meet the requirements of sustainable operation. Hence, the future research directions should fulfill the current challenges while meeting the requirements of both cost effective and green processing.

2.28.6

Closing Remarks

Minimization of corrosion induced degradation is one of the current challenges in energy materials. In some cases, it becomes vital to prevent the corrosion attack to the energy materials, since the maintenance and replacement of parts are extremely costly. In order to minimize corrosion and degradation of energy materials, coating and/or similar preventive processes need to be introduced. Although the preventive processes are expensive, the overall cost of the energy harvesting and storage units can be still

Anti-Corrosive Materials

941

economically viable in the long terms. The research toward development of new energy materials and processes shows significant progress; however, corrosion prevention of the energy materials remains overlooked by the researchers. Consequently, research toward corrosion prevention and minimization of corrosion effects still requires excessive efforts to meet the industrial and domestic demands. Coating technologies incorporating nano-scale processes is one of the promising solutions for the corrosion prevention. In addition, the development of new corrosion resistance materials in the frame of energy applications is also one of the solutions to the corrosion problem in energy sector. Nevertheless, the recent progresses toward the corrosion prevention are promising and could be implemented at large scales when the preventive technology becomes economically viable.

Acknowledgment The authors acknowledge the financial support of King Fahd University of Petroleum and Minerals (KFUPM) through Project# MIT11111-11112 to accomplish this work.

References [1] Harju M, Järn M, Dahlsten P, Rosenholm JB, Mäntylä T. Influence of long-term aqueous exposure on surface properties of plasma sprayed oxides Al2O3, TiO2 and their mixture Al2O3–13TiO2. Appl Surf Sci 2008;254(22):7272–9. [2] Gao Y-l, Wang C-s, Yao M, Liu H-b. The resistance to wear and corrosion of laser-cladding Al2O3 ceramic coating on Mg alloy. Appl Surf Sci 2007;253(12):5306–11. [3] Lawrence J, Li L. Surface treatment of an Al2O3-based refractory with CO2 and high power diode lasers for improved mechanical and chemical resistance characteristics. Surf Coat Technol 2003;162(1):93–100. [4] Schacht M, Boukis N, Dinjus E. Corrosion of alumina ceramics in acidic aqueous solutions at high temperatures and pressures. J Mater Sci 2000;35(24):6251–8. [5] Chong P, Liu Z, Skeldon P, Thompson G. Large area laser surface treatment of aluminium alloys for pitting corrosion protection. Appl Surf Sci 2003;208:399–404. [6] Xu Z, Leong KH, Sanders PG. Laser surface alloying of silicon into aluminum casting alloys. J Laser Appl 2000;12(4):166–70. [7] Paduru SK, Abu-Safe HH, Naseem HA, Al-Shariah A, Brown WD. CW Argon-ion laser initiated aluminum induced crystallization of amorphous silicon thin films. MRS Online Proc Library Arch 2004;808:339–44. [8] Zhu X, Huang Y, Liu D, et al. Surface properties of aluminum–silicon alloy irradiated by picosecond laser. Phys Procedia 2011;22:435–41. [9] McCay M, Kennedy N, Hopkins J, Dahotre NB. Wear enhancement of aluminum by laser surface alloying with silicon. Lasers Eng 2000;10(2):107–22. [10] Pokhmurs’ ka H, Student M, Chervins' ka N, et al. Structure and properties of aluminum alloys modified with silicon carbide by laser surface treatment. Mater Sci 2005;41(3):316–23. [11] Qu C, Li P, Fan J, et al. Aluminum oxide and silicon nitride thin films as anti-corrosion layers. Ceram Eng Sci Proc 2010;31(3):123–34. [12] Pardo A, Merino M, Arrabal R, Viejo F, Carboneras M, Coy A. Improvement of corrosion behavior of aluminum alloy/silicon carbide composites with lanthanum surface treatments. Corrosion 2006;62(2):141–51. [13] Ding H, Hihara LH. Localized corrosion of silicon-reinforced aluminum composites in Na2SO4 Solution. ECS Trans 2007;2(9):99–110. [14] Etzion R, Metson JB. Factors affecting corrosion resistance of silicon nitride bonded silicon carbide refractory blocks. J Am Ceram Soc 2012;95(1):410–5. [15] Schoukensa I, De Graeve I, De Strycker J, Terryn H. Effect of local surface microstructure and composition on the electrochemical behavior of hot dip aluminum–silicon coatings on steel. Corrosion 2012;68(11):982–93. [16] Ahmad Z, Paulette P, Aleem BA. Mechanism of localized corrosion of aluminum–silicon carbide composites in a chloride containing environment. J Mater Sci 2000;35 (10):2573–9. [17] Tang C, Cheng F, Man H. Laser surface alloying of a marine propeller bronze using aluminium powder. Part II: corrosion and erosion-corrosion synergism. Surf Coat Technol 2006;200(8):2594–601. [18] Kwok C, Wong P, Chan W, Cheng F, Man H. Laser surface alloying of Mn–Ni–Al bronze for cavitation erosion resistance. In: 29th int. congress on applications of lasers and electro-optics, vol 103; 2010. p. 1518–24. [19] Klassen R, Hyatt C, Roberge P. Passivation of laser-treated nickel aluminum bronze as measured by electrochemical impedance spectroscopy. Can Metall Q 2000;39 (2):235–46. [20] Mazurkevich A, Shiganov I, Tretyakov R, et al. Study of corrosion resistance of coatings of different composition prepared by laser surfacing. Chem Petroleum Eng 2012;48(5-6):380–3. [21] Garbacz H, Fortuna-Zalesna E, Marczak J, et al. Effect of laser treatment on the surface of copper alloys. Appl Surf Sci 2011;257(17):7369–74. [22] Kawazoe T, Ura A, Saito M, Nishikido S. Erosion characteristics of surface hardened Ni–Al bronze. Surf Eng 1997;13(1):37–40. [23] Hua Y, Bai Y, Ye Y, et al. Hot corrosion behavior of TC11 titanium alloy treated by laser shock processing. Appl Surf Sci 2013;283:775–80. [24] Shanjin L, Yang W. An investigation of pulsed laser cutting of titanium alloy sheet. Opt Lasers Eng 2006;44(10):1067–77. [25] Jeong Y-H, Kim W-G, Choe H-C. Electrochemical behavior of nano and femtosecond laser textured titanium alloy for implant surface modification. J Nanosci Nanotechnol 2011;11(2):1581–4. [26] Blanco-Pinzon C, Liu Z, Voisey K, et al. Excimer laser surface alloying of titanium with nickel and palladium for increased corrosion resistance. Corrosion Sci 2005;47 (5):1251–69. [27] Cui B, Song C, Wang G, et al. Reversible ferromagnetic phase transition in electrode-gated manganites. Adv Funct Mater 2014;24(46):7233–40. [28] Razavi RS, Gordani G, Man H. Corrosion behavior of laser nitrided Ti–6Al–4V alloy: a review. Defect and Diffusion Forum 2011;312-315:376–80. [29] Zhang CH, Tang LW, Hu F, Zhang S, Man HC. Study on wear and corrosion properties of Ti6Al4V alloy by laser gas nitriding. Adv Mater Res 2011;188–92. [30] Wong P, Kwok C, Man H, Cheng F. Corrosion behavior of laser-alloyed copper with titanium fabricated by high power diode laser. Corrosion Sci 2012;57:228–40. [31] Yu HJ. Research development of the laser surface modification to improve the corrosion-resistance of titanium and titanium alloys. Adv Mater Res 2013;748:184–7. [32] Sathish S, Geetha M, Pandey N, Richard C, Asokamani R. Studies on the corrosion and wear behavior of the laser nitrided biomedical titanium and its alloys. Mater Sci Eng: C 2010;30(3):376–82. [33] Kooi B, De Hosson JTM. Interfacial reactions in Ti–6Al–4V with laser-embedded sic particles and the origin of intergranular corrosion susceptibility of an Al–Mg alloy. MRS Online Proc Library Arch 2000;586:93–8. [34] Geetha M, Kamachi Mudali U, Pandey N, Asokamani R, Raj B. Microstructural and corrosion evaluation of laser surface nitrided Ti–13Nb–13Zr alloy. Surf Eng 2004;20 (1):68–74. [35] Yang M, Wilmott M, Luo J. Analysis of the electrochemical noise for localized corrosion of type A516-70 carbon steel. Corrosion 1998;54(11):869–76.

942

Anti-Corrosive Materials

[36] Newman R. The dissolution and passivation kinetics of stainless alloys containing molybdenum – 1. Coulometric studies of Fe
Cr and Fe
Cr
Mo alloys. Corrosion Sci 1985;25(5):331–9. [37] Marshall P, Burstein G. Effects of alloyed molybdenum on the kinetics of repassivation on austenitic stainless steels. Corrosion Sci 1984;24(5):463–78. [38] Sato N, Cohen M. The kinetics of anodic oxidation of iron in neutral solution: I. Steady growth region. J Electrochem Soc 1964;111(5):512–9. [39] Cabrera N, Mott N. Theory of the oxidation of metals. Rep Prog Phys 1949;12(1):163. [40] Cho E-A, Kim C-K, Kim J-S, Kwon H-S. Quantitative analysis of repassivation kinetics of ferritic stainless steels based on the high field ion conduction model. Electrochim Acta 2000;45(12):1933–42. [41] Frankel GS. Mechanism of Al alloy corrosion and the role of chromate inhibitors. Columbus, OH: Ohio State University; 2001. [42] Kwon H, Cho E, Yeom K. Prediction of stress corrosion cracking susceptibility of stainless steels based on repassivation kinetics. Corrosion 2000;56(1):32–40. [43] Kwon HS, Yeom K, Cho E, Kim JJ. Predicting SCC susceptibility of austenitic stainless steel by rapid scratching electrode technique. Mater Sci Forum 1998;969–78. [44] Mansfeld F, Shih H. Detection of pitting with electrochemical impedance spectroscopy. J Electrochem Soc 1988;135:1171–2. [45] Mansfeld F. Electrochemical impedance spectroscopy (EIS) as a new tool for investigating methods of corrosion protection. Electrochim Acta 1990;35(10):1533–44. [46] Mansfeld F, Lin S, Kim K, Shih H. Pitting and surface modification of SIC/Al. Corrosion Sci 1987;27(9):997–1000. [47] Mansfeld F, Lin S, Kim S, Shih H. Pitting and passivation of Al alloys and Al-based metal matrix composites. J Electrochem Soc 1990;137(1):78–82. [48] Mansfeld F, Wang Y, Lin S, Xiao H, Shih H. Detection and monitoring of localized corrosion by EIS, Electrochemical impedance: analysis and interpretation. West Conshohoken, PA: ASTM International; 1993. [49] Hunkeler F, Bohni H. Determination of pit growth rates on aluminum using a metal foil technique. Corrosion 1981;37(11):645–50. [50] Scully J, Frankenthal R, Hanson K, Siconolfi D, Sinclair J. Localized corrosion of sputtered aluminum and Al–0.5% Cu alloy thin films in aqueous HF solution I. Corrosion phenomena. J Electrochem Soc 1990;137(5):1365–73. [51] Mansfeld F, Little B. A technical review of electrochemical techniques applied to microbiologically influenced corrosion. Corrosion Sci 1991;32(3):247–72. [52] Oltra R, Keddam M. Application of impedance technique to localized corrosion. Corrosion Sci 1988;28(1):17–518. [53] Oltra R, Keddam M. Application of EIS to localized corrosion. Electrochim Acta 1990;35(10):1619–29. [54] Wenger F, Cheriet S, Talhi B, Galland J. Electrochemical impedance of pits. Influence of the pit morphology. Corrosion Sci 1997;39(7):1239–52. [55] Zhang P, Wu J, Zhang W, Lu X, Wang K. A pitting mechanism for passive 304 stainless steel in sulphuric acid media containing chloride ions. Corrosion Sci 1993;34 (8):13431351–91354. [56] Park J-J, Pyun S-I. Pit formation and growth of alloy 600 in Cl ion-containing thiosulphate solution at temperatures 298–573 K using fractal geometry. Corrosion Sci 2003;45(5):995–1010. [57] Darowicki K. Theoretical description of the measuring method of instantaneous impedance spectra. J Electroanal Chem 2000;486(2):101–5. [58] Darowicki K, Orlikowski J, Lentka G. Instantaneous impedance spectra of a non-stationary model electrical system. J Electroanal Chem 2000;486(2):106–10. [59] Darowicki K, Orlikowski J, Arutunow A. Investigations of the passive layer cracking by means of dynamic electrochemical impedance spectroscopy. Electrochim Acta 2003;48(28):4189–96. [60] Darowicki K, Ślepski P. Instantaneous electrochemical impedance spectroscopy of electrode reactions. Electrochimica Acta 2004;49(5):763–72. [61] Darowicki K, Ślepski P. Dynamic electrochemical impedance spectroscopy of the first order electrode reaction. J Electroanal Chem 2003;547(1):1–8. [62] Hoar T, Mears D, Rothwell G. The relationships between anodic passivity, brightening and pitting. Corrosion Sci 1965;5(4):279–89. [63] Marcus P, Herbelin J-M. The entry of chloride ions into passive films on nickel studied by spectroscopic (ESCA) and nuclear (36Cl radiotracer) methods. Corrosion Sci 1993;34(7):1123–45. [64] Yang W, Costa D, Marcus P. Resistance to pitting and chemical composition of passive films of a Fe–17% Cr alloy in chloride-containing acid solution. J Electrochem Soc 1994;141(10):2669–76. [65] Lin L, Chao C, Macdonald D. A point defect model for anodic passive films II. Chemical breakdown and pit initiation. J Electrochem Soc 1981;128(6):1194–8. [66] Chao C, Lin L, Macdonald D. A point defect model for anodic passive films I. Film growth kinetics. J Electrochem Soc 1981;128(6):1187–94. [67] Kolotyrkin JM. Pitting corrosion of metals. Corrosion 1963;19(8):261t–8t. [68] Strehblow H-H, Marcus P. Mechanisms of pitting corrosion. In: Oudar MP, editors. Corrosion mechanisms in theory and practice. New York: Marcel Dekker; 1995. p. 201–38. [69] Vetter K, Strehblow HH. Entstehung und Gestalt von Korrosionslöchern bei Lochfraß an Eisen und theoretische Folgerungen zur Lochfraßkorrosion. Berichte Bunsengesellschaft Physikalische Chem 1970;74(10):1024–35. [70] Sato N, Kudo K, Noda T. The anodic oxide film on iron in neutral solution. Electrochim Acta 1971;16(11):1909–21. [71] Macdonald DD. Passivity – the key to our metals-based civilization. Pure Appl Chem 1999;71(6):951–78. [72] Uhlig H. Passivity of metals. Frankenthal RP, Kruger J, editors. Princeton, NJ: The Electrochemical Society; 1978, p. 1. [73] Macdonald DD. The point defect model for the passive state. J Electrochem Soc 1992;139(12):3434–49. [74] Schmuki P, Bohni H. Semiconductive properties of passive films and susceptibility to localized corrosion, (German). Werkstoffe Korrosion 1991;42(1991):203–7. [75] Streicher M. Pitting corrosion of 18Cr–8Ni stainless steel. J Electrochem Soc 1956;103(7):375–90. [76] Rhodin TN. Oxide films on stainless steels. Corrosion 1956;12(3):41–53. [77] Tomashov N, Chernova G, Marcova O. Effect of supplementary alloying elements on pitting corrosion susceptibility of 18Cr–14Ni stainless steel. Corrosion 1964;20 (5):166t–73t. [78] Llewellyn D. Copper in steels. Ironmaking Steelmaking 1995;22(1):25–34. [79] Yilbas B, Khaled M, Karatas C. Laser remelting of alumina tile surfaces: corrosion testing in aqueous solution. Corrosion Eng Sci Technol 2011;46(4):477–80. [80] Meneau C, Andreazza P, Andreazza-Vignolle C, Goudeau P, Villain J-P, Boulmer-Leborgne C. Laser surface modification: structural and tribological studies of AlN coatings. Surf Coat Technol 1998;100:12–6. [81] Yilbas BS, Toor I-u-H, Malik J. Corrosion resistance of laser treated titanium alloy with B4C particles at the surface. Int J Mater Res 2014;105(10):975–82. [82] Bloyce A, Morton P, Bell T. Surface engineering of titanium and titanium alloys. ASM Handbook 1994;5:835–51. [83] Mudali UK, Reynders B, Stratmann M. Localised corrosion behaviour of Fe–N model alloys. Corrosion Sci 1999;41(1):179–89. [84] Halada G, Kim D, Clayton C. Influence of nitrogen on electrochemical passivation of high-nickel stainless steels and thin molybdenum-nickel films. Corrosion 1996;52 (1):36–46. [85] Vanini AS, Audouard J-P, Marcus P. The role of nitrogen in the passivity of austenitic stainless steels. Corrosion Sci 1994;36(11):1825–34. [86] Razavi RS, Salehi M, Ramazani M, Man H. Corrosion behaviour of laser gas nitrided Ti–6Al–4V in HCl solution. Corrosion Sci 2009;51(10):2324–9. [87] Toor IH, Yilbas BS, Karatas C, Hussein MA, Zafar MN. Electrochemical investigation of the effect of different laser surface treatments on Hastelloy G alloy. Int J Mater Res 2013;104(10):1007–12. [88] Yilbas¸ B, S¸Sahin A. Turbulent boundary layer approach allowing chemical reactions for CO2 laser oxygen-assisted cutting process. Proc Inst Mech Eng, Part C: J Mech Eng Sci 1994;208(4):275–84. [89] Osorio W, Cheung N, Spinelli J, Cruz K, Goulart P, Garcia A. Thermally and chemically induced microstructual modifications affecting the electrochemical corrosion behavior of an Al–9WT% Si casting alloy. J New Mater Electrochem Syst 2008;11(3):205. [90] Shabestari S, Moemeni H. Effect of copper and solidification conditions on the microstructure and mechanical properties of Al–Si–Mg alloys. J Mater Process Technol 2004;153:193–8.

Anti-Corrosive Materials

943

[91] Goulart P, Osório W, Spinelli J, Garcia A. Dendritic microstructure affecting mechanical properties and corrosion resistance of an Al–9 wt% Si alloy. Mater Manuf Process 2007;22(3):328–32. [92] Wong T, Liang G, Tang C. The surface character and substructure of aluminium alloys by laser-melting treatment. J Mater Process Technol 1997;66(1-3):172–8. [93] Ćurković L, Jelacˇa MF, Kurajica S. Corrosion behavior of alumina ceramics in aqueous HCl and H2SO4 solutions. Corrosion Sci 2008;50(3):872–8. [94] Yilbas B, Toor I-u-H, Malik J, Patel F. Laser treatment of high strength low alloy steel and electrochemical response of the surface. Ind Lubrication Tribol 2015;67 (2):166–71. [95] Aqida S, Calosso F, Brabazon D, Naher S, Rosso M. Thermal fatigue properties of laser treated steels. Int J Mater Forming 2010;3:797–800. [96] Khaled M, Yilbas B, Shirokoff J. Electrochemical study of laser nitrided and PVD TiN coated Ti–6Al–4V alloy: the observation of selective dissolution. Surface Coat Technol 2001;148(1):46–54. [97] Yilbas B, Malik J, Patel F, Karatas C. Electrochemical testing of laser treated bronze surface. J Alloys Compounds 2013;563:180–5. [98] Willenbruch R, Clayton C, Oversluizen M, Kim D, Lu Y. An XPS and electrochemical study of the influence of molybdenum and nitrogen on the passivity of austenitic stainless steel. Corrosion Sci 1990;31:179–90. [99] Ha H, Jang H, Kwon H, Kim S. Effects of nitrogen on the passivity of Fe–20Cr alloy. Corrosion Sci 2009;51(1):48–53. [100] Laycock N, Noh J, White S, Krouse D. Computer simulation of pitting potential measurements. Corrosion Sci 2005;47(12):3140–77. [101] Burstein G, Vines S. Repetitive nucleation of corrosion pits on stainless steel and the effects of surface roughness. J Electrochem Soc 2001;148(12):B504–16.

Further Reading Ahmad Z. Principles of corrosion engineering and corrosion control, ISBN: 978-0-7506-5924-6. Munger CG. Corrosion prevention by protective coatings, ISBN-13: 978-1575900889. Revie RW. Corrosion and corrosion control: an introduction to corrosion science and engineering, R. Winston Revie, ISBN: 978-0-470-27725-6.

Relevant Websites http://www.chemistryexplained.com/Co-Di/Corrosion.html Chemistry Explained: foundations and Applications. http://www.electrochem.org/corrosion-science ECS, The Electrochemical Society. http://www.sciencedirect.com/science/journal/0010938X?sdc=2 ScienceDirect. http://www.springer.com/gp/book/9783540006268 Springer.

2.29 Desulfurization Materials Mashallah Rezakazemi, Shahrood University of Technology, Shahrood, Iran Zhien Zhang, Chongqing University of Technology, Chongqing, China; Ningde Normal University, Ningde, China; and Chongqing University, Chongqing, China r 2018 Elsevier Inc. All rights reserved.

2.29.1 Introduction 2.29.2 Current Desulfurization Methods 2.29.2.1 Wet Type 2.29.2.2 Dry Type 2.29.2.3 Semiwet Type 2.29.3 Adsorptive Desulfurization 2.29.4 Adsorption Kinetics and Adsorption Isotherms 2.29.4.1 Adsorption Capacity 2.29.4.2 Activation Energy 2.29.4.3 Adsorption Selectivity 2.29.4.4 Adsorption Kinetics 2.29.4.4.1 Pseudo first-order kinetic model 2.29.4.4.2 Pseudo second-order kinetic model 2.29.4.5 Equilibrium Adsorption 2.29.4.5.1 Langmuir isotherm 2.29.4.5.2 Freundlich isotherm 2.29.5 Mechanism of Adsorptive Desulfurization 2.29.6 Adsorbent Properties 2.29.6.1 Dynamic Properties 2.29.6.2 Chemical Properties 2.29.6.3 Physical Properties 2.29.7 Challenges that Adsorptive Desulfurization Process Is Facing 2.29.8 Adsorptive Desulfurization Materials 2.29.8.1 Zeolites 2.29.8.2 Activated Carbon 2.29.8.3 Graphite 2.29.8.4 Silica 2.29.8.5 Calcium Oxide 2.29.8.6 Copper Oxide 2.29.8.7 Magnesia 2.29.8.8 Manganese Oxide 2.29.8.9 Titania 2.29.8.10 Zirconia 2.29.8.11 Ceria 2.29.8.12 Mixed Oxides 2.29.8.13 Metal Organic Frameworks 2.29.8.14 Nickel-Based Adsorbents 2.29.8.15 Adsorptive Polymers 2.29.8.16 Sludge Derived Adsorbents 2.29.8.17 Porous Glass Materials 2.29.8.18 Clay Mineral Adsorbents 2.29.9 Factors Affecting the Adsorption Process 2.29.10 Conclusions and Future Work Acknowledgments References Further Reading Relevant Websites

Nomenclature AC

944

Activated carbon

ACFs ADS

945 947 947 948 948 950 951 951 952 952 952 952 952 953 953 954 954 955 955 955 956 956 956 957 958 960 960 961 963 963 964 964 964 965 965 965 966 966 969 970 971 972 972 973 973 979 979

Activated carbon fibers Adsorptive desulfurization

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00263-7

Desulfurization Materials

AFC AM AMSA AS AZ BDS BET BT BTO CaO CFB CNT COS DAC DBT DBTO DFT DGC DMDBT DMF EPA EPR ED FCC FGD FTIR HDS HPAs JP-5

2.29.1

Alkaline fuel cell Activated manganese oxide Amino methane sulfonic acid Ammonium sulfate Activated zinc oxide Biodesulfurization Brunauer–Emmett–Teller Benzothiophene Benzothiophene sulfone Calcium oxide Circulating fluidized bed Carbon nanotube Carbonyl sulfide Demineralized activated carbon Dibenzothiophene Dibenzothiophene sulfone Density functional theory Dry gel conversion Dimethyl dibenzothiophene N, N-dimethyl formamide Environmental Protection Agency Electron paramagnetic resonance Ethylene diamine Fluid catalytic cracking Flue gas desulfurization Fourier transform infrared spectroscopy Hydrodesulfurization HeteroPOLYACIds Jet Propellant 5

945

JP-10 Jet Propellant 10 LNG Liquefied natural gas MBT Methylated benzothiophenes MCFC Molten carbonate fuel cell MCM-41 Mobil Composition of Matter No. 41 MIP Molecular imprinted polymer MLAC Manganese oxides-loaded activated carbon MMT Metal Impregnated montmorollonite clay MOFs Metal organic frameworks MSN Mesoporous silica nanoparticle ODS Oxidative desulfurization PAC Powdered activated carbon PADS Physical adsorption desulfurization PAFC Phosphoric acid fuel cell PEMFC Proton-exchange membrane fuel cell PGB Porous Glass Bead PSAC Palm Shell Activated Carbon PWA Phosphotungstic Acid RADS Reactive Adsorption Desulfurization RTI Research Triangle Institute SBA-15 Santa barbara amorphous SOFC Solid oxide fuel cell SRT Sulfur removal technology TDDFT Time-dependent density functional theory TP Thiophene TPO Temperature programmed oxidation TReND Transport reactor naphtha desulfurization 4, 6-DMDBT 4, 6-dimethyl dibenzo thiophene

Introduction

Desulfurization is removal of sulfur or sulfur derivative compounds commonly from flue gas or fuels. The most usually needed desulfurization is in natural gas and for coal, oil, and flue gas. Sulfur compounds in natural gas, crude oil, and natural gas liquids (LNG) can take numerous forms such as carbonyl sulfide (COS), hydrogen sulfide (H2S), sulfur oxides (SOx), and all derivatives of mercaptans. Organosulfur derivative compounds in fuels can be simply changed to sulfur oxides (SO2) and fine powder of metal sulfated during combustion. All those organosulfur derivative compounds are taken into account as major pollutants in the air. SO2 is one of the major acid gas emissions emitted from industries. SO2 from the burning of fossil fuels has resulted in negative impacts on the environment and human health such as acid rain and respiratory diseases due to SO2 formation upon combustion of fuels. It is one of the most harmful gases among the gases released from the industrial processes such as power plants, syngas plant, fertilizer, etc. Among them, the thermal power plants contribute most to the SO2 emissions [1–3]. Those powders of metal-sulfur create huge injuries to human health. Consequently, the greatest challenge of industries is the demand for fuels having a low content of organosulfur derivative compounds. The US Environmental Protection Agency (EPA) has established regulations that decree petrochemical complexes and refineries to decrease sulfur contents of gasoline down to 30 ppmw and diesel down to 15 ppmw. The sulfur content in different countries’ produced crude oil worldwide in the year 2010 is reported in Table 1 [4]. Desulfurization is a technology used to separate SO2 from its emitting sources such as exhaust flue gases of fossil-fuel power plants, oil refineries, etc. The crucial requirement to decrease sulfur loading in fuels to nearly zero content is imposed by mandatory environmental and health protocols as well as the accurate tolerance needed for their use in fuel applications. Numerous methods have been applied in the desulfurization process including wet scrubbing by means of slurry of alkaline sorbent, commonly lime or limestone, or seawater, spray-dry scrubbing, SNOX desulfurization, etc. Desulfurization efficiency and energy consumption during the process are two main factors when choosing the appropriate materials for SO2 separation [5–7]. Jet fuels are the kerosene-based fuels for different applications. Jet propellant 5 (JP-5) is designed for navy shipboard activities. JP-5 is identified by its high flash point and low volatility. JP-8 is one of the most challenging fuels to desulfurize. Table 2 represents a volumetric and gravimetric analysis of various common fuels sources including diesel, gasoline, hydrogen, biodiesel oil, methanol, etc. As can be seen from the table, the gravimetric energy density of the JP-8 fuel is noteworthy compared with hydrogen gas and particularly compared with lithium ion battery.

946

Table 1

Desulfurization Materials

The sulfur content in different countries’ produced crude oil worldwide in the year 2010

Region

Crude oil gravity (API)

Sulfur weight (%, 1990)

Production (tpd)

Crude oil gravity (API)

Sulfur weight (%, 2010)

Production (tpd)

Alaska Canada California Rest of United States Africa Europe Latin America Middle East Far East World Average

26.970 31.400 17.430 35.110 31.280 33.200 25.060 33.730 33.800 31.300

1.11 1.52 1.59 0.86 0.17 1.09 1.62 1.69 1.09 1.13

1954 2000 970 4510 7000 16,330 7770 29,100 16,330 70,800

28.340 32.000 18.730 36.930 32.640 33.700 27.100 34.350 37.300 32.810

0.99 1.62 2.60 0.88 0.18 1.10 1.82 1.71 1.10 1.27

1645 2500 951 2470 6100 15,530 9850 35,760 15,530 83,450

Source: Reproduced with permission from Chandra Srivastava V. An evaluation of desulfurization technologies for sulfur removal from liquid fuels. RSC Adv 2012;2:759–83. doi:10.1039/C1RA00309G.

Table 2

Volumetric and gravimetric analysis of various common fuels sources (at 1 bar)

Fuel

Gravimetric energy density (MJ/kg)

Volumetric energy density (MJ/L)

Ref.

JP-8 JP-5 Diesel Biodiesel oil Gasoline (petrol) Hydrogen (liquid) Methanol Lithium ion battery

43.4 42.4 42.5 42.2 44 8 19.7 0.6

– – 37 33 35 0.011 18 –

[8] [9] [8] [9] [10] [11] [12] [13]

JP-8 involves various sulfur derivative compounds. Lee and Ubanyionwu [14] determined the total sulfur concentration of four JP-8 from Fort Belvoir, Virginia. The analysis of JP-8 fuel is reported in Table 3. As observed, 2, 3-dimethylbenzothiophene and 2, 3, 7-trimethyl benzo thiophene (Fig. 1) are the major sulfur contributors. Desulfurized fuel can be used as a fuel for an internal combustion engines, as a raw material used in fluid catalytic cracking (FCC) processes, as fuel for boilers for power generation, to start up the fuel processor at low temperatures, and as fuel for various types of fuel cells such as solid oxide fuel cells (SOFCs), phosphoric acid fuel cells (PAFCs), proton-exchange membrane fuel cells (PEMFCs), molten carbonate fuel cells (MCFCs), and alkaline fuel cells (AFCs) and jet engines [15]. Organosulfur derivative compounds in the different fuels may simply poison the anode catalysts in the different types of fuel cell processors, and anode electrodes may also be damaged by organosulfur derivative species. Hence, various types of fuel cells have dissimilar restrictions in total sulfur compositions as reported in Table 4 [16]. Fuel cells need a basically sulfur-free feedstock to avoid poisoning of anode catalyst. Conoco Phillips Company has presented the S Zorb sulfur removal technology (SRT) to produce fuel with a low level of organosulfur derivative compounds by reactive adsorption. The S Zorb SRT process is schematically depicted in Fig. 2 [17]. In the S Zorb SRT technology, organosulfur derivative compounds react with the S Zorb sorbent, which preserves the sulfur element from the compound while the hydrocarbon part of the compound is released back into the flow stream. This exceptional reaction mechanism does not produce free H2S, and, hence prevents issues related to mercaptan recombination and H2S inhibition [18]. In Fig. 3, the S Zorb SRT process is compared with a hydrotreating process [18]. The S Zorb technology works at 340–420°C temperatures and 30– 300 psi pressure in a fluidized bed reactor with hydrogen and a solid adsorbent to separate thiophenic based species [19]. The adsorbent is made up of alumina, zinc oxide, nickel oxide, and silica. The adsorbent is then regenerated firstly by nitrogen, and then by oxidation with air, and lastly reduced. A suitable desulfurization adsorbent should be regenerated for multiple process cycles. In Fig. 4, the reaction rates for various organosulfur derivative compounds that are naturally present in cracked naphthas are compared. This figure, which is promising in the order of organosulfur reactivity, compares the reaction rates on a relative basis, rather than on an absolute basis, where the first-order reaction rate of organosulfur compounds has been relatively normalized to thiophene (T). As can be seen in the figure, thiophene is the hardest compound to separate for both S Zorb SRT and hydrotreating processes but the relative reactivity for alkylated benzothiophene, alkylated thiophenes, and benzothiophene is increased for the S Zorb SRT process [18]. Also, the Research Triangle Institute (RTI) has presented a reactive adsorption process called the transport reactor naphtha desulfurization (TReND) technology for fuel desulfurization, which is similar to S-Zorb process, but utilizes copper or iron oxide endorsed by Al2O3-ZnO adsorbent at temperatures higher than 400°C without or with hydrogen to separate organosulfur derivative compounds from FCC naphtha [17].

Desulfurization Materials

Table 3

947

Sulfur restrictions in various fuel cells

Retention time (min)

Compound

Amount (% of total sulfur concentration) June 2003

– – – 2, 7-Dimethyl benzo thiophene Dimethyl benzo thiophene isomers Dimethyl benzo thiophene isomers Dimethyl benzo thiophene isomers 3, 5-Dimethyl benzo thiophene 2, 3-Dimethyl benzo thiophene Trimethyl benzo thiophene isomers 2, 5, 7-Trimethyl benzo thiophene Trimethyl benzo thiophene isomers Trimethyl benzo thiophene isomers 2, 3,7-Trimethyl benzo thiophene 2, 3, 5-Trimethyl benzo thiophene/ 2, 3, 6-Trimethyl benzo thiophene – – Di benzo thiophene –

o6.726 6.726 6.789 7.263 7.324 7.391 7.430 7.490 7.534 7.909 7.961 8.021 8.059 8.129 8.237 8.505 8.715 9.998 11.11 Total ppmw S

October 2004

July 2005

April 2006

36.01 1.8 1.4 2.03 2.57 1.8 1.54 1.24 5.36 1.68 0.96 1.22 1.68 5.72 2.35

32.75 1.69 1.97 2.32 2.12 1.84 1.39 1.39 5.36 1.47 1.21 1.22 1.39 5.3 1.96

37.1 1.66 1.75 2.93 3.02 1.99 2.52 1.09 5.98 1.5 1.51 2.11 1.09 3.92 2.19

34.75 1.91 1.73 2.49 2.19 1.89 2.12 1.29 5.39 1.51 1.1 1.92 1.34 4.11 2.61

0.82 1.23 0.89 0.74 325

0.96 0.9 0.77 0.75 396

0.72 0.74 0.23 0.09 1096

1.12 0.94 0.44 0.2 648

Source: Reproduced with permission from Sun X, Tatarchuk BJ. Photo-assisted adsorptive desulfurization of hydrocarbon fuels over TiO2 and Ag/TiO2. Fuel 2016;183:550–6.

S

S

Fig. 1 Structure of 2, 3-dimethylbenzothiophene (left) and 2, 3, 7-trimethyl benzo thiophene (right). Table 4

Chemical adsorption and physical adsorption properties

Sulfur (i.e., H2S and COS)

PEMFC

SOFC

PAFC

MCFC

AFC

o0.1 ppm

o1 ppm

o50 ppm

o0.5 ppm

Unknown

Source: Reproduced with permission from Reed NJ. A comparative study of adsorption desulfurization of liquid transportation fuels over different sorbents for fuel cell applications, The Pennsylvania State University; 2008.

Black and Veatch Pritchard Inc. and Alcoa Industrial Chemicals have presented the IRVAD process for low content of the organosulfur fuel. The IRVAD technology is also similar to the S Zorb SRT process while operating at less severe conditions. The IRVAD and S Zorb SRT technologies have been explained by Ito and Rob van Veen [17] using the reactive adsorption mechanism presented in Fig. 5. In this reactive adsorption mechanism, the process commonly uses Ni functioning as an HDS site to create ZnS from ZnO.

2.29.2

Current Desulfurization Methods

Up to now, numerous attempts have been made to propose new methods that either stand alone or can be coupled with the conventional desulfurization technologies, to effectively capture sulfur compounds in fuels. The SO2 separation methods are mainly divided into several wet, semiwet (or called semidry), and dry based processes (Fig. 6). Some methods for SO2 separation have been used for more than 160 years.

2.29.2.1

Wet Type

Generally, the most common desulfurization methods are the wet types like ammonia [20–24], limestone [25–27], and seawaterbased techniques [28–30]. The total and the running costs of the dry methods are much less, owing to their simplicity, easier

948

Desulfurization Materials

Product Sorbent receiver Sorbent storage

To SRU

Reactor

Regenerator Feed + H2

Air

H2

Fig. 2 A schematic of S Zorb sulfur removal technology (SRT) process. Reproduced with permission from Ito E, van Veen JAR. On novel processes for removing sulphur from refinery streams. Catal Today 2006;116:446–60.

Hydrotreating Catalyst H2S +

+ 3H2 S S Zorb treatment + 3H2 + Sorbent

Sorbent-S + H2O +

S Fig. 3 Comparison between the S Zorb sulfur removal technology (SRT) process and hydrotreating process. Reproduced with permission from Laan JV. ConocoPhillips S Zorb gasoline sulfur removal technology: unique chemistry, proven performance, and optimized design. Available from: http://www.icheh.com/Files/Posts/Portal1/S-Zorb.pdf.

installation, and less space and water consumption. Thus, they are good choices for the current desulfurization processes. For example, the dual alkali process [31] is a widely used method for SO2 separation. The process includes two steps as follows: Absorption step Regeneration step

SO2 þ H2 O þ Na2 SO3 →2NaHSO3

2NaHSO3 þ Na2 CO3 →2Na2 SO3 þ CO2 þ H2 O

ð1Þ ð2Þ

Another common SO2 separation technology, the ammonium sulfate (AS) process, i.e., ammonia-based flue gas desulfurization (FGD), is accomplished with either aqueous or anhydrous ammonia and converts the absorbed SO2 into crop fertilizers. It is advantageous due to low investment cost, reuse of AS fertilizers, and no greenhouse gas emissions produced [32–34]. This technique has been widely and commercially applied by companies such as Dakota Gasification, Zaklady Azotowe Pulawy Heating and Power Plant, etc.

2.29.2.2

Dry Type

In terms of low SO2 concentrations and temperature, the dry methods are superior to the wet ones due to less SO2 recovery efficiency [35,36]. In addition, this process reduces the water and power consumption, and the produced wastewater. The water and reagent could be operated separately [37]. On the downside, the disadvantages are low SO2 removal efficiency and high sorbent cost [38].

2.29.2.3

Semiwet Type

The semiwet method could be used for the small to medium sized industrial boilers because the adsorber is pneumatically and directly injected as a dry powder [39,40]. This technique is also widely used in the multipollutant removal areas. However, the removal efficiency by the semiwet method is lower than that by a wet scrubber [41,42].

Desulfurization Materials

949

10.0 Relative reactivity for S Zorb SRT

10

7.0

4.0

3.7 2.1 1.0 1 Thiophene

C1−C2 Thiophenes

C3+ Thiophenes

AlkylBenzothiophenes

Benzothiophene

Mercaptans and sulfides

5.5

5.5

Benzothiophene

Mercaptans and sulfides

Relative reactivity for HDS

10

2.0 0.3

1.0

1.0

Thiophene

C1−C2 Thiophenes

1 C3+ Thiophenes

AlkylBenzothiophenes

Fig. 4 Comparison between the reaction rates for various organosulfur derivative compounds that are naturally present in cracked naphthas. SRT, sulfur removal technology; hydrodesulfurization. Reproduced with permission from Laan JV. ConocoPhillips S Zorb gasoline sulfur removal technology: unique chemistry, proven performance, and optimized design. Available from: http://www.icheh.com/Files/Posts/Portal1/S-Zorb.pdf.

NiO

NiSsurf

Ni + H2 ZnO

ZnO − H2O

+ C4H4S

ZnO

− C4H6 in H2 −H2S/ +H2S − H2O

NiSsurf Ni

Ni ZnO

−H2S/ +H2S

ZnO

− H2O in H2 + C4H4S

+ C4H4S

ZnO

ZnS

− C4H6 NiSsurf

− C4H6 ZnS Fig. 5 Schematic description of reactive adsorption mechanism with NiO–Zn adsorbent. Reproduced with permission from Ito E, van Veen JAR. On novel processes for removing sulphur from refinery streams. Catal Today 2006;116:446–60.

950

Desulfurization Materials

Sodium sulfite

Calcium hydroxide Limestone magnesium oxide Wet

Desulfurization method

Citres process

Ammonium sulfite Semi-wet

Sulfuric acid and catalyst Copper oxide

Dry

Catalytic oxidation Manganese oxide application

Fig. 6 Desulfurization method classifications.

Several approaches have been proposed for desulfurization fuels including biodesulfurization (BDS) [43], extraction–oxidation [44], extraction using impregnant [45], oxidative desulfurization (ODS) [46], hydrodesulfurization (HDS) [47], alkylation [48], and adsorption. Fuels that have larger sulfur derivative species such as methylated benzothiophenes (MBT) and dibenzo thiophenes (DBT) may not possible to desulfurize by HDS. When sulfur in these species is sterically hindered, it is not able to bind to attach and therefore remain in the fuel, which is recognized as refractory sulfur compounds. JP-8 is full of MBT and therefore HDS is not a practical technique to desulfurize it. Fig. 7 exhibits the common organosulfur compounds available in various fuels, classified by their difficulty to be separated by HDS. As organosulfur species become larger and more methylated, the sulfur element’s tendency to bind the catalyst becomes less.

2.29.3

Adsorptive Desulfurization

Among the mentioned technologies used to effectively capture sulfur compounds, the selective adsorption process presents one of the most effective techniques available for stand-alone desulfurization, since it presents an easy to handle, economical, compact, and regenerable process. Adsorption desulfurization is on the basis of selective adsorbing sulfur compounds on a solid surface. Numerous selective materials comprising one or a blend of transition metals are commonly used for the adsorptive removal of sulfur compounds. The development of these types of adsorptive desulfurization (ADS) materials is the main focus of this chapter. ADS is based on the capability of a nanoadsorbent to selectively remove sulfur derivative compounds from the fuel/flue gas. The effectiveness of this technique is directly related to the characteristics of the adsorbent nanomaterials: adsorption capacity, selectivity to sulfur derivative compounds relative to fuels, regenerability, and durability. The adsorption capacities of various materials such as activated alumina, metal supported activated carbons, metal supported zeolites, and MOF-based adsorbents for removing DBT are reported in Table 5. Two different kinds of adsorption desulfurization technique can be taken [58]: 1. Reactive adsorption desulfurization (RADS): this approach contains a chemical reaction between sulfur derivative compounds and adsorbent surface. Organosulfur derivative compounds are commonly linked to the adsorbent surface as a sulfide. The adsorbent regeneration may be performed by thermal treatment or by flushing spent adsorbent with desorbent. In this method, organosulfur derivative compounds are commonly separated as SOx, H2S, and S element [58]. Ni/ZnO-based adsorbents showed a high desulfurization capability and were extensively employed in RADS processes [59–61]. 2. Physical adsorption desulfurization (PADS): in this approach, the organosulfur derivative compounds are not chemically changed during the desulfurization. The required energy for the adsorbent regeneration is directly related to the strength of the ADS [58].

Desulfurization Materials

R

951

Thiols

SH

Gasoline contaminants

R

Thiophenes

S R S R Relative reaction rate

S

Jet fuel contaminants

Benzothiophenes

R S R S

R

Diesel contaminants

S R S Dibenzothiophenes

R S

Difficulty to remove by HDS Fig. 7 Common organosulfur compounds available in various fuels, classified by their difficulty to be separated by hydrodesulfurization (HDS) Reproduced with permission from Palomino JM. Mesoporous inorganic materials for the desulfurization of jet fuel; 2014.

Table 5

The adsorption capacities of various materials for removing dibenzo thiophenes (DBT)

Adsorbent

Adsorptive capacity (mg (S) g

Porous glass Cu (I)-zeolite Cu (I)-zeolite Nanoporous carbons Ag/Cu/Fe-supported activated carbons Pt/Ag-supported activated carbons Ni/SiO2–Al2O3 Activated alumina Microporous coordination polymers Mo-supported MOF-5 Ni-zeolite

8.58 12.64 8.90 10.77 8.40 7.50 3.40 2.40 25.10 16 2.10

2.29.4 2.29.4.1

adsorbent)

Initial concentration (ppmw S)

Temperature (K)

Ref.

500 364 297 20 376 500 687 687 300 1000 14.5

303 298 298 298 303 298 298 298 298 298 493

[49] [50] [50] [51] [52] [53] [54] [54] [55] [56] [57]

Adsorption Kinetics and Adsorption Isotherms Adsorption Capacity

The adsorption capacity of organosulfur species on adsorbent at any time and at equilibrium are measured as follows, respectively [62]: qt ¼ ðC0 −Ct Þ

V W

ð3Þ

952

Desulfurization Materials

qe ¼ ðC0 −Ce Þ

V W

ð4Þ

where q is the amount of organosulfur species adsorbed per gram of an adsorbent at any time (qt) and at equilibrium (qe), Co (mg/l) and Ct (mg/l) are the concentration at initial and time t, respectively. V (l) and W (g) are the volume of a solution and the weight of an adsorbent, respectively. The desulfurization efficiency (organosulfur removal efficiency), η (%), is determined as follows [62]:   C0 −Ct η¼  100 ð5Þ C0

2.29.4.2

Activation Energy

The reaction rate of physical adsorption is usually fast since the activation energy is not involved in the physical adsorptive process. As the activation energy is involved in chemical adsorptive process, the reaction rate is related to the rate coefficient by the Arrhenius equation [63,64]:   Ea k ¼ Aexp − ð6Þ RT where Ea and A are the activated energy and the preexponential or frequency factor, respectively. K and T are the rate coefficient for chemical adsorption, and temperature, respectively. R is the universal gas constant (R¼ 8.314 J/mol K). In order to determine the activation energy, a plot of the ln(k) against (1/T) must be depicted to find the slope that is −Ea/R. For the organosulfur adsorptive process, which takes place by chemical adsorption, the Arrhenius graph explains how variations in temperature influence the reaction rate [63,64].

2.29.4.3

Adsorption Selectivity

The selectivity of a nanoadsorbent expresses the removal of organosulfur derivative species as an explanation of species tendency for competing reactions to adsorb on the nanoadsorbent surface. The relative selectivity parameter, α, is also defined as the kinetic adsorptive capacity at breakthrough for species I (Wi) and a reference species, j (Wj) as follows [63,64]: α¼

2.29.4.4

Wi Wj

ð7Þ

Adsorption Kinetics

Understanding the governing adsorption kinetics makes it possible to control the mechanism of the adsorption process including chemical reaction and mass transfer and find the time needed to reach equilibrium conditions during the adsorption process as well [65,66].

2.29.4.4.1

Pseudo first-order kinetic model

The pseudo first-order rate kinetic model of the Lagergren expression is given below [62]: dqt ¼ k1 ðqe −qt Þ ð8Þ dt where k1 (min) is the constant for pseudo first-order adsorption rate. By integrating this expression and taking the initial conditions of qt ¼ 0 at t ¼ 0 and q ¼ qt at t¼ t, the following expression is yielded: lnðqe −qt Þ ¼ lnqe −k1 t

ð9Þ

A linear relationship of ln(qe−qt) in terms of t shows that pseudo first-order rate kinetic model is valid to correlate the experimental data. The amount of organosulfur species adsorbed per gram of an adsorbent at equilibrium (qe) and the adsorption rate (k1) are determined from the intercept and slope of the line, respectively.

2.29.4.4.2

Pseudo second-order kinetic model

The pseudo second-order rate kinetic model is given below [67]: dqt ¼ k2 ðqe −qt Þ2 ð10Þ dt where k2 (g/mg min) is the constant for pseudo first-order adsorption rate. By integrating this expression and taking the initial conditions of qt ¼ 0 at t ¼ 0 and q ¼ qt at t¼ t, the following expression is yielded: t 1 1 ¼ þ t qt k2 q2e qe

ð11Þ

A linear relationship of t/qt in terms of t shows that pseudo second-order rate kinetic model is valid to correlate the experimental data. The second model is based on the hypothesis that the rate-limiting step must be followed the chemical adsorption mechanism [65].

Desulfurization Materials 2.29.4.5

953

Equilibrium Adsorption

The isotherm of an adsorption is a mathematical model that explains a relation between the amount of a compound adsorbed on the surface of an adsorbent and its composition in the liquid at equilibrium condition and constant temperature [62]. The isotherm describes how the adsorbed compound distributes between the solution and the adsorbent at equilibrium condition. This adsorption isotherm is assumed that is related to the heterogeneity/homogeneity of the adsorbents, and the conceivable interaction between the compounds willing to adsorb on the surface of an adsorbent [65]. The two adsorption isotherm models that are regularly used are the Freundlich and Langmuir models.

2.29.4.5.1

Langmuir isotherm

The Langmuir isotherm model is based on the assumptions that the adsorption process occurs at particular homogeneous sites on the surface of the adsorbent [62]. Moreover, this model is also based on the assumptions that when a molecule is adsorbed onto a site, other molecules are not able to adsorb at that site. When the adsorption of organosulfur from fuels obeys the Langmuir isotherm, the theoretical relationship between the volume of organosulfur in the fuels and the volume of organosulfur attached on the surface of adsorbent at equilibrium conditions, the concentration of adsorbed organosulfur can also be described as follows [63,64]: q¼

Kqm Ce 1 þ KCe

ð12Þ

where q and K are the concentration of adsorbed organosulfur and the adsorption equilibrium constant, respectively. qm and Ce are the maximum adsorption capacity and the concentration of the organosulfur compound in fuel, respectively [63,64]. The Eq. (12) can be rearranged into a linear form as shown below: Ce 1 Ce ¼ þ q Kqm qm

ð13Þ

Since the concentration of the organosulfur compound in fuels is relatively low (i.e., Ce{1), therefore, the reaction rate can be considered be first order. A graph of Ce/q against Ce would yield a straight line, which determines the adsorption capacity term, qm, from the slope (1/qm) as well as the adsorption equilibrium constant, K, from the intercept (1/qmK). The adsorption terms qm and K relatively specify the tendency of the adsorbate to the surface of nanoadsorbent and explains the physical, chemical, and dynamic characteristic of a nanoadsorbent [63,64]. The separation factor is useful to determine the value of the dimensionless adsorption constant as follows [68]. r¼

1 1 þ KC0

ð14Þ

The value of r shows the type of the adsorption isotherm. With increasing temperature, the value of r decreases indicating the adsorption process is desirable at high temperatures [66]. In Table 6, the values of r and the relevant types of adsorption isotherm are given. The Brunauer–Emmett–Teller (BET) model is a developed model of the Langmuir isotherm that considers the multilayers of coverage that are not allowed in the Langmuir isotherm. The BET expression is illustrated in Eq. (3). The BET expression is extensively used to measure the surface area of adsorbent surfaces by calculating the physical adsorption of nitrogen in terms of pressure [63,64].   1 c−1 P0 1  P0
 ¼ ð15Þ þ vm c P vm c ν P −1

where P and P0 are equilibrium pressure and saturation pressure, respectively. Here, v and vm are adsorbed gas quantity and monolayer adsorbed gas quantity, respectively [63,64]. Table 6 Total sulfur concentration of four JP-8 from Fort Belvoir, Virginia Values of r

Type of adsorption isotherm

r40 r ¼1 0oro1 r ¼0

Unfavorable Linear Favorable Irreversible

Source: Reproduced with permission from LeeIC, Ubanyionwu HC. Determination of sulfur contaminants in military jet fuels, Fuel 2008;87:312–18. Available from: http://dx.doi.org/10.1016/j.fuel.2007.05.010.

954

Desulfurization Materials

2.29.4.5.2

Freundlich isotherm

The Freundlich isotherm explores the adsorption process on heterogeneous adsorbents at equilibrium condition. In this model, it is assumed that adsorbate is adsorbed on various sites of heterogeneous adsorbents with different energy [65]. In the Freundlich isotherm, in contrast to the Langmuir isotherm, a monolayer capacity is not assumed [62]. The theoretical Freundlich isotherm is expressed as follows [69]: qe ¼ Qf Ce1=n

ð16Þ

where Qf (mg/g) and 1/n (l/mg) are an indicator of the adsorption capacity and the adsorption intensity, respectively. This equation can be expressed in a linear form below: 1 lnqe ¼ lnQf þ lnCe n

ð17Þ

The Freundlich parameters (1/n and Qf) are calculated from intercepts and slopes of a graph of ln qe versus ln Ce. The value of n shows the desirability of the adsorption process in which n41 reveals the adsorption process is desirable at certain circumstances [70].

2.29.5

Mechanism of Adsorptive Desulfurization

The adsorption of organosulfur compounds onto adsorbent surfaces takes place through chemical adsorption (chemisorption) or physical adsorption (physisorption). The basis of physical adsorption is on the fundamental of Van der Waals forces and electrostatic forces in molecules with a permanent dipole moment. The forces that physically adsorb an organosulfur compound into the nanoadsorbent surface do not change the adsorbate species and are commonly weak forces. While in chemical adsorption, a chemical bond must be formed between the surface of an adsorbent nanomaterial and the adsorbate species. The surface of an adsorptive nanomaterial has numerous free valencies due to the broken covalent bonds among elements at the surface. In Table 7, the physical adsorption is compared with the chemical adsorption [63]. The volume of organosulfur compound attached to the surface of adsorbent is determined by an isotherm that shows the volume of adsorbed organosulfur compounds in terms of concentration or pressure at a certain temperature. The Langmuir isotherm, which is the most basic model for chemical adsorption, is commonly used to explain this relationship for monolayer adsorption. Langmuir adsorption is responsible based on the following assumptions [63,64]:

• • • • • •

The whole site must be energetically equivalent. The interaction between organosulfur compounds adsorbed on adjacent sites is negligible. There is an equilibrium condition between the adsorption process and the desorption process. Organosulfur compounds are chemically adsorbed at a fixed number of sites. The adsorption occurs in the monolayer sites. Each site must keep one adsorbate molecule.

There are two different mechanisms of sulfur compounds’ adsorption toward ADS materials: selective π-complexation and direct sulfur-adsorbent (S–M) binding mechanisms. Indeed, thiophene has two different lone pairs of electrons on the S element: a pair of electrons lies on a six-electron π system and another electron lies in a plane related to the ring. Thiophene can be donated to those lone pairs of electrons that lie in the plane of the ring to the nanoadsorbent (direct S–M σ bond) or donated by delocalized π electrons of the aromatic ring (π bond) to make π-complexation with nanometal ions [71]. Therefore, thiophene may play either as an n-type donor (direct S–M binding) or as a π-type donor (π-complexation mechanism). To reveal which mechanism is involved Fourier transform infrared spectroscopy (FTIR) must be performed. Based on the π-complexation mechanism and competitive adsorption of aromatic compounds, although the π-complexation nanoadsorbents show low selectivity, they have larger adsorption capacity for sulfur compounds. Moreover, based on the direct S–M interaction mechanism, although the nanoadsorbents show high selectivity, the steric interruption makes it hard to separate sulfur from dimethyl dibenzothiophene (DMDBT), etc. The conflict of these mechanisms can be used correctly [72]. The π-complexation mechanism has been confirmed experimentally and theoretically [73,74]. In Fig. 8, the well-known coordination structures of thiophene with metal in organometallic complexes are represented [75]. The selectivity and adsorption capacity of a nanoadsorbent may be more enhanced by introducing different types of active sites on the nanoadsorbent surface such as functional groups, Lewis acid sites, microstructural defects, electronic defect centers, etc. [72]. Table 7

The adsorption capacities of activated carbon loaded with different metals

Property

Chemical adsorption

Physical adsorption

Enthalpy of adsorption,-AHads Activation energy, Ea Temperature No. of layers adsorbed Trends

40–800 kJ/mol Small Depends on Ea, commonly low One Activated, can be slow or irreversible, involves e-transfer allowing bond formation

8–20 kJ/mol Zero Depends on BP, commonly low One or more Rapid, nonactivated, reversible, van der Waals forces, electrostatic interactions

Desulfurization Materials

955

S M S

S S

M M

M S-3

1S

M

4

M

M

5

M S

S

S

S

M 4, S-2

M 4, S-3

M

M 2



1C

Fig. 8 Well-known coordination structures of thiophene with metal in organometallic complexes. Reproduced with permission from Ma X, Sun L, Song C. A new approach to deep desulfurization of gasoline, diesel fuel and jet fuel by selective adsorption for ultra-clean fuels and for fuel cell applications. Catal Today 2002;77:107–16. Available from: http://doi.org/10.1016/S0920-5861(02)00237-7.

It has been found that large surface area of nanoadsorbents is useful to achieve high sulfur-adsorption capacity by means of chemical or physical adsorption. Based on the Lewis acid–base adsorption mechanism, major thiophene derivative compounds in fuels are Lewis base, which are easily adsorbed at Lewis acid sites. This mechanism indeed is the interaction between the active acid sites available on the nanoadsorbent surface and thiophene-based compounds. Furthermore, sulfur derivative compounds have further affinity to oxidation than hydrocarbons compounds. Hence, redox characteristic of nanoadsorbents may increase oxidization of sulfur derivative compounds into sulfoxides and sulfones. High conversion of sulfides to sulfoxides and sulfones forms stronger polarities, which improves selective of nanoadsorbents to remove sulfur derivative compounds. Therefore, redox characteristics of nanoadsorbents may improve the adsorption capacity and selectivity toward sulfur derivative compounds.

2.29.6

Adsorbent Properties

Nanoadsorbent properties can be classified in three categories: physical properties, chemical properties, or dynamic properties [76]. For the separation of organosulfur compounds, the physical properties include pore size and distribution, surface area, adsorbent shape and size, and also comprise mechanical characteristics including attrition resistance and crush strength. Chemical properties comprise electronic characteristics, nanoadsorbent structure and composition, acid/base characteristics, and electrostatic characteristics. The dynamic properties of nanoadsorbents are based on the behavior during the ADS process like regenerability, selectivity, and capacity. Nanoadsorbent selection and design must aim to maximize the physical, chemical, and mechanical characteristics of the nanomaterials to reach the goal of favorable industrial applications [63,64].

2.29.6.1

Dynamic Properties

An excellent nanoadsorbent must have three important features: high selectivity, high capacity, and desirable regenerability. The dynamic properties of a nanoadsorbent include these mentioned significant factors. The breakthrough capacity of nanoadsorbent explains the volume of fuel that may be desulfurized before any detectable organosulfur is detected at the outlet. The breakthrough capacity and the saturation capacity respectively correspond to the kinetic capacity and the thermodynamic limit of adsorption. The nanoadsorbent capacity is normalized to the mass of nanoadsorbent that shows chemical and physical properties. The nanoadsorbent capacity can be also normalized to the volume of nanoadsorbent as well as surface area [63,64].

2.29.6.2

Chemical Properties

The chemical properties of nanoadsorbent are commonly based on the surface properties at the molecular level. One of the most important parameters for investigating a nanoadsorbent that acts via chemical adsorption is the number of active sites and their density [73]. Further to the impact of surface properties and the number of the active sites on adsorptive capacity, the electronic characteristic of a nanoadsorbent has been recommended to influence the adsorption of organosulfur from fuels for some nanomaterials. The electronic characteristic of a nanomaterial is the bond order, dipole, ionization energy, and atomic partial charge. The ionization energy explains the amount of energy needed to remove an electron that is held within a molecule or atom. Fringuelli et al. [77] studied the aromatic characteristic of thiophene, tellurophene, furan, and selenophene, and revealed that the aromaticity of components are decreased as the order of benzene4thiophene4selenophene4tellurophene4furan. As can be seen, the aromaticity of thiophene is lower than benzene and the ionization energy of benzene is lower than thiophene.

956

Desulfurization Materials

The electric potential (also named the electrostatic potential or the electric field potential) is the amount of energy required to transfer a unit positive charge from a specific location to another specific location under a static electric field without making any acceleration. An electron pair and π-bond in a compound make a high electron density that appeals protons relating to a negative electric field potential. The electric field potential can be recommended as a potential contributing parameter to the mechanism for organosulfur adsorption on activated alumina but not for activated carbon [73]. Moreover, the electric field potential of a species also specifies the polarity of species, whereas the polarity is affected by the electronegativity of elements. Electronegativity is a measure of the affinity of an element to appeal electrons in a covalent bond. Fluorine is the highest electronegative atom and in general the electronegativity of atoms reduces upon moving farther away from fluorine on the periodic table. In a bond between two atoms, a higher difference in electron affinity makes a higher polarity in a molecule. Indeed, an atom attracts the electron pair rather more than the second atom. This indicates that the second atom has higher than its fair share of electron density and therefore becomes a little negative. Furthermore, this partial charge shows the electronegativity of element within the covalent bonding. Hence, an electronegative atom has higher negative partial charges in a dipole. This is explained as a polar bond. The base/acid characteristic of a nanomaterial has significant effect on the selective adsorptive of organosulfur by some adsorptive nanomaterials. For instance, the acidity of the adsorbent has significant impact on the selective adsorption of activated alumina nanoadsorbents [73]. The nickel supported on SiO2–Al2O3 was more adsorptive to indole (neutral) over quinoline (basic), revealing that the base–acid interaction is not significant in the adsorption over Ni/SiO2-Al2O3 [73]. The influence of chemical characteristic varies according to the type of nanoadsorbent used. The functional groups and surface properties are those factors that have significant impact on the adsorption capacity of activated carbon. The capacity of a nanoadsorbent is attributed to its surface characteristic due to the fact that a small part of the surface is covered by organosulfur derivative species [78]. The selectivity of a nanoadsorbent is directly related to the properties of functional groups at the nanoadsorbent’s surface, particularly oxygen functional groups [78]. Therefore, for activated carbon, the selectivity of the nanoadsorbent material cannot be fitted to the electronic interaction and the methyl groups are also attributed for increasing the adsorption capacity by enhancing the electron density of the aromatic molecule [73].

2.29.6.3

Physical Properties

Physically, the tendency of organosulfur to nanoadsorbent is relatively restricted by the volume of accessible surface area that is dependent on the pore structure of the nanomaterial and the surface area of the nanoadsorbent. The mass transfer rate of organosulfur from and to the surface is proportionate to the surface area of the nanoadsorbent, which is faster than the adsorption reaction. Overall, enhancing the adsorption surface area enhances the area accessible for an adsorption reaction, therefore adsorption must be enhanced [79]. By enhancing the porosity of a nanomaterial, the accessible surface area could be enhanced, making some sufficient channels for nanoadsorbate to pass in. For the activated carbon adsorbent, enhancing adsorption surface area enhances the adsorption capacity, particularly for the surface area of the micropores [78]. The distribution of pore size and pore volume of a nanoadsorbent specifies the effective and available surface area in a certain pore size range, that is useful to optimize fuel penetration and available surface area of a nanoadsorbent. Pore size of nanoadsorbents is categorized into three different classes according to the pore size diameters: micropores (o2 nm), mesopores (2–50 nm), and macropores (450 nm). Various nanomaterials have various distributions and pore sizes. For instance, the surface area of the activated carbon is relatively high and most of the surface area is in the range of the micropores, whereas most of the pore volume is in the range of the macropores. The surface area of alumina is relatively low, while most of its pores are in the range of the macropores [78]. Further to the pore size and surface area, the shape and real nanoadsorbent particle size as well as particle size distribution have significant effects on the nanoadsorbent capacity. A nanoadsorbent with more circular structure has smaller surface area. The uniform particle size distribution influences adsorption factors like pressure drop and fuel flow rate, therefore influencing economics and product yield [78]. Larger particle size decreases pressure drop and can obstruct effective contact between fuel and nanoadsorbent, whereas smaller particle size enhances surface area while reduce the pore penetration path [78].

2.29.7

Challenges that Adsorptive Desulfurization Process Is Facing

Although the ADS process is a well-engineered process, there still exist some issues/challenges which research community is still facing:

• • • •

The development of the nanoadsorbents’ selectivity and the adsorption capacity [19,80,81]. The improvement of less energy-intense regeneration approaches [19,80–82]. Developing nanoadsorbents that are not simply blocked during the ADS process [83]. Proposing nanoadsorbents that can do desulfurization without being influenced by the existence of olefins and aromatics [19,80–82,84,85].

2.29.8

Adsorptive Desulfurization Materials

There are numerous adsorbent materials for desulfurization including zeolite, metal-doped zeolites, alumina, activated carbon, silica, Mobil Composition of Matter No. 41 (MCM-41), Santa Barbara Amorphous (SBA-15), aluminosilicates, titania, titania

Desulfurization Materials

High sulfur-adsorption capacity

Mesoporous structure

Redox properties

Big specific surface area

Perfect adsorbent

High selectivity for sulfur compounds

Strong ionic polarity

957

Desulfurization at room temperature

Lewis acid sites

Easy to be regenerated

Fig. 9 An ideal nanoadsorbent for desulfurization process. Reproduced with permission from Shen Y, Xu X, Li P. A novel potential adsorbent for ultra deep desulfurization of jet fuels at room temperature. RSC Adv 2012;2:6155–60. doi:10.1039/C2RA20224G.

supported on metal oxides, zirconia, and so on. The nature of the adsorbent, active sites for sulfur moieties to bind, existence of particular functional groups, porosity (micro-/mesoporous), surface area, etc., have various roles in the whole desulfurization process. These factors offer abundant opportunity for improving novel ADS materials, modification/tailoring of the current nanomaterials with higher capturing capacity. Numerous studies have addressed the requirements for a study to show how more detailed insights into the mechanism of selective ADS materials over various adsorptive materials increase our understanding in the development of these materials. Numerous literature surveys have endeavored for a desulfurization mechanism by providing hydrophobic interaction, van der Waals interaction, π-complexation, H-bonding, and acid–base interaction mechanism. Adsorption technique is relatively less cost intensive and as a possible opportunity for commercial applications, it may be combined with existing commercial techniques of desulfurization to provide a practical approach to improving desulfurization while meeting the current rules and standards. An ideal nanoadsorbent can be featured as represented in Fig. 9.

2.29.8.1

Zeolites

Zeolites are inorganic porous materials having a highly regular structure of pores and chambers that allows some molecules to pass through and causes others to be either excluded or broken down [86,87]. They are any of a large group of minerals consisting of hydrated aluminosilicates, used as cations exchanger and molecular sieves. In general, there are two types of zeolites such as natural and synthetic zeolites. Natural zeolites are cheaper and more abundant but have smaller channels as compared with synthetic zeolites. As a natural mineral resource, natural zeolites are easy to obtain and cause negligible chemical pollutions during the production process [88–92]. Moreover, their inorganic framework keeps them from photodecay. Generally, a decrease in particle size could be expected to lead to a higher efficiency in photocatalysis. This is because the bulk charge recombination of photogenerated electrons and holes, dominant in the well crystallized large semiconductor particles, was reduced by decreasing particle size. Reduction in particle size could also lead to a larger surface area and increased available surface active sites. Synthetic zeolites are also three-dimensional, microporous, crystalline solids with well-defined structure made by crystallization of sodium alumina-silicate gels prepared from pure sodiumaluminate, sodium silicate, and sodium hydroxide solutions. Ion exchange, thermal stability, catalytic properties, and modifications of the surface and pores of zeolites make them attractive candidates for various applications [93–96]. Synthetic zeolites are alumina-silicates and are represented as MO4 (M ¼ Si and Al) with a three-dimensional microporous crystalline structure of interconnected framework of tunnels and cages. The silicon to aluminum ratio in zeolites varies and this directly determines their acidic properties. Also, zeolites contain water molecules alongside different cations, which are mainly from group I and II elements present in the porous structure of zeolites frameworks. As shown in the formula the cations balance with the negative charge of the zeolite lattice [93–96]. MX=nO½ðAlO2 X⋅ðSiO2 ÞÞY Š⋅WH2 O where M is the exchangeable cation with the valence of n, W is the number of water molecules, Y/X is the stoichiometric factor between 1 and 5 depending upon the structure, and (X+Y) is the total number of tetrahedra in the unit cell. Natural zeolites are colorless but impurities such as Fe can give them color. In synthetic zeolites, the presence of alkali or alkaline earth cations impact on the color. Zeolites are porous alumino silicates having a uniform pore structure. They exhibit pore sizes from 0.3 to 1.0 nm and pore volumes from about 0.10 to 0.35 cm3/g. Water moves freely in and out of these pores but the zeolite framework remains rigid and does not swell unlike clays. Another special aspect of this structure is that the pore and channel sizes are nearly uniform, allowing the crystal to act as a molecular sieve at the Angstrom scale. The porous zeolites are host

958

Desulfurization Materials

Thiophene Thiophene 1-hexene 1-hexene Toluene

Gasoline model compounds

Toluene

Competitive adsorption π

C+

S−M

Surface Active metals Complex adsorption

Brönsted acid sites Catalytic reaction

Fig. 10 Competitive adsorption mechanism desulfurization performance over K-doped NiY zeolite. Reproduced with permission from Li H, Han X, Huang H, et al. Competitive adsorption desulfurization performance over K – doped NiY zeolite. J Colloid Interface Sci 2016;483:102–8. Available from: http://dx.doi.org/10.1016/j.jcis.2016.08.024.

to water and other molecules, but only those of appropriate molecular size fit into the pores, creating the molecular sieving property [93–96]. The cations in the zeolites are exchangeable with other cations giving zeolites an ion exchange property. The Si in the structure has a valence of +4 making the SiO4 tetrahedra neutral, while the AlO4 tetrahedra are negatively charged because Al has a valence of +3, creating a Brönsted acid site due to the resulting charge imbalance in the framework structure, which imparts exchangeable sites to the zeolite structure. Therefore the ion exchange capacity of zeolites depends on its chemical composition. Ion exchange capacity is inversely proportional to the Si/Al ratio. The specific ion exchange capacity depends on the structure of the zeolite, the Si/Al ratio and the ions to be exchanged. Zeolites are the most used desulfurization supports [93–96]. The introduction of transition metal ions into numerous zeolite structures has gained considerable attention in the modification of the current zeolites with high selectivity sorption capacity, and regenerability. Li et al. [97] synthesized NiY and KNiY adsorbents in a miniature fixed-bed flow (Fig. 10) by impregnation technique for desulfurization of gasoline containing toluene or 1-hexene. The KNiY exhibited better performance in sulfur removal with 5 vol% aromatics or olefins owing to the high selectivity of incorporated active metals. Indeed, K+ increased dispersion and loading of active Ni because a protonation reaction of 1-hexene and thiophene is happened on Brönsted acid sites of NiY by doping K+ on NiY, resulting in the coverage of active centers and pore blockage. Velu et al. [71] synthesized five metals (including Cu, Ni, Zn, Pd, and Ce) doped nanoadsorbents over ion-exchanged Y zeolites for the ADS fuels. The results revealed that Ce-exchanged Y zeolites showed larger selectivity in comparison with aromatics selectivity. The sulfur derivative compounds are adsorbed on Ce-doped zeolite through direct S-M interaction rather than through π-complexation. It can be understood that some particular nanometal ions with robust ionic polarities are capable of increasing direct S-M interactions and more increase desulfurization. Velu et al. [98] synthesized potassium exchanged NiY zeolites with various Ni contents using incipient wetness impregnation and ion exchange methods. The performance of NiY zeolites, as adsorbents, was evaluated for separating organosulfur derivative compounds from a jet fuel at atmospheric circumstances. The results showed that at the adsorptive temperature of 80°C, NiY zeolite comprised of 30 wt% Ni prepared by impregnation of NH4Y zeolite was capable of cleaning near 10 mL of a fuel per gram of the nanoadsorbent to generate a free-sulfur fuel having less than 1 ppmw sulfur. Moreover, Lee and Valla [99] investigated the sulfur removal from liquid fuels using metal-exchanged mesoporous Y zeolites, which exhibited high sulfur removal performance. Especially, CeSAY zeolites showed the highest sulfur removal performance because of its high selectivity and adsorption capacity.

2.29.8.2

Activated Carbon

Activated carbon (AC) is a good alternative of catalyst for fuel desulfurization processes including several amorphous carbonaceous materials. Activated carbon is extensively applied in various industries such as wastewater treatment, gas pollution control, food production, and chemical processes due to its low cost. It provides a big surface area and a high porosity [100,101]. Yu et al. [102] investigated diesel fuels desulfurization with hydrogen peroxide and studied the adsorption catalytic property of various activated carbons for DBT. The presence of activated carbon promoted the hydrogen peroxide–formic acid system, and the

Desulfurization Materials

959

percent of residual sulfur in oxidized oils was 142 wt. ppm after the oxidation process. In addition, up to 98% sulfur was removed from diesel fuel with 96.5% oil recovery in this work. Meanwhile, the activated carbons could be reused after water washing, which saves the investment and running costs and reduces energy consumption. López et al. [103] examined the SO2 adsorption behaviors on activated carbon at low operating temperatures. The presence of the copper catalyst improved the SO2 adsorption capacity. They found that the demineralized activated carbon (DAC) was better than the activated carbon for adsorption of SO2. Activated carbons impregnated with various metals, such as Cu, Fe, V, Mg, Co, Mn, and Ni, are a way to improve the SO2 absorption performance. Carabineiro et al. [104] claimed that the binary mixtures of Cu, Fe, and V were the best additives on activated carbons for the adsorption of SO2. Gao et al. [105] also studied the SO2 removal performance with different metal-doped samples. They figured out that the good performance of SO2 adsorption was dependent on the metal redox pairs. Fe was also found to be helpful to improve the SO2 adsorption of ACs [106]. Similar results have also been obtained by other researchers. Sumathi et al. [107] developed a model for SO2 capture from the simulated flue gas using palm shell activated carbon (PSAC) impregnated with Ce. It was found that the SO2 adsorption capacities of Ce/PSAC could be well fitted by Langmuir in comparison with Freundlich isotherms. Meanwhile, a sustainable ADS process by the sucrose-derived ACs was extensively introduced, which showed promising applications in industries [108]. Chen et al. [109] studied the adsorption capacities of benzothiophene sulfone (BTO) and dibenzothiophene sulfone (DBTO) by means of unmodified and modified activated carbon. The modified adsorbent was impregnated with nanometal ions including Fe3+, Cu2+, and Ni2+, and the impacts on their adsorption properties was investigated. The experimental data revealed a remarkable enhancement in the adsorption capacity of the modified activated carbon impregnated with nanometal ions as comparison with the unmodified activated carbon. For BTO, the enhancing adsorption capacity was in order of Fe3+/AC4Ni2+/AC4Cu2+/AC, while for DBTO, the enhancing adsorption capacity was in order of Cu2+/AC4Fe3+/ AC4Ni2+/AC. The whole materials have numerous surface functional groups and the properties of the surface functional groups determine the properties of the materials. Table 8 shows the adsorption performance of activated carbons loaded with various metals. Activated carbon fibers (ACFs) have a higher adsorption surface (1300–2000 m2/g), a faster adsorption rate, and a uniform micropore structure in comparison with ACs. They are made by pyrolysis and activation of precursor polymers. Mangun et al. [117] used different ACFs and ammonia-treated ACFs for SO2 adsorption. The results showed that incorporating the basic functional groups into ACFs enhanced the SO2 adsorption performance. It was also proved that ACFs are excellent adsorptive materials for SO2 removal. Daley et al. [118] also studied untreated and heat-treated ACFs prepared using phenolic fiber precursors. They found that the total amount of SO2 adsorbed was relevant to the pore volume and the pore size. Among the different types of ACFs, ACF15 showed the best SO2 adsorption performance. Hong et al. [119] found that ACFs impregnated with various metals improved the DeSOx performance compared to ACFs. In addition, ACFs prepared from raw materials also have better SO2 adsorption rates than powdered activated carbon (PAC) [120–122]. It is also possible to use unburned carbon material from lignite fly ash to remove SO2 from flue gases. Kisiela et al. [123] investigated three fractions of unburned carbon, fabricated during working of the Bełchatów Power Station (PGE GiEK) in Poland. Various characterizations revealed that a large amount of carbonaceous wastes produced from fast lignite oxidation in a boiler showed excellent properties for SO2 adsorption. The chemical properties of unburned carbon material have an important effect on the adsorbate selectivity and on the SO2 adsorption. Zhang et al. [124] also used fly ash from a fluidized bed coal gasifier to synthesize adsorbent for SO2 removal. Table 8

Silica-based materials incorporated with metal ions and metal oxides used for fuel desulfurization

AC impregnated metal oxides AC/Ce AC/Ce AC/Co AC/Cu AC/Fe AC/Fe AC/Fe AC/Mn AC/Mn (MLAC005)b AC/Ni AC/V AC/Ce/Fe a

Surface area (m2/g) 448 525 515 520 432 940.13 1023.6 160.98 912 533 589 432

Total pore volume (cm3/g) 0.670 0.28 0.349 0.31 0.675 0.27 0.5512 0.141 0.286 0.33 0.36 0.644

Not Reported. Manganese oxides-loaded activated carbon (MLAC). Abbreviations: BT, benzothiophene; DBT, dibenzothiophene; TP, Thiophene.

b

Average pore diameter (nm) 6 2.12 2.72 3.404 6.2 2.904 2.25 9.27 4.955 3.222 3.356 6

Adsorption capacity (mg/g)

Ref.

TP

BT

DBT

6.7 –a – – 10 – – 4.5 – – – 33

14.4 – – – 18.4 – – 5.7 – – – 32

73.4 – – – 82.8 – – 11.4 – – – 76.7

[110] [111] [112] [112] [110] [113] [114] [115] [116] [112] [112] [110,115]

960 2.29.8.3

Desulfurization Materials Graphite

Graphite is a stable crystalline form of carbon. Carbon materials with oxygen-containing functional groups and pores size less than 7 Å are significant properties for ADS. The oxygen-containing functional groups enhance acidity because the interaction of sulfur compounds with acidic center may increase adsorptive characteristic of carbon materials. Although oxygen-containing functional group improves the adsorptive characteristic, it reduces the thermal resistance of the carbon materials. For example, activated carbon may be reactivated at 300°C while it is modified with oxygen functional groups is decomposed less than 300°C. Therefore, improvement of a stable carbon material at high temperatures is needed. Carbon material with heteroatom doping is promising since it is commonly stable at high temperatures. Moreover, the adsorptive energy of sulfur derivative compounds could be controlled by means of different dopant atoms [125]. Dai et al. [126] studied the adsorption of SO2 on graphene doped with boron (B), nitrogen (N), aluminum (Al), and sulfur (S) doped graphene using density functional theory (DFT) and found that the adsorptive energy reduces as Al4N4S4B. Shimoyama and Baba [125] studied the adsorption properties of phosphorus- and nitrogen-doped graphites for thiophene removal. They explored the effects of heteroatom doping on the adsorption characteristics of π-conjugated carbon adsorbents.

2.29.8.4

Silica

Silica, also known as silicon dioxide, is an oxide of silicon (i.e., SiO2). There are two kinds of crystalline silica and amorphous silica in nature. Normally, silica is mainly obtained from mining and purification of quartz. It is reported that around 95% of silica produced is consumed in the area of construction industry [5]. In addition, silica materials are also used for gas separation [127]. Silica has been extensively used for sulfur removal, with emphasis on mesoporous silica including silica gel, MCM-41 (a hexagonal array of 1D pores, ~2–4 nm), and SBA-15 (analogous to MCM-41, ~4.5–30 nm). Fig. 11 shows the general route for preparing mesoporous silica in four steps: (1) self-assembling of a surfactant; (2) incorporation of a silica precursor; (3) condensing precursor around the surfactant; and (4) removing the surfactant by calcination or reflux. Silica-based materials incorporated with metal ions and metal oxides used for fuel desulfurization are listed in Table 9. Jabeen et al. [131] used metal doped silica nanoparticles for desulfurization process. To synthesis adsorbent, a solution of each nickel, copper, and zinc metal is gelated with tetraethoxysilane by means of glycerol using the sol–gel method. These nanometals are stabilized on the large surface areas of silica. The synthesized nanomaterials were evaluated in adsorptive separation of benzothiophene and thiophene as organosulfur compounds. Sol–gel synthesis of silica nanometals presents better anchorage of nanometals into silica structure.

1

2

3

4

Fig. 11 General route for preparing mesoporous silica in four steps. Reproduced with permission from Palomino JM. Mesoporous inorganic materials for the desulfurization of jet fuel; 2014.

Desulfurization Materials

Table 9

961

Use of mixed oxides for desulfurization

Framework

Metal

Metal loading (mmol/g)

Surface area (m2/g)

Breakthrough capacity (mgS/g) at 50 ppmw S

Saturation capacity (mgS/g)

Ref.

SBA-15 SBA-15 MCM-41 MCM-41 MCM-41 SBA-15 MCM-41 MCM-41 SBA-15

Pd2+ Cu+ Pd2+ Cu+ Ag+ Ag+ Cu2Oa Cu2Ob Cu2Oa

2.6 5.1 3.1 5.7 2.21 1.77 6.3 6.3 4.7

358 411 502 456 490 408 490 523 400

32.1 19.9 10.9 7.7 15.7e 10.3e 9.9 5.1 5.1

38.5c 25.7c 16.0c 14.4c 32.1d 29.2d 12.8c 10.3c 9.6c

[128] [128] [128] [128] [129] [129] [130] [130] [130]

a

Calcined at 7001C. Calcined at 5501C. c Tested with JP-5 light 841 ppmw S. d Tested on JP-5 1172 ppmw S. e Breakthrough capacity at 10 ppmw S. b

M

M M S

S

Fig. 12 Possible coordination geometries between a nanometal and thiophene.

Palomino et al. [132] represented an optimized method using mesoporous silica nanoparticles (MSN) for desulfurization of JP8 fuel. The bulk of MSN exhibited three times higher capacity of desulfurization. They also compared MSN with bulk MCM-41 and silver-impregnated nanoparticles. Their results showed that MCM-41 and silver-impregnated nanoparticles have saturation capacity of 25.4 and 32.6 mgS/g, respectively. MSN showed the larger capacity to separate 4, 6-DMDBT with a two-time increment in the breakthrough capacity of 0.98 mgS/g at 10 ppmw S. Molecular orbital modeling on thiophene derivative compounds showed that the largest occupied orbital is positioned mainly on the S element. This location offers that direct interaction between S element and nanometal is promising [75]. The two wellknown coordination geometries between a nanometal and thiophene are depicted in Fig. 12. This interaction has a significant effect on desulfurization efficiency. Ligand functionalized silica as an ADS material can be used for both gas and liquid phase fuels as well as for ODS. Song et al. [133] functionalized MCM-41 with aminopropyl groups and incorporated Cu2+ for sulfur removal. The results indicated that a metal ion anchored by ligand group is more effective than MCM-41 comprised of Cu2+ at desulfurization process. The ligand functionalized silica separated nearly 1.8 mgS/mmol Cu while the ligand-free silica separated nearly 0.7. This increment is related to the ligand’s capability for better dispersion of the adsorption sites. The ligand-free silica sample has copper in oxide form while ligand functionalized silica has copper in ion form, which can also account for the difference in efficiency. Therefore, adsorptive materials containing Cu2+ have higher adsorption capacity than materials containing CuO in their structure. The general silica functionalization and metal incorporating for organosulfur adsorption are shown in Fig. 13. Shen et al. [72] designed and prepared a nanoadsorbent, Ni–Ce/Al2O3–SiO2, for ADS of real jet fuel at ambient temperature and atmospheric pressure. The structure of the Al2O3–SiO2 based on Tanabe’s assumption is depicted in Fig. 14. Based on Tanabe’s assumption, for each bond the charge difference is 0.5, and the valence unit of 23 for all the bonds is additional. The SiO2–Al2O3 substrate has large surface area, and its binary oxide shows the Lewis base or the Brønsted acidity, which reveals a similar acid nature to repulse adsorbing thiophene organosulfur compound. Therefore, the large specific surface area of SiO2–Al2O3 is not capable of making large organosulfur adsorption capacity at low temperatures. Shen et al. [72] synthesized Ni–Ce/Al2O3–SiO2 adsorbents by various methods to determine desulfurization efficiency of real jet-A fuels at ambient temperature and atmospheric pressure (Table 10).

2.29.8.5

Calcium Oxide

Calcium oxide (CaO), generally recognized as burnt lime or quicklime, is a widely used material for desulfurization. The dry method of desulfurization with dolomite or lime involves in the solid–gas phase heterogeneous reaction that happens at high

962

Desulfurization Materials

H2N

NH2

O Si O

Si O

SiO2

O

O

SiO2

Mx+

S Mx+

Mx+

HN

HN

S

Si O

O

Si

O

O

SiO2

O

O

SiO2

Fig. 13 General silica functionalization and metal incorporating for organosulfur adsorption. Reproduced with permission from Palomino JM. Mesoporous inorganic materials for the desulfurization of jet fuel; 2014.

O O

O

O

A1

O

Si

O

O O

O

Fig. 14 The structure of Al2O3–SiO2 based on Tanabe’s hypothesis. Reproduced with permission from Shen Y, Xu X, Li P. A novel potential adsorbent for ultra deep desulfurization of jet fuels at room temperature. RSC Adv 2012;2:6155–60. doi:10.1039/C2RA20224G.

temperatures. The process includes the below reactions [134]: At first, lime is calcinated as follows; this takes place at temperatures higher than 850°C: CaCO3 →CaO þ CO2

ð18Þ

SO2 then reacts at the surface of the CaO at temperatures higher than 600°C as follows: SO2 þ CaO→CaSO3

ð19Þ

Afterward, CaSO3 is oxidized in a fast reaction as follows: 4CaSO3 →3CaSO4 þ CaS

ð20Þ

Finally, calcium sulfide is also oxidized to calcium sulfate as follows: CaS þ 2O2 →CaSO4

ð21Þ

Mura et al. [135] studied the kinetics of the reaction between CaO and SO2 in dry desulfurization process using a differential, fixed-bed reactor. They investigated the effects of particle size (0.65–2.0 mm), temperature (500–950°C), and exposure time (up to 20 min) on the degree of sulfation of CaO. Yang and Chen [136] and Hartman and Trnka [137] also studied the reaction occurring between CaO and carbonyl sulfide (COS) using CaO for the following reactions: 2COS→2CO þ S2

ð22Þ

CaO þ COS→CaS þ CO2

ð23Þ

Numerous works on SO2 removal using CaO have been carried out [138–141]. Kevin et al. [94] studied the sulfation rate of CaO particles experimentally. CaO derived from Ca(OH)2 showed a higher performance than that derived from CaCO3. That was

Desulfurization Materials

Table 10

963

MOFs used in adsorptive desulfurization (ADS) Concentration after adsorption (S-mg kg−1)

Desulfurization efficiency (%)

Sample no.

Preparation method

Sintering condition

Original concentration (S-mg kg−1)

1

Extrusion

h, under

949.03

62.26

93.47

2

Sol–gel

h, under

949.03

79.76

91.60

3

Wet impregnation

h, under

949.03

734.83

22.57

4

Extrusion

h, under

949.03

84.02

91.15

5

Sol–gel

h, under

949.03

196.98

79.24

6

Wet impregnation

h, under

949.03

648.96

31.62

7 8 9 10 11 12 13

Extrusion Sol–gel Wet impregnation Extrusion Sol–gel Wet impregnation Extrusion

949.03 949.03 949.03 949.03 949.03 949.03 949.03

593.74 264.25 709.93 590.85 219.97 739.62 33.88

37.44 72.16 25.19 37.74 76.82 22.07 96.43

14

Extrusion

h, under

949.03

34.64

96.35

15

Extrusion

600°C×3 helium 600°C×3 helium 600°C×3 helium 800°C×3 helium 800°C×3 helium 800°C×3 helium 600°C×3 600°C×3 600°C×3 800°C×3 800°C×3 800°C×3 650°C×3 helium 700°C×3 helium 750°C×3 helium

h, under

949.03

41.09

95.67

h, h, h, h, h, h, h,

under under under under under under under

air air air air air air

why Ca(OH)2 was widely used in industries for SO2 removal. Siriwardane and Cook [140] investigated the interactions between SO2 and CaO at different temperatures. Clean CaO and sodium deposited CaO were used as the absorbing materials. As the added amount of sodium increased, the SO2 adsorption performance significantly improved. In addition, regarding the modeling of the SO2 adsorption with CaO, a comprehensive model for SO2 reacting with Ca(OH)2 or CaCO3 at high temperatures was proposed by Mahuli et al. [142]. On the basis of the grain–subgrain method, the calcination, sintering, and sulfation reactions, and the interactive effects among them were considered in the simulations. Wang et al. [143] also developed a model for SO2 capture by CaO particles inside a circulating fluidized bed (CFB) combustor. They found that increasing the cluster porosity of the CaO particles and the feed gas flow rate could improve the SO2 removal.

2.29.8.6

Copper Oxide

Copper (II) oxide (also known as cupric oxide) is a black and stable oxide of copper (i.e., CuO). Turbevillle et al. [144] developed a copper-containing catalyst for the ADS from an olefin feed stream that exhibited high sulfur removal performance.

2.29.8.7

Magnesia

Magnesia is also well known as magnesium oxide (MgO). MgO reacting with SO2 in a spray dryer was reported by Egan and Felker [145]. Magnesium sulfite was formed in the following reactions. MgO þ H2 O→MgðOHÞ2

ð24Þ

MgðOHÞ2 þ SO2 →H2 O þ MgSO3

ð25Þ

Ziolek et al. [146] studied SO2 adsorption on the surface properties of various metal oxides. The results indicated that the strength of these acid sites depended on the metal oxide being weak on MgO. Freitag et al. [147] also measured the SO2 adsorption isotherms on MgO via in situ TPD and XANES at temperatures from 170 to 205K. Moreover, Stark et al. [148] used the nanoscale MgO particles for SO2 adsorption and compared their surface adsorptive properties with microscale MgO particles. The experiment results showed that the nanoparticles adsorbed more SO2 due to their high surface areas and surface reactivity. In terms of the theoretical studies of SO2 adsorption on MgO, Eid and Ammar [149] conducted a DFT study on SO2 adsorption performance on a Li atom deposited MgO. The adsorption energy (Eads) of SO2 in different positions on both of O−2 and Fs sites was considered. It was also found that SO2 was strongly adsorbed on the substrate surfaces of MgO containing Fs-center. Francesco et al. [150] simulated the core excitation spectra of model systems for SO2 adsorption on the MgO (100) surface using the timedependent density functional theory (TDDFT) method. The cluster models were used for illustrating the various interaction modes

964

Desulfurization Materials

of SO2 with the acidic and basic sites of the surface. This approach provided a useful tool for investigating the origins of the spectral features. In addition, Pacchioni et al. [151] investigated numerically the adsorption performance of SO2 at the step sites of the MgO (100) surface. They claimed that the sulfite could form by the interaction of the sulfur atom in SO2 with two surface fivecoordinated O2− anions.

2.29.8.8

Manganese Oxide

Manganese (II) oxide (MnO) is an inorganic compound and nonstoichiometric. Adeyi and Abekanmi [152] compared the desulfurization performance of crude oil using two different metal oxides including activated zinc oxide (AZ) and activated manganese oxide (AM). They analyzed the kinetics between these materials and sulfur. MnO showed a better reactivity than ZnO.

2.29.8.9

Titania

Titania, also known as titanium dioxide, is a naturally occurring oxide of titanium (i.e., TiO2). Saleh et al. [153] compared the desulfurization activity of carbon nanotube (CNT), titania, and mechanical mixture of titania and CNTs (MM). The prepared CNT/ TiO2 nanomaterials (Fig. 15) showed better desulfurization performance for model fuel oil than other materials. It was indicated that the CNT/TiO2 materials are a promising adsorbent in petroleum engineering. Guo et al. [154] performed a DFT study to better understand the mechanisms and interactions on adsorption of thiophene derivative compounds on the anatase (0 0 1) surface of TiO2-based materials for sulfur removal from liquid hydrocarbon fuels (Fig. 16). The study showed that on complete anatase, the robust interactions happen between the titanium cation and sulfur atom. Generally, thiophene has the biggest affinity toward the oxygen-rich environment. TiO2 as an adsorbent is commonly made from either mixed metal titanium oxide or titania supported on other metal oxides for desulfurization. Hussain et al. [155] synthesized different metal oxide supports and used titanium (IV) chloride, titanium (IV) isopropoxide, and titanyl oxide sulfate as titania precursors. Titanium (IV) isopropoxide showed the excellent performance, owing to easy hydrolysis of this precursor, therefore, providing the amount of titania distributed on the surface. The results also showed that Ti:Al ¼1:4.4 was the optimum ratio. Ag content experiments showed that the optimal content is between 8 and 12 wt%. In this optimum condition, adsorption sites are balanced without beginning to block pores and decreasing the available surface area of the support.

2.29.8.10

Zirconia

Zirconia, also known as zirconium dioxide, is a crystalline oxide of zirconium (i.e., ZrO2). It has good adsorptive capability of thiophene from n-octane and n-heptane [156,157]. An example of sulfated ZrO2 was also proved to be a good catalyst for the reactions by Wang et al. [157]. 100% desulfurization efficiency was obtained in this study. Kumar et al. [158] synthesized different zirconia-based adsorbents including normal dried zirconia, calcined zirconia, sulfated and calcined zirconia for desulfurization of model oil. Adsorbent dose, contact time and temperature were optimized for the adsorption efficiency. The adsorption kinetics and different kinetic models were studied. Thermodynamics of desulfurization process were also explored and enthalpy, entropy, and Gibbs free energy were determined.

Titania

(1) Oxidation of MWCNT (2) Dispersion of o-MWCNT (3) Reflux the precursor of titania with MWCNT (4) Calcination

MWCNT

MWCNT/TiO2

Fig. 15 Schematic description of preparing the multiwalled carbon nanotubes (MWCNTs)–titania nanomaterial Reproduced from Saleh TA, Siddiqui MN, Al-Arfaj AA. Synthesis of multiwalled carbon nanotubes-titania nanomaterial for desulfurization of model fuel. J Nanomater 2014;2014:194. Available from: http://dx.doi.org/10.1155/2014/940639.

Desulfurization Materials

O1

1.65

1.65

965

O2

1.60 O3

Fig. 16 Adsorption conformation of thiophene on O-rich (dis) anatase (0 0 1) surface. Reproduced with permission from Guo J, Watanabe S, Janik MJ, Ma X, Song C. Density functional theory study on adsorption of thiophene on TiO2 anatase (0 0 1) surfaces. Catal Today 2010;149:218– 223. Available from: http://dx.doi.org/10.1016/j.cattod.2009.05.002.

2.29.8.11

Ceria

Ceria, also known as cerium (IV) oxide or cerium dioxide, is an oxide of the metal cerium (i.e., CeO2). Kylhammar et al. [159] studied SO2 removal from the lean exhausts using the regenerable ceria-based SOx traps. It was indicated that the presence of Pt increased the SOx storage capacity at 250°C for the CeO2-based sample. Meanwhile, in comparison with fresh samples, the samples preexposed to high amount of SO2 had a lower SOx storage capacity. SO2 adsorption on CeO2 solid using Raman spectroscopy, electron paramagnetic resonance (EPR), thermal analysis approaches have also been investigated by Flouty et al. [160]. Recently, Tumuluri et al. [161] comprehensively reviewed the applications of CeO2 in the SO2 removal processes. The interactions of SO2 with pure CeO2 and CeO2-based catalysts were introduced in detail. It was also found that the sulfate/sulfite species on these materials were stable above 350°C. Regarding the investigations of SO2 removal performance by the CeO2-based catalysts, Flytzani-Stephanopoulos et al. [162] carried out the experiments and found that the Cu– or Ni–CeO2 system improved the SO2 reduction performance. The cocapture SO2 and NO were also studied in this work.

2.29.8.12

Mixed Oxides

Mixed oxides are an alternative material for SO2 removal, which could improve the SO2 adsorption performance compared to the single oxides [163]. Table 11 summarizes the use of mixed oxides for desulfurization. The surface area and the removal capacity of SO2 of different mixed oxides were compared. Zhang et al. [163] performed reactive ADS over metal oxide nanoadsorbents to separate the organosulfur compound of actual FCC gasoline in a batch reactor in the presence of hydrogen at low pressures. Their findings revealed that high adsorption temperature allows for a profound enhancement in the ability of the nanoadsorbents in reactive ADS. With the increase of the initial composition of thiophene in the fuel, the adsorption rate reduced whereas the adsorption capacity of the nanoadsorbents enhanced and the sulfur-adsorption capacity and the highest organosulfur removal rate were 13.72 mg/g and 86.33%, respectively.

2.29.8.13

Metal Organic Frameworks

Metal organic frameworks (MOFs) are a new class of materials with high adsorption capacity. A MOF is built by a metal site attached to an organic ligand (as bridging ligands between the metal ions) therefore creating a 3D structure [172]. MOFs are attractive adsorbents for industrial usage particularly for desulfurization. MOFs have excellent characteristics such as high BET surface area, large pore volume, high porosity, and high inner surface area [173,174]. Table 12 reports the MOFs that have been used for desulfurization and their maximum uptake related to thiophene (TP); benzothiophene (BT), dibenzothiophene (DBT); 4, 6-dimethyl dibenzo thiophene (4, 6-DMDBT) are summarized. One of the most used MOFs in fuel desulfurization is HKUST-1 with formula C18H6Cu3O12, also known as Basolite C300 CuMOF [188]. Tan et al. [189] synthesized HKUST-1/Fe3O4 adsorbents for desulfurization by a dry gel conversion (DGC) method. In

Desulfurization Materials

966

Table 11 Ni–Ce/Al2O3–SiO2 adsorbents prepared by various method to determine desulfurization efficiency of real jet-A fuels at ambient temperature and atmospheric pressure Cads (mg/g)

Nature

a

CuxAl MgAl MgFe NiAl ZnAl Ce/MgAl 5% CuNiAl MgAlFe MgCuAl MgFeAl Cu/MgAlFe CuMgAlFe 10% MnMgAlFe

b

– 1150 1010 – 670 715 – – 340 1070 1600 1620 –

SBET (m2/g)

93–108 204 75 143 48 153 155 118 183 169 154 196 112

T (°C)

700 700 700 500 700 700 500 800 700 700 700 700 800

Gas composition (vol%)

Ref.

SO2

O2

Balanced gas

0.01 1 1 1 1 0.15 1 0.3 1 1 2 2 0.3

5 0 0 0 0 1.5 0 5.05 0 0 8 8 5.05

Ar Air Air N2 Air He N2 He Air Air N2 N2 He

[164] [165] [165] [166] [165] [167] [168] [169] [96] [165] [170] [171] [169]

a

X¼ 2–5. NOT reported.Source: Reproduced with permission from Shen Y, Xu X, Li P. A novel potential adsorbent for ultra deep desulfurization of jet fuels at room temperature. RSC Adv 2012;2:6155–60. doi:10.1039/C2RA20224G. b

this method, the vapor of N, N-dimethyl formamide (DMF) penetrates and links with the parent nanomaterials, allowing to the growth of HKUST-1 in an exclusive atmosphere. Therefore, HKUST-1 is combined with Fe3O4. A small of MOFs may grow on the Fe3O4 surface via the solvothermal technique. Fig. 17 shows schematically the preparation method of (A) HKUST-1 by DGC, and HKUST-1/Fe3O4 by (B) DGC and (C) solvothermal technique. Shi et al. [190] investigated the sulfur removal performance using various modified MOFs. It was found that the highest H2S and dimethyl sulfide (CH3SCH3) removal efficiency were 8.46% and 8.53% when incorporating MOFs with 2% AC, respectively. The adsorption mechanisms for MOF-based desulfurization are coordination bond formation, acid–base interactions, hydrogen-bonding, π-complexation, and van der Waals forces. A schematic of these mechanisms is shown in Fig. 18. MOFs can play as Lewis acids due to coordinatively unsaturated sites that can admit a pair of electrons. Based on the Pearson concept, bases are categorized into two groups: polarizable and nonpolarizable, which are known as “soft” and “hard” bases, respectively. Acids are also categorized according to their preferential interactions with soft or hard bases. Acids that make robust interactions with soft and hard bases are recognized as soft and hard acids, respectively. Therefore, sulfur-containing bases are relatively intermediate to soft; hence, they have tendency to attract soft Lewis acids like Zn2+, Cu2+, and Co2+ [178]. Numerous functional nanomaterials (such as ethylene diamine (ED) and amino methane sulfonic acid (AMSA) [186]) may be grafted to the Lewis acid of the MOFs, which provides appropriate sites for several compounds. Besides the functionalization of central nanometal sites, MOFs can be acidic via appropriate modification. Heteropolyacids (HPAs) such as phosphotungstic acid (PWA) can play as acidic functional moieties in MOF structure without leaching out throughout the desulfurization process. After impregnation, Lewis acidic nanometal salts have been revealed to be useful for developing the desulfurization process. One major challenge faced in grafting or impregnation is the decreased porosity that imposes negative impacts on the adsorptive capacity of the MOFs [191].

2.29.8.14

Nickel-Based Adsorbents

One of the most favorable classes of nanomaterials for organosulfur desulfurization is nickel-based metal nanoadsorbents, due to their high selectivity and capacity. The nickel-based nanoadsorbents have been investigated as better for separating organosulfur derivative compounds from fuels/flue gas/hydrocarbon gas streams [192–194]. By comparing the capacity of various nanoadsorbents, it can be found that nickel-based nanoadsorbents are better than activated carbon, copper exchanged zeolites, and activated alumina [81]. Dasgupta et al. [192] investigated experimentally the desulfurization process of diesel using the regenerable nickel-based adsorbents. At the conditions of 4 bar pressure and 350°C, SO2 concentration could be decreased to 50 ppm from 450 ppm by the NiMCM. As the inlet SO2 concentration reduced to 150 ppm, the NiY showed better performance than the NiMCM. A temperature programmed oxidation (TPO) study was performed to determine the optimal regeneration conditions.

2.29.8.15

Adsorptive Polymers

Various classes of adsorptive polymers exist that have been investigated for the purpose of the ADS including microporous coordination polymers (MCPs), and imidation agents. Shiraishi et al. [195] was firstly used an imidation agent (chloramine T, sodium N-chloro-p-toluene sulfonamide) for the desulfurization of both industrial and model fuel in the existence of acetic acid

Desulfurization Materials

Table 12

The fundamental on which clay minerals can be categorized

MOF

HKUST-1

UMCM-150 MOF-505 MOF-5 MOF-177 CPO-27 (Ni)

CPO-27 (Co) Cu3 (NAPANA) rho-ZMOF ZIF-8 ZIF-76 MIL-53 (Al) (A100) MIL-53 (Al)

MIL-53 (Al)/CuCl2 MIL-53 (Cr) MIL-53 (Cr)/CuCl2 MIL-53 (Fe) hydrated MIL-53 (Fe) dehydrated MIL-47 (V)

MIL-47 (V)/CuCl2 MIL-100 (Fe)

MIL-100 (Al)

MIL-100 (V)a MIL-100 (Cr)

MIL-101 (Cr)

967

Solvent

Isooctane Isooctane/toluene (85/15) n-Octane/toluene (95/5) 2, 2, 4-Trimethyl pentane Heptane/toluene (80/20) Heptane/toluene (20/80) Isooctane Isooctane/toluene (85/15) Isooctane Isooctane/toluene (85/15) Isooctane Isooctane/toluene (85/15) Isooctane n-Octane/toluene (95/5) Heptane/toluene (20/80) Heptane/toluene (80/20) Heptane/toluene (20/80) Isooctane n-Octane/toluene (95/5) n-Octane/toluene (95/5) n-Octane/toluene (95/5) 2, 2, 4-Trimethyl pentane Heptane/toluene (80/20) Heptane/toluene (20/80) n-Octane n-Octane/toluene (75/25) n-Octane n-Octane/toluene (75/25) n-Octane n-Octane/toluene (75/25) n-Octane n-Octane/toluene (75/25) Heptane Isopropanol Heptane Isopropanol Heptane/toluene (80/20) Heptane/toluene (20/80) n-Octane n-Octane/toluene (75/25) n-Octane n-Octane/toluene (75/25) 2, 2, 4-Trimethylpentane Heptane Heptane/toluene (80/20) Heptane/toluene (20/80) Heptane Heptane/toluene (80/20) Heptane/toluene (20/80) Heptane Heptane Heptane/toluene (80/20) Heptane/toluene (20/80) Heptane/toluene (80/20) Heptane/toluene (20/80) Gas oil Isooctane n-octane/p-xylene (25/75)

Maximum uptake (wt% S)

Ref.

TP

BT

DBT

4, 6-DMDBT

– – 8.0 – 3.8 1.9 – – – – – – – 9.6 4.6 11.4 4.9 – 1.3 1.0 1.9 – o0.4 o0.4 – – – – – – – – – – – – o0.4 o0.4 – – – – – 3.2 1.5 0.4 3.4 0.8 o0.4 1.9 2.7 0.8 o0.4 1.2 0.4 – – –

2.5 – – 4.0 2.4 0.5 4.0 – 5.1 – 1.6 – 0.8 – 4.5 12.0 4.8 2.8 – – – – – – 0.9 0.5 0.8 0.4 2.4 0.9 1.9 0.8 3.6 11.2 5.3 7.4 – – 5.7 1.2 7.4 1.4 – – 1.4 o0.2 – 1.4 0.2 – – 1.7 o0.2 1.2 o0.2 – – 0.7

4.5 1.0 – 4.5 2.1 0.4 8.3 2.5 3.9 3.3 3.3 1.0 1.8 – 3.5 6.9 3.5 3.3 – – – 3.8 o0.2 o0.2 – – – – – – – – – – – – o0.2 o0.2 – – – – 2.5 – 1.2 o0.2 – 1.0 o0.2 – – 1.2 0.2 1.4 0.2 0.5 1.7 –

1.6 0.8 – 1.4 – – 4.1 0.7 2.7 0.9 1.4 0.7 0.8 – – 3.6 – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – 1.1 –

[175] [55] [176] [177] [178] [178] [175] [55] [175] [55] [175] [55] [175] [176] [178] [179] [178] [180] [176] [176] [176] [177] [178] [178] [181] [181] [181] [181] [181] [181] [181] [181] [182] [182] [182] [182] [178] [178] [181] [181] [181] [181] [177] [183] [178] [178] [183] [178] [178] [183] [183] [178] [178] [178] [178] [184] [184] [185] (Continued )

968

Table 12

Desulfurization Materials

Continued

MOF

MIL-101 MIL-101 MIL-101 MIL-101

Solvent

(Cr)/1% phosphotungstic acid (Cr)/ethylene diamine (Cr)/amino methane sulfonic acid (Cr)/0.25% graphene oxide

n-octane/p-xylene n-octane/p-xylene n-octane/p-xylene n-octane/p-xylene

Maximum uptake (wt% S)

(25/75) (25/75) (25/75) (25/75)

Ref.

TP

BT

DBT

4, 6-DMDBT

– – – –

0.6 0.6 0.7 0.7

– – – –

– – – –

[185] [186] [186] [187]

a

MIL-100 (V) pretreated under vacuum at 423 K for 12 h. Abbreviations: BT, benzothiophene; DBT, dibenzothiophene; 4, 6-DMDBT, 4, 6-dimethyl dibenzo thiophene; TP, thiophene. Source: Reproduced with permission from Church M, Coniglio M, Hardie L, Longstaffe F, Middleton V. Encyclopedia of sediments and sedimentary rocks, Berlin Springer, 2005.

(A)

Grind

Wash HKUST-1

Fe3O4

Cu2+

+

(B)

Grind

Wash HKUST-1/Fe3O4

Solvothermal method

BTC (C)

HKUST-1 and Fe3O4 Fig. 17 Preparation method of (A) HKUST-1 by dry gel conversion (DGC), and HKUST-1/Fe3O4 by (B) DGC and (C) solvothermal technique. Reproduced with permission from Tan P, Xie X-Y, Liu X-Q, et al. Fabrication of magnetically responsive HKUST-1/Fe3O4 composites by dry gel conversion for deep desulfurization and denitrogenation. J Hazard Mater 2017;321:344–52. Available from: http://doi.org/10.1016/j. jhazmat.2016.09.026.

and methanol. The organosulfur species in model fuel were changed into N-tosylsulfimides through imidation with chloramine T and these species were polar and therefore they can dissolve in methanol leading them to be simply separated from the fuel. This phenomenon was not the case for the industrial fuel because the alkyl substituted DBT existing in industrial fuel is hydrophobic and remains in the fuel. Hence, fuel desulfurization was performed using silica or alumina to remove more of the organosulfur species from the fuel. As adsorption and imidation processes can be combined, Shiraishi et al. [196] recommended that to make this process energy efficient, the adsorption of sulfimides must be prevented. This new approach was found to be feasible and chloramine T reduced the organosulfur loading of fuel from 1900 ppm to 400 ppm. In order to prevent subsequent ADS of organosulfur species, Shiraishi et al. [197] studied the use of a polymer-supported imidation agent (PI, sodium N-chloro-polystyrene sulfonamide). Once a solution of n-tetradecane comprising dibenzo thiophene (DBT) was mixed with MEoH in the existence of acetic acid and PI at 50°C, DBT sticks to the PI, through the imidation of sulfur element on DBT by means of PI, and separated easily from the fuel. The PI imidation agent is insoluble within the fuel mixture and the yielded sulfimides were easily anchored on PI and therefore they can be successfully separated by microfiltration. A schematic of PI and chloramine T is depicted in Fig. 19. By comparing the activity of the chloramine T and PI, it can be realized that PI performs much better. Moreover, the organosulfur loading in industrial light oil reduced from 400 to 54 ppm for 40 h reaction time [198]. Furthermore, Fadhel [199] studied the organosulfur separation efficiency of both chloramine T and PI and investigated the impact of initial organosulfur concentration, sorbent dose, and reaction time. The results showed that the separation efficiency enhances with reducing organosulfur concentrations as well as enhancing the sorbent dose. For industrial hydrocracked diesel, organosulfur loading was reduced from 1900 ppm to 180 ppm. This desulfurization method was not applicable for a fuel containing 12,354 ppm organosulfur loading. MCPs are another type of polymers that have been found to be attractive in the ADS process. Kitagawa et al. [200] showed that the adsorption capacity of MCP is higher than that of NaY zeolites and activated carbon during the fuel desulfurization process.

Desulfurization Materials

I. Acid-base interaction MOFs

SO3 H

MOFs

N:

969

II. II-complexation

−B

H A

Cu(I) or Ag(I)

H

MOFs III. Coordination bond IV. Hydrogen bond MOFs

M

:ligand MOFs

OH

CUS or OMS

N-H

H O

MOFs

H-N-

V. Van der Waals interaction δ− S/NCCs MOFs δ+ δ− δ+ S/NCCs S/NCCs

MOFs

Fig. 18 Schematic of various adsorption mechanisms applied for the desulfurization process. MOFs, metal organic framework. Reproduced with permission from Ahmed I, Jhung SH. Adsorptive desulfurization and denitrogenation using metal-organic frameworks. J Hazard Mater 2016;301:259–76. Available from: http://doi.org/10.1016/j.jhazmat.2015.08.045.

O

S

O

O

S

O

Cl

N

Na

Cl

N

Na

Fig. 19 A schematic of PI (right) and chloramine T (left).

Cychosz et al. [175] studied various types of MCPs with difference in shapes, metal clusters, and pore sizes. The results showed that MCPs have high affinity toward DMDBT than BT and DBT at 300 ppmw S in isooctane solution. Their findings were motivating because the DMDBT species were hard to separate from fuel in comparison with DBT and BT. Cychosz et al. [55] also studied fuel desulfurization using packed-bed breakthrough experiments. From screening the results, it can be realized that MCPs were efficient adsorbents and may be regenerated for subsequent use by using solvent and heat. There is a novel approach that has appeared as the molecular imprinting method and it leads to the making of tailor-made binding sides for certain molecules [201]. The resulting molecular imprinted polymers (MIPs) are employed in an extensive range of applications including in membrane fabrication, biosensors, etc. [202]. A MIP is synthesized through polymerization in which a solution of an objective molecule (template), functional monomers that interact with the functional groups of the objective molecule, and an additional amount of cross-linkers are polymerized. Finally, the template is separated from the cross-linked polymer network [203]. MIPs have been extensively utilized in ADS process due to the easy synthesis approaches, as well as their low cost and stability. The conventional MIPs have smaller adsorption capacity. To enhance the binding rate and the adsorption capacity of the conventional MIPs, thin polymers are grafted onto different substrate like silica-based nanomaterials, polymer beads, and carbon nanotubes [203].

2.29.8.16

Sludge Derived Adsorbents

The most available industrial adsorbents for SO2 removal are expensive or have other related issues; consideration has been paid to employing different sludge derived materials as adsorbents [204]. Sludge can be obtained from wastewater treatment plants in the petrochemical, refinery, oil, and gas industries. Sludge can undergo a pyrolysis reaction to yield a mesoporous material with high active surface area with chemistry that can facilitate the H2S oxidation to sulfur element [205,206]. The proposed mechanisms for H2S removal presented by Yan et al. [205] can be used to explain sludge derived adsorbents. Sludge derived adsorbents have

970

Desulfurization Materials

complex behavior, and they have various reactive components that can recommend an alternative to employing nonimpregnated activated carbon materials. The adsorption efficiency of sludge derived adsorbents for desulfurization is similar to iron based adsorbents, while less efficient than impregnated activated carbon adsorbents [207]. A major concern with utilizing sludge as adsorbents is that it may have species that adversely influence the desulfurization process. Some species are derived from metal sludge that is formed during industrial applications. Yuan and Bandosz [205] studied a mixed of sewage sludge and metal sludge with different weight ratios obtained from an industrial galvanizing process to investigate H2S removal from simulated digester gas mixture as biogas. They pyrolyzed the mixed sewage sludge and metal sludge at different temperatures. The results indicated that the adsorption capacity of the mixed sewage sludge and metal sludge in the ADS process is comparable to the adsorption capacity of impregnated activated carbon materials. Moreover, their findings showed that the adsorption capacity depends on the pyrolysis temperature and the overall sludge composition. Sewage sludge adsorbents pyrolyzed at high temperatures (800 and 950°C) with high level of sludge exhibited excellent adsorption efficiency. The maximum adsorption capacity of the mixed sewage sludge and metal sludge was found to be less than 21 mg H2S/g sludge, which is less than the adsorption capacity of unimpregnated activated carbon materials. Bagreev et al. [208] also studied sewage sludge derived adsorbents for H2S removal from moist air. They found that the adsorption capacity of the sewage sludge derived adsorbents enhances with enhancing carbonization temperature. With carbonization, a mineral-like phase is created that comprises catalytically active metals such as copper, iron, and zinc. They also found that the existence of iron oxide considerably enhances the adsorption capacity of industrial activated alumina and carbon materials. Their results also indicated that the sewage sludge derived materials are effective for H2S capture until the entrance pores of adsorbents are not closed with organosulfur compounds. When the catalytic impact is predominant for adsorbents, H2S is removed until all entrance pores of adsorbents are blocked with organosulfur compounds. In these types of sewage sludge derived adsorbents, the chemical adsorption rather than physical adsorption acts with a remarkable role in H2S removal from moist air. Ros et al. [209] investigated H2S separation using sewage sludge derived adsorbents that are recognized by their promising textural characteristics when compared to other adsorbents fabricated by activation and/or pyrolysis of similar precursors. They used alkaline hydroxide as activation precursor to fabricate catalysts/adsorbents with a wide range of porosities. Besides the large mesoporous adsorbents, a basic pH and high metallic level of adsorbents are needed to attain excellent desulfurization performance. Bandos and Block [210] synthesized composite adsorbents by pyrolysis of sewage sludge and waste oil sludge at various pyrolysis time and temperature. The resulting adsorbents were examined in H2S reactive adsorption through a dynamic experiment. The synthesized mesopore materials showed high adsorption capacity for H2S separation and high selectivity for H2S conversion to sulfur as element. Various characterizations indicated that elemental sulfur is trapped in the mesopores of the adsorbent. In some samples, different pores are shaped after the oxidation process within the deposited sulfur. Wallace et al. [211] prepared fish/sludge waste mixtures and then pyrolyzed at inert environment by means of CO2 as activation precursor. The adsorbents were finally employed for H2S removal at atmospheric conditions. An increment in the adsorption capacity of H2S using the composite adsorbents was found to be attributed to the synergistic impacts between fish and sludge waste. The surface reactivity of samples was found to be attributed to the existence of Ca, K, Na, Fe, and Mg in different shapes. The results confirmed that the activation has a negative impact on the adsorption properties of the composite for H2S removal. The species of sludge waste forms an active iron phase that is significant for the H2S oxidation process. The highest synergistic impact was related to the sample with the lowest level of the fish species. The synergistic impact decreases with increasing in the level of the fish waste due to the reactions of silica with alkaline metals during pyrolysis. Ansari et al. [212] mixed dewatered sludge with polystyrene sulfonic acid-co-maleic acid sodium salt at variety of sludge to polymer ratios to enhance the carbonaceous phase level in sewage sludge derived adsorbents. They carbonized the materials at high temperature in the presence of the virgin precursors and then washed in water to separate the additional hydroxide/sodium salt. The separation properties of the samples as H2S adsorbents were evaluated using dynamic breakthrough experiments. Dissimilarities in the adsorption properties were found to be related to the surface properties. The experimental results also indicated that a mixture containing polymer and sludge enhances the volume of H2S oxidized/adsorbed as compared with the materials prepared from virgin precursors (pure polymer or sludge). A sewage sludge makes the catalytic centers available for the H2S oxidation process while a carbonaceous phase plays significant roles to an increment in the distribution of catalytic centers and makes additional “storage space” in the micropores of the synthesized adsorbents. There exists an optimum proportion in the concentration of the precursors where the highest performance can be attained.

2.29.8.17

Porous Glass Materials

The unique porous structure of porous glass beads makes it flexible and extremely versatile for numerous applications particularly for desulfurization owing to the low cost (cheap), and that it is pollutant free synthesized, environmentally friendly, and recyclable. Shen et al. [49] prepared porous glass beads by subcritical water treatment for removing organosulfur species from liquid hydrocarbon mixtures. Treating porous glass beads with subcritical water treatment is a green method. In order to find the mechanism of adsorption, they studied the impacts of the surface chemistry of the porous glass beads and the structure of the sulfur compounds on adsorption capacity variations. The impacts of initial concentration of the mixture, the temperature of desulfurization process, and regeneration performance were investigated as well on the adsorption capacity variations. Shen et al. [213] also prepared chitosan supported on porous glass beads adsorbent. Furthermore, they investigated the catalytic performance of monodispersed Pd or Ni nanoparticles supported on porous glass beads prepared by combining two methods of reduction and

Desulfurization Materials

50 μm

(A)

971

10 μm

(B)

50 μm

(C)

10 μm

(D)

Fig. 20 Morphology of porous glass beads ((A) and (B)) treated with hydrochloric acid ((C) and (D)), supported with chitosan. Reproduced with permission from Shen C, Wang Y, Xu J, Luo G. Chitosan supported on porous glass beads as a new green adsorbent for heavy metal recovery. Chem Eng J 2013;229:217–24. Available from: https://doi.org/10.1016/j.cej.2013.06.003.

ion exchange for cyclohexene hydrogenation [49,214–216]. The in situ technique prevents performing at high temperature and stabilizers that can decrease catalytic activity. This technique can also be used with other base nanometal materials. The proposed approach can be easily altered to synthesize other supported transition nanometal materials with catalytic characteristics and opens new windows to coordination catalysis. The morphology of porous glass beads treated with hydrochloric acid is depicted in Fig. 20.

2.29.8.18

Clay Mineral Adsorbents

Some of the most distinctive materials near the surface of the earth are clay minerals, which are created by hydrothermal and diagenetic variation of rocks in the sediments and soils. For creation of clay minerals, water is necessary, and most of them are labeled as hydrous aluminum silicates. The clay minerals adsorbents are commonly comprised of cation planes, organized in a sheetlike structure that can be octahedrally or tetrahedrally coordinated with oxygen, which in turn are sandwiched between two layers regularly labeled as 2:1 when clay minerals consist of units with 2 tetrahedral and 1 octahedral sheet or 1:1 when clay minerals consist of units of periodic octahedral and tetrahedral sheets. Furthermore, some of those with 2:1 structure have an interlayer structure among 2:1 units that can be filled by hydrated cations. The clay minerals are often comprised of remarkable quantities of alkali metals, iron, or alkaline earths [217–220]. The fundamentals by which clay minerals can be categorized are listed in Table 13. Choi et al. [221] prepared three different clay mineral adsorbents including activated clay, kaolinite, and bentonite and used them for benzothiophene sulfone (BTO) removal in an ODS process. Activated clay mineral adsorbent revealed a favorable performance in comparison with activated carbon and alumina. Moreover, activated clay mineral adsorbent exceeds kaolinite and bentonite based on adsorption capacity. Enhancing adsorptive capacity followed kaoliniteobentoniteoactivated clay order. The equilibrium adsorption of BTO over clay mineral adsorbents followed and Freundlich Langmuir models confirming a heterogeneous and monolayer adsorption. They also investigated the impact of different parameters including temperature, time, and concentration on the efficiency of adsorption desulfurization process. Ahmad et al. [222] studied desulfurization of diesel oil and kerosene using metals (Zn, Mn, Cr, Fe, Ni, Co, Ag, and Pb) impregnated montmorollonite clay (MMT) as adsorbents. MMTs were

972

Table 13

Desulfurization Materials

Values of r and the relevant type of adsorption isotherm

Layer type

Layer charge (q)

Group

Subgroup

Species (e.g.)

1:1

q≈0

Kaolin-Serpentine

2:1

q≈0 q≈1

Kaolin Serpentine Pyrophyllite Talc Di-smectite Tri-smectite Di-vermiculite Tri-vermiculite Di-mica Tri-mica Di-chlorite Tri-chlorite Sepiolite Palygorskite Di-mica-di-smectite Tri-chlorite-tri-smectite

Kaolinite Berthierine Pyrophyllite Talc Montmorillonite Saponite Di-vermiculite Tri-vermiculite Illite, Muscovite Biotite Sudoite Chamosite Sepiolite Palygorskite Rectorite Corrensite

Enhancing layer charge

Pyrophyllite-talc Smectite (q≈0.2–0.6) Vermiculite (q≈0.6–0.9) Mica (q≈1.0)

q variable

Chlorite Sepiolite-Palygorskite

Variable

q variable

Mixed layer

synthesized by wet impregnation technique. The adsorption desulfurization of MMTs revealed that high sulfur removal can be achieved by Zn-MMT. For kerosene and diesel, the highest desulfurization by mean of Zn-MMT adsorbent was 76% and 77%. Shakirullah et al. [223] studied various clay minerals such as montmorollonite, kaolinite, vermiculite, and palygorskite for the ADS of kerosene crude oil, and diesel oil. Kaolinite showed the highest yield in sulfur removal.

2.29.9

Factors Affecting the Adsorption Process

The ADS process is influenced by several parameters including nature of nanoadsorbate, surface area, volume of pores, temperature, solution pH, moisture, inferring substances, and nanoadsorbent’s dose [224–226]. Other parameters that affect the ADS process are concentration or pressure, mixing rate, residence time [19,81,85], degree of ionization, and the size of the organosulfur according to the pore size. ADS is a surface based process and therefore an excellently divided surface area and extra microporous adsorbent suggest a higher adsorption rate [225]. Moreover, restriction of internal diffusion rate and mass transport diffusion rate is decreased by means of nanoadsorbents with smaller particle size [225]. Physical ADS phenomenon is commonly exothermic and therefore enhancing the temperature reduces the adsorption rate whereas for endothermic reaction high temperatures are desirable [225]. Inorganic and organic components, which exist in the fuel, can interfere with one another or can commonly increase desulfurization. For instance, Wang et al. [85] explored that the existence of aromatics compounds in the fuel reduces the ADS capacity. The nature of the nanoadsorbent is also vital because the nanoadsorbent with a higher adsorption capacity decreases the extent of adsorption [225]. Lu [227] synthesized different microporous adsorbent materials from solid waste carbonaceous materials such as sewage sludge, coal reject, and sawdust for simultaneous NOx and SOx removal. Pore structural evolution and surface area of adsorbents were studied to show the significance of processing parameters including hold time and pyrolysis temperature, carbon burn-off and activation methods, etc. From these parameters, activation chemicals, pyrolysis temperature, and carbon burn-off have remarkable impacts on the pore structure evolution and the surface area enhancement. Optimum hold time and pyrolysis temperatures were found to achieve large surface area for whole precursor materials. The hold time and carbon burn-off are vital parameters in developing the surface area of the adsorbent. The composition of the activating agent is a significant parameter in chemical modification of sewage sludge.

2.29.10

Conclusions and Future Work

The following conclusions can be also drawn from this study: numerous nanoadsorbents including zeolites, activated carbon, graphite, silica, calcium oxide, copper oxide magnesia, manganese oxide titania, zirconia, ceria, mixed oxides, MOFs, nickel-based adsorbents, adsorptive polymers, sludge derived adsorbents, porous glass materials, and clay mineral adsorbents have been assessed for selective adsorption of sulfur compounds. Some of the adsorptive nanomaterials have been found regenerable, in numerous cases even reaching 100% regenerability. Even though the adsorptive properties of some adsorbent nanomaterials are notable, the regenerability of the nanomaterials requires enhancement. Nanomaterials that use nanometal oxides instead of metal ions have also presented desirable regeneration. Even though these kinds of nanomaterials exhibit lower adsorption capacity, it may be valuable to investigate f- and d-block nanometal oxides as a coating for high surface area nanomaterials. Preferably, no loss in capacity must be witnessed during several cycles. With the available nanomaterials, the interaction between the structure and

Desulfurization Materials

973

nanometal is very weak to avoid metal loss. An ideal approach to avoid this could be to better anchor the nanometal sites to the structure. Noble nanometals such as gold, silver, and palladium are found to be most suitable active sites for future desulfurization. The structure of pore size requires being higher than trimethyldibenzothiophene compounds and larger that it is not useful (e.g., mesoporosity of MCM-41 is adequate while macropores of SBA-15 is inessential). Furthermore, nanomaterials’ characteristics (ligand functionalization or chemistry of nanometal oxide surface) that lead for widely distributed nanometals typically lead for larger desulfurization capability. Thermal regeneration performs better once the nanomaterials are incorporated with a metal oxide, but in the thermal regeneration technique, sulfur compounds are released into the environment. Solvent regeneration separates the sulfur compounds, avoiding additional contamination. This method may be applied on functionalized nanomaterials and avoids the oxidation of nanometal ions and the requirement for an inert environment. As of now, the selective adsorption of sulfur compounds still has some problems: (1) regeneration and restricted saturated adsorption capacity of the nanoadsorbents make the desulfurization hard to use in industrial applications; (2) the impacts of nanomaterial structure and its defects on desulfurization have not been presented yet; and (3) investigations on the adsorption mechanism reaction have been focused on some particular nanoadsorbents, and a global model for all nanoadsorbents is still not presented. The zeolites, activated carbons, mixed oxides, MOF, adsorptive polymers, sludge derived adsorbents, and clay mineral adsorbents have shown good desulfurization performance. The usage of the overviewed nanoadsorbents can help different engineers to remove extra organosulfur derivative compounds; therefore, energy and money would be saved. To achieve this goal, interaction between the organosulfur derivative compounds and nanoadsorbents material needs to be assessed and comprehended. One further future direction takes account of the usage of green nanoadsorbent, which has motivated engineers to start their examination to find a cost-effective nanoadsorbent. Consequently, it is very important to choose a high efficiency and more economic ADS approach.

Acknowledgments The authors would like to acknowledge the financial support from Open Fund of Key Laboratory of Low-Grade Energy Utilization Technologies and Systems, Ministry of Education of China (Grant Nos. LLEUTS-201708, LLEUTS-201307), Scientific Research Fund of Chongqing University of Technology (Grant No. 2016ZD07), XingHuo Support Plan for Youth Scientific Research of Chongqing University of Technology (Grant No. 2014XH20), Open Fund of Fujian Province University Key Laboratory of Green Energy and Environment Catalysis (Grant No. FJ-GEEC201702), and Scientific and Technological Research Program of Chongqing Municipal Education Commission (Grant Nos. KJ1709193, KJ1500940).

References [1] Fasihi M, Shirazian S, Marjani A, Rezakazemi M. Computational fluid dynamics simulation of transport phenomena in ceramic membranes for SO2 separation. Math Comput Modelling 2012;56:278–86. doi:10.1016/j.mcm.2012.01.010. [2] Shirazian S, Rezakazemi M, Marjani A, Rafivahid MS. Development of a mass transfer model for simulation of sulfur dioxide removal in ceramic membrane contactors. Asia-Pac J Chem Eng 2012;7:828–34. doi:10.1002/apj.641. [3] Zhang Z, Yan Y, Wood DA, et al. Influence of the membrane module geometry on SO2 removal: a numerical study. Ind Eng Chem Res 2015;54:11619–27. doi:10.1021/ acs.iecr.5b03374. [4] Chandra Srivastava V. An evaluation of desulfurization technologies for sulfur removal from liquid fuels. RSC Adv 2012;2:759–83. doi:10.1039/C1RA00309G. [5] Bashkova S, Bagreev A, Locke DC, Bandosz TJ. Adsorption of SO2 on sewage sludge-derived materials. Environ Sci Technol 2001;35:3263–9. doi:10.1021/es010557u. [6] Savage M, Cheng Y, Easun TL, et al. Selective adsorption of sulfur dioxide in a robust metal–organic framework material. Adv. Mater. (Weinheim, Ger.) 2016;28:8705–11. doi:10.1002/adma.201602338. [7] Sun F, Gao J, Zhu Y, et al. Adsorption of SO2 by typical carbonaceous material: a comparative study of carbon nanotubes and activated carbons. Adsorption 2013;19:959–66. doi:10.1007/s10450-013-9504-9. [8] Johnson G, Hunter G. How to deal with fuel found in theater: AVL cypress-cylinder pressure based combustion control for consistent performance with varying fuel properties, DTIC Document; 2012. [9] Mazloomi K, Gomes C. Hydrogen as an energy carrier: prospects and challenges. Renew Sustain Energy Rev 2012;16:3024–33. Available from: https://doi.org/10.1016/j. rser.2012.02.028. [10] Van de Krol R, Grätzel M. Photoelectrochemical hydrogen production. Springer; 2012. http://dx.doi.org/10.1007/978-1-4614-1380-6 [11] Satyapal S, Petrovic J, Read C, Thomas G, Ordaz G. The U.S. Department of Energy’s National Hydrogen Storage Project: progress towards meeting hydrogen-powered vehicle requirements. Catal Today 2007;120:246–56. Available from: http://dx.doi.org/10.1016/j.cattod.2006.09.022. [12] Campanari S, Manzolini G, Garcia de la Iglesia F. Energy analysis of electric vehicles using batteries or fuel cells through well-to-wheel driving cycle simulations. J. Power Sources 2009;186:464–77. Available from: http://dx.doi.org/10.1016/j.jpowsour.2008.09.115. [13] Gomez A, Berry JJ, Roychoudhury S, Coriton B, Huth J. From jet fuel to electric power using a mesoscale, efficient Stirling cycle. Proc Combust Inst 2007;31:3251–9. [14] Lee IC, Ubanyionwu HC. Determination of sulfur contaminants in military jet fuels. Fuel 2008;87:312–8. Available from: http://dx.doi.org/10.1016/j.fuel.2007.05.010. [15] Rezakazemi M, Sadrzadeh M, Mohammadi T, Matsuura T. Methods for the preparation of organic–inorganic nanocomposite polymer electrolyte membranes for fuel cells. In: Inamuddin D, Mohammad A, Asiri AM, editors. Organic–inorganic composite polymer electrolyte membranes: preparation, properties, and fuel cell applications. Cham: Springer International Publishing; 2017. p. 311–25. Available from: http://dx.doi.org/10.1007/978-3-319-52739-0_11. [16] Sun X, Tatarchuk BJ. Photo-assisted adsorptive desulfurization of hydrocarbon fuels over TiO2 and Ag/TiO2. Fuel 2016;183:550–6. [17] Ito E, van Veen JAR. On novel processes for removing sulphur from refinery streams. Catal Today 2006;116:446–60. [18] Laan JV. ConocoPhillips S Zorb gasoline sulfur removal technology: unique chemistry, proven performance, and optimized design. Available from: http://www.icheh.com/ Files/Posts/Portal1/S-Zorb.pdf. [19] Ma X, Velu S, Kim JH, Song C. Deep desulfurization of gasoline by selective adsorption over solid adsorbents and impact of analytical methods on ppm-level sulfur quantification for fuel cell applications. Appl Catal B: Environ 2005;56:137–47. Available from: http://doi.org/10.1016/j.apcatb.2004.08.013.

974

Desulfurization Materials

[20] Gao X, Ding H, Du Z, et al. Gas–liquid absorption reaction between (NH4)2SO3 solution and SO2 for ammonia-based wet flue gas desulfurization. Appl Energy 2010;87:2647–51. Available from: http://doi.org/10.1016/j.apenergy.2010.03.023. [21] Jia Y, Zhong Q, Fan X, Wang X. Kinetics of oxidation of total sulfite in the ammonia-based wet flue gas desulfurization process. Chem Eng J 2010;164:132–8. Available from: http://doi.org/10.1016/j.cej.2010.08.041. [22] Wang SJ, Zhu P, Zhang G, et al. Numerical simulation research of flow field in ammonia-based wet flue gas desulfurization tower. J Energy Inst 2015;88:284–91. Available from: http://doi.org/10.1016/j.joei.2014.09.002. [23] Yan J, Bao J, Yang L, Fan F, Shen X. The formation and removal characteristics of aerosols in ammonia-based wet flue gas desulfurization. J Aerosol Sci 2011;42:604–14. Available from: http://doi.org/10.1016/j.jaerosci.2011.05.005. [24] Gao X, Ding H-l, Wu Z-l, et al. Analysis on leaching characteristics of iron in coal fly ash under ammonia-based wet flue gas desulfurization (WFGD) conditions. Energy Fuels 2009;23:5916–9. [25] Gutiérrez Ortiz FJ, Vidal F, Ollero P, et al. Pilot-plant technical assessment of wet flue gas desulfurization using limestone. Ind Eng Chem Res 2006;45:1466–77. doi:10.1021/ie051316o. [26] Ukawa N, Takashina T, Oshima M, Oishi T. Effects of salts on limestone dissolution rate in wet limestone flue gas desulfurization. Environ Prog 1993;12:294–9. doi:10.1002/ep.670120410. [27] Warych J, Szymanowski M. Optimum values of process parameters of the “wet limestone flue gas desulfurization system”. Chem Eng Technol 2002;25:427–32. [28] Huang S, Yuan D, Lin H, Sun L, Lin S. Fractionation of mercury stable isotopes during coal combustion and seawater flue gas desulfurization. Appl Geochem 2017;76:159–67. Available from: http://doi.org/10.1016/j.apgeochem.2016.12.002. [29] Oikawa K, Yongsiri C, Takeda K, Harimoto T. Seawater flue gas desulfurization: Its technical implications and performance results. Environ Prog 2003;22:67–73. doi:10.1002/ep.670220118. [30] Vidal B F, Ollero P, Gutiérrez Ortiz FJ, Villanueva A. Catalytic seawater flue gas desulfurization process: an experimental pilot plant study. Environ Sci Technol 2007;41:7114–9. doi:10.1021/es0706899. [31] Chang JCS, Kaplan N. SO2 removal by limestone dual alkali. Environ Prog 1984;3:267–74. doi:10.1002/ep.670030412. [32] Chmielewski AG, Licki J, Pawelec A, Tymin´ski B, Zimek Z. Operational experience of the industrial plant for electron beam flue gas treatment. Radiat Phys Chem 2004;71:441–4. Available from: http://dx.doi.org/10.1016/j.radphyschem.2004.03.020. [33] Chou M-IM, Bruinius JA, Benig V, Chou S-FJ, Carty RH. Producing ammonium sulfate from flue gas desulfurization by-products. Energy Sources 2005;27:1061–71. doi:10.1080/00908310490479510. [34] Srivastava RK, Jozewicz W, Singer C. SO2 scrubbing technologies: a review. Environ Prog 2001;20:219–28. doi:10.1002/ep.670200410. [35] Zhang Q, Gui K. A novel semidry flue gas desulfurization process with the magnetically fluidized bed reactor. J Hazard Mater 2009;168:1341–5. Available from: http://dx. doi.org/10.1016/j.jhazmat.2009.03.019. [36] Gutiérrez Ortiz FJ, Ollero P. Modeling of the in-duct sorbent injection process for flue gas desulfurization. Sep Purif Technol 2008;62:571–81. Available from: http://dx. doi.org/10.1016/j.seppur.2008.03.012. [37] Jozewicz W, Chang JCS, Sedman CB, Brna TG. Characterization of advanced sorbents for dry SO2 control. Reactivity Solids 1988;6:243–62. Available from: http://dx.doi. org/10.1016/0168-7336(88)80065-2. [38] Wang N, Zhang X. Effect of humidification water on semi-dry flue gas desulfurization. Procedia Environ Sci 2011;11:1023–8. Available from: http://dx.doi.org/10.1016/j. proenv.2011.12.156. [39] Zhong Y, Gao X, Huo W, et al. A model for performance optimization of wet flue gas desulfurization systems of power plants. Fuel Process Technol 2008;89:1025–32. Available from: http://dx.doi.org/10.1016/j.fuproc.2008.04.004. [40] Yan K, Li R, Zhu T, et al. A semi-wet technological process for flue gas desulfurization by corona discharges at an industrial scale. Chem Eng J 2006;116:139–47. Available from: https://doi.org/10.1016/j.cej.2005.09.030. [41] Xu G, Guo Q, Kaneko T, Kato K. A new semi-dry desulfurization process using a powder-particle spouted bed. Adv Environ Res 2000;4:9–18. Available from: https://doi. org/10.1016/S1093-0191(00)00003-4. [42] Ma X, Kaneko T, Guo Q, Xu G, Kato K. Removal of SO2 from flue gas using a new semidry flue gas desulfurization process with a powder-particle spouted bed. Can J Chem Eng 1999;77:356–62. doi:10.1002/cjce.5450770222. [43] Agarwal M, Dikshit PK, Bhasarkar JB, Borah AJ, Moholkar VS. Physical insight into ultrasound-assisted biodesulfurization using free and immobilized cells of Rhodococcus rhodochrous MTCC 3552. Chem Eng J 2016;295:254–67. Available from: http://dx.doi.org/10.1016/j.cej.2016.03.042. [44] Zhang C, Pan X, Wang F, Liu X. Extraction–oxidation desulfurization by pyridinium-based task-specific ionic liquids. Fuel 2012;102:580–4. Available from: http://dx.doi. org/10.1016/j.fuel.2012.07.040. [45] Gao L, Wan H, Han M, Guan G. Deep desulfurization of model oil by extraction with a low-viscosity ionic liquid [BMIM]SCN. Pet Sci Technol 2014;32:1309–17. doi:10.1080/10916466.2011.653466. [46] Tian Y, Wang G, Long J, et al. Ultra-deep oxidative desulfurization of fuel with H2O2 catalyzed by phosphomolybdic acid supported on silica. Chin J Catal 2016;37:2098–105. Available from: http://dx.doi.org/10.1016/S1872-2067(16)62558-5. [47] Ren X, Miao G, Xiao Z, et al. Catalytic adsorptive desulfurization of model diesel fuel using TiO2/SBA-15 under mild conditions. Fuel 2016;174:118–25. Available from: http://dx.doi.org/10.1016/j.fuel.2016.01.093. [48] Guo B, Wang R, Li Y. Gasoline alkylation desulfurization over Amberlyst 35 resin: Influence of methanol and apparent reaction kinetics. Fuel 2011;90:713–8. Available from: http://dx.doi.org/10.1016/j.fuel.2010.10.010. [49] Shen C, Wang YJ, Xu JH, Lu YC, Luo GS. Porous glass beads as a new adsorbent to remove sulfur-containing compounds. Green Chem 2012;14:1009–15. doi:10.1039/C2GC16559G. [50] Hernández-Maldonado AJ, Yang FH, Qi G, Yang RT. Desulfurization of transportation fuels by π-complexation sorbents: Cu(I)-, Ni(II)-, and Zn(II)-zeolites. Appl Catal B: Environ 2005;56:111–26. Available from: https://doi.org/10.1016/j.apcatb.2004.06.023. [51] Seredych M, Bandosz TJ. Adsorption of dibenzothiophenes on nanoporous carbons: identification of specific adsorption sites governing capacity and selectivity. Energy Fuels 2010;24:3352–60. doi:10.1021/ef9015087. [52] Xiao J, Bian G, Zhang W, Li Z. Adsorption of dibenzothiophene on Ag/Cu/Fe-supported activated carbons prepared by ultrasonic-assisted impregnation. J Chem Eng Data 2010;55:5818–23. doi:10.1021/je1007795. [53] Tang X-l, Qian W, Hu A, et al. Adsorption of thiophene on Pt/Ag-supported activated carbons prepared by ultrasonic-assisted impregnation. Ind Eng Chem Res 2011;50:9363–7. doi:10.1021/ie2008778. [54] Kim JH, Ma X, Zhou A, Song C. Ultra-deep desulfurization and denitrogenation of diesel fuel by selective adsorption over three different adsorbents: A study on adsorptive selectivity and mechanism. Catal Today 2006;111:74–83. Available from: https://doi.org/10.1016/j.cattod.2005.10.017. [55] Cychosz KA, Wong-Foy AG, Matzger AJ. Enabling cleaner fuels: desulfurization by adsorption to microporous coordination polymers. J Am Chem Soc 2009;131:14538–43. doi:10.1021/ja906034k. [56] Shi F, Hammoud M, Thompson LT. Selective adsorption of dibenzothiophene by functionalized metal organic framework sorbents. Appl Catal B: Environ 2011;103:261–5. Available from: https://doi.org/10.1016/j.apcatb.2010.07.016. [57] Sentorun-Shalaby C, Saha SK, Ma X, Song C. Mesoporous-molecular-sieve-supported nickel sorbents for adsorptive desulfurization of commercial ultra-low-sulfur diesel fuel. Appl Cataly B: Environ 2011;101:718–26. Available from: https://doi.org/10.1016/j.apcatb.2010.11.014.

Desulfurization Materials

975

[58] Javadli R, de Klerk A. Desulfurization of heavy oil. Appl Petrochem Res 2012;1:3–19. doi:10.1007/s13203-012-0006-6. [59] Ahmad W, Ahmad I, Ishaq M, Ihsan K. Adsorptive desulfurization of kerosene and diesel oil by Zn impregnated montmorollonite clay. Arabian J Chem 2017;2:S3263–9. Available from: https://doi.org/10.1016/j.arabjc.2013.12.025. [60] Ullah R, Bai P, Wu P, et al. Comparison of the reactive adsorption desulfurization performance of Ni/ZnO–Al2O3 adsorbents prepared by different methods. Energy Fuels 2016;30:2874–81. doi:10.1021/acs.energyfuels.6b00232. [61] Wang L, Zhao L, Xu C, Wang Y, Gao J. Screening of active metals for reactive adsorption desulfurization adsorbent using density functional theory. Appl Surf Sci 2017;399:440–50. Available from: https://doi.org/10.1016/j.apsusc.2016.11.160. [62] Jiwalak N, Rattanaphani S, Bremner JB, Rattanaphani V. Equilibrium and kinetic modeling of the adsorption of indigo carmine onto silk. Fibers Polym 2010;11:572–9. doi:10.1007/s12221-010-0572-2. [63] Reed NJ. A comparative study of adsorption desulfurization of liquid transportation fuels over different sorbents for fuel cell applications, The Pennsylvania State University; 2008. [64] Seader J. Adsorption, lon exchange and chromatography. In: Seader JD, Henley EJ, editors. Separation process principles. New York, NY: John Wiley and Sons; 1998 [chapter 15]. [65] Kumar PS, Kirthika K. Equilibrium and kinetic study of adsorption of nickel from aqueous solution onto bael tree leaf powder. J Eng Sci Technol 2009;4:351–63. [66] Mittal A, Kurup L, Mittal J. Freundlich and Langmuir adsorption isotherms and kinetics for the removal of Tartrazine from aqueous solutions using hen feathers. J Hazard Mater 2007;146:243–8. Available from: https://doi.org/10.1016/j.jhazmat.2006.12.012. [67] Lakshmi UR, Srivastava VC, Mall ID, Lataye DH. Rice husk ash as an effective adsorbent: Evaluation of adsorptive characteristics for Indigo Carmine dye. J Environ Manag 2009;90:710–20. Available from: https://doi.org/10.1016/j.jenvman.2008.01.002. [68] Weber TW, Chakravorti RK. Pore and solid diffusion models for fixed-bed adsorbers. AIChE J. 1974;20:228–38. doi:10.1002/aic.690200204. [69] Hameed BH, Mahmoud DK, Ahmad AL. Equilibrium modeling and kinetic studies on the adsorption of basic dye by a low-cost adsorbent: Coconut (Cocos nucifera) bunch waste. J Hazard Mater 2008;158:65–72. Available from: https://doi.org/10.1016/j.jhazmat.2008.01.034. [70] Chiou M-S, Li H-Y. Equilibrium and kinetic modeling of adsorption of reactive dye on cross-linked chitosan beads. J Hazard Mater 2002;93:233–48. Available from: https://doi.org/10.1016/S0304-3894(02)00030-4. [71] Velu S, Ma X, Song C. Selective adsorption for removing sulfur from jet fuel over zeolite-based adsorbents. Ind Eng Chem Res 2003;42:5293–304. doi:10.1021/ ie020995p. [72] Shen Y, Xu X, Li P. A novel potential adsorbent for ultra deep desulfurization of jet fuels at room temperature. RSC Adv 2012;2:6155–60. doi:10.1039/C2RA20224G. [73] Kim JH, Ma X, Zhou A, Song C. Ultra-deep desulfurization and denitrogenation of diesel fuel by selective adsorption over three different adsorbents: a study on adsorptive selectivity and mechanism. Catal Today 2006;111:74–83. Available from: http://doi.org/10.1016/j.cattod.2005.10.017. [74] Yang RT, Takahashi A, Yang FH. New sorbents for desulfurization of liquid fuels by π-complexation. Ind Eng Chem Res 2001;40:6236–9. doi:10.1021/ie010729w. [75] Ma X, Sun L, Song C. A new approach to deep desulfurization of gasoline, diesel fuel and jet fuel by selective adsorption for ultra-clean fuels and for fuel cell applications. Catal Today 2002;77:107–16. Available from: http://doi.org/10.1016/S0920-5861(02)00237-7. [76] Khajeh M, Laurent S, Dastafkan K. Nanoadsorbents: classification, preparation, and applications (with emphasis on aqueous media). Chem Rev (Washington, DC, US) 2013;113:7728–68. doi:10.1021/cr400086v. [77] Fringuelli F, Marino G, Taticchi A, Grandolini G. A comparative study of the aromatic character of furan, thiophen, selenophen, and tellurophen. J Chem Society, Perkin Trans 1974;2:332–7. doi:10.1039/P29740000332. [78] Zhou A, Ma X, Song C. Liquid-phase adsorption of multi-ring thiophenic sulfur compounds on carbon materials with different surface properties. J Phys Chem B 2006;110:4699–707. doi:10.1021/jp0550210. [79] Bartholomew CH, Farrauto RJ. Fundamentals of industrial catalytic processes. New York, NY: John Wiley & Sons; 2011. [80] Ma C-Q, Feng J-H, Zeng Y-Y, et al. Methods for the preparation of a biodesulfurization biocatalyst using Rhodococcus sp. Chemosphere 2006;65:165–9. Available from: https://doi.org/10.1016/j.chemosphere.2006.03.010. [81] Ma X, Sprague M, Song C. Deep desulfurization of gasoline by selective adsorption over nickel-based adsorbent for fuel cell applications. Ind Eng Chem Res 2005;44:5768–75. doi:10.1021/ie0492810. [82] Hernández-Maldonado AJ, Yang RT. Desulfurization of liquid fuels by adsorption via π complexation with Cu(I)−Y and Ag−Y zeolites. Ind Eng Chem Res 2003;42:123–9. doi:10.1021/ie020728j. [83] Park JG, Ko CH, Yi KB, et al. Reactive adsorption of sulfur compounds in diesel on nickel supported on mesoporous silica. Appl Catal B: Environ 2008;81:244–50. Available from: https://doi.org/10.1016/j.apcatb.2007.12.014. [84] Wang L, Sun B, Yang FH, Yang RT. Effects of aromatics on desulfurization of liquid fuel by π-complexation and carbon adsorbents. Chem Eng Sci 2012;73:208–17. Available from: https://doi.org/10.1016/j.ces.2012.01.056. [85] Wang J, Xu F, Xie W-J, et al. The enhanced adsorption of dibenzothiophene onto cerium/nickel-exchanged zeolite Y. J Hazard Mater 2009;163:538–43. Available from: https://doi.org/10.1016/j.jhazmat.2008.07.027. [86] Shahverdi M, Baheri B, Rezakazemi M, Motaee E, Mohammadi T. Pervaporation study of ethylene glycol dehydration through synthesized (PVA–4A)/polypropylene mixed matrix composite membranes. Polym Eng Sci 2013;53:1487–93. doi:10.1002/pen.23406. [87] Baheri B, Shahverdi M, Rezakazemi M, Motaee E, Mohammadi T. Performance of PVA/NaA mixed matrix membrane for removal of water from ethylene glycol solutions by pervaporation. Chem Eng Commun 2015;202:316–21. doi:10.1080/00986445.2013.841149. [88] Rezakazemi M, Shahidi K, Mohammadi T. Hydrogen separation and purification using crosslinkable PDMS/zeolite A nanoparticles mixed matrix membranes. Int J Hydrogen Energy 2012;37:14576–89. doi:10.1016/j.ijhydene.2012.06.104. [89] Rezakazemi M, Shahidi K, Mohammadi T. Sorption properties of hydrogen-selective PDMS/zeolite 4A mixed matrix membrane. Int J Hydrogen Energy 2012;37:17275–84. doi:10.1016/j.ijhydene.2012.08.109. [90] Rezakazemi M, Mohammadi T. Gas sorption in H2-selective mixed matrix membranes: Experimental and neural network modeling. Int J Hydrogen Energy 2013;38:14035–41. Available from: http://dx.doi.org/10.1016/j.ijhydene.2013.08.062. [91] Rostamizadeh M, Rezakazemi M, Shahidi K, Mohammadi T. Gas permeation through H2-selective mixed matrix membranes: experimental and neural network modeling. Int J Hydrogen Energy 2013;38:1128–35. Available from: http://dx.doi.org/10.1016/j.ijhydene.2012.10.069. [92] Rezakazemi M, Dashti A, Asghari M, Shirazian S. H2-selective mixed matrix membranes modeling using ANFIS, PSO-ANFIS, GA-ANFIS. Int J Hydrogen Energy 2017;42:15211–25. Available from: https://doi.org/10.1016/j.ijhydene.2017.04.044. [93] Anis SF, Khalil A, Saepurahman , Singaravel G, Hashaikeh R. A review on the fabrication of zeolite and mesoporous inorganic nanofibers formation for catalytic applications. Microporous Mesoporous Mater 2016;236:176–92. Available from: https://doi.org/10.1016/j.micromeso.2016.08.043. [94] Ilić B, Wettstein SG. A review of adsorbate and temperature-induced zeolite framework flexibility. Microporous Mesoporous Mater 2017;239:221–34. Available from: https://doi.org/10.1016/j.micromeso.2016.10.005. [95] Ma Y, Tong W, Zhou H, Suib SL. A review of zeolite-like porous materials. Microporous Mesoporous Mater 2000;37:243–52. Available from: https://doi.org/10.1016/ S1387-1811(99)00199-7. [96] Weckhuysen BM, Yu J. Recent advances in zeolite chemistry and catalysis. Chem. Soc. Rev. 2015;44:7022–4. doi:10.1039/C5CS90100F. [97] Li H, Han X, Huang H, et al. Competitive adsorption desulfurization performance over K – Doped NiY zeolite. J Colloid Interface Sci 2016;483:102–8. Available from: http://dx.doi.org/10.1016/j.jcis.2016.08.024.

976

Desulfurization Materials

[98] Velu S, Song C, Engelhard MH, Chin Y-H. Adsorptive removal of organic sulfur compounds from jet fuel over K-exchanged NiY zeolites prepared by impregnation and ion exchange. Ind Eng Chem Res 2005;44:5740–9. doi:10.1021/ie0488492. [99] Lee KX, Valla JA. Investigation of metal-exchanged mesoporous Y zeolites for the adsorptive desulfurization of liquid fuels. Appl Catal B: Environ 2017;201:359–69. Available from: https://doi.org/10.1016/j.apcatb.2016.08.018. [100] Rezakazemi M, Ebadi Amooghin A, Montazer-Rahmati MM, Ismail AF, Matsuura T. State-of-the-art membrane based CO2 separation using mixed matrix membranes (MMMs): an overview on current status and future directions. Prog Polym Sci 2014;39:817–61. Available from: http://dx.doi.org/10.1016/j.progpolymsci.2014.01.003. [101] Baltzopoulou P, Kallis KX, Karagiannakis G, Konstandopoulos AG. Diesel Fuel Desulfurization via Adsorption with the aid of activated carbon: laboratory- and pilot-scale studies. Energy Fuels 2015;29:5640–8. doi:10.1021/acs.energyfuels.5b01133. [102] Yu G, Lu S, Chen H, Zhu Z. Diesel fuel desulfurization with hydrogen peroxide promoted by formic acid and catalyzed by activated carbon. Carbon 2005;43:2285–94. Available from: http://dx.doi.org/10.1016/j.carbon.2005.04.008. [103] López D, Buitrago R, Sepúlveda-Escribano A, Rodríguez-Reinoso F, Mondragón F. Low temperature catalytic adsorption of SO2 on activated carbon,. J Phys Chem C 2008;112:15335–40. doi:10.1021/jp802809c. [104] Carabineiro SAC, Ramos AM, Vital J, et al. Adsorption of SO2 using vanadium and vanadium–copper supported on activated carbon. Catal Today 2003;78:203–10. Available from: http://doi.org/10.1016/S0920-5861(02)00335-8. [105] Gao X, Liu S, Zhang Y, Luo Z, Cen K. Physicochemical properties of metal-doped activated carbons and relationship with their performance in the removal of SO2 and NO. J Hazard Mater 2011;188:58–66. DOI: http://doi.org/10.1016/j.jhazmat.2011.01.065. [106] Davini P. Influence of surface properties and iron addition on the SO2 adsorption capacity of activated carbons. Carbon 2002;40:729–34. Available from: http://doi.org/ 10.1016/S0008-6223(01)00161-0. [107] Sumathi S, Bhatia S, Lee KT, Mohamed AR. Adsorption isotherm models and properties of SO2 and NO removal by palm shell activated carbon supported with cerium (Ce/PSAC). Chem Eng J 2010;162:194–200. Available from: https://doi.org/10.1016/j.cej.2010.05.028. [108] Shi Y, Zhang X, Liu G. Activated carbons derived from hydrothermally carbonized sucrose: remarkable adsorbents for adsorptive desulfurization. ACS Sustain Chem Eng 2015;3:2237–46. doi:10.1021/acssuschemeng.5b00670. [109] Chen T-C, Agripa ML, Lu M-C, Dalida MLP. Adsorption of sulfur compounds from diesel with ion-impregnated activated carbons. Energy Fuels 2016;30:3870–8. doi:10.1021/acs.energyfuels.6b00230. [110] Danmaliki GI, Saleh TA. Effects of bimetallic Ce/Fe nanoparticles on the desulfurization of thiophenes using activated carbon. Chem Eng J 2017;307:914–27. Available from: https://doi.org/10.1016/j.cej.2016.08.143. [111] Xiong L, Chen F-X, Yan X-M, Mei P. The adsorption of dibenzothiophene using activated carbon loaded with cerium. J Porous Mater 2012;19:713–9. doi:10.1007/ s10934-011-9523-4. [112] Fan L, Jiang X, Jiang W, Guo J, Chen J. Physicochemical properties and desulfurization activities of metal oxide/biomass-based activated carbons prepared by blending method. Adsorption 2014;20:747–56. doi:10.1007/s10450-014-9618-8. [113] Altıntıg E, Altundag H, Tuzen M, Sarı A. Effective removal of methylene blue from aqueous solutions using magnetic loaded activated carbon as novel adsorbent. Chem Eng Res Des 2017;122:151–63. Available from: https://doi.org/10.1016/j.cherd.2017.03.035. [114] Wang G-J, Zhang J-K, Liu Y. Catalytic oxidative desulfurization of benzothiophene with hydrogen peroxide over Fe/AC in a biphasic model diesel-acetonitrile system. Korean J Chem Eng 2013;30:1559–65. doi:10.1007/s11814-013-0052-5. [115] Saleh TA, Sulaiman KO, Al-Hammadi SA, Dafalla H, Danmaliki GI. Adsorptive desulfurization of thiophene, benzothiophene and dibenzothiophene over activated carbon manganese oxide nanocomposite: with column system evaluation. J Clean Prod 2017;154:401–12. Available from: https://doi.org/10.1016/j.jclepro.2017.03.169. [116] You F-T, Yu G-W, Wang Y, et al. Study of nitric oxide catalytic oxidation on manganese oxides-loaded activated carbon at low temperature. Appl Surf Sci 2017;413:387–97. Available from: https://doi.org/10.1016/j.apsusc.2017.04.044. [117] Mangun CL, DeBarr JA, Economy J. Adsorption of sulfur dioxide on ammonia-treated activated carbon fibers. Carbon 2001;39:1689–96. Available from: http://dx.doi.org/ 10.1016/S0008-6223(00)00300-6. [118] Daley MA, Mangun CL, DeBarrb JA, et al. Adsorption of SO2 onto oxidized and heat-treated activated carbon fibers (ACFs). Carbon 1997;35:411–7. Available from: http://dx.doi.org/10.1016/S0008-6223(97)89612-1. [119] Hong I, Jiang H, Park Y-D, Kim J-Y, Ha B-H. Metal dispersed activated carbon fibers and their application for removal of SOx. Chem. Phys. Lett. 2002;366:572–7. Available from: http://doi.org/10.1016/S0009-2614(02)01631-7. [120] Liu W, Adanur S. Desulfurization properties of activated carbon fibers. J Eng Fibers Fabrics 2014;9:70–3. [121] Gaur V, Sharma A, Verma N. Preparation and characterization of ACF for the adsorption of BTX and SO2. Chem Eng Process: Process Intensification 2006;45:1–13. Available from: http://doi.org/10.1016/j.cep.2005.04.006. [122] Song X, Wang Z, Li Z, Wang C. Ultrafine porous carbon fibers for SO2 adsorption via electrospinning of polyacrylonitrile solution. J Colloid Interface Sci 2008;327:388–92. Available from: http://doi.org/10.1016/j.jcis.2008.02.070. [123] Kisiela AM, Czajka KM, Moron´ W, Rybak W, Andryjowicz C. Unburned carbon from lignite fly ash as an adsorbent for removal. Energy 2016;116(Part 3):1454–63. Available from: https://doi.org/10.1016/j.energy.2016.02.143. [124] Yongqi Z, Yitian F, Jiejie H, Yang W. Utilization of fly ash to prepare adsorbent for SO2 removal in flue gas. Energy Sources, A: Recovery Utilization Environ Effects 2006;28:727–41. doi:10.1080/009083190523226. [125] Shimoyama I, Baba Y. Thiophene adsorption on phosphorus- and nitrogen-doped graphites: Control of desulfurization properties of carbon materials by heteroatom doping. Carbon 2016;98:115–25. Available from: http://dx.doi.org/10.1016/j.carbon.2015.10.094. [126] Dai J, Yuan J, Giannozzi P. Gas adsorption on graphene doped with B, N, Al, and S: A theoretical study. Appl Phys Lett 2009;95:232105. doi:10.1063/1.3272008. [127] Rezakazemi M, Vatani A, Mohammadi T. Synergistic interactions between POSS and fumed silica and their effect on the properties of crosslinked PDMS nanocomposite membranes. RSC Adv 2015;5:82460–70. doi:10.1039/C5RA13609A. [128] Wang Y, Yang RT, Heinzel JM. Desulfurization of jet fuel by -complexation adsorption with metal halides supported on MCM-41 and SBA-15 mesoporous materials. Chem Eng Sci 2008;63:356–65. Available from: http://dx.doi.org/10.1016/j.ces.2007.09.002. [129] Chen H, Wang Y, Yang FH, Yang RT. Desulfurization of high-sulfur jet fuel by mesoporous π-complexation adsorbents. Chem Eng Sci 2009;64:5240–6. Available from: http://dx.doi.org/10.1016/j.ces.2009.08.031. [130] Wang Y, Yang RT, Heinzel JM. Desulfurization of jet fuel JP-5 light fraction by MCM-41 and SBA-15 supported cuprous oxide for fuel cell applications. Ind Eng Chem Res 2009;48:142–7. doi:10.1021/ie800208g. [131] Jabeen B, Rafique U. Synthesis and application of metal doped silica particles for adsorptive desulphurization of fuels. Environ Eng Res 2014;19:205–14. doi:10.4491/ eer.2014.017. [132] Palomino JM, Tran DT, Hauser JL, Dong H, Oliver SRJ. Mesoporous silica nanoparticles for high capacity adsorptive desulfurization. J Mater Chem A 2014;2:14890–5. doi:10.1039/C4TA02570A. [133] Song L, Bu T, Zhu L, et al. Synthesis of organically–inorganically functionalized MCM-41 for adsorptive desulfurization of C4 hydrocarbons. J Phys Chem C 2014;118:9468–76. doi:10.1021/jp500687v. [134] Hlincik T, Buryan P. Desulfurization of boiler flue gas by means of activated calcium oxide. Fuel Process Technol 2013;111:62–7. Available from: http://doi.org/10.1016/ j.fuproc.2013.01.018.

Desulfurization Materials

977

[135] Mura G, Lallai A, Olla P. On the kinetics of dry desulfurization with calcium oxide. Chem Eng J 1991;46:119–28. Available from: http://dx.doi.org/10.1016/0300-9467 (91)87003-S. [136] Yang RT, Chen JM. Kinetics of desulfurization of hot fuel gas with calcium oxide. Reaction between carbonyl sulfide and calcium oxide. Environ Sci Technol 1979;13:549–53. doi:10.1021/es60153a010. [137] Hartman M, Trnka O. Reactions between calcium oxide and flue gas containing sulfur dioxide at lower temperatures. AIChE J 1993;39:615–24. doi:10.1002/ aic.690390410. [138] Bruce KR, Gullett BK, Beach LO. Comparative SO2 reactivity of CaO derived from CaCo3 and Ca(OH)2. AIChE J 1989;35:37–41. doi:10.1002/aic.690350104. [139] Sasmaz E, Wilcox J. Mercury species and SO2 Adsorption on CaO(100). J Phys Chem C 2008;112:16484–90. doi:10.1021/jp801250h. [140] Siriwardane RV, Cook JM. Interactions of SO2 with sodium deposited on CaO. J Colloid Interface Sci 1986;114:525–35. Available from: http://dx.doi.org/10.1016/00219797(86)90438-8. [141] Stinespring CD, Cook JM. A study of SO2 adsorption on CaO (100) surfaces using X-ray photoelectron spectroscopy. J Electron Spectrosc Relat Phenom 1983;32:113–22. Available from: http://dx.doi.org/10.1016/0368-2048(83)85089-0. [142] Mahuli SK, Agnihotr R, Jadhav R, Chauk S, Fan LS. Combined calcination, sintering and sulfation model for CaCo3–SO2 reaction. AIChE J 1999;45:367–82. doi:10.1002/aic.690450216. [143] Shuyan W, Lijie Y, Huilin L, et al. Numerical analysis of particle clustering effects on desulphurization and NO emission in a circulating fluidized bed combustor. Fuel 2008;87:870–7. Available from: https://doi.org/10.1016/j.fuel.2007.05.052. [144] Turbevillle W, Yap N, Cai Y, Ladebeck J. Process for sulfur adsorption using copper-containing catalyst, Google Patents; 2010. [145] Egan BZ, Felker LK. Removal of sulfur dioxide from simulated flue gas by magnesia spray absorption: parameters affecting removal efficiency and products. Ind Eng Chem Process Desand Dev 1986;25:558–61. doi:10.1021/i200033a037. [146] Ziolek M, Kujawa J, Saur O, Aboulayt A, Lavalley JC. Influence of sulfur dioxide adsorption on the surface properties of metal oxides. J Mol Catal A: Chem 1996;112:125–32. Available from: http://dx.doi.org/10.1016/1381-1169(96)00222-1. [147] Freitag A, Hanson J, Larese J. Adsorption properties of SO2 on MgO (100) using in-situ x-ray diffraction/absorption and adsorption isotherm techniques. In: American Physical Society, Annual March Meeting, Washington State Convention Center Seattle, Washington; 2001. p. 12011. [148] Stark JV, Park DG, Lagadic I, Klabunde KJ. Nanoscale metal oxide particles/clusters as chemical reagents. unique surface chemistry on magnesium oxide as shown by enhanced adsorption of acid gases (sulfur dioxide and carbon dioxide) and pressure dependence. Chem Mater 1996;8:1904–12. doi:10.1021/cm950583p. [149] Eid KM, Ammar HY. Adsorption of SO2 on Li atoms deposited on MgO (1 0 0) surface: DFT calculations. Appl Surf Sci 2011;257:6049–58. Available from: https://doi. org/10.1016/j.apsusc.2011.01.122. [150] De Francesco R, Stener M, Fronzoni G. TDDFT Calculations of NEXAFS spectra of model systems for SO2 adsorbed on the MgO (100) surface. J Phys Chem C 2007;111:13554–63. doi:10.1021/jp072710y. [151] Pacchioni G, Clotet A, Ricart JM. A theoretical study of the adsorption and reaction of SO2 at surface and step sites of the MgO(100) surface. Surf Sci 1994;315:337–50. Available from: http://dx.doi.org/10.1016/0039-6028(94)90137-6. [152] Adeyi AA, Aberuagba F. Comparative analysis of adsorptive desulphurization of crude oil by manganese dioxide and zinc oxide. Res J Chem Sci 2012;2:14–20. [153] Saleh TA, Siddiqui MN, Al-Arfaj AA. Synthesis of multiwalled carbon nanotubes-titania nanomaterial for desulfurization of model fuel. J Nanomater 2014;2014:194. Available from: http://dx.doi.org/10.1155/2014/940639. [154] Guo J, Watanabe S, Janik MJ, Ma X, Song C. Density functional theory study on adsorption of thiophene on TiO2 anatase (0 0 1) surfaces. Catal Today 2010;149:218–23. Available from: http://dx.doi.org/10.1016/j.cattod.2009.05.002. [155] Shahadat Hussain AHM, Tatarchuk BJ. Adsorptive desulfurization of jet and diesel fuels using Ag/TiOx–Al2O3 and Ag/TiOx–SiO2 adsorbents. Fuel 2013;107:465–73. Available from: http://dx.doi.org/10.1016/j.fuel.2012.11.030. [156] Baeza P, Aguila G, Gracia F, Araya P. Desulfurization by adsorption with copper supported on zirconia. Catal Commun 2008;9:751–5. Available from: http://dx.doi.org/ 10.1016/j.catcom.2007.08.020. [157] Wang B, Zhu J, Ma H. Desulfurization from thiophene by SO42−/ZrO2 catalytic oxidation at room temperature and atmospheric pressure. J Hazard Mater 2009;164:256–64. Available from: http://dx.doi.org/10.1016/j.jhazmat.2008.08.003. [158] Kumar S, Srivastava VC, Badoni RP. Studies on adsorptive desulfurization by zirconia based adsorbents. Fuel 2011;90:3209–16. Available from: http://dx.doi.org/ 10.1016/j.fuel.2011.06.029. [159] Kylhammar L, Carlsson P-A, Ingelsten HH, Grönbeck H, Skoglundh M. Regenerable ceria-based SOx traps for sulfur removal in lean exhausts. Appl Catal B: Environ 2008;84:268–76. Available from: http://doi.org/10.1016/j.apcatb.2008.04.003. [160] Flouty R, Abi-Aad E, Siffert S, Aboukaïs A. Formation of cereous sulphate phase upon interaction of SO2 with ceria at room temperature. J Therm Anal Calorim 2003;73:727–34. doi:10.1023/A:1025818127541. [161] Tumuluri U, Rother G, Wu Z. Fundamental understanding of the interaction of acid gases with CeO2: from surface science to practical catalysis. Ind Eng Chem Res 2016;55:3909–19. doi:10.1021/acs.iecr.5b05014. [162] Flytzani-Stephanopoulos M, Zhu T, Li Y. Ceria-based catalysts for the recovery of elemental sulfur from SO2-laden gas streams. Catal Today 2000;62:145–58. Available from: https://doi.org/10.1016/S0920-5861(00)00416-8. [163] Zhang XM, Shen BX, Hou XM, Zhao JG. Research on reactive adsorption desulfurization over metal oxides adsorbent. Energy Sources, Part A: Recovery, Utilization Environ Effects 2015;37:209–16. doi:10.1080/15567036.2011.582602. [164] Yan Q, Nie Y, Yang R, et al. Highly dispersed CuyAlOx mixed oxides as superior low-temperature alkali metal and SO2 resistant NH3-SCR catalysts. Appl Catal A: Gen 2017;538:37–50. Available from: https://doi.org/10.1016/j.apcata.2017.03.021. [165] Sanchez-Cantu M, Perez-Diaz LM, Maubert AM, Valente JS. Dependence of chemical composition of calcined hydrotalcite-like compounds for SOx reduction. Catal Today 2010;150:332–9. Available from: http://doi.org/10.1016/j.cattod.2009.09.010. [166] Zhao L, Li X, Qu Z, et al. The NiAl mixed oxides: the relation between basicity and SO2 removal capacity. Sep Purif Technol 2011;80:345–50. Available from: https://doi. org/10.1016/j.seppur.2011.04.035. [167] Polato CMS, Henriques CA, Neto AA, Monteiro JLF. Synthesis, characterization and evaluation of CeO2/Mg, Al-mixed oxides as catalysts for SOx removal. J Mol Catal A: Chem 2005;241:184–93. [168] Zhao L, Li X, Zhao J. Correlation of structural and chemical characteristics with catalytic performance of hydrotalcite-based CuNiAl mixed oxides for SO2 abatement. Chem Eng J 2013;223:164–71. Available from: https://doi.org/10.1016/j.cej.2013.02.103. [169] Kang HT, Lv K, Yuan SL. Synthesis, characterization, and SO2 removal capacity of MnMgAlFe mixed oxides derived from hydrotalcite-like compounds. Appl Clay Sci 2013;72:184–90. Available from: https://doi.org/10.1016/j.clay.2013.01.015. [170] Cheng WP, Zhao JZ, Yang JG. MgAlFeCu mixed oxides for SO2 removal capacity: Influence of the copper and aluminum incorporation method. Catal Commun 2012;23:1–4. Available from: http://doi.org/10.1016/j.catcom.2012.02.024. [171] Cheng WP, Yang JG, He MY. Evaluation of crystalline structure and SO2 removal capacity of a series of MgAlFeCu mixed oxides as sulfur transfer catalysts. Catal Commun 2009;10:784–7. Available from: http://doi.org/10.1016/j.catcom.2008.11.038. [172] Li Y-X, Jiang W-J, Tan P, et al. What matters to the adsorptive desulfurization performance of metal-organic frameworks? J Phys Chem C 2015;119:21969–77. doi:10.1021/acs.jpcc.5b07546.

978

Desulfurization Materials

[173] Kuppler RJ, Timmons DJ, Fang Q-R, et al. Potential applications of metal-organic frameworks. Coord. Chem Rev 2009;253:3042–66. Available from: https://doi.org/ 10.1016/j.ccr.2009.05.019. [174] Li J-R, Kuppler RJ, Zhou H-C. Selective gas adsorption and separation in metal-organic frameworks. Chem Soc Rev 2009;38:1477–504. doi:10.1039/B802426J. [175] Cychosz KA, Wong-Foy AG, Matzger AJ. Liquid phase adsorption by microporous coordination polymers: removal of organosulfur compounds. J Am Chem Soc 2008;130:6938–9. doi:10.1021/ja802121u. [176] Achmann S, Hagen G, Hämmerle M, et al. Sulfur removal from low-sulfur gasoline and diesel fuel by metal-organic frameworks. Chem Eng Technol 2010;33:275–80. doi:10.1002/ceat.200900426. [177] Blanco-Brieva G, Campos-Martin JM, Al-Zahrani SM, Fierro JLG. Effectiveness of metal–organic frameworks for removal of refractory organo-sulfur compound present in liquid fuels. Fuel 2011;90:190–7. Available from: http://doi.org/10.1016/j.fuel.2010.08.008. [178] Maes M, Trekels M, Boulhout M, et al. Selective removal of N-heterocyclic aromatic contaminants from fuels by Lewis acidic metal–organic frameworks. Angew Chem Int Edn 2011;50:4210–4. doi:10.1002/anie.201100050. [179] Van de Voorde B, Hezinova M, Lannoeye J, et al. Adsorptive desulfurization with CPO-27/MOF-74: an experimental and computational investigation. Phys Chem Chem Phys 2015;17:10759–66. doi:10.1039/C5CP01063B. [180] Li S-L, Lan Y-Q, Sakurai H, Xu Q. Unusual regenerable porous metal–organic framework based on a new triple helical molecular necklace for separating organosulfur compounds. Chem – Eur J 2012;18:16302–9. doi:10.1002/chem.201203093. [181] Khan NA, Jhung SH. Effect of central metal ions of analogous metal-organic frameworks on the adsorptive removal of benzothiophene from a model fuel. J Hazard Mater 2013;260:1050–6. Available from http://doi.org/10.1016/j.jhazmat.2013.06.076. [182] Van de Voorde B, Munn AS, Guillou N, et al. Adsorption of N/S heterocycles in the flexible metal-organic framework MIL-53(FeIII) studied by in situ energy dispersive X-ray diffraction. Phys Chem Chem Phys 2013;15:8606–15. doi:10.1039/C3CP44349C. [183] Van de Voorde B, Boulhout M, Vermoortele F, et al. N/S-heterocyclic contaminant removal from fuels by the mesoporous metal–organic framework MIL-100: the role of the metal ion. J Am Chem Soc 2013;135:9849–56. doi:10.1021/ja403571z. [184] Nuzhdin AL, Kovalenko KA, Dybtsev DN, Bukhtiyarova GA. Removal of nitrogen compounds from liquid hydrocarbon streams by selective sorption on metal-organic framework MIL-101. Mendeleev Commun 2010;20:57–8. Available from: http://dx.doi.org/10.1016/j.mencom.2010.01.022. [185] Ahmed I, Khan NA, Hasan Z, Jhung SH. Adsorptive denitrogenation of model fuels with porous metal-organic framework (MOF) MIL-101 impregnated with phosphotungstic acid: effect of acid site inclusion. J Hazard Mater 2013;250–251:37–44. Available from: http://doi.org/10.1016/j.jhazmat.2013.01.024. [186] Ahmed I, Hasan Z, Khan NA, Jhung SH. Adsorptive denitrogenation of model fuels with porous metal-organic frameworks (MOFs): effect of acidity and basicity of MOFs. Appl Catal B: Environ 2013;129:123–9. Available from: http://doi.org/10.1016/j.apcatb.2012.09.020. [187] Ahmed I, Khan NA, Jhung SH. Graphite oxide/metal–organic framework (MIL-101): remarkable performance in the adsorptive denitrogenation of model fuels. Inorg Chem 2013;52:14155–61. doi:10.1021/ic402012d. [188] Peralta D, Chaplais G, Simon-Masseron A, Barthelet K, Pirngruber GD. Metal–organic framework materials for desulfurization by adsorption. Energy Fuels 2012;26:4953–60. doi:10.1021/ef300762z. [189] Tan P, Xie X-Y, Liu X-Q, et al. Fabrication of magnetically responsive HKUST-1/Fe3O4 composites by dry gel conversion for deep desulfurization and denitrogenation. J Hazard Mater 2017;321:344–52. Available from: http://doi.org/10.1016/j.jhazmat.2016.09.026. [190] Shi R-H, Zhang Z-R, Fan H-L, et al. Cu-based metal–organic framework/activated carbon composites for sulfur compounds removal. Appl Surf Sci 2017;394:394–402. Available from: https://doi.org/10.1016/j.apsusc.2016.10.071. [191] Ahmed I, Jhung SH. Adsorptive desulfurization and denitrogenation using metal-organic frameworks. J Hazard Mater 2016;301:259–76. Available from: http://doi.org/ 10.1016/j.jhazmat.2015.08.045. [192] Dasgupta S, Gupta P, Aarti, et al. Adsorptive desulfurization of diesel by regenerable nickel based adsorbents. Fuel 2013;108:184–9. Available from: https://doi.org/ 10.1016/j.fuel.2012.12.060. [193] Zhang J, Qiu G, Fan L, et al. Enhanced sulfur capacity of durable and regenerable mesoporous sorbents for the deep desulfurization of diesel. Fuel 2015;153:578–84. Available from: https://doi.org/10.1016/j.fuel.2015.03.056. [194] Potter G, Wilson GE, MacLeod N, Lara AC, Lopez YS. Desulfurization materials, Google Patents; 2012. [195] Shiraishi Y, Taki Y, Hirai T, Komasawa I. A novel desulfurization process for fuel oils based on the formation and subsequent precipitation of S-alkylsulfonium salts. 1. light oil feedstocks. Ind Eng Chem Res 2001;40:1213–24. doi:10.1021/ie000547m. [196] Shiraishi Y, Naito T, Hirai T, Komasawa I. Novel desulfurization process for light oils based on the formation and subsequent adsorption of N-tosylsulfimides. Fuel Chem Division Preprint 2002;47:701–2. [197] Shiraishi Y, Naito T, Hirai T. Desulfurization process for light oil based on chemical adsorption of sulfur compounds on polymer-supported imidation agent. J Chem Eng Jpn 2003;36:1528–31. [198] Shiraishi Y, Naito T, Hirai T, Komasawa I. A desulfurization process for light oils based on the formation and subsequent adsorption of N-tosylsulfimides. Ind Eng Chem Res 2002;41:4376–82. doi:10.1021/ie010620o. [199] Fadhel, ZS. Desulfurization of light diesel fuel using Chloramine T and polymer supported imidation agent, Chemical Engineering Department, University of Technology, Iraq; 2010. Available from: http://www.uotechnology.edu.iq/dep-chem-eng/thesis/Zainab%20Subhi%20Thesis.pdf. [200] Kitagawa S, Kitaura R, Noro S-I. Functional porous coordination polymers. Angew Chem Int Edn 2004;43:2334–75. doi:10.1002/anie.200300610. [201] Lee E, Park D-W, Lee J-O, et al. Molecularly imprinted polymers immobilized on carbon nanotube. Colloids Surf A: Physicochem Eng Aspects 2008;313–314:202–6. Available from: https://doi.org/10.1016/j.colsurfa.2007.04.093. [202] Yang Y, Liu X, Guo M, et al. Molecularly imprinted polymer on carbon microsphere surfaces for adsorbing dibenzothiophene. Colloids Surfaces A: Physicochem Eng Aspects 2011;377:379–85. Available from: https://doi.org/10.1016/j.colsurfa.2011.01.032. [203] Lee HY, Kim BS. Grafting of molecularly imprinted polymers on iniferter-modified carbon nanotube. Biosensors Bioelectron 2009;25:587–91. Available from: https://doi. org/10.1016/j.bios.2009.03.040. [204] Nunthaprechachan T, Pengpanich S, Hunsom M. Adsorptive desulfurization of dibenzothiophene by sewage sludge-derived activated carbon. Chem Eng J 2013;228:263–71. Available from: https://doi.org/10.1016/j.cej.2013.04.067. [205] Yuan W, Bandosz TJ. Removal of hydrogen sulfide from biogas on sludge-derived adsorbents. Fuel 2007;86:2736–46. Available from: https://doi.org/10.1016/j. fuel.2007.03.012. [206] Lu GQ, Lau DD. Characterisation of sewage sludge-derived adsorbents for H2S removal. Part 2: Surface and pore structural evolution in chemical activation. Gas Sep Purif 1996;10:103–11. Available from: http://dx.doi.org/10.1016/0950-4214(96)00011-4. [207] Abatzoglou N, Boivin S. A review of biogas purification processes. Biofuels, Bioproducts Biorefining 2009;3:42–71. doi:10.1002/bbb.117. [208] Bagreev A, Bashkova S, Locke DC, Bandosz TJ. Sewage sludge-derived materials as efficient adsorbents for removal of hydrogen sulfide. Environ Sci Technol 2001;35:1537–43. doi:10.1021/es001678h. [209] Ros A, Lillo-Ródenas MA, Canals-Batlle C, et al. A new generation of sludge-based adsorbents for H2S abatement at room temperature. Environ Sci Technol 2007;41:4375–81. doi:10.1021/es062358m. [210] Bandosz TJ, Block K. Effect of pyrolysis temperature and time on catalytic performance of sewage sludge/industrial sludge-based composite adsorbents. Appl Catal B: Environ 2006;67:77–85. Available from: https://doi.org/10.1016/j.apcatb.2006.04.006.

Desulfurization Materials

979

[211] Wallace R, Seredych M, Zhang P, Bandosz TJ. Municipal waste conversion to hydrogen sulfide adsorbents: investigation of the synergistic effects of sewage sludge/fish waste mixture. Chem Eng J 2014;237:88–94. Available from: https://doi.org/10.1016/j.cej.2013.10.005. [212] Ansari A, Bagreev A, Bandosz TJ. Effect of adsorbent composition on H2S removal on sewage sludge-based materials enriched with carbonaceous phase. Carbon 2005;43:1039–48. Available from: https://doi.org/10.1016/j.carbon.2004.11.042. [213] Shen C, Wang Y, Xu J, Luo G. Chitosan supported on porous glass beads as a new green adsorbent for heavy metal recovery. Chem Eng J 2013;229:217–24. Available from: https://doi.org/10.1016/j.cej.2013.06.003. [214] Shen C, Wang YJ, Xu JH, Luo GS. Synthesis of TS-1 on porous glass beads for catalytic oxidative desulfurization. Chem Eng J 2015;259:552–61. Available from: https://doi.org/10.1016/j.cej.2014.08.027. [215] Shen C, Wang YJ, Xu JH, Wang K, Luo GS. Size control and catalytic activity of highly dispersed Pd nanoparticles supported on porous glass beads. Langmuir 2012;28:7519–27. doi:10.1021/la300825s. [216] Shen C, Li Y, Wang YJ, Xu JH, Luo GS. Monodispersed Ni nanoparticles supported on porous glass: composition and size controllable synthesis. Ind Eng Chem Res 2015;54:2910–8. doi:10.1021/ie5047522. [217] Church M, Coniglio M, Hardie L, Longstaffe F, Middleton V. Encyclopedia of sediments and sedimentary rocks. Berlin: Springer; 2005. [218] Vinati A, Mahanty B, Behera SK. Clay and clay minerals for fluoride removal from water: a state-of-the-art review. Appl Clay Sci 2015;114:340–8. Available from: http:// dx.doi.org/10.1016/j.clay.2015.06.013. [219] Lv P, Liu C, Rao Z. Review on clay mineral-based form-stable phase change materials: preparation, characterization and applications. Renew Sustain Energy Rev 2017;68 (Part 1):707–26. Available from: https://doi.org/10.1016/j.rser.2016.10.014. [220] Uddin MK. A review on the adsorption of heavy metals by clay minerals, with special focus on the past decade. Chem Eng J 2017;308:438–62. Available from: https:// doi.org/10.1016/j.cej.2016.09.029. [221] Choi AES, Roces S, Dugos N, Wan M-W. Adsorption of benzothiophene sulfone over clay mineral adsorbents in the frame of oxidative desulfurization. Fuel 2017;205:153–60. Available from: https://doi.org/10.1016/j.fuel.2017.05.070. [222] Ahmad W, Ahmad I, Ishaq M, Ihsan K. Adsorptive desulfurization of kerosene and diesel oil by Zn impregnated montmorollonite clay. Arabian J Chem 2017;10:S3263–9. Available from: http://dx.doi.org/10.1016/j.arabjc.2013.12.025. [223] Shakirullah M, Ahmad W, Ahmad I, Ishaq M, Khan M. Desulphurization of liquid fuels by selective adsorption through mineral clays as adsorbents. J Chil Chem Soc 2012;57:1375–80. [224] Seredych M, Lison J, Jans U, Bandosz TJ. Textural and chemical factors affecting adsorption capacity of activated carbon in highly efficient desulfurization of diesel fuel. Carbon 2009;47:2491–500. Available from: https://doi.org/10.1016/j.carbon.2009.05.001. [225] Lofrano G. Emerging compounds removal from wastewater: natural and solar based treatments. New York, NY: Springer Science & Business Media; 2012. [226] Xiao J, Song C, Ma X, Li Z. Effects of aromatics, diesel additives, nitrogen compounds, and moisture on adsorptive desulfurization of diesel fuel over activated carbon. Ind Eng Chem Res 2012;51:3436–43. doi:10.1021/ie202440t. [227] Lu GQ. Preparation and evaluation of adsorbents from waste carbonaceous materials for SOx and NOx removal. Environ Prog 1996;15:12–8. doi:10.1002/ep.670150114.

Further Reading Babich V, Moulijn JA. Science and technology of novel processes for deep desulfurization of oil refinery streams: a review. Fuel 2003;82:607–31. Do DD. Adsorption analysis: equilibria and kinetics. London: Imperial College Press; 1998. El-Gendy NS, Speight JG. Handbook of refinery desulfurization. Boca Raton, FL: CRC Press; 2015. Hernández‐Maldonado AJ, Yang RT. Desulfurization of transportation fuels by adsorption. Catal Rev 2004;46:111–50. Ishihara A, Wang D, Dumeignil F, et al. Oxidative desulfurization and denitrogenation of a light gas oil using an oxidation/adsorption continuous flow process. Appl Catal A: Gen 2005;279:279–87. Khan NA, Hasan Z, Jhung SH. Adsorptive removal of hazardous materials using metal-organic frameworks (MOFs): a review. J Hazard Mater 2013;244:444–56. Kowsari E. Recent advances in the science and technology of desulfurization of diesel fuel using ionic liquids. In: Kadokawa J, editor. Ionic liquids-new aspects for the future. InTech Open Access Publisher; 2013. Kumar DR, Srivastava VC. Studies on adsorptive desulfurization by activated carbon. Clean – Soil Air Water 2012;40:545–50. Lee SH, Kumar R, Krumpelt M. Sulfur removal from diesel fuel-contaminated methanol. Sep Purif Technol 2002;26:247–58. Mathieu Y, Tzanis L, Soulard M, et al. Adsorption of SOx by oxide materials: a review. Fuel Process Technol 2013;114:81–100. Sharma M, Vyas RK, Singh K. A review on reactive adsorption for potential environmental applications. Adsorption 2013;19:161–88. Song CS. An overview of new approaches to deep desulfurization for ultra-clean gasoline, diesel fuel and jet fuel. Catal Today 2003;86:211–63. Stanislaus A, Marafi A, Rana MS. Recent advances in the science and technology of ultra low sulfur diesel (ULSD) production. Catal Today 2010;153:1–68. Yang RT. Gas Separation by Adsorption Processes. Stoneham: Butterworth-Heinemann; 2013. Yang RT, Hernández-Maldonado AJ, Yang FH. Desulfurization of transportation fuels with zeolites under ambient conditions. Science 2003;301:79–81.

Relevant Websites http://www.bechtel.com/bhts/sulfur/– Bechtel, Sulfur Technology Center. http://hdl.handle.net/1721.1/60646,1983– DSpace@MIT. https://hub.globalccsinstitute.com/publications/coal-quality-impacts-and-gas-quality-control-oxy-fuel-technology-carbon-capture-and-storage-cost-impacts-and-coal-value,2014– Global CCS Institute. http://www.idc-online.com/technical_references/pdfs/chemical_engineering/Flue_Gas_Desulfurization.pdf– IDC Technologies. https://www.rti.org/impact/warm-gas-desulfurization-process-technology,2017– RTI International. https://www.osti.gov/scitech/servlets/purl/7703,1998– SciTech Connect. http://ieaghg.org/docs/General_Docs/Reports/2011-18.pdf,2011– The IEA Greenhouse Gas R&D Programme (IEAGHG), Addressing SO2/SO3/Hg/Corrosion Issues in Oxyfuel Combustion Boiler and Flue Gas Processing Units. http://www.trcsolutions.com/resources/blog/managing-flue-gas-desulfurization-materials-a-brief-physio-chemical-lesson,2016– TRC. https://www.fhwa.dot.gov/publications/research/infrastructure/structures/97148/fgd1.cfm,2016– U.S. Department of Transportation, Federal Highway Administration Research and Technology.

2.30 Novel Building Materials Claudia Fabiani and Anna L Pisello, University of Perugia, Perugia, Italy Halime Paksoy, Çukurova University, Adana, Turkey r 2018 Elsevier Inc. All rights reserved.

2.30.1 Introduction 2.30.2 Principles of Building Physics 2.30.3 Fundamentals 2.30.3.1 Basic Thermal Properties 2.30.3.2 Basic Acoustic Properties 2.30.3.3 Basic Optical Properties 2.30.3.4 Opaque Materials: Required Physical Properties 2.30.3.5 Transparent Materials: Required Physical Properties 2.30.4 Traditional Versus Novel Materials: New Trends 2.30.4.1 Evolution of the Building Sector 2.30.4.2 New Trends of Material Science 2.30.5 Systems and Applications 2.30.5.1 Novel Materials for High Performance Concretes 2.30.5.1.1 Concretes with enhanced mechanical property 2.30.5.1.2 Concretes for environmental sustainability 2.30.5.2 Thermal Insulation Materials 2.30.5.3 Passive Cooling Materials 2.30.5.4 New Materials for Shading Elements 2.30.5.5 Advanced Materials and Coatings for High Performance Glazing 2.30.5.5.1 Aerogel 2.30.5.5.2 Smart materials 2.30.5.5.3 Antireflective coatings 2.30.5.6 Integration of Novel Materials in Building Applications 2.30.5.6.1 Novel materials integration in structural elements 2.30.5.6.2 Novel materials integration in roofs and ceilings 2.30.5.6.3 Novel materials integration in fenestration systems 2.30.6 Assessment Methods 2.30.6.1 Life Cycle Assessment Methodology 2.30.6.2 Energy and Exergy Analysis 2.30.7 Case Studies 2.30.7.1 Existing Buildings Implementing Innovative Materials 2.30.7.2 Prototype Research Applications 2.30.8 Future Directions 2.30.9 Closing Remarks Acknowledgments References Further Reading Relevant Websites

Abbreviations ECMs GC-DGU GCMs GSAs HSC LCA LCIA LCs

980

Electrochromic materials Gasochromic double glazing unit Gasochromic materials Granular silica aerogel High strength concrete Life cycle assessment Life cycle impact assessment Liquid crystals

MSA NIR PCM SMAs SMHs SMMs SMPs SPs TCMs

981 982 982 983 984 985 987 987 988 988 988 988 988 988 990 994 996 997 998 998 998 1000 1000 1001 1002 1003 1005 1005 1007 1008 1008 1009 1013 1013 1013 1013 1017 1017

Monolithic silica aerogel Near infrared Phase change materials Shape memory alloys Shape memory hybrids Shape memory materials Shape memory polymers Suspended particles Thermochromic materials

Comprehensive Energy Systems, Volume 2

doi:10.1016/B978-0-12-809597-3.00257-1

Novel Building Materials

Subscripts a e i l p

r t v 0 1 2

Absorbed Electron Incident Lattice Constant pressure

981

Reflected Transmitted Constant volume Initial/Static Medium 1 Medium 2

Symbol

Quantity

International system of unit

a c C E f I k L P Q r t T TL V W

Sound absorption coefficient Speed of sound Heat capacity Emissive power Frequency Intensity of radiation Thermal conductivity Latent heat Pressure Amount of heat Sound reflection coefficient Sound transmission coefficient Temperature Sound transmission loss Volume Power

– m/s J/K W/m2 Hz W/sr W/mK J/g Pa J – – K dB m3 W

Greek letter

Quantity

International system of unit

a e Z y yD l j r

Thermal expansion coefficient Emissivity Refractive index Plane angle Debye temperature Wavelength Plane angle Density

K 1 – – rad K m rad kg/m3

2.30.1

Introduction

Buildings are currently consuming 35% of the earth’s energy. Regional climates play a large role in the energy consumption behaviors of buildings. A considerable part of this goes to space heating, which is about 60% for buildings in cold climates and 43% for those in moderate and/or warm climates [1]. Fossil fuels have still the highest share with 66.7% among all the energy sources used globally [2]. This fossil fuel dominant picture increases CO2 emissions and buildings are the major contributor to this share. Reducing energy consumption and increasing efficiency of energy use has become the most important target among the efforts of mitigating CO2 emissions to combat climate change. The recent severe results of climate change seen around the world necessitate rapid actions. Energy consumption of buildings is affected by the changes in building stock. Growth of building stock is directly related to the population growth. In countries where population growth is nearly zero, construction of new buildings is on minor level. The renovation of existing buildings to increase energy savings becomes important in those countries. In countries where population and/or standard of living is increasing demand for new housing also rapidly increases. Population migrating to metropolitan areas also increases demand for new buildings. Fifty-four percent of the world’s population lives in cities. This level is expected to increase to 66% by 2050 (3n). Addressing the energy conservation issues in buildings may differ based on the building being new or existing.

982

Novel Building Materials

The buildings have to meet thermal comfort and indoor environmental quality expectations of end-users. In addition, energy sources used for delivering services like heating and cooling in buildings should be economic and environmentally friendly. These issues are accounted in building energy and environmental performance. This performance can be improved by more integration of renewables and using new materials and technologies to increase energy efficiency. These improvements require innovative research, new policies, and social awareness of end-users and technical people. The recent IEA Technology Roadmap on energy efficient building envelopes shows that improvements in building envelopes can reduce the building sector’s total consumption by a factor of 20% [3]. Choosing the right building materials not only effects the energy consumption, but also sustainability of the building. The architects or builders often make this choice and end-user is not usually involved. Considering the long service life of buildings, this decision has a long-term effect on the energy consumption of buildings. Renovation that can be made afterwards will require additional investment, consumption, and demolition of excessive amount of materials. This chapter gives an overview on basic principles of building physics and new trends on novel building materials. The objective is to give properties of these building materials and how they can be applied in buildings for energy savings. After an introduction to principles of building physics and fundamental properties of building materials, traditional versus new trends are introduced. Systems and Applications section is about the new building materials and their integration methods in buildings. In the following section, assessment methods are covered. Some case studies from monitored examples are given in the last section. The chapter ends with recommendations on Sections 2.30.8 and 2.30.9.

2.30.2

Principles of Building Physics

Understanding building physics and using the correct building material are important for increasing the number of sustainable buildings and decreasing global energy consumption. Current demanding problems of climate change make the need for sustainable buildings more urgent than ever. For realization of sustainable buildings, a comprehensive understanding of buildings covering different fields of expertize is essential. Building physics is a relatively new discipline that originated to fulfill this demand. It is a multidisciplinary field bringing building services engineering, applied physics and building construction engineering together. The ultimate aim consists of designing and constructing high performance buildings that are functional and comfortable for the occupants, at the same time use resources efficiently and minimize the environmental impacts of their construction and operation [4]. The principal aspects of building engineering physics include topics on [5]:

• • • • • • • •

thermal performance acoustics air movement climate construction technology building services control of moisture lighting The benefits of building engineering physics include [5]:

• • • •

Capital cost reduction: better design decisions and reduced design fees. Operating cost reduction: energy efficiency, resulting in lower energy bills and lower exposure to energy price rises. Creative design focused on real life building performance rather than compliance. Occupant satisfaction: high performance buildings can result in better productivity and comfort of the occupants.

Diverse materials are needed to meet the requirements of the aspects of building physics. Together with the choice of material the processing techniques to provide the desired property has to be considered. Some of these properties can be interrelated and may require detailed analysis tools and approaches. Among these are:

• • • • •

thermal dynamic simulation computational fluid dynamics (CFD) energy analysis to assess carbon reduction strategies, feasibility of renewable technologies, and operational performance daylighting and glare analysis solar shading analysis

2.30.3

Fundamentals

Building envelope can be described as the set of building components creating a physical boundary between the outdoor and the indoor conditioned environment. Such physical boundary is characterized by a threefold function: firstly, it has to support the

Novel Building Materials

983

building structure in resisting and transferring structural loads; secondly, it must control mass and energy flows through the building; and finally, as a finishing element, it needs to meet human expectations and desires on both the inside and the outside. In order to properly respond to all these requirements, besides the purely architectural design, building envelope must also be optimized in terms of material characteristics. This is particularly true when the control function is taken into account. Mass and energy flows are indeed affected by the intrinsic nature of a component and can be properly controlled from a material science perspective. In this section, the fundamental physical properties that must be considered when dealing with material science and engineering related to building envelope are investigated and their proper application to both opaque and transparent systems is presented.

2.30.3.1

Basic Thermal Properties

Thermal properties refer to the interaction of a material with thermal energy. At a microscopic scale, thermal energy is the expression of the vibrational energy possessed by the atoms: the higher the temperature, the higher the thermal agitation of the atoms. Whenever heat is applied to a solid material, thermal energy is absorbed by it, temperature rises and dimensions increase. Furthermore, according to the second thermodynamics principle, if a temperature gradient exists inside a material, the energy will be spontaneously transported to cooler regions inside the specimen, and ultimately, the specimen may melt. These are the main thermal properties that are often critical for solid materials: heat capacity, thermal conductivity and thermal expansion. Heat capacity (referred to as C, measured in J/K) is defined as the amount of energy that needs to be supplied to a given material in order to produce a unit temperature rise within it. It can be expressed by the following relationship: C¼

dQ dT

ð1Þ

where dQ is the amount of energy required to produce a dT temperature change. Measuring the heat capacity of a material is not a trivial procedure since such property is a function of temperature, volume, and pressure. In order to properly define its value, the heat capacity is measured either by (1) maintaining the specimen volume constant (Cv), or (2) keeping a constant external pressure (Cp). The difference between Cv and Cp is very slight for most solids at room temperature, however, the heat capacity at constant volume is most usually preferred since keeping a constant volume during an experimental procedure is generally easier than maintaining a fixed pressure. The main contribution to heat capacity is the vibrational contribution, represented in Fig. 1 at constant volume for many crystalline solids. Cv is zero at 0K, but rapidly raises with increasing temperature until the Debye temperature (yD) is reached. Above such value, Cv becomes essentially independent of temperature and reaches the value of about 3R (with R being the gas constant). Latent heat (L measured in J/g) is the general term used for energy released or absorbed by a material at constant temperature during phase change. For solid–liquid phase change it is called latent heat of fusion and that for liquid–gas is heat of vaporization. Latent heat can also be associated with transition between different crystalline forms of the same material. Latent heat arises from the energy needed to overcome the molecular forces that hold the atoms/molecules together during a phase change. The value of latent heat is positive for endothermic changes like melting and evaporation and negative for exothermic changes like freezing and condensation. Thermal conductivity (k measured in W/m/K) is the physical property that characterizes heat conduction and represents the amount of heat per unit time and per unit area that can be conducted through a plate of unit thickness of a given material (1 m), the faces of the plate differing by one unit of temperature (1K). As stated in Eq. (2), k is associated with two basic heat transfer

Heat capacity, Cv

3R

0

0

D Temperature (K)

Fig. 1 Temperature dependence of the heat capacity at constant volume; yD is the Debye temperature.

984

Novel Building Materials

phenomena: lattice vibration waves (phonons) and free electrons k ¼ kl þ ke

ð2Þ

where kl and ke represent the lattice vibration and the electron thermal conductivity, respectively. The first contribution results from a net movement of phonons, while the second one from the kinetic energy of free electrons. In homogeneous materials, i.e., of uniform composition in all directions, the thermal conductivity is a property of the material itself. Generally, it is a function of temperature, pressure, and chemical composition. In anisotropic materials, on the other hand, for example, oriented molecular crystals, the thermal conductivity can be a function of the flux direction. Lastly, in composite materials an equivalent thermal conductivity is generally defined, which can even be a function of the history of the material, of the processing undergone by it, or of its superficial characteristics. The last thermal property considered in this section is the thermal expansion, generally described in terms of thermal expansion coefficient (a). Upon heating or cooling, most solid materials experience a change in dimension, i.e., expansion or contraction, respectively. Such change in dimension affects the body in all directions causing an overall change in volume, computed by: DV ¼ av DT V0

ð3Þ

where DV is the volume change, V0 is the initial volume, and av is the volume coefficient of thermal expansion. The increase in dimensions generally occurring upon heating is the macroscopic effect of the increase in the average distances between atoms. More in details, the macroscopic expansion is due to the asymmetric curvature of the energy-versus-interatomic spacing curve for a solid material (see Fig. 2). When a symmetric potential energy curve is considered, in fact, no net change in the interatomic distance can be found.

2.30.3.2

Basic Acoustic Properties

Sound is a mechanical vibration or oscillation of a substrate produced through the sympathetic vibration of an elastic medium. Its presence induces a perturbation of pressure inside a medium, in addition to its original static pressure (P0). If a sound wave impinges on an obstacle which is large compared to its wavelength, it is partially reflected (Wr) (and possibly diffracted and scattered), partially allowed to pass through (Wt), partially transmitted as structure-borne sound (Wf), as well as absorbed (Wa) (Fig. 3): Wi ¼ Wr þ Wt þ Wf þ Wa

ð4Þ

The part of sound wave which is transferred to the material enters its pores and a certain amount of its energy is converted into heat because of the friction and the viscosity resistance between the air molecules and the wall of pores. In this way, the sound energy is absorbed. The effectiveness of a sound absorber is quantified by the absorption coefficient Eq. (5), which defines the part of acoustical energy of the incident wave that is actually absorbed by the material.

Potential energy

a¼

Wa Wi

Symmetric rs ra2→ra7

ð5Þ

Interatomic distance (r) Asimmetric

E7 E6 E5 E4 E3 E2 E1

r

Fig. 2 Potential energy vs. interatomic distance showing the increase in interatomic distance with potential energy for the asymmetric curve (in red).

Novel Building Materials

985

Wr Wi

Wf

Wr Wi

Wt

Wa

Wr

Wf

Fig. 3 Sound wave impinging on a solid surface.

Together with the absorption coefficient, it is also possible to define two other acoustic parameters: the reflection (r) and the transmission coefficient (t): r¼

Wr Wt ; t¼ Wi Wi

ð6Þ

These parameters can only assume values between 0 and 1 and their cumulate value always equals 1. In particular, the transmission coefficient t, defines the percentage of an impinging sound wave intensity, which truly passes through the material and it is a function of different values. First of all, it depends on the incidence angle (f) between the wave and the material surface, the wave frequency (f), the sound speed (c) and the material density (r). tðf; f ; c; rÞ ¼

Wt ðf; f ; c; rÞ Wi ðf; f ; c; rÞ

ð7Þ

This physical property, however, is not very used in acoustic, because of the typical nonlinearity of sound phenomena. Transmission loss (TL) measured in decibels (dB), on the other hand, is a commonly used factor for the quantification of the insulation properties of acoustic materials. It is defined at a logarithmic scale and is related to the sound transmission coefficient (t) by the law presented in Eq. (8):   1 TLðf; f ; c; rÞ ¼ 10log ð8Þ tðf; f ; c; rÞ Acoustic absorption, quantified by the absorption coefficient and acoustic insulation, quantified by the TL, are two key properties for acoustic materials, and are the ones that need to be considered in order to properly design building components.

2.30.3.3

Basic Optical Properties

Optical properties of a material refer to its way of interacting with the electromagnetic radiation, and in particular with visible light and infrared radiation. In general, the optical performance of a material can be described by means of five optical properties: refractivity, emissivity, absorptivity, reflectivity, and transmissivity. When light moves from one medium to another, it faces a physical phenomenon called refraction, consisting of a change in direction of the light ray at the interface between the two media. More in detail, if the light ray moves from a medium with refractive index Z1 to one with refractive index Z2, with an incidence angle to the surface normal equal to y1, the final refraction angle y2 can be calculated from Snell's law as: Z1 cosy1 ¼ Z2 cosy2

ð9Þ

When light enters a material with higher refractive index, the angle of refraction will be smaller than the angle of incidence and the light will be refracted toward the normal of the surface (see Fig. 4). The higher the refractive index, the closer to the normal direction the light will travel. When passing into a medium with lower refractive index, the light will instead be refracted away from the normal, toward the surface. Emissivity, absorptivity, reflectivity, and transmissivity are used to describe the radiative behavior of real materials, which will certainly emit and absorb less than a theoretical black body. The emissivity of a real body surface is then described as the ratio between the radiative energy emitted by the considered surface and the radiative energy corresponding to a black body surface at the same temperature. Beside temperature, emissivity is also a function of wavelength and direction, so it is possible to define a spectral directional emissivity (el,W) of a surface at a given temperature as: el;W ðl; W; j; TÞ ¼

Il;e ðl; W; j; T Þ Il;b ðl; T Þ

ð10Þ

where Il,e(l,W,j,T) is the spectral intensity of the considered surface and Il,b(l,T) the spectral intensity of the black body at the given temperature T.

986

Novel Building Materials

Refractive index 1

2 2

Normal

Incident wave

Interface

1

Fig. 4 Refraction of light at the interface between two media of different refractive indices, with Z24Z1.

 Blackbody, T Real surface, T Real surface I (,,T )= , I,b (,T )

E,b (,T ) E

Blackbody, I,b

E (,T ) = E,b (,T )  (A)

(B)

Fig. 5 (A) Spectral directional emissivity of a gray body. (B) Spectral hemispherical emissivity of a gray body.

It is also possible to introduce the spectral hemispherical emissivity (el) (defined as a directional integration of the previous property), which is only a function of the considered radiation wavelength and surface temperature: el ðl; TÞ ¼

El ðl; T Þ El;b ðl; T Þ

ð11Þ

Of course, as it can be seen in Fig. 5, a common gray body is always associated to a lower and less uniform spectral directional emissivity and spectral hemispherical emissivity value, if compared to a black body. ðTÞ ¼

EðT Þ Eb ðT Þ

In technical applications, however, the total hemispherical emissivity (e), defined by integrating el on the overall spectrum of the considered radiation, is generally considered: eðTÞ ¼

E ðT Þ Eb ðT Þ

ð12Þ

A given material does not only emits radiative energy as a function of its temperature and superficial and bulk properties: it also reacts to the radiative energy impinging on it. In fact, when a semitransparent body is exposed to a given intensity of monochromatic, directional radiation (Il,i(l,W,j)), some of this radiation will be reflected (Il,r), some will be absorbed (Il,a), and finally, a part of it will be transmitted through the material. By considering the body as the control volume and the system to be stationary, it is possible to apply the energy conservation principle: Il;i ¼ Il;r þ Il;a þ Il;t

ð13Þ

and dividing both terms of Eq. (12) by Il,i the following equation is obtained: rl;W ðl; W; jÞ þ al;W ðl; W; jÞ þ tl;W ðl; W; jÞ ¼ 1

ð14Þ

where rl,W (l,W, j), al,W (l,W, j), and tl,W (l,W, j) are the spectral directional reflection, absorption, and transmission coefficients, respectively, which can be transformed in total hemispherical reflection (r), absorption (a), and transmission (t) coefficients, by integrating on the overall spectrum and directions.

Novel Building Materials 2.30.3.4

987

Opaque Materials: Required Physical Properties

Depending on their way of interaction with light, materials can be classified in transparent, translucent, and opaque media. More in details, materials that are capable of transmitting a huge percentage of the impinging light wave with little or no absorption or reflection are called transparent materials. Those that are capable of transmitting light in a diffuse way are called translucent materials, while opaque materials can be said to be impervious to the transmission of visible light. Opaque, transparent, and translucent materials are used in different building components, and consequently need to guarantee different physical properties. Opaque materials are used in order to produce both internal partitions and opaque building envelope components. From a thermal point of view, building envelope and internal partitions components need to guarantee similar properties. First of all, thermal expansion must be relatively low, since it could produce the development of internal strains in the building structure, thus modifying its mechanical response to structural loads and producing serious threats to the overall stability of the building. In general, values in the order of 10 5/K of the volumetric thermal expansion coefficient (at 201C) are considered and used for the most common structural materials. Furthermore, in both building envelope and internal partitions, materials are indeed selected in order to guarantee the minimum heat transfer between the outdoor and the indoor environment, but also between adjacent indoor environments. In this context, materials with high-specific heat and low thermal conductivity values are, generally, selected and implemented. A good thermal insulation material for opaque components, for example, is characterized by Cp values higher than 1000 J/kg K, and l values at least below 0.06 W/mK. Successful noise control and acoustic comfort of the indoor environment, heavily rely on the implementation of efficient sound absorbers and insulation systems, covering a broad range of frequencies. Materials with high density are generally used in order to maximize the insulation capability of the building envelope, particularly on the outdoor finishing. Common TL values for dense materials, such as concrete or brick walls, vary in the range 30–60 dB, depending on the specific frequency and thickness of the material. Lightweight media with a huge number of interconnected pores, on the other hand, are used to absorb acoustic energy in the indoor environment. Acoustic absorbers can be bulk materials, like porous or fibrous absorbers, or geometrical systems, such as reverberating or resonant panels. In both cases, a good acoustic absorber is associated with a values equal or above 0.8 over most sound frequencies. Lastly, from an optical point of view, opaque materials are by definition characterized by negligible transmission coefficient, but can be optimized in terms of thermal emissivity and reflection coefficient. Such optimization is particularly important for outdoor surfaces directly exposed to solar radiation, and its study developed a fundamental class of materials for passive cooling techniques, such as cool materials. Cool materials are characterized by high solar reflectance values, i.e., higher than 0.65 and also high thermal emittance values, i.e., around 0.80–0.90. Building envelopes components having such optical properties are able to reduce the amount of solar energy absorbed by the material, causing a drastic surface temperature decrease, and consequently an appreciable reduction in the bordering air temperature. Common materials, on the other hand, are generally associated to similar e values but to solar reflectance values even lower than 0.2.

2.30.3.5

Transparent Materials: Required Physical Properties

Transparent materials are capable of transmitting an impinging light wave with little or no absorption or reflection. These materials are generally used to produce building envelope components, such as windows, or more generally fenestration systems. Although such building components generally only constitute a small part of the building surface, they highly contribute to the overall heat transfer and sound transmission between the indoor and the outdoor environment. Furthermore, glazing systems also play a significant role in low-energy buildings using thermal gains from solar radiation as a passive heating technique, and the solar radiation itself in order to reduce lighting energy demand. All things considered, glazing systems need to fulfill several requirements. First of all, they must guarantee the correct amount of light entering the indoor environment, secondly, they must prevent heat to easily transfer through the building envelope, while still ensuring solar heating phenomena for passive heating purposes, and lastly, they must also assure adequate insulation in terms of sound TL. In order to fulfill all these requirements, glazing units are generally constituted of two layers of glass, separated by an air cavity and sealed together as a single component. The air within the cavity of glass units is often replaced by insulating gas as argon, krypton, or xenon in order to reduce heat losses. The air cavity of glass units can also be filled with special gels, such as aerogels, for increasing the thermal insulation capability of the component and/or the fire resistance properties or the protection from glare effect in the case of large-glazed cladding areas. The double glazing system can anyhow also produce a significant decrease on the acoustic TL of these components. This is due to the air- or gel-filled gap between the two layers of glass acting like a spring and transferring vibrational energy from one layer to the other (mass-air-mass resonance). Such effect can be reduced by increasing the gap dimension [6,7]. From an optical point of view, low-emissivity (low-e) glazing can also be produced by applying thin metal coatings on the glass. Such transparent metallic coatings allow transmission of visible light through the glass pane but they also significantly reduce radiation losses. The mean emissivity of the ordinary glass is about 0.83–0.84 [6] and the average reflectance reaches values in the range 0.16–0.17. Low-e glazing, on the other hand, generally guarantee very high visible transmittance and infrared reflectance r(l)-1 in the wavelength range from 5 to 50 mm [8].

988

Novel Building Materials

2.30.4

Traditional Versus Novel Materials: New Trends

2.30.4.1

Evolution of the Building Sector

Evolution of building design in history shows that early human beings were smarter in finding a shelter that will protect them from the environment in the most efficient way. The use of natural material, which was the only choice at that time were used to adjust to the climate conditions. The vernacular building design evolved to represent the most efficient response to the climate given the local availability of resources. The adobe beehive houses that are still used in Harran, Southeast of Turkey have a dome shape that allows natural ventilation [9]. The excavations in Mesopotamia show that domed building forms were used as early as the 7th century BC [10]. Such domeshaped buildings are also seen in Apulia region in Italy today. The use of under floor thermal labyrinths to store thermal energy is also a heritage from ancient times. This structure commonly seen in thermal baths of Roman cities were used at least 2000 years ago. With the developments in the manufacturing industry brought together many new building materials. These developments together with cheap and abundant fossil fuels lead to the uncontrollable growth of the building sector in an unsustainable way. Some of the reasons can be given as:

• • • •

new building designs with more esthetic concern and less emphasis on energy use increasing population and low-cost housing demand – use of cheap and non-efficient materials increasing urbanization – urban heat island effects increasing living standards and consumer behavior

Today’s buildings are expected to meet the highest quality of life, while preserving environment during its life-time. This requires new way of thinking and approach to building design, construction, and use.

2.30.4.2

New Trends of Material Science

Decreasing “carbon foot print” became an indicator of moving to low carbon technologies in building sector also. In the evaluation of this indicator, not only reduction of CO2 emissions from direct use of energy is considered, but also avoided embodied energy plays an important role. In this approach, energy consumption of all the processes involved in bringing the material to the final user is taken into account. Cement, iron, and steel industries are among the most energy intensive industries. Reinforced concrete that is the most widely used building material has the highest amount of embodied energy. The new trend in building materials is trying to go back to basics. Using natural materials, building with less material as much as possible and imitating nature are becoming important. The choice of building material should be done to optimize the building performance over its life-time. The following topics have become important in the new trends of material science and building engineers: 1. sustainability including economic, environmental, and societal respects; 2. decreasing embodied energy by increasing energy efficiency in buildings, as well as the whole value chain of building material production; 3. using recycled material as building material or additive to building materials; 4. life cycle assessment (LCA) examining the total environmental impact of a material; 5. developing low-energy retrofit building materials for existing building stock; 6. use of multifunctional materials serving thermal, optic, hygienic, and acoustic purposes; 7. internet of things (IoTs) applications in smart buildings, which may require complex combinations of telecommunications, electronics, informatics, and human behavior; and 8. maximizing renewable energy use in buildings with an effort to reach zero energy or buildings.

2.30.5

Systems and Applications

2.30.5.1

Novel Materials for High Performance Concretes

Concrete is the most widespread construction material worldwide. It is used for buildings, infrastructural systems, geotechnical works, road pavements, and more. In fact, it was calculated that about 1 t of concrete is produced every year for every person in the world [11]. Therefore, the development of novel concrete mixture plays a fundamental role in the material science panorama. Two main research fields can be analyzed regarding concrete materials:

• •

concretes with enhanced mechanical properties; concretes for environmental sustainability.

2.30.5.1.1

Concretes with enhanced mechanical property

The development of concretes with high mechanical properties mainly focuses on the study of high strength concretes (HSCs), higher durability concretes, enhanced fire performance concretes, self-healing concretes, and self-sensing concretes.

Novel Building Materials

989

HSCs allow to reduce the cross-section of the structural members of a building and the overall volume of the structure [12]. The increase in mechanical strength of such material can be attributed to the use of higher quantities of cement, and consequently produces higher CO2 emissions in the atmosphere. Anyway, the introduction of several industrial by-products, such as fly ash, as potential substitutes for cement in HSCs is currently being investigated in order to avoid its major environmental drawbacks [13]. HSCs can claim a massive application in the main load-bearing components of innovative buildings. They are widely used because of their enhanced mechanical characteristics in common conditions, but a vast body of literature demonstrated a huge decrease of such properties under fire conditions [14–17]. This vast reduction in the mechanical resistance in HSC is basically due to the occurrence of the spalling effect, which can be considered as the combination of two different phenomena: the thermosmechanical and the thermo-hydral process [18]. Both these phenomena are a consequence of the high permeability reduction, which characterizes HSCs. The first process concerns the development of thermal dilatation/shrinkage gradients within the material. The second concerns the development of high pressure fields in the porous cementitious matrix by water vapor and enclosed air [14–21]. Fig. 6 shows the development of the spalling effect, within the cementitious matrix of a concrete member exposed to fire. As it can be seen, condensed water accumulates in a fully saturated layer (“the moisture clog”) in response to the huge amount of heat transferred to the composite during fire exposition. Consequently, large pore pressures and thermal stresses are located close to the saturated layer, eventually causing the failure of the fire-exposed building component. In order to prevent or at least reduce the spalling effect in HSC, several factors need to be taken into account at the material scale, the most important are: moisture content, material density, and presence of silica fumes. In general, in fact, higher moisture content and lower density are associated to higher vapor pressure and spalling. In order to improve the material response to fire, the most appreciated solution is to include in the cementitious matrix small percentages of polypropylene fibers, nylon fibers, and steel fibers which were shown to be able to increase the tensile strength of the mixtures, while also providing free channels for steam and high pressure air after melting [21–24]. Durability issues also represent a crucial factor concerning real applications in buildings. Indeed, the exposition to severe environmental boundary conditions vigorously affects the durability of the structural material [25,26]. Consequently, the subject of durability is of great importance both in terms of structural safety and environmental sustainability of constructions affected by life cycle maintenance and repair activities. Two main phenomena critically influence the performance of concrete: carbonationinduced reinforcement corrosion and freeze–thaw cycles. In this context, Limbachiya et al. showed that the finer cementitious components used in binary and ternary systems in concrete production enhance the material density, resulting in improved durability performance [26]. Furthermore, the use of alkali-activated binders (aaB) with pozzolanic materials was also found to attain good mechanical properties at early ages of curing [27].

Nearly saturated porous medium

Heat flow into porous medium Vapor migration out of medium

Vapor migration towards lower pressure region (A)

Vaporization of capillary water

Accumulation of condensed water vapor

(B)

Vaporization zone

(D)

Spalling due to (1) large pore pressures and (2) Thermal stresses

Vapor cannot pass the saturated layer Build-up of large pore pressure at saturated layer

(C)

‘Dry’ zone ‘Moisture clog’ (fully saturated layer)

Fig. 6 Illustration of mechanism of spalling of concrete as a result of fire loading. Reproduced from Zeiml M, Leithner D, Lackner R, Mang HA. How do polypropylene fibers improve the spalling behavior of in-situ concrete? Cem Concr Res 2006;36:929–42, with permission from Elsevier Ltd.

990

Novel Building Materials

Lastly, excellent results were obtained in terms of mechanical strength, resistance to chloride-ion penetration and resistance to freeze–thaw cycles by using fly ash doped concretes [28]. A more pioneering research in terms of durability is associated to self-healing mixtures. Self-healing is the capacity of a material to automatically and autonomously heal (recover/repair) damages when they happen. The most suitable technique for concrete maintenance appears to be the use of bacterial-based self-healing [29]. In this context, Wang et al. investigated the performance of encapsulated bacterial spores for self-healing concrete applications and found that water permeability of the novel bio-doped concrete was 10 times lower than that of the common one [30]. Fig. 7 shows the phenomenon of crack healing (photographical images of the specimens before and after healing) for five different concrete mixes:

• • • • •

No.1, No.4, No.5, No.8, No.9,

without healing agent, healing in water; without healing agent, healing in L-Ca media (calcium salt) with bacteria; without healing agent, healing in G-Ca media (calcium salt) with bacteria; with L-Ca and bacteria as healing agents, healing in water; with G-Ca and bacteria as healing agents, healing in water.

No obvious healing occurred in the control specimen (sample No.1), while in cases of external treatment by nutritional calcium source media (L-Ca and G-Ca) with bacteria, the cracks were sealed and a layer of precipitates due to external applied healing covered the surface of the specimen. The same authors also investigated the effect of hydrogel-encapsulated spores on the self-healing capability of cement-based samples. Also in this case water permeability showed a huge percentage decrease (68%), furthermore, the bio-doped mixture was able to heal cracks till the maximum width of about 5 mm [31]. Another class of interesting novel cementitious mixtures is the self-sensing category. Self-sensing composites are able to detect variations of applied loads or displacements, but also temperature and damages without the use of common external transducers [32]. The nano-doped cementitious matrix itself indeed behaves as a sensor. More in details, the piezoelectric fillers dispersed in the engineered mixture allow the concrete to obtain a self-monitoring capability and convert every change of external strain and stresses into a change in the overall electrical properties of the material. Carbon based, metallic, polymeric, or even hybrid materials can be used as self-sensing fillers being also capable of enhancing concrete density and consequently its mechanical properties. Two are the main features affecting the monitoring capability of such mixtures: filler concentration and dispersion method [33]. The correct functioning of the self-sensing material is, in fact, strictly correlated to the establishment of an optimally conductive network with the proper dispersion and distance between fibers. The presence of concentration gradients, for example, could produce an electric potential well within the matrix and a consequent loss of signal.

2.30.5.1.2

Concretes for environmental sustainability

Given the widespread use of concrete all over the world, it is clear that a proper engineering of such material could help in reducing the environmental impact of the construction sector. This reduction can be attained in a threefold way. First of all, concrete can be used to actively reduce the percentage of organic and inorganic pollutants in urban areas by using smart additives with photocatalytic properties. Secondly, the inner environmental impact of concrete as a construction material can be reduced by lowering its embodied energy during the production phase, which can be achieved by substituting some of its classical components with recycled materials. Lastly, novel concretes can be produced with enhanced thermal properties, in order to reduce the energy need of buildings. The first environmentally sustainable concretes category is that of concretes with photocatalytic capability. Photochemical catalysis together with super-hydrophilicity and photo-electrochemistry is a physical phenomenon occurring when a semiconductor is exposed to an electromagnetic radiation with an energy content equal or higher to that of its band gap [34]. The interaction between the semiconductor and the electromagnetic wave produces an electron–hole pair, which can be used to produce electric energy (photo-electrochemistry), produce an effective change on the material surface (super-hydrophilicity), or initiate chemical reactions (photochemical catalysis). In photochemical catalysis, once the electron–hole pairs are generated, they interact with molecular oxygen (O2) and water (H2O), producing two highly reactive radicals: superoxide radical anions (O2 ) and hydroxyl radicals (OH ), respectively. Such radicals can be used in order to remove organic and inorganic pollutants produced by anthropogenic sources and saturating urban areas. In this view, various types of sulfides and oxides have been studied and integrated in cementitious matrixes CdS, ZnS, TiO2, ZnO, CeO2, etc.) in order to improve air quality in urban areas, but the best performance are generally associated with titanium dioxide molecules (TiO2) [35]. In particular, the production of TiO2doped concretes was shown to produce positive effects in terms of self-cleaning properties (associated with super-hydrophilicity features), and indoor and outdoor air pollution concentration reduction (given by the photocatalytic effect). Titanium dioxide was effectively used in concrete products both as additive [36–38] and as a superficial coating layer [39–41]. Most of the experimental analyses conducted on such materials showed a negative effect of relative humidity on the reaction effectiveness and positive correlations with nitrogen oxides concentration and irradiance intensity. Fig. 8(A) shows the NO concentration reduction caused by the activation of the photocatalytic reaction in a TiO2-additivated double-layer paving stone and the dependency of the reaction on the UV-A irradiance in case of an outside monitoring experiment conducted on a rainy day, using a reactor cell. Fig. 8(B), on the other hand, presents the influence of the pollutant concentration on the NO degradation on the same kind of paving blocks. In particular, from Fig. 8(B) it is clear that the increase in the inlet concentration results in lower degradation rates, whereas, lower pollutants concentrations are associated to increasing final performance.

Novel Building Materials

(A) No. 1, before healing 5 mm

(C) No. 4, before healing 5 mm

(B) No. 1, after healing 5 mm

(D) No. 4, after healing 5 mm

(I) No. 9, before healing 5 mm

(E) No. 5, before healing 5 mm

(G) No. 8, before healing 5 mm

991

(F) No. 5, after healing 5 mm

(H) No. 8, after healing 5 mm

(J) No. 9, after healing 5 mm

Fig. 7 Photographical images of crack healing process. (A) No. 1, before healing. (B) No. 1, after healing. (C) No. 4, before healing. (D) No. 4, after healing. (E) No. 5, before healing. (F) No. 5, after healing. (G) No. 8, before healing. (H) No. 8, after healing. (I) No. 9, before healing. (J) No. 9, after healing. Reproduced from Xu J, Yao W. Multiscale mechanical quantification of self-healing concrete incorporating non-ureolytic bacteriabased healing agent. Cem Concr Res 2014;64:1–10, with permission from Elsevier Ltd.

Beside active pollutants reduction, it is possible to develop highly sustainable concretes by using recycled materials as alternative components of aggregates, sand, or cement. Such strategy can be considered as a proper way to both reduce the extensive depletion of natural resources associated with concrete-manufacturing, and positively deal with the production of industrial and agricultural wastes, which are more and more becoming a huge concern in an environmental perspective. In this panorama, waste materials to be integrated in cementitious matrixes can be classified in two main categories: by-products, i.e., secondary products derived from a manufacturing process or a chemical reaction, and properly said waste materials, i.e., unwanted or unusable materials discarded after primary use [34]. The use of by-products in the production of concretes dates back to 1970s, when silica fumes, or super-pozzolans where firstly substituted to cement in the production of HSCs up to 9–15% in mass content [42]. Fly ash, from coal combustion processes and

Novel Building Materials

992

1.2

18

1.0

15

0.8

12

0.6

9

0.4

6

100

0 (A)

Concentration NO UV-A irradiance

0.2

0

5

10

15

20 25 Time (min)

3

80 Degradation rate (%)

UV-A irradiance (W m−2)

Concentration NO (ppmv)

90

70

68.4

66.6

60 50 40

47.6 38.3

36.9

30

36.9 22.1

20 10

30

35

40

0 45

0 (B)

0.1 0.3 0.5 1.0 Concentration NO (ppmv)

1.0 3.0 5.0 Flow rate (L/min)

Fig. 8 (A) Results of an outside measurement of NO concentration vs. measured UV-A irradiance, and (B) influence of the pollutant concentration and flow rate Q on the degradation rate, in photocatalytic concretes. Readapted from Hüsken G, Hunger M, Brouwers HJH. Experimental study of photocatalytic concrete products for air purification. Build Environ 2009;44:2463–74, with permission from Elsevier Ltd.

slags, from iron manufacturing, are also very commonly used as supplementary materials in the production of mortars and concretes. They have similar silica-alumina content to pozzolanic materials and because of this, can guarantee an interesting pozzolanic activity. Silica fumes, fly ash, and slags are all fine materials, generally used as cement substitutes in concretes, there are, however, other industrial by-products which can be used to replace sand and coarse aggregates. Among these products papermill residues and wood shavings have been shown to appropriately bind in cementitious mixtures and are widely used for the production of lightweight concretes (LWC) [43–46]. Given the huge diffusion of such kind of composites, Coatanlem et al. investigated the durability of wood fiber lightweight concrete in different environmental conditions. Results show that the material properties are improved when wood chippings saturated with sodium silicate solution are considered. In particular, Fig. 9 shows the formation of a huge amount of CSH I ettringite, responsible for an improved bond between the chippings and cement paste. The scanning electron microscopy (SEM) observation proved that after 16 months these formations are widely developed on both the surface of wood chippings and within wood capillaries. Concerning the production of concrete from waste materials, three different categories can be identified: demolition wastes concretes, industrial wastes concretes, and agricultural wastes concretes. Being able to successfully reuse the huge amount of debris and wastes from the demolition activities in the building sector could represent an effective way of decreasing its overall impact. As a matter of fact, the use of such demolition wastes in order to produce new concrete composites can be considered as a widely investigated application. Recycled concrete aggregates (RCAs) have been proven to be of poor quality, when compared to the original product, given the presence of impurities, such as glass, metal, gypsum, or wood, coming from the demolition process. They are indeed associated to higher porosity and lower density [47–49], and the bond between the original aggregate and the attached mortar residue generally presents a weak quality. Nevertheless, different research studies proved that the use of specific pretreatments procedures can highly reduce such harmful decrease, i.e., RCA coating procedure, impurities removal procedure, oven curing technique, and RCA calcination [50–53]. The use of industrial waste as concrete components can rely on a massive number of possible candidates with completely different characteristics. Among all these substances, plastic materials and tire rubber are probably the most interesting ones. Such wastes are in fact characterized by low biodegradability and represent a serious problem during the end phase of their life cycle. Plastics are generally used as coarse aggregates substitutes and produce a sensible increase in both tensile and flexural strength of the composite. As a drawback, compressive strength and workability tend to decrease with increasing plastic content in the mixture. Tire rubber, on the other hand, which seem to be intrinsically compatible with asphalt mixtures, are very often associated with poor behavior if used in common concrete mixtures, this is probably due to both chemical incompatibility and significant stiffness variation between the two materials [54–56]. Beside plastic and rubber, further waste materials were investigated in concrete mixtures with the main purpose to produce ecoconcretes. Among these materials, ceramic and glass are particularly important. Ceramic materials seem to produce no adverse effect on concrete consistency when combined with it, and to increase its mechanical properties [57]. Interesting results also come from the implementation of glass residues in autoclaved aerated concretes, where it produces composite materials with similar mechanical properties compared to the original ones [58]. In particular, as it is shown in Fig. 10, the average compressive strength of concrete samples produced with using 20% and 25% of recycled aggregates in weight, i.e., CC-20 and CC-25 sample, respectively, follows a similar trend in time. Agricultural wastes can, as well as the industrial ones, guarantee a massive variety of residues and they are deeply rooted all around the world. In general, agricultural wastes can be combined with concrete as partial cement replacement material, partial aggregate replacement, or even as fiber reinforcement. The former application is justified by the high number of minerals and

Novel Building Materials

993

Ettringite Ettringite C-S-H

Ca(OH)2

(A)

(B)

Ettringite and C-S-H

(C)

Wood shaving surface

Ettringite

C-S-H

(D)

Fig. 9 Scanning electron microscopy (SEM) observations of the long-term conserved samples: (A) and (B) the matrix of a WCwh sample, (C) the shaving surface, and (D) a detail of ettringite and C–S–H crystals. Reproduced from Coatanlem P, Jauberthie R, Rendell F. Lightweight wood chipping concrete durability. Constr Build Mater 2006;20:776–81, with permission from Elsevier Ltd.

silicates that an annually grown plant can absorb from the earth, yielding to resources with high pozzolanic reactivity. The use of agricultural wastes as partial aggregates replacement, on the other hand, seems to be an encouraging way to reduce the dependence on gravels and natural mining sand [59]. Lastly, their use in fiber-reinforced concrete seems to be an environmentally friendly alternative to the synthetic fibers, which can be harmful to both the environment and the human health, such as carbon fibers and polypropylene fibers. In general, the use of agricultural wastes can maintain or even improve some mechanical properties, such as tensile strength, nevertheless, a major drawback lies in the fact that their use in cementitious mixtures always increases the overall water demand of the composite, with negative effects in terms of workability. The last category of concrete composites is that of concretes with enhanced thermal properties. From a thermal point of view, concretes can be modified in order to achieve an increase in its sensible or in its latent heat storage capability. In sensible heat storage, the energy is stored by raising the temperature of the storage medium. In this view, sensible storage materials need to guarantee high-specific heat values. Latent heat storage, on the other hand, relies on the phase change transition of the storage medium, for example, solid–liquid transformation (phase change materials (PCMs)), to store heat in the form of latent heat of fusion. High sensible storage concretes mainly sum up on the class of LWC. Such composites can be described as low density concretes for structural and nonstructural elements, characterized by an enhanced thermal insulation capability, due to the introduction of a gas or a foam agent within the concrete matrix, or to a partial replacement of ordinary aggregates with low weight ones [60–62]. LWCs can be produced by using natural or artificial minerals or biological natural fibers or particulate. In the first case, materials; such as porous volcanic stones, expanded clay, expanded shale, and perlite, can produce a sensible increase in thermal insulation capability of the composite [63,64], but a significantly decrease is generally obtained by using expanded polystyrene (EPS) beads or foams, which can lead to thermal conductivity values of about 0.06 W/mK, quite remarkable if compared to the

994

Novel Building Materials

Compressive strength (MPa)

50.00

7 days

28 days

90 days

45.00 40.00 35.00 30.00 25.00 20.00 15.00 10.00 5.00 0.00

RC

CC-20 Type concrete

CC-25

Fig. 10 Compressive strength of reference concrete (RC), concrete with 20% recycled aggregate (CC-20), and concrete with 25% recycled aggregate (CC-25). Reproduced from Medina C, de Rojas MS, Thomas C, Polanco JA, Frías M. Durability of recycled concrete made with recycled ceramic sanitary ware aggregate. Inter-indicator relationships. Constr Build Mater 2016;105:480–6, with permission from Elsevier Ltd.

general range of common concrete materials, i.e., 0.62 and 3.3 W/mK [65]. Appreciable results can also be obtained by considering the inclusion in LWCs of glass beads, plastic and rubber. A huge piece of literature also assesses the thermal properties of LWC produced by means of wood chippings, barley straw and hemp fibers inclusion. In particular, hemp fiber concrete is the reference material when new bio-based LWCs are considered. In all these composites, the overall percentage of bio-based material included in the mixture clearly influences the final thermal response of the final component. In particular, Al Rim et al. found thermal conductivity values ranging from 0.24 to 0.08 W/mK with varying wood proportions from 10% to 50% in weight [66]. In recent years, the possibility to overcome the simple sensible heat storage in favor of combined sensible-latent heat storage is receiving more and more attention in the scientific community. In order to do so, several researchers have been studying the chance of producing a new family of concretes by using PCMs as doping agents. PCMs are a particular kind of latent heat storage media, using melting and solidification processes to store heat and release it later on, when such a heat will be needed. A PCM should be characterized by a large latent heat, in order to store the higher amount of energy in the transition process, and a high thermal conductivity, in order to guarantee a proper distribution of heat within the material. Furthermore, they should also be characterized by a certain chemical and cycling stability, non-toxicity, non-corrosiveness, and small volume change values. They should of course be chosen in order to have a melting temperature lying in the practical range of operation of common building applications, and to reduce as possible the subcooling effect. Subcooling is a severe problem, capable of seriously reducing heat recovery from PCMs, and particularly serious when salt hydrates are considered [67]. When subcooling appears, in fact, the selected PCM starts to solidify at temperatures below its congealing temperature. PCMs can be distinguished in two main groups: organic and inorganic materials. Both categories include eutectics, characterized by a single melting temperature value, and mixtures, characterized by a melting interval. Organic mixtures can be distinguished in paraffins and fatty acids, while the inorganic ones only consist of hydrated salts. Organic PCMs possess a reliable chemical and thermal stability, no or low subcooling, and can be considered as noncorrosive materials. On the other hand, they are generally associated with lower phase change enthalpy, low thermal conductivity, and flammability. Inorganic PCMs have several disadvantages, such as corrosion, phase separation and segregation, lack of thermal stability, and common subcooling, nevertheless, they are usually associated with greater phase change enthalpy with respect to organic materials. The incorporation of a PCM within concrete can be made in different ways and it will be further discussed in Section 2.30.5.2, however, whatever technique will be chosen, the final effect of the PCM inclusion in concrete composites, at least until leakage occurrence, will be the reduction of the composite thermal conductivity and the increase of its specific heat capacity consistently with the PCM percentage content.

2.30.5.2

Thermal Insulation Materials

The enhancement of the insulation performance of concrete is a particularly interesting field of research, nevertheless, until now it was not possible to define the proper mix design able to both maintain its overall mechanical performance and drastically reduce heat transfer through its matrix. In this context, the building sector has been working for years in developing additional insulation components in the form of panels, flocks, or foams, to be implemented in the building envelope in order to produce an effective change in its overall thermal response. Given the strict control of national standards in terms of thermal transmittance values of building envelope components, the use of conventional insulation materials, such as stone and glass wool, EPS, and extruded polystyrene (XPS), phenolic foam (PF), polyurethane (PUR), and more is a widely diffused practice. The implementation of such components in the building envelope creates a thin but high thermal resistance layer, characterized by thermal conductivity values

Novel Building Materials

995

ranging from 0.018 to 0.040 W/mK. In addition to the previously presented materials, natural substances, such as cellulose, cork, hemp, and kenaf fibers, are starting to be used with higher frequency. They are indeed considered as more sustainable alternatives to artificial insulation materials but are generally associated with lower thermal performance, i.e., l values between 0.037 and 0.075 and more, depending on the specific application and its overall density. Beside such common applications, recently, several innovative solutions have been investigated by researchers all over the world, with the focus of both reducing thermal conductivity values, and thickness and weight of the final component. Particularly interesting examples are PCMs, aerogels, vacuum insulation panels (VIPs), and gas filled panels (GFPs). PCMs were already presented in Section 2.30.5.1.2. They are a particular kind of materials which undergo a phase transition, i. e., solid to liquid and liquid to solid, in the temperature range commonly occurring in building applications. As it can be seen in Fig. 11, during phase change, heat is absorbed or released and used to produce or break interatomic and intermolecular bonds, whether melting or solidification phenomena are occurring, respectively. The use of such materials enables to store heat in a latent form reducing the overall sensible temperature gain of the material. Two are the main PCMs which can be used in passive building applications: organic, mostly paraffin and acids and inorganic PCMs, generally hydrated salt. Although hydrated salts are associated with high volumetric heat storage and good thermal conductivity, their very high volume change and subcooling effect highly limit their effective use in most components. Paraffin and acids, on the other hand, which generally combine a high latent heat storage capacity with low volume change, have been widely and effectively incorporated in concrete composites [67,68]. Aerogels are synthetic, porous, ultralight materials derived from a gel, whose liquid component has been extracted through supercritical drying, and replaced with a gas. They can be made from a variety of chemical compounds (silica, carbon, metal oxide, etc.), but they always present a high open porosity with pores ranging from 2 to 50 nm [68–72]. Such an extremely high porosity generally results in densities of about 70–150 kg/m3 for common building applications, but can even produce values as low as 3 kg/m3. Given the highly variable density range, aerogels thermal conductivities are also spread over a broad range of variation. For example, Beatens et al. found that silica aerogel, whose porosity value is between 85% and 99.8%, are associated with l values between 0.0131 and 0.0136 W/mK for the monolithic material [73]. Neugebauer et al., on the other hand, measured a thermal conductivity decrease from 0.024 W/mK for an 88 kg/m3 dense sample to 0.013 W/mK for a 150 kg/m3 dense one, by means of transient plane source method (see Fig. 12) [74]. The production costs of aerogels are still very high. Aerogels have a relatively high compression strength, but are very fragile due to their very low tensile strength. These materials can be produced as either opaque, translucent or transparent, thus enabling a wide range of possible building applications. VIPs are a particular kind of insulation panels that can reach thermal conductivity values ranging between 0.004 W/mK in fresh conditions, to typically 0.008 W/mK after 25 years ageing, and thicknesses of about 5 mm. From a technical point of view, VIPs use no insulation material, since their thermal properties are due to vacuum generation, thus to the complete absence or, more realistically, huge concentration reduction of any atomic compound within a finite volume. Such panels consist of three different layers: a membrane wall, used to prevent air from entering the panel; a panel of a rigid, highly porous material, such as fumed silica, aerogel, perlite, or glass fiber, used to support the membrane walls against the atmospheric pressure once the air is evacuated; some chemicals, added to VIPs with glass fiber or foam cores, since cores with bigger pore size require a higher vacuum during the planned service life. The unavoidable thermal conductivity increase upon aging associated to VIPs is due to water vapor and air diffusion into the VIP core material, which has an open pore structure. Such phenomenon, together with high sensitivity to any structural damage of the panel, represent a major drawback of all VIPs, and is capable of enhancing thermal conductivities up to values of about 0.020 W/(mK).

(L Sen iquid) sible hea t

id) (Liqu heat sible

(Melting) Latent heat

Cooling cycle Sen

(Freezing) Latent heat

Se (So ns lid) ibl eh ea

d) eat oli (S le h ib ns

t

Se

Temperature

Heating cycle

Energy absorbed (heating) - energy released (cooling) Fig. 11 Theoretical latent heat curve for solid/liquid phase transition.

996

Novel Building Materials

Thermal conductivity, k (mW/(m K))

25

20

Cabot granules MIT granules MIT monolithic

46 kg m−3 66 kg m−3 55 kg m−3 79 kg m−3 71 kg m−3

15

99 kg m−3 90 kg m−3 132 kg m−3

10

150 kg m−3

5

0 0.1

1.0 10.0 Vacuum chamber pressure (kPa)

100.0

Fig. 12 Thermal conductivity of granular silica aerogel (GSA) from hot-wire testing under vacuum. Bed or monolithic densities are listed for each sample. Reproduced from Neugebauer A, Chen K, Tang A, et al. Thermal conductivity and characterization of compacted, granular silica aerogel. Energy Build 2014;79:47–57, with permission from Elsevier Ltd.

Close to VIPs in principle, GFPs use noble gasses, such as argon (Ar), krypton (Kr), and xenon (Xe), as thermal insulation materials. A GFP structure consists of a barrier foil and a baffle structure inside of it. To keep the optimal thermal performance, it is crucial to maintain the low-conductive gas concentration inside the structure and avoid air and moisture penetration into it. Obviously, vacuum is a better thermal insulator than the various gases employed in this panels, but on the other hand, the GFP grid structure does not have to withstand an inner vacuum-generated stress as the VIPs do. Furthermore, the use of low-e surfaces inside GFPs structure can also produce a significant reduction in the overall radiative heat transfer phenomenon. Beatens et al. produced 45-mm thick air and argon-filled GFPs prototypes showing thermal conductivity values of about 0.046 and 0.040 W/mK, respectively [75]. Lower values were instead obtained by Griffith et al. in the development of a 44.5-mm thick krypton filled panel, i.e., about 0.012 W/mK [76].

2.30.5.3

Passive Cooling Materials

Passive cooling is defined as the limitation of heat inside buildings by means of natural processes aimed at reflecting or expelling such heat into the atmosphere or the ground beneath constructions [77]. There are different kinds of passive cooling techniques: natural ventilation, evaporative cooling, radiant cooling, reflective cooling, and more. In particular, in the last decade researchers all over the world have investigated the effect of reflective cooling on the overall thermal and energy balance of buildings. Reflective cooling is a passive solution aimed at decreasing the amount of thermal solar gains through the building envelope. It is obtained by applying materials with enhanced thermal-optical properties on the external surface of the building envelope, i.e., cool materials. Cool materials are high solar reflectance and high thermal emittance materials, i.e., associated to solar reflectance index (SRI) values higher than 0.65 and e values around 0.80–0.90, respectively. They can be applied as surface coatings on existing building envelop components, or directly incorporated within the building element during the production phase, such as in concrete or common roof tiles or paving materials, for outdoor environments. In any case, their main effect consists of the reduction of the total amount of solar energy absorbed by the material, consequently causing the drastic abatement of the local surface temperature, and finally an appreciable air temperature decrease in the surrounding environment [78]. Beside the thermal effect on the indoor environment, cool materials have been shown to also produce significant benefits in terms of climate change and urban heat island mitigation [79,80]. Since cool materials were initially developed in order to interact with the visible part of the solar spectrum, the original prototypes were mostly white pigment finishing to be applied as an external layer onto existing elements. However, the increasing awareness about solar radiation characteristics and energy content, and particularly the fact that almost 50% of the global solar energy reaching the earth surface can be found in the infrared region, drove researchers’ attention toward the production of new cool materials. As a matter of fact, four different classes of cool materials can currently be depicted: light-colored, highly reflective colored, retroreflective, and thermochromic materials (TCMs). Novel light-colored materials preserve the visual appearance of the original prototypes, but they also implement a special kind of white-colored pigments, such as zinc dioxide and titanium dioxide, or transparent polymers, such as acrylic, characterized by a high solar reflectance value also in the near-infrared region of the solar spectrum, i.e., about 0.7–0.85 [81–84]. Highly reflective colored materials, on the other hand, preserve the original interaction

Novel Building Materials

997

between the surface and the visible part of the solar radiation, but because of the implementation of small percentages of high infrared reflectance pigments, such as the ones used in the previous case, can guarantee competitive optical performance in the near-infrared wavelength range. Retroreflective materials and TCMs present a more elaborated reflection capability. The former are highly reflective coatings, for example, glass-beaded type and cube corner type, reflecting solar radiation in the same direction of the incoming one. The latter can dynamically alter their thermal-optical properties in response to varying environmental conditions. Both these materials can be very useful when integrated in building envelope components. Retroreflective materials can, indeed, reduce mutual reflection phenomena characterizing urban canyon and canopy layers, producing a sensible benefit in terms of outdoor and indoor air temperature in hot summer conditions. TCMs, on the other hand, using a reversible reaction of the pigments molecular structure could be used to mitigate both summer and winter building thermal loads, given their ability to reflect most of the incoming radiation in summer conditions, and absorb most of it in winter conditions [85,86].

2.30.5.4

New Materials for Shading Elements

A particular kind of reflective passive cooling strategy consists of the implementation of shading elements on the building façade. Such components are used in order to control the amount of solar radiation received by the building and are particularly effective when façade cavities, such as fenestration systems, are taken into account. Anyway, the use of passive shading systems reduces summer heat gains but it also decreases daylight availability [87]. In this context, the use of dynamic solar shading devices is becoming more and more popular, and smart materials are gathering greater attention. Smart materials are substances capable of changing one or more of their properties, in response to an external stimulus [88,89]. They can act both as sensor and actuator elements, i.e., they are capable of analyzing and detecting the variation of an external boundary condition, and in response to such variation, they are also able to operate an effective change in one of their properties. Given their outstanding characteristics, smart materials can be considered as the perfect candidates for shape morphing solar shading applications. Clearly, not all the smart materials can be successfully applied in shading devices: they have to fulfill some basic selection criteria [90,91]:

• • • • •

they they they they they

need need need need need

to to to to to

be responsive to specific boundary conditions, such as solar radiation, air temperature, or electrical stimuli; properly interact with the outdoor environment without any corrosion or chemical degradation; ensure suitable workability and adaptability features; withstand mechanical stresses and guarantee adequate handling; and preserve their characteristics in time.

In particular, focusing on the shape change physical phenomenon, smart materials can be grouped in two main categories: shape change materials (SCMs) and shape memory materials (SMMs). The former, are capable of operating a change in their shape when exposed to the proper stimulus, for example, piezoelectric and electro-active polymers. The latter, are able to maintain a specific shape until the proper stimulus, generally temperature difference, is applied, activating the shape recovery cycle [92,93]. Given their intrinsic correlation with temperature, and the inevitable thermal effect of the incoming solar radiation on the receiving surface, SMMs can be considered as the most suitable substrates for shading elements applications. There are three fundamental SMMs categories, i.e., shape memory polymers (SMPs), shape memory alloys (SMAs), and shape memory hybrids (SMHs), each of which can be designed in order to be thermally activated by the outdoor environment. SMPs are low density and highly biocompatible SMMs, constituted by chemically bonded molecular chains of monomers [88]. On the same level of PCMs presented in Sections 2.30.5.2 and 2.30.5.3, SMPs due their unique properties to a phase transition trigger within their matrix. However, in this case such transition is not used in order to store heat in a latent form, but to alter the mechanical properties of the material. Indeed, once the specific transition temperature is reached, SMPs can be easily modeled is a desired temporary shape by applying an external force on the softened bulk material. If the material is then kept at the temporary desired shape upon cooling, such shape, will be fixed and preserved until the specimen is again heated above its transformation temperature [90]. There is a huge variety of SMPs, but depending on their chemical composition, they can be distinguished in two main categories: SMPs based on glass transition, i.e., chemically cross-linked glassy thermosets, and physically cross-linked amorphous thermoplastics, and SMPs based on melting transition, i.e., chemically cross-linked semicrystalline rubbers, and physically cross-linked semicrystalline block copolymers. In any case the considered material should never fully complete the transition, particularly in the melting case, but only reach a softened and malleable form, in order to guarantee a proper deformation process, as it is shown in Fig. 13. SMAs are metal alloys using the martensitic thermoelastic transformation in order to preserve the memory of a previous state. They are characterized by a good biocompatibility and more interestingly by a high corrosion and electrical resistance [88]. If a SMA initially in the parent phase condition, i.e., higher temperature, is cooled to the martensite phase and an external load is applied, the specimen is apparently deformed in a plastic way. In fact, the deformation takes place by a reorientation of the martensite variants toward those favored by the external load. If the specimen is heated up till the reverse transformation takes place, the parent phase crystal structure and shape is spontaneously restored and the material spontaneously adopts its original shape. According to their base alloy, SMAs can be divided in three different categories: Cu-based, Fe-based, and NiTi-based, which is characterized by the higher recoverable strain and corrosion resistance [90].

Novel Building Materials

998

(A)

(B)

(C)

Fig. 13 Visual demonstration of shape memory properties for NGDE2. (A) Original/permanent shapes, (B) fixed temporary shapes, and (C) recovered shapes. Reproduced from Xie T, Rousseau IA. Facile tailoring of thermal transition temperatures of epoxy shape memory polymers. Polymer 2009;50(8–9):1852–6, with permission from Elsevier Ltd.

The last considered group of SMMs is that of SMHs. SMHs are produced by the combination of well-known materials separately having no memory properties and maintaining their characteristic chemical nature even when combined. The final composite can be obtained by combining metals, organic, or inorganic materials. A well representative SMH is the thermo-responsive silicon/wax one. Wax, which becomes quite soft upon heating, works as the transition inclusion, while silicon, which generally is more stable in terms of temperature gradients, is the shape stabilizing agent, keeping the elasticity of the material. During the heating phase, wax is modeled and the acquired temporary shape is preserved upon cooling, causing a certain amount of elastic energy to be stored in the silicon matrix. Once the SMH is reheated, the embedded elastic energy is finally released and the original shape restored [90,92].

2.30.5.5

Advanced Materials and Coatings for High Performance Glazing

Glazing constitute the largest proportion of a window area, therefore, they play a major role on the overall thermal-optic performance of a fenestration system and consequently on the global thermal-energy performance of buildings. In such view, glazing materials need to be properly designed in terms of thermal insulation performance, solar gain control, and daylighting management. Technologies, such as multilayer glazing filled with low U-value inert gasses, i.e., Argon or Krypton, vacuum glazing, and low-emittance coatings made of metals or metallic oxides are, nowadays, widely used in common applications. Anyway, in recent years, several kinds of novel materials have been investigated in search of the optimum balance between visual and thermal comfort of the indoor environment. In this context, some of the most promising materials and coatings are aerogels, smart materials, and antireflective coatings.

2.30.5.5.1

Aerogel

Aerogel is a highly appreciated thermal superinsulation material characterized by an open-celled mesoporous structure, which can lead to thermal conductivity values of about 0.013 W/mK and porosity values between 50% and 90%. Different kind of aerogel can be produced, but silica aerogel is the most common application for fenestration systems and for the building sector in general. Silica aerogel is derived from supercritical drying of alcogels and can be shaped in the monolithic (monolithic silica aerogel, MSA) or in the granular (granular silica aerogel, GSA) form. The former, is usually associated with high light transmittance values, i.e., around 0.9, while the latter is characterized by very low light transmittance [94]. Both MSA and GSA are generally used as filling agent between multipane windows and glazing material for windows. Because of their translucent nature, aerogel glazing can seriously improve the indoor visual comfort level and reduce the annoying glare effect [95]. Furthermore, monolithic aerogel has also proven good thermal performance both in cooling and heating dominant climates [96].

2.30.5.5.2

Smart materials

Smart materials, introduced in Section 2.30.5.5, can also be applied as glazing materials. In this case, it is useful to distinguish between materials for passive and active glazing applications. In passive glazing, the material undergoes a modification in its original properties when exposed to heat, but it is not possible to directly modulate the rate of change of such properties. Common examples of materials for passive glazing applications are TCMs and PCMs. In active glazing applications, on the other hand, the materials optical properties are directly tuned by means of an external stimulus, such as an electric field, heat, or ion diffusion. Common examples of materials for active glazing applications are electrochromic materials (ECMs), gasochromic materials (GCMs), suspended particles (SPs), and liquid crystals (LCs). Among the materials for passive glazing, PCMs were already described in the previous sections. Their use as thermal energy storage and shape memory media has already been explained, nevertheless, their application in glazing reveals their peculiar interaction with solar radiation. PCMs can indeed be implemented in translucent windows. They will mostly absorb infrared radiation, store its energy content in a latent way and undergo a solid to liquid transition. Their effect will be particularly

Novel Building Materials

999

Fig. 14 Two of the phase change material (PCM) glazing samples, with PCM in solid state (left) and PCM in liquid state (right). Reproduced from Goia F, Zinzi M, Carnielo E, Serra V. Spectral and angular solar properties of a PCM-filled double glazing unit. Energy Build 2015;87:302–12, with permission from Elsevier Ltd.

interesting in terms of thermal insulation capability, nevertheless, as it is shown in Fig. 14, their poor visual transmittance will inevitably reduce indoor visual comfort in buildings. The second kind of passive materials, i.e., TCMs, on the other hand, can change its optical properties in response to temperature variation. When their temperature increases, and reaches a critical transition value, materials, such as vanadium oxide (VO2), undergo a semiconductor to metal transition. As a consequence of such transition, the material is able to reflect the incoming infrared radiation and reduce the indoor solar gains [98]. The main drawbacks of such technology lie on the fact that generally, materials like VO2 have a high transition temperature, i.e., about 681C, and a low visible light transmittance. For these reasons, they need to be doped in order to decrease the transition threshold and antireflective coatings need to be implemented on their surface to increase visible transmission. Despite the huge effort in implementing such kind of coatings, it was not possible to develop a suitable prototype for industrial applications. Passive glazing materials react to an external stimulus with a direct change in their optical properties, but such change cannot be tuned. By using active glazing materials, on the other hand, it is possible to directly modulate the change in absorbance, reflectance, or scattering of the smart matrix. Diverse kinds of active glazing materials are triggered by different stimulus and change their optical properties in a unique way. ECMs, for example, change their optical properties, and consequently their visible and infrared transmittance, upon a small DC voltage application. The change in transmittance is simultaneously activated in the visible and the infrared range of the solar spectrum. However, recent researches are trying to develop EC materials in which infrared transmittance can be altered without changing the visible one [99–101]. ECMs can be distinguished in organic and inorganic. The former due their unique attitude to oxidation/reduction processes of bipyridinium systems, conducting polymers and other ion-susceptible composite, the latter to the variation of the oxidation state of metal oxides, such as Wolfram and Nichel. ECMs are highly appreciated since they can guarantee a high modulation over visible transmittance, i.e., more than 60%, and switching time varying from several seconds to several minutes [102]. A transparency range similar to that of ECMs can also be reached by using GCMs. GCMs are a particular kind of materials capable of changing their color and transparency to solar radiation, when exposed to diluted hydrogen gas, i.e., mixed with Argon. In these materials, light transmission is a function of hydrogen content, and the inverse reaction is activated by the introduction of diluted oxygen in the system. The most common GCM is tungsten oxide, coated with a thin catalyst layer, for example, Platinum (Pt) or Palladium (Pd). Such combination allows the transmittance coefficient to vary from 0.77 to 0.06 but, when implemented, generally needs more control equipment if compared to electrochromic systems [103,104]. Gasochromic glazing systems are highly affected by the overall thickness of the tungsten oxide film, the total sputtering pressure and the total pressure. In particular, Vitry et al. found that increasing values of the former two variables directly affects the coloring rate of the gasochromic, while higher porosity is obtained by producing an increase in total pressure [105]. Lastly, SPs and LCs are a particular kind of active smart materials, reacting to electric gradients. In the absence of an electric field SP and LC glazing are constituted by randomly oriented particles: the former behaves as a light absorbing material seriously reducing light transmission; the latter, behave as highly scattering substrates and appears as a white, translucent media. The application of an AC voltage, causes the particles to realign and the window to become transparent (see Fig. 15). SP media are a suspension of dipolar particles with high optical anisotropy, i.e., polyodides, in an organic fluid. Such composite is then interposed between two insulator and conductive layers. The overall transmittance of the system is a function of its thickness and of the particles density. Furthermore, it is possible to directly tune the optical response of the glazing and the time response is considerably faster than that of passive ECMs. Nevertheless, SP layers are generally associated to higher energy consumption level [107]. LCs, on the other hand, are a composite of nematic LC droplets in a polymeric matrix. Their highly refractive nature in absence of an external electric field is a consequence of the huge mismatch between the droplets and the matrix refractive index.

1000

Novel Building Materials

Two glass panes

Two glass panes

SPD particles Randomly oriented SPD particles Light Light

V

V

SPD switch off/ ‘opaque’ state

SPD switch on/ ‘transparent’ state

(A)

(B)

Fig. 15 Schematic of a suspended particle device (SPD) window (A) upon the applied voltage (B) in the absence of voltage. Reproduced from Ghosh A, Norton B, Duffy A. Measured overall heat transfer coefficient of a suspended particle device switchable glazing. Appl Energy 2015;159:362–9, with permission from Elsevier Ltd.

2.30.5.5.3

Antireflective coatings

Aerogel and smart materials are generally implemented within the gap between two window panes. The overall thickness of such additional layer can vary from a few millimeters to a few centimeters, and of course such a variation affects both thermal and optical properties of the system. Antireflective coatings, on the other hand, are generally associated with smaller thicknesses and can be applied to both sides of the glass. They are basically aimed at increasing daylight and solar energy transmittance, without altering the overall thermal transmittance of the glazing element. Antireflective properties can be achieved by producing a multilayered structure with increasing refractive index materials or by using patterned and porous structure coatings. The first approach is based on the consideration that if a substrate with a refractive index Z2 is coated with a film with a refractive index Z1, light that is reflected at the interface with air can be expressed by the following index: R¼

Zair Zair þ

Z1 2 Z2 Z1 2 Z2

ð15Þ

As a consequence of the above equation, a gradual variation of the refractive index is required in order to minimize the medium change induced reflectance. Unfortunately, it is not always possible to identify the material with the proper refractive index, since it is a function of the light wavelength and of the incidence angle, consequently, it is not possible to identify a perfect reflective coating for visible light. Nevertheless, using multilayered coatings can help in reducing the overall reflectance of the considered glazing (see Fig. 16). As for patterned and porous structure coatings, they are generally produced by means of nano-imprint lithography and can guarantee superior antireflective properties over a wide range of frequencies and incidence angles. However, production scalability and production costs of the final product represent a challenge for patterned structure coatings nowadays.

2.30.5.6

Integration of Novel Materials in Building Applications

Novel materials with advanced thermal-optic and acoustic performance introduced in the previous section are specifically developed and optimized in order to be integrated in high performance building components.

Novel Building Materials

1001

Air Incident light

n1

Reflected light Air

n2

n3

n3

Air

n2

70% of incident light transmitted

nair