132 29 18MB
English Pages 516 [511] Year 2021
Patricia Horcajada Cortés Sara Rojas Macías Editors
Metal-Organic Frameworks in Biomedical and Environmental Field
Metal-Organic Frameworks in Biomedical and Environmental Field
Patricia Horcajada Cortés • Sara Rojas Macías Editors
Metal-Organic Frameworks in Biomedical and Environmental Field
Editors Patricia Horcajada Cortés Parque Tecnológico de Móstoles IMDEA Energy Foundation Móstoles, Madrid, Spain
Sara Rojas Macías Parque Tecnológico de Móstoles IMDEA Energy Foundation Móstoles, Madrid, Spain
ISBN 978-3-030-63379-0 ISBN 978-3-030-63380-6 https://doi.org/10.1007/978-3-030-63380-6
(eBook)
© Springer Nature Switzerland AG 2021 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors, and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Switzerland AG The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
Preface
Metal-Organic Frameworks (MOFs), also known as Porous Coordination Polymers (PCPs), first emerged as a new class of porous materials, but later came to an extensive new domain of research. MOFs are synthetic materials based on inorganic nodes (e.g., atoms, clusters, chains) and polycomplexant organic linkers (e.g., carboxylates, phosphonates, azolates) that assemble into multidimensional periodic lattices, giving rise to crystalline structures with channels and/or cavities of different size and shape. Compared with other classical porous materials (e.g., active carbons, zeolites, silicas), the modular nature of MOFs, with a large diversity of cations and organic linkers, allows certain level of design, affording both fine chemical and structural control. For example, the appropriate selection of endogenous, biocompatible, and/or even bioactive metals and linkers has been proposed as an interesting approach for producing promising MOFs in biomedicine. Another structural feature that is also characteristic of most MOFs is large porosity (surface area, pore volume, pore size) associated with exceptional sorption capacities of different guest molecules (e.g., greenhouse gases, aqueous contaminants, active pharmaceutical ingredients). Further, MOFs contain coordinatively unsaturated sites (CUS) that exhibit unique adsorption and catalytic properties. Considering their potential applicability, MOFs show an adequate stability profile under relevant working conditions, remaining stable enough to carry out their function under a large range of environments, can be synthetized at large scale, and be adequately shaped (e.g., monoliths, pellets, membranes, columns). The combination of all these features turns the initial research, focused on the description of novel crystalline structures, to the application of these hybrid porous solids to relevant fields, including storage, separation, sensing, fuel cells, drug delivery, imaging, (photo-, electro-, bio-) catalysis, etc. In the present book, we mainly focus on the potential of MOFs in applications such as the leading-edge environmental (energy-related) and biomedical fields. The successful application of MOFs in these areas takes place after the physi- or chemi-sorption of active ingredients or substrates in the porosity or active sites, the sorption process being a crucial step. The use of MOFs in the biomedical and environmental field is v
vi
Preface
currently progressing at a fast pace due to the large possibilities that MOFs offer in terms of composition, topology, incorporation of active species (in their porosity, on their external surface, or within the framework), and post-synthetic modification, among others. The present book contains a collection of 14 chapters written by recognized authors that have importantly contributed to the development of their respective fields. The book is divided in two main parts corresponding to the environmental (from Chaps. 1 to 8) and biomedical application of MOFs (from Chaps. 9 to 14). Chapter 1 deals with the synthesis of robust and environmentally friendly MOFs, highlighting the pivotal relevance of their stability and green nature to promote their use in industry. This chapter considers the overall operating conditions including the time-scale of the targeted process and more sustainable and/or economically viable synthetic routes. In this sense, the cooperation between industrial and academic partners is contributing to pave the way of MOFs to reach the doors of industrial production. In this sense, Chap. 2 summarizes the large-scale synthesis and specific shaping of MOFs, primarily by granulation methods. Chapters 3 and 4 propose MOFs in green energy production and storage. Chapter 3 is divided between the two main strategies of pure MOFs as photoand/or electro-catalysts around up-conversion of molecules for energy application. Then, Chap. 4 gives a brief overview of the potential applications of MOFs in the field of secondary (rechargeable) batteries and discusses how their unique characteristics might be exploited to improve the performances of the current electrochemical energy storage devices. Continuing with the environmental applications of MOFs, the subsequent chapters are focused on the environmental remediation, particularly Chaps. 5 and 6 on the adsorption of the main greenhouse gas (CO2), and Chaps. 7 and 8 on water remediation. Particularly, and after more than a decade of research into MOFs for selective CO2 uptake, Chap. 5 highlights several highly promising MOF candidates, which have emerged operating at a few percent of CO2 or even down to levels in the air. Authors review the significant developments in MOF-containing membranes for CO2 separation from more or less complex gas mixtures. Further, considering the rapid increase in the synthetized MOF structures and the challenge of assessing their CO2 adsorption capacity using purely experimental methods, Chap. 6 summarizes the efficient computational screening of MOFs in these applications. Recent progress in computational tools, high-throughput molecular simulations, and machine learning algorithms provide great opportunities for the effective MOFs screening and identification of the most promising adsorbents for CO2 capture prior to experimental studies. Large-scale computational screening studies and quantitative structure– performance relationships, obtained from molecular simulations, are discussed here. Considering the ultrahigh surface area and modular structure of MOFs, Chap. 7 highlights the potential of MOFs and MOF composites for the elimination of three heavy metals (mercury, lead, and cadmium) from water, along with an overview of their removal efficiencies and different kind of strategies used for designing these multifunctional materials. Complementary, Chap. 8 concerns the removal of hazardous organic contaminants from water using MOF. Authors discuss here the effect
Preface
vii
of functional groups on adsorption/purification, giving also brief methods to introduce functionality onto MOFs. The second part of the book, focused on biomedical applications of MOFs, is divided into six chapters. Chapter 9 details the emergence of biomolecules-based MOFs, particularly those constructed using nucleobases, amino acids, peptides/proteins, and carbohydrates building blocks. This chapter summarizes the unique properties of Bio-MOFs, or MOFs constructed using at least one biomolecule, for environmental and biological uses. In addition, MOFs can be further introduced into biocompatible architectures, enabling the integration of MOFs into real world applications. Thus, Chap. 10 summarizes recent developments on the MOFs incorporation into natural polymers like cellulose, cotton, pulp, silk, and chitin/chitosan for relevant uses, ranging from antibacterial to filtration. Chapter 11 deals with the quintessential bioapplication of MOFs: drug delivery. This chapter covers the recent progress of MOFs for delivery of small molecules and biologically active gases, with an emphasis on the importance of the biological context and pharmacological direction. The significance of material components and synthetic methods for optimal safety profiles, and recent progress in characterizing cellular uptake and internalization pathways of nanoscaled MOF are deeply evaluated. On a further step, Chap. 12 discusses the integration of MOFs with biomacromolecules to enhance targeting capabilities and improved biodistribution. In this chapter, authors summarize the principal synthetic methods for the preparation of MOF bio-composites using proteins, carbohydrates, nucleic acids, and lipids, and disclose relevant examples for their application in drug delivery, biopreservation, and biosensing. Finally, the use of MOF as protective exoskeletons of virus and cells for their potential applications to vaccine transportation/delivery and regenerative medicine is also discussed. Chapter 13 summarizes the advances on MOFs as sensing platforms for biomedicine, disclosing the application of MOFs in monomodal magnetic resonance imaging and multimodal imagining based on fluorescence imaging and magnetic resonance imaging for precise diagnosis of diseases. Finally, Chap. 14 gives a new perspective of the wide diversity of MOF-based biosensors, providing a significant number of examples and focusing on both the transduction mechanism and the nature of the analytes. Overall, the present selection of chapters aims to provide a balanced view of the current areas related with the environmental (energy/remediation) and biomedical fields in which MOFs have become in a short time period among the most prominent materials, highlighting the reasons why MOFs exhibit advantages over other existing materials. The aim here is to provide future research goals that emphasize relevant nuances to this class of materials as a whole. Finally, we hope that the reader enjoys this great selection and benefits from the pertinent views of leading researchers in their respective fields to learn first-hand the achievements and future targets in each of the topics covered by the book. Móstoles, Spain
Patricia Horcajada Cortés Sara Rojas Macías
Contents
1
Robust and Environmentally Friendly MOFs . . . . . . . . . . . . . . . . . Raquel Del Angel, Georges Mouchaham, Farid Nouar, Antoine Tissot, and Christian Serre
2
Large-Scale Synthesis and Shaping of Metal-Organic Frameworks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . U-Hwang Lee, Sachin K. Chitale, Young Kyu Hwang, and Jong-San Chang
1
33
3
Green Energy Generation Using Metal-Organic Frameworks . . . . . Giacomo Armani-Calligaris, Sara Rojas Macías, Víctor Antonio de la Peña O’Shea, and Patricia Horcajada Cortés
55
4
The Potential of MOFs in the Field of Electrochemical Energy Storage . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111 Thomas Devic
5
Carbon Capture Using Metal–Organic Frameworks . . . . . . . . . . . . 155 Ram R. R. Prasad, Qian Jia, and Paul A. Wright
6
Computational Screening of MOFs for CO2 Capture . . . . . . . . . . . 205 Cigdem Altintas, Ilknur Erucar, and Seda Keskin
7
Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks (MOFs) . . . . . . . . . . . . . . . . . . . 239 Sanjit Nayak
8
Adsorptive Purification of Water Contaminated with Hazardous Organics by Using Functionalized Metal-Organic Frameworks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269 Dong Kyu Yoo, Biswa Nath Bhadra, and Sung Hwa Jhung
9
MOFs Constructed from Biomolecular Building Blocks . . . . . . . . . 291 Zachary M. Schulte and Nathaniel L. Rosi ix
x
Contents
10
Natural Polymer-Based MOF Composites . . . . . . . . . . . . . . . . . . . . 321 Tanay Kundu, Bikash Garai, and Stefan Kaskel
11
Metal-Organic Frameworks as Delivery Systems of Small Drugs and Biological Gases . . . . . . . . . . . . . . . . . . . . . . . . 349 Emily Linnane and David Fairen-Jimenez
12
MOFs and Biomacromolecules for Biomedical Applications . . . . . . 379 Francesco Carraro, Miriam de J. Velásquez-Hernández, Mercedes Linares Moreau, Efwita Astria, Christopher Sumby, Christian Doonan, and Paolo Falcaro
13
Diagnosis Employing MOFs (Fluorescence, MRI) . . . . . . . . . . . . . . 433 Jie Yang and Ying-Wei Yang
14
Biosensing Using MOFs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457 Javier Cepeda, Antonio Rodríguez-Diéguez, and Alfonso Salinas-Castillo
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
Contributors
Cigdem Altintas Department of Chemical and Biological Engineering, Koc University, Istanbul, Turkey Efwita Astria Institute of Physical and Theoretical Chemistry, Graz University of Technology, Graz, Austria Biswa Nath Bhadra Department of Chemistry and Green-Nano Materials Research Center, Kyungpook National University, Daegu, South Korea Giacomo Armani-Calligaris Advanced Porous Materials Unit, IMDEA Energy, Madrid, Spain Francesco Carraro Institute of Physical and Theoretical Chemistry, Graz University of Technology, Graz, Austria Javier Cepeda Departamento de Química Aplicada, Facultad de Química, Universidad del País Vasco/Euskal Herriko Unibertsitatea (UPV/EHU), DonostiaSan Sebastián, Spain Jong-San Chang Research Center for Nanocatalysts, Korea Research Institute of Chemical Technology (KRICT), Daejeon, Korea Department of Chemistry, Sungkyunkwan University, Suwon, Korea Sachin K. Chitale Research Center for Nanocatalysts, Korea Research Institute of Chemical Technology (KRICT), Daejeon, Korea Department of Green Chemistry & Biotechnology, University of Science and Technology (UST), Daejeon, Korea Raquel Del Angel Institut des Matériaux Poreux de Paris, Ecole Normale Supérieure, ESPCI Paris, CNRS, PSL University, Paris, France Thomas Devic Institut des Matériaux Jean Rouxel (IMN), Université de Nantes, Nantes, France
xi
xii
Contributors
Christian Doonan Department of Chemistry and Centre for Advanced Nanomaterials, The University of Adelaide, Adelaide, Australia Ilknur Erucar Department of Natural and Mathematical Sciences, Faculty of Engineering, Ozyegin University, Istanbul, Turkey David Fairen-Jimenez Adsorption & Advanced Materials Laboratory (A2ML), Department of Chemical Engineering and Biotechnology, University of Cambridge, Cambridge, UK Paolo Falcaro Institute of Physical and Theoretical Chemistry, Graz University of Technology, Graz, Austria Bikash Garai Technische Universität Dresden, Dresden, Germany Patricia Horcajada Cortés Advanced Porous Materials Unit, IMDEA Energy, Madrid, Spain Young Kyu Hwang Research Center for Nanocatalysts, Korea Research Institute of Chemical Technology (KRICT), Daejeon, Korea Department of Green Chemistry & Biotechnology, University of Science and Technology (UST), Daejeon, Korea Sung Hwa Jhung Department of Chemistry and Green-Nano Materials Research Center, Kyungpook National University, Daegu, South Korea Qian Jia EaStCHEM School of Chemistry, University of St Andrews, St Andrews, UK Stefan Kaskel Technische Universität Dresden, Dresden, Germany Seda Keskin Department of Chemical and Biological Engineering, Koc University, Istanbul, Turkey Tanay Kundu Technische Universität Dresden, Dresden, Germany U-Hwang Lee Research Center for Nanocatalysts, Korea Research Institute of Chemical Technology (KRICT), Daejeon, Korea Department of Green Chemistry & Biotechnology, University of Science and Technology (UST), Daejeon, Korea Mercedes Linares Moreau Institute of Physical and Theoretical Chemistry, Graz University of Technology, Graz, Austria Emily Linnane Adsorption & Advanced Materials Laboratory (A2ML), Department of Chemical Engineering and Biotechnology, University of Cambridge, Cambridge, UK Georges Mouchaham Institut des Matériaux Poreux de Paris, Ecole Normale Supérieure, ESPCI Paris, CNRS, PSL University, Paris, France
Contributors
xiii
Sanjit Nayak Lecturer in Chemistry, School of Chemistry and Biosciences, University of Bradford, West Yorkshire, UK Farid Nouar Institut des Matériaux Poreux de Paris, Ecole Normale Supérieure, ESPCI Paris, CNRS, PSL University, Paris, France Victor Antonio de la Peña-O’Shea Photoactivated Process Unit, IMDEA Energy, Madrid, Spain Ram R. R. Prasad Department of Chemistry, Dainton Building, The University of Sheffield, Sheffield, UK Antonio Rodríguez-Diéguez Departamento de Química Inorgánica, Facultad de Ciencias, Universidad de Granada, Granada, Spain Sara Rojas Macías Advanced Porous Materials Unit, IMDEA Energy, Madrid, Spain Nathaniel L. Rosi Department of Chemistry, Kenneth P. Dietrich School of Arts and Sciences, University of Pittsburgh, Pittsburgh, PA, USA Alfonso Salinas-Castillo Departamento de Química Analítica, Facultad de Ciencias, Universidad de Granada, Granada, Spain Zachary M. Schulte Department of Chemistry, Kenneth P. Dietrich School of Arts and Sciences, University of Pittsburgh, Pittsburgh, PA, USA Christian Serre Institut des Matériaux Poreux de Paris, Ecole Normale Supérieure, ESPCI Paris, CNRS, PSL University, Paris, France Christopher Sumby Department of Chemistry and Centre for Advanced Nanomaterials, The University of Adelaide, Adelaide, Australia Antoine Tissot Institut des Matériaux Poreux de Paris, Ecole Normale Supérieure, ESPCI Paris, CNRS, PSL University, Paris, France Miriam de J. Velásquez-Hernández Institute of Physical and Theoretical Chemistry, Graz University of Technology, Graz, Austria Paul A. Wright EaStCHEM School of Chemistry, University of St Andrews, St Andrews, UK Jie Yang College of Chemistry, Jilin University, Changchun, People’s Republic of China Ying-Wei Yang College of Chemistry, Jilin University, Changchun, People’s Republic of China Dong Kyu Yoo Department of Chemistry and Green-Nano Materials Research Center, Kyungpook National University, Daegu, South Korea
Chapter 1
Robust and Environmentally Friendly MOFs Raquel Del Angel, Georges Mouchaham, Farid Nouar, Antoine Tissot, and Christian Serre
List of Abbreviations ABTC BDC BET BPDC DABCO DMF DUT HKUST HPLC IBU ICP IRMOF MDIP MIL MIP MOF NMR NU PCN SALI TCPP TGA
3,30 5,50 -Azobenzenetetracarboxylate 1,4-Benzenedicarboxylic acid Brunauer-Emmett-Teller 4,40 -Biphenyldicarboxylate 1,4-Diazabicyclo[2.2.2]octane Dimethylformamide Dresden University of Technology Hong Kong University of Science and Technology High-performance liquid chromatography Inorganic building unit Inductively coupled plasma Isoreticular MOF 3,30 ,5,50 -Tetra-carboxylatediphenylmethane Materials of Institute Lavoisier Materials of the Institute of porous materials from Paris Metal-organic framework Nuclear magnetic resonance Northwestern University Porous coordination network Solvent-assisted ligand incorporation Meso-tetra(4-carboxyphenyl)porphyrin Thermogravimetric analysis
R. Del Angel · G. Mouchaham (*) · F. Nouar · A. Tissot · C. Serre (*) Institut des Matériaux Poreux de Paris, Ecole Normale Supérieure, ESPCI Paris, CNRS, PSL University, Paris, France e-mail: [email protected]; [email protected] © Springer Nature Switzerland AG 2021 P. Horcajada Cortés, S. Rojas Macías (eds.), Metal-Organic Frameworks in Biomedical and Environmental Field, https://doi.org/10.1007/978-3-030-63380-6_1
1
2
TzGal UiO XRD ZIF
1.1
R. Del Angel et al.
5,50 -(1,2,4,5-Tetrazine-3,6-diyl)bis(benzene-1,2,3-triolate) University of Oslo X-ray diffraction Zeolitic imidazolate framework
Introduction
Porous materials have attracted the attention of scientists for several decades. The presence of pores and high surface areas make this type of solids propitious for a wide variety of applications such as molecules capture and storage, filtration, or catalysis, among others. Zeolites, activated carbons, and silica are examples of benchmark porous materials that have been employed in industries for long time [1]. The impact of porous materials for industrial and technological applications can be of prime importance. Therefore, many scientists have embarked on the quest of finding new solids with higher surface areas and new properties that can make them appealing for novel applications or to improve the efficiency of current ones. Since the early 1990s, a new series of hybrid materials, porous polymeric coordination networks, has emerged as a promising class of crystalline porous solids [2–5]. These solids have been denoted as metal-organic frameworks (MOFs) and have received increasing attention, promoting them as a very popular research field among the scientific community [6–8]. MOFs are a family of porous crystalline materials consisting of inorganic building units or IBUs (metal ion or, more often, metal-oxo-clusters/metal-oxo-chains) that are bridged together by means of organic building units, also called ligands or linkers (bearing coordination functional groups such as carboxylates, phosphonates, azolates, etc.), through coordination bonds, leading to a periodic porous 2D or 3D structure (Fig. 1.1) [9, 10]. MOFs have aroused great interest by presenting numerous features that have shown considerable potential of efficiency enhancement in a variety of applications, when compared to conventional porous materials such as carbon, zeolites, clays, etc. Indeed, their highly modular crystalline porous structures (with tunable surface areas and pore sizes, pore geometries, and chemical tunability) make these versatile materials appealing for a wide range of potential applications related to the environment, energy, health, catalysis, sensing, electronics, gas adsorption, and separation [6–8]. Nevertheless, despite the evident and numerous advantages that these hybrid materials may offer for a plethora of potential applications, most MOFs still suffer from specific limitations that have slowed down their integration into industrial applications. Albeit substantial progresses over the last few years, stability, scalability, and sustainability as well as cost limitations remain important barriers that MOFs still shall overcome to reach industrialization. Indeed, their limited stability (chemical, thermal, or mechanical) in some cases and the use of expensive and/or toxic
1 Robust and Environmentally Friendly MOFs
3
Fig. 1.1 General overview summarizing the modularity and the versatility of the crystalline structures of MOFs
chemicals (e.g., precursors, organic solvents) together with the use of sometimes complex and/or energy-demanding synthetic conditions (e.g., solvothermal) are among the practical aspects that, for several years, have hampered the great potential of these materials. They have significantly contributed to consider MOFs as “unsuitable” for industrial developments in comparison with commercial activated carbons or zeolites, already in use in industry since decades. In this regard, significant efforts are devoted, nowadays, to unravel these issues. The discovery of MOFs with enhanced stability is a priority for many scientific groups. In parallel, developing more environmentally friendly scalable synthetic routes is also being increasingly favored. This book chapter provides an overview on the considerable advances that have been made in this line. It gives some insights on the strategies that have been followed to develop robust MOFs and greener synthetic routes, paving the way for MOFs toward higher levels of developments.
1.2
Stability of Metal-Organic Frameworks
The stability, also called robustness, is one the main feature that any material/ compound should present at least in the timeframe of defined operating conditions. Subsequently, it can be designated as a multifold feature (also depending on the required conditions), and its definition can be further divided into three main
4
R. Del Angel et al.
subcategories. Accordingly, one can associate to a given MOF a chemical, thermal, and/or mechanical stability. Nevertheless, even though the definition of stability seems to be simple and straightforward, the actual concept is, in reality, more subjective. Nowadays, there is no strict parameter/standard to delimit what can be considered, in absolute terms, as a chemically, thermally, or mechanically stable MOF. Consequently, the so-called stability of a MOF stands only under certain tested conditions or, in some cases, that a MOF is more stable relatively to another when tested under similar parameters. For these reasons, one shall consider a MOF as robust when it can retain its crystalline structure and properties (including porosity) after being exposed to the operation conditions required for a desired application. Assessing the stability of a MOF cannot rely on only one characterization technique, but it must be confirmed after corroborating the results of different analyses. Indeed, while X-ray diffraction (XRD) is a natural tool to assess the crystallinity of a structure, this is far from being enough. Other techniques are required such as N2 porosimetry to check if the material has maintained its porosity or not. Moreover, additional analysis such as infrared spectroscopy, solid-state NMR, and thermogravimetric analysis (TGA), especially when studying the thermal stability, can also be of great help. Besides, chemical analysis (i.e., elemental analysis, ICP) and chromatography (i.e., HPLC) can be, for instance, of high need to identify if the MOF degradation is very slow. Finally, electron microscopy is an important tool to verify that the particle morphology (size and shape) has not been impacted throughout the exposition of the MOF crystals during the practical use. Even if all these characterizations are in favor of the stability of the MOF after its exposition, one shall ultimately verify that the properties of the MOF are kept under operating conditions. In the following sections, a general overview will be given in terms of the aforementioned three subcategories of stability. Definitions for each type of stability will be drawn up together with an outline of structural features that can potentially confer an enhanced chemical, thermal, or mechanical stability to a MOF. It is worth mentioning, though, that the given stability cannot be absolute but depends on the complex interplay between the chemical natures of the building units, their arrangement (topology of the MOF), the shape and the size of the cavities, as well as a series of sub-factors such as the coordination geometry of the cation, the nuclearity and the connectivity of the IBU, the rigidity and the topicity of the linker, the hydrophobic/ hydrophilic character of the network (particularly with water), the redox behavior, the presence of open metal sites or defects, framework catenation, inter- and intramolecular interaction, the particle size, and so on.
1.2.1
Chemical Stability
The chemical stability of a MOF can be defined as the resistance of its structure to degradation (or coordination bond breakage) when exposed to a certain chemical
1 Robust and Environmentally Friendly MOFs
5
environment. This latter may include the presence of water, either as a liquid or as a vapor; other molecules or ions, corrosive or not, such as phosphates, carbonates, ammonia, SOx, NOx, etc.; or even different ranges of pH, temperatures, or concentrations. In general, the chemical stability of a MOF strongly depends on the strength of the iono-covalent bond that links organic and the inorganic parts of the structure. In other words, a stronger coordination bond between the organic linker and the metal oxo-cluster is expected to lead to the formation of a more chemically stable structure. This section will discuss, although in a non-exhaustive manner, the different approaches that have been followed to enhance the chemical stability of MOFs, with particular attention dedicated to the stability in aqueous media. It will be divided into two main categories: the first will focus on strategies allowing strengthening the interaction between the metal cation and the organic ligand, while the second will be showing alternatives aiming to preserve the coordination bonds.
1.2.1.1
Reinforcing the Coordination Bond
Because water (or humidity) is an essential element of our environment, this part will address strategies leading to enhanced chemical stability through examples mostly dealing with aqueous conditions (neutral, acidic, and basic). However, stabilities toward other species than water will be also highlighted in some cases [11–14]. Indeed, stability of MOFs under aqueous conditions (or hydrothermal stability) becomes of great importance for numerous applications, due to the fact that most of the processes (at room temperature or higher) take place in atmospheres where a certain degree of moisture (if not liquid water) cannot be prevented. This is, for instance, the case for water purification/desalination (the most prominent one) [15– 17], for gas/vapor separation (although corrosive contaminants are also present) [18, 19], or even for drug delivery and other biomedical applications, where the robustness toward humid environments under different pH and in presence of highly complexing species (i.e., phosphates, enzymes, etc.) becomes critical [20, 21]. The main reason for which a MOF would present a low stability under an aqueous environment is the existence of a competitive reaction between the linker (or the metal ions) and the water molecules or any other species present in it, such as hydroxide ions. This competition reaction can be easily explained considering the strength of the Lewis pair – metal ion as acid and ligand as base – that forms the link between the building units of the MOF. Subsequently, a hard base, such as a carboxylate group, exhibits a stronger bond with a hard acid (metal ion with high valence state) such as Zr4+, Ti4+, and Cr3+, rather than divalent cation such as Cu2+ or Zn2+. Similarly, low-valence metal ions (soft acids) such as Zn2+, Cu2+, and Ni2+ give stronger bonds when combined with soft ligands such as imidazolates, triazolates, or pyrazolates. One can give, for instance, the striking example of zinccarboxylate MOFs, such as IRMOFs, highly unstable in water, while Ni-pyrazolates MOFs remain intact after exposure to water or even very harsh conditions [22, 23].
6
R. Del Angel et al.
In fact, when MOF particles are in contact with water, this molecule enters into the pores of the MOF and gets in competition with organic linker, resulting in a possible break and hydrolysis of the metal-ligand bond. This phenomenon is even further amplified in a non-neutral pH. In a basic medium, hydroxide ions (hard base) present in the solution can easily react with high-valence metal ions, resulting in the replacement of the organic linker to produce metal oxides or hydroxides. In a similar scenario, in acidic media, MOFs formed via a low-valence metal and N-donor linker are expected to be more easily degraded due to the affinity between the azolate groups and the protons present in solution [24].
Acidic medium Basic medium
High-valence metal ion (M) + carboxylate (L) Higher resistance
Low-valence metal ion (M) + Ncontaining linker (L) MnL + H+ ! Mn + LH
MnL + nOH ! M(OH)n + L
Higher resistance
Moreover, one should also remind that the complexing ability of the functional group has also a considerable effect on the strength of the interaction: the higher the pKa, the stronger is the interaction and, hence, the less vulnerable the bond is. In this regard, three approaches have been followed (where at least one of the building units is showing high complexing potential) to reinforce the metal-ion/organic ligand interaction, hence enhancing the chemical stability: 1. The use of high-valence cations or high-charge density cations (directly related to the charge over the ionic radius of the metal ion) 2. The use of organic linkers bearing highly complexing functional groups (showing higher pKa values than that of carboxylic acids (~4.5)) 3. The combination, when chemically possible, of both cations of high oxidation states and highly complexing ligands Strengthening the interaction between the metal ion and the organic bridging ligand is however not without any consequence on the formation of MOFs. Indeed, a stronger interaction is associated with a faster kinetics of crystallization between the chemical species (ligands and solvents), hence decreasing the possibilities of a controlled MOFs nucleation and growth. To circumvent this problem, reactions combining high-valence cations and polycarboxylic acids are generally performed under slightly acidic conditions (such as in the presence of HF [25, 26] or HCl [27]). This is done in particular when the metal ion is highly oxophilic in order to limit the formation of metal oxides or hydroxides and to maintain a minimal concentration of metal cations needed for the formation of hybrid frameworks in solution. Moreover, the use of monotopic carboxylic acids [28] or more recently oxalates [29] has also shown to be very useful for a better control of the MOF crystallization. These molecules (generally monotopic ligands, but not limited to) are known as modulators (or inhibitors) that are not expected to bridge the IBUs but to compete with the polytopic linker helping in controlling and slowing down the nucleation-growth
1 Robust and Environmentally Friendly MOFs
7
processes and yielding to larger crystallites. One would also notice that strategies using less reactive precursors such as ester form of the ligand [30], metal oxides, metal hydroxides [31], and even pure metal [32] have been in some cases reported to be successful in controlling the crystallization process. Since the early 2000s, trivalent metal ions (such as Cr3+, Al3+, and Fe3+) have proven to be prone to yield relatively highly chemically stable MOFs suitable for many applications. Materials of Institute Lavoisier (MIL) compounds are prominent successful examples of this strategy. One can cite, for instance, the benchmark MIL-53 [22, 33–35] and MIL-100 [36–38] (M ¼ Al3+, Fe3+, or Cr3+) analogs that are built of benzene-1,4-dicarboxylate (terephthalate) and M-oxo-chains for the former, and M-oxo-trimers and benzene-1,3,5-tricarboxylate (trimesate) for the latter, showing 1D microporous channels or mesoporous cages, respectively (Fig. 1.2). They have shown very good stability in aqueous media, at different pH (ranging from 2 to 12 in the case of MIL-53(Cr)), and turned to be successful candidates for dehumidification or separation [39]. More recently, the Al-dicarboxylate MIL-160 has also proven to be a very promising material for application related to water [40]. It is constructed from Al-oxo-chains and furan2,5-dicarboxylate (Fig. 1.2), and its synthesis can be easily scaled up in green conditions and at high space time yields. Due to its very high stability in water and inherent properties toward water uptake, it has shown high promises for application in heat reallocation [41]. Other relatively hydro-chemically stable MOFs based on more extended ligands and/or higher topicities than the aforementioned ones have also emerged more recently [42–44]. This is, for instance, the case of MOFs constructed from mesotetra(4-carboxyphenyl)porphyrin (TCPP) and the Al-oxo-chains [45] or Fe-oxotrimer (labeled PCN-600) [46] that have been reported to be stable in water as well as into slightly acidic pH (~5) and pH values ranging from 2 to 11, respectively. Moreover, MIL-127(Fe) (also known as soc-MOF(Fe) or PCN-250), built up from Fe-oxo-trimers and 3,30 5,50 -azobenzenetetracarboxylate (ABTC), has been shown to be hydrothermally resistant [47–49]. It should be highlighted that chemical stabilities of MOFs based on trivalent metals can drastically drop from one analog to another. This is due to differences (i) in charge density (Z/r2) between the different ions (i.e., Al3+ vs In3+) [50], (ii) in kinetics of ligand exchange (i.e., 2.4 106 s1 for Cr(H2O)63+ vs 1.6 102 s1 for Fe(H2O)63+) [51, 52], and/or (iii) in their redox behaviors (for instance, Al3+ vs V3+) [34, 35, 47–49]. Kinetic crystalline phases can also have a lower chemical stability compared to thermodynamic ones. This can be exemplified by comparing MIL-88B (Fe)’s and MIL-101(Fe)’s both built up from Fe-oxo-trimers and dicarboxylates, where building units are arranged in different fashion. These phases tend to be produced using shorter synthesis times and/or lower temperatures and in some cases to undergo phase transition toward denser polymorphs such as MIL-53(Fe)’s [42]. One can also, on the whole, consider that for a given metal and IBU, the chemical stability depends on the number of coordinating groups per organic spacer; one representative example is the lower hydrolytic stability of metal dicarboxylates in comparison with tricarboxylates or tetracarboxylates that is observed for
Fig. 1.2 Selected M(III-IV)-carboxylate IBUs and MOFs: (a) MIII-oxo-trimer, (b) MIL-53’s MIIIoxo-chains, (c) MIL-160 MIII-oxo-chain, (d) ZrIV-hexanuclear oxo-cluster, (e) MIL-140’s ZrIVoxo-chain, and (f) TiIV-dodecanuclear oxo-cluster. Crystalline structures of (g) MIL-100, (h) MIL-53 (i) MIL-160(Al), (j) MIL-140A, (k) MIP-200, and (l) MIP-177-LT. Color code:
1 Robust and Environmentally Friendly MOFs
9
MIL-88B(Fe)’s and MIL-101(Fe)’s versus MIL-100(Fe) or MIL-127(Fe) (these two latter being even stable under hydrothermal conditions). This can be further associated with chemical and geometrical constraints such as the nuclearity of the IBUs (i.e., 0D vs 1D), the presence of “vulnerable” sites (i.e., open metal sites), MOF topology, pore dimensions, etc. Another considerable breakthrough in the quest for chemically stable MOFs has been accomplished with the discovery of the UiO-66 (UiO stands for University of Oslo), which exhibits a fcu-topology built of 12-connected hexanuclear ZrIV-oxocluster connected by means of terephthalate ligands [53]. Since then, a great number of MOFs containing Zr6-oxo-clusters, as well as analogous structures to UiO-66, have been produced. Markedly, to the tremendous number of Zr-carboxylate-based MOFs (such as the chemically stable NU-1000 and PCNs, MOF-801, MOF-808, and some of their analogs) [54], one can cite as highly chemically stable the recent MIP-200 (MIP stands for Materials of the Institute of porous materials from Paris) which presents a high potential for application in water-sorption-driven refrigeration (Fig. 1.2) [55]. Formulated [Zr6(μ3-O)4(μ3-OH)4(MDIP)2(formate)4]n, this MOF shows a csq-topology arising from the association of an eight-connected Zr6-oxocluster bridged together with a tetracarboxylate ligand (MDIP or 3,30 ,5,50 -tetracarboxylatediphenylmethane). The exposure of MIP-200 to a wide variety of extreme chemical conditions, such as boiling water, highly extremely acidic conditions, and even a buffer solution at pH 12, results in the retention of its crystalline structure and N2 sorption capacity, revealing its astonishing chemical resistance. Although much scarcer than Zr-carboxylate MOFs based on discrete IBUs, MOFs built of Zr-oxo-chains have been reported to exhibit even higher stability (chemical, mechanical). This is mainly the case of MIL-140’s family that can be built with a variety of linear dicarboxylate ligands (Fig. 1.2) [56–58]. In this specific case, in addition to the strength of the coordination bond, the fact that IBU is chain-like (rather than discrete) with inaccessible oxo-groups together with hydrophobic channels helps considerably enhancing the hydrothermal stability. TiIV is another prominent tetravalent ion that, when combined with carboxylate linkers, often results in the formation of chemically stable MOFs such as MIL-125 (based on Ti8 oxo-cluster and terephthalates) that have shown a very good stability in presence of water at room temperature [59]. However, the number of Ti-MOFs in literature still remains very limited, although an increasing number of examples is observed since the last 5 years [60–64]. The main issue that restricts the synthesis of more Ti-MOFs arises, in fact, not only from the strength of the Ti–O bond but also from the lack of control of the kinetic of crystallization and polycondensation processes. The high affinity between Ti(IV) and oxygen makes almost impossible the process of bond dissociation that is needed in the mechanism of MOF formation in order to get a crystalline solid. As a result, most of the synthesis attempts result in ⁄ Fig. 1.2 (continued) MIII ¼ Fe3+, Al3+, etc., dark gray; Zr, blue; Ti, blue-gray; C, gray; N, blue; O, red. Cages of MIL-100 are represented by colored spheres. Hydrogen atoms are not represented for the sake of clarity
10
R. Del Angel et al.
the formation of either titanium dioxide or powders with a very poor crystallinity [65]. In addition, among the few dozens of Ti-MOFs reported so far, almost all of them exhibit a different IBU, making the design of Ti-MOFs highly challenging [60, 66]. Recently, the synthesis of a highly stable three-dimensional Ti-MOF was reported, using MDIP as linker. MIL-177-LT (where LT refers to the pristine structure before the calcination step leading the MIL-177-HT), built of dodecanuclear Ti-oxo-clusters and MDIP ligands (Fig. 1.2), presents an outstanding chemical stability under extreme acidic treatments. The obtained results demonstrated no apparent damage to its crystallinity and porosity under boiling water, aqua regia at room temperature, and even under a 6 M H3PO4 solution at room temperature for 1 week. Regardless of the excellent stability of the material under acid conditions, 24 h in contact of a NaOH/NaHCO3 buffer (pH ¼ 10) solution at room temperature however start to damage the crystalline structure [67]. Moreover, involving high-valence cations (hard Lewis acids) and hard Lewis bases linkers bearing highly complexing functional groups has also proven to be very successful strategy for the synthesis of highly chemically stable MOFs. However, one should highlight here the even more pronounced difficulties in controlling the crystallization process of these materials. In this category, one can recall the examples of MOFs obtained through the association of ZrIV and highly complexing galates (or deprotonated trihydroxyphenyl) groups have been recently described [68–71] and show excellent chemical stability. For instance, MIL-163 (Fig. 1.3) [71], constructed from Zr-oxo-chains connected by means of TzGal (5,50 -(1,2,4,5tetrazine-3,6-diyl)bis(benzene-1,2,3-triolate)), can resist to boiling water for days. More interestingly, its crystalline structure can stand in presence of phosphate buffer solution for days, in contrast with Zr-carboxylate MOFs, where even the highly chemically stable MIL-140 can’t resist even for 1 day. The chain-like IBU free of oxo/hydroxo bridges together with the high chelate effect of the ligand and the high strength of Zr–O bonds can be some of the elements to rationalize this exceptional chemical stability. In addition, successful examples encompass also Zr-phosphonates MOFs such as UPG-1 [72, 73] and SZ-1-2-3 [74] that present remarkable stability in extremely harsh conditions (such as highly concentrated nitric acid or aqua regia in the case of SZ-3). This includes also Ti-phosphonates MOFs, reported in the late 1990s or early 2000s, that are also highly chemically stable. For instance, MIL-91(Ti) (Fig. 1.3), obtained through the association or TiIV-oxo-chains bridged by highly complexing ligands, N,N-piperazinebismethylphosphonates, exhibits an outstanding hydrothermal stability [75, 76]. Finally, scarce examples combining low-valent cations (soft Lewis acid) and highly complexing ligands have been reported, attesting, as expected, an enhanced chemical stability. Here, it should be highlighted that this is mainly the case of linkers based on soft ligands (more suitable to react with soft Lewis acids). Among them, one can cite MOFs based on divalent metal ions and azolate derivatives such as imidazolates – including zeolitic imidazolate frameworks (ZIFs) [77] and zeolitelike MOFs [78] that on the whole exhibit moderate to high tolerance to humidity
1 Robust and Environmentally Friendly MOFs
11
Fig. 1.3 Selected IBUs and MOFs based on highly complexing ligands: (a) TiIV-oxo-phosphonate chain, (b) ZIV-galates-chain, (c) octanuclear NiII-oxo-pyrazolate cluster, (d) MIL-91 and (e) MIL-163, (f) fcu-Ni(DP), (g) PCN-601. Color code: NiII, green; Zr, blue; Ti, blue-gray; C, gray; N, blue; O, red. Cages are represented by colored spheres. Hydrogen atoms are not represented for the sake of clarity
and/or water [79–81] – and triazolates [82, 83] that sometimes show relatively good chemical stability even in presence of halogens or highly corrosive NH3. Besides, metal pyrazolate-based MOFs [23, 84–89] are also a clear demonstration for remarkable stabilities not only in water but also in extremely basic conditions. This is, for instance, the case of MOFs built of 12-connected octanuclear NiII-oxo-
12
R. Del Angel et al.
cluster and linear di-pyrazolate ligands, such as fcu-Ni(DP) (Fig. 1.3), capable to withstand to water, harmful organic compounds, or SO2 [23, 89]. More recently, PCN-601, built of the same oxo-cluster assembled with square-shaped porphyrintetrapyrazolate ligands (Fig. 1.3), and its extended isoreticular analog (PCN-602) have also been reported to be remarkably stable in aqueous media, at basic pH and in presence of highly complexing species such as phosphates [85, 90].
1.2.1.2
Preserving the Coordination Bond
Besides the strengthening of the coordination bond between the IBU and the organic linker, there are other strategies in order to improve the chemical stability of MOFs. This relies on the use of a protective “shield” near the metal-ligand bond in order to limit the effect of a competitive reaction. This strategy can be helpful, for instance, when a given MOF shows high potential for a given particular application, but presents a limited chemical stability. Beyond the chemical stability, embedding MOFs in a (flexible) polymeric matrix can also provide additional mechanical stability (see Sect. 1.2.3). In both cases, the inherent stability of the MOF is not changing, but the degradation kinetics are sufficiently slowed down for the application in play because the MOF is more protected. This approach has been proven to be relevant to enhance the chemical resistance of MOF in presence of water/humidity. It mainly consists in increasing the hydrophobicity of the MOF, therefore limiting the hydrolysis of the framework. In this regard, two main approaches can be considered: 1. The introduction of bulky groups within the crystalline framework (such as hydrophobic groups to limit the diffusion of water molecules) 2. The incorporation of the MOF into a protecting matrix (such as coating the crystallites or simply embedding it into a (polymeric) support) The introduction of hydrophobic functional groups can be done at the beginning of the synthesis of the MOF (during its crystallization) by choosing the correct organic precursor, or in a subsequent step, as part of a post-synthetic modification. In both cases, the result of the modification should be the same: the presence of linkers with hydrophobic groups that can limit the access of water molecules into the framework. Fluorinated groups such as –CF3 or alkyl groups are the most common options. Note that in any case, the resulting materials are necessarily (much) more expensive and less scalable compared with the starting bare MOF due to the (often) high cost of the hydrophobic functional groups and/or the need of a post-synthetic treatment. Walton et al. studied the influence of the addition of different functional groups on the hydrothermal stability of a MOF. For this, [Zn(BDC)(DABCO)0.5] was chosen as a model case, where BDC corresponds to 1,4-benzenedicarboxylic acid and DABCO to 1,4-diazabicyclo[2.2.2]octane. Different functionalized BDC linkers were tested, arriving to the conclusion that polar functional groups such as nitro, bromo, and hydroxy groups would have a negative effect on the MOF stability in
1 Robust and Environmentally Friendly MOFs
13
comparison to the parent MOF, while nonpolar (hydrophobic) groups like methyl groups enhance the stability of the MOF in presence of humidity. This systematic study in which the topology and the porosity where maintained constant allowed to see the beneficial shielding effect that nonpolar functional groups have on the Zn-O coordination bond stability [91]. Similarly, MIL-88B(Fe)-(CF3)2 or UiO-66-(CF3)2 have shown an enhancement in hydrothermal stability compared with their bare analogs due to the presence of the hydrophobic fluorinated groups [92, 93]. Another interesting example is the case of phosphonate monoester-based MOFs where the presence of one ester group allows phosphonates to play a double role: providing to phosphonates a carboxylate-like coordination mode (only two over three oxo-groups will coordinate but with stronger interaction) and in the same time providing a shielding effect (due to the presence of alkyl groups on the third O-atom) enhancing the stability to moisture [94, 95]. Following a different approach, Farha et al. used the solvent-assisted ligand incorporation (SALI) method to attach perfluoroalkane carboxylates, as well as some other hydrophobic groups, into the channels of NU-1000. It was observed that the presence of the different functional carboxylates increased the hydrophobicity of the material, maintaining their crystallinity and surface area even after 20 cycles of water vapor adsorption-desorption. The main drawback of this method is the significant reduction of the pore size of the material due to the presence of functional groups that occupy the channels of the MOF. However, this reduction in pore size means also a reduction in the size of water clusters that can be formed inside the pores, therefore limiting the accessibility of water [96]. A post-synthetic functionalization approach was also used to exchange the monocarboxylic ligands coordinated to the metal oxo-clusters of DUT-67 by a series of different fluorinated monocarboxylates. All the prepared materials presented an increase in their hydrophobic character, evidenced by higher values of contact angles, especially when functionalized with perfluorooctanoic acid (110 ). At the same time, it was observed that only one type of pore was able to be functionalized, causing a change in the water adsorption mechanism when compared to the pristine material [97]. A very recent study has demonstrated the improvement of chemical stability, thanks to electrostatic interactions. It was demonstrated on isoreticular NiII-azolatebased MOFs that PCF-8, built up with 1,4-bis(4H-1,2,4-triazol-4-yl)benzene and showing a cationic skeleton (in comparison with a neutral analog PFC-9, built using the deprotonated form of 1,4-di(1H-pyrazol-4-yl)benzene), is resistant to extremely acidic, oxidative, reductive, and high ionic strength conditions such as 12 M HCl, aqua regia, H2O2, and seawater (30 days). This has been explained by the presence of repulsive interaction and steric hindrance toward the positively charged species, thus protecting the vulnerable dative bonds in structure [98]. The second approach for increasing water stability of MOFs is based on increasing the hydrophobicity of the material by the formation of a hydrophobic layer on the surface of the MOF. This coating, usually consisting of a polymeric material, is expected to protect the MOF from water without affecting the inner porosity (and the performance) of the pristine framework. In 2014, Zhang et al. worked on the external
14
R. Del Angel et al.
coating of MOF-5, HKUST-1, and ZnBT with a thin layer of hydrophobic polydimethylsiloxane (PDMS) by using a thermal vapor deposition technique. This method improved the water resistance of the MOFs by increasing their hydrophobicity, passing from a water contact angle of almost 0 to angles of 130 2 , while maintaining their porosity and surface area [99]. Since then, some other coating matrices have been tested as well as some other coating techniques. Park et al. reported the carbon-coating of IRMOF-1 [100], Maspoch et al. described the coating of HKUST-1 (Cu-BTC) with polystyrene using a spray-drying synthesis method [101], while, more recently, Xu et al. used the simple solution-immersion technique to deposit a hydrophobic layer of organosilicon on the surface of DUT-4 [102]. As an extension to this approach, the formation of a hybrid composite consisting of a mixture of a MOF and a second substrate which can result in better and enhanced properties for both materials has been tested. Substrates such as carbon nanotubes (CNTs), silica, clays [103], and graphite oxide (GO) have been tried for the formation of composites with metal-organic frameworks giving encouraging results. Park et al. prepared a hybrid composite by incorporating CNTs into MOF-5. The synthesis method consisted in the dispersion of the carbon nanotubes in a DMF solution containing the required precursor for the synthesis of MOF-5 followed by a heating step. The resulting MOF composite showed to be stable after 1 week in air with a relative humidity of 33%, while the pristine MOF-5 was almost completely decomposed [104]. Sun et al. reported that at the optimal incorporation of 8.7 wt % of graphite oxide in HKUST-1, both the hydrothermal stability of the material and the surface area were increased. The growth of the MOF was observed in-between the graphite layers, which enhanced the porosity. The hydrothermal stability of the sample was improved due the incorporation of the graphite oxide when exposed to water vapor at 363 K for 12 h. Nevertheless, at concentrations of graphite oxide higher than 34.3 wt %, the hydrothermal stability decreased again, due probably to the formation of distortions on the structure that create “easier targets” for the “attack” of water molecules [105]. On the other hand, Decoste et al. achieved the formation of a hydrophobic form of HKUST-1 by the chemical vapor deposition of perfluorohexane (PFH). Grand canonical Monte Carlo (GCMC) simulations showed that, even though the position of the PFH in the pores would not prevent the coordination of water molecules to the Cu atoms in the oxo-cluster, its hydrophobic nature could help preventing the formation of water clusters that are needed to break the Cu-BTC bond, ensuring the preservation of the sample in the presence of water [106].
1.2.2
Thermal Stability
Industrial applications where MOFs are used at high temperatures (one should keep in mind that for MOFs, high temperature should not be exceeding ca. 400 C) are typically catalysis or in some cases gas or vapor phase separation or other energy-
1 Robust and Environmentally Friendly MOFs
15
related applications. Besides, withstanding relatively high temperatures is, in some cases, important not only during the application itself but also for the activation and/or the regeneration process of the material. In this regard, the thermal stability – the capacity of the framework to maintain its crystalline structure and porosity at high temperatures – is a significant parameter to consider. The most common technique used to study the thermal stability of a given material is thermogravimetric analysis (TGA). Nevertheless, while the TGA of a sample can give important information directly related to its chemical composition (i.e., loss of guest molecules, dehydration of the oxo-cluster, loss of specific molecule) or even the temperature at which the calcination of the organic linker occurs, this technique finds some limitations by not being able to show the precise temperature at which the amorphization of the crystalline structure begins. In this case, a more extensive study of the thermal stability of a MOF can be achieved by using variable temperature X-ray diffraction (VT-XRD) as well as other typical characterization techniques (IR, NMR, porosimetry, SEM, etc.). VT-XRD is a powerful tool that can give important information about the structural changes that MOFs undergo at variable temperature, even though the conditions at which the measurement takes place may not faithfully represent the real working conditions. Various measurement conditions can be chosen, such as the temperature sweep rate, the scanning time, and the atmosphere (sealed conditions or under gas flux), not taking into account, though, that the MOF particle size (and the defect content) can also play an important role in the kinetics of thermal degradation, etc. All these factors make rather difficult the comparison of thermal results among different reported studies, which means that the information obtained by this technique must be analyzed with caution. Compared to the chemical stability of a MOF, the thermal properties of a framework are even more difficult to predict. However, some common rules can be found, giving an idea of the different aspects that may influence the robustness of the structure at relatively high temperatures. In summary, a higher stability is directly related to a stronger coordination bond between the Lewis acid and the Lewis base, which means that a stronger metal-ligand interaction is translated into a higher thermal stability. A dense packing will also produce the desired effect, whereas the presence of defects in framework usually results in a decrease of the thermal stability of the material. In fact, Lillerud et al. showed that by increasing the synthesis temperature, the number of missing-linker defects (sites of weakness) in UiO-66 can be decreased. Consequently, as shown in Fig. 1.4, the most “ideal” UiO-66 obtained presented the highest thermal stability by retaining its crystallinity until 450 C [107, 108]. Concerning the nature of the metal ions that build the MOF, it has been observed that a higher thermal stability can be obtained for materials containing metals at their most stable oxidation state. For example, MOFs containing Fe(III) and Cu(II) are more thermally stable than those based on Fe(II) [109, 110] and Cu(I) [111, 112]. Another important factor to be taken into consideration when analyzing the thermal stability is the nuclearity of the IBU. When comparing the thermal properties of a series of MIL-140, which consists of a long infinite zirconium oxide chain,
16
R. Del Angel et al.
Fig. 1.4 PXRD patterns (Cu-Kα radiation 1.5418 Å) recorded after heating UiO-66 samples for 12 h in air at various temperatures. The two reflections labeled “FR” are symmetry forbidden. The red curves correspond to a more “ideal” UiO-66, showing a higher thermal stability. (Figure reproduced with permission from reference [108]. Copyright (2014) American Chemical Society)
against the thermal stability of a series of UiO-66 materials (Zr-oxo-cluster), it can be observed that the formers present a relatively higher stability, preserving their crystallinity up to 500 C, while the materials formed by the Zr6O4(OH)4 oxo-clusters lose it around 450 C [58]. A similar trend was observed for MIL-125 with a Ti8-oxo-cluster against COK-69, consisting of Ti3-oxo-clusters. This observation leads to the conclusion that MOFs with IBU of higher nuclearity, including infinite chains, will present higher thermal stabilities. Indirectly, this enhanced thermal stability can be also related to condition in which the metal oxide (or the IBU) can be formed. The more thermodynamically stable is the IBU, the more thermally stable will be the corresponding MOF. Concerning the organic linkers that form the MOF, some tendencies have also been observed. When exposed to high temperatures in presence of oxygen, carboxylates and phenolates undergo decarboxylation and oxidation reactions, respectively, which makes these two kinds of linkers less suitable for the synthesis of a highly thermally stable MOF, although to a much lower extent for the carboxylates. In comparison, phosphonates and sulfates linkers present a certain chemical inertness that will be favorable their thermal stability [113–115]. It can lead to sensitively
1 Robust and Environmentally Friendly MOFs
17
high thermal stabilities such as the one of the Zr methylene bisphosphonate MIL-57 that is thermally stable up to 700 C under air atmosphere. Finally, when choosing the organic linker for a thermally stable MOF, short aromatic molecules are usually favored due to their rigidity and robustness in comparison to flexible aliphatic linkers. Indeed, it has been observed that the thermal stability is usually lower for these systems [116].
1.2.3
Mechanical Stability
The number of studies focusing on the mechanical stability of MOFs is considerably lower when compared to those aiming to enhance their chemical or thermal stability. Nevertheless, this type of stability becomes of great importance when a certain material needs to pass through a process associated with a mechanical stress such as compacting and shaping. This step is usually imperative before the material can be used in an industrial process and commercialized. The mechanical stability of MOFs refers also to the resistance of the material to collapse under vacuum or when in presence of a certain external pressure. Most of the first MOFs synthesized presented a relatively low mechanical stability, resulting in the collapse of the porous structure once the guest molecules were removed [117]. However, with time and after the development of more chemically stable MOFs, the mechanical properties of these materials have also seen great improvement. Even now, the activation step or the removal of guest molecules in the pores is still an important step to take into consideration. The collapse of the structure during the activation of a sample is due to the capillary forces created by the presence of solvent molecules trapped in the pores. Some values are summarized in Fig. 1.5. In order to prevent this problem, the molecules in the pores might be ideally exchanged with another solvent that presents a lower surface tension, or in an ideal situation, supercritical CO2 activation can be used to guaranty almost zero-capillary force [118, 119].
Fig. 1.5 Surface tension of different organic solvents (DCM dichloromethane, DMF N,Ndimethylformamide, DMSO dimethyl sulfoxide) compared to liquid CO2. (Figure reproduced with permission from reference [118]. Copyright (2017) American Chemical Society)
18
R. Del Angel et al.
The mechanical properties of a solid material can be evaluated using different parameters. Young’s modulus corresponds to the capacity of a solid to undergo compression in one direction, the shear modulus corresponds to the withstand to deformation when applying a force parallel to one face of a solid, Poisson’s ratio is the ratio of transverse strain to longitudinal strain in the direction of the applied force, while the bulk modulus represents the deformation of the material in all directions when a uniform pressure is applied. One or several of these parameters can be reported as a measurement for the mechanical stability of the MOFs. While it may be difficult to determine exactly what makes one MOF more mechanically stable than another, several studies along the years have provided some general tendencies. For this purpose, structural and topological factors must be considered. For example, the computational study performed by Coudert et al. showed that the presence of missing linkers in UiO-66 results in the decrease of its shear and bulk moduli, demonstrating the negative effect of defects in the mechanical stability of the material [120]. In general, MOFs with higher porosity tend to have a lower mechanical stability. It means that denser materials can present better mechanical properties. Numerous factors can influence the mechanical stability of a MOF. The coordination number, the functionalization [121] and structure of the organic ligand, the presence of guest molecules in the pores [122], the strength of the chemical bonds, and the topology [122] of the framework are some of the parameters that must be taken into account when studying the mechanical stability of a MOF. Zhou et al. predicted by the use of density functional theory (DFT) calculations that MOFs with higher connectivity between organic-inorganic nodes present a higher mechanical stability. In this light, UiO-66, a MOF based on 12-connected Zr6-oxo-clusters, will present higher bulk and shear moduli in comparison to other Zr6-oxo-cluster-based MOFs with lower network connections [123]. In the same study, the influence of the linker’s length was evaluated by comparing the same mechanical parameters calculated for isoreticular structures. Again, UiO-66 (linker with one-phenyl ring) presented a much higher mechanical stability compared to UiO-67 (linker with two-phenyl ring) and UiO-68 (linker with three-phenyl ring), evidencing the negative effect that longer organic linkers have in the mechanical properties of a MOF. As mentioned before, frameworks with larger pore sizes can be quite unstable. Subsequently, the insertion of bridging ligands that can split the bigger channels into smaller segments can be a plausible solution to enhance the rigidity of the MOF and its mechanical stability. Indeed, it is very rational that installing molecular struts, which help better holding the walls, will minimize the tension of the porous framework [124, 125]. For instance, Yaghi et al. achieved the improvement of the mechanical stability of MOF-520 [(Al8(μ-OH)8(HCOO)4(BTB)4)]n by what they call the “molecular retrofitting” of the material. In this method, an additional linker is incorporated into the framework in order to reinforce its structure. As shown for MOF-520 in Fig. 1.6, a molecule of 4,40 -biphenyldicarboxylate (BPDC) was integrated between two IBUs. This molecule acts as a “rafter” to further support the structure. The reinforced MOF-520-BPDC was stable to a pressure up to 5.5 GPa,
1 Robust and Environmentally Friendly MOFs
19
Fig. 1.6 Left: structure of pristine MOF-520. Right: retrofitted MOF-520-BPDC. BPDC girders are shown in green [126]
while its pristine counterpart (MOF-520) was destroyed at 2.8 GPa [126]. Similarly, self-catenated or interpenetrated structures are also expected to show enhanced mechanical stability [127]. In summary, a MOF will more likely exhibit a better mechanical stability when it possesses smaller pores, short linkers, strong linker-metal interactions, defect-free, and dense structures. Moreover, the flexibility (either of the linker or of the MOF framework and inducing structural phase transition without bond breakage) [128] is a nontrivial parameter that one should also keep in mind. Indeed, it is expected that the flexibility helps the material to support better the external stimuli (such as pressure). However, this complex aspect has not been discussed in this section focusing only on rigid structures (without phase transition).
1.3
Environmentally Friendly MOFs
Making a flawlessly environmentally friendly MOF is not an easy endeavor. Indeed, one of the most important targets should be to help driving MOF properties into real applications. In fact, when trying to lower the impact to the environment, several interconnected parameters are at play, all of which closely related to economic necessary compromises. The production of MOFs requires, as for other chemicals, various important aspects such as safety, health, environmental impacts, synthesis, and engineering processes (e.g., product recovery, purification, etc.) [129]. Moreover, it is important to consider different production scales for a given material, i.e.: 1. Lab scale: This is generally around the 100 mg to a few g scale. Here not much efforts are dedicated to optimizations in terms of energy used, chemicals toxicity, processes, etc. The goal is to obtain a pure and performant product, and efforts on fundamental science and understanding are the most important. 2. Kilo scale: This scale from a few hundred g up to a few kg is crucial, and the target is here to optimize the fabrication for large-scale purposes, prior to further industrial manufacture end targets. The goal is to reduce the cost of fabrication using chemicals that are safe, nontoxic, and with minimal impacts to the
20
R. Del Angel et al.
environment but also in most cases to produce enough MOF to integrate them into various process demonstration tests at the laboratory pilot scale; it often involves the shaping of the MOF powder to avoid any diffusion issues. Most of the improvement of the synthesis and the material recovery and purification are done in this step. 3. Pilot scale: This scale (from a few dozen kg to a few hundred kg) precedes the industrial scale where further improvement is made in the process engineering (e.g., solvent and unreacted precursor recirculation) and produces in all cases shaped MOFs. 4. Industrial scale: Normally most of the possible synthesis and engineering processes improvements are validated and the production exceeds the ton scale; at this stage note that industry shall overcome important regulation issues such as those associated with REACH or other economical perquisites, etc.
1.3.1
Chemicals
Lowering the impact to the environment requires a careful choice of the types and quantities of chemicals to be employed during the whole process of fabrication of the final material [130]. The common synthetic methods involve the use of various metal-ion precursors, organic linkers, solvents, and sometimes additives [131]. An environmentally friendly MOF should be built of components that have as low as possible impact on the ecological system during its utilization and/or after (recycling of the material).
1.3.1.1
Metal-Ion Precursors
Some metal ions that have been thoroughly used for the synthesis of MOFs and that can be accepted as safe for the environment are aluminum, iron, zirconium, titanium, copper, zinc, etc. [132]. However, these metals are mostly incorporated in the material via a reaction involving metal salts and linkers. Therefore, the choice of the metal source is crucial in the fabrication of the materials as the counter ions are not part of the materials. Sulfates and oxides metal sources are preferred. However, they are usually not so trivial to use generally because of the lower solubility of the reactants [131]. Chlorides and nitrates are more frequently employed as, for instance, their reactivity toward carboxylates or phosphonates is favorable. Nevertheless, they have to be avoided for safety and corrosion reasons and as a result are not favorable for large-scale production [131].
1 Robust and Environmentally Friendly MOFs
1.3.1.2
21
Linkers
If one analyzes the final cost of a MOF material, the part related to the linker is usually prominent. The fabrication process of these linkers is typically well established as in most cases commercially available ligands are already considered for other industrial applications. These molecules are often produced at large scale and their price is in some cases rather attractive. In addition, it is better if their fabrication process exhibits a minimal environmental impact. Thus, bio-sourced ligands or biomass-derived linkers are usually preferred due to their lower environmental impact [40, 131, 133]. In addition, these latter are usually more soluble in water and thus the related MOF synthesis is greener, which leads to a reduced final cost of the material. Larger and often costly ligands with poor solubilities in water or alcohols have therefore to be discarded as in most cases the corresponding MOF synthesis requires the use of toxic solvents and/or larger amounts of solvents (dilution) that negatively impact their production (cost, low space time yields, etc.) [55].
1.3.1.3
Solvent
The choice of the best solvent for a given MOF synthesis is not trivial. The optimal solvent should be safe, nontoxic, environmentally friendly, recyclable, and cheap. Although finding a solvent fulfilling these requirements is ideal, when one deals with lab-scale synthesis, the choice is more driven by other considerations such as (i) maximal solubilization of the precursors to enable an appropriate and successful metal-ligand reaction and (ii) a possible template effect to favor certain topological arrangements. Other solvents are also typically required during purification processes such as the washing and the so-called activation steps of the material. Unfortunately, lab-scale activation often requires larger solvent quantities than the synthesis itself. Reducing the volume of solvent is therefore of utmost importance to minimize safety, health, environmental, and cost issues [131, 134]. Solvent waste can indeed be a source of pollution itself via VOC emissions or as liquid during the various phases of the manufacture processes. Therefore, water is the best economic and environmental choice while having no negative impact on health and safety. Other preferred solvents are alcohols and esters. However, one shall take into account that the use of high boiling point solvents can have a negative impact on the environment because of the energy demand to recycle.
1.3.1.4
Additives
Additives are often used as inhibitors during the synthesis reactions since they have an effect on the metal-ligand reactions, acting as competitors for the MOF linker and therefore allowing a more controlled material synthesis. However, they are usually
22
R. Del Angel et al.
used in large excess which raises economic issues even if greener ones are selected [40, 133, 135]. Additives are also used to improve the solubility of the linkers by promoting the formation of linkers salts. Many reports on MOF synthesis involve the use of inorganic or organic bases however in a stoichiometric amount. Typical bases are inorganic ones such as NaOH or KOH that have a minimized negative environmental impact. However, when using organic solvents, organic bases are considered, and, therefore, attention should be given to select bases (e.g., amines, etc.) that possess a minimum effect on safety, health, and environment. Thus, as general rule, it is preferable to use water, alcohols, or esters with inorganic bases to promote the solubility than using a toxic solvent with higher solubilizing capabilities.
1.3.2
Synthesis and Purification Processes
The preparation of MOFs involves mostly two steps, i.e., the fabrication of the material in its raw form so-called as-synthesized form and a purification process in which the material is cleaned of any unreacted precursors and the pore content is exchanged with a clean, preferably low boiling point and as green as possible solvent.
1.3.2.1
Synthesis Process
In a typical MOF reaction preparation, precursors are mixed together in order to promote metal-ligand bond formation. Batch- and flow-type synthesis can be distinguished. However, the great majority of the reported material syntheses use batchtype methods. This class of preparation involves a closed vessel in which precursors are mixed together using a single solvent or a mixture. The reaction takes place in an auto-generated pressure environment, i.e., at temperatures exceeding the boiling point of the solvent, or at ambient pressure at temperatures below the boiling point. For economical and safety reasons, ambient pressure synthesis is clearly advantageous, and is therefore a method of choice for large-scale production [131]. As of today, this is the only reported method for an industrial batch ton-scale production of an Al-based MOF (Basolite A520) synthesized in water, with environmentally friendly and economically favorable precursors and a space time yield as high as 3600 kg/m3/day [131]. Among these types of processes (batch and flow), some methods can also be of great economic and environmental interest. For instance, microwave (hydro/solvothermal) synthesis is a very attractive method and is often used to produce small particle size with a minimum polydispersity. The heating process is rapid and very efficient in terms of thermal energy loss and is therefore environmentally and economically favorable. This mode of synthesis is beneficial for fast and efficient heating. It enables the uniform formation of crystallization nuclei in the whole reaction mixture. Crystal growth is therefore controlled, and the entire MOF formation process involves very short time. However, the safety
1 Robust and Environmentally Friendly MOFs
23
and the high cost of the instrumentation are limiting a possible industrial-scale production. One interesting though less used method is the electrochemical synthesis. Even if it often requires organic solvents, this type of synthesis can be very advantageous as it is a metal salt-free synthesis. The materials can also be obtained in shorter times and a continuous synthesis mode can be envisaged [136, 137]. In contrast to batch-type syntheses, flow-type process is a very attractive method to scale up MOFs as associated with a continuous production together with a very high space time yield and a minimal solvent use. It can therefore be of a high economic and environmental value and thus promising for MOF industrialization. Numerous reports have devoted to date efforts to produce various MOFs following this route [138]. One can consider extrusion as a successful method to produce through thermomechanical effect the large-scale production of MOFs associated with high space time yields [139]. However, as this method involves short reaction contact times as well as low mixing efficiency, precursors poor solubilities can be an issue. Therefore, this method still needs further optimization to be applied to a larger variety of MOFs. Another attractive method is spray drying; this allows the synthesis and shaping of small objects at the same time [140, 141]. This is therefore an environmentally and economically attractive method. High space time yields may be obtained using as green as possible precursors and solvent, making this process a method of choice for the industrial preparation of shaped MOFs. As for the extrusion method, one cannot however produce any type of MOFs using this technique. More generally, solvent-free-type syntheses (or minimal solvent amount) are indeed very favorable in terms of environmental impact and cost of production [142]. However, these methods still need further developments to include more industrially relevant materials as well as purification of the crude materials that often requires the use of solvent.
1.3.2.2
Purification Processes
Purification processes in MOF synthesis have not, to date, been well addressed. These should however not be neglected because of the time allocated and the expensive necessary materials such as the important quantities of solvent required to perform all the necessary steps. Therefore, the final space time yield of the entire processes (synthesis and purification) can be greatly reduced. Typical purification steps comprise filtration or centrifugation when the MOF particle size is small associated with clogging effects. The impact on the environment is therefore not negligible and could even be worse than the synthesis process itself. Reducing or eliminating these steps when possible can be a remarkable step toward a greener and cheaper manufacturing cost. Purification processes include recuperation of the material as a powder and elimination of any impurities from the as-synthesized material (i.e., unreacted precursors, solvents, or additives). For mostly economic but also environmental reasons, a chemical recirculation process shall be in place when addressing scale-up of materials. An ideal system would be a flow-type water-based and ambient pressure synthesis process with highly soluble and green precursors
24
R. Del Angel et al.
along with a maximized yield and minimum unreacted precursors that require little to no extra purification steps. When dealing with a typical batch synthesis process, one shall however consider the use of a precursor recirculation process to avoid any chemical loss.
1.4
Concluding Remarks
It is now clear that the era of MOFs has substantially evolved over the last decade demonstrating the great potential of this new family of materials in addressing serious societal and industrial challenges. If achieving chemically stable MOFs seems to follow clear guidelines, thermal and mechanical stabilities are still less rationalized. Though, whatever is the required stability for each application, the robustness of MOFs is governed by a complex interplay between chemical and structural parameters that one can tune to achieve a satisfying compromise. Moreover, it is worth mentioning here that stability assessment is now needed to be more widened and not be limited to only one or two parameters. Indeed, it is essential now to look into a more global picture and take into consideration the overall operating conditions including the time scale of the targeted process. In this regard, serious efforts are still needed. To our opinion, this task cannot be achieved without a cooperation between industrial partners and academic researchers from different horizons, to first identify the process “real” needs and then evaluate the MOF stability at the micro- and macroscopic levels. Besides, novel routes toward greener, more sustainable, and/or more economically viable synthesis have also started to emerge. Rising startups are also contributing to pave the way of MOFs to reach the doors of industrialization production. If routes to achieve more environmentally friendly MOFs can be successfully developed, the cost (mainly, of the organic raw materials) remains still one of the most challenging parameters that can be very probably overcame only when a real commercialization strategy for a wide potential market will be defined. To conclude, it is obvious that the achieved progress in terms of robustness and respect of environmentally friendly syntheses is promoting MOFs closer to realworld applications. Though, one should also highlight that in order to be considered as socioeconomically viable for a given application, the superiority in terms of performances of these relatively new materials is expected to be significantly higher (i.e., by at least one order of magnitude) compared to existing solutions. This being said, addressing novel challenges (such new fields of application, niche applications, new criteria of updated regulations, etc.) can be also a wise alternative. Acknowledgments RDA, GM, and CS would like to acknowledge CONACYT (grant number 2018-000003-01EXTF-00075) for the financial support.
1 Robust and Environmentally Friendly MOFs
25
References 1. Liu PS, Chen GF (2014) General introduction to porous materials. In: Liu PS, Chen GF (eds) Porous mater. Butterworth-Heinemann, Boston, pp 1–20 2. Abrahams BF, Hoskins BF, Robson R (1990) A honeycomb form of cadmium cyanide. A new type of 3D arrangement of interconnected rods generating infinite linear channels of large hexagonal cross-section. J Chem Soc Chem Commun 1:60–61 3. Abrahams BF, Hoskins BF, Liu J et al (1991) The archetype for a new class of simple extended 3D honeycomb frameworks. The synthesis and X-ray crystal structures of Cd(CN)5/3(OH)1/ 3.1/3(C6H12N4), Cd(CN)2.1/3(C6H12N4), and Cd(CN)2.2/3H2O.tBuOH (C6H12N4 ¼ Hexamethylenetetramine) revealing two topologically equivalent but geometrically different frameworks. J Am Chem Soc 113(8):3045–3051 4. Rosi NL, Eckert J, Eddaoudi M et al (2003) Hydrogen storage in microporous metal-organic frameworks. Science 300(5622):1127–1129 5. Millange F, Serre C, Férey G (2002) Synthesis, structure determination and properties of MIL-53as and MIL-53ht: the first CrIII hybrid inorganic–organic microporous solids: CrIII(OH){O2C–C6H4–CO2}{HO2C–C6H4–CO2H}x. Chem Commun 8:822–823 6. Zhou HC, Long JR, Yaghi OM (2012) Introduction to metal-organic frameworks. Chem Rev 112(2):673–674 7. Zhou HCJ, Kitagawa S (2014) Metal-organic frameworks (MOFs). Chem Soc Rev 43 (16):5415–5418 8. Maurin G, Serre C, Cooper A et al (2017) The new age of MOFs and of their porous-related solids. Chem Soc Rev 46(11):3104–3107 9. Schoedel A, Li M, Li D et al (2016) Structures of metal-organic frameworks with rod secondary building units. Chem Rev 116(19):12466–12535 10. Kalmutzki MJ, Hanikel N, Yaghi OM (2018) Secondary building units as the turning point in the development of the reticular chemistry of MOFs. Sci Adv 4(10):eaat9180 11. Barea E, Montoro C, Navarro JAR (2014) Toxic gas removal-metal-organic frameworks for the capture and degradation of toxic gases and vapours. Chem Soc Rev 43(16):5419–5430 12. DeCoste JB, Peterson GW (2014) Metalorganic frameworks for air purification of toxic chemicals. Chem Rev 114(11):5695–5727 13. Bobbitt NS, Mendonca ML, Howarth AJ et al (2017) Metal-organic frameworks for the removal of toxic industrial chemicals and chemical warfare agents. Chem Soc Rev 46 (11):3357–3385 14. Horcajada P, Gref R, Baati T et al (2012) Metal–organic frameworks in biomedicine. Chem Rev 112(2):1232–1268 15. Wang C, Cheng P, Yao Y et al (2020) In-situ fabrication of nanoarchitectured MOF filter for water purification. J Hazard Mater 392:1–7 16. Lin S, Song Z, Che G et al (2014) Adsorption behavior of metal-organic frameworks for methylene blue from aqueous solution. Microporous Mesoporous Mater 193:27–34 17. Haque E, Jun JW, Jhung SH (2011) Adsorptive removal of methyl orange and methylene blue from aqueous solution with a metal-organic framework material, iron terephthalate (MOF-235). J Hazard Mater 185(1):507–511 18. Zhao X, Wang Y, Li DS et al (2018) Metal–organic frameworks for separation. Adv Mater 30 (37):1–34 19. Xue DX, Belmabkhout Y, Shekhah O et al (2015) Tunable rare Earth fcu-MOF platform: access to adsorption kinetics driven gas/vapor separations via pore size contraction. J Am Chem Soc 137(15):5034–5040 20. Wu MX, Yang YW (2017) Metal–Organic Framework (MOF)-based drug/cargo delivery and cancer therapy. Adv Mater 29(23):1–20 21. Wang JL, Wang XY, Wang YH et al (2020) Room-temperature preparation of coordination polymers for biomedicine. Coord Chem Rev 411:213256
26
R. Del Angel et al.
22. Low JJ, Benin AI, Jakubczak P et al (2009) Virtual high throughput screening confirmed experimentally: porous coordination polymer hydration. J Am Chem Soc 131 (43):15834–15842 23. Padial NM, Quartapelle Procopio E, Montoro C et al (2013) Highly hydrophobic isoreticular porous metal-organic frameworks for the capture of harmful volatile organic compounds. Angew Chem Int Ed 52(32):8290–8294 24. Yuan S, Feng L, Wang K et al (2018) Stable metal–organic frameworks: design, synthesis, and applications. Adv Mater 30(37):1–35 25. Serre C, Millange F, Thouvenot C, et al (2002) Very large breathing effect in the first nanoporous chromium(III)-based solids: MIL-53 or CrIII(OH)‚{O2C-C6H4-CO2} ‚{HO2C-C6H4-CO2H}x‚H2Oy. J Am Chem Soc 124(45):13519–13526 26. Férey G, Mellot-Draznieks C, Serre C et al (2005) A chromium terephthalate-based solid with unusually large pore volumes and surface area. Science 309(5743):2040–2042 27. Wiersum AD, Soubeyrand-Lenoir E, Yang Q et al (2011) An evaluation of UiO-66 for gas-based applications. Chem Asian J 6(12):3270–3280 28. Schaate A, Roy P, Godt A et al (2011) Modulated synthesis of Zr-based metal-organic frameworks: from nano to single crystals. Chem Eur J 17(24):6643–6651 29. Canossa S, Gonzalez-Nelson A, Shupletsov L et al (2020) Overcoming crystallinity limitations of aluminium metal-organic frameworks by oxalic acid modulated synthesis. Chem Eur J 26 (16):3564–3570 30. García Márquez A, Demessence A, Platero-Prats AE et al (2012) Green microwave synthesis of MIL-100(Al, Cr, Fe) nanoparticles for thin-film elaboration. Eur J Inorg Chem 100 (32):5165–5174 31. Zhan G, Zeng HC (2017) Alternative synthetic approaches for metal-organic frameworks: transformation from solid matters. Chem Commun 53(1):72–81 32. Nouar F, Devic T, Chevreau H et al (2012) Tuning the breathing behaviour of MIL-53 by cation mixing. Chem Commun 48(82):10237–10239 33. Hu Y, Dong X, Nan J et al (2011) Metal-organic framework membranes fabricated via reactive seeding. Chem Commun 47(2):737–739 34. Zhao X, Liu D, Huang H et al (2014) The stability and defluoridation performance of MOFs in fluoride solutions. Microporous Mesoporous Mater 185:72–78 35. Kang IJ, Khan NA, Haque E et al (2011) Chemical and thermal stability of isotypic metalorganic frameworks: effect of metal ions. Chem Eur J 17(23):6437–6442 36. Küsgens P, Rose M, Senkovska I et al (2009) Characterization of metal-organic frameworks by water adsorption. Microporous Mesoporous Mater 120(3):325–330 37. Cunha D, Ben Yahia M, Hall S et al (2013) Rationale of drug encapsulation and release from biocompatible porous metal-organic frameworks. Chem Mater 25(14):2767–2776 38. Yoon JW, Seo YK, Hwang YK et al (2010) Controlled reducibility of a metal-organic framework with coordinatively unsaturated sites for preferential gas sorption. Angew Chem Int Ed 49(34):5949–5952 39. Seo YK, Yoon JW, Lee JS et al (2012) Energy-efficient dehumidification over hierarchically porous metal-organic frameworks as advanced water adsorbents. Adv Mater 24(6):806–810 40. Cadiau A, Lee JS, Damasceno Borges D et al (2015) Design of hydrophilic metal organic framework water adsorbents for heat reallocation. Adv Mater 27(32):4775–4780 41. Permyakova A, Skrylnyk O, Courbon E et al (2017) Synthesis optimization, shaping, and heat reallocation evaluation of the hydrophilic metal–organic framework MIL-160(Al). ChemSusChem 10(7):1419–1426 42. Horcajada P, Chevreau H, Heurtaux D et al (2014) Extended and functionalized porous iron (III) tri- or dicarboxylates with MIL-100/101 topologies. Chem Commun 50(52):6872–6874 43. Halis S, Reimer N, Klinkebiel A et al (2015) Four new Al-based microporous metal-organic framework compounds with MIL-53-type structure containing functionalized extended linker molecules. Microporous Mesoporous Mater 216(2):13–19
1 Robust and Environmentally Friendly MOFs
27
44. Feng D, Liu TF, Su J et al (2015) Stable metal-organic frameworks containing single-molecule traps for enzyme encapsulation. Nat Commun 6(1):5979 45. Fateeva A, Chater PA, Ireland CP et al (2012) A water-stable porphyrin-based metal-organic framework active for visible-light photocatalysis. Angew Chem Int Ed 51(30):7440–7444 46. Wang K, Feng D, Liu TF et al (2014) A series of highly stable mesoporous metalloporphyrin Fe-MOFs. J Am Chem Soc 136(40):13983–13986 47. Dhakshinamoorthy A, Alvaro M, Chevreau H et al (2012) Iron(III) metal-organic frameworks as solid Lewis acids for the isomerization of α-pinene oxide. Cat Sci Technol 2(2):324–330 48. Chevreau H, Permyakova A, Nouar F et al (2016) Synthesis of the biocompatible and highly stable MIL-127(Fe): from large scale synthesis to particle size control. CrystEngComm 18 (22):4094–4101 49. Wongsakulphasatch S, Nouar F, Rodriguez J et al (2015) Direct accessibility of mixed-metal (III/II) acid sites through the rational synthesis of porous metal carboxylates. Chem Commun 51(50):10194–10197 50. Devic T, Serre C (2014) High valence 3p and transition metal based MOFs. Chem Soc Rev 43 (16):6097–6115 51. Helm L, Merbach AE (2005) Inorganic and bioinorganic solvent exchange mechanisms. Chem Rev 105(6):1923–1959 52. Erras-Hanauer H, Clark T, Van Eldik R (2003) Molecular orbital and DFT studies on water exchange mechanisms of metal ions. Coord Chem Rev 238:233–253 53. Cavka JH, Jakobsen S, Olsbye U et al (2008) A new zirconium inorganic building brick forming metal organic frameworks with exceptional stability. J Am Chem Soc 130 (42):13850–13851 54. Bai Y, Dou Y, Xie LH et al (2016) Zr-based metal-organic frameworks: design, synthesis, structure, and applications. Chem Soc Rev 45(8):2327–2367 55. Wang S, Lee JS, Wahiduzzaman M et al (2018) A robust large-pore zirconium carboxylate metal–organic framework for energy-efficient water-sorption-driven refrigeration. Nat Energy 3(11):985–993 56. Butova VV, Budnyk AP, Charykov KM et al (2019) Water as a structure-driving agent between the UiO-66 and MIL-140A metal-organic frameworks. Chem Commun 55 (7):901–904 57. Zhang Q, Wahiduzzaman M, Wang S et al (2019) Multivariable sieving and hierarchical recognition for organic toxics in nonhomogeneous channel of MOFs. Chem 5(5):1337–1350 58. Guillerm V, Ragon F, Dan-Hardi M et al (2012) A series of isoreticular, highly stable, porous zirconium oxide based metal-organic frameworks. Angew Chem Int Ed 51(37):9267–9271 59. Dan-Hardi M, Serre C, Frot T et al (2009) A new photoactive crystalline highly porous titanium(IV) dicarboxylate. J Am Chem Soc 131(31):10857–10859 60. Wang S, Reinsch H, Heymans N et al (2020) Toward a rational design of titanium metalorganic frameworks. Matter 2(2):440–450 61. Wang C, Liu C, He X et al (2017) A cluster-based mesoporous Ti-MOF with sodalite supercages. Chem Commun 53(85):11670–11673 62. Keum Y, Park S, Chen YP et al (2018) Titanium-carboxylate metal-organic framework based on an unprecedented Ti-oxo chain cluster. Angew Chem Int Ed 57(45):14852–14856 63. Li C, Xu H, Gao J et al (2019) Tunable titanium metal-organic frameworks with infinite 1D Ti-O rods for efficient visible-light-driven photocatalytic H2 evolution. J Mater Chem A 7 (19):11928–11933 64. Assi H, Mouchaham G, Steunou N et al (2017) Titanium coordination compounds: from discrete metal complexes to metal-organic frameworks. Chem Soc Rev 46(11):3431–3452 65. Bosch M, Yuan S, Zhou H-C (2016) Group 4 metals as secondary building units: Ti, Zr, and Hf-based MOFs. In: Kaskel S (ed) The chemistry of metal–organic frameworks: synthesis, characterization, and applications, 1st edn. Wiley-VCH, New York, pp 137–141 66. Yuan S, Qin JS, Xu HQ et al (2018) [Ti8Zr2O12(COO)16] Cluster: an ideal inorganic building unit for photoactive metal-organic frameworks. ACS Cent Sci 4(1):105–111
28
R. Del Angel et al.
67. Wang S, Kitao T, Guillou N et al (2018) A phase transformable ultrastable titaniumcarboxylate framework for photoconduction. Nat Commun 9(1):1660 68. Chen EX, Qiu M, Zhang YF et al (2018) Acid and base resistant zirconium polyphenolatemetalloporphyrin scaffolds for efficient CO2 photoreduction. Adv Mater 30(2):1704388 69. Chen EX, Xu G, Lin Q (2019) Robust porphyrin-spaced zirconium pyrogallate frameworks with high proton conduction. Inorg Chem 58(6):3569–3573 70. Mouchaham G, Abeykoon B, Giménez-Marqués M et al (2017) Adaptability of the metal(III, IV) 1,2,3-trioxobenzene rod secondary building unit for the production of chemically stable and catalytically active MOFs. Chem Commun 53(54):7661–7664 71. Mouchaham G, Cooper L, Guillou N et al (2015) A robust infinite zirconium phenolate building unit to enhance the chemical stability of Zr MOFs. Angew Chem Int Ed 54 (45):13297–13301 72. Taddei M, Costantino F, Marmottini F et al (2014) The first route to highly stable crystalline microporous zirconium phosphonate metal-organic frameworks. Chem Commun 50 (94):14831–14834 73. Taddei M, Costantino F, Vivani R et al (2014) The use of a rigid tritopic phosphonic ligand for the synthesis of a robust honeycomb-like layered zirconium phosphonate framework. Chem Commun 50(43):5737–5740 74. Zheng T, Yang Z, Gui D et al (2017) Overcoming the crystallization and designability issues in the ultrastable zirconium phosphonate framework system. Nat Commun 8(1):1–11 75. Serre C, Groves JA, Lightfoot P et al (2006) Synthesis, structure and properties of related microporous N,N0 -piperazinebismethylenephosphonates of aluminum and titanium. Chem Mater 18(6):1451–1457 76. Benoit V, Pillai RS, Orsi A et al (2016) MIL-91(Ti), a small pore metal-organic framework which fulfils several criteria: an upscaled green synthesis, excellent water stability, high CO2 selectivity and fast CO2 transport. J Mater Chem A 4(4):1383–1389 77. Wang B, Côté AP, Furukawa H et al (2008) Colossal cages in zeolitic imidazolate frameworks as selective carbon dioxide reservoirs. Nature 453(7192):207–211 78. Eddaoudi M, Sava DF, Eubank JF et al (2015) Zeolite-like metal-organic frameworks (ZMOFs): design, synthesis, and properties. Chem Soc Rev 44(1):228–249 79. Banerjee R, Phan A, Wang B et al (2008) High-throughput synthesis of zeolitic imidazolate frameworks and application to CO2 capture. Science 319(5865):939–944 80. Park KS, Ni Z, Côté AP et al (2006) Exceptional chemical and thermal stability of zeolitic imidazolate frameworks. Proc Natl Acad Sci U S A 103(27):10186–10191 81. Zhang J-P, Zhang Y-B, Lin J-B et al (2012) Metal azolate frameworks: from crystal engineering to functional materials. Chem Rev 112(2):1001–1033 82. Rieth AJ, Tulchinsky Y, Dincă M (2016) High and reversible ammonia uptake in mesoporous azolate metal-organic frameworks with open Mn, Co, and Ni sites. J Am Chem Soc 138 (30):9401–9404 83. Tulchinsky Y, Hendon CH, Lomachenko KA et al (2017) Reversible capture and release of Cl2 and Br2 with a redox-active metal-organic framework. J Am Chem Soc 139 (16):5992–5997 84. Colombo V, Galli S, Choi HJ et al (2011) High thermal and chemical stability in pyrazolatebridged metal-organic frameworks with exposed metal sites. Chem Sci 2(7):1311–1319 85. Wang K, Lv XL, Feng D et al (2016) Pyrazolate-based porphyrinic metal-organic framework with extraordinary base-resistance. J Am Chem Soc 138(3):914–919 86. Burtch NC, Jasuja H, Walton KS (2014) Water stability and adsorption in metal-organic frameworks. Chem Rev 114(20):10575–10612 87. Canivet J, Fateeva A, Guo Y et al (2014) Water adsorption in MOFs: fundamentals and applications. Chem Soc Rev 43(16):5594–5617 88. Masciocchi N, Galli S, Colombo V et al (2010) Cubic octanuclear Ni(II) clusters in highly porous polypyrazolyl-based materials. J Am Chem Soc 132(23):7902–7904
1 Robust and Environmentally Friendly MOFs
29
89. Rodríguez-Albelo LM, López-Maya E, Hamad S et al (2017) Selective sulfur dioxide adsorption on crystal defect sites on an isoreticular metal organic framework series. Nat Commun 8 (1):14457–14466 90. Lv X, Wang K, Wang B et al (2017) A base-resistant metalloporphyrin MOF for C–H bond halogenation. J Am Chem Soc 139(1):211–217 91. Jasuja H, Huang YG, Walton KS (2012) Adjusting the stability of metal-organic frameworks under humid conditions by ligand functionalization. Langmuir 28(49):16874–16880 92. Devic T, Horcajada P, Serre C et al (2010) Functionalization in flexible porous solids: effects on the pore opening and the host-guest interactions. J Am Chem Soc 132(3):1127–1136 93. Planchais A, Devautour-Vinot S, Salles F et al (2014) A joint experimental/computational exploration of the dynamics of confined water/Zr-based MOFs systems. J Phys Chem C 118 (26):14441–14448 94. Iremonger SS, Liang J, Vaidhyanathan R et al (2011) Phosphonate monoesters as carboxylatelike linkers for metal organic frameworks. J Am Chem Soc 133(50):20048–20051 95. Gelfand BS, Lin JB, Shimizu GKH (2015) Design of a humidity-stable metal-organic framework using a phosphonate monoester ligand. Inorg Chem 54(4):1185–1187 96. Deria P, Chung YG, Snurr RQ et al (2015) Water stabilization of Zr6-based metal-organic frameworks via solvent-assisted ligand incorporation. Chem Sci 6(9):5172–5176 97. Drache F, Bon V, Senkovska I et al (2016) Postsynthetic inner-surface functionalization of the highly stable zirconium-based metal-organic framework DUT-67. Inorg Chem 55 (15):7206–7213 98. Huang G, Yang L, Yin Q et al (2020) A comparison of two isoreticular metal–organic frameworks with cationic and neutral skeletons: stability, mechanism, and catalytic activity. Angew Chem 132(11):4415–4420 99. Zhang W, Hu Y, Ge J et al (2014) A facile and general coating approach to moisture/waterresistant metal-organic frameworks with intact porosity. J Am Chem Soc 136 (49):16978–16981 100. Yang SJ, Park CR (2012) Preparation of highly moisture-resistant black-colored metal organic frameworks. Adv Mater 24(29):4010–4013 101. Carné-Sánchez A, Stylianou KC, Carbonell C et al (2015) Protecting metal-organic framework crystals from hydrolytic degradation by spray-dry encapsulating them into polystyrene microspheres. Adv Mater 27(5):869–873 102. Qian X, Zhang R, Chen L et al (2019) Surface hydrophobic treatment of water-sensitive DUT-4 metal–organic framework to enhance water stability for hydrogen storage. ACS Sustain Chem Eng 7(19):16007–16012 103. Yuan B, Yin XQ, Liu XQ et al (2016) Enhanced hydrothermal stability and catalytic performance of HKUST-1 by incorporating carboxyl-functionalized attapulgite. ACS Appl Mater Interfaces 8(25):16457–16464 104. Yang SJ, Choi JY, Chae HK et al (2009) Preparation and enhanced hydrostability and hydrogen storage capacity of CNT@MOF-5 hybrid composite. Chem Mater 21(9):1893–1897 105. Zu DD, Lu L, Liu XQ et al (2014) Improving hydrothermal stability and catalytic activity of metal-organic frameworks by graphite oxide incorporation. J Phys Chem C 118 (34):19910–19917 106. Decoste JB, Peterson GW, Smith MW et al (2012) Enhanced stability of Cu-BTC MOF via perfluorohexane plasma-enhanced chemical vapor deposition. J Am Chem Soc 134 (3):1486–1489 107. Bárcia PS, Guimarães D, Mendes PAP et al (2011) Reverse shape selectivity in the adsorption of hexane and xylene isomers in MOF UiO-66. Microporous Mesoporous Mater 139 (1–3):67–73 108. Shearer GC, Chavan S, Ethiraj J et al (2014) Tuned to perfection: ironing out the defects in metal-organic framework UiO-66. Chem Mater 26(14):4068–4071
30
R. Del Angel et al.
109. Wang Z, Zhang Y, Liu T et al (2007) [Fe3(HCOO)6]: a permanent porous diamond framework displaying H2/N2 adsorption, guest inclusion, and guest-dependent magnetism. Adv Funct Mater 17(9):1523–1536 110. Weber B, Betz R, Bauer W et al (2011) Crystal structure of iron(II) acetate. Z Anorg Allg Chem 637(1):102–107 111. Liu W, Fang Y, Wei GZ et al (2015) A family of highly efficient CuI-based lighting phosphors prepared by a systematic, bottom-up synthetic approach. J Am Chem Soc 137(29):9400–9408 112. Yu Y, Zhang X-M, Ma J-P et al (2014) Cu(I)-MOF: naked-eye colorimetric sensor for humidity and formaldehyde in single-crystal-to-single-crystal fashion. Chem Commun 50 (12):1444–1446 113. Shimizu GKH, Vaidhyanathan R, Taylor JM (2009) Phosphonate and sulfonate metal organic frameworks. Chem Soc Rev 38(5):1430–1449 114. Serre C, Férey G (2002) Hydrothermal synthesis, structure determination from powder data of a three-dimensional zirconium diphosphonate with an exceptionally high thermal stability: Zr (O3P-(CH2)-PO3) or MIL-57. J Mater Chem 12(8):2367–2369 115. Barthelet K, Merlier C, Serre C et al (2002) Microporous hybrid compounds: hydrothermal synthesis and characterization of two zinciomethylenediphosphonates with 3D structures, structure determination of their dehydrated forms. J Mater Chem 12(4):1132–1137 116. Kaskel S (2002) Porous metal-organic frameworks. In: Schüth F, Sing KSW, Weitkamp J (eds) Handbook porous solids, 1st edn. Wiley, New York, pp 1190–1249 117. Liu J, Chen L, Cui H et al (2014) Applications of metal–organic frameworks in heterogeneous supramolecular catalysis. Chem Soc Rev 43(16):6011–6061 118. Howarth AJ, Peters AW, Vermeulen NA et al (2017) Best practices for the synthesis, activation, and characterization of metalorganic frameworks. Chem Mater 29(1):26–39 119. Farha OK, Hupp JT (2010) Rational design, synthesis, purification, and activation of metalorganic framework materials. Acc Chem Res 43(8):1166–1175 120. Bennett TD, Cheetham AK, Fuchs AH et al (2016) Interplay between defects, disorder and flexibility in metal-organic frameworks. Nat Chem 9(1):11–16 121. Moosavi SM, Boyd PG, Sarkisov L et al (2018) Improving the mechanical stability of metalorganic frameworks using chemical caryatids. ACS Cent Sci 4(7):832–839 122. Tan JC, Bennett TD, Cheetham AK (2010) Chemical structure, network topology, and porosity effects on the mechanical properties of zeolitic imidazolate frameworks. Proc Natl Acad Sci U S A 107(22):9938–9943 123. Wu H, Yildirim T, Zhou W (2013) Exceptional mechanical stability of highly porous zirconium metal-organic framework UiO-66 and its important implications. J Phys Chem Lett 4(6):925–930 124. Zhai QG, Bu X, Mao C et al (2016) An ultra-tunable platform for molecular engineering of high-performance crystalline porous materials. Nat Commun 7(1):1–9 125. Zhao X, Bu X, Zhai QG et al (2015) Pore space partition by symmetry-matching regulated ligand insertion and dramatic tuning on carbon dioxide uptake. J Am Chem Soc 137 (4):1396–1399 126. Kapustin EA, Lee S, Alshammari AS et al (2017) Molecular retrofitting adapts a metal-organic framework to extreme pressure. ACS Cent Sci 3(6):662–667 127. Ding M, Cai X, Jiang HL (2019) Improving MOF stability: approaches and applications. Chem Sci 10(44):10209–10230 128. Schneemann A, Bon V, Schwedler I et al (2014) Flexible metal-organic frameworks. Chem Soc Rev 43(16):6062–6096 129. Slater CS, Savelski M (2007) A method to characterize the greenness of solvents used in pharmaceutical manufacture. J Environ Sci Heal – Part A Toxic/Hazardous Subst Environ Eng 42(11):1595–1605 130. Czaja A, Leung E, Trukhan N et al (2011) Industrial MOF synthesis. In: Farrusseng D (ed) Metal-organic frameworks: applications from catalysis to gas storage, 1st edn. WileyVCH Verlag, New York, pp 337–352
1 Robust and Environmentally Friendly MOFs
31
131. Gaab M, Trukhan N, Maurer S et al (2012) The progression of Al-based metal-organic frameworks – From academic research to industrial production and applications. Microporous Mesoporous Mater 157:131–136 132. Wang S, Serre C (2019) Toward green production of water-stable metal-organic frameworks based on high-valence metals with low toxicities. ACS Sustain Chem Eng 7(14):11911–11927 133. Motagamwala AH, Won W, Sener C et al (2018) Toward biomass-derived renewable plastics: production of 2,5-furandicarboxylic acid from fructose. Sci Adv 4(1):1–8 134. Byrne FP, Jin S, Paggiola G et al (2016) Tools and techniques for solvent selection: green solvent selection guides. Sustain Chem Process 4(1):1–24 135. Peh SB, Wang Y, Zhao D (2019) Scalable and sustainable synthesis of advanced porous materials. ACS Sustain Chem Eng 7(4):3647–3670 136. Lanchas M, Arcediano S, Aguayo AT et al (2014) Two appealing alternatives for MOFs synthesis: solvent-free oven heating vs. microwave heating. RSC Adv 4(104):60409–60412 137. Mueller U, Schubert M, Teich F et al (2006) Metal-organic frameworks – Prospective industrial applications. J Mater Chem 16(7):626–636 138. Rubio-Martinez M, Avci-Camur C, Thornton AW et al (2017) New synthetic routes towards MOF production at scale. Chem Soc Rev 46(11):3453–3480 139. Crawford D, Casaban J, Haydon R et al (2015) Synthesis by extrusion: continuous, large-scale preparation of MOFs using little or no solvent. Chem Sci 6(3):1645–1649 140. Avci-Camur C, Troyano J, Pérez-Carvajal J et al (2018) Aqueous production of spherical Zr-MOF beads: via continuous-flow spray-drying. Green Chem 20(4):873–878 141. Garcia Marquez A, Horcajada P, Grosso D et al (2013) Green scalable aerosol synthesis of porous metal-organic frameworks. Chem Commun 49(37):3848–3850 142. Garay AL, Pichon A, James SL (2007) Solvent-free synthesis of metal complexes. Chem Soc Rev 36(6):846–855
Chapter 2
Large-Scale Synthesis and Shaping of Metal-Organic Frameworks U-Hwang Lee, Sachin K. Chitale, Young Kyu Hwang, and Jong-San Chang
2.1
Introduction
Currently, MOF research and development trends are exhibiting an increasing demand for scale-up and shaping technologies owing to the increasing expectations with regard to potential industrial applications. To satisfy such industrial demands, production efficiency, performance reproducibility, and price competitiveness are simultaneously required, and various mass production methods and shaping technologies have been reported in this regard. Mass production technologies of MOFs can be classified into traditional batch production technologies and continuous production technologies as a platform for the development and modification of MOF products. Recent studies have demonstrated that the “scaled-up” or “large-scale” synthesis of MOFs can ensure a high production efficiency with a compact installation. Although most studies report high space-time yields (STYs), further research is required to meet industrial demands. Therefore, the ultimate goal of current MOF mass synthesis studies is to develop a low-cost and stable production technology for MOFs that can replace the porous materials currently used in the industry. The
U.-H. Lee · S. K. Chitale · Y. K. Hwang Research Center for Nanocatalysts, Korea Research Institute of Chemical Technology (KRICT), Daejeon, Korea Department of Green Chemistry & Biotechnology, University of Science and Technology (UST), Daejeon, Korea J.-S. Chang (*) Research Center for Nanocatalysts, Korea Research Institute of Chemical Technology (KRICT), Daejeon, Korea Department of Chemistry, Sungkyunkwan University, Suwon, Korea e-mail: [email protected] © Springer Nature Switzerland AG 2021 P. Horcajada Cortés, S. Rojas Macías (eds.), Metal-Organic Frameworks in Biomedical and Environmental Field, https://doi.org/10.1007/978-3-030-63380-6_2
33
34
U.-H. Lee et al.
Fig. 2.1 Schematic of batch-type and continuous-flow production of MOFs
quality of the fabricated product, STY, efficient use of resources and energy, waste reduction, and safety are important parameters to be considered [19–21, 27, 69, 77, 80, 87, 112, 118] (Fig. 2.1). Shaping of MOFs is essential for their practical applications such as adsorption/ separation and catalysis, which require the structuring of nano-/microcrystalline fine powders of MOFs into a macroscopic body [2, 24, 63, 97, 103, 104, 113]. When a catalysis or adsorptive separation process is scaled up to the pilot or commercial scale, the form in which an active catalyst or adsorbent is used is decisive in determining the ultimate performance of the unit [58, 82, 83]. Therefore, the size and shape of the target material require a compromise between minimizing the pore diffusion effects in the material particles (which requires particle sizes to be small) and minimizing the pressure drop across the reactor (which requires particle sizes to be large) [65]. The shaped MOFs should retain the excellent performance exhibited by the original powder forms based on their high porosity, crystallinity, and flexibility [68]. Therefore, the proper selection of the shaping or forming process is important in MOF research with regard to chemical engineering and industrial applications. As shaping methods are often proprietary and more of an art than a science, the exact procedures are rarely disclosed. Shaped MOFs can also be prepared using techniques that have typically been utilized for synthesizing catalysts and adsorbents from porous materials, such as pressing, extrusion, granulation, spray-drying, foaming, casting, and coating. Well-shaped bodies have been produced in the form of pellets, tablets, monoliths, granules, and spheres [9, 16, 31, 43]. However, for the effective utilization of shaped MOFs for industrial applications, they should possess a sufficient mechanical strength, chemical stability, attrition resistance, and maximal bulk density as well as exhibit a minimal wasted space in storage vessels [34–36, 45, 62, 79, 90, 109]. As the performance of shaped MOFs with regard to industrial or large-scale applications depends on several parameters, including mass and heat transfer properties, gas diffusion kinetics, pressure drop across the material, mechanical strength, and volumetric efficiency, achieving a high mechanical integrity of structured MOFs is critical to their performance in real engineering processes (Fig. 2.2).
2 Large-Scale Synthesis and Shaping of Metal-Organic Frameworks
35
Fig. 2.2 Examples of shaped MOF products prepared using granulation and pressing methods
2.2 2.2.1
Scale-Up Synthesis of MOFs Batch-Type Production
The mass production of MOFs can be classified into two parts: batch production and continuous production. Batch production is a method of producing traditional organometallic compounds and metal oxide chemicals, such as hydrothermal synthesis, solvent thermal synthesis, and reflux. Therefore, large-scale reaction facilities are essential for mass production (Fig. 2.1). Only several research groups around the world have successfully mass-produced MOFs. BASF in Germany has produced more than five MOFs on the pilot scale and one MOF (Basolite A520) on the multitone scale, as shown in Table 2.1. Some products are sold through SigmaAldrich Inc. and Strem Chemicals Inc. Additionally, Framergy and NuMat in the USA are known to produce MOFs on a kilogram scale, but the exact amount produced by Framergy and NuMat is undisclosed (www.basf.com, www. moftechnology.com, www.mofapps.com, www.mofgen.com, www. mosaicmaterial.com, www.numat-tech.com, www.novomof.com, www. prometheanparticles.co.uk, and www.stream.com). In addition to these examples, KRICT also reported the production of tens of kilograms of MIL-100(Fe) through a 200 L scaled hydrothermal reaction using suitable conditions without HF. MIL-100(Fe) comprises hierarchically mesoporous iron (III) trimesate with a zeolite-MTN topology. Although such conditions are narrow, the concurrent change of the iron precursor and an increase in the concentration of the reaction mixture causes a synergetic effect leading to an increase in the crystallinity of F-free MIL-100(Fe) (called KRICT F100; Fig. 2.3). This method,
36
U.-H. Lee et al.
Table 2.1 List of MOFs produced by BASF MOF Basolite A100 Basolite C300 Basolite F300 Basolite Z1200 Basolite M050 Basolite A520
Langmuir surface area (m2/g) 1100–1500
Space-time yield (kg/m3/day) 160
Cu-benzene-1,3,5-tricarboxylate (HKUST-1) Fe-benzene-1,3,5-tricarboxylate
1500–2100
225
1300–1600
20
Zn-2-methyl-imidazolate (ZIF-8)
1300–1800
100
Mg-formate
400–600
> 300
Aluminum fumarate
1300
–
Types of MOFs Al-terephthalate (MIL-53(Al))
The all above MOF products belong to BASF Company (www.basf.com)
Fig. 2.3 Photographs of 200 L glass-lined reactor and a plastic container containing 15 kg F-free MIL-100(Fe) powder produced by KRICT. (Reprinted with permission from Ref. [93]. Copyright 2012 Elsevier Inc)
combined with two purification steps (solvent extraction and chemical treatment with NH4F), leads to a highly porous F-free material with a very high STY (> 1700 kg/m3/day). The resulting material exhibits similar physicochemical properties to those of the MOFs prepared in the presence of HF, except for a slight difference in sorption capacities of gases and liquid vapors corresponding to the difference in pore volume. In the recent years, KRICT has carried out a pilot production of several hundred kilograms of MOFs [93].
2 Large-Scale Synthesis and Shaping of Metal-Organic Frameworks
2.2.2
37
Continuous-Flow Production of MOFs
The continuous-flow production method can utilize electrochemical, microwave, and mechanochemical methods as energy sources for the reaction, in addition to the conventional hydrothermal synthesis or solvothermal synthesis methods (Fig. 2.4). In particular, the MOF crystal growth reaction depends on the residence time, and MOFs can be synthesized in a very short period using this method. Hence, the empty time yield is 2–3 times higher compared to that of batch-type production. Moreover, unlike the batch production method, this method reduces the production cost as high productivity can be achieved by employing simple, small production facilities. Table 2.2 shows the various continuous production methods for producing MOF reported to date. Commercially available industrial-scale batch synthesis of HKUST-1 and CPO-27 was achieved by Richard et al. They used a continuous-flow reactor with a rapid mixing stream of a preheated solvent and reagent solution. This methods delivered a high STY despite the short residence time, and the product obtained had a high purity and crystallinity which were elucidated using a continuous-flow hydrothermal, solvothermal, and microwave-assisted system to achieve the desired crystal size and phase control. A high rate of synthesis of a Cu-BTC MOF with a BET surface area of more than 1600 m2/g (Langmuir surface area of more than 2000 m2/g) and a production yield of 97% could be achieved with a total reaction time of 5 min [4, 48]. Additionally, several studies have reported that a high crystallinity and specific surface area have been achieved in MIL-53(Al), UiO-66 (Zr), UiO-66(Zr)-NH2 [52], CAU-13, NOTT-100, ZIF-8, etc., synthesized with high STY using the continuous synthesis method [3, 8, 23, 30, 40, 56, 64, 73, 88, 89, 101, 107].
Fig. 2.4 Schematic diagram of the continuous-flow production of MOFs
38
U.-H. Lee et al.
Table 2.2 Continuous MOF production methods Method Continuous-flow hydrothermal
Continuous-flow solvothermal
Continuous-flow microwave
MOF HKUST-1 CPO-27 ZIF-8 MIL-53(Al) Cu-BTC Al-Fumarate UiO-66 NOTT-400 HKUST-1 HKUST-1 MOF-5 IRMOF-3 UiO-66 UiO-66-NH2 UiO-66 UiO-66-NH2 CAU-13 MOF-74(Ni) UiO-66 MIL-53(Al) HKUST-1
Residence (time) 1 min 1 min 1 min 5–6 min 5 min 1 min 10 min 15 min 1.2 min 15 min
44 s 1.1 min 20 min 1s 7s 4s 1s
SBET (m2/g) 1950 1030 1806 919 1600 1054 1186 1078 1805
STY (kg/m3/day) 4399 1501 3875 1300 97,159 672 741 4533
References [23, 30, 69]
[8] [40] [89] [88]
[64] 3185 2428 1059 827 1204 922 401 840 1352 1376 1550
[73] [101] 3049 2160 7204 3618 64,800
[107] [3] [99]
The CSIRO group studied the scalability of the kilo-scale production of MOFs at 5.6 kg/h through the continuous-flow production of aluminum fumarate at a pilot scale. Their water-based production method was validated and re-optimized for the scale-up from the laboratory to the pilot-plant scale. By controlling the flow rate from 10 mL to 1.3 L, they achieved a high production rate and STY of over 5.5 kg/h and 9000 kg/m3/day, respectively (Fig. 2.5; [89]). Microwave- and electrochemistry-assisted continuous-flow production can achieve the highest STY as well as the fastest conversion of reagents. Rapid heating provides excellent control over material properties owing to the occurrence of uniform nucleation, separation of nucleation, growth steps, and the phase formation of MOFs in a short reaction time. Despite these advantages, only laboratory-scale studies have been reported in this regard. The extrusion of ZIF-8 at a rate of 4 kg/h has been achieved via continuous neat melt phase synthesis. The STY of this method (up to 144 103 kg/m3/day) is an order of magnitude greater than the STYs of other methods used to produce MOFs. This extrusion method clearly enables the scalability of mechanochemical and melt phase synthesis under solvent-free or low-solvent conditions, which may generally
2 Large-Scale Synthesis and Shaping of Metal-Organic Frameworks
39
occur during synthesis. This technique could be an effective method for covalent chemical synthesis under solvent-free, minimal solvent, or melt phase conditions. The authors suggest that, subsequent to a degree of optimization, synthesis by extrusion can be performed at a large scale in a continuous process, and an outcome similar to that of synthesis by ball milling at a small scale in a batch process can be achieved. Additionally, phases other than those obtained by ball milling can in some cases be obtained. It is notable that high-throughput rates of several kilograms per hour were readily achieved, and much higher rates of several hundred kilograms per hour should be possible via the use of a large-scale extrusion equipment [23]. As discussed above, the shaping of MOFs via extrusion has several advantages, including a possibly high throughput, relatively low cost, various possible extrudate shapes, and the possibility of continuous mass production with a high STY. However, extrudates are normally less uniform and less resistant to abrasion compared to pressed materials. Several parameters govern the extrusion process. One of the most important parameters is the flow behavior of the MOF mixture paste. The mixture paste is prepared by mixing the MOF, inorganic or organic additives, and solvent. If the mixture paste is too viscous, it can block the extruder. The mixture paste must exhibit suitable properties, such as plasticity, to allow the paste to be extruded through the die; furthermore, it should exhibit sufficient cohesion to avoid the formation of surface and bulk defects in the extruded product. To produce better MOF extrudates, the effects of the porosity of the MOF with additives have to be studied (Fig. 2.5) [38].
Fig. 2.5 A photograph of twin screw extruder for mechanochemical synthesis of MOFs. The two screws that convey and knead the reactants are housed in the barrel. (Reprinted with permission from Ref. [38]. Copyright 2018 American Chemical Society)
40
2.3
U.-H. Lee et al.
Shaping of MOF
Most porous materials require binders to achieve a sufficient strength after shaping to avoid sand-like behavior. Therefore, it is important to add binders in the paste as they provide a sufficient strength to the shaped bodies after drying and thermal treatments. Binders maintain the integrity of the paste and the green body. For shaping inorganic materials, the final strength must be derived from inorganic binders because most of the shaped inorganic materials are calcined at relatively high temperatures after their production. However, organic binders or plasticizers can be considered for shaping MOFs because these materials consist of organic linkers so that high-temperature treatments can be avoided. Alumina, silica sols, or clay are generally used as inorganic binders [33]. Plasticizers are additives that improve the rheological behavior of the paste for the shaping process. These materials generally include polymers, such as polyethylene oxide, polyethylene glycol, polyvinyl alcohol, polyacrylamide, polyvinylpyrrolidine, alginates, or different types of cellulose [26, 39, 96]. Organic plasticizers can also be used as organic binders for MOFs [59, 74]. The main processing steps for porous materials and MOFs are primarily the same and can be described as follows: mixing of the porous powder with inorganic and organic additives, shaping the powders into the desired shape, removing temporal additives, and creating a mechanically robust structure via thermal treatment. The efforts made in recent studies in terms of processing routes for porous materials have been reviewed [1, 71, 81, 91, 94, 102, 116, 119]. An important point in shaping MOFs is using the most suitable method for shaping. The current status of scalable shaping methods, including extrusion, pressing, granulation, spray-drying, foaming, and alginate, will be discussed in detail in this chapter.
2.3.1
Conventional Methods of Powder Shaping (Fig. 2.6)
2.3.1.1
Granulation
Granulation is an example of designing shaped bodies via the agglomeration of a powder into large aggregated granules, similar to the formation of a snowball, as shown in Figs. 2.2 and 2.7. It is the simplest and cheapest shaping method, as a shaped material can be produced continuously in a granulation apparatus with uniform quality. The granulation of powder can serve as a simple and inexpensive method for producing shaped particles, although the shape and size of the particles is irregular compared to those achieved using other shaping methods. The granulation process is governed by the interactions between the powder and the liquid binder
2 Large-Scale Synthesis and Shaping of Metal-Organic Frameworks
41
Fig. 2.6 Schematic of conventional powder shaping methods
Fig. 2.7 Photographs and N2 isotherms, at 77 K, of granulated MOFs: (a) MIL-100(Fe), (b) MIL-101(Cr), (c) UiO-66(Zr), and (d) UiO-66(Zr)-NH2. (Reprinted with permission from Ref. [105]. Copyright 2017 Royal Society of Chemistry)
based on the following key elementary steps [37]: (a) wetting and nucleation, (b) consolidation, and (c) attrition and breakage. In wet granulation, liquid bridges between particles and capillarity provide the major binding force [11, 67, 76]. The mechanical strength of the granules formed is determined by the liquid content, surface tension, and viscosity of the liquid binder. The particles grow until the destructive forces exceed the strength and an equilibrium particle size is attained. The properties of the powder and liquid binder are relevant to their interactions. The properties of the fine MOF powder may change the mechanical function and physical property of the composition. Therefore, the desired body shapes are controlled by a combination of the feed MOF powder, additive (binder and solvent) properties, and the selection of the types of granulators and processes. The
42
U.-H. Lee et al.
granulation process is generally performed by wetting the solvent and binder on the MOF powders in a granulator apparatus, such as a tumbling drum, fluidized bed, shear mixer, or rotating fan (Fig. 2.6) [117]. In the granulation method, the selection of the binder is a critical factor. Both inorganic and organic binders can be used, depending on the purpose. MIL-100(Fe) and MIL-160(Al) were granulated using silica sol as a binder, which is one of the most well-known inorganic binders [41]. To maintain the inherent characteristics, such as the high porosity and high surface area of MOFs, research on using a porous inorganic binder has been reported. As shown in Fig. 2.7, Valekar et al. prepared millimeter-scale spheres of MIL-100(Fe), MIL-101(Cr), UiO-66(Zr), and UiO-66(Zr)-NH2 using mesoporous r-alumina (MRA) as a binder that resulted in optimally shaped MOF bodies that retained their intrinsic properties after shaping. The loss of weight-specific surface area for the MIL-100(Fe), MIL-101(Cr), UiO-66(Zr), and UiO-66(Zr)-NH2 MOFs was 7.7%, 4.6%, 8.7%, and 1%, respectively [105]. There is a lot of development on research granulated using a polymer binder. For example, Z-MOFs using a sucrose binder; UiO-66(Zr), UiO-66(Zr)-NH2, MIL-100 (Fe), MIL-127(Fe), and UiO-66-COOH using the organic binder polyvinyl alcohol (PVA); and anion-pillared ultramicroporous MOFs using a polyvinyl butyral binder have been produced [106].
2.3.1.2
Extrusion
The term “extrusion” refers to a family of continuous processing techniques in which materials are forced through constrained spaces or dies. The products obtained via extrusion are generally called “extrudates.” Drawing material is the main method of producing wires, sheets, bars, and tubes. Extrusion is one of the most widely used manufacturing methods for shaping porous powders for adsorption and catalytic applications. It has been established as a commercial method for producing mechanically strong and attrition-resistant granules, pellets, and honeycomb structures of industrially important adsorbents and catalysts, such as zeolites, MOFs, and porous carbon for adsorption, air separation, and catalytic applications (Fig. 2.2) [10, 32, 56, 85, 86, 114, 115]. The main advantage of this process over other manufacturing processes is its ability to produce various cross-sectional shapes, of which a cylinder is the simplest, as the material encounters only compressive and shear stresses. However, rings, trilobates, stars, starrings, or monolithic honeycombs can also be produced using suitable dies. After leaving the die, the extrudate is either left to break, resulting in homogeneously sized extrudates, or it is cut off by a rotating blade, which provides extrudates that are all almost identical (Fig. 2.8). Extruded MOF monoliths with reasonable mechanical stability were produced by Kaskel et al. [55]. They reported an essential step toward the use of MOF monoliths
2 Large-Scale Synthesis and Shaping of Metal-Organic Frameworks
43
Fig. 2.8 Preparation of extrudates of MIL-100(Cr). (Reprinted with permission from Ref. [56]. Copyright 2016 Elsevier Ltd)
Fig. 2.9 Schematic of the preparation procedure of MOF disk tablets and ZIF-8 (left) and MIL-53 (Al) (right) adsorbent particles (1–2 mm) prepared via binder-free compression, crushing, and sieving. (Reprinted with permission from Ref. [84]. Copyright 2018 Elsevier Inc)
in industrial separation processes, gas storage, and catalysis. The monolith was prepared using a ram extruder that resulted in a stable monolithic structure with a specific surface area of up to 484 m2/g. The fabrication of Cu-BTC monoliths employed silicone resin (SILRES® MSE100) as the binding agent and cellulose as the plasticizer. The mixture paste was extruded into a monolithic shape using a ram extruder and microwave drying. Ramírez et al. demonstrated the complete shaping of Cu-BTC into a monolith with millimeter-sized extrudates. Their synthetic approach exhibited many decisive advantages for the valuable advancement of the lab-scale and industrial production of Cu-BTC. The scale-up performed in their study resulted in a product that was obtained using a Caleva Mini Screw Extruder with a 2 mm die after mixture optimization with 20 wt.% kaolinite as an additive. The kaolinite additive did not affect the crystallinity but resulted in a surface area 900 m2 g 1 [25, 55, 60, 66]. Cellulose additives are also used for extrusion, e.g., MIL-101(Cr) was pelletized using carboxymethyl cellulose; furthermore, flexible MOFs MIL-53(Al) and MIL-53(Al)-NH2 were obtained using a methylcellulose binder (Fig. 2.9) [42].
44
2.3.1.3
U.-H. Lee et al.
Pressing
The pressing method (including the compressing, tableting, and pelleting methods) results in the most uniformly shaped bodies of pellets or tablet types with a high mechanical strength. This method is widely used for industrial applications because it can be used for mass production. The aim of this shaping method is to transform fluffy powders into a desired compact shape with a maximal overall density. This method involves filling a die with powder and pressing the powder into the desired shape. The powder fills a metallic mold, and pressure is applied from the top and bottom (uniaxial press) to achieve high-density compaction, minimal variation in density, packing homogeneity, sufficient strength, and integrity. With regard to MOF pressing, certain criteria should be followed to maintain their unique properties. The basics of pressing are covered in substantial detail in various papers and patents that focus on MOFs. The quality of pellets depends, to a substantial extent, on the mechanical properties of the MOF powders. The deformation properties of a MOF considerably affect the pressing process as the response of the material to the applied stress is the main factor controlling the properties of the resulting pellets [5, 6, 14, 15, 44, 95]. The ability of MOF to produce binder-free robust pellets has already been reported. In particular, MOF-5 when pressed at approximately 0.18 GPa produces hard disks. However, the crystalline structure of the MOF changes radically subsequent to pressing and becomes more similar to that of an amorphous structured material. This evidence is consistent with the breakdown of the structure of the framework. Tagliabue et al. produced hard disk pellets of CPO-27-Ni without any binder (binder-free). The tablet obtained by compacting CPO-27-Ni at 0.1 GPa retained the color of the original powder (yellow), whereas a dramatic color change (from yellow to dark brown) occurred after the sample was pressed at 1 GPa. This finding suggests deep modifications in the crystal structure of the MOF. A remarkable decrease in crystallinity was observed after the sample was pressed at a pressure of 1 GPa. This substantial crystallinity loss is most likely associated with the breakdown of the MOF structure. The Cu-BTC (HKUST-1) pellet produced by BASF has been evaluated with regard to the separation of propane and propylene mixtures using the adsorption process. The pure-component adsorption isotherms revealed that the adsorption capacity of Cu-BTC crystals was reduced by 21% owing to the shaping process [98–103]. Recently, MIL-53(Al) was shaped by the mechanical compression of commercial ZIF-8 and MIL-53(Al) powders at two different pressures of 62 MPa and 125 MPa, and the consequent impact on the mechanical, structural, and textural characteristics of the MOFs was evaluated (Fig. 2.10) [4, 70, 75, 78, 84, 100, 111]. Our previous paper discussed pressed pellets of microporous UiO-66(Zr) and mesoporous MIL-100(Fe) with additives [105] (Fig. 2.9). The shaped pellets were prepared using a two-step process: first, the material was pre-compacted, grinded, and sieved, and the resulting granulated material was shaped into pellets using a
2 Large-Scale Synthesis and Shaping of Metal-Organic Frameworks
45
Fig. 2.10 Schematic (a) and photograph (b) of a spray-drying apparatus for continuous-flow synthesis of high-nuclearity MOFs. (Reprinted with permission from reference [29]. Copyright 2016 Royal Society of Chemistry)
rotary press tabletizer. The MOF powders were uniformly mixed with a small amount of graphite powder (1%) using ball milling. Then, the solid mixture was fed into a rotary press tabletizer to form a shaped body consisting of 99 wt.% MOFs and 1 wt.% graphite. Cylindrical pellets were formed by the combined pressing action of two punches and a die with holes that were 3 mm in diameter.
2.3.2
Solidifying Methods
2.3.2.1
Spray-Drying
The spray-drying method offers many advantages because the particle size of the powder, and morphology and density of the particles are controlled in one step. Spray-drying is a process by which particles are produced from a liquid or slurry via atomization and drying using a hot gas. In this process, shaping occurs during the
46
U.-H. Lee et al.
atomization of the feed slurry into granules (called droplets), which are circulated in a chamber. The separation of the shaped bodies is performed based on their size distribution. This method enables the production of small particles on the micrometer scale to large particles on the millimeter scale, depending on the initial feed and operating conditions. However, medium-sized droplets (agglomerated particles with a size of 10–300 μm) are commonly obtained. Owing to the presence of large particles or powders, millimeter-sized droplets might not be free-flowing. It has been reported that spray-drying can be exploited as a general, low-cost, rapid, and scalable method for the synthesis and self-assembly of MOFs [28]. Serre et al. proposed the one-pot synthesis and processing of several MOFs using spraydrying via the self-assembly method by using templating agents. They used three concentric fluid nozzles on the system. Two separate solutions of organic ligand and metallic salts were simultaneously injected into the same flow of hot air. Using this simple setup, it was possible to produce shaped bodies of Fe-BTC, HKUST-1, and ZIF-8, with submicron and nano-sized particles. They demonstrated the possibility of industrial batch production through this method, which is economical and environmentally friendly. Additionally, this method can be used for continuous production [7, 12, 13, 22, 28, 72, 98, 108]. Maspoch et al. described a highly versatile and effective methodology for synthesizing either nano-MOFs or hollow MOF superstructures that enabled the largescale production of sub-5 μm hollow spherical superstructures via the localized crystallization of nano-MOFs [29]. The atomized droplets of MOFs were manufactured from a liquid precursor solution of metal salts and organic ligands. The MOF droplets produced during spray-drying could also be used as individual reactors to confine the large-scale synthesis and assembly of nano-MOFs. The resulting MOF superstructures were well-dispersed and discrete nano-MOFs, owing to disassembly using sonication. In 2016, D. Maspoch et al. have produced commonly known MOFs and their hollow superstructures, e.g., HKUST-1; Cu-BDC; NOTT-100; MIL-88A; MIL-88B; MOF-14; MOF-74 [M¼Zn(II), Ni(II), and Mg(II)]; UiO-66; ZIF-8; a Prussian blue analogue of Cu(II), that is, (Cu-PB); MOF-5; and IRMOF-3 [13]. Additionally, an updated version was reported based on incorporating a continuous-flow reaction that combined the advantages of the continuous-flow and spray-drying methods to provide MOFs with good yields, an excellent porosity, and highly dense cores. Furthermore, it is amenable to the fabrication of MOFs, thereby providing new avenues for fine-tuning the porosity of these materials (Fig. 2.10) [29].
2.3.2.2
Foaming
The foaming method was developed to shape MOFs into a variety of robust and stable bodies, such as beads or foams, from transformable fluids, gels, and hydrocolloids. Eddaoudi et al. reported a facile and quick shaping method by foaming
2 Large-Scale Synthesis and Shaping of Metal-Organic Frameworks
47
Fig. 2.11 Light and robust foam with hierarchical porosity of HKUST-1@Fe3O4 core-shell nanoparticles. (Reprinted with permission from Ref. [17]. Copyright 2016 Royal Society of Chemistry)
[61]. They prepared a mixture of UiO-66, ZIF-8, HKUST-1, and two distinct polymers, namely, polyethelyneglycol (PEG, a rubbery polymer) and poly(methyl methacrylate) (PMMA, a glassy polymer). To corroborate the universality of the proposed shaping strategy, MOF beads were successfully prepared with a high mechanical stability [18, 92]. Moreover, a strategy for the continuous and reversible processing of MOF-based materials was successfully applied to various MOF and composite structures, such as HKUST-1@Fe3O4 core-shell nanoparticles. A magnetic fluid with a high particle content (25.0–45.4 wt.%) was obtained. Based on this magnetic fluid, a cup reactor was fabricated along with a light and robust foam with hierarchical porosity using a solvent-induced hardening process (Fig. 2.11) [17].
2.3.2.3
Alginate
A modified granulation method denoted as molecular gastronomy was recently proposed to form MOFs into spheres using hydrocolloids. Hydrocolloids were used to prepare uniformly shaped spheres of CPO-27(Ni). The mixed slurry of CPO-27(Ni) dispersed in aqueous alginate or chitosan solution was dropped on gelling agent solutions. Droplets that were approximately 2.5–3.5 mm in size were instantaneously formed; these could be dried into spheres containing >84 wt.% of CPO-27(Ni) [28]. Grande et al. studied the effect of the most important process
48
U.-H. Lee et al.
Fig. 2.12 MOF beads in the gelation solution (left) and after activation (right). Different shapes of beads obtained from different concentrations of calcium cations in the gelation bath. (Reprinted with permission from Ref. [57]. Copyright 2019 Elsevier B.V)
variables to produce UiO-66 beads using the alginate method. Furthermore, the possibility of using other gel-inducing cations was evaluated to characterize the produced beads with respect to mechanical strength and surface area (Fig. 2.12) [57].
2.4
Summary
In this chapter, current status of large-scale synthesis of MOFs and several MOF shaping methods that primarily focus on granulation were discussed. The granulation method is appropriate for shaping MOFs because MOFs are generally mechanically weak and the granulation method performs shaping in a mechanical shock-free manner. However, several parameters, such as the type of binder, liquid content, surface tension, and viscosity of the liquid binder/powder mixture, are crucial for increasing the mechanical strength of shaped MOFs obtained via wet granulation. Air calcination of MOF granulates is not suitable for increasing the mechanical strength because MOFs are generally unstable in air at high temperatures. Hence, additional studies on the selection of binders and interfacial conditions between the surfaces of MOF powders and binders are necessary.
2 Large-Scale Synthesis and Shaping of Metal-Organic Frameworks
49
References 1. Akhtar F, Andersson L, Ogunwumi S et al (2014) Structuring adsorbents and catalysts by processing of porous powders. J Eur Ceram Soc 34:1643–1666. https://doi.org/10.1016/j. jeurceramsoc.2014.01.008 2. Alaerts L, Kirschhock CEA, Maes M et al (2007) Selective adsorption and separation of xylene isomers and ethylbenzene with the microporous vanadium(IV) terephthalate MIL-47. Angew Chem Int Ed 46:4293–4297. https://doi.org/10.1002/anie.200700056 3. Albuquerque GH, Fitzmorris RC, Ahmadi M et al (2015) Gas-liquid segmented flow microwave-assisted synthesis of MOF-74(Ni) under moderate pressures. CrystEngComm 17:5502–5510. https://doi.org/10.1039/c5ce00848d 4. Alcañiz-Monge J, Trautwein G, Pérez-Cadenas M, Román-Martínez MC (2009) Effects of compression on the textural properties of porous solids. Microporous Mesoporous Mater 126:291–301. https://doi.org/10.1016/j.micromeso.2009.06.020 5. Ribeiro AM, Campo M, Narin G, Santos J, Alexandre Ferreira J-S, Hwang YK, You-Kyong Seo B, U-Hwang Lee, Loureiro AER JM (2013) Pressure swing adsorption process for the separation of nitrogen and propylene with a MOF adsorbent MIL-100(Fe). Sep Purif Technol Sustain Futur 110:101–111 6. Ardelean O, Blanita G, Borodi G et al (2013) Volumetric hydrogen adsorption capacity of densified MIL-101 monoliths. Int J Hydrog Energy 38:7046–7055. https://doi.org/10.1016/j. ijhydene.2013.03.161 7. Avci-Camur C, Troyano J, Pérez-Carvajal J et al (2018) Aqueous production of spherical Zr-MOF beads: via continuous-flow spray-drying. Green Chem 20:873–878. https://doi.org/ 10.1039/c7gc03132g 8. Bayliss PA, Ibarra IA, Pérez E et al (2014) Synthesis of metal-organic frameworks by continuous flow. Green Chem 16:3796–3802. https://doi.org/10.1039/c4gc00313f 9. Boettger J, Deussen O, Ziezold H (2011) (12) Patent Application Publication (10) Pub. No.: US 2011/0141115A1. 1 10. Brandani F, Ruthven DM (2004) The effect of water on the adsorption of CO2 and C 3H8 on type X zeolites. Ind Eng Chem Res 43:8339–8344. https://doi.org/10.1021/ie040183o 11. Cameron IT, Wang FY, Immanuel CD, Stepanek F (2005) Process systems modelling and applications in granulation: a review. Chem Eng Sci 60:3723–3750. https://doi.org/10.1016/j. ces.2005.02.004 12. Carné-Sánchez A, Imaz I, Cano-Sarabia M, Maspoch D (2013) A spray-drying strategy for synthesis of nanoscale metal-organic frameworks and their assembly into hollow superstructures. Nat Chem 5:203–211. https://doi.org/10.1038/nchem.1569 13. Carné-Sánchez A, Stylianou KC, Carbonell C et al (2015) Protecting metal-organic framework crystals from hydrolytic degradation by spray-dry encapsulating them into polystyrene microspheres. Adv Mater 27:869–873. https://doi.org/10.1002/adma.201403827 14. Chapman KW, Halder GJ, Chupas PJ (2009) Pressure-induced amorphization and porosity modification in a metal-organic framework. J Am Chem Soc 131:17546–17547. https://doi. org/10.1021/ja908415z 15. Chapman KW, Halder GJ, Chupas PJ (2008) Guest-dependent high pressure phenomena in a nanoporous metal-organic framework material. J Am Chem Soc 130:10524–10526. https:// doi.org/10.1021/ja804079z 16. Charkhi A, Kazemeini M, Ahmadi SJ, Kazemian H (2012) Fabrication of granulated NaY zeolite nanoparticles using a new method and study the adsorption properties. Powder Technol 231:1–6. https://doi.org/10.1016/j.powtec.2012.06.041 17. Chen Y, Huang X, Zhang S et al (2016) Shaping of metal-organic frameworks: from fluid to shaped bodies and robust foams. J Am Chem Soc 138:10810–10813. https://doi.org/10.1021/ jacs.6b06959
50
U.-H. Lee et al.
18. Cui X, Sun X, Liu L et al (2019) In-situ fabrication of cellulose foam HKUST-1 and surface modification with polysaccharides for enhanced selective adsorption of toluene and acidic dipeptides. Chem Eng J 369:898–907. https://doi.org/10.1016/j.cej.2019.03.129 19. Czaja AU, Trukhan N, Müller U (2009) Industrial applications of metal-organic frameworks. Chem Soc Rev 38:1284–1293. https://doi.org/10.1039/b804680h 20. Das AK, Vemuri RS, Kutnyakov I et al (2016) An efficient synthesis strategy for metalorganic frameworks: dry-gel synthesis of MOF-74 framework with high yield and improved performance. Sci Rep 6:1–7. https://doi.org/10.1038/srep28050 21. Dunne PW, Lester E, Walton RI (2016) Towards scalable and controlled synthesis of metalorganic framework materials using continuous flow reactors. React Chem Eng 1:352–360. https://doi.org/10.1039/c6re00107f 22. Elversson J, Millqvist-Fureby A (2005) Particle size and density in spray drying – effects of carbohydrate properties. J Pharm Sci 94:2049–2060. https://doi.org/10.1002/jps.20418 23. Faustini M, Kim J, Jeong GY et al (2013) Microfluidic approach toward continuous and ultrafast synthesis of metal-organic framework crystals and hetero structures in confined microdroplets. J Am Chem Soc 135:14619–14626. https://doi.org/10.1021/ja4039642 24. Férey G (2008) Hybrid porous solids: past, present, future. Chem Soc Rev 37:191–214. https:// doi.org/10.1039/b618320b 25. Ferreira AFP, Santos JC, Plaza MG et al (2011) Suitability of Cu-BTC extrudates for propanepropylene separation by adsorption processes. Chem Eng J 167:1–12. https://doi.org/10.1016/ j.cej.2010.07.041 26. Finsy V, Ma L, Alaerts L et al (2009) Separation of CO2/CH4 mixtures with the MIL-53(Al) metal-organic framework. Microporous Mesoporous Mater 120:221–227. https://doi.org/10. 1016/j.micromeso.2008.11.007 27. Gaab M, Trukhan N, Maurer S et al (2012) The progression of Al-based metal-organic frameworks – from academic research to industrial production and applications. Microporous Mesoporous Mater 157:131–136. https://doi.org/10.1016/j.micromeso.2011.08.016 28. Garcia Marquez A, Horcajada P, Grosso D et al (2013) Green scalable aerosol synthesis of porous metal–organic frameworks. Chem Commun 49:3848–3850. https://doi.org/10.1039/ c3cc39191d 29. Garzón-Tovar L, Cano-Sarabia M, Carné-Sánchez A et al (2016) A spray-drying continuousflow method for simultaneous synthesis and shaping of microspherical high nuclearity MOF beads. React Chem Eng 1:533–539. https://doi.org/10.1039/c6re00065g 30. Gimeno-Fabra M, Munn AS, Stevens LA et al (2012) Instant MOFs: continuous synthesis of metal–organic frameworks by rapid solvent mixing. Chem Commun 48:10642–10644. https:// doi.org/10.1039/c2cc34493a 31. Gordina NE, Prokof’ev VY, Il’in AP (2005) Extrusion molding of sorbents based on synthesized zeolite. Glas Ceram (English Transl Steklo i Keramika) 62:282–286. https://doi.org/10. 1007/s10717-005-0092-3 32. Grande CA, Águeda VI, Spjelkavik A, Blom R (2015) An efficient recipe for formulation of metal-organic frameworks. Chem Eng Sci 124:154–158. https://doi.org/10.1016/j.ces.2014. 06.048 33. Heilig ML (1994) United States patent office. ACM Siggraph Comput Graph 28:131–134 34. Holt EM (2004) The properties and forming of catalysts and absorbents by granulation. Powder Technol 140:194–202. https://doi.org/10.1016/j.powtec.2004.01.010 35. Howarth A, Liu Y, Li P et al (2016) Chemical, thermal and mechanical stabilities of metal– organic frameworks. Nat Rev Mater 1:15018. https://doi.org/10.1038/natrevmats.2015.18 36. Iveson SM, Beathe JA, Page NW (2002) The dynamic strength of partially saturated powder compacts: the effect of liquid properties. Powder Technol 127:149–161. https://doi.org/10. 1016/S0032-5910(02)00118-3 37. Iveson SM, Litster JD, Hapgood K, Ennis BJ (2001) Nucleation, growth and breakage phenomena in agitated wet granulation processes: a review. Powder Technol 117:3–39. https://doi.org/10.1016/S0032-5910(01)00313-8
2 Large-Scale Synthesis and Shaping of Metal-Organic Frameworks
51
38. Karadeniz B, Howarth AJ, Stolar T et al (2018) Benign by design: green and scalable synthesis of zirconium UiO-metal-organic frameworks by water-assisted mechanochemistry. ACS Sustain Chem Eng 6:15841–15849. https://doi.org/10.1021/acssuschemeng.8b04458 39. Khabzina Y, Dhainaut J, Ahlhelm M et al (2018) Synthesis and shaping scale-up study of functionalized UiO-66 MOF for ammonia air purification filters. Ind Eng Chem Res 57:8200–8208. https://doi.org/10.1021/acs.iecr.8b00808 40. Kim KJ, Li YJ, Kreider PB et al (2013) High-rate synthesis of Cu-BTC metal-organic frameworks. Chem Commun 49:11518–11520. https://doi.org/10.1039/c3cc46049e 41. Kim PJ, You YW, Park H et al (2015) Separation of SF6 from SF6/N2 mixture using metalorganic framework MIL-100(Fe) granule. Chem Eng J 262:683–690. https://doi.org/10.1016/ j.cej.2014.09.123 42. Kriesten M, Schmitz JV, Siegel J et al (2019) Shaping of flexible metal-organic frameworks: combining macroscopic stability and framework flexibility. Eur J Inorg Chem 2019:4700–4709. https://doi.org/10.1002/ejic.201901100 43. Kulprathipanja S (2010) Chemistry of zeolites and related porous materials molecular heterogeneous catalysis concepts of modern catalysis and kinetics handbook of porous solids catalysts for fine chemical synthesis, vol 6. Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim 44. Kumar P, Vejerano E, Khan A et al (2019) Metal organic frameworks (MOFs): currents trends and challenges in control and management of air quality. Korean J Chem Eng 36:1839–1853. https://doi.org/10.1007/s11814-019-0378-8 45. Kunowsky M, Marco-Lozar JP, Cazorla-Amorós D, Linares-Solano A (2010) Scale-up activation of carbon fibres for hydrogen storage. Int J Hydrog Energy 35:2393–2402. https://doi. org/10.1016/j.ijhydene.2009.12.151 46. www.basf.com 47. www.moftechnologies.com 48. www.mofapps.com 49. www.mofgen.com 50. www.mosaicmaterials.com 51. www.numat-tech.com 52. www.novomof.com 53. www.prometheanparticles.co.uk 54. www.strem.com 55. Küsgens P, Zgaverdea A, Fritz HG et al (2010) Metal-organic frameworks in monolithic structures. J Am Ceram Soc 93:2476–2479. https://doi.org/10.1111/j.1551-2916.2010. 03824.x 56. Laredo GC, Vega-Merino PM, Ascención Montoya-De La Fuente J et al (2016) Comparison of the metal-organic framework MIL-101 (Cr) versus four commercial adsorbents for nitrogen compounds removal in diesel feedstocks. Fuel 180:284–291. https://doi.org/10.1016/j.fuel. 2016.04.038 57. Lee DW, Didriksen T, Olsbye U et al (2020) Shaping of metal-organic framework UiO-66 using alginates: effect of operation variables. Sep Purif Technol 235:116182. https://doi.org/ 10.1016/j.seppur.2019.116182 58. Lee J, Farha OK, Roberts J et al (2009) Metal-organic framework materials as catalysts. Chem Soc Rev 38:1450–1459. https://doi.org/10.1039/b807080f 59. Li L, Yao J, Xiao P et al (2013) One-step fabrication of ZIF-8/polymer composite spheres by a phase inversion method for gas adsorption. Colloid Polym Sci 291:2711–2717. https://doi.org/ 10.1007/s00396-013-3024-8 60. Majano G, Pérez-Ramírez J (2013) Scalable room-temperature conversion of copper (II) hydroxide into HKUST-1 (Cu3(btc)2). Adv Mater 25:1052–1057. https://doi.org/10. 1002/adma.201203664
52
U.-H. Lee et al.
61. Mallick A, Mouchaham G, Bhatt PM et al (2018) Advances in shaping of metal-organic frameworks for CO2 capture: understanding the effect of rubbery and glassy polymeric binders. Ind Eng Chem Res 57:16897–16902. https://doi.org/10.1021/acs.iecr.8b03937 62. Marco-Lozar JP, Juan-Juan J, Suárez-García F et al (2012) MOF-5 and activated carbons as adsorbents for gas storage. Int J Hydrog Energy 37:2370–2381. https://doi.org/10.1016/j. ijhydene.2011.11.023 63. McKinlay AC, Morris RE, Horcajada P et al (2010) BioMOFs: metal-organic frameworks for biological and medical applications. Angew Chem Int Ed 49:6260–6266. https://doi.org/10. 1002/anie.201000048 64. McKinstry C, Cathcart RJ, Cussen EJ et al (2016) Scalable continuous solvothermal synthesis of metal organic framework (MOF-5) crystals. Chem Eng J 285:718–725. https://doi.org/10. 1016/j.cej.2015.10.023 65. Mitchell S, Michels NL, Majano G, Pérez-Ramírez J (2013) Advanced visualization strategies bridge the multidimensional complexity of technical catalysts. Curr Opin Chem Eng 2:304–311. https://doi.org/10.1016/j.coche.2013.04.005 66. Moreira MA, Santos JC, Ferreira AFP et al (2012) Reverse shape selectivity in the liquidphase adsorption of xylene isomers in zirconium terephthalate MOf UiO-66. Langmuir 28:5715–5723. https://doi.org/10.1021/la3004118 67. Mort PR (2005) Scale-up of binder agglomeration processes. Powder Technol 150:86–103. https://doi.org/10.1016/j.powtec.2004.11.025 68. Mueller U, Schubert M, Teich F et al (2006) Metal-organic frameworks – prospective industrial applications. J Mater Chem 16:626–636. https://doi.org/10.1039/b511962f 69. Munn AS, Dunne PW, Tang SVY, Lester EH (2015) Large-scale continuous hydrothermal production and activation of ZIF-8. Chem Commun 51:12811–12814. https://doi.org/10.1039/ c5cc04636j 70. Munusamy K, Sethia G, Patil DV et al (2012) Sorption of carbon dioxide, methane, nitrogen and carbon monoxide on MIL-101(Cr): volumetric measurements and dynamic adsorption studies. Chem Eng J 195–196:359–368. https://doi.org/10.1016/j.cej.2012.04.071 71. Musyoka NM, Ren J, Langmi HW, Rogers DEC, North BC, Mathe M, Bessarabov D (2014) Synthesis of templated carbons starting from clay and clay-derived zeolites for hydrogen storage applications. Int J Energy Res 39:494–503. https://doi.org/10.1002/er.3261 72. Oberacker R (2013) Powder compaction by dry pressing. Ceram Sci Technol 3–4:1–37. https://doi.org/10.1002/9783527631940.ch32 73. Schoenecker PM, Belancik GA, Grabicka BE, KSW (2013) Kinetics study and crystallization process design for scale-up of UiO-66-NH2 synthesis. AICHE J 59:1255–1262 74. Pinto ML, Dias S, Pires J (2013) Composite MOF foams: the example of UiO-66/ polyurethane. ACS Appl Mater Interfaces 5:2360–2363. https://doi.org/10.1021/am303089g 75. Plaza MG, Ferreira AFP, Santos JC et al (2012a) Propane/propylene separation by adsorption using shaped copper trimesate MOF. Microporous Mesoporous Mater 157:101–111. https:// doi.org/10.1016/j.micromeso.2011.06.024 76. Plaza MG, Ribeiro AM, Ferreira A et al (2012b) Propylene/propane separation by vacuum swing adsorption using Cu-BTC spheres. Sep Purif Technol 90:109–119. https://doi.org/10. 1016/j.seppur.2012.02.023 77. Polyzoidis A, Altenburg T, Schwarzer M et al (2016) Continuous microreactor synthesis of ZIF-8 with high space-time-yield and tunable particle size. Chem Eng J 283:971–977. https:// doi.org/10.1016/j.cej.2015.08.071 78. Purewal JJ, Liu D, Yang J et al (2012) Increased volumetric hydrogen uptake of MOF-5 by powder densification. Int J Hydrog Energy 37:2723–2727. https://doi.org/10.1016/j.ijhydene. 2011.03.002 79. Quinn DF, MacDonald JA (1992) Natural gas storage. Carbon NY 30:1097–1103. https://doi. org/10.1016/0008-6223(92)90141-I
2 Large-Scale Synthesis and Shaping of Metal-Organic Frameworks
53
80. Ren J, Dyosiba X, Musyoka NM et al (2017) Review on the current practices and efforts towards pilot-scale production of metal-organic frameworks (MOFs). Coord Chem Rev 352:187–219. https://doi.org/10.1016/j.ccr.2017.09.005 81. Ren J, North B (2014) Shaping porous materials for hydrogen storage applications: a review. J Technol Innov Renew Energy 3:12–20. https://doi.org/10.6000/1929-6002.2014.03.01.3 82. Rezaei F (2009) Optimum structured adsorbents for gas separation processes. Chem Eng Sci 64:5182–5191. https://doi.org/10.1016/j.ces.2009.08.029 83. Rezaei F, Webley PA (2012) Optimal design of engineered gas adsorbents: pore-scale level. Chem Eng Sci 69:270–278. https://doi.org/10.1016/j.ces.2011.10.039 84. Ribeiro RPPL, Antunes CL, Garate AU et al (2019) Binderless shaped metal-organic framework particles: impact on carbon dioxide adsorption. Microporous Mesoporous Mater 275:111–121. https://doi.org/10.1016/j.micromeso.2018.08.002 85. Richardson JJ, Tardy BL, Guo J et al (2019) Continuous metal-organic framework biomineralization on cellulose nanocrystals: extrusion of functional composite filaments. ACS Sustain Chem Eng 7:6287–6294. https://doi.org/10.1021/acssuschemeng.8b06713 86. Richardson JT, Peng Y, Remue D (2000) Properties of ceramic foam catalyst supports: pressure drop. Appl Catal A Gen 204:19–32. https://doi.org/10.1016/S0926-860X(00) 00508-1 87. Rubio-Martinez M, Avci-Camur C, Thornton AW et al (2017) New synthetic routes towards MOF production at scale. Chem Soc Rev 46:3453–3480. https://doi.org/10.1039/c7cs00109f 88. Rubio-Martinez M, Batten MP, Polyzos A et al (2014) Versatile, high quality and scalable continuous flow production of metal-organic frameworks. Sci Rep 4:5443. https://doi.org/10. 1038/srep05443 89. Rubio-Martinez M, Hadley TD, Batten MP et al (2016) Scalability of continuous flow production of metal-organic frameworks. ChemSusChem 9:938–941. https://doi.org/10. 1002/cssc.201501684 90. S-a S (1981) The selection of Pelletisers meets the requirements of subsequent of pellet growth. Powder Technol 29:211–216 91. Schwab MG, Senkovska I, Rose M et al (2008) MOF@PolyHIPEs. Adv Eng Mater 10:1151–1155. https://doi.org/10.1002/adem.200800189 92. Semino R, Moreton JC, Ramsahye NA et al (2018) Understanding the origins of metal-organic framework/polymer compatibility. Chem Sci 9:315–324. https://doi.org/10.1039/c7sc04152g 93. Seo YK, Yoon JW, Lee JS et al (2012) Large scale fluorine-free synthesis of hierarchically porous iron(III) trimesate MIL-100(Fe) with a zeolite MTN topology. Microporous Mesoporous Mater 157:137–145. https://doi.org/10.1016/j.micromeso.2012.02.027 94. Shah BB, Kundu T, Zhao D (2019) Mechanical properties of shaped metal–organic frameworks. Springer, Cham 95. Spencer EC, Angel RJ, Ross NL et al (2009) Pressure-induced cooperative bond rearrangement in a zinc imidazolate framework: a high-pressure single-crystal X-ray diffraction study. J Am Chem Soc 131:4022–4026. https://doi.org/10.1021/ja808531m 96. Spjelkavik AI, Aarti DS et al (2014) Forming MOFs into spheres by use of molecular gastronomy methods. Chem – A Eur J 20:8973–8978. https://doi.org/10.1002/chem. 201402464 97. Spokoyny AM, Kim D, Sumrein A, Mirkin CA (2009) Infinite coordination polymer nanoand microparticle structures. Chem Soc Rev 38:1218–1227. https://doi.org/10.1039/b807085g 98. Sternberg CN, Hawkins RE, Wagstaff J et al (2013) Re: a randomised, double-blind phase III study of pazopanib in patients with advanced and/or metastatic renal cell carcinoma: final overall survival results and safety update. J Urol 190:2017. https://doi.org/10.1016/j.juro. 2013.08.043 99. Taddei M, Steitz DA, Van Bokhoven JA, Ranocchiari M (2016) Continuous-flow microwave synthesis of metal-organic frameworks: a highly efficient method for large-scale production. Chem – A Eur J 22:3245–3249. https://doi.org/10.1002/chem.201505139
54
U.-H. Lee et al.
100. Tagliabue M, Rizzo C, Millini R et al (2011) Methane storage on CPO-27-Ni pellets. J Porous Mater 18:289–296. https://doi.org/10.1007/s10934-010-9378-0 101. Tai S, Zhang W, Zhang J et al (2016) Facile preparation of UiO-66 nanoparticles with tunable sizes in a continuous flow microreactor and its application in drug delivery. Microporous Mesoporous Mater 220:148–154. https://doi.org/10.1016/j.micromeso.2015.08.037 102. Tan JC, Bennett TD, Cheetham AK (2010) Chemical structure, network topology, and porosity effects on the mechanical properties of zeolitic imidazolate frameworks. Proc Natl Acad Sci U S A 107:9938–9943. https://doi.org/10.1073/pnas.1003205107 103. Tranchemontagne DJ, Mendoza-Cortés JL, O’Keeffe M, Yaghi OM (2009) Secondary building units, nets and bonding in the chemistry of metal-organic frameworks. Chem Soc Rev 38:1257–1283. https://doi.org/10.1039/b817735j 104. Uemura T, Yanai N, Kitagawa S (2009) Polymerization reactions in porous coordination polymers. Chem Soc Rev 38:1228–1236. https://doi.org/10.1039/b802583p 105. Valekar AH, Lee SG, Cho KH et al (2017) Shaping of porous metal-organic framework granules using mesoporous ρ-alumina as a binder. RSC Adv 7:55767–55777. https://doi.org/ 10.1039/c7ra11764g 106. Valizadeh B, Nguyen TN, Stylianou KC (2018) Shape engineering of metal–organic frameworks. Polyhedron 145:1–15. https://doi.org/10.1016/j.poly.2018.01.004 107. Waitschat S, Wharmby MT, Stock N (2015) Flow-synthesis of carboxylate and phosphonate based metal-organic frameworks under non-solvothermal reaction conditions. Dalton Trans 44:11235–11240. https://doi.org/10.1039/c5dt01100k 108. Walker WJ, Reed JS, Verma SK (1999) Influence of slurry parameters on the characteristics of spray-dried granules. J Am Ceram Soc 82:1711–1719. https://doi.org/10.1111/j.1151-2916. 1999.tb01990.x 109. Webley P (2010) Structured adsorbents in gas separation processes. Sep Purif Technol 70:243–256. https://doi.org/10.1016/j.seppur.2009.10.004 110. https://www.basf.com 111. Yaghi O, Angeles L, Arbor A et al (2009) Shaped bodes containing metal-organic frameworks. 2 112. Yilmaz B, Trukhan N, Müller U (2012) Industrial outlook on zeolites and metal organic frameworks. Cuihua Xuebao/Chin J Catal 33:3–10. https://doi.org/10.1016/s1872-2067(10) 60302-6 113. Yoon JW, Seo YK, Hwang YK et al (2010) Controlled reducibility of a metal-organic framework with coordinatively unsaturated sites for preferential gas sorption. Angew Chem Int Ed 49:5949–5952. https://doi.org/10.1002/anie.201001230 114. Yu L, Dean K, Li L (2006) Polymer blends and composites from renewable resources. Prog Polym Sci 31:576–602. https://doi.org/10.1016/j.progpolymsci.2006.03.002 115. Zacahua-Tlacuatl G, Pérez-González J, Castro-Arellano JJ, Balmori-Ramírez H (2010) Rheological characterization and extrusion of suspensions of natural zeolites. Appl Rheol 20:1–10. https://doi.org/10.3933/ApplRheol-20-34037 116. Zacher D, Shekhah O, Wöll C, Fischer RA (2009) Thin films of metal-organic frameworks. Chem Soc Rev 38:1418–1429. https://doi.org/10.1039/b805038b 117. Zhang Y, Cai J, Zhang D et al (2018) Shaping metal-organic framework materials with a honeycomb internal structure. Chem Commun 54:3775–3778. https://doi.org/10.1039/ c8cc01289j 118. Zhao T, Jeremias F, Boldog I et al (2015) High-yield, fluoride-free and large-scale synthesis of MIL-101(Cr). Dalton Trans 44:16791–16801. https://doi.org/10.1039/c5dt02625c 119. Zheng J, Cui X, Yang Q et al (2018) Shaping of ultrahigh-loading MOF pellet with a strongly anti-tearing binder for gas separation and storage. Chem Eng J 354:1075–1082. https://doi.org/ 10.1016/j.cej.2018.08.119
Chapter 3
Green Energy Generation Using Metal-Organic Frameworks Giacomo Armani-Calligaris, Sara Rojas Macías, Víctor Antonio de la Peña O’Shea, and Patricia Horcajada Cortés
Abbreviations 2bpy 4bpy AB AC AcH AgE Azene bib cat CB CNTs CPE CPs Cyclam dc DFT DMF DoS Eg EPhD
2,20 -Bipyridine 4,40 -Bipyridine Acetylene black Activated carbon Acetic acid Silver chloride electrode, Ag|AgCl|Cl (E)-1,2-Di(pyridin-4-yl)diazene 1,4-Bis(imidazol)butane Catalyst Carbon black Carbon nanotubes Carbon paste electrode Coordination polymers 1,4,8,11-Tetraazacyclotetradecane Drop casting Density-functional theory N,N-Dimethylformamide Density of state analysis Energy bandgap Electrophoretic deposition
G. Armani-Calligaris Advanced Porous Materials Unit, IMDEA Energy Institute, Móstoles, España Photoactivated Process Unit, IMDEA Energy Institute, Móstoles, España S. Rojas Macías · P. Horcajada Cortés (*) Advanced Porous Materials Unit, IMDEA Energy Institute, Móstoles, España e-mail: [email protected] V. A. de la Peña O’Shea Photoactivated Process Unit, IMDEA Energy Institute, Móstoles, España © Springer Nature Switzerland AG 2021 P. Horcajada Cortés, S. Rojas Macías (eds.), Metal-Organic Frameworks in Biomedical and Environmental Field, https://doi.org/10.1007/978-3-030-63380-6_3
55
56
FCs FTO GCE H21,4-NDC H25-bdc H2adip H2bbta H2BDC H2BDC-NH2 H2BPDC H2DSPTP H2isop H2L3 H2mna H2TD H3BPTC H3BTB H3BTC H3BTT H3dcpna H3pdc H3TATB H3TCBPA H3TCPA H4abtc H4TBP H4TBpyz H6BHB H6HIB H6HITP H8phth HBA Hbim HER HHTP HKUST HL4 HL7 HOMO H-pycz HR-TEM Htz IPCC IUPAC
G. Armani-Calligaris et al.
Fuel cells Fluorine-doped tin oxide Glassy carbon electrode Naphthalene-1,4-dicarboxylic acid 5-Nitroisophthalic acid Adipic acid 1H,5H-Benzo-(1,2-d:4,5-d0 )bistriazole 1,4-Benzenedicarboxylic or terephthalic acid Aminoterephthalic acid 1,10 -Biphenyl-4,40 -dicarboxylic acid 40 -(2,4-Disulfophenyl)-3,20 :60 ,300 -terpyridine Isophthalic acid 4,5-Di(40 -carboxylphenyl)phthalic acid 2-Mercaptonicotinic acid 1-Thia-3,4-diazole-2,5-dithiol Biphenyl-3,40 ,5-tricarboxylic acid 1,3,5-Tris(4-carboxylphenyl)benzene Trimesic acid 5,50 ,500 -(1,3,5-Phenylene)tris(1H-tetrazole) 5-(30 ,50 -Dicarboxylphenyl)nicotinic acid Pyrazole-3,5-dicarboxylic acid 4,40 ,400 -(1,3,5-Triazine-2,4,6-triyl)tribenzoic acid Tris(40 -carboxybiphenyl)amine Tris(4-carboxylphenyl)amine 3,30 ,5,50 -Azobenzenetetracarboxylic acid 5,10,15,20-Tetra( p-benzoic acid)porphyrin 2,3,5,6-Tetra(4-carboxyphenyl)pyrazine 5,50 ,500 -Benzene-1,3,5-triyl-isophthalic acid Hexaminobenzene 2,3,6,7,10,11-Hexaminotriphenylene 2,3,9,10,16,17,23,24-Octaaminophthalocyaninato N-(2-Hydroxybenzyl)-alanine Benzimidazole Hydrogen evolution reaction Hexahydroxytriphenylene Hong Kong University of Science and Technology 4-(5-(Pyridin-4-yl)-4H1,2,4-triazol-3-yl)benzoic acid 5-Sulphidyl-1-methyltetrazole acetic acid Highest occupied molecular orbital 3-(Pyrid-40 -yl)-5-(400 -carbonylphenyl)-1,2,4-triazolyl High-resolution transmission electron microscopy Triazole International Panel on Climate Change International Union of Pure and Applied Chemistry
3 Green Energy Generation Using Metal-Organic Frameworks
L1 L2 L6 LLCT LMCT LUMO MeCN MeOH MIL MIm MLCT MOE MOFs NHE NIR NPs NU OER ORR ox PBS PCN phen POMOFs POMs PXRD py PYZ pyz RD RDS RHE RT sat SBET SBU SCE SHE SP STh TBA TEA TEOA tib
4H-4Amino-1,2,4-triazole 3,5-Dimethyl-4-amino-4H-1,2,4-triazole 1,4-Bis(3-pyridylaminomethyl)benzene Ligand-to-ligand charge transfer Ligand-to-metal charge transfer Lowest unoccupied molecular orbital Acetonitrile Methanol Matériaux Institut Lavoisier 2-Methylimidazole Metal-to-ligand charge transfer Mercury oxide electrode, Hg|HgO|OH Metal-organic frameworks Normal hydrogen electrode Near infrared Nanoparticles Northwestern University Oxygen evolution reaction Oxygen reduction reaction Oxalate Phosphate buffer solution Porous coordination network 1,10-Phenanthroline POM-based MOFs Polyoxometalates Powder X-ray diffraction Pyridine Pyrazine Pyridazine Rotating disk electrode Rate-determining step Reversible hydrogen electrode Room temperature Saturated Brunauer-Emmett-Teller surface area Secondary building unit Saturated calomel electrode Standard hydrogen electrode Super P, Timcal Solvothermally deposited Tetrabutylammonium Triethylamine Triethanolamine 1,3,5-Tris(1-imidazolyl)benzene
57
58
TOF TOFeq TON UiO Vp WS wt ZSTU
3.1
G. Armani-Calligaris et al.
Turnover frequency Equivalent TOF Turnover number Universitetet i Oslo Pore volume Water splitting Weight ratio Zhejiang Sci-Tech University
General Introduction
Energy is essential to biological life and for its technological use by humankind. Regrettably, the current total energy demand is mostly supplied by either inefficiently burning biomass or fossil fuels. In both cases, harmful by-products and greenhouse gases are produced, generating air pollution and the climate change we are experiencing in the last decades [1]. Because of these alarms and the depletion of fossil fuels, society is now facing the major challenge of moving from nonrenewable carbon energy sources to more sustainable ones. From a biological and technological point of view, most of the energy production procedures are related with electron transfer processes (i.e., electron transport chain in cells). Electricity is one of the most common and renewable ways to produce energy. However, the main drawback is the need of storage systems to make it available through time. Among the different possibilities to store it, chemical ones lead to the unique advantages of allowing undefined shelf-life and the direct utilization of by-products, so they can be converted into value-added compounds. On the other hand, H2 completely independent from any carbon requirement has emerged as a potential energy carrier to substitute fossil fuels, as its combustion produces just water. As molecular hydrogen reservoirs do not exist naturally on Earth, this must be generated. However, green H2 technologies are not quickly accessible, as they are associated with low generation efficiencies and high costs. In 2019, 68% of the world’s 70 million tons of H2 consumed yearly in industrial processing was produced by steam reforming, which uses fuels such as natural gas [2], while less than 0.1% of global H2 production came from water electrolysis [3]. Always bearing in mind the free access to drinking water of all humans, water splitting (WS, Eq. 3.1) is the ideal reaction to produce H2, because of the accessible character of water and the only other product is O2. Indeed, O2 is the real source of electrons to produce H2. In order to recover the generated gases separately, WS is usually performed by activating the two half reactions far from each other, so it is common to refer to the hydrogen evolution reaction (HER, Eq. 3.2, at standard conditions in alkaline medium) and to the oxygen evolution reaction (OER, Eq. 3.3, at standard conditions in alkaline medium), being energy-demanding processes that require external electricity or solar energy input:
3 Green Energy Generation Using Metal-Organic Frameworks
59
energy
2H2 O ! 2H2 þ O2
2H2 O þ 2e ! H2 ðgÞ þ 2OH 4OH ! O2 þ 2H2 O þ 4e
ð3:1Þ
ð3:2Þ ð3:3Þ
On the other side, in the use of H2 as fuel generating H2O, the two previous half reactions are still involved, but in the opposite direction, as hydrogen oxidation (HOR) and oxygen reduction reactions (ORR). ORR is involved in classical combustions, in animal and plants respiration and, from a technological point of view, in the development of fuel cells (FCs) and metal-air batteries, where oxygen acts as the final e acceptor. In all these reactions, catalysts are needed, because of the inertness of the involved molecules. Traditionally, noble metals (e.g., Pt, RuOX, and IrOX [4– 6]) are regarded as the best HER and OER catalysts. Unfortunately, their high cost and limited resources have also made these catalysts the primary limitation for practical use. Until now, non-noble metal-based inorganic nanomaterials and functional molecular materials have been thoroughly explored for WS, like transition metal (Mo, Ni, Co, and Fe)-based sulfides [7, 8], carbides, and nitrides [9, 10], novel carbon materials (graphene and nanotube structures [11]), and also clusters from molecules to single-atom levels [12–14]. However, low light utilization efficiency, improper band position, fast recombination of charge carriers, and photocorrosion have accelerated the investigation on new strategies to design more efficient catalysts and, thus, to advance into this appealing technology. Besides their traditional applications in the chemical industry, porous materials, which are able to store energy carriers or facilitate fast mass and electron transportation for energy storage and conversion, have been explored for the development of innovative energy technologies. Among them, metal-organic frameworks (MOFs) are good candidates in energy applications, benefitting from the ordered structural diversity and tailorability. Indeed, the International Union of Pure and Applied Chemistry (IUPAC) has recently described MOFs as one of the top 10 emerging technologies in chemistry [15]. Comparing with classical inorganic porous materials, MOFs have several features that attracted researchers’ interest: (i) their natural porous structure with tunable pore sizes and topologies can facilitate diffusion and accelerate transportation of charge carriers, improving reaction rate but also selectivity; (ii) their large surface area can offer a large amount of active sites (i.e., metal ion or cluster, defects) to participate in the catalytic process; (iii) MOFs can be used not only as light-harvesting materials and catalysts but also as rigid and stable supports for other catalytically active species (e.g., metal [16], salt or chalcogenide/pnictide nanoparticles (NPs) [17], polyoxometalates (POMs) [18], enzymes [19]); and (iv) confinement effects within MOFs porosity allow selective adsorption of reagents, stabilizing encapsulated catalytic species, and precise location of catalytic groups (improving selectivity and efficiency, tandem reactions, etc.). Finally, MOFs can also be used as precursors of carbon catalysts [11].
60
3.2
G. Armani-Calligaris et al.
Initial Considerations
This book chapter is focused on the main catalytic processes in which MOFs are studied for up-conversion of molecules in energy applications, photocatalysis and electrocatalysis, as well as on the design strategies to improve MOFs performance for these applications. For clarity, each section is organized considering the studied reaction (i.e., HER, OER, and ORR) and some general (e.g., active metal in photocatalysis, reaction medium in electrocatalysis) and/or specific parameters involved. From an IUPAC recommendation, coordination polymers (CPs) have been defined as “coordination compounds with repeating coordination entities extending in 1, 2, or 3 dimensions,” while “a metal–organic framework, abbreviated to MOF, is a coordination network with organic ligands containing potential voids” [20]. These controversial definitions do not take into account the crystallinity of the network. In our particular case, we have included here only crystalline CPs (1, 2, and 3D) and MOFs (only potentially porous 3D structures) in order to highlight the importance of the structure-property correlation. Although the most striking results in this field have been obtained using MOFbased composites (by the association of different species with the MOF) or MOF-derived carbons (from MOF degradation) [21], our attention will be devoted to pure MOFs, unless these strategies (composites and carbons) have also provided us valuable discussion. Indeed, these derivatives are inherently expensive because of involving multistep processes or the use of high-value metals [22]. Pure MOFs, as well as their defective versions, have demonstrated their potential as photo- and electrocatalysts, even if this research field is still at its infancy. Comparison between the published results is however a hard task since different conditions are tested (e.g., reaction medium, irradiation density and quality, current collector), even for similar materials. In addition, previous works were either focused to the structural characterization or to the application properties, lacking sometimes from useful details for an accurate comparison. Apart from the specific activity comparison, there is also the need to find the optimal working conditions for each entire system, so the “best catalysts” are often comparable just if the whole catalyst system (catalyst, reactor or electrode design, reagents) is taken into account and not only individual contributions (e.g., catalysts’ mechanism). On the other hand, MOFs can be designed in so many different ways, being often deeply different for photo-and electrocatalytic applications. Thus, here we have performed a specific analysis of the main properties influencing each catalysis type. Moreover, it has to be mentioned here that the stability to reactants in specific reaction conditions is one of the main limitations of the use of MOFs as catalyst, so a careful survey of the MOFs stability under each reaction conditions should be mandatory [23]. In WS and generation reactions, MOFs need to be stable also along all the pH range. Particularly, HER is generally favored in acidic medium, while OER is favored in alkaline conditions, for both thermodynamic (e.g., reagent concentrations) and kinetic reasons (e.g., adsorption, species involved, acid-base reactions). In order to reach a complete device that can perform both OER and
3 Green Energy Generation Using Metal-Organic Frameworks
61
HER/ORR, catalysts must be robust along all the pH range. While there are some examples of HER in basic media [24] or HER/OER/ORR at around neutral pH, the vast majority of reports use incompatible reaction media to support a dual WS.
3.2.1
Parameters Affecting Photocatalysis
Although quantum yield (Φ ¼ number of events per photons absorbed) is an appropriate parameter to evaluate the photocatalytic activity of a heterogeneous catalyst, MOFs activity is usually reported as productivity (mmol or mL per hour per gram of catalyst), a comfortable unit for industrial purposes, but an inappropriate unit if one wants to compare the catalytic activity of the molar structure. TOF is defined as the turnover number (TON, Eq. 3.4) per time (Eq. 3.5): np nc
ð3:4Þ
TON t
ð3:5Þ
TON ¼ TOF ¼
with np as the produced moles of product, nc as the moles of precatalyst in the system (as from unit formula), and t as the reaction time (s). However, this is not well-adapted to compare the activity of MOFs with different number of active sites (i.e., activity of a single active site). Thus, the discussion on MOFs’ effectiveness as photocatalysts will be discussed by calculating the here denoted “equivalent TOF” (TOFeq, Eq. 3.6), which takes into account the number of reported active metal sites within a specific MOF: TOFeq ¼
TOF m
ð3:6Þ
with m as the total metal centers per unit formula potentially involved in the catalytic activity. In some publications, produced gases are expressed in volume. In these cases, the ideal gas equation was used in the conversion (assuming a pressure of 1 atm and a temperature of 20 C when the reaction conditions were not expressed). Photocatalytic activity of MOFs is significantly affected by different parameters, being here some of them further discussed.
3.2.1.1
Surface Area Effect
One of the main advantages of using MOFs in catalysis is the exploitation of their high surface areas. When used as photocatalysts, this property has been repeatedly exploited, even when textural properties were not deeply analyzed. However, one should consider both the outer particle surface and the porous surface, as well as the
62
G. Armani-Calligaris et al.
diffusion of reagents and products through them. An enlightening work dealing with the effect of external particle surface in catalysis, instead of just the microporous surface area, used the benchmark submicrometric MIL-125-NH2 (MIL stands for Matériaux Institut Lavoisier, [Ti8O8(OH)4(BDC-NH2)6]; H2BDC-NH2, aminoterephthalic acid [25]; particle size ~ 600 nm) exposing different planes of the particle surface [26]. It was shown that photocatalysis happened mainly at the surfaces, as the activity changed with the exposed MOF planes. Planes with the highest surface free energy ((110) planes) led to an increase of about 48% of the H+ reduction activity (using TEOA – triethanolamine – as sacrificial agent), compared to the most commonly exposed plane (001) (see Table 3.1). On the other side, particles exhibiting only (111) planes showed a halved TOFeq (1.15106 s1). The predominance of the particle surface activity can be rationalized considering the hindered diffusion of the large TEOA through the MIL-125-NH2 pores, as known for other MOF-based catalysts [27, 28].
3.2.1.2
Active Cluster
Metal cations are the main component around which HER, OER, and ORR have been developed, as their redox properties and chemical strength allow their repeated use. The choice of a proper metal to synthetize a photocatalyst is not trivial, as its orbital levels must match those required by the target reaction, while allow a favorable charge transfer from or to the antenna. In this regard, a computational comparison of UiO-66 (UiO stands for Universitetet i Oslo; [M6O4(OH)4(BDC)6] nH2O; H2BDC, 1,4-benzenedicarboxylic acid) based on different metals (M: Ti, Zr, Hf, Ce, Th, U) revealed the effect of their energy levels on the photocatalytic activity of UiO-66(M) [29]. Orbitals associated with the Zr-, Hf-, Th-, and U-based structures do not favor a ligand-to-metal charge transfer (LMCT), thus limiting possible photocatalytic reduction processes. In contrast, UiO-66(Ti) may show a good mixing of its 3D orbitals with the ligand π* ones, giving an ELMCT ¼ 0 eV, thus enhancing the charge transfer. Even better, the low positions of CeIV orbitals lead to a highly favored LMCT, as CeIV has a lowest unoccupied molecular orbital (LUMO) of lower energy than BDC ligand and is easily reduced to CeIII, thus giving negative ELMCT. However, the very low energy levels of UiO-66(Ce) are associated with a high energy bandgap (Eg) (mainly caused by linker orbitals), thus limiting its application. It is worth noting that UiO-66(Ce) has indeed been experimentally obtained [30]. In a second study, the Zr- or Ti-doping effect was studied on the UiO-66(Ce). These doping metal oxides can produce solid solutions with CeO2, causing a decrease of the LUMO in the pristine UiO-66(Ce) [31], thus again reducing its Eg. The results showed that the bandgap decreased when introducing Zr or Ti, although visible light absorption was not reached (Eg < 3.4 eV). On the other side, the work of Wu et al. [32] showed a photoactive Gd-MOF, which was synthesized from the triphenylamine ligand tris(4-carboxyphenyl)amine (H3TCPA) for the proton and carbon dioxide reduction. The combination of proton reduction and/or carbon dioxide reduction catalyst (i.e., the Fe-Fe hydrogenase active-site
[ZII(2bpy)3] [MIRuIII(ox)3] Gd-MOF [Cu(py)]4[βMo8O26]
Cu-X-bpy (X ¼ Cl, Br, I)
Material MIL-125-NH2 MIL-125 UiO-66 UiO-66-NH2 [Cd(TD)] [Cd(TD)(H2O)] Dy2(abtc) (H2O)2(OH)2 Ru-TBP Ru-TBP-Zn [Cd(DSPTP) (H2O)2]2H2O [Cu(DSPTP) (H2O)2] [Cu (HDSPTP)2(H2O)3]
2.35 1.8
1.85 1.90 2.00 1.67–2.68
2.85
2.39
1.86
a
3.2 3.4 2.17
a
a
Bandgap (eV)
H2O, TEA (41.1% v/v), Hg lamp 500 W, λ < 366 nm (or Xe, 150 W, λ > 417 nm), cooled H2O, TEOA (40% v/v), Xe arc lamp 300 W, pyrex wall H2O:MeOH ¼ 4:1, Xe lamp 1000 W (1 Wcm2), λ > 200 nm or > 420 nm
H2O, pH ¼ 7, TEOA (10% v/v), Xe lamp 300 W (800 > λ > 420 nm), but also just NIR (monochromatic 700, 800, 900 nm), RT H2O, pH ¼ 11.5, TEA (5% v/v), Hg lamp 500 W (UV light), 20 C
H2O, pH ¼ 7, TEOA (10% v/v), Xe lamp 300 W (800 > λ > 420 nm), RT, for Cu also just NIR (λ > 800 nm), RT
MeCN + H2O (3.8% v/v), TEOA (19.2% v/v), solid state, 230 W (λ > 400 nm)
pH ¼ 9, TEOA (15% v/v), H2O, Xe lamp 300 W, 20 C pH ¼ 10, TEOA (10% v/v), H2O, Xe lamp 300 W, 20 C H2O, TEOA (30% v/v), Xe lamp 300 W (λ > 320 nm), RT
MeOH:H2O (1:3), Hg (Xe doped) lamp 200 W, pyrex filtered, 38 C
Experimental conditions H2O, TEOA (0.01 M), Xe lamp 500 W, >420 nm
154 316
3760 5340 7090 Various
1154
1397
130 240 313
Production rate (μmolg1h1) 167 0 200 348 26100 6900 2.4
Table 3.1 Overview of the key information of pure MOFs and CPs with photoactivity against HER found in literature
1.66
36.0 60.9 421.4
3.7
42.6
21.2 39.4 2.6
38.9 9.6 8.6103
TON 2.3104 0
1.93105
2.67104 4.44104 6.83104
3.38104
2.20104
3.07105 5.70105 5.67105
TOFeq (eqs1) 9.57106 0 1.54105 2.83105 1.89103 5.34104 2.38107
(continued)
[87] [84]
[78]
[81]
[39]
[39, 80]
[77]
[86]
[83]
[85]
Refs. [22]
3 Green Energy Generation Using Metal-Organic Frameworks 63
a
Bandgap (eV)
Experimental conditions MeCN, H2O (3.3% v/v), TEOA (5% v/v), Xe lamp 300 W (400 nm λ), pyrex TON 3.03 0.81 0.33 0.68
Production rate (μmolg1h1) 60.8 49 20 41
2.81106 1.15106 2.35106
TOFeq (eqs1) 3.49106 Refs. [26]
H2TD, 1,3,4-thiadiazole-2,5-dithiol; H4abtc, 3,30 ,5,50 -azobenzenetetracarboxylic acid; H2DSPTP, 40 -(2,4-disulfophenyl)-3,20 :60 ,300 -terpyridine; ox, oxalate; TEA, trimethylamine; TEOA, triethanolamine; MeOH, methanol; MeCN, acetonitrile; NIR, near infrared; RT, room temperature a Datum not reported
Material MIL-125-NH2 T110 T100 T111 T001
Table 3.1 (continued)
64 G. Armani-Calligaris et al.
3 Green Energy Generation Using Metal-Organic Frameworks
65
model and Ni(cyclam) complexes – cyclam: 1,4,8,11-tetraazacyclotetradecane) initiated a photoinduced single-electron transfer from its excited state to the substrate. Gd3+ was intentionally selected to avoid any energy transfer to the metal center, thanks to the lack of available orbitals, allowing a more effective energy transfer to the homogeneous catalyst. Another important effect on the metal cluster is indirectly provided by its chemical environment, formed by the coordinated ligands. The presence of functional groups on the aromatic linkers can strongly influence the intrinsic catalytic activity of the metal nodes through inductive effects [33]. This strategy was confirmed in OER by a series of MIL-88B structures or [Fe3O(OH)(BDC)3] based on the BDC ligand bearing different substituents (2CH3, 4CH3, OH, 2OH, NH2) [34]. These results are further discussed in Sect. 3.3.2.
3.2.1.3
Light Absorption
Ligand selection and its engineering strongly affect the MOF absorption bandgap [35, 36]. For instance, a series of Ti-Zr mixed-metal MOFs, based on the cluster [Ti8Zr2O12(COO)16] and different proportions between BDC, BDC-NH2, and 2,5-diaminoterephthlate (BDC-(NH2)2) ligands (PCN-415; PCN, porous coordination network; Fig. 3.1), was applied for HER using Pt NPs as cocatalyst [37]. Despite the higher visible light absorption of BDC-(NH2)2, it resulted in a lower activity toward HER compared to the BDC-NH2-based material. This was rationalized saying that the diaminated ligand was not able to provide an efficient LMCT, probably having internal π- π* transitions that would decay back to its fundamental state.
3.2.1.4
Excitation Lifetime/Rate-Determining Step
Although a photoreaction depends on a variety of contributions, its rate is determined by the rate-determining step (RDS). Very recently, the first isoreticular expansion of a Ti-MOF series (ZSTU-1, ZSTU-2, and ZSTU-3; ZSTU, Zhejiang
Fig. 3.1 The PCN-415 cluster seen as expansion of the UiO-66 cluster. (Reproduced with permission from Ref. [37]. Further permissions related to the material excerpted should be directed to the ACS)
66
G. Armani-Calligaris et al.
Sci-Tech University) was successfully constructed from infinite [Ti6(μ-O)6(μ-OH)6]n secondary building units (SBUs) and different tritopic carboxylate linkers, H3TCPA, 1,3,5-tris(4-carboxylphenyl)benzene (H3BTB), and tris(40 -carboxybiphenyl)amine (H3TCBPA), respectively (Fig. 3.2) [38]. Their photocatalytic activity was tested for HER with visible light irradiation (λ > 420 nm) in the presence of Pt as cocatalyst and indirectly correlated to their RDSs. ZSTU-1 and ZSTU-3 were the most active species, thanks to their low bandgaps (Eg ¼ 2.3 and 2.2 eV, respectively, in comparison with Eg ¼ 3.1 eV for ZSTU-2), resulting from the triphenylamine moiety. In addition, ZSTU-3 showed a higher excitation lifetime compared to ZSTU-1 (τ ¼ 453.6 vs. 193.1 ps, respectively), reported as a possible consequence of its extended π-conjugation, leading to the highest H2 production among the series. Therefore, RDS seems to be related to the charge separation state, deriving from the poor visible light absorption of ZSTU-2 and the different excitation lifetimes of ZSTU-1 and ZSTU-3. In 2017, two CPs based on Cu and a 3,300 -terpyridine ligand 0 (4 -(2,4-phenyldisulfonate)-3,20 :60 ,300 -terpyridine, H2DSPTP) were synthesized [39], leading to a 2D [Cu(HDSPTP)2(H2O)3] and a 3D [Cu(HDSPTP)(H2O)2] structures. Upon HER evaluation, the 3D-CP presented a higher H2 productivity than the 2D-CP (see Table 3.1). Indeed, the 3D-CP presented a slightly lower Eg, a lower photoluminescence spectrum intensity, and a higher photocurrent compared to the 2D-CP. The authors rationalized these data by the improved π-π interactions between the ligands in the 3D-CP, which might avoid a fast recombination of the generated charges. However, the lower observed τ value for 3D-CP does not seem to support this rationalization (τ ¼ 720 vs. 190 ps for the 2D- and 3D-CP, respectively). Indeed, converting productivities in TOFeqs, it turns out that the 2D-CP catalyst is 53% more active than the 3D-CP. Thus, the lower excitation lifetimes would derive
Fig. 3.2 Ligands used in Ref. [38], from the left: H3TCPA (ZSTU-1), H3TCBPA (ZSTU-3), H3BTB (ZSTU-2), respectively. (Figure performed using MolView web page and MS Paint software)
3 Green Energy Generation Using Metal-Organic Frameworks
67
from a weaker metal-ligand interaction, as a consequence of the higher water content associated with the coordination sphere of CuII in the 2D-CP than in the 3D-CP.
3.2.1.5
Sacrificial Agents
Sacrificial agents, or electron/hole scavengers, play a prominent role in photocatalytic applications, as they help to direct the desired half reaction of WS. Although their use is often relegated as mere reaction assistants, their active and different role in facilitating the studied reactions has been highlighted using general catalysts [40–42], as well as MOFs. The active role and specificity of the sacrificial agent (i.e., TEOA, lactic acid, and MeOH) was strongly evidenced when using the Cd(TD) MOF (H2TD, 1-thia-3,4-diazole-2,5-dithiol) in H2 production (TOFeqs ¼ 1.89103, 1.9104, and 6105 s1, respectively). Sacrificial agents however should be gradually replaced by other species, as they do not allow the WS but just the half reactions and are meant for study purposes, while they are usually associated with higher economic and environmental costs. To avoid their use, a study of H2 photoproduction using PCN-777 [Zr6O4(OH)10(H2O)6(TATB)2] (H3TATB, 4,40 ,400 -(1,3,5-triazine-2,4,6-triyl)tribenzoic acid) took advantage of their role in the reaction mechanism to obtain a double harvest as both the targeted gas (H2) and a commercially interesting molecule (N-benzylbenzaldimine) [43]. Despite that this work still used Pt-NPs as cocatalyst, the catalytic reaction enables the synthesis of value-added organic chemicals. Considering OER, a clear effect of different sacrificial agents was demonstrated for MIL-100(Fe) ([Fe3O(OH) (H2O)2(BTC)3]; H3BTC, trimesic acid) [44]. In this study, OER activity was remarkably not correlated with the redox potentials of the sacrificial agents. Indeed, AgNO3 was by far the most active oxidant (resulting in TOFeq ¼ 1.10105 s1, Ered ¼ 0.771 V vs. normal hydrogen electrode – NHE) even if much stronger oxidants were tested (Na2S2O8, Ered ¼ 1.96 V vs. NHE; note the production of just oxygen traces), demonstrating how more holistic studies should be undertaken in this regard.
3.2.2
Parameters Affecting Electrocatalysis
In MOF electrocatalysts, electric potential measurements generally use different reference electrodes and reaction conditions, limiting their comparison. In the last few years, the tendency relied on the use of the reversible hydrogen electrode (RHE), which eliminates the pH effect in the potential measurement, as a first approximation. Furthermore, electrochemical reactions are governed not only by the Nernst equation but also by kinetic limitations, which produce overpotentials (η), expressed by the Butler-Volmer equation, and by its simplified version at high η, the Tafel equation [45]. Overpotential is a very important indicator of the reaction barriers, being involved in the intrinsic activity of the catalyst. The activity is then dependent
68
G. Armani-Calligaris et al.
on the reaction kinetics which are modified by the pH (e.g., adsorption, involved species, acid-base reactions). Thus, η is not constant under pH changes, even referring it to RHE. Despite that many studies do not consider mechanism variations with the pH, here we will use η as a first comparison between pHs, as other factors cannot be linearly added. Electrochemical reactions are usually determined by voltammetry, showing the current density at a set potential. On the other side, literature often relies on the indication of the “current onset potential” (i.e., the potential at which one reaction shows the first current increase). This definition is however incomplete as current growth is exponential, leading to misunderstandings. Rarely, onset potential has been better referred to a low fixed current density (usually from 0.1 to 2 mAcm2). Fortunately, this value has been gradually substituted by the potential to reach 10 mAcm2, based on estimations of the conversion efficiency of a photoelectrode for HER and OER [46]. Consequently, for an appropriate comparison, we have homogenized all the reported potentials (Eq. 3.7) and overpotentials (Eq. 3.8) following the equations: ERHE ¼ EREF þ ΔEREF þ 0:05916 pH
ð3:7Þ
η ¼ jE RHE EðRHEÞj
ð3:8Þ
where ERHE is the recorded potential referred to RHE, EREF is the recorded potential vs. the reference electrode, ΔEREF is the reference potential referred to the standard hydrogen electrode (SHE), and E(RHE) is the thermodynamic potential of the reaction versus RHE (equal to 0.0 V for HER and + 1.229 V for ORR and OER). Accordingly, in our discussion, we have reported the “current onset overpotential” (ηon) and the overpotential to reach 10 mAcm2 (η10) or the indicated “x” current density (ηx). In the case of ORR, the half-wave overpotential (η1/2) has also been calculated. In this account, temperature has been excluded as most of the reports do not state it, so all the potential values, despite the obvious possible error, have here been standardized at 25 C. Other error sources could derive from those works where Hg|HgO|OH (MOE) or Ag|AgCl|Cl (AgE) were used as reference electrodes or when solution pH was not declared, thus assuming pH from concentrations. For example, AgE is commercially available as saturated KCl (ΔEREF ¼ +0.197 V [47]), 3.0 M, or even at 1 M (ΔEREF +0.222 V [48]). Wherever not stated, it has been assumed here that the electrode was a saturated KCl one, giving a possible error of up to 25 mV in the calculations. A doubtful situation we faced in the survey is that some works used 0.5 M H2SO4 as electrolyte, while others used 1 M H2SO4 or 1 M HCl. In the case of 0.5 M H2SO4, the resulting pH would be about 0.29, as pKa2 for H2SO4 is 1.99 [49]. As properties are popularly measured at integer pH values, it is possible that H2SO4 was assumed to completely dissociate to give pH ¼ 0. If pKa2 of H2SO4 was not taken into account, it would bring to a possible error of about 17 mV in the reported ERHE. For other acids and bases, complete dissociation was assumed as the calculated error on pH was less than 0.1 points. Moreover, we want to point out that η is not the only important parameter in electrocatalysis. Thus,
3 Green Energy Generation Using Metal-Organic Frameworks
69
Fig. 3.3 Representation of a CP deposited on the current collector. Zoom on the main contributions to the goodness of an electrode catalyst. The numbers reflect the paragraphs division in Sect. 3.2.2
we have also reported the Tafel slope since it sheds light on how good an electrode is considering the whole system (e.g., catalyst, contact, reagent concentration). Finally, the faradaic efficiency (YF) has also been reported as the number of electrons (ne) involved following Eq. 3.9: YF ¼
ne 2 100 2
ð3:9Þ
An ideal electrode should (i) allow reducing to zero the energy losses (associated with reagent diffusion, activation energies, low selectivity, etc.); (ii) allow accelerating the target reactions, keeping them indefinitely stable over time; and (iii) be efficiently and economically fabricated. In order to near this utopia, the electrode needs (i) a high surface area with exposed catalytic sites (electrode reactions are interphasic processes); (ii) a high catalytic activity; (iii) to be electrically connected to the current collector with a fast charge transfer; and (iv) to be performant even during reagent scarcity. The design of the perfect electrode is difficult, being the research of CPs electrocatalysts mainly focused on specific zones of the electrode (Fig. 3.3). In a similar way than photocatalysis, we have here discussed some of the main parameters influencing the electrocatalytic activity of MOFs (e.g., active sites, conductivity, material-collector interface).
3.2.2.1
Catalytic Activity of the Reaction Site
In the electrocatalytic field, research has been mainly focused on the substitution of precious metals (Pt for HER and Pt, IrO2, and RuO2 for OER/ORR) with redox activity, more abundant and cheaper ones. The most common metals used when studying MOFs as electrocatalysts are (i) Mo and Zn (structured in POMs) and Co and Cu for HER (see Table 3.2), and (ii) mainly Co (with both carboxylate and
70
G. Armani-Calligaris et al.
nitrogenated ligands), followed by mixed Fe-Ni, pure Fe and Ni, and, finally, mixed Fe-Co, for oxygen reactions (see Tables 3.3 and 3.4). Some works put the accent on the cluster structure, like a [M2(OOC)4] paddle wheel [50–52] or [M3O (OH)x(H2O)y(OCC)6] [53–55], and their selection based on their robustness without considering their mechanistic behaviors. The availability of active sites is of paramount importance in order to enhance the catalyst activity. In MOFs, this is accomplished by either designing metal clusters with dangling bonds or by creating defective structures. An interesting approach produced defects on ZIF-67 ([Co (MIm)2]; MIm, 2-methylimidazole) using a dielectric barrier discharge nitrogen plasma etching [56]. This material presented improved performances (lower Tafel slope, lower η10) for OER in basic medium once etched, thanks to the ligand defects provoked by the plasma treatment. Also, the mixed-metal strategy, with an already demonstrated synergetic effect in the inorganic catalyst field, has been extensively applied in the use of MOFs as electrocatalysts (see Table 3.2). In this case, the interactions between the multiple metals and the ligand environment strongly modify the catalytic center. However, up to now the research was mainly focused in an optimization activity by mixing more or less randomly the most acknowledged metals. Apart from this parameter, the influence of the organic ligand on the MOF/CP electroactivity has also been poorly explored. Indeed, despite the large variety of polycomplexant ligands, most works used carboxylate (BDC, BTC, other benzene derivatives) or nitrogenated aromatic ligands (imidazoles, triazoles, bipyridines). An interesting in-depth study on the reaction mechanism and the proton transfer kinetics was done using a pH-degraded NiFe-BDC [57]. This research not only showed a pH-independent η in the small studied pH range (13–14), but also demonstrated the effect of the outer coordination sphere of the catalyst (uncoordinated ligands), helping the proton relaying far from the electrode surface.
3.2.2.2
Intrinsic Conductivity of the Material
In general, MOFs are known to be electrical insulators or semiconductors. This property leads to an intrinsic difficulty in using them directly as electrodes. The classic internal resistance dependences (specific resistivity, ρ, and conductor length, l) still apply to the charge transfer inside CPs. Focusing on the specific resistivity reduction, this can be pursued by controlled pyrolysis, which degrades the insulating ligands and can form structures that are much more conductive than the original polymer. The drawback of this strategy is the partial collapsing of the pores, reducing the access to internal catalytic sites. Having no more the original CP as the active material, this strategy is beyond the scope of this survey. On the other side, π-conjugation is the main mechanism involved in the conductivity of graphite and conductive polymers [58]. Therefore, the use of ligands that favor this conduction mechanism could reduce the influence of internal charge transfer on the RDS of the system. Few works have focused on this strategy [59– 61]. For example, the use of hexaminobenzene (H6HIB) and Fe and Mn resulted in three 2D-CPs [M3(HIB)2] with multiwalled hollow sphere morphology [60]. These
3 Green Energy Generation Using Metal-Organic Frameworks
71
spheres showed astonishing electrical conductivities of about 108 and 149 Scm1 for the Mn and Fe 2D-CP, respectively, and an amazing 359 Scm1 for the mixedmetal one, which was then reflected on the OER/ORR activity of these materials. The other strategy to improve the charge transfer is to reduce the depth of the insulating catalyst to lower its overall resistance. Moreover, this strategy preserves the pore structure and enhances the accessibility of internal active sites. This is the case of MOFs thin films, nanosheets, and monolayers. For example, a Ni-MOF grown from Ni foam through a mediating layer of Fe carbonate hydroxide showed much smaller particle size than the unmediated grown one [62], leading to better performances, thanks to the enhanced charge transfer. In this sense, reducing the particle size of the catalyst is particularly important for nonporous CPs, in order to increase the accessibility of their active centers [63].
3.2.2.3
Electrical Contact to the Current Collector/CP-Collector Interface
Being the catalyst either conductive or not, it will be connected somehow to a current collector. The type of collector, the type of interactions between CPs and the collector, and the adhesion amount between them have demonstrated to hugely influence the overall performance of the electrode [64]. The current collector has changed over time, from just an inert support to more developed materials able to enhance the final performances. Thus, glassy carbon electrode (GCE) has been partially substituted by metals or by other conductors (e.g., fluorine-doped tin oxide (FTO), carbon paper, carbon cloth, carbon paste). Although these supports contribute to enhance the overall performance, they limit and mislead the full assessment of the catalyst behavior itself, unless a good comparison with the bare support and subtraction of its contribution is done. In this sense, an accurate evaluation was carried out in some recent works, by depositing different CPs on GCE or by growing them on metal foams [57, 65], confirming thus the beneficial effect of the current collector. As MOFs and CPs are mostly insulators, the electrical contact of the entire particle with the current collector is often an issue, reducing the general performances. To compensate this limitation, CPs have been either mixed in carbon paste electrodes (CPE, used mainly in HER) or deposited on large surface area electrodes, like metal foams (Ni, Fe, Cu). The latter strategy is useful in depositing large amounts of catalyst in reduced spaces, but the determination of the electrode area is difficult, making inaccurate the calculation of the current density. Moreover, the recovery of the used catalyst for its subsequent characterization is complicated. The use of carbon paste electrodes or other conducting additives (i.e., acetylene black, carbon black, activated carbon, graphite) can improve the electric contact through the entire particle surface. However, it dilutes the catalyst mass and could limit the reagent diffusion (in CPE, paraffins are often used as hydrophobic glues), leading to possible concentration overpotentials. Its use can be beneficial in case one wants to limit the specific surface area effect of the catalyst [66].
72
G. Armani-Calligaris et al.
Catalyst-collector adhesion is also critical in the evaluation of electrochemical properties. Low adhesion leads on the one side to catalyst detachment, giving underestimations of electrode lifetime (usually compensated with the use of polymeric glues like Nafion™, at the expenses of some catalyst activity), and on the other side to just partial electrical contact, giving lower active catalyst and thus errors in calculations. Drop casting of the microcrystalline catalyst suspension was initially applied, being later substituted by other techniques, like electrophoretic deposition [67, 68] or in situ growth on the collector surface. The first enhances the homogeneity of the dispersion on the collector, by minimizing the overlapping of particles caused by the surface tension of drop casting, while the second can be either electrochemical [69, 70] or solvothermal [65, 71]. MOFs thin films of controlled thickness and morphology can be thus grown, with durable adhesion and electrical connection. Overall, one should take into account two main factors: the specific catalyst behavior and the overall electrode performance. In this sense, the electrochemical performances of the catalyst itself must be considered independently from the charge transfer to the current collector, and, obviously, the general resistance of the electrode should be reduced to achieve a good activity of the prepared device. Thus, an extensive study using a mixed-metal MIL-53(Fe/Ni) [FexNi1-x(OH)(BDC)] electrophoretically deposited on a carbon cloth [68] and its pyrolyzed derived carbons succeeded to distinguish the different contributions to the catalyst activity. The improved overall performance of pristine MIL-53(Fe/Ni) was explained by its higher availability of exposed sites rather than their specific activity, which was better in pyrolyzed samples (probably because of the easier charge transfer along the generated carbon). This work confirms that many different contributions affect the electrocatalyst behavior and all of them should be taken into account in order to highlight the specific behaviors of the tested structures.
3.3
Photocatalysis
Sun is our most important energy source and daily we received 1120 Wm2 solar irradiance in the crust of Earth [72]. Photocatalytic processes to produce fuels and chemicals are an attractive way to store solar energy [73]. TiO2 is the most used catalyst for WS [74]. However, the main problem is that it absorbs only UV light, reducing sunlight exploitation to less than 7% [75]. The development of photochemical systems that use visible sunlight to drive chemical reactions is then of great practical significance. Among others, MOFs have been proposed as photocatalysts, thanks to the combination of their large surface areas with the potential ability of harnessing sunlight. Moreover, their adjustable structure allows to modulate the light absorption to use visible light.
2 (98%) (97%) 2 (99%) 2 (98%) 2 (98%) 2 (98%) 2
0.52
0.45 0.13 0.6 0.75 0.80 0.259 0.255 0.274 0.300 0.180 0.304 0.298 0.349
96 (0.036) 137 (0.015) 125 (0.017) 176 (0.37)
0.237 (10) 0.392 (10) 0.388 (10) 0.425 (10)
a
0.85 (1.01) 0.85 (1.19) 0.85 (0.75) 0.85 (0.52)
[113] [106]
H2SO4 (0.5 M, pH ¼ 0.16); CPE GCE (POMOF: CB ¼ 1:1 wt), Pt foil, Ag|AgCl H2SO4 (0.5 M, pH ¼ 0.29 calc); GCE (0.14 mgcm2), C rod, RHE
a
a
(continued)
[104]
[112]
[111]
[110]
Refs. [105]
HCl (pH ¼ 1), LiCl (1 M); CPE, Pt, SCE
PBS (0.05 M, pH ¼ 6.8); GCE (20 mgcm2), Pt foil, SCE
PBS (0.2 M, pH ¼ 6.8); GCE (dc, 20 mgcm2), Pt foil, SCE
Experimental conditions; cell design – W, C, R HCl (pH ¼ 1), LiCl 1 M; W, large Pt mesh, SCE, W: CPE (cat, 3.1108 molcm2) HCl (pH ¼ 1), LiCl 1 M; W, large Pt mesh, SCE, W: CPE HCl (pH ¼ 1), LiCl 1 M; W, large Pt mesh, SCE, W: CPE H2SO4 (0.5 M, pH ¼ 0.29 calc); GCE (dc), Pt foil, Ag|AgCl
a
a
0.46
a
Cu3(Mo8O26) (H2O)2(OH)2(L1)4 Ag4(Mo8O26) (L2)2.5(H2O) Co4(L3)2(4bpy)(H2O)6 Co2(L3)(azene)(H2O)3 Co2(L4)4 Cu2(L4)4 Zn2(L4)4 ε2(BPDC)3 ε(bim)2(isop) ε(bim)8/3(BDC)2/3 ε(bim)2(BTC)2/3 NENU-500 NENU-501 CTGU-5 CTGU-6
a
a
a
a
ε(BTC)
a
Duration, h (faradaic efficiency) 5 (95%)
a
ηon, V 0.323
ηappl, V ( j @ ηappl, mAcm2) 0.503
ε2(BTC)2
Material ε(BTC)4/3
Tafel slope, mVdec1 ( j0, mAcm2) 142 (0.0024)
Table 3.2 Overview of the key information of pure MOFs and CPs tested for HER
3 Green Energy Generation Using Metal-Organic Frameworks 73
0.45
0.101
0.348 0.362 0.336 0.319 0.244 0.350 0.155
Cu3(pdc)2(phen)2(H2O)2
Co2(H-pycz)4
Cd(H2TBpyz) Pb2(TBpyz) La(H2TBpyz) Sr3(HTBpyz)2 Ce(HTBpyz) Mn2(TBpyz) MIL-53(Fe)
98 145
400 253 485 362 273 360 71.6 (0.00025)
121 (0.162)
62
Tafel slope, mVdec1 ( j0, mAcm2)
0.417 (10) 0.535 (10)
0.423 (10) 0.449 (10) 0.545 (10) 0.487 (10) 0.457 (10) 0.452 (10)
0.223 (10)
0.552 (10)
ηappl, V ( j @ ηappl, mAcm2) 0.365 (10) 0.522 (10) 0.398 (10) 0.563 (10) 0.281 (10) 0.283 (10) 0.590 (10)
12
a
a
72
8
Duration, h (faradaic efficiency) 3 (52%, 1 h) 3 (57%, 1 h) 3 (54%, 1 h) 3 (71%, 1 h) 3 (49%, 1 h) 3 (52%, 1 h) 12
NaOH (0.1 or 1 M, pH ¼ 13–14 calc); GCE (1.416 mgcm2), Pt, Ag|AgCl (NaCl 3 M) KOH (1 M, pH ¼ 14 calc); GCE (0.0376 mgcm2), graphite rod, SCE
[69]
H2SO4 (0.5 M, pH ¼ 0.29 calc); C paper (ED, 0.7 mgcm2), GCE, Ag|AgCl (KCl sat) H2SO4 (1 M, pH ¼ 0 calc); CPE (cat/graphite/ paraffin ¼ 1:6.9:2.8 wt, 9.2%wt MOF), Pt, SCE H2SO4 (0.5 M, pH ¼ 0.29 calc); GCE (0.14 mgcm2), C rod, SCE KOH (1 M, pH ¼ 14 calc); GCE (RD, 1200 rpm), Pt, Ag|AgCl
[108]
[24]
[109]
[107]
[66]
Refs. [114]
Experimental conditions; cell design – W, C, R H2SO4 (0.1 M, pH ¼ 1 stated, 0.96 calc); CPE GCE (cat:AC ¼ 1:2 wt; RD, 2000 rpm), Pt, Ag| AgCl (KCl sat)
W, working electrode; C, counter electrode; R, reference electrode; CPE, carbon paste electrode; GCE, glassy carbon electrode; dc, drop casting; PBS, phosphate buffer solution; SCE, saturated calomel electrode; CB, carbon black; RD, rotating disk electrode; L1, 4H-4amino-1,2,4-triazole; L2, 3,5-dimethyl-4-amino-4H1,2,4-triazole; H2L3, 4,5-di(40 -carboxylphenyl)phthalic acid; azene, (E)-1,2-di(pyridin-4-yl)diazene; HL4, 4-(5-(pyridin-4-yl)-4H1,2,4-triazol-3-yl)benzoic acid; H3pdc, pyrazole-3,5-dicarboxylic acid; phen, 1,10-phenanthroline; H-pycz, 3-(pyrid-40 -yl)-5-(400 -carbonylphenyl)-1,2,4-triazolyl; HBA, N-(2hydroxybenzyl)-alanine a Datum not reported
Cu2(HBA)2(H2O) Cu(HBA)
ηon, V 0.056 0.196 0.156 0.196 0.006 0.016 0.400
Material Co-ε(BTB)4/3 ε(BTB)4/3 Co-ε(BTC)(bipy)2 Ru-ε2(BTC)2 Ru-ε2(BPDC)2 PPh4-ε2(BTC)2 HKUST-1
Table 3.2 (continued) 74 G. Armani-Calligaris et al.
a
a
209
NiCo-UMOFNs
Fe-TCA Al-TCA
2200; 0.544
a
1070; 10–15 Å
a
a
a
a
a
a
a
a
NU-1000 thin film
MIL-100(Fe)
Co2(L4)43H2O Cu2(L4)43H2O Zn2(L4)43H2O Co4(L3)2(4bpy)(H2O)6 Co2(L3)(azene)(H2O)3 Fe/Co-MOF
Material Co2(L6)(adip)2 Co2(L6)2(5-bdc)2(H2O)2 Co-ZIF-9
SBET, m2g1; Vp, cm3g1; pore size (Å)
0.42 0.35
0.19
0.17
0.52
0.47 0.79 0.93 0.437 0.237 0.487
ηon, V 0.47 0.58 0.371
87 140
42
193
Tafel slope, mVdec1
0.458 (10) Low
0.250 (10)
0.62
0.82 (0.48) 0.82 (1.50) 0.82 (2.70) 0.64 (2.97)
a
0.82 (1.47)
ηappl, V ( j @ ηappl, mAcm2) 0.82 (1.65) 0.82 (1.07) 0.74 (1)
Table 3.3 Overview of the key information of pure MOFs and CPs tested for OER
(99.3%)
(100%)
2 (98%) 2 (98%) 2 2 (99%) 2 (98%)
Duration, h (faradaic efficiency) 2 (96%) 2 (95%) 3
[126]
[63]
[67]
[117]
[125]
[112]
[113]
[122]
Refs. [124]
(continued)
KOH (0.1 M, pH ¼ 13); C paper (hydrophobic, 0.2 μg + 0.05 μg SP), Pt, Ag|AgCl (KCl sat) KCl (0.1 M), NaOH added (pH ¼ 11.3); Pt (0.038 mgcm2), a, Ag|AgCl (3 M NaCl) NaHCO3 (0.05 M), NaOH 0.1 M (pH ¼ 11); FTO (thin film growing, 0.03 mgcm2, 1 cm2), Pt, Ag|AgCl (KCl sat) 200 h, KOH (1 M, pH ¼ 14); GCE (0.2 mgcm2), Pt foil, Ag|AgCl KOH (1 M, pH ¼ 14); GCE (cat: AB ¼ 9:1 wt), graphite, Hg|HgO
2 h, PBS (0.2 M, pH ¼ 6.8); GCE (20 mgcm2), Pt foil, SCE
PBS (0.1 M, pH ¼ 7.0); FTO (1.7 μmolcm2), graphite, Ag|AgCl PBS (0.05 M, pH ¼ 6.8); GCE (20 mgcm2), Pt foil, SCE
Experimental conditions; cell design – W, C, R PBS (pH ¼ 6.8), degassed; GCE (20 mgcm2), Pt foil, SCE
3 Green Energy Generation Using Metal-Organic Frameworks 75
a
a
Co1.5(tib)(dcpna)6H2O
ZIF-5(CoCo)
ZIF-9(Co)
3.9 5.1 Å2
a
a
0.28
0.27
89
82
68.3
60
0.28
1514 (SLangmuir); 0.55
MAF-X27-OH
63
106.2
a
100
a
Tafel slope, mVdec1 47
180; 0.103; 11.3 8.3 Å2
a
0.170
a
a
a
ηon, V
SBET, m2g1; Vp, cm3g1; pore size (Å)
Pb2TBP
Ni-BTC@NF
Fe-BTC@NF
Fe/Ni-BTC STS
Material Fe/Ni(1/12)-BTC@NF
Table 3.3 (continued)
0.36 (10)
0.67 (10)
0.363 (10)
0.39 (10)
0.47 (10)
0.330 (10)
0.350 (10)
a
ηappl, V ( j @ ηappl, mAcm2) 0.270 (10)
(50%)
2
24
a
a
a
Duration, h (faradaic efficiency) 15 (95%) Experimental conditions; cell design – W, C, R KOH (0.1 M, pH ¼ 13); W, Pt wire, Ag|AgCl (KCl sat), W ¼ Ni foam (1 2 cm2) KOH (0.1 M, pH ¼ 13); W, Pt wire, Ag|AgCl (KCl sat), W ¼ GCE (0.5 mgcm2) KOH (0.1 M, pH ¼ 13); W, Pt wire, Ag|AgCl (KCl sat), W ¼ Ni foam (1 2 cm2) KOH (0.1 M, pH ¼ 13); W, Pt wire, Ag|AgCl (KCl sat), W ¼ Ni foam (1 2 cm2) KOH (1 M, pH ¼ 14), O2 sat (also 0.1 M, pH ¼ 13) and PBS (0.2 M, pH ¼ 7.4); GCE (0.199 mgcm2), Pt foil, SCE KOH (1 M, pH ¼ 14), O2 sat (also PBS 1 M, pH ¼ 7); GCE (0.2 mgcm2; RD, 1200 rpm), C rod, SCE KOH (1 M, pH ¼ 14), O2 sat; GCE (0.127 mgcm2; RD, 1600 rpm), Pt wire, Ag|AgCl NaOH (1 M, pH ¼ 14); FTO (0.393 mgcm2), Pt, SCE KOH (1 M, pH ¼ 14), O2 sat; GCE (0.70 mgcm2), Pt foil, SCE
[130]
[129]
[128]
[121]
[127]
Refs. [70]
76 G. Armani-Calligaris et al.
1720
1539; 0.639
ZIF-67(Co)-unsat
MCF-37 MCF-49 Ni-BTC
Cu(L7)(4bpy)(ClO4) Cu(L7)(phen)(ClO4)(H2O) Ni3(OH)2(BDC)2(H2O)4
NNU-21 NNU-22 NNU-23 NNU-24 Ni-BDC Ni-Fe (3,5:1)-BDC
MIL-53(Fe/Ni/Mn0.4)-NF
MIL-53(Fe/Ni2.4/Co0.4)
1820
ZIF-67(Co)
a
a
a
a
1000; 13 Å 990; 13 Å 950; 13 Å 990; 13 Å
a
a
a
a
924; 0.33; 14.6 Å
UTSA-16
185–53.7
0.27–a
122 243.5 123
122.7 77.2 81.8 121.8 139 82
71.3
53.5
112 43 64
317–74.9
0.35–a
0.403 0.233
77
0.38
0.32 (100)
0.555 (10) 0.376 (10) 0.365 (10) 0.522 (10) 0.370 (10) 0.265 (10)
0.238 (100)
0.219 (10)
0.346 (10)
0.551 (10)– 0.401 (10) 0.411 (10)– 0.311 (10)
0.421 (10)
20
(99.1%)
a
15 15 15 15
60
a
96
a
24 (95.9%)
a a
–
7 (95%)
KOH (1 M, pH ¼ 14); Ni foam (2 3 cm2, 4.5 mgcm2), graphite plate, Hg|HgO
KOH (1 M, pH ¼ 14), O2 sat; GCE (0.2 mgcm2; RD, 1600 rmp), Pt mesh, Hg|HgO KOH (1 M, pH ¼ 14), O2 sat; GCE (0.05 mg), a, SCE
[136]
[135]
[134]
[53]
[133]
[132]
[50]
[56]
[131]
(continued)
KOH (1 M, pH ¼ 14); C paper (0.48 mgcm2), graphite rod, SCE KOH (1 M, pH ¼ 14), O2 sat; W, Pt, Ag|AgCl (KCl sat), W ¼ GCE (cat: Cpowder ¼ 1:1 wt, 0.032 mgcm2) KOH (1 M, pH ¼ 14), O2 sat; W, Pt, Ag|AgCl (KCl sat), W ¼ Ni foam (0.5x2 cm2) KOH (0.1 M, pH ¼ 13); C cloth (1x1 cm2, cat:AC ¼ 1:1 wt, 1 mgcm2), Pt, Ag|AgCl
KOH (0.1 M, pH ¼ 13); GCE (0.22 mgcm2), Pt wire, SCE
KOH (1 M, pH ¼ 14), O2 sat; GCE (0.35 mgcm2; RD, 1600 rpm), Pt foil, Ag|AgCl (3 M KCl) K2B4O7 (KBi) (0.5 M, pH ¼ 9.2) – KOH (1 M, pH ¼ 14), O2 sat; GCE (0.097 mgcm2), Pt mesh 1 1 cm2, SCE
3 Green Energy Generation Using Metal-Organic Frameworks 77
2298; 1.63; 8 Å
Mn/Fe-HIB
a
a
a
200 ~10 14 Å2
a
a
SBET, m2g1; Vp, cm3g1; pore size (Å) 231 302 430 378 426 396 593
MIL-88-A-Fe MIL-88-A-Co MIL-88-A-FeCo1.6 MIL-88-B-FeCo1.6 MIL-88-A2.7B-FeCo1.6 BIF-91
BIF-89
Material NiNi-PYZ FeFe-PYZ CoCo-PYZ NiFe-PYZ NiCo-PYZ CoFe-PYZ NiPc-MOF
Table 3.3 (continued)
0.191
0.34
0.37
0.27
ηon, V
45
186 76 53 78 39 117
165
Tafel slope, mVdec1 161 62 59 53 70 44 74
0.281 (10)
0.541 (10) 0.378 (10) 0.330 (10) 0.323 (10) 0.288 (10)
0.392 (10)
ηappl, V ( j @ ηappl, mAcm2) 0.399 (10) 0.693 (10) 0.353 (10) 0.560 (10) 0.362 (10) 0.300 (10)
100 (98.4%)
10.5
a
a
a
a
12
50 (94%)
Duration, h (faradaic efficiency)
KOH (1 M, pH ¼ 14); GCE (0.23 mgcm2), Ag|AgCl (4 M KCl) KOH (0.1 M, pH ¼ 13), O2 sat; GCE (0.15 mgcm2, RD), Pt, Ag|AgCl (KCl sat)
KOH (1 M, pH ¼ 14); FTO (0.0076 mgcm2), Pt wire, Ag|AgCl (3 M KCl) KOH (1 M, pH ¼ 14); Ni foam (1 1 cm2, 0.19 mgcm2), Pt wire, Ag|AgCl (4 M KCl) KOH (1 M, pH ¼ 14), O2 sat; GCE (0.35 mgcm2), Pt mesh, Ag|AgCl (KCl sat)
Experimental conditions; cell design – W, C, R KOH (0.1 M, pH ¼ 13); GCE (0.21 mgcm2), graphite plate, Ag| AgCl
[60]
[139]
[54]
[138]
[59]
Refs. [137]
78 G. Armani-Calligaris et al.
512; 0.238; 18.6 Å
145; 20 Å
Fe/Ni-MOF
a
a
a
a
a
a
a
a
a
17;
a
a
a
MOF-74(Ni)
Co/NiMOF@Fe
Co/NiMOF
CoMOF@Fe
CoMOF
NiMOF@Fe
NiMOF
Ni/Fe-MOF/FeCH-NF Ni-MOF-NF Ni/Fe-CP/NF, MIL-53(Fe/Ni)
MIL-101(Fe/Ni) MIL-101(Fe) MIL-100(Fe) Co2(OH)2(BDC)
27.6
84
54.3
110.5
65.5
132.6
70.3
161.9
51.3 78.4 29
44.2 41.5 41.4 63
0.190 (10)
0.45 (50)
0.264 (10)
0.377 (10)
0.289 (10)
0.406 (10)
0.327 (10)
0.433 (10)
0.200 (10) 0.270 (10) 0.188 (10)
0.237 (20) 0.253 (20) 0.251 (20) 0.318 (100)
40
16
67
67
a
a
a
a
17 (98.4%)
50
80
a
a
[71]
[140]
[65]
[57]
[62]
[120]
[119]
(continued)
KOH (1 M, pH ¼ 14); Ni foam, Pt, Hg| HgO KOH (0.1 M, pH ¼ 13), O2 sat; W, Pt, Ag|AgCl, W ¼ GCE (0.35 mgcm2) KOH (0.1 M, pH ¼ 13), O2 sat; W, Pt, Ag|AgCl, W ¼ Fe foam (1 3 cm2) KOH (0.1 M, pH ¼ 13), O2 sat; W, Pt, Ag|AgCl, W ¼ GCE (0.35 mgcm2) KOH (0.1 M, pH ¼ 13), O2 sat; W, Pt, Ag|AgCl, W ¼ Fe foam (1 3 cm2) KOH (0.1 M, pH ¼ 13), O2 sat; W, Pt, Ag|AgCl, W ¼ GCE (0.35 mgcm2) KOH (0.1 M, pH ¼ 13), O2 sat; W, Pt, Ag|AgCl, W ¼ Fe foam (1 3 cm2) KOH (1 M, pH ¼ 14); RDE (cat/carbon ¼ 1:1 wt), Pt wire, Ag| AgCl (KCl sat) KOH (1 M, pH ¼ 14); Ni foam (1x3 cm2) STh, Pt rod, Hg|HgO
KOH (1 M, pH ¼ 14); Ni foam (2 mgcm2), Pt, Ag|AgCl KOH (1 M, pH ¼ 14); Ni foam (1 2 cm2), Pt mesh, Hg|HgO
KOH (1 M, pH ¼ 14); Ni foam (1 1 cm2, 3 mgcm2), Pt, Hg|HgO (20% KOH)
3 Green Energy Generation Using Metal-Organic Frameworks 79
MAF-6-Co
MAF-69-W
MAF-69-W
MAF-69-Mo
Material CTGU-10a1 CTGU-10b1 CTGU-10c1 CTGU-10d1 CTGU-10a2 CTGU-10b2 CTGU-10c2 CTGU-10d2 PCN-Fe2Co-Fe2Ni (1:1) PCN-250-Fe2Co PCN-250-Fe2Ni MAF-69-Mo
Table 3.3 (continued)
a
a
a
a
a
1200; 0.65 1250; 0.72 1185; 0.61
a
a
a
a
a
a
a
a
SBET, m2g1; Vp, cm3g1; pore size (Å)
0.435
a
0.366
a
0.271
ηon, V
215
375
178
219
Tafel slope, mVdec1 102 95 87 140 92 81 58 127 67.7 80.6 71.6 144
0.559 (2)
a
0.482 (2)
0.210 (1)
ηappl, V ( j @ ηappl, mAcm2) 0.36 0.34 0.33 (10) 0.44 (10) 0.36 (10) 0.33 (10) 0.18 (10) 0.46 (10) 0.318 (10) 0.307 (10) 0.271 (10) 0.388 (2)
a
a
a
a
15 (98%)
a
a
50 50 50 50 16 (90%)
a
a
a
a
Duration, h (faradaic efficiency)
GCE (0.209 mgcm2; RD, 1600 rpm), graphite rod, Ag|AgCl (KCl sat), PBS (0.1 M, pH ¼ 7), O2 sat GCE (0.209 mg cm2; RD, 1600 rpm), graphite rod, Ag|AgCl (KCl sat), NaHCO3 (0.5 M, pH ¼ 7.2), CO2 sat GCE (0.209 mg cm2; RD, 1600 rpm), graphite rod, Ag|AgCl (KCl sat), PBS (0.1 M, pH ¼ 7), O2 sat GCE (0.209 mg cm2; RD, 1600 rpm), graphite rod, Ag|AgCl (KCl sat), NaHCO3 (0.5 M, pH ¼ 7.2), CO2 sat GCE (0.209 mg cm2; RD, 1600 rpm), graphite rod, Ag|AgCl (KCl sat), PBS (0.1 M, pH ¼ 7), O2 sat
KOH (1 M, pH ¼ 14); C cloth (1 mgcm2, cat:AB ¼1:1 wt), C rod, Ag|AgCl
Experimental conditions; cell design – W, C, R KOH (0.1 M, pH ¼ 13); GCE (0.141 mgcm2), Crod, Ag|AgCl
[123]
[141]
Refs. [55]
80 G. Armani-Calligaris et al.
a
a
a
a
a
a
a
a
83; 0.262 51; 0.286
a
a
a
a
a
a
a
a
a
a
a
0.45
0.25
273 237 214 147 82 41.1 44.5
173 101 65
105 125
167 68
51
71 60 49 87.5 75.7
0.621 (10) 0.541 (10) 0.448 (10) 0.344 (10) 0.296 (10) 0.207 (10) 0.220 (10)
0.410 (10) 0.250 (20)
a
0.537 (5) 0.687 (5)
0.600 (200) 0.210 (200)
0.280 (10)
0.296 (10) 0.277 (10) 0.236 (10) 0.364 (10) 0.309 (10)
a
5 72
a
a
a
a
20
a
(87%)
200 (99.5%)
a
30 (96.4%, 1 h) 13
a
15
a
a
KOH (1 M, pH ¼ 14), O2 sat; Ni foam (1 2.5 cm2), carbon rod, Hg|HgO
KOH (1 M, pH ¼ 14); C cloth (EPhD, 2 mgcm2), graphite rod, Hg|HgO (NaOH 1 M) KOH (1 M, pH ¼ 14); GCE (0.28 mgcm2), carbon rod, Hg|HgO
KOH (1 M, pH ¼ 14); GCE (cat: AB ¼ 1:2 wt; 0.25 mgcm2; RD, 1600 rpm), Pt wire, Ag|AgCl KOH (0.1 M, pH ¼ 13); Ni foam (2 1 cm2), graphite rod, Ag|AgCl (KCl sat) PBS (Na+) (0.1 M, pH ¼ 7.0); graphite (0.357 mgcm2; RD, 4000 rpm), Pt rod, Ag|AgCl (KCl sat) KOH 1 M (pH ¼ 14); FTO (0.2 mgcm2), a, Hg|HgO
KOH (1 M, pH ¼ 14); C paper (1 mgcm2), Pt mesh, Ag|AgCl (KCl 1 M)
KOH (1 M, pH ¼ 14), O2 sat; Ni foam (0.48 mgcm2), graphite plate, Ag| AgCl (KCl sat)
[146]
[145]
[68]
[144]
[52]
[102]
[51]
[143]
[142]
W, working electrode; C, counter electrode; R, reference electrode; PBS, phosphate buffer solution; FTO, fluorine-doped tin oxide glass; GCE, glassy carbon electrode; RD, rotating disk electrode; SCE, saturated calomel electrode; cat, catalyst; sat, saturated; wt, weight ratio; AB, acetylene black; SP, Super P,Timcal; AC, activated carbon; CNTs, carbon nanotubes; EPhD, electrophoretic deposition; STh, solvothermally deposited; L6, 1,4-bis(3-pyridylaminomethyl)benzene; H2adip, adipic acid; H25-bdc, 5-nitroisophthalic acid; tib, 1,3,5-tris(1-imidazolyl)benzene; H3dcpna, 5-(30 ,50 -dicarboxylphenyl)nicotinic acid; HL7, 5-sulphidyl1-methyltetrazole acetic acid; phen, 1,10-phenanthroline a Datum not reported
(Ni1Zn3)3(OH)2(BDC)2(H2O)4 (Ni1Zn2)3(OH)2(BDC)2(H2O)4 (Ni1Zn1)3(OH)2(BDC)2(H2O)4 (Ni2Zn1)3(OH)2(BDC)2(H2O)4 (Ni3Zn1)3(OH)2(BDC)2(H2O)4 FeII-MOF-74 NAs FeIII-MOF-74 NAs
MIL-88B-(Fe)-NH2 MIL-88B-(Fe0.5-Co0.5)-NH2 MIL-53(Fe/Ni)
Co2-MOF Co3(BTC)2(H2O)3
Ni-Fe-NDC Ni-Fe-NDC 4.3% strains
Co-MOF
HKUST-1(Ni) HKUST-1(Ni-Fe 10:1) HKUST-1(Ni-Fe 3:1) Co-MOF Co-MON
3 Green Energy Generation Using Metal-Organic Frameworks 81
PCN-223
55; 13 Å
Co-CAT
a
80; 13 Å
Ni-CAT
ZIF-8
710; 0.34 740; 0.34
a
a
1150
HKUST-1 ST HKUST-1 EC
HKUST-1 Cu-bipy-BTC
Al-Co-PMOF
a
1600
Material PCN-223-FeIII
PCN-222
SBET (m2g1) Vp (cm3g1) Pore size (Å) 1600
0.51
0.78 0.88
0.48
ηon, V (Tafel slope, mVdec1) 0.40 vs. NHE 0.55 vs. NHE
1.1 (3.60)
1.194 (0.45)
1.284 (0.36)
0.79 (8.1) 0.77 (29.6)
1.035 (current peak)
ηdrop, V ( jcathodic @ ηdrop, mAcm2)
Table 3.4 Overview of the key information of pure MOFs and CPs tested for ORR
0.94
0.854 (5.02) (8)
0.84
a
η1/2, V ( jlimit, mAcm2)
90 days in device
6
Duration, h (current retention) 6
H2SO4 (0.5 M, pH ¼ 0.29), O2 sat; GCE, Pt wire, Ag|AgCl (KCl 3 M) H2SO4 (0.1 M, pH ¼ 0.96), O2 sat; GCE (2.4107 molcm2, cat/carbon ¼ 60:40 wt, RD), GCE plate, RHE PBS (pH ¼ 6), O2 sat, GCE (dc; RD, 400 rpm), Pt wire, Ag|AgCl (KCl sat) PBS (0.1 M, pH ¼ 6), O2 sat; GCE (0.21 mgcm2), Pt rod, Ag|AgCl (KCl 3 M) PBS (0.02 M, pH ¼ 7), NaClO4 (0.10 M), O2 sat, GCE (0.200 mgcm2; cat/carbon ¼ 30:70 wt; RD, 1600 rpm), Ag|AgCl (KCl sat) PBS cellular solution (50 mM, pH neutral), O2 sat, steel plate (0.5 mgcm2), Pt foil, Ag|AgCl
Experimental conditions; cell design – W, C, R DMF (distilled), LiClO4 (0.1 M), AcH (0.3 M), O2 sat; FTO (STh, 2 cm2), Pt mesh, Ag|AgCl (KCl sat)
[159]
[103]
[154]
[155]
[157]
[100]
Refs. [158]
82 G. Armani-Calligaris et al.
a
Co-MOF
Fe-HIB 0.38
a
a
a
a
2298; 1.63; 8Å
Mn/Fe-HIB
Mn-HIB
0.25 (36)
248
0.53
0.409 (128)
0.39
0.289
MOF-5(Zn)
a
630 (Rouquerol method); 20 Å
Ni3(HITP)2
ZIF-67
1070; 10–15 Å
a
Fe/Co-MOF
(Cu4Cl)3(H0.5BTT)8(H2O)12
0.72 (current peak)
0.71 (current peak)
0.56 (1.19)
0.346 (6.37) 0.49 (4.08) 0.41 (5.39)
0.451
26 (90% up 25 h, then 72%)
a
a
100
4
8 (88%)
12 (84%)
[51]
[60]
[102]
[160]
[61]
[124]
[150]
(continued)
KOH (0.1 M, pH ¼ 13), O2 sat, GCE (0.25 mgcm2; cat: AB ¼ 2:1; RD, 1600 rpm), Pt wire, Ag|AgCl
KOH (0.05 M, pH ¼ 12.7), O2 sat, GCE (0.032 mgcm2; cat/carbon ¼ 1:1 wt; RD, 800 rpm), C felt, Ag|AgCl KOH (0.1 M, pH ¼ 13), hydrophobic C paper (cat +SP ¼ 0.2μg + 0.05μg), Pt, Ag| AgCl (KCl sat) KOH (0.1 M, pH ¼ 13.0), O2 sat; GCE (120 nm thin film; 0.254 mgcm2; RD, 2000 rpm), Pt mesh, Hg|HgO (KOH 1.00 M) KOH (0.1 M, pH ¼ 13.0), O2 sat; GCE (0.20 mgcm2; RD, 1600 rpm), Pt wire, Ag|AgCl (KCl sat) KOH (0.1 M, pH ¼ 13), O2 sat, GCE (0.016 mgcm2), Pt wire, Ag|AgCl (KCl sat) KOH (0.1 M, pH ¼ 13), GCE (dc, 0.15 mgcm2; RD, 1600 rpm), Pt, Ag|AgCl (KCl sat)
3 Green Energy Generation Using Metal-Organic Frameworks 83
412 15 Å
a
a
a
SBET (m2g1) Vp (cm3g1) Pore size (Å)
a
a
a
a
(61)
0.329
ηon, V (Tafel slope, mVdec1) 0.43 (220) 0.31 (70) ηdrop, V ( jcathodic @ ηdrop, mAcm2)
0.40 (5.3) 0.56 (3.3) 0.54 (4.3) 0.48 (4.0) 0.53 (2.7)
0.469
0.40 (5.2)
a
η1/2, V ( jlimit, mAcm2)
24
200 (97%)
a
Duration, h (current retention) Experimental conditions; cell design – W, C, R KOH (0.1 M, pH ¼ 13), O2 sat, GCE (0.25 mgcm2, RD), graphite rod, Ag|AgCl (KCl sat) KOH (0.1 M, pH ¼ 13), O2 sat, GCE (0.25 mgcm2, on AC), glassy carbon rod, RHE KOH (0.1 M, pH ¼ 13), O2 sat, GCE (cat:CNTs ¼ 2:1 wt, 0.48 mgcm2, STh; RD, 1600 rpm), Pt wire, Ag|AgCl
[151]
[161]
Refs. [101]
W, working electrode; C, counter electrode; R, reference electrode; FTO, fluorine-doped tin oxide glass; GCE, glassy carbon electrode; STh, solvothermally deposited; dc, drop casting; RD, rotating disk electrode; cat, catalyst; sat, saturated; wt, weight ratio; SP, Super P, Timcal; AB, acetylene black; AC, activated carbon; CNTs, carbon nanotubes; AcH, acetic acid; PBS, phosphate buffer solution; H3BTT, 1,3,5-benzenetris(1H-tetrazole); H6HITP, 2,3,6,7,10,11hexaaminotriphenylene, H6HIB hexaaminobenzene a Datum not reported
PcCu-O8-Co PcCu-O8-Ni PcCu-O8-Cu PcCu-O8-Fe Pc-O8-Co
Ag-BTC
Material NiFe-NDC NiFe-NDC 4.3% strains
Table 3.4 (continued)
84 G. Armani-Calligaris et al.
3 Green Energy Generation Using Metal-Organic Frameworks
3.3.1
85
Hydrogen Evolution Reaction
In 2009, the first example of a MOF for the reduction of water into H2 was reported [76]. This Ru-based 2D porous [Ru2(BDC)2]n structure (Brunauer-Emmett-Teller surface area (SBET) ~ 800 m2g1, pore size ~ 4.8 Å) showed a high H2 productivity (1178μmolg1h1) as a result of the whole interaction chain between the [Ru (2bpy)3]2+ photosensitizer (2bpy, 2,20 -bipyridine), methylviologen as redox relay, and the Ru24+ paddlewheel catalytic centers. A similar concept was later used in 2018, when two new 3D MOFs Ru-TBP [Ru2(TBP-Ru-DMF)(H2O)2]X+ and RuTBP-Zn [Ru2(TBP-Zn)(H2O)2]X+ (H4TBP, 5,10,15,20-tetra( p-benzoic acid)porphyrin; DMF, N,N-dimethylformamide; SBET ¼ 441 and 422 m2g1, respectively) were synthesized [77]. It was suggested that the integration of photosensitizing porphyrin ligands and Ru2 SBUs in the MOFs could facilitate a LMCT to drive HER. Proton reduction was tested in acetonitrile (MeCN)/water, using TEOA as reductant and under visible light, reaching catalytic TONs after 4 days (21.2 and 39.4 for Ru-TBP and Ru-TBP-Zn, respectively). In a further study, a series of new 3D oxalate-bridged (ox) Ru-based anionic CPs with formula [MIRuIII(ox)3]2 (MI: Na or Li) was synthesized, using photosensitizer templates [ZII(2bpy)3]2+ (ZII: Zn, Cu, Ru, or Os) as counterions [78]. Among the tested photosensitizers, [Os(2bpy)32+], with the lowest bandgap of 1.68 eV, evolved the highest amount of H2, reaching however just a TON of 0.123 after 8 h of reaction (see Table 3.1 for further details). To be noted that the large honeycombed channels were fully occupied by the counterion, so incorporating the photosensitizer as a template counterion could be a good strategy, but larger pore structures are required. Although Ti-MOFs are depicted as ideal platforms for photocatalysis, few structures are available, limiting the number of reports using them for HER. Indeed, just MIL-125 (SBET ~ 1300 m2g1, pore volume (Vp) ¼ 0.65 cm3g1, pore size ¼ 5–7 Å [79]) and its aminated version MIL-125-NH2 (SBET ~ 1050 m2g1, Vp ~ 0.49 cm3g1) were evaluated as photocatalysts in HER. Their photocatalytic properties were assessed under visible light (λ > 420 nm) with TEOA as reductant and in presence or not of Pt NPs as cocatalyst [22]. While pure MIL-125 was inactive to HER, pure MIL-125-NH2 showed some activity, reaching a calculated TOFeq of 9.57106 s1. Further studies on MIL-125-NH2 evidenced that TOFeq depends on the exposed MOF planes (i.e., (110), (100), (111), and (001)), ranging from 1.15106 s1 to 3.49106 s1 for the (001) and the (110) plane, respectively [26]. The different TOFeq values found in the two studies, probably related with experimental conditions, evidence the complex comparison of photocatalytic activity, even using the same material. More extensive work has been performed in HER using Cu- and Cd-CP photocatalysts. The effect of the excitation lifetime on the activity of two Cu-CPs based on H2DSPTP ligand has already been analyzed in Sect. 3.2.1. In a second study using H2DSPTP, two isoreticular MOFs were synthesized, one based on CuII [Cu(DSPTP)(H2O)2] and the other based on CdII [Cd(DSPTP)(H2O)2] [80]. Although CdII showed lower Eg (1.39 vs. 2.36; see Table 3.1) and higher τ
86
G. Armani-Calligaris et al.
(4460 vs. 190 ps) compared to CuII, the latter demonstrated a TOFeq of 3.9 times the CdII one, with Φ of 0.076% at 900 nm. This behavior was rationalized calculating the position of the conduction bands, which seemed to be localized on the metal, giving 0.96 eV for CdII and 1.49 eV for CuII. Thus, CuII would have a much higher reduction potential than CdII, enhancing the reaction rate. It should be noted that both MOFs absorb light up to 1200 nm and show a monochromatic HER even at 900 nm, reaching improved activity than under a more energetic light (700–800 nm). Although this behavior was justified by a metal-to-ligand charge transfer (MLCT), it remains unclear how the MLCT could provide a H+ reduction, considering that the catalytic center and the conduction band would be positioned at the metal. The same year, another series of Cu-CPs [Cu2X2(4bpy)2] (4bpy, 4,40 -bipyridine; X, halides from Cl to I) revealed interesting photocatalytic properties [81]. These materials form almost planar rhombi of Cu2X2 (where Cu is in tetrahedral field [82]) connected by 4bpy, resulting in a honeycomb 2D structure that is interwoven perpendicularly in the 16 26 Å2 windows, with packages of three sheets in both directions. When tested under UV-visible light irradiation, these CPs revealed a trend of higher production and stability increasing halogens’ size. Furthermore, the degradation of [Cu2Cl2(4bpy)2] and [Cu2Br2(4bpy)2] after 18 h of reaction was confirmed by high-resolution transmission electron microscopy (HR-TEM), showing the presence of possible Cu-NPs. Unfortunately, the quantitative stability of these materials and their textural properties and the possible relation between Cu-X and Cu-Cu distances were not reported for a deeper discussion. It is remarkable that [Cu2I2(4bpy)2] bandgap was slightly higher than that of the two other CPs (2.00 eV vs. 1.90 and 1.85 eV for I, Br, and Cl, respectively), in any case leading to the highest TOFeq. This could be explained by a lower degradation rate or, as previously mentioned [80], by the shift of the LUMO level to more reducing values (moving toward iodide), while the highest unoccupied molecular orbital (HOMO) would stay. The authors proposed a direct reduction of CuI to Cu0 by triethylamine (TEA), supported by the formation of Cu NPs as degradation product. Very recently, two Cd-based MOFs, [Cd(TD)] and [Cd(TD)(H2O)], were synthesized [83]. The H2TD ligand was designed to favor H-bond interactions with the sacrificial agent TEOA, favoring the e transfer to neutralize the ligand. [Cd(TD)] reached the highest productivity found for a CP (26100 μmolg1h1 under optimized conditions), corresponding to a quite high TOFeq of 1.89103 s1 and a calculated Φ of 1.38% at 365 nm (although the absorption maximum is at 280 nm). Remarkably, such high H2 productivity was obtained at pH ¼ 9, when acidic conditions would be preferred to reduce protons. Despite a quite high Eg, a strong electronic connection seems to be established between the metal and the ligand (not existing a clear LMCT), suggesting that RDS is not produced by the charge separation lifetime. After three testing cycles (3 h each), the activity was reduced to ca. 60%. An in-depth physicochemical evaluation of the solid before and after cycling would be required to rationalize the occurring reactions. Moreover, a potential direct interaction between TEOA-Cd could be considered, as well as a study of H-bonding of TEOA to the ligand in a coordinated state (exclusively done considering the free state). On the other hand, [Cd(TD)(H2O)] material showed just a
3 Green Energy Generation Using Metal-Organic Frameworks
87
slightly higher bandgap than [Cd(TD)] but still a much lower activity, which was explained by the weaker interaction between TD and CdII (penta- vs. hexacoordinated, respectively). Finally, a 2D-POM-CP was reported in 2014 based on CuII, octamolybdates [βMo8O26], and pyridazine (pyz), where pyz coordinates to two Cu centers forming dimers that are further connected to oxygen of octamolybdates, giving rise to [Cu (pyz)]4[β-Mo8O26] (SBET ¼ 1.5 m2g1) [84]. It is worth pointing out that compared to other reports, an indirect lower bandgap (1.8 eV) was obtained and TOFeq reached 4.93103 s1 (based on the measured exposed POMs) or 1.93105 s1 (based on Mo from the gross formula). Furthermore, the estimated TON was about 380 (based on the exposed POMs) or 1.66 (based on Mo from the gross formula) after 3-h testing in two cycles. Although the POM plays an active role in the catalysis, this was unclear for the Cu centers.
3.3.2
Oxygen Evolution Reaction
Even if the term “artificial photosynthesis” is commonly attributed to CO2 photoreduction, natural photosynthesis incorporates CO2 just in “dark” reactions, while in chloroplasts the photoredox reaction converts water into O2 [88]. This sophisticated system consists of a five-ring chain that spatially separates the photo-generated charges. In contrast, in usual MOFs geometry, there are just two of these chain rings, the metal cluster and the antenna. As the photo-excited antenna usually donates an e to the metal cluster, the depleted ligand is more susceptible of degradation if not quickly reduced back to its neutral state. On the other hand, when using metal centers that allow d-d* transitions, a MLCT with OER from metalcoordinated water becomes possible. OER has been extensively pursued using both heterogeneous [89] and homogeneous [90] catalysis, mainly based on Ru and Ir [91]. However, these scarce transition metals suffer of high and volatile prices. So, cheaper solutions are foreseen using more abundant d- and f-group metals (Table 3.5). A series of Fe-MOFs, based on the [Fe3O(OH)(H2O)2(OOC)6]6 cluster, were studied in 2016 as photocatalysts for OER under visible light [44]. These materials included the flexible microporous MIL-88B(Fe) [Fe3O(OH)(H2O)2(BDC)3] (pore size ¼ 8 Å), MIL-100(Fe) [Fe3O(OH)(H2O)2(BTC)3] (SBET ~ 1400 m2g1, pore size ¼ 5.5 and 8.6 Å), MIL-101(Fe) [Fe3O(OH)(H2O)2(BDC)3] (SBET ~ 2320 m2g1, pore size ¼ 12 and 16 Å), and MIL-126(Fe) [Fe3O(OH)(H2O)2(BPDC)3] (H2BPDC, 1,10 -biphenyl-4,40 -dicarboxylic acid; SBET ~ 1460 m2g1; pore size ¼ 10.2 Å). The flexible microporous MIL-53(Fe) [Fe(OH)(BDC)] (pore size ¼ 8 Å), based on a 1D metal oxide cluster, was also evaluated. On the whole, MIL-101(Fe) was the most active material for OER, reaching a TON of 1.07 after 9 h. However, their stabilities upon reaction conditions were assessed by N2 sorption, concluding that MIL-101 (Fe) was the most affected solid, with a 66% reduction of its initial SBET. This was also related to the formation of Ag0 NPs from the used AgNO3 sacrificial agent. In
H2O, AgNO3 (0.012 M), UV-vis light, glass vessel, 5 C
a
H2O, Xe lamp 1000 W (1 Wcm2), λ > 200 nm or λ > 420 nm
H2O, AgNO3 (0.1 M), Xe lamp 500 W (λ > 420 nm), 122 Wcm2 for λ > 420 nm, of which 48.2 Wcm2 for 420 < λ < 660 nm
a
1.8
H2O, AgNO3 (0.006 M), Xe lamp 300 W, glass wall, 5 C
Experimental conditions H2O, pH ¼ 7, SO32 (103 M), W lamp 200 W, λ ¼ 400–800 nm, 50 C Water, AgNO3 (2 gL1, 0.012 M), Xe lamp 300 W, λ > 420 nm filter, 5 C H2O, AgNO3 (0.1 M), Xe lamp 500 W (600 mWcm2), λ > 420 nm filter
a
a
a
Bandgap (eV) 2.58
a
Htz, triazole; H2mna, 2-mercaptonicotinic acid Datum not reported
MIL-88 MIL-88B-4F MIL-88B4CH3 MIL-88B2CH3 MIL-88B-2OH MIL-88B-NH2 MIL-88B-OH MIL-88B-NO2 MIL-88B-Br MIL-53(Al) MIL-53(Al)NH2 MIL-101(Fe) MIL-126(Fe) MIL-100(Fe) MIL-53(Fe) MIL-88B(Fe) [Bi(BTC) (DMF)] [Cu(py)]4[βMo8O26]
Material Cu3PO4(tz)2OH Bi-mna
Table 3.5 List of pure CPs with photoactivity against OER
90
0.79
1.072 0.325 0.016 0.035 0.370
5.49106
1.10015 3.35106 1.27107 3.84107 3.80106 1.34104
1.53106 3.94106 1.98106 1.80105 1.56105 4.31105 6.25105 20 55 27.5 225 175 745 1008 163 38 2.1 16 56 771
3.01106
40
TOFeq (eqs1) 6.43 105 2.09105 1.01105 6.56105 2.28106
TON 0.69 0.38
150 750 27.5
Production rate (μmolg1h1) 1583.9 177
[84]
[96]
[44]
[94]
[98]
Refs. [97] [95]
88 G. Armani-Calligaris et al.
3 Green Energy Generation Using Metal-Organic Frameworks
89
addition, although the different photocatalytic behavior was not fully elucidated, suggestions were given based on the pore sizes of the structures, having MIL-101 (Fe) the largest ones. To note isoreticular MIL-88B(Fe) and MIL-126(Fe) showed very similar activity, suggesting that the interweavement of the larger structure of MIL-126(Fe) would reduce and compensate its activity with respect to MIL-88B (Fe), which possesses smaller but more accessible pores. In 2019, an extensive work investigated the e-donating and e-withdrawing effect of ligand substituents on the flexible MIL-88B(Fe)-X (X: H, Br, NO2, NH2, OH, 4F, 2CH3, 4CH3, 2OH) for OER [34]. The deactivating substituents (X: Br, NO2, 4F) had the double effect of changing the band position and of stabilizing the structure against OH, thus preventing ligand degradation. On the other side, ringactivating substituents (X: 2CH3, 4CH3, OH, 2OH, NH2) were more prone to degradation and led to the lowest O2 production. This is in line with a MLCT, using d electrons available in FeIII. Indeed, the 4F-functionalized ligand showed by far a much higher activity in O2 generation (see Table 3.5) compared to all the other samples, being the only one reaching a catalytic production in the tested time (TON ¼ 1.42 in 2 h). Remarkably, one can compare the activity of MIL-88B(Fe) in these two latter studies of 2016 and 2019, respectively (see Table 3.5). Although the testing conditions were almost the same (note a different test duration 9 vs. 2 h – and the absence of details for the reactor setup and particle sizes), the 2016 study reported a less active MIL-88B(Fe) than 2019 one (TOFeq of 3.80106 s1 vs. 10.1106 s1). The difference could be tentatively assigned to the MIL-88B(Fe) synthetic conditions: the lower active MOF was synthesized from a more diluted solution, which could be associated with a different crystal size [92]. This fact evidences the difficulty to compare the activity of MOFs since many details substantially affect the final activity. Finally, another study using MIL-53(Fe) and its aminated version showed the reverse effect in the presence of [Ru(bpy)3]2+ as photosensitizer, so that the aminated MOF was more active for OER [93]. However, the photosensitizer could act itself as the OER catalyst. Concerning Al-MOFs, OER activity of MIL-53(Al) [Al(OH)(BDC)] (SBET ¼ 747 m2g1), as well as its aminated derivative [Al(OH)(BDC-NH2)] (SBET ¼ 847 m2g1), was also investigated [94]. Both structures exhibited comparable kinetics in O2 production to that of MIL-88B(Fe)-4F under similar conditions. MIL-53(Al)-NH2 generated more O2 than MIL-53(Al), as it was found for the Fe counterparts. Using a density of state analysis (DoS), authors highlighted that the Al-hydroxo chains do not participate in the photocatalysis, despite that Al would behave as an e-tunnel toward Ag+. In 2015, the photocatalytic OER was firstly assessed using Bi-MOFs, in particular using the newly synthesized [Bi(mna)]2NMe2H (H2mna, 2-mercaptonicotinic acid; SBET ¼ 35 m2g1) [95]. Despite its low porosity, this MOF was highly active for OER, probably related with its long photoluminescence lifetime of 1.1 ns. A second Bi-MOF was later synthesized using BTC as linker, leading to [Bi(BTC)(DMF)] DMF2MeOH [96], in which Bi dimers are coordinated to six carboxylate groups and one DMF molecule. This solid demonstrated an effective photocatalytic OER
90
G. Armani-Calligaris et al.
under UV-vis light, reaching the highest TOFeq among the reviewed panorama (1.34104 s1). The proposed mechanism involved a ligand-to-ligand charge transfer (LLCT). However, under the studied conditions, this Bi-MOF quickly degrades as a possible consequence of the solvent exchange in water solution. Furthermore, although potentially porous, its structure also collapses upon solvent removal. Among Cu-MOFs, the same [Cu(pyz)]4[β-Mo8O26] 2D-CP [84] previously described for HER was also tested for OER under visible light. Its OER effectivity was lower with respect to HER (considering HER was tested in the presence of a sacrificial donor), giving a final TON of about 215, again calculated referring to the [Mo8O26]4 available to the reagents, while the average TON did not reach the unit. The available octamolybdate TOF was calculated to be 102 s1, while TOFeq was 5.49106 s1 (based here on single Mo ions). The authors rationalized the charge transfer as deriving from CuI2(pyz)2 to [Mo8O26]4, Cu centers being able to catalyze the OER, even if this last was not explicitly indicated (TOFeq ¼ 1.10105 s1 using CuI as the reference metal). Overall, from the reported data, it is not easy to identify the trends that affect OER. Although not fully clear from other reported works [93, 94], the comparative studies using Fe-MOFs suggest a potential energy-level tunability of the ligand through simple functionalization [33]. Also, the effect of the inorganic clusters seems important (discrete units or infinite rods), as well as pores accessibility. To this respect, the poor textural information of some MOFs with flexible porosity (partially or fully open in water) hinders the extraction of solid conclusions. Finally, among the tested metals (Fe, Al, Bi, and Cu), Bi-MOFs have demonstrated to be highly active (co-rationalizing their activity with the donating-withdrawing behavior of their ligands). Although not reported, Mn- and Co-MOFs, based on relatively abundant metals and a priori active in this reaction [91], could be promising candidates for photocatalytic OER.
3.4
Electrocatalysis
Electrochemistry has the enormous advantage of spatially splitting reactions between redox species and allowing a precise control of their rates. Moreover, unlike photocatalysis, sustainable electrocatalysis allows the use of far generators and to be independent on local Sun variability in the generation of endergonic molecules. In the reverse reaction, like in FCs and batteries, electrocatalysis is fundamental to allow the fast generation of electrical power. An electrode can catalyze the desired reaction, but also a series of unwanted ones, like those that lead to catalyst deactivation or to by-products formation. In the case of O2 reduction, this can be converted to either H2O by an overall 4e mechanism or to H2O2 by a 2e mechanism, according to [99]: Acidic conditions
O2 þ 4Hþ þ 4e ! 2H2 O O2 þ 2Hþ þ 2e ! 2H2 O2
E ¼ 1.229 VSHE E ¼ 0.695 VSHE
3 Green Energy Generation Using Metal-Organic Frameworks
H2 O2 þ 2Hþ þ 2e ! 2H2 O Alkaline conditions O2 þ 2H2 O þ 4e ! 4OH O2 þ H2 O þ 2e ! HO2 þ OH HO2 þ H2 O þ 2e ! 3OH
91
E ¼ 1.763 VSHE E ¼ 0.401 VSHE E ¼ 0.065 VSHE E ¼ 0.867 VSHE
Maximizing the faradaic efficiency toward the desired reaction is then of paramount importance, playing the catalyst the central role in giving the appropriate selectivity among the whole η range considered (in general, selectivity changes with η [100–103]), in order to reduce the electric cost and avoid the degradation of the material caused by reactive species. Also, a very interesting concept to evaluate an electrocatalyst is the reversibility of the reaction involved [60]. By that, analyzing both the anodic and the cathodic waves of a reaction would help to understand its mechanism, for example, in situations where there is a negative overpotential (like in the case of some POMs-based MOFs – POMOFs [104, 105]) or where there are adsorbed products.
3.4.1
Hydrogen Evolution Reaction
When fed by renewable electricity, electrolysis of water is a clean method for H2 production. Current MOF-based strategies for HER use redox metals (i.e., Cu and Co) or POMs, these lasts surpassing the performances of other studied compounds. Although HER experiments are mostly performed in strongly acidic conditions, basic media are also employed.
3.4.1.1
Acidic Medium
One notable work dealing with CPs for HER described a series of POMOFs using the ε-Keggin POM structure [ε-PMoV8MoVI4O40Zn4] (named ε) [104], leading to a 3D [ε2(BPDC)3] with two interwoven networks, a 2D [ε(bim)2(BTC)2/3], and three 1D-CPs [ε(bim)2(isop)], [ε(bim)8/3(BDC)2/3], and [ε(bim)2(BTC)2/3] (Hbim, benzimidazole; H2isop, isophthalic acid). All these CPs showed a strong anodic shift of the potential (a negative η ranging from 0.255 to 0.300 V), giving results that are apparently better than thermodynamics. This behavior was justified by a local higher H+ “confinement” effect, which would shift η to more positive potentials, joined to an exchange reaction between H+ and Li+, used as the electrolyte. Particularly for [ε(bim)2(BTC)2/3], a pH-dependent study was performed, revealing an expected negative shift of the potentials with increasing pH, but no quantitative values were reported. In 2018, the same group synthesized another series of POMOFs based on the same ε-Keggin POM and the BTC ligand [105]. In particular, (TBA)3[PMoV8MoVI4O36(OH)4Zn4][BTC]4/36H2O (TBA, tetrabutylammonium) showed a remarkable activity at pH ¼ 1 in CPE and in the presence of LiCl (1 M), with an η of 0.323 V. This behavior was justified by both the presence of Li+, which might intercalate into the pores, and the confinement effect that might increase
92
G. Armani-Calligaris et al.
the local concentration of H+ near the POMs, thus changing the equilibrium reaction toward the gas generation. Regarding the Cu-based materials, the 2D-CP [Cu3(pdc)2(phen)2(H2O)2] (H3pdc, pyrazole-3,5-dicarboxylic acid; phen, 1,10-phenanthroline), consisting of trinuclear mixed-ligand units, was hydrothermally prepared and further evaluated for HER [66]. A CPE based on 9.2% wt of this solid gave an η10 of 0.552 V. After the irreversible HER, scanning back to more positive potentials, it appeared a large anodic peak, rationalized as Cu-H oxidation, not appearing when stopping the cyclic voltammetry at +0.1 V vs. RHE. This was supported by FTIR, with an additional IR band at 715 cm1, corresponding to O-H out-of-plane bending absorption. This hydrogen seems to migrate by spillover effect. The mechanism proposed is a double reduction for CuII, with a subsequent H+ adsorption-reduction step and a final H+ adsorption with further reduction to give H2. Among the Co-based materials, it was observed that the 2D CTGU-5 [Co (1,4-NDC)(bib)(H2O)] (H21,4-NDC, naphthalene-1,4-dicarboxylic acid; bib, 1,4-bis(imidazol)butane) had a higher catalytic ability for HER (with a 50 mV less reductive shift of the overpotential) than the 3D CTGU-6 [Co(1,4-NDC)(bib)]H2O and with also a lower Tafel slope [106]. On the other side, CTGU-6 presented a much higher exchange current density, and this relates with the performance of the catalyst. CTGU-5 was then blended with acetylene black to improve the electrical conductivity, giving better HER activity. Unfortunately, no studies of the electrical conductivities were reported. Recently, an eightfold interwoven network [Co2(Hpycz)4]H2O (H2pycz, 3-(pyrid-40 -yl)-5-(400 -carbonylphenyl)-1,2,4-triazolyl) was synthesized and tested on GCE, giving good values of overpotential and a low Tafel slope value [107]. Its ability to catalyze HER was assessed during 72 h, revealing almost no variation over this time. Regrettably, no structural analysis was performed after the reaction to assess its robustness. The eightfold interweavement nature could in principle help to increase the material stability, but it could also reduce the accessibility of reagents, thus reducing the number of active sites.
3.4.1.2
Alkaline Medium
In WS, the kinetic and energetic bottleneck is given by OER, so moving to more favoring OER conditions would increase the overall reaction yield, even if depleting the H+ concentration. Moreover, working under acidic conditions leads to the need of more robust and expensive containers, representing alkaline solutions a strategy for using cheaper materials in the cell design. Regarding the Cu-based compounds, two Cu-CPs built from N-(2hydroxybenzyl)-alanine (HBA) were synthesized by a surfactant effect of different metal cations (CoII and NiII) [108]. [Cu2(HBA)2(H2O)] was prepared from a Cu sheet in the presence of a NiII salt, and it could be converted to the anhydrous one [Cu(HBA)] by simple heating, while in the presence of a CoII salt, the anhydrous structure was directly formed. [Cu2(HBA)2(H2O)] showed lower overpotential and Tafel slope for HER, indicating better catalytic behavior than [Cu(HBA)]. The
3 Green Energy Generation Using Metal-Organic Frameworks
93
proposed mechanism involves multiple reactions, as its Tafel slope does not match (98 mVdec1) that of more classical mechanisms in alkaline solution. On the other side, [Cu(HBA)] showed a higher Tafel slope (121 mVdec1), suggesting that water dissociation was the rate-determining step. This work highlights the importance of unsaturated metal centers in electro- and photocatalysis. Using 2,3,5,6-tetra(4-carboxyphenyl)pyrazine (H4TBpyz) as ligand, a series of different nonporous CPs was synthesized changing the metal center (i.e., CdII, PbII, LaIII, SrII, CeIII, MnII) [109]. All exhibiting a η10 ranging from 423 to 545 mV, [Pb2(TBpyz)] and [Ce(HTBpyz)]H2O demonstrated to be stable at pH ¼ 14 for at least 4 h, exhibiting lower Tafel slopes than the others.
3.4.2
Oxygen Evolution Reaction
OER is very foreseen as the probably definitive renewable e source to produce H2, but also for a more general use in rechargeable batteries. The reason for the huge research to catalyze the total oxidation of the oxide ion to O2 is due to the required high energy and the difficult associated kinetics, with the possible generation of partially oxidized species, like hydrogen peroxide and superoxide. From bibliography, Fe, Co, Ni, and their mixed metal oxides are widely proposed as catalysts. In particular, a large number of Fe-Ni hydroxide structures is recognized for their high catalytic activity and stability at high pH [115]. Indeed, these systems are often described as NiII, being the active center, and FeII/III, increasing the conductivity of the system. Concerning CPs and MOFs, porosity was not an important criterion, ruling out the potential of high surface area MOFs. Moreover, some structures have been synthesized knowing their subsequent degradation to metal hydroxides once immersed in strong alkaline solutions, thus evidencing the focus on the dispersion of the metal sites more than the chemical environment modulation.
3.4.2.1
Alkaline Medium
Among Fe-MOFs, MIL-100(Fe) was deposited on a Pt electrode and evaluated as catalyst [116]. The suggested mechanism was chemical-electrochemical, related with the dissolution of FeIII as MIL-100(Fe) shows a low stability under neutral and basic conditions [117]. η was 0.52 V, even using a Pt collector, but no considerations were made about a possible passivation. In 2019, a mixed-metal MIL-101(Fe/Ni) was synthesized using different ratios of Ni/Fe and tested in KOH (1 M) on Ni foam [118]. All synthetized MIL-101(Fe/Ni) showed low Tafel slopes (ranging 40–45 mVdec1) and quite low η20 (230–250 mV). The best performing material had a Fe/Ni molar ratio of about 2, thus supposing an average Fe2NiO metal cluster. However, no stability considerations were reported, making possible that the given values refer to mixed metals oxides-hydroxides coming from the framework degradation at high pH. In this regard, the mixed-metal MIL-53
94
G. Armani-Calligaris et al.
(Fe/Ni), grown on Ni foam, degrades after OER in KOH (1 M) [57], even maintaining very good performances (Tafel slope of 29 mVdec1, η10 ¼ 0.188 mV). On the other side, this type of Fe cluster demonstrated to resist prolonged time in KOH 0.1 M (pH ¼ 13) using biphenyl-3,40 ,5-tricarboxylic acid (H3BPTC) as ligand [53], so it is unclear whether the instabilities of MIL-101(Fe/Ni) and MIL-53(Fe/Ni) come from the high pH and/or from OER. A series of mixed-metal structures using the same [M3(μ3-OH)(H2O)3(OOC-R)6] cluster and 5,50 ,500 -benzene-1,3,5-triyl-isophthalic acid (H6BHB) was synthesized by changing the ratio between NiII and CoII (0:3, 1:2, 2:1, 3:0) [55]. Tested on GCE, the best ratio was Co:Ni ¼ 1:2, reaching low Tafel slope and η10. By tuning the synthetic conditions, a desert rose crystal morphology was obtained, leading to a significant improvement on their electrochemical properties (i.e., lowering the Tafel slope and, for Co:Ni ¼ 1:2, also shifting η10 to 0.18 V, the lowest we found for GCE). After 50 h of reaction, the samples partially retained their crystalline structures, as confirmed by powder X-ray diffraction (PXRD). Very recently, a multiwalled sphere 2D-CP, [(Fe50%wtNi50%wt)3(HIB)2] (SBET ¼ 2298 m2g1, dpore ¼ 8 Å), showed a fivefold hollow sphere morphology, thus giving curved 2D surfaces [60]. Presenting pores perpendicular to the layers with very high surface area, species diffusion should be favored inside the spheres. At pH ¼ 13, it showed a remarkable low ηon of 191 mV and an η10 of just 280 mV, improving the performances when compared to its single-metal analogues and retaining 94.8% of its activity for at least 100 h. Concerning Co-based structures, a nonporous 3D-CP made from Co(OH) sheets connected by pillaring BDC was deposited on Ni foam and tested in OER [119], resulting in super-low η100 (318 mV). The catalyst was also evaluated on GCE, with poorer performances. Moreover, modified benzoic acids (4-aldehyde, 4-nitro) and ferrocenecarboxylic acid were used as “dopants” to form defective sites at different concentrations. In all cases, the catalytic activity was enhanced and, for ferrocenecarboxylic acid, η10 was just 178 mV. The presence of defective sites is rationalized as enhancing the hydroxide free energy of adsorption and the electrical conductivity of the Co(OH) layers. A further study reported a postsynthetic ion exchange of MAF-X27-Cl [Co2(μ-Cl)2(bbta)] (H2bbta, 1H,5H-benzo-(1,2-d:4,5-d0 ) bis-triazole; Langmuir surface area of 1514 m2g1; Fig. 3.4 [120]) possessing open metal sites on its porous surface. Under alkaline conditions, this yields to [Co2(μ-OH)2(bbta)] (MAF-X27-OH), functionalized by both open metal sites and hydroxide ligands, which drastically improved electrocatalytic activities for the OER. It is worth to note the high stability of this MOF at pH ¼ 14 (1 week) and the low Tafel slope of 60 mVdec1, even if η10 is on the average values for MOFs (390 mV). Moreover, the electrical conductivity decreased with the ion exchange (from 2.2107 to 2.2109), demonstrating that the performances have to be ascribed to the more active metal centers rather than a higher intrinsic conductivity. A mixed Co-Ni-BDC MOF ultrasonicated nanosheet [Co2Ni2(OH)3(BDC)2], formed by 2D metal oxide sheets connected by pillaring BDCs, was drop casted on GCE and then OER-tested at pH ¼ 14 [63]. It was theorized that the surface dangling bonds of the metals layer were responsible of the catalytic effect. Despite
3 Green Energy Generation Using Metal-Organic Frameworks
95
Fig. 3.4 Coordination environment for MAF-X27-Cl. Co, violet; C, gray; N, blue; O, red; Cl, green; H, white. (Adapted with permission from Ref. [120], Copyright © 2020, American Chemical Society)
the low micropore area (SBET ¼ 209 m2g1), this MOF-based ultrathin film presented a quite low ηon (250 mV, defined at j ¼ 0.1 mAcm2) and performed better than both the single-metal counterparts and the solvothermally synthesized bulk MOF. Remarkably, the current retention was as high as 97% and the activity was maintained for at least 200 h. Among the pure Ni-based compounds, Ji et al. reported the 2D-CP [Ni2(phth) (Ni)] (H8phth, 2,3,9,10,16,17,23,24-octaamino-phthalocyaninato; SBET ¼ 593 m2g1) supported on FTO glass [59]. This compound revealed a high conductivity (0.2 Scm1, activation energy of 0.11 meV), derived from the π-conjugation of its moiety. The square coordination mode of the ligand and the eclipsed stacking of the layers gave a series of square 1D channel. The correlation of η1 change with pHs lower than 7 revealed a possible OER mechanism of 1 e – 2 H+.
3.4.2.2
Neutral Medium
Lowering the pH of the reaction medium will necessarily decrease the reactant concentration, and this will increase the η needed for the reaction to flow. This has been evidenced in the bibliography, where much higher η and Tafel slopes were declared at neutral pH with respect to the structures tested in alkaline medium. Despite of this, as for the case of HER, working at neutral pH has the economic advantage of avoiding costly corrosion-resistant materials. There are very few works using neutral pH, focused just on CoII as the metal center. From them, we would like to highlight an interesting study of the electrocatalytic OER using ZIF-9 [Co(bim)2] on FTO glass, which showed how this type of structure can be effective [121]. Through a density-functional theory (DFT) study, it was suggested that OH coordinates to CoII by a change on the metal coordination geometry from tetrahedral to bipyramidal trigonal. Despite that the Tafel slope was not so low (193 mVdec1 at pH ¼ 7.0), in this work the optimum deposition amount of catalyst was assessed, and, more importantly, the variation of potential with pH was studied. Remarkably, η seemed to depend almost linearly on the pH, with a slope of about 34 mVpH 1 (at 1 mAcm2). Regrettably, no discussion about the involved mechanism was done. On the other side, the faradaic efficiency increased approximating the unity after 3 h, suggesting a gradual catalyst activation. In a different
96
G. Armani-Calligaris et al.
Fig. 3.5 Active-site structures and OER activities of MAF-69-Mo (or Co4Mo), MAF-69-W (or Co4W), and MAF-6-Co. Each model is extracted from the corresponding crystal structure, and ethyl groups and hydrogen atoms are omitted for clarity. (Reproduced with permission from Ref. [122])
study, a series of isostructural Co-ZIFs (ZIF-67 like) using 2-ethylimidazole (namely, MAF-69-Mo, MAF-69-W, and MAF-6-Co) was obtained, partially replacing the ligand by [MoO4]2 or [WO4]2 (in MAF-69-Mo or MAF-69-W, respectively) (Fig. 3.5) [122]. These ZIFs were tested with both phosphate buffer saline (pH ¼ 7.0, PBS) and CO2-saturated HCO3 buffer (pH ¼ 7.2) solutions. It has to be noted that the systems performed much better in PBS than in HCO3. The best performing material was MAF-69-Mo, and the relatively high activity was rationalized by DFT, suggesting a modulation of the electronic structure of CoII giving a high adsorption energy for water and a low adsorption energy for O2, which would lead to the reaction.
3.4.3
Oxygen Reduction Reaction
Applications of ORR in the energy field are mainly focused on metal-air batteries and FCs. This brings very different chemical environments where this reaction is required, from acidic (for proton exchange FCs) to strongly alkaline (for alkaline FCs and metal-air batteries), the last condition favored by higher reaction kinetics [146]. The catalyst must stay in a three-phase boundary or at least in air [147] to ensure good mass transport kinetics. Moreover, in the case of rechargeable metal-air batteries, dual catalysts for both ORR and OER are required. Again, despite its cost, the current commercial catalysts are based on Pt due to its high performances and selectivity over a wide range of conditions [148, 149]. For ORR, η has the same direction (negative) than for HER, but the significance is different as, in this case, it is a galvanic process instead of an electrolytic one. So, for ORR, η receives the meaning of voltage drop of a device with respect to the theoretical potential, and this leads to a lower power that can be extracted from the reaction. During this review process, the panorama found in ORR was similar to
3 Green Energy Generation Using Metal-Organic Frameworks
97
OER in terms of metals, as the first transition metals row was mostly implicated. However, for ORR, authors have selected porphyrin/phthalocyanine moieties, as well as 2D-CPs, enhancing the electrical conductivity, even though sometimes at the expenses of the faradaic efficiency. Performances are here quite enhanced in basic media, although further studies in acidic media could deepen the mechanisms involved and would also be beneficial in applications such as proton exchange FCs.
3.4.3.1
Alkaline Medium
Recently, a Cu-MOF [(Cu4Cl)3(H0.5BTT)8(H2O)12]based on a tetrazole version of trimesic acid (H3BTT, 5,50 ,500 -(1,3,5-phenylene)tris(1H-tetrazole)) and square planar Cu4Cl clusters was used in ORR [150]. As active site, each octahedral CuII presents a dangling bond with a coordinated water. In a physical mixture with carbon, this material showed a quite low η1/2 value and a very high faradaic efficiency, producing just 0.5–1.5% of peroxides as by-products (at η ¼ 0.28–0.83 V). In contrast with other Cu catalysts, this material proved a good tolerance to methanol in its activity. Despite its proven structural integrity after 12-h reaction, the current retention lowered to 84% of the initial value. In a further report using CuII-metallated phthalocyanine-derived tetracatecholates (PcCu), a series of 2D-CP structures was synthesized based on different cations (PcCu-O8-M, M: CoII, FeII, CuII, NiII) [151]. PcCu-O8-Co (SBET ¼ 412 m2g1, with square 1D channels) had the best performances among the series (Co > Fe ~ Cu > Ni) in ORR. High activities were described as deriving from the 2D π conjugation that would increase the electrical conductivity. Despite the potential intrinsic electrical conductivity, all the tests were performed using carbon nanotubes as electrical connector, and no conductivity values were given. In 2014, a new 2D-CP Ni hexaminosemiquinonate [Ni3(HITP)2] (H6HITP, 2,3,6,7,10,11-hexaminotriphenylene) was proposed as a possible graphene-type semiconducting material [152]. This slipped eclipsed structure revealed a very high conductivity for a CP, ranging from 2 to 40 Scm1 when used as pellet or as thin film on quartz, respectively. The compound was subsequently tested for ORR [61], revealing an important selectivity for the peroxide formation (63%–100%, depending on η) and a quite high Tafel slope. The peroxide selectivity was later rationalized by DFT [153], concluding that the calculated catalytic center was formed by the two iminic protons rather than the Ni cation, as favored by a higher adsorption energy. On the other side, a galvanostatic probe indicated a pH-dependent H+ kinetic order (0 order at pH > 12.8 and non-0 order for lower pHs), showing a proton-coupled electron transfer or proton-dependent chemical steps before or during the RDS. Finally, isoreticular to [Ni3(HITP)2], the (Mn-Fe)3(HIB)2 previously analyzed for OER was also tested for ORR. Again, its performances were notable and higher than the single-metal counterparts, reaching a η1/2 of 0.346 V and a current density at diffusional limit ( jlimit) of 6.37 mAcm2 (at 1600 rpm of the rotating disk electrode, RD). Moreover, this material showed a high reversibility of this sluggish reaction, as
98
G. Armani-Calligaris et al.
the ΔV for the peak current of the anodic and cathodic waves was of 0.627 V, a relevant parameter usually not taken into account.
3.4.3.2
Neutral Medium
Note here the few works regarding the activity of CPs or MOFs under neutral conditions since overpotentials are much larger than in basic media. Among them, one nice example is the electrochemically synthetized HKUST-1 (Hong Kong University of Science and Technology or [Cu3(BTC)2]), which was tested at buffered pH ¼ 6 [154]. The electrochemical synthesis led to a lower η compared to the solvothermal one. Regrettably, from parallel studies, it seems that HKUST-1 is not stable under these conditions, these results being potentially associated with the degradation products [155]. Another example deals with two 2D-CP catecholates, based on the 2,3,6,7,10,11-hexahydroxytriphenylene (HHTP) ligand and NiII or CoII as the metals, demonstrating very high voltage drops and just medium faradaic efficiencies [103]. Despite the possibility of having a good conductivity, this was not tested.
3.4.3.3
Acidic Medium
Again, scarce CPs or MOFs have been tested for ORR under acidic conditions, although this medium should thermodynamically favor the reaction. A porphyrincarboxylate-based MOF with catalytically inert Al2(OH)2 chains (named Al-CoPMOF, metallated from [Al2(OH)2(TBP-H2)] [156]) was tested at pH ¼ 1 and compared with the corresponding free TBP-Co ligand [157]. It was noted how, despite having a similar ηon, the faradaic efficiency was about 50% for the Al-CoPMOF, and 80% for Co-TBP, along all the potential range tested. This behavior was however not rationalized in terms of a possible low conductivity of the Al2(OH)2 chains, but on the base of the relative distances between Co centers, which are the catalytic active sites. Other porphyrin-derivative MOFs, PCN-223 [Zr6O4(OH)4(TBP-H2)3] and its FeIII metallated analogous, were evaluated in DMF solutions using a supporting electrolyte [158]. The effect of the addition of different proton sources (acetic acid and trichloroacetic acid) was studied, evidencing that the proton sources favored both the onset of the reaction and the faradaic efficiency. However, the stronger acid led to the formation of a deposit on the electrode associated with a significant dropping of the performances within an hour. This behavior is apparently in contrast with the generally accepted better performances of ORR in alkaline solutions, but in agreement with the thermodynamic behavior.
3 Green Energy Generation Using Metal-Organic Frameworks
3.5
99
Conclusions and Perspectives
Despite the many challenges faced in the development of new photo- and electrochemical catalysts, a great research effort is being dedicated in pursuing a more sustainable planet. From the abundant work dealing with WS and generation using direct solar energy or electrical power, there are many factors that affect the mechanism and performances of the catalyst (e.g., active-site activity, given by the metal and ligand selection and by their coordination geometry, reaction medium, sacrificial agent for photocatalysis, reactor design, as well as overpotential and applied potential for electrocatalysis), so it is advisable to take all of them into account, even if the focus is centered just on some. Moreover, actual focus is often directed in finding high materials’ performances but with strategies or reaction media not applicable in real working conditions, so shifting the design of the materials and studying their performances under real conditions to be implemented into the practical device could help to tackle some lab issues (e.g., reactor design [162]). The general trend using CPs or MOFs as catalyst is that electrocatalysis reaches much higher TOFs than photocatalytic processes (103–102 s1 vs. 107–100 s1), also related with the higher energy input in the system than in photocatalysis. Moreover, it relies on simpler reaction designs, avoiding, for instance, the use of sacrificial agents. On the other side, it suffers from low specific activity and available active sites of the material. One should also consider that the infrastructure to use electrocatalysts (a renewable energy installation) could make more complex the installation of such technology. A crucial point that should be considered in this kind of catalytic processes is the material stability under specific working conditions and its characterization after the process. One could tentatively propose a different and standardized/comparable way to determine the durability of a catalyst by evaluating the time needed to reach a halfperformance than the starting one. In addition, among the parameters affecting the performances (regardless of the type of catalysis), porosity emerges as an important one and should be considered in more detail (particularly for electrocatalytic applications, where its potential does not seem to be actually highlighted). Complementary to the material micro-/mesoporosity, also the external surface area, which is usually excluded in CP/MOF photocatalysts, has revealed a major role. A particular limitation in photocatalysis is the difficulty to compare several works as performed under different conditions (e.g., lamps, reactor size and material). Also, particularly important for studies conducted under UV light is the material of the reactor screen (e.g., type of glass, quartz), not indicated in many researches. Other factors such as reagent diffusion parameters (length, rate) could be beneficially considered. Furthermore, the study of pure MOF catalysts (without the presence of cocatalysts) would be interesting from a fundamental (better understanding of degradation kinetics and cocatalyzed systems) and practical point of view (low-cost efficient catalysts). On the other hand, in electrocatalysis, many of the currently considered parameters are dependent on the potential ( j, TOF, faradaic efficiency), being beneficial to
100
G. Armani-Calligaris et al.
shift to more intrinsic parameters or plots in the future. For instance, pH dependency (within the stability range) of η and Tafel slope might be good parameters to evaluate the applicability of catalysts under different environments, as well as indicators of the catalytic mechanism. In addition, many studies have highlighted the support effect on the η value. In this sense, the use of electrochemical impedance spectroscopy can indeed help to distinguish the contributions to the final required η, from the active site to the resistivity at the collector surface. Despite that CPs/MOFs are generally known to be insulators, conductivity measurements are great complementary tools to electrochemical impedance spectroscopy to understand this effect. Another important factor that could be considered is to monitor the CPs stability under the ink preparation (e.g., drop casting, sonication) and working conditions. Also, many works reported catalysts based on a mixed-metal approach with higher performances. Complementary studies in this line could reveal larger composition ranges and correlate also multicomponent behaviors, not just to optimize the performances through composition but also to understand the process (e.g., solid solutions formation, generated magnetic effects). On the whole, although CPs and MOFs have proven a great potential in energetic applications, this field is still at its infancy and requires further and extensive development. Acknowledgments The work has been supported by IMDEA Energy Foundation, BBVA Foundation (ref: IN [17] CBB_QUI_0197), Ramón Areces Foundation (H+MOFs), Raphuel project (ENE2016-79608-C2-1-R, MINECOAEI /FEDER, UE), and M-ERA.net (2019, C-MOF.cell, MCI/AEI/FEDER, UE). S.R. and P.H. acknowledge the financial support of Madrid Community (Recruitment of Young doctors 2017, Talento Modality 2, 2017-T2/IND-5149). P.H. acknowledges the Spanish Ramón y Cajal Programme (Grant Agreements 2014-15039). G.A.C. thanks the FPI fellowship (BES-2017-082749) associated with the Raphuel project (ENE2016-79608-C2-1-R, M INECOAEI/FEDER, UE). Finally, we would like to thank Dr. Edgar Ventosa from the Electrochemical Processes Unit in IMDEA Energy for the kind suggestions and useful comments.
References 1. IPCC (2013) Climate change 2013: the physical science basis. Contribution of Working Group I to the fifth assessment report of the Intergovernmental Panel on Climate Change. Cambridge University Press. Available at https://www.ipcc.ch/site/assets/uploads/2018/02/WG1AR5_ SPM_FINAL.pdf. Accessed 20 Apr 2020 2. Grand View Research, Inc. (2020) Hydrogen generation market size, share & trends analysis report by application (coal gasification, steam methane reforming), by systems (merchant, captive), by technology, and segment forecasts, 2020–2027. Available at https://www. grandviewresearch.com/industry-analysis/hydrogen-generation-market 3. Agency IE (2019) The future of hydrogen. Available at https://www.iea.org/reports/the-futureof-hydrogen 4. Wen T, Hu C (1992) Hydrogen and oxygen evolutions on Ru-Ir binary oxides. J Electrochem Soc 139(8):2158. https://doi.org/10.1149/1.2221195
3 Green Energy Generation Using Metal-Organic Frameworks
101
5. Di Blasi A, D’Urso C, Baglio V et al (2009) Preparation and evaluation of RuO2–IrO2, IrO2–Pt and IrO2–Ta2O5 catalysts for the oxygen evolution reaction in an SPE electrolyzer. J Appl Electrochem 39:191–196. https://doi.org/10.1007/s10800-008-9651-y 6. Esposito DV, Hunt ST, Stottlemyer AL et al (2010) Low-cost hydrogen-evolution catalysts based on monolayer platinum on tungsten monocarbide substrates. Angew Chem Int Ed 49:9859–9862. https://doi.org/10.1002/anie.201004718 7. Ouyang C, Wang X, Wang C et al (2015) Hierarchically porous Ni3S2 nanorod array foam as highly efficient electrocatalyst for hydrogen evolution reaction and oxygen evolution reaction. Electrochim Acta 174:297–301. https://doi.org/10.1016/j.electacta.2015.05.186 8. Yin X, Yan Y, Miao M et al (2018) Quasi-emulsion confined synthesis of edge-rich ultrathin MoS2 nanosheets/graphene hybrid for enhanced hydrogen evolution. Electrochemistry 24:556–560. https://doi.org/10.1002/chem.201703493 9. Chen W, Muckerman JT, Fujita E, Chen W (2013) Recent developments in transition metal carbides and nitrides as hydrogen evolution electrocatalysts. Chem Commun 49:8896–8909. https://doi.org/10.1039/c3cc44076a 10. Li J, Wang Y, Liu C et al (2016) Coupled molybdenum carbide and reduced graphene oxide electrocatalysts for efficient hydrogen evolution. Nat Commun 7:1–8. https://doi.org/10.1038/ ncomms11204 11. Xia BY, Yan Y, Wang X et al (2014) Recent progress on graphene-based hybrid electrocatalysts. Mater Horizon 1:379–399. https://doi.org/10.1039/c4mh00040d 12. Zou SX, Zhang Y (2015) Noble metal-free hydrogen evolution catalysts for water splitting. Chem Soc Rev 44:5148–5180. https://doi.org/10.1039/c4cs00448e 13. Kibsgaard J, Jaramillo TF, Besenbacher F (2014) Building an appropriate active-site motif into a hydrogen-evolution catalyst with thiomolybdate [Mo3S13]2- clusters. Nat Chem 6:248–253. https://doi.org/10.1038/nchem.1853 14. Tran PD, Tran TV, Orio M et al (2016) Coordination polymer structure and revisited hydrogen evolution catalytic. Nat Mater 15:640–647. https://doi.org/10.1038/NMAT4588 15. International Union of Pure and Applied Chemistry (2019) Top ten emerging technologies in chemistry. Available at https://iupac.org/what-we-do/top-ten/. Accessed 18 Feb 2020 16. Yang Q, Xu Q (2017) Metal–organic frameworks meet metal nanoparticles: synergistic effect for enhanced catalysis. Chem Soc Rev 46:4774–4808. https://doi.org/10.1039/C6CS00724D 17. Rahmani A, Emrooz HBM, Abedi S, Morsali A (2018) Synthesis and characterization of CdS/ MIL-125 (Ti) as a photocatalyst for water splitting. Mater Sci Semicond Process 80 (2017):44–51. https://doi.org/10.1016/j.mssp.2018.02.013 18. Samaniyan M, Mirzaei M, Khajavian R et al (2019) Heterogeneous catalysis by polyoxometalates in metal-organic frameworks. ACS Catal 9(11):10174–10191. https://doi. org/10.1021/acscatal.9b03439 19. Drout RJ, Robison L, Farha OK (2019) Catalytic applications of enzymes encapsulated in metal–organic frameworks. Coord Chem Rev 381:151–160. https://doi.org/10.1016/j.ccr. 2018.11.009 20. Batten SR, Champness NR, Chen X-M et al (2013) Terminology of metal–organic frameworks and coordination polymers (IUPAC recommendations 2013). Pure Appl Chem 85 (8):1715–1724. https://doi.org/10.1351/PAC-REC-12-11-20 21. Wang H-F, Chen L, Pang H et al (2020) MOF-derived electrocatalysts for oxygen reduction, oxygen evolution and hydrogen evolution reactions. Chem Soc Rev 49:1414–1448. https:// doi.org/10.1039/C9CS00906J 22. Horiuchi Y, Toyao T, Saito M et al (2012) Visible-light-promoted photocatalytic hydrogen production by using an amino-functionalized Ti(IV) metal–organic framework. J Phys Chem C 116(39):20848–20853. https://doi.org/10.1021/jp3046005 23. Dhakshinamoorthy A, Navalon S, Asiri AM, Garcia H (2019) Metal organic frameworks as solid catalysts for liquid-phase continuous flow reactions. Chem Commun 56(1):26–45. https://doi.org/10.1039/c9cc07953j
102
G. Armani-Calligaris et al.
24. Nivetha R, Kollu P, Chandar K et al (2019) Role of MIL-53(Fe)/hydrated–dehydrated MOF catalyst for electrochemical hydrogen evolution reaction (HER) in alkaline medium and photocatalysis. RSC Adv 9(6):3215–3223. https://doi.org/10.1039/C8RA08208A 25. Zlotea C, Phanon D, Mazaj M et al (2011) Effect of NH2 and CF3 functionalization on the hydrogen sorption properties of MOFs. Dalton Trans 40:4879–4881. https://doi.org/10.1039/ c1dt10115c 26. Guo F, Guo J-H, Wang P et al (2019) Facet-dependent photocatalytic hydrogen production of metal–organic framework NH2-MIL-125(Ti). Chem Sci 10(18):4834–4838. https://doi.org/ 10.1039/C8SC05060K 27. Ma L, Falkowski JM, Abney C, Lin W (2010) A series of isoreticular chiral metal–organic frameworks as a tunable platform for asymmetric catalysis. Nat Chem 2(10):838–846. https:// doi.org/10.1038/nchem.738 28. Tahmouresilerd B, Moody M, Agogo L, Cozzolino AF (2019) The impact of an isoreticular expansion strategy on the performance of iodine catalysts supported in multivariate zirconium and aluminum metal–organic frameworks. Dalton Trans 48(19):6445–6454. https://doi.org/ 10.1039/C9DT00368A 29. Wu X-P, Gagliardi L, Truhlar DG (2018) Cerium metal–organic framework for photocatalysis. J Am Chem Soc 140(25):7904–7912. https://doi.org/10.1021/jacs.8b03613 30. Lammert M, Wharmby MT, Smolders S et al (2015) Cerium-based metal organic frameworks with UiO-66 architecture: synthesis, properties and redox catalytic activity. Chem Commun 51 (63):12578–12581. https://doi.org/10.1039/C5CC02606G 31. Wu X-P, Gagliardi L, Truhlar DG (2018) Metal doping in cerium metal-organic frameworks for visible-response water splitting photocatalysts. J Chem Phys 150(4):41701–41711. https:// doi.org/10.1063/1.5043538 32. Wu P, Guo X, Cheng L et al (2016) Photoactive metal–organic framework and its film for light-driven hydrogen production and carbon dioxide reduction. Inorg Chem 55 (16):8153–8159. https://doi.org/10.1021/acs.inorgchem.6b01267 33. Vermoortele F, Vandichel M, Van De Voorde B et al (2012) Electronic effects of linker substitution on Lewis acid catalysis with metal-organic frameworks. Angew Chem Int Ed 51 (20):4887–4890. https://doi.org/10.1002/anie.201108565 34. Lionet Z, Kim T-H, Horiuchi Y et al (2019) Linker engineering of iron-based MOFs for efficient visible-light-driven water oxidation reaction. J Phys Chem C 123(45):27501–27508. https://doi.org/10.1021/acs.jpcc.9b06838 35. Hendon CH, Tiana D, Fontecave M et al (2013) Engineering the optical response of the titanium-MIL-125 metal-organic framework through ligand functionalisation. J Am Chem Soc 135(30):10942–10945. https://doi.org/10.1021/ja405350u 36. Chambers MB, Wang X, Ellezam L et al (2017) Maximizing the photocatalytic activity of metal–organic frameworks with aminated-functionalized linkers: substoichiometric effects in MIL-125-NH2. J Am Chem Soc 139(24):8222–8228. https://doi.org/10.1021/jacs.7b02186 37. Yuan S, Qin J-S, Xu H-Q et al (2018) [Ti8Zr2O12(COO)16] cluster: an ideal inorganic building unit for photoactive metal–organic frameworks. ACS Cent Sci 4(1):105–111. https://doi.org/ 10.1021/acscentsci.7b00497 38. Li C, Xu H, Gao J et al (2019) Tunable titanium metal–organic frameworks with infinite 1D Ti–O rods for efficient visible-light-driven photocatalytic H2 evolution. J Mater Chem A 7 (19):11928–11933. https://doi.org/10.1039/C9TA01942A 39. Song T, Zhang L, Zhang P et al (2017) Stable and improved visible-light photocatalytic hydrogen evolution using copper(II)–organic frameworks: engineering the crystal structures. J Mater Chem A 5(13):6013–6018. https://doi.org/10.1039/C7TA00095B 40. Pellegrin Y, Odobel F (2017) Sacrificial electron donor reagents for solar fuel production. Comptes Rendus Chim 20(3):283–295. https://doi.org/10.1016/j.crci.2015.11.026 41. Schneider J, Bahnemann DW (2013) Undesired role of sacrificial reagents in photocatalysis. J Phys Chem Lett 4(20):3479–3483. https://doi.org/10.1021/jz4018199
3 Green Energy Generation Using Metal-Organic Frameworks
103
42. Kumaravel V, Imam MD, Badreldin A et al (2019) Photocatalytic hydrogen production: role of sacrificial reagents on the activity of oxide, carbon, and sulfide catalysts. Catalysts 9 (3):276–311. https://doi.org/10.3390/catal9030276 43. Liu H, Xu C, Li D, Jiang H-L (2018) Photocatalytic hydrogen production coupled with selective benzylamine oxidation over MOF composites. Angew Chem Int Ed 57 (19):5379–5383. https://doi.org/10.1002/anie.201800320 44. Horiuchi Y, Toyao T, Miyahara K et al (2016) Visible-light-driven photocatalytic water oxidation catalysed by iron-based metal–organic frameworks. Chem Commun 52 (29):5190–5193. https://doi.org/10.1039/C6CC00730A 45. Bard AJ, Faulkner LR (2000) Electrochemical methods: fundamentals and applications, 2nd edn. Wiley 46. Walter MG, Warren EL, McKone JR et al (2010) Solar water splitting cells. Chem Rev 110 (11):6446–6473. https://doi.org/10.1021/cr1002326 47. EDAQ (2016) Reference electrode potentials. Available at https://www.edaq.com/wiki/ Reference_Electrode_Potentials. Accessed 28 Apr 2020 48. Bates RG, Macaskill JB (1978) Standard potential of the silver-silver chloride electrode. Pure Appl Chem 50(11–12):1701–1706. https://doi.org/10.1351/pac197850111701 49. Haynes WM (ed) (2014) CRC handbook of chemistry and physics, 95th edn. CRC Press 50. Shen J-Q, Liao P-Q, Zhou D-D et al (2017) Modular and stepwise synthesis of a hybrid metal– organic framework for efficient electrocatalytic oxygen evolution. J Am Chem Soc 139 (5):1778–1781. https://doi.org/10.1021/jacs.6b12353 51. Tripathy RK, Samantara AK, Behera JN (2019) A cobalt metal–organic framework (Co-MOF): a bi-functional electro active material for the oxygen evolution and reduction reaction. Dalton Trans 48(28):10557–10564. https://doi.org/10.1039/C9DT01730E 52. Gutiérrez Tarriño S, Olloqui-Sariego JL, Calvente JJ et al (2019) Cobalt metal-organic framework based on two dinuclear secondary building units for electrocatalytic oxygen evolution. ACS Appl Mater Interfaces 11(50):46658–46665. https://doi.org/10.1021/acsami. 9b13655 53. Wang X-L, Dong L-Z, Qiao M et al (2018) Exploring the performance improvement of the oxygen evolution reaction in a stable bimetal–organic framework system. Angew Chem Int Ed 57(31):9660–9664. https://doi.org/10.1002/anie.201803587 54. Xue Z, Li Y, Zhang Y et al (2018) Modulating electronic structure of metal-organic framework for efficient electrocatalytic oxygen evolution. Adv Energy Mater 8(29):1801564–1801571. https://doi.org/10.1002/aenm.201801564 55. Zhou W, Huang D-D, Wu Y-P et al (2019) Stable hierarchical bimetal–organic nanostructures as high-performance electrocatalysts for the oxygen evolution reaction. Angew Chem Int Ed 58(13):4227–4231. https://doi.org/10.1002/anie.201813634 56. Tao L, Lin C-Y, Dou S et al (2017) Creating coordinatively unsaturated metal sites in metalorganic-frameworks as efficient electrocatalysts for the oxygen evolution reaction: insights into the active centers. Nano Energy 41:417–425. https://doi.org/10.1016/j.nanoen.2017.09. 055 57. Li W, Li F, Yang H et al (2019) A bio-inspired coordination polymer as outstanding water oxidation catalyst via second coordination sphere engineering. Nat Commun 10 (1):5074–5085. https://doi.org/10.1038/s41467-019-13052-1 58. Das TK, Prusty S (2012) Review on conducting polymers and their applications. Polym-Plast Technol Eng 51(14):1487–1500. https://doi.org/10.1080/03602559.2012.710697 59. Jia H, Yao Y, Zhao J et al (2018) A novel two-dimensional nickel phthalocyanine-based metal–organic framework for highly efficient water oxidation catalysis. J Mater Chem A 6 (3):1188–1195. https://doi.org/10.1039/C7TA07978H 60. Shinde SS, Lee CH, Jung J-Y et al (2019) Unveiling dual-linkage 3D hexaiminobenzene metal–organic frameworks towards long-lasting advanced reversible Zn–air batteries. Energy Environ Sci 12(2):727–738. https://doi.org/10.1039/C8EE02679C
104
G. Armani-Calligaris et al.
61. Miner EM, Fukushima T, Sheberla D et al (2016) Electrochemical oxygen reduction catalysed by Ni3(hexaiminotriphenylene)2. Nat Commun 7(1):10942–10949. https://doi.org/10.1038/ ncomms10942 62. Du J, Xu S, Sun L, Li F (2019) Iron carbonate hydroxide templated binary metal–organic frameworks for highly efficient electrochemical water oxidation. Chem Commun 55:14773–14776. https://doi.org/10.1039/C9CC07433C 63. Zhao S, Wang Y, Dong J et al (2016) Ultrathin metal–organic framework nanosheets for electrocatalytic oxygen evolution. Nat Energy 1(12):16184–16194. https://doi.org/10.1038/ nenergy.2016.184 64. Zhang K, Kirlikovali KO, Van LQ et al (2020) Extended metal-organic frameworks on diverse supports as electrode nanomaterials for electrochemical energy storage. ACS Appl Nano Mater 3:3964–3990. https://doi.org/10.1021/acsanm.0c00702 65. Luo P, Li S, Zhao Y et al (2019) In-situ growth of a bimetallic cobalt-nickel organic framework on Iron foam: achieving the electron modification on a robust self-supported oxygen evolution electrode. ChemCatChem 11:6061–6069. https://doi.org/10.1002/cctc. 201900972 66. He D-H, Liu J-J, Wang Y et al (2019) Electrocatalysis of the first electron transfer in hydrogen evolution reaction with an atomically precise CuII-organic framework catalyst. Electrochim Acta 308:285–294. https://doi.org/10.1016/j.electacta.2019.04.038 67. Kung C-W, Mondloch JE, Wang TC et al (2015) Metal–organic framework thin films as platforms for atomic layer deposition of cobalt ions to enable electrocatalytic water oxidation. ACS Appl Mater Interfaces 7(51):28223–28230. https://doi.org/10.1021/acsami.5b06901 68. Singh C, Liberman I, Shimoni R et al (2019) Pristine versus pyrolyzed metal–organic framework-based oxygen evolution electrocatalysts: evaluation of intrinsic activity using electrochemical impedance spectroscopy. J Phys Chem Lett 10(13):3630–3636. https://doi. org/10.1021/acs.jpclett.9b01232 69. Li X-F, Lu M-Y, Yu H-Y et al (2019) Copper-metal organic frameworks electrodeposited on carbon paper as an enhanced cathode for the hydrogen evolution reaction. ChemElectroChem 6(17):4507–4510. https://doi.org/10.1002/celc.201901153 70. Wang L, Wu Y, Cao R et al (2016) Fe/Ni metal–organic frameworks and their binder-free thin films for efficient oxygen evolution with low overpotential. ACS Appl Mater Interfaces 8 (26):16736–16743. https://doi.org/10.1021/acsami.6b05375 71. Wang Q, Liu F, Wei C et al (2019) High efficiency FeNi-metal-organic framework grown in-situ on nickel foam for Electrocatalytic oxygen evolution. ChemistrySelect 4 (19):5988–5994. https://doi.org/10.1002/slct.201901709 72. Newport Corp. Introduction to solar radiation. In: Newport Corp. Available at https://www. newport.com/t/introduction-to-solar-radiation. Accessed 21 May 2020 73. Ciamician G (1912) The photochemistry of the future. Science 36(926):385–394. https://doi. org/10.1126/science.36.926.385 74. Fujishima A, Honda K (1972) Electrochemical photolysis of water at a semiconductor electrode. Nature 238(5358):37–38. https://doi.org/10.1038/238037a0 75. National Renewable Energy Laboratory Reference Air Mass 1.5 Spectra. Available at https:// www.nrel.gov/grid/solar-resource/spectra-am1.5.html. Accessed 2 May 2020 76. Kataoka Y, Sato K, Miyazaki Y et al (2009) Photocatalytic hydrogen production from water using porous material [Ru2( p-BDC)2]n. Energy Environ Sci 2(4):397–400. https://doi.org/10. 1039/B814539C 77. Lan G, Zhu Y-Y, Veroneau SS et al (2018) Electron injection from Photoexcited metal– organic framework ligands to Ru2 secondary building units for visible-light-driven hydrogen evolution. J Am Chem Soc 140(16):5326–5329. https://doi.org/10.1021/jacs.8b01601 78. Dikhtiarenko A, Villanueva-Delgado P, Valiente R et al (2016) Tris(bipyridine)metal(II)templated assemblies of 3D alkali-ruthenium oxalate coordination frameworks: crystal structures, characterization and photocatalytic activity in water reduction. Polymers (Basel) 8 (2):48–69. https://doi.org/10.3390/polym8020048
3 Green Energy Generation Using Metal-Organic Frameworks
105
79. Dan-Hardi M, Serre C, Frot T et al (2009) A new photoactive highly porous titanium (IV) dicarboxylate. J Am Chem Soc 131(31):10857–10859. https://doi.org/10.1021/ ja903726m 80. Song T, Zhang P, Zeng J et al (2017) Tunable conduction band energy and metal-to-ligand charge transfer for wide-spectrum photocatalytic H2 evolution and stability from isostructural metal-organic frameworks. Int J Hydrog Energy 42(43):26605–26616. https://doi.org/10. 1016/j.ijhydene.2017.09.081 81. Shi D, Zheng R, Sun M-J et al (2017) Semiconductive copper(I)–organic frameworks for efficient light-driven hydrogen generation without additional photosensitizers and cocatalysts. Angew Chem Int Ed 56(46):14637–14641. https://doi.org/10.1002/anie.201709869 82. Batten SR, Jeffery JC, Ward MD (1999) Studies of the construction of coordination polymers using linear pyridyl-donor ligands. Inorg Chim Acta 292(2):231–237. https://doi.org/10.1016/ S0020-1693(99)00203-0 83. Dong D, Yan C, Huang J et al (2019) An electron-donating strategy to guide the construction of MOF photocatalysts toward co-catalyst-free highly efficient photocatalytic H2 evolution. J Mater Chem A 7(42):24180–24185. https://doi.org/10.1039/C9TA06141J 84. Luo L, Lin H, Li L et al (2014) Copper-organic/octamolybdates: structures, bandgap sizes, and photocatalytic activities. Inorg Chem 53(7):3464–3470. https://doi.org/10.1021/ic402910a 85. Gomes Silva C, Luz I, Llabrés i, Xamena FX et al (2010) Water stable Zr– benzenedicarboxylate metal–organic frameworks as photocatalysts for hydrogen generation. Chem Eur J 16(36):11133–11138. https://doi.org/10.1002/chem.200903526 86. Yu Q, Dong H, Zhang X et al (2018) Novel stable metal–organic framework photocatalyst for light-driven hydrogen production. CrystEngComm 20(23):3228–3233. https://doi.org/10. 1039/C8CE00386F 87. Sun X, Yu Q, Zhang F et al (2016) A dye-like ligand-based metal–organic framework for efficient photocatalytic hydrogen production from aqueous solution. Cat Sci Technol 6 (11):3840–3844. https://doi.org/10.1039/C5CY01716E 88. Yano J, Kern J, Yachandra VK et al (2015) Light-dependent production of dioxygen in photosynthesis BT. In: Kroneck PMH, Sosa Torres ME (eds) Sustaining life on planet earth: metalloenzymes mastering dioxygen and other chewy gases. Springer, Cham, pp 13–43 89. Fukuzumi S, Jung J, Yamada Y et al (2016) Homogeneous and heterogeneous photocatalytic water oxidation by persulfate. Chem An Asian J 11(8):1138–1150. https://doi.org/10.1002/ asia.201501329 90. Llobet A, Romain S (2005) Oxygen production from water: molecular catalysts. Encycl Inorg Chem. https://doi.org/10.1002/0470862106.ia804 91. Blakemore JD, Crabtree RH, Brudvig GW (2015) Molecular catalysts for water oxidation. Chem Rev 115(23):12974–13005. https://doi.org/10.1021/acs.chemrev.5b00122 92. Wang X-G, Cheng Q, Yu Y, Zhang X-Z (2018) A general and scalable approach with controlled nucleation and controlled growth for size predicable synthesis of nanoscale MOFs. Angew Chem Int Ed 57(26):7836–7840. https://doi.org/10.1002/anie.201803766 93. Qu L-L, Wang J, Xu T-Y et al (2018) Iron(III)-based metal–organic frameworks as oxygenevolving photocatalysts for water oxidation. Sustain Energy Fuels 2(9):2109–2114. https:// doi.org/10.1039/C8SE00311D 94. An Y, Li H, Liu Y et al (2016) Photoelectrical, photophysical and photocatalytic properties of Al based MOFs: MIL-53(Al) and MIL-53-NH2(Al). J Solid State Chem 233:194–198. https:// doi.org/10.1016/j.jssc.2015.10.037 95. Wang G, Sun Q, Liu Y et al (2015) A bismuth-based metal–organic framework as an efficient visible-light-driven photocatalyst. Chem Eur J 21(6):2364–2367. https://doi.org/10.1002/ chem.201405047 96. Wang G, Liu Y, Huang B et al (2015) A novel metal–organic framework based on bismuth and trimesic acid: synthesis, structure and properties. Dalton Trans 44(37):16238–16241. https:// doi.org/10.1039/C5DT03111G
106
G. Armani-Calligaris et al.
97. Bagtache R, Rekhila G, Abdmeziem K, Trari M (2014) Characterization of a copper phosphate triazole metal organic framework material (Cu3PO4(C2N3H2)2OH) and oxygen evolution studies. Mater Sci Semicond Process 23:144–150. https://doi.org/10.1016/j.mssp.2014.02.018 98. Lionet Z, Kim T-H, Horiuchi Y et al (2019) Linker engineering of iron-based MOFs for efficient visible-light-driven water oxidation reaction. J Phys Chem C 123(45):27501–27508. https://doi.org/10.1021/acs.jpcc.9b06838 99. Robertson R (2017) Electrogeneration of hydrogen peroxide for applications in water/wastewater treatment. University of Waterloo 100. Sohrabi S, Ghalkhani M, Dehghanpour S (2019) The Electrocatalytic stability investigation of a proton manager MOF for the oxygen reduction reaction in acidic media. J Inorg Organomet Polym Mater 29(2):528–534. https://doi.org/10.1007/s10904-018-1025-2 101. Cheng W, Zhao X, Su H et al (2019) Lattice-strained metal–organic-framework arrays for bifunctional oxygen electrocatalysis. Nat Energy 4(2):115–122. https://doi.org/10.1038/ s41560-018-0308-8 102. Das D, Raut V, Kireeti KVMK, Jha N (2018) Non-platinum metal-organic framework based electro-catalyst for promoting oxygen reduction reaction. AIP Conf Proc 1942 (1):140049–140054. https://doi.org/10.1063/1.5029180 103. Liu X-H, Hu W-L, Jiang W-J et al (2017) Well-defined metal–O6 in metal–catecholates as a novel active site for oxygen electroreduction. ACS Appl Mater Interfaces 9(34):28473–28477. https://doi.org/10.1021/acsami.7b07410 104. Rousseau G, Rodriguez-Albelo LM, Salomon W et al (2015) Tuning the dimensionality of polyoxometalate-based materials by using a mixture of ligands. Cryst Growth Des 15 (1):449–456. https://doi.org/10.1021/cg501524a 105. Nohra B, El Moll H, Rodriguez Albelo LM et al (2011) Polyoxometalate-based metal organic frameworks (POMOFs): structural trends, energetics, and high electrocatalytic efficiency for hydrogen evolution reaction. J Am Chem Soc 133(34):13363–13374. https://doi.org/10.1021/ ja201165c 106. Wu Y-P, Zhou W, Zhao J et al (2017) Surfactant-assisted phase-selective synthesis of new cobalt MOFs and their efficient electrocatalytic hydrogen evolution reaction. Angew Chem Int Ed 56(42):13001–13005. https://doi.org/10.1002/anie.201707238 107. Zhou Y-C, Dong W-W, Jiang M-Y et al (2019) A new 3D 8-fold interpenetrating 66-dia topological Co-MOF: syntheses, crystal structure, magnetic properties and electrocatalytic hydrogen evolution reaction. J Solid State Chem 279:120929–120934. https://doi.org/10. 1016/j.jssc.2019.120929 108. Muthukumar P, Moon D, Anthony SP (2019) The Co2+/Ni2+ ion-mediated formation of a topochemically converted copper coordination polymer: structure-dependent electrocatalytic activity. CrystEngComm 21(43):6552–6557. https://doi.org/10.1039/C9CE01178A 109. Wang T, Huang K, Peng M et al (2019) Metal–organic frameworks based on tetraphenylpyrazine-derived tetracarboxylic acid for electrocatalytic hydrogen evolution reaction and NAC sensing. CrystEngComm 21(3):494–501. https://doi.org/10.1039/C8CE01868E 110. Gong Y, Wu T, Jiang PG et al (2013) Octamolybdate-based metal–organic framework with unsaturated coordinated metal center as electrocatalyst for generating hydrogen from water. Inorg Chem 52(2):777–784. https://doi.org/10.1021/ic3018858 111. Gong Y, Hao Z, Meng J et al (2014) Two CoII metal–organic frameworks based on a multicarboxylate ligand as electrocatalysts for water splitting. ChemPlusChem 79 (2):266–277. https://doi.org/10.1002/cplu.201300334 112. Gong Y, Shi H-F, Jiang P-G et al (2014) Metal(II)-induced coordination polymer based on 4-(5-(Pyridin-4-yl)-4H-1,2,4-triazol-3-yl)benzoate as an electrocatalyst for water splitting. Cryst Growth Des 14(2):649–657. https://doi.org/10.1021/cg401529u 113. Qin J-S, Du D-Y, Guan W et al (2015) Ultrastable polymolybdate-based metal–organic frameworks as highly active electrocatalysts for hydrogen generation from water. J Am Chem Soc 137(22):7169–7177. https://doi.org/10.1021/jacs.5b02688
3 Green Energy Generation Using Metal-Organic Frameworks
107
114. Salomon W, Paille G, Gomez-Mingot M et al (2017) Effect of cations on the structure and electrocatalytic response of polyoxometalate-based coordination polymers. Cryst Growth Des 17(4):1600–1609. https://doi.org/10.1021/acs.cgd.6b01600 115. Burke MS, Enman LJ, Batchellor AS et al (2015) Oxygen evolution reaction electrocatalysis on transition metal oxides and (oxy)hydroxides: activity trends and design principles. Chem Mater 27(22):7549–7558. https://doi.org/10.1021/acs.chemmater.5b03148 116. Kim HK, Shin I-S (2015) Electrochemical investigation of metal-organic framework MIL-100 (Fe) and its electrocatalytic activity towards hydroxide oxidation. Bull Kor Chem Soc 36 (3):1051–1053. https://doi.org/10.1002/bkcs.10185 117. Bezverkhyy I, Weber G, Bellat J-P (2016) Degradation of fluoride-free MIL-100(Fe) and MIL-53(Fe) in water: effect of temperature and pH. Microporous Mesoporous Mater 219:117–124. https://doi.org/10.1016/j.micromeso.2015.07.037 118. Wang Q, Wei C, Li D et al (2019) FeNi-based bimetallic MIL-101 directly applicable as an efficient electrocatalyst for oxygen evolution reaction. Microporous Mesoporous Mater 286:92–97. https://doi.org/10.1016/j.micromeso.2019.05.040 119. Xue Z, Liu K, Liu Q et al (2019) Missing-linker metal-organic frameworks for oxygen evolution reaction. Nat Commun 10(1):5048–5056. https://doi.org/10.1038/s41467-01913051-2 120. Lu X-F, Liao P-Q, Wang J-W et al (2016) An alkaline-stable, metal hydroxide mimicking metal–organic framework for efficient Electrocatalytic oxygen evolution. J Am Chem Soc 138 (27):8336–8339. https://doi.org/10.1021/jacs.6b03125 121. Wang S, Hou Y, Lin S, Wang X (2014) Water oxidation electrocatalysis by a zeolitic imidazolate framework. Nanoscale 6(17):9930–9934. https://doi.org/10.1039/C4NR02399D 122. Xu Y-T, Ye Z-M, Ye J-W et al (2019) Non-3d metal modulation of a cobalt Imidazolate framework for excellent electrocatalytic oxygen evolution in neutral media. Angew Chem Int Ed 58(1):139–143. https://doi.org/10.1002/anie.201809144 123. Gong Y, Shi H-F, Hao Z et al (2013) Two novel Co(II) coordination polymers based on 1,4-bis (3-pyridylaminomethyl)benzene as electrocatalysts for oxygen evolution from water. Dalton Trans 42(34):12252–12259. https://doi.org/10.1039/C3DT50697E 124. Wang H, Yin F, Li G et al (2014) Preparation, characterization and bifunctional catalytic properties of MOF(Fe/Co) catalyst for oxygen reduction/evolution reactions in alkaline electrolyte. Int J Hydrog Energy 39(28):16179–16186. https://doi.org/10.1016/j.ijhydene. 2013.12.120 125. Dhara B, Sappati S, Singh SK et al (2016) Coordination polymers of Fe(III) and Al(III) ions with TCA ligand: distinctive fluorescence, CO2 uptake, redox-activity and oxygen evolution reaction. Dalton Trans 45(16):6901–6908. https://doi.org/10.1039/C6DT00009F 126. Dai F, Fan W, Bi J et al (2016) A lead–porphyrin metal–organic framework: gas adsorption properties and electrocatalytic activity for water oxidation. Dalton Trans 45(1):61–65. https:// doi.org/10.1039/C5DT04025F 127. Zhang D, Zhang J, Bai H et al (2016) ZIF-9 with enhanced surpercapacitor and electrocatalytic for oxygen evolution reaction performances in alkaline electrolyte. Int J Electrochem Sci 11 (9):7519–7526. https://doi.org/10.20964/2016.09.01 128. Flügel EA, Lau VW, Schlomberg H et al (2016) Homonuclear mixed-Valent cobalt Imidazolate framework for oxygen-evolution electrocatalysis. Chem Eur J 22 (11):3676–3680. https://doi.org/10.1002/chem.201504151 129. Meng Q, Yang J, Ma S et al (2017) A porous cobalt (II) metal–organic framework with highly efficient electrocatalytic activity for the oxygen evolution reaction. Polymers (Basel) 9 (12):676–687. https://doi.org/10.3390/polym9120676 130. Jiang J, Huang L, Liu X, Ai L (2017) Bioinspired cobalt–citrate metal–organic framework as an efficient electrocatalyst for water oxidation. ACS Appl Mater Interfaces 9(8):7193–7201. https://doi.org/10.1021/acsami.6b16534
108
G. Armani-Calligaris et al.
131. Maruthapandian V, Kumaraguru S, Mohan S et al (2018) An insight on the electrocatalytic mechanistic study of pristine Ni MOF (BTC) in alkaline medium for enhanced OER and UOR. ChemElectroChem 5(19):2795–2807. https://doi.org/10.1002/celc.201800802 132. Li F-L, Shao Q, Huang X, Lang J-P (2018) Nanoscale trimetallic metal–organic frameworks enable efficient oxygen evolution electrocatalysis. Angew Chem Int Ed 57(7):1888–1892. https://doi.org/10.1002/anie.201711376 133. Rui K, Zhao G, Chen Y et al (2018) Hybrid 2D dual-metal–organic frameworks for enhanced water oxidation catalysis. Adv Funct Mater 28(26):1801554. https://doi.org/10.1002/adfm. 201801554 134. Liu X, Wang Y, Liu W et al (2018) Two water stable copper metal-organic frameworks with performance in the electrocatalytic activity for water oxidation. MATEC Web conference 142:01004 international conference on Materials Applications and Engineering 2017 (ICMAE2017). https://doi.org/10.1051/matecconf/201814201004 135. Liu Q, Xie L, Shi X et al (2018) High-performance water oxidation electrocatalysis enabled by a Ni-MOF nanosheet array. Inorg Chem Front 5(7):1570–1574. https://doi.org/10.1039/ C7QI00808B 136. Gao J, Cong J, Wu Y et al (2018) Bimetallic Hofmann-type metal–organic framework nanoparticles for efficient electrocatalysis of oxygen evolution reaction. ACS Appl Energy Mater 1(10):5140–5144. https://doi.org/10.1021/acsaem.8b01229 137. Wen T, Zheng Y, Xu C et al (2018) A boron imidazolate framework with mechanochromic and electrocatalytic properties. Mater Horizon 5(6):1151–1155. https://doi.org/10.1039/ C8MH00859K 138. Wen T, Zheng Y, Zhang J et al (2019) Co(II) boron imidazolate framework with rigid auxiliary linkers for stable electrocatalytic oxygen evolution reaction. Adv Sci 6(9):1801920. https:// doi.org/10.1002/advs.201801920 139. Zheng F, Xiang D, Li P et al (2019) Highly conductive bimetallic Ni–Fe metal organic framework as a novel electrocatalyst for water oxidation. ACS Sustain Chem Eng 7 (11):9743–9749. https://doi.org/10.1021/acssuschemeng.9b01131 140. Dong H, Zhang X, Yan X-C et al (2019) Mixed-metal-cluster strategy for boosting electrocatalytic oxygen evolution reaction of robust metal–organic frameworks. ACS Appl Mater Interfaces 11(48):45080–45086. https://doi.org/10.1021/acsami.9b14995 141. Zheng F, Zhang Z, Xiang D et al (2019) Fe/Ni bimetal organic framework as efficient oxygen evolution catalyst with low overpotential. J Colloid Interface Sci 555:541–547. https://doi.org/ 10.1016/j.jcis.2019.08.005 142. Wang X, Zhang H, Yang Z et al (2019) Ultrasound-treated metal-organic framework with efficient electrocatalytic oxygen evolution activity. Ultrason Sonochem 59:104714–104720. https://doi.org/10.1016/j.ultsonch.2019.104714 143. Iqbal B, Saleem M, Arshad SN et al (2019) One-pot synthesis of heterobimetallic metal– organic frameworks (MOFs) for multifunctional catalysis. Chem Eur J 25(44):10490–10498. https://doi.org/10.1002/chem.201901939 144. Tang Y, Zheng S, Xue H, Pang H (2019) Regulation of the Ni2+ content in a hierarchical urchin-like MOF for high-performance electrocatalytic oxygen evolution. Front Chem 7:411–417. https://doi.org/10.3389/fchem.2019.00411 145. Wang Q, Wei F, Manoj D et al (2019) In situ growth of Fe(II)-MOF-74 nanoarrays on nickel foam as an efficient electrocatalytic electrode for water oxidation: a mechanistic study on valence engineering. Chem Commun 55(75):11307–11310. https://doi.org/10.1039/ C9CC05087F 146. Bidault F, Brett DJL, Middleton PH, Brandon NP (2009) Review of gas diffusion cathodes for alkaline fuel cells. J Power Sources 187(1):39–48. https://doi.org/10.1016/j.jpowsour.2008. 10.106 147. Osmieri L (2019) Transition metal–nitrogen–carbon (M–N–C) catalysts for oxygen reduction reaction. Insights on synthesis and performance in polymer electrolyte fuel cells. ChemEngineering 3(1):16–48. https://doi.org/10.3390/chemengineering3010016
3 Green Energy Generation Using Metal-Organic Frameworks
109
148. Stolten D (ed) (2010) Hydrogen and fuel cells: fundamentals, technologies and applications. Wiley 149. Stephens IEL, Bondarenko AS, Grønbjerg U et al (2012) Understanding the electrocatalysis of oxygen reduction on platinum and its alloys. Energy Environ Sci 5(5):6744–6762. https://doi. org/10.1039/C2EE03590A 150. Mani P, Devadas S, Gurusamy T et al (2019) Sodalite-type Cu-based three-dimensional metal–organic framework for efficient oxygen reduction reaction. Chem An Asian J 14 (24):4814–4818. https://doi.org/10.1002/asia.201901242 151. Zhong H, Ly KH, Wang M et al (2019) A phthalocyanine-based layered two-dimensional conjugated metal–organic framework as a highly efficient electrocatalyst for the oxygen reduction reaction. Angew Chem Int Ed 58(31):10677–10682. https://doi.org/10.1002/anie. 201907002 152. Sheberla D, Sun L, Blood-Forsythe MA et al (2014) High electrical conductivity in Ni3(2,3,6,7,10,11-hexaiminotriphenylene)2, a semiconducting metal–organic graphene analogue. J Am Chem Soc 136(25):8859–8862. https://doi.org/10.1021/ja502765n 153. Sun F, Chen X (2017) Oxygen reduction reaction on Ni3(HITP)2: a catalytic site that leads to high activity. Electrochem Commun 82:89–92. https://doi.org/10.1016/j.elecom.2017.07.028 154. Jabarian S, Ghaffarinejad A, Kazemi H (2018) Electrochemical and solvothermal syntheses of HKUST-1 metal organic frameworks and comparison of their performances as electrocatalyst for oxygen reduction reaction. Anal Bioanal Electrochem 10(12):1611–1619 155. Mao J, Yang L, Yu P et al (2012) Electrocatalytic four-electron reduction of oxygen with Copper (II)-based metal-organic frameworks. Electrochem Commun 19:29–31. https://doi. org/10.1016/j.elecom.2012.02.025 156. Fateeva A, Chater PA, Ireland CP et al (2012) A water-stable porphyrin-based metal–organic framework active for visible-light Photocatalysis. Angew Chem Int Ed 51(30):7440–7444. https://doi.org/10.1002/anie.201202471 157. Lions M, Tommasino J-B, Chattot R et al (2017) Insights into the mechanism of electrocatalysis of the oxygen reduction reaction by a porphyrinic metal organic framework. Chem Commun 53(48):6496–6499. https://doi.org/10.1039/C7CC02113E 158. Usov PM, Huffman B, Epley CC et al (2017) Study of electrocatalytic properties of metal– organic framework PCN-223 for the oxygen reduction reaction. ACS Appl Mater Interfaces 9 (39):33539–33543. https://doi.org/10.1021/acsami.7b01547 159. Xue W, Zhou Q, Li F, Ondon BS (2019) Zeolitic imidazolate framework-8 (ZIF-8) as robust catalyst for oxygen reduction reaction in microbial fuel cells. J Power Sources 423:9–17. https://doi.org/10.1016/j.jpowsour.2019.03.017 160. Xia H, Zhang J, Yang Z et al (2017) 2D MOF nanoflake-assembled spherical microstructures for enhanced supercapacitor and electrocatalysis performances. Nano-Micro Lett 9(4):43–54. https://doi.org/10.1007/s40820-017-0144-6 161. Gonen S, Lori O, Fleker O, Elbaz L (2019) Electrocatalytically active silver organic framework: Ag(I)-complex incorporated in activated carbon. ChemCatChem 11:6124–6130. https:// doi.org/10.1002/cctc.201901604 162. Po R, Bernardi A, Calabrese A et al (2014) From lab to fab: how must the polymer solar cell materials design change? – An industrial perspective. Energy Environ Sci 7(3):925–943. https://doi.org/10.1039/C3EE43460E
Chapter 4
The Potential of MOFs in the Field of Electrochemical Energy Storage Thomas Devic
Abbreviations AQDC BArF BDC BET BTB BTC CMC Cp* CPO DANT DBBQ DEC DEF DMC DME DMF DOBPDC DPNDI EC EMIM HHTP HKUST IR KB
Anthraquinone dicarboxylate Tetrakis(perfluorophenyl)borate 1,4-Benzenedicarboxylate Brunauer-Emmett-Teller 1,3,5-Benzenetribenzoate 1,3,5-Benzenetricarboxylate Carboxymethylcellulose 1,2,3,4,5-Pentamethylcyclopentadienyl anion Coordination polymer of Oslo 2,5-(Dianilino)terephthalate Dibutylbenzoquinone Diethylcarbonate N,N-diethylformamide Dimethylcarbonate Dimethoxyethane N,N-dimethylformamide 4,40 -Dioxidobiphenyl-3,30 -dicarboxylate N,N0 -di(4-Pyridyl)-1,4,5,8-naphthalenediimide Ethylene carbonate 1-Ethyl-3-methylimidazolium 2,3,6,7,10,11-Hexahydroxytriphenylene Hong Kong University of Science and Technology Infrared Ketjen black
T. Devic (*) Institut des Matériaux Jean Rouxel (IMN), Université de Nantes, CNRS UMR 6502, Nantes, France e-mail: [email protected] © Springer Nature Switzerland AG 2021 P. Horcajada Cortés, S. Rojas Macías (eds.), Metal-Organic Frameworks in Biomedical and Environmental Field, https://doi.org/10.1007/978-3-030-63380-6_4
111
112
T. Devic
LFP LMNO LP30 LTO MIL MOF MTN NMC NMR OER ORR PAN PC PDF PMMA PTFE PVA PVDF TCNQ TEM TFSI TGA TTF UiO XAS XPS XRD ZIF
4.1
Lithium iron phosphate Lithium manganese nickel oxide 1 M solution of LiPF6 in a 1:1 EC:DMC electrolyte Lithium titanate oxide Material Institute Lavoisier Metal-organic framework Mobile technology network Nickel manganese cobalt oxide Nuclear magnetic resonance Oxygen evolution reaction Oxygen reduction reaction Polyaniline Propylene carbonate Pair distribution function Poly(methyl methacrylate) Poly(tetrafluoroethene) Polyvinyl alcohol Polyvinylidene fluoride 7,7,8,8-Tetracyano-p-quinodimethane Transmission electron microscopy Bis(trifluoromethanesulfonyl)imide Thermogravimetric analysis Tetrathiafulvalene University of Oslo X-ray absorption spectroscopy X-ray photoelectron spectroscopy X-ray diffraction Zeolitic imidazolate framework
Introduction
Metal-organic frameworks (MOFs) (we focus here mostly on crystalline materials) were initially developed for their porosity and hence, primarily considered for sorption-related applications. Nevertheless, the interest of such materials in the field of batteries and supercapacitors was identified as early as 2006, with the first report on the use of a Zn-carboxylate MOF (MOF-177) as an electrode in a lithium half cell. This solid exhibited a reversible redox activity at low potential ( 104, respectively). They thus appear as complementary rather than competitive energy storage solutions [21].
4.1.2
Devices
Whatever their potential practical use (as a positive or negative electrode), new battery-type materials are typically evaluated vs. a lithium metal electrode in half cells. Batteries made of metallic lithium anodes (Li-metal technology) are also commercially available, but for safety reasons, the Li-on technology is currently highly preferred [22]. In this last device, the positive and the negative electrodes are based on different redox-active materials, and lithium is only present in its cationic form. Testing new materials in such a full-cell condition thus requires proper lithium management. Eventually, whereas lithium is clearly the most used element, notably because of its light weight, there is also a strong interest for other elements, notably the more abundant alkaline ions (sodium [28] and potassium), as well as more charged cations (magnesium, [29, 30] calcium, zinc, aluminum [31]). Batteries based on such elements are not obtained by a simple transposition of those based on lithium, and their development usually requires the preparation of dedicated active materials, electrolytes, and so on. Regarding SCs, EDLCs are traditionally made of two identical electrodes separated by an aqueous electrolyte. Asymmetric supercapacitors, made of two different capacitive electrodes, can also be built, notably aiming at expanding the operating voltage window and hence the amount of energy stored [26]. Eventually, hybrid energy storage devices (hybrid ion capacitors) combine one faradic electrode with one capacitive electrode in the same device to improve both the energy and power densities [32, 33]. A scheme of an idealized electrochemical energy storage device is shown Fig. 4.2. It is typically built up from a stack of two electrodes stuck on current collectors (typically metal foils), separated by a separator impregnated with an electrolyte, in order to guarantee the ionic conduction while preventing the shuttling of redox-active species between the electrodes. If one looks at the electrodes themselves, they are composite materials, comprising not only the active material, optionally coated (see later), but also in most cases a carbonaceous additive to reach a suitable electronic conductivity as well as a binder to guarantee the integrity of the electrode.
116
T. Devic
current collector 1
binder
coang / SEI
electrode 1 electrolyte + separator electrode 2
carbon addive
acve material
current collector 2 Fig. 4.2 Schematized architecture of a rechargeable battery. SEI solid electrolyte interphase (see below)
Because of their tunable composition and structure, MOFs could be of interest in various parts of such devices. Table 4.1 summarizes which characteristics (composition, structure, and porosity) should be tuned to access to a specific property and where this property could be of interest in a charge storage device. As a (very obvious) example, a porous MOF of a suitable chemical composition could present a reversible redox activity and is thus of potential interest as an active electrode material. Alternatively, a redox inactive MOF presenting small polar pores facilitating ion diffusion while preventing solvent adsorption could be of interest as a protecting coating on a given active material, acting as an artificial passivation layer. Although MOFs could appear as ideal candidates for applications in the field of electrochemical energy storage, they suffer from two important drawbacks: (i) Their moderate chemical stability, which is known to be very composition and structure dependent [34–36]. Here, materials should obviously be stable in working conditions, i.e., in the presence of the electrolyte (which could be a strongly basic aqueous medium for SCs) on the whole operating voltage window. (ii) A weak electronic conductivity, which could also be managed again by playing on the composition and structure of the MOF [37]. In the next sections, by focusing on selected examples, we will discuss each of these potential applications in the battery field, deciphering when available the mechanisms and the exact nature of the materials in play (MOF vs. degradation products), with the aim at drawing when possible structure-properties relationships. Note here that measured performances are very condition dependent and that battery evaluation is a multiparameter task (electrolyte, potential window, current density, electrode content, thickness, density, etc.); comparing data from different articles is thus often of limited interest. The reader interested in precise numerical performances should refer to the original articles.
FeIII/II (XAS) FeIII/FeII
FeIII/FeII
FeIII/FeII (XAS, magnetometry) C¼O/C–O
FeIII(OH,F)(BDC)
VIVO(BDC)
FeIII3OX(H2O)2 (BDC)6
FeIII3OX(H2O)2 (BTC)4
Li1.4FeII6.8[CH2(PO3)2]3 [CH2(PO3)(PO3H)]
MIL-68 (Fe) MIL-47 (V) MIL-101 (Fe)
MIL-100 (Fe)
Cd(ClO4)2(DPNDI)2
C¼O/C–O (IR)
VIV/III
FeIII(OH,F)(BDC)
MIL-53 (Fe)
Cd(NO3)2(DPNDI)
FeIII/II (Mössbauer, XAS [46]) FeIII/II
Formula
Label
Redox couple in play (exp. evidence)
47–37 (93)
42–32 (82)
85–128 (168)
(93)
70 (108) 60–7 (108)
82 (116)
30 (113)
70 mAh (113)
Exp. capacity (theo.)
Table 4.2 Selected MOF materials used as active materials in half cells
1.8–3.4
1.8–3.4
1.5–4.5
1.5–4.0
2–3.5 2.0–4.2
1.5–4
1.5–3.5
1.5–3.5
Potential range (V)
1 M LiPF6 in a 1:1:1 EC: EMC:DMC 1 M LiPF6 in a 1:1:1 EC: EMC:DMC
LP30 1 M LiPF6 in 1:2 EC:DMC Various, best results with 1 M NaClO4 in 1:1:1 EC: PC:DME LP30
LP30
LP30
LP30
Electrolyte
50
50
200
30
Irrev 100
50
12
50
Nb of cycles
Li
100 mA g1
Li
Li
20 mA g1
100 mA g1
Na
Li Li
Li
Li
Li
Counter electrode
C/10
C/10–C/60 C/5
C/12
C/10–C/40
C/10–C/50
Rate
Refs.
[52]
[45]
[42]
[44] [43]
[49]
[41]
[2]
(continued)
45/45(SP)/10(PVDF)
45/45(SP)/10(PVDF)
50:40(CB):10(PVDF)
60/25(SP)/15 (hexafluoropropylenevinylidene fluoride copolymer)
65/30(ACB)/5 (PVDF) 70/30(SP)/0 33/33(KB)/33(PTFE)
70/30(KB)/0
85/15(SP)/0
Composition (wt%) of the electrode active material/carbon additive/binder
Zn0/ZnLi
ZnII/Zn0/ZnLi (TEM) MII/M0 (XAS) R(CO2)22/R (CO2)2n, n > 2 (NMR) CuII/I (XAS) RC¼O/RC– O
Zn4O2(DAnT)3
Zn4O(BTB)2
Zn3(HCO2)6
ZnDAnT
MOF177
FOR-1
CuAQDC
N+/N
Formula
Label
Cu(2,7-AQDC)
M(BDC) (Ni, Fe)
Redox couple in play (exp. evidence)
Table 4.2 (continued)
147–100 (162)
1100
560
1st charge: 112.8, first disch 59.3, second charge 77.1, 100th 40.1 (63.1) 105 (190)
Exp. capacity (theo.)
1.7–4.0
0–3.0
0.005–3.0
0.05–1.6
2.5–4
Potential range (V)
LP30
LP30
1 M LiPF6 in EC:PC:DEC 3:1:1. LP30
Electrolyte
50
30
60
50
100
Nb of cycles
Li
30 mA g1
Li
Li
Li
Li
Counter electrode
0.11C
50 mA g1
100–1000 mA g1 (~16C)
Rate
10/70(CB)/20(PVDF)
40/30(SP)/10(CMC)
70/15(SP)/15(Kynar)
85/10(AB)/5(PTFE)
70/20(SP)/10(PVDF)
Composition (wt%) of the electrode active material/carbon additive/binder
[72]
[57]
[55]
[1]
[53]
Refs.
CuCAT-1
MnAQDC
FeIII/II (Mössbauer)
CuII/I (XPS) CuII/I and C¼O/C–O (XPS)
CuII3(HHTP)2
TCNQ2// 0 (XPS, XRD)
TCNQ2// 0 (IR, UV–vis) CuII/I (XPS) TCNQ2// 0 (IR)
MnIII/II (XAS) RC¼O/RC– O CuII/I (XPS)
FeII2(DOPBDC)
CuTCNQ
Mn7(2,7-AQDC)6 (2,6-AQDC)
95–80 215–120
1.7–3.5 0.5–1.3
2–3.6
~90
2.4–4.4
225–205
2.4–3.9 and 3.4–3.9 1.8–4
2.8–4.3
~145
120–170
2–4.1
2–4.1
1.3–4.5
~170
255–214 (300)
190(205)
0.6 M NaPF6 in 3:7 (EC/DMC) LP30 0.25 M Zn (CF3SO3)2 in H2O
Various
1 M KPF6 In 1:1 EC:PC 1–7 M LiClO4 in 1:1 EC:PC 1 M LiClO4 in 1:1 EC:EC or EC:PC LP30
1 M NaClO4 in 1:1 EC:PC
LP30
5000
50
50
50
50
50
50
50
Li
Li
50 mA g1
20 mA g1
C/2-20C 50–4000 (18 C) mA g1
C/20-2C
Li Zn
Li Na Na
Li
30 mA g1
C/60
K
Na
Li
50 mA g1
20 mA g1
1 mA (30 mg of electrode)
80/0/20(PVDF) 60/20(AB)/20(PVDF) and 90/5/5
60/30(SP)/10(PVDF)
60/30(SP)/10(PVDF)
75/12.5(SP)/12.5 (PVDF)
60/30(SP)/10(PVDF)
PAN/MOF composite
60/30(SP)/10(CMC)
60/30(SP)/10(PVDF)
10/70(CB)/20(PVDF)
[90]
[88]
[88]
[81]
[80]
[79]
[78]
[77]
[73]
120
4.2
T. Devic
MOF as Active Materials
The use of a porous coordination network to electrochemically store charges was proposed as soon as 2002 by Huang et al. [38], but, as mentioned in the introduction, the first experimental evidences came only a few years later [1, 2]. The section is divided into three parts depending on the mechanism in play: 1. High-potential (cathode) materials, whose activity typically relies on insertion mechanism 2. Low-potential (anode) materials, whose activity usually relies on conversion or alloying reaction 3. Multiredox materials combining redox-active inorganic cations and organic ligands Electrochemical evaluation was typically carried out either in Swagelok or coin cells. For each material discussed in this section, experimental details (nature of the electrolyte, potential window, etc.) are summarized in Table 4.2. Please note that the frameworks built up from oxalate, which can be considered as inorganic polyanionic materials [39], as well as Prussian blue analogues, which are promising active materials notably for aqueous batteries, will not be discussed here.
4.2.1
High-Potential Materials: Insertion Mechanism
Initial reports focused on the use of MOFs built up from redox-active inorganic ions such as Fe(III/II) and V(V/IV/III), which are also found conventional inorganic active materials. Here, the main advantage of MOFs relates to their porosity, which should allow the fast diffusion of ion within the structure during the oxidation/reduction process. Although cavities are typically larger in MOFs than in inorganic solids, size limitation could also occur within microporous MOFs. Such steric effects were nicely demonstrated by Saouma et al. when studying the chemical reduction of the Ti MOF TiIV8O8(OH)4(BDC)6 or MIL-125 [40]. Whereas this solid presents a high porosity (SBET ~ 1500 m2 g1), the complete reduction of all Ti(IV) ions to Ti(III) by M(Cp*)2 (M ¼ Co, Cr), which are bulky reducing agents, requires the presence of the smaller Na+ cations to enter the pore and balance the charge. The crystal structures of few representative MOFs used as positive electrodes are shown Fig. 4.3. This includes the two polymorphs MIL-53 and MIL-68 formulated FeIII(OH,F)(BDC), which are built up of identical chains of corner-sharing FeO6 octahedrons connected through the ligands to define 1D channels (lozenge shaped vs. triangular and hexagonal shaped in MIL-53 [2] and MIL-68 [41], respectively; see Fig. 4.3a, b), the mesoporous MIL-100 [42] or FeIII3OX (H2O)2(BTC)4 and MIL-101 [43, 44] or FeIII3OX(H2O)2(BDC)6 (X ¼ OH, Cl, F),
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
121
Fig. 4.3 Crystal structure of few representative Fe-based redox-active MOFs. (a) MIL-53 and (b) MIL-68, both formulated FeIII(OH)(BDC); (c) hybrid supertetrahedron (top) found in MIL-100 or FeIII3OX(H2O)2(BTC)4 (left) and MIL-101 or FeIII3OX(H2O)2(BDC)6 (X ¼ OH, Cl, F) (right) and their assembly to define a MTN network (bottom); (d) Li1.4FeII6.8[CH2(PO3)2]3[CH2(PO3)(PO3H)]
which are built up from hybrid supertetrahedra which assemble to define zeotypic MTN networks (Fig. 4.3c), and an iron(II) diphosphonate formulated Li1.4FeII6.8[CH2(PO3)2]3[CH2(PO3)(PO3H)] [45]. The electrochemical activity of MIL-53(Fe) (in its anhydrous form) vs. Li was evaluated in a Swagelok cell, using a standard LP 30 electrolyte and carbon SP as a conducting additive. This solid is able to reversibly insert up to 0.6 Li/Fe at a rather constant potential of ca. 3 V vs. Li+/Li. Mössbauer spectroscopy (and later XAS experiments [46]) unambiguously confirmed that such a behavior is associated with the reduction of FeIII to FeII. The polymorph MIL-68 presents a similar behavior, except that in this case only 0.35 Li/Fe could be inserted reversibly. This result appears rather surprising, as the accessible pore volume is a priori higher in MIL-68 (Fe) than in MIL-53(Fe), but suggests that the limitation of capacity does not relate to steric effects. Extensive electronic structure calculations carried out on MIL-53 (Fe) indeed reveal that its limited capacity is directly related to the structure of the chain-like inorganic motif: the insertion of 0.5 Li/Fe leads to a stable state with localized Fe(III) and Fe(II) site, while further reduction is accompanied by a strong
122
T. Devic
weakening of the ligand-Fe interaction and hence a compete destabilization of the framework leading to an irreversible conversion reaction [47]. The vanadium analogue of MIL-53 was also investigated. This solid labeled MIL-47(V) and formulated VIV(O)BDC is prepared in the V(IV) form, the charge being balanced by the presence of oxygen rather than hydroxo inorganic bridges [48]. This solid again reversibly inserts up to 0.7 Li/V at an averaged slightly lower potential of 2.7 V vs. Li+/Li [49]. When switching to other Fe(III) carboxylates, especially those based on molecular rather than extended inorganic secondary building units, the results are more contrasted. The electrochemical behavior of MIL-101(Fe) was first investigated by Shin et al. [44]. This solid was able to insert up to 0.6 Li/Fe, but the capacity dropped down within few cycles. In situ XRD as well as in and ex situ XAS analyses revealed that such a loss of capacity is not associated with the degradation of the framework but to the instability of the mixed-valence reduced state, which naturally evolves with time to a FeIII form, while Li+ ions remain stuck in the pores. The rapid capacity fading was then attributed to a loss of Li insertion sites. Working with the same material, but in another electrolyte and in the presence of a larger amount of carbon additive, Yamada et al. were nevertheless able to maintain capacity of 0.6–0.7 Li/Fe over 100 cycles [43]. The properties of the structurally related MIL-100(Fe) were also investigated, this time against Na. The initial capacity and capacity retention were found to strongly depend on the nature of the electrolyte (salt and solvent). In the optimized conditions (1 M NaClO4 in 1:1:1 EC:PC:DME), up to 0.6 Na/Fe were inserted at the first cycle. Nevertheless, upon cycling a huge polarization was observed (>1 V), together with a rapid capacity fading (~50% of the initial capacity remaining after 30 cycles). Surprisingly, the best performances were observed when working with the hydrated form of the material, although water is known to be usually involved in parasitic side reactions leading to degraded performances. PDF and XRD analyses suggest that such a loss is again not related to the degradation of the material but to the formation of an additional, amorphous organic/carbonaceous phase in the electrode, likely located within the pores and hence impeding the diffusion of Na+ ions. Eventually, the redox properties of the small pore Fe phosphonate Li1.4FeII6.8[CH2(PO3)2]3[CH2(PO3)(PO3H)] were evaluated starting this time mainly from the reduced state, as confirmed by Mössbauer spectroscopy. Upon cycling up to 4.5 V vs. Li, the capacity slowly increased up to 0.6 Li/Fe after 200 cycles. The oxidation occurred at a higher potential than Fe carboxylates (~3.54 V), but a strong hysteresis is observed (reduction below 3.2 V). The authors suggest that the partial reduction is associated with either (i) the inactivity of one of the three crystallographically independent Fe sites, (ii) the lack of electrical connection of some particles, or (iii) the lower Li content of the pristine solid [45]. In all these solids, only a portion of the redox-active cations was found to be electrochemically accessible (and not always reversibly). Considering the size of the pores, far larger than the one of standard inorganic materials, this is clearly not related to a lack of space for ion diffusion (except if by-product accumulate within
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
123
the porosity; see above). The lack of electronic conductivity, leading to a poor accessibility of the redox centers, is likely one of the major issues. Authors mainly handled this issue by using a large amount of conductive carbon additive (see Table 4.2). It appears that the best results (in terms of reversibility and polarization) were obtained with the materials built up from infinite (MIL-47(V), MIL-53(Fe), MIL-68(Fe), Fe diphosphonate) rather than molecular inorganic building units (MIL-100(Fe), MIL-101(Fe)). Whereas electron transport through carboxylatebased cation-ligand bonds is very unlikely, infinite inorganic motifs are prone to (moderately) conduct electrons, at least in their mixed-valence state, as proven by impedance spectroscopy for MIL-47(V) at different stages of oxidation [50]. Nevertheless, even if one is able to electrochemically access to all the redox centers, these solids will still poorly compete with standard inorganic materials, their low density and large “dead” weight (coming from the ligand) leading to moderate theoretical volumetric and gravimetric capacities, respectively. The activity of few MOFs built up from redox-active ligands was also investigated. Depending on the nature of the redox center, such ligands could act either as n- or p-type materials [51]. For n-type materials, the redox process involves a neutral (oxidized) and an anionic (reduced) state, their activity thus relies on cation migration (typically Li+). On the opposite, p-type materials involve a neutral (reduced) and a cationic (oxidized) states and hence require anion migration. Typical examples of n- and p-type systems are the phenolate/quinone and arylamine/arylammonium couples, respectively. Two representative examples of MOFs based on such ligands are discussed below. The structure of these ligands as well as their expected redox activity is shown Fig. 4.4. Upon reacting N,N0 -di(4-pyridyl)-1,4,5,8naphthalenediimide (DPNDI) with Cd(II) and Co(II) salts, various 1D and 2D coordination polymers were first produced, which were evaluated against Li in coin cells [52]. Under galvanostatic conditions, a plateau centered at ca. 2.5 V vs. Li+/Li was observed during the first cycle, as expected from DPNDI
Fig. 4.4 Redox-active ligands incorporated in MOFs and their expected redox activity. The initial state (as found in the as-synthesized MOF) is shown on the left. Top, N,N0 -di(4-pyridyl)-1,4,5,8naphthalenediimide (DPNDI); bottom, 2,5-(dianilino)terephthalate (DAnT)
124
T. Devic
core, but the retention of capacity was found to strongly depend on the chemical composition of the MOF. Whereas the solids made of Co(II) suffer from a rapid capacity fading, the ones built up from Cd(II) exhibited about half of the theoretical capacity during the first cycle and more than 75% of this capacity after 50 cycles. The authors attributed this phenomenon to the higher chemical stability of the Cd (II)-based materials. Cyclic voltammetry revealed that only two out of the four C¼O were involved in the reversible redox process, as already observed with other naphthalenediimide derivatives, and IR and XP spectroscopies confirmed that the redox activity was exclusively related to the ligand. The authors suggested that this activity is associated with the insertion of Li+, although, considering the formula of the materials (Cd(NO3)2(DNPI) and Cd(ClO4)2(DNPI)2), the release of anions upon reduction could not be totally ruled out. Regarding p-type systems, 2,5-(dianilino)terephthalic acid (H2DANT; see Fig. 4.4) was reacted with Zn(II) in DMF to afford a 3-D pcu MOF formulated Zn4O2(DAnT)3 and presenting open channels that shrink upon solvent departure [53]. Its redox activity was again evaluated against Li. After a first oxidation associated with a strong irreversibility (attributed to the irreversible insertion of some PF6 anion, the gradual collapse of the MOF together with some decomposition of the electrolyte), the solid was able to reversibly store ca. 60 mAh g1, a value close to the theoretical one (63 mAh g1). This occurred at an average potential of 3.5 V vs. Li+/Li, a value higher than the one commonly observed with redox-active MOFs. Two thirds of this capacity was maintained after 100 cycles a 1.6 C; even at a very high rate (16 C), half of the capacity was retained. XPS analysis confirmed that the redox process involved the ligand and not Zn(II) and suggested that it is accompanied by the migration of PF6 anions, although both Li+ and PF6 were found to be present in the solid independently of its redox state, possibly in the form of the neutral LiPF6 entity. Although this MOF was found to partially degrade upon cycling, the capacity retention was better than the one measured for the salt Li2DANT, likely because of its lower solubility. The authors noticed that switching to a more stable MOF should increase the capacity retention, but no other report involving this ligand is currently available.
4.2.2
Low-Potential Materials: Conversion and Alloying
When MOFs are evaluated as low-potential active materials (i.e., to be ultimately used as negative electrodes in Li-ion batteries), the redox process involves at least the reduction of the metallic cation to its neutral state, a phenomenon which is not compatible with the retention of the crystalline structure. The system must thus evolve from monophasic to (at least) biphasic (the metal and a lithium salt of the ligand). The reversibility of the redox process then strongly depends on the nanostructuration of this reduced biphasic mixture, which could sometimes be tentatively related to the initial crystal structure of the MOF. Here, almost all chemical compositions might be suitable; as a consequence, the number of reports dealing with MOFs as anode materials is far higher than the one related to cathode
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
125
materials. We will here discuss few selected examples; more exhaustive listing can be found in dedicated reviews [3, 4, 8, 54]. As mentioned in the introduction, the first report dealt with MOF-177 or Zn4O (BTB)3, which exhibits a very high surface area (>4000 m2.g1). This material was cycled against Li between 1.6 and 0.05 V in its solvated form (i.e., with N,Ndiethylformamide (DEF) and water molecules entrapped in the pores) [1]. Particles of different shapes were evaluated, without any significant effect on the electrochemical performances. A capacity of ca. 400 mAh g1 was obtained during the first discharge but dropped down to 100 mAh g1 during the first charge and then remained constant for at least 50 cycles. Such an initial capacity loss was attributed to the irreversible reduction of the solvent molecules entrapped in the pores. The first reduction is also accompanied with the complete degradation of the framework and formation of Zn(0) nanoparticles, as confirmed by TEM and XPS analyses. It is then proposed that the reversible activity found after the first cycle is associated with the allowing reaction Zn + Li+ + e $ LiZn. Such an approach was then extended to FOR-1 or Zn3(HCO2)6, which is based on formate ligands [55]. Here, upon cycling between 0.005 and 3 V, after again a high capacity loss during the first cycle, a reversible capacity of ca. 560 mAh g1 (close to the theoretical value) was obtained. Based on IR, TEM, and XRD analyses, the authors suggested the occurrence of two reversible redox reactions: A conversion reaction : Zn3 ðHCO2 Þ6 þ 6Liþ þ 6e $ 3Zn þ 6LiðHCO2 Þ Followed by an alloying reaction : 3Zn þ 3Liþ þ 3e $ 3LiZn Hence, when compared to the previous example, when working with a larger potential range (here 0.005–3 V vs. 0.05–1.6 V in Ref. [1]) and a smaller ligand, the conversion reaction was found to be reversible. The authors specifically highlighted that the formation of Li formate as an intermediate (rather than the conventional Li2O) was likely at the origin of such a good reversibility, as lithium formate could be more easily converted back to zinc formate. Indeed, although powder XRD indicated that the first reduction and further reoxidation lead to a bulk amorphization, crystalline particles were still detected by TEM after oxidation, and their crystallographic parameters matched with the ones of the initial zinc formate. Similar experiments carried out on the Co(II) analogue, for which conversion is possible, but not alloying, ultimately confirmed this two-step mechanism. Hence, this study revealed that a judicious choice of ligand could lead to reversible conversion reactions with MOF materials. On another side, it was found that the lithium salts of few conjugated dicarboxylate ligands were able to reversibly store two electrons at low potential. As an example, lithium terephthalate formulated Li2(BDC) can be reversibly reduced at about 1 V vs. Li+/Li with a capacity of 250–230 mAh g1 (see the redox process Fig. 4.5) [56].
126
T. Devic
Fig. 4.5 Reversible reduction of Li2BDC occurring at ca. 1 V vs. Li+/Li
Hence, considering that terephthalate is a very common organic building block in MOF chemistry, the low-potential electrochemical activity of metal terephthalates was evaluated, with the aim at combining conversion reaction with such an organic reduction to achieve high capacities. The initial report by Lee et al. focused on Ni and Fe terephthalates [57]. Their activity vs. Li was evaluated between 0 and 3.0 V. As commonly observed with low-potential materials (see above), a high irreversible capacity loss is observed during the first cycle, associated with the irreversible consumption of lithium and the formation of the solid-electrolyte interface (SEI); nevertheless, after five cycles, a reversible capacity of about 1100 mAh g1 was obtained, corresponding to about 10 electrons per M(BDC) motif. XANES experiments were carried out at various potentials on the Ni derivative and revealed that the reduction of Ni(II) to Ni(0) occurred within the range 1–1.5 V vs. Li+/Li but is not sufficient to explain the full capacity. Indeed, the reduction of the terephthalate ligand was also detected by solid-state 13C NMR. To account for the unusually high capacity, the authors proposed that such a reduction did not simply involve two electrons (as shown Fig. 4.5) but that a higher reduction state involving eight electrons was reached. This approach was then extended to other 3D transition metal cations (Mn [58], Zn [59], Ti [60]), p-block elements (Sn [61]), other conjugated polycarboxylate (trimesate [62], thiophenedicarboxylate [63]), mixed pyridine N-oxide/carboxylate [64], croconate [54], and even diphosphonate [65] ligands. In few cases, the role of the initial structuration of the pristine MOF was highlighted: notably, nanosheets of coordination networks were found to be more efficient than bulk materials, especially at high rates [58, 63]. Eventually, the same strategy was also applied vs. Na [66] and K [67, 68]. Although impressive reversible capacities were obtained with these systems, two critical points must be taken into account: (i) The carbon additive (typically Csp) which is often added in a large amount (see Table 4.2) to afford a decent electronic conductivity has a significant contribution at low potential; this should be taken into account when determining the capacity associated with the MOF itself [57, 67, 68]. (ii) To compensate for the absence of clear plateaus on the charge and discharge curves and the high hysteresis between both processes, such materials are typically cycled on a very large potential range (typically 0.05–2.5–3 V); this is incompatible with their use in practical Li-ion full cells in front of “real” cathode materials.
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
4.2.3
127
Combining Organic and Inorganic Redox Activity: From Redox-Active Core to Non-innocent Ligands
Obviously, one of the advantages of MOFs when compared with other materials is the possibility to exploit their hybrid nature by combining both organic and inorganic redox activities or, in other words, combine cationic and anionic redox, as found in some specific metallic oxides [69]. This might be indeed the way to compete with standard electrode materials, at least in terms of gravimetric capacity. The aim here is commonly to rely on insertion mechanisms, which could request both cation and anion migration from the material to the electrolyte (and the opposite). Nevertheless, as discussed below, deciphering the exact mechanism in play could be challenging when multiple redox centers are involved, and partial solubilization and/or conversion should sometimes be taken into account (see the example of Cu-TNCQ below). We will here first discuss few examples of MOFs built up from ligands made of a redox-active core and then switch to non-innocent ligands. In the first case, the organic redox centers are electronically isolated from the metallic cations, whereas, in the second one, the frontier crystal orbitals (at least either the highest occupied or the lowest unoccupied crystal orbital) which are involved in the electron transfer are of a hybrid nature, i.e., combine contributions from the cation and the ligand (d and π orbitals, respectively). As a consequence, the “exact” oxidation state of each component is ambiguous [70] (and even sometimes meaningless). Few of the redox-active fragments found in MOFs are shown Fig. 4.6. One of the first examples deals with the tetrathiafulvalene (TTF) core. This motif could be reversibly oxidized to a fully delocalized monocationic radical state, followed by a dicationic state, and is at the origin of the field of molecular conductors (and supraconductors). The tetracarboxylate derivative TTF-TC (see Fig. 4.6a) was reacted with Ni(II) to yield to a coordination complex [Ni(H2O)4]2(TTF-TC)4H2O, which evolves upon dehydration to a 2D coordination polymer formulated Ni2(H2O)5(TTF-TC)H2O [71]. Solid-state cyclic voltammetry revealed that both solids present similar features in the 2–4 V vs. Li+/Li range, with three redox events associated with (i) the double oxidation of the TTF core and (ii) the reduction Ni (II) to Ni(I). Nevertheless, the corresponding capacities were found to be very low (10–45 mAh g1), likely because only a minor part of the redox centers was accessible, in line with the absence of porosity. In 2014, Zhang et al. reported a new MOF formulated Cu(2,7-AQDC) built up from the dicarboxylate derivative of anthraquinone (2,7-AQDC; see Fig. 4.6a) and Cu(II) paddle wheels [72]. Although 2D, this solid was found to present a significant accessible surface area (SBET ~ 630 m2 g1) with micropores of ca. 12 Å. When cycling vs. Li between 4 and 1.7 V, two plateaus centered at ca. 3 V and 2.1 V were clearly visible on both the charge and discharge curves. XAS revealed that the first reduction is associated with the stepwise reduction of the CuII2 paddle wheels to CuI2, whereas the second one is attributed to the two-electron reduction of the anthraquinone motifs. The initial capacity was found to be close to the theoretical one (147 vs. 162 mAh g1, respectively) but dropped down within few cycles to reach ~100 mAh g1 after
128
T. Devic
Fig. 4.6 Examples of redox-active motifs found in MOFs. The redox-active moieties are shown in red. (a) Isolated organic and inorganic redox centers; left, TTF-TC; right, 2,7-AQDC; (b) non-innocent ligands; left, TCNQ; right, catecholate (Q ¼ O), dithiolene (Q ¼ S), diaminobenzene (Q ¼ NH)
50 cycles. XRD analysis suggested that this capacity fading is not related to the collapse of the framework but rather to an insufficient extraction of Li+ upon reoxidation. Although this study focused on extremely poorly loaded electrodes to get rid of electronic transport limitation (the electrode is here composed of 10 and 70 and 20 wt% of MOF, carbon additive, and binder, respectively), this could be considered as the first evidence of a truly hybrid redox activity within an MOF. The activity was here likely related solely to lithium migration. Using a mixture of anthraquinone dicarboxylate isomers and Mn(II), another solid formulated Mn7(2,7-AQDC)6(2,6-AQDC) was reported by the same group [73]. The activity of this solid against Li was evaluated between 4.5 and 1.3 V vs. Li+/Li, using again a large amount of carbon additive. Upon discharging (i.e., reducing), a single plateau centered at ~2.4 V was observed and attributed again to the two-electron reduction of the anthraquinone. When charging, a second event was detected at ca. 3.6 V. This later is associated with the reversible oxidation of Mn(II) to Mn(III), as confirmed by XANES experiments. While the reduction of the ligand is again associated with Li+ migration, the oxidation of Mn(II) requests the insertion of PF6 anion, as suggested by XRD and solid-state 19F NMR analyses. Hence, the redox processes occurring in both the Cu- and Mn- based MOFs can be summarized as follows (initial state in bold format): CuII(2,7-AQDC)
+e-, +Li+ -e-, -Li+
LiCuI(2,7-AQDC)
MnIII7(2,7-AQDC)6(2,6-AQDC)(PF6)7
+7e-, -7PF6-7e-, +7PF6-
+2e-, +2Li+ -2e-, -2Li+
Li3CuI(2,7-AQDCred)
MnII7(2,7-AQDC)6(2,6-AQDC)
+14e-, +14Li+ -14e-, -14Li+
Li14MnII7(2,7-AQDCred)6(2,6-AQDCred)
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
129
For the Mn derivative, a capacity close to the theoretical one (205 vs. 190 mAh g1, respectively) was observed and remained almost constant for 50 cycles when the current was fixed to 1 mA (for ~30 mg of electrode). Upon increasing the current five times, 70% of the capacity was maintained. Interestingly, even at high rate, the two plateaus were still discernable on the discharge curve, suggesting that PF6 diffusion was not significantly slower than Li+ diffusion. A mentioned above, one of the key limitations of the aforementioned MOFs is their poor electrical conductivity. To solve such a matter, non-innocent ligands are particularly appealing. Apart from their multiredox activity, the electronic delocalization between the ligand and the metallic cation could expand through the whole framework, thus offering a suitable electron transport path. Indeed, most of the semiconductive MOFs reported to date (as well as the very rare examples of metallic [74] and even superconducting [75] MOFs) are made of non-innocent ligands [37]. Because of this non-innocence, deciphering accurately their redox behavior is often challenging. We will here discuss two sets of non-innocent ligands: tetracyanoquinodimethane or TNCQ, which can be reversibly reduced twice, and catecholate and the related dithiolene and diaminobenzene motifs, which can be oxidized twice (Fig. 4.6b). TCNQ is an electron donor that has been extensively used in the field of molecular conductors, notably because of its fully delocalized, stable anionic radical state. It has also been considered as a suitable ligand, even before the rise of the MOF era. Indeed, two polymorphs formulated Cu(TCNQ) were reported as soon as 1999 [76]; both extended coordination networks are built up from Cu(I) and the anion radical TCNQ. One of them presents a rather high electrical conductivity (0.25 and 1.3 105 S cm1 for polymorphs I and II, respectively), associated with short interligand π–π contacts (3.24 Å). The redox activity of polymorph II against Na was first evaluated by Fang et al. in 2017 [77]. Upon cycling between 2 and 4.1 V at 50 mA g1, an impressive capacity of 214 mAh g1 was measured, with a limited capacity fading up to 50 cycles. This capacity corresponds almost to three electrons exchanged per Cu(TCNQ) and is associated with a complex redox behavior, which was studied by a set of spectroscopies (IR, UV-vis, XPS). The following redox pathway was proposed: the initial oxidation of the material leads to its complete transformation to TCNQ0 and Cu(I), this later being oxidized to Cu(II) above 3.8 V. These two steps are reversible, and upon reduction, CuI(TCNQ) is reformed (likely polymorph I) and further reversibly reduced to NaCuI(TCNQ2). TCNQx entities were found to solubilize in the electrolyte; hence, adapting a strategy developed for Li-S batteries, a super P carbon interlayer was deposited between the cathode and the separator to prevent redox shuttling of the soluble species and dramatic capacity fading. Upon narrowing the potential window to 2.5–4.1 V, a capacity retention of ca. 90% after 200 cycles was reached, suggesting that the formation of TNCQ2 should be avoided to maintain a high cyclability. Focusing directly on the most conductive polymorph I, similar results were obtained against potassium [78] as well as lithium [79–81]. In the latter case, further improvements of the capacity retention were obtained in half cells either by preventing even more the redox shuttling through the use of a graphene oxide-modified separator [80] or by decreasing the
130
T. Devic
solubility of the molecular species through the use of very concentrated electrolytes (typically 7 M LiClO4 in a mixture of EC and PC) [79]. By a combination of XRD and XPS studies, it is further proven that, at least when working with a LiPF6containing electrolyte, the redox event occurring at high potential (>3.85 V vs. Li+/ Li) is associated with the reversible oxidation of TNCQ to TCNQ0 and the formation of a new crystalline phase, likely containing CuI, PF6, and TCNQ0 [81]. Eventually, upon cycling in the 0.01–3.0 V range against Li, the potential of such a material as an anode material was also evaluated [82]. As expected, the redox activity here relies of the most reduced states of the constituents, namely, the TCNQ/TCNQ2 and CuI/Cu0 redox couples. The second class of non-innocent ligands is the phenolates (especially catecholates) and S and NH analogues. Here, properly describing even the as-synthesized form of the derived MOFs (i.e., the exact oxidation state of the ligands and cations, the charge of the network with the possible presence of counterions, etc.) is often challenging [83, 84], notably because these solids are commonly produced with a poor crystallinity, preventing any accurate structure determination by standard XR diffraction techniques. Proper characterization then requests additional structural information gained from various spectroscopies (XAS, Raman, IR, solid-state NMR, Mössbauer, etc.). Such ligands (at least the catechol (ate) ones) could moreover bear protons even when coordinated [85–87]. These latter are not localized by XRD techniques; their identification again requests additional characterization tools (e.g., IR or solid-state NMR spectroscopies, neutron diffraction). A first report by Aubrey et al. focused on an expanded version of CPO-27, built up from a bis-salicylate (i.e., o-hydroxycarboxylate) and Fe(II) and formulated FeII2(DOBPDC) (Fig. 4.7) [88]. As this solid is prepared in a neutral, fully reduced state (both the ligand and metallic cation), it is anticipated that charge-discharge cycling will request anion migration; the very large channels found in this structure (~20 Å diameter; see Fig. 4.7) appear particularly appealing. Under galvanostatic oxidation up to 4 V vs. Li, a linear increase of the potential was observed, in line with a process based on the formation of a solid solution rather than biphasic; up to two electrons per formula unit could be reversibly extracted. Mössbauer analysis indicated that this process is related to the inorganic center only (oxidation of Fe(II) to Fe (III)). The average potential is close to 3.5 V, i.e., significantly higher than the one found in other Fe(II) materials, possibly because the process relies on anion insertion rather than cation departure. When increasing the potential, a second event, located this time on the ligand, was observed at 4.1 V but was found to be irreversible. Focusing on the inorganic redox activity, the authors found that the electrolyte has a strong impact on the electrochemical performances. When the size of the anion increases (from BF4 to TFSI and BArF), the maximal capacity decreases, but at the same time the hysteresis between the charge and discharge curve decreases. While the first effect is likely associated with the steric hindrance of the anions within the pores, the second one suggests that the bulkier anions diffuse more easily
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
131
Fig. 4.7 MOFs based on non-innocent, phenolate-derived ligands. Left, FeII2(DOPBDC); right, Cu3(HHTP)2. The initial redox state of the ligand is shown at the bottom
within the pores because of weaker electrostatic interactions with the Fe(III/II) centers. The cation was also found to affect the capacity, suggesting that the reduction of the material could be associated not only with anion release but also with cation insertion. The nature of the solvent also impacts on the capacity, in line with its adsorption within the pore and potential interaction with coordinatively unsaturated Fe(III/II) sites. Eventually, when switching from a Li to Na counter electrode and cycling between 2 and 3.6 V, a reversible capacity of ca. 80 mAh g1 was maintained up to 50 cycles, again likely associated solely with an inorganic centered redox process. Further examples focused on catecholate ligands. The structure of Cu-CAT or Cu3(HHTP)2 was first reported by Hmadeh et al. in 2012. This solid is built up from tricatecholate ligands, which are connected to isolated octahedral Cu(II) centers to afford a 2D honeycomb network (Fig. 4.7). In its as-synthesized form, Cu adopts a + II oxidation state, while the ligand is partially oxidized to a radical form and presents a net charge of 3. In this report, the authors briefly mentioned that this solid exhibited a high electronic conductivity (~0.2 S cm1) and is able to deliver a capacity of 80 mAh g1, but no further details were given [89]. Later on, Gu et al. showed that this solid could be cycled between 1.7 and 3.5 V against Li and presents a maximum capacity of 95 mAh g1 (about one electron per Cu) and that 50% of this capacity was retained after 500 cycles [90]. Interestingly, no carbon additive is needed. XPS revealed that the redox activity is associated with the reduction of Cu (II) to Cu(I). Further experiments showed that this solid could be cycled even at very high rate (20 C); nevertheless, considering that this solid and related catecholate MOFs are also suitable in supercapacitive electrodes [91–93], the existence of a
132
T. Devic
capacitive contribution is likely. The redox behavior of this solid down to 0.01 V vs. Li+/Li was further evaluated by Guo et al. Surprisingly [94], the authors did not notice any reduction of Cu(II) and suggest that Li insertion is here associated with the reduction of the C6 aromatic cycles. This solid was further evaluated in a Zn battery in aqueous medium [95]. Again, upon cycling between 0.5 and 1.3 V vs. Zn2 + /Zn at 50 mA g1, a reversible capacity above 200 mAh g1 is obtained with a good capacity retention for at least 30 cycles. Upon increasing the rate up to 20 times, more than 50% of the capacity is maintained for at least 100 cycles. On the opposite, no reversible activity is found in an organic (acetonitrile) electrolyte. XPS analysis showed that the charges were stored both by the inorganic (Cu(II/I)) and organic (quinolate/phenolate) redox centers, while XRD confirmed that the structure is maintained. Both XPS and microscopy EDX confirmed that the reduction was associated with the bulk insertion of Zn2+ cations within the MOF crystallites, while further electrochemical analysis revealed that the charge storage is mainly of a capacitive nature, suggesting that this material acts as a pseudocapacitor. While only a single example MOF was presented here, it should be noted that other MOFs built up from different polycatecholate ligands were also investigated [96–99]. Interestingly, the electrochemical reduction of such solids was found to be a suitable method to tune their magnetic properties [96, 99]. The same approach was applied to closely related ligands, namely, bisdithiolenes [84] as well as tetra- and hexa-aminobenzenes [83, 100–102] vs. either Li [100, 102] or Na [83, 84, 101]. Depending on the nature of the ligand and cation, the redox processes were found to be either solely ligand centered [84, 101] or were taking place both on the organic and inorganic centers [83, 100]. It should be noted that the initial oxidation state could strongly depend on the synthesis conditions [84] and the nature of the constituents; hence, charge storage could occur either in oxidation or in reduction and request cation or anion migration. The presence or remaining acidic proton in aminobenzene-based MOFs was found to lead to side reaction and irreversible capacity losses [102]; hence, catecholate and dithiolene ligands might be more suitable for long-term storage. To conclude, non-innocent ligands – especially the anionic ones – allow producing MOFs combining a multiredox activity and a good stability in the electrolyte, together with decent electronic and ionic conductivities. They thus allow reaching high capacities even when used as cathode materials, but often on a large potential window, in line with the presence of multiple redox centers. They still deliver decent capacities even at high rates, where the capacitive contribution to charge storage becomes predominant.
4.3
MOFs as Host for Active Species
The aim here is to exploit the huge porosity of MOFs to immobilize redox-active materials, which are of potential interest (e.g., because of their high theoretical capacity) but soluble in the electrolyte in at least one of their oxidation state, and
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
133
hence, present a very rapid capacity fading when used in their native form. Ideally, MOFs should here combine (i) large pores to store large amounts of active materials and allow ion diffusion, (ii) small windows and/or strong interaction sites to prevent (or at least slow down significantly) the leaching during the electrochemical process, (iii) a decent electronic conductivity to manage charge transport, and obviously (iv) a suitable stability in the voltage range of interest. Two examples will be discussed below: – The encapsulation of redox-active molecules – The encapsulation of sulfur
4.3.1
Organic Molecules
Molecular redox organic compounds have come to light as promising and potentially “greener” electrode materials, notably because of the high abundancy and availability of their constituents (when compared to metals), as well as the tunability of their redox potential through simple organic chemistry. They nevertheless typically suffer for a high solubility in standard polar electrolytes, especially in their neutral state. The encapsulation of benzoquinone in MIL-53(Fe) was first investigated, with the aim at combining the redox activity of the framework (see above) with the one of the guest [103]. Encapsulation was carried out in solvent-free conditions by simply hand-milling the activated MOF with benzoquinone. XRD and TGA unambiguously confirmed the insertion of about 1 quinone per FeIII(OH)(BDC) unit. Upon reducing this solid down to 1.8 V vs. Li+/Li, new features are observed on the discharge curve when compared to the quinone-free material. Notably, a new plateau centered at 2.7 V is detected and attributed to the reduction of quinone to quinolate. On the whole, 1.2 electron per formula unit could be stored (vs. 0.65 in the quinone-free material), corresponding to a capacity of 93 mAh g1. This process is found to be reversible, but a very rapid capacity fading was observed; after six to eight cycles, the capacity became identical to the one of the quinone-free MIL-53(Fe). XRD and solid-state NMR revealed that this loss of capacity is associated with the solubilization of the organic guest. More precisely, it is shown that the oxidized, neutral quinone molecules are easily replaced by solvent molecules within the pores, whereas the reduced anionic quinolates remained entrapped. To the best of our knowledge, this represents the only example of such an approach; improvements will obviously request the strengthening of the host-guest interactions (at all oxidation states) to prevent leaching. Alternatively, the encapsulation of redox-active molecules larger than the pore apertures could be envisioned but will require a suitable one pot “ship-in-a-bottle” preparation method.
134
4.3.2
T. Devic
Sulfur
The Li-S battery technology (i.e., cathode ¼ S0, anode ¼ Li0) is considered as a promising alternative to the Li-ion one. Indeed, sulfur could be reduced to Li2S (at 2–2.4 V vs. Li+/Li) and thus presents a very high theoretical capacity (>1600 mAh g1). Apart from safety issues related with the use of metallic lithium metal, sulfur suffers from serious drawbacks, notably (i) the electrically insulating character of both the oxidized (S0) and reduced (Li2S) states, (ii) the large volume expansion associated with the reduction, and (ii), before the full reduction, the intermediate formation of long-chain polysulfides (S82, S62, etc.) which are soluble in standard electrolytes. All of this leads to reduced experimental capacities when compared to the theoretical value and rapid capacity fading preventing the practical development of such devices. The use of MOFs and MOF-derived materials has been proposed to solve these matters (for recent reviews on MOFs for Li-S batteries, see references [104, 105]); we will here exclusively discuss the use of native MOFs as host matrices. The first report dealt with the encapsulation of sulfur within the mesoporous MOF MIL-100(Cr) (see the crystal structure of the Fe analogue, Fig. 4.3) [106]. The S-MOF composite was prepared through a melting diffusion strategy, in which the desolvated MOF was mixed with elemental sulfur and heated at 155 C to allow the diffusion of liquid sulfur within the pores. The final composite contained ca. 50 wt% of sulfur exclusively located within the porosity (no recrystallized by-product). The BET surface area is then reduced from ~1500 to 360 m2 g1, the residual porosity being a priori able to buffer the volume expansion associated with sulfur lithiation. This solid was then cycled against Li in the 1–3 V range. Two plateaus located at 2.4 and 2 V are clearly visible on the first discharge curve; they correspond to the formation of long-chain polysulfides and Li2S, respectively. This process is found to be reversible upon oxidation. After the optimization of the amount of carbon additive (50 wt%), up to 1.85 electron per S could be exchanged. The capacity reached 800 mAh g1 at the second cycle and slowly decreases down to ca. 400 mAh g1 after 60 cycles. Such a capacity retention was far better than the one obtained by simply mixing sulfur with MIL-100(Cr) or with a sulfur-loaded mesoporous carbon. This relates to the unique porosity of MIL-100, which comprises large spherical pores (free diameter > 20 Å) connected through small windows (5–8 Å), this latter being able to reduce the leaching of polysulfides. Although not fully deciphered, XPS analysis furthermore suggested the presence of relatively strong sulfur-host interactions, which were again not found in mesoporous carbon and could also explain the better capacity retention of the MOF-based composite. Zhou et al. further studied various MOFs (ZIF-8, HKUST-1, MIL-53(Al), MIL-53(Al)-NH2) and showed that the capacity retention is mainly governed by the pore morphology, namely, the size of the pore windows and the dimensionality (1D vs. 3D) of the porosity [107]. Further studies have indeed focused on the improvement of the capacity retention through the tuning of the MOF-guest (sulfur and sulfides) interactions. The critical role of the coordinatively unsaturated metal sites (i.e., Lewis
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
135
acid sites) was confirmed by studies on Cu- [108], Ni- [109], and Mn S-loaded MOFs [110], which showed that these sites have a beneficial impact on the capacity retention, with strong polysulfide-cation interactions preventing the fast release of lithium polysulfides in the electrolyte. Apart from the metallic site, the postfunctionalization of MOFs was also found to be beneficial [111]. Focusing on the Zr terephthalate UiO-66 and its derivatives, Baumann et al. studied the effect of the functionalization of defective Zr6 oxoclusters, either through the exchange of proton by Li+ on defective OH/OH2 sites [112] or through their substitution by PS43 anions [113]. While additional Li+ ions seem to facilitate ion diffusion during the cycling process, Zr-bound PS43 anions also established covalent S-S bonds with polysulfide ions; both strategies lead to an enhancement of the capacity and capacity retention. The effect of the MOF particle size was also investigated. When downsizing ZIF-8 crystallites from 3 μm down to 150 nm, the reversible capacity of 30 wt% Sloaded ZIF-8 increased, but the capacity retention remains rather constant [107]. Further studies suggest the existence of a “golden particle size” to take into account two antagonist effects: the maximization of the use of sulfur through short internal diffusion distance requests downsizing, while such downsizing increases the external surface exposed to the electrolyte, hence facilitating the escape of polysulfides [114]. When switching back to HKUST-1, the capacity and retention capacity were improved upon downsizing [115]. Through the use of UV-visible and X-ray absorption spectroscopies, the authors showed that specific coordinatively unsaturated metal sites are present in larger amount at the surface of the smaller particles, and that these sites interact favorably with S42 and thus play a role in promoting the uptake of these anions during cycling. Eventually, recent studies also showed that it is possible to combine the electronic properties of the MOF with the one of sulfur to improve the performances: it was, for example, proposed to use a semiconductive MOF as a host (built up from a hexaiminotriphenylene ligand; see above) [116] or an MOF comprising a redoxactive anthraquinone derivative as a ligand [117]. In order to account for the poor electronic conductivity of most host MOF matrices, further improvements were proposed, which rely on the preparation of sulfur-loaded MOFs coated with conducting additives. Both carbonaceous species (graphene sheets [118, 119], carbon nanotubes [120], N- and P-doped carbon [121]) and conducting polymers (polypyrrole [122] or polythiophene [123]) were investigated. Improvements were observed mostly at high current rate and/or upon long electrochemical cycling (>100 cycles).
4.4
MOFs for as Coatings of Active Materials
The aim here is to exploit the tunable sieving properties of MOFs through their coating on inorganic active electrode materials, in order to address the three issues discussed below:
136
T. Devic
(i) Standard electrolytes are unstable toward active materials working at low potential (typically 4 V) cathode materials could present cathode-electrolyte interphase (CEI). MOFs were then proposed as artificial stable SEI/CEI, as the fine-tuning of their porosity might allow ion transport while voiding contacts between the other components of the electrolyte and the active material. (ii) In Li metal batteries, dissolution/redeposition of lithium at the anode is usually accompanied with the growth of Li0 dendrites, which can ultimately lead to electrical short circuits. This major safety issue is at the origin of the limited development of this technology when compared to the Li-ion one. Coating was then proposed to favor the homogeneous plating of lithium during the charge process. (iii) Alloy negative electrodes are often subject to huge volume variations upon electrochemical cycling. As an example, silicon, a promising anode material, presents a volume expansion of 280% between the oxidized (Si0) and reduced (Li15Si4) states [129]. This results in high mechanical strains inducing electrical disconnections, the destabilization of the SEI, and ultimately rapid capacity fading. Suitable coating might buffer such volume variations. MOF-derived coatings obtained by thermal treatments (either oxide or carbonaceous species) will not be discussed here; the reader interested in this field could refer to recent articles [124–128] and cited references.
4.4.1
Coating on Cathode Materials
To the best of our knowledge, a single example reports the use of a native MOF coating on a high potential cathode material. This could be related to the chemical nature of most of the standard MOFs, which are built up from aromatic polycarboxylate ligands: such molecules are easily electrochemically decarboxylated upon oxidation at >4 V; hence, the derived materials would not sustain such conditions. Qiao et al. reported the functionalization of the Li-rich oxide Li(Li0.17Ni0.20Co0.05Mn0.58)O2 [130]. The preparation is done as follows: the oxide is suspended in a solution containing a Mn(II) salt and H2DOBDC, heated under solvothermal conditions and further washed. The formation of an amorphous layer of 2–3 nm on the oxide particles is observed by TEM. Although the formation of a coordination network could not be univocally evidenced, the presence of Mn(II) at the surface of the particles was detected by XPS, suggesting its deposition on the surface. When cycling at low rate between 2 and 4.8 V vs. Li+/Li, a slight increase of the reversible capacity was detected. O 1 s XPS analysis carried out after the first charge revealed differences between the pristine material and the coated one; the
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
137
authors suggest that the coated layer could store gaseous oxygen which is irreversibly formed during this first step, leading to a better stability of the “MOF-modified” solid. Eventually, the analysis of the Mn 2p spectrum suggests that such a layer also participates to the redox process, in line with the increase of capacity.
4.4.2
Coating on Anode Materials
Coating of anode materials was studied more deeply. Silicon electrodes have been first investigated. The large volume variation occurring during the lithiation/ delithiation processes indeed leads to a destabilization of the SEI and ultimately a rapid capacity loss. A suitable MOF coating should allow Li+ and electron conduction while preventing the contact of Si or LixSi with the anions and solvent molecules [129]. Two methods of preparation have been proposed: the first one relies on the direct synthesis of the MOF on the silicon substrate [131, 132], and the second one on the deposition of pre-synthesized MOF nanoparticles [133]. In this last report, Han et al. prepared a sandwich Si electrode, which was obtained by casting ZIF-8 nanoparticles on a prebuilt three-layer electrode made of (1) a copper foil, (2) a carbon conducting additive, and (3) Si particles. The thickness of the MOF coating was about 50 μm; an improvement of the capacity retention was observed and attributed both to the formation of a more stable SEI and the facilitated diffusion of Li+ through the MOF. Nevertheless, this approach was efficient only for poorly loaded electrodes (0.4–0.7 mgSi cm2). This study was extended to other standard MOFs (HKUST-1, MOF-5, MIL-53) and concludes that the improvement of the capacity retention requests a pore aperture close in size to that of Li+. A similar approach, based that time on the direct synthesis of ZIF-8 from solubilized precursors, was then applied to pure Si- [131] and TiN/Ti-coated Si nanorods [132]. By playing on the experimental conditions, it was possible to grow a broad variety of coordination coatings. Electrochemical characterizations revealed that an amorphous but homogenous coating leads to a significant improvement of the capacity retention, while a highly crystalline but incomplete coating has a very limited impact [131]. This suggests that the main benefit of the MOF coating relates to its stability (when compared to natural SEI), rather than to the improved diffusion of Li+ through the micropores. A similar approach was recently applied to black phosphorus, which is another anode material of potential interest [134]. Deposition on conversion materials was also carried out. Han et al. reported the solution-based deposition of – again – ZIF-8 on ZnO and MnO2 [135]. A high amount of coating was used (10–20 wt%), allowing to detect its microporosity by N2 sorption isotherm measurements. When compared to the pristine oxides in the 0.01–3 V potential range, a similar conversion reaction was observed but with a higher initial capacity and a better capacity retention for the MOF-coated oxides. Although powder XRD analysis suggested that the structure of ZIF-8 is maintained, its participation to the redox reaction could not be ruled out. Electrochemical impedance spectroscopy (EIS) experiments suggested a facilitated charge transfer
138
T. Devic
for the coated materials, in line with the porosity. It should nevertheless be pointed that, upon pyrolyzing the MOF-coated oxides at 700 C to favor the formation of a conductive carbonaceous coating, the improvement is far more marked. MOF coating was also proposed to suppress Li dendrite growth in Li-metal batteries. Fan et al. reported the deposition on a Cu current collector of a mixed coating made of a Zn MOF and an organic polymer (PVA) [136]. While the MOF is expected to favor again Li+ diffusion and to improve the mechanical strength, the polymer should afford some flexibility preventing the formation of cracks upon cycling. This 3-μm-thick coating is electrically insulating; hence, it is expected that upon charging, Li+ ions migrate through the MOF and Li deposits on the current collector (hence behind the coating). EIS evidenced that such a coating led to the formation of a stable SEI within few cycles and that dendrite growth was suppressed, as confirmed by scanning electron microscopy. As a consequence, a high coulombic efficiency was obtained even for hundreds of cycles, not only in symmetric Li||Li cells but also with a full cell (with LiFePO4 as a cathode material). Yuan et al. proposed another approach, in which they combine the porous HKUST-1 MOF with Ag nanoparticles [137]. Whereas the MOF should afford numerous homogenous adsorption sites for Li+, the Ag nanoparticles bring suitable low-energy nucleation sites for Li as well as a good electronic conduction. With this approach, again a high coulombic efficiency (>97%) was obtained for at least 300 cycles. Note here that a recent study relates that the simple addition of MOF particles in the electrolyte (i.e., without the need of the preparation of proper coating) could also prevent dendrite growth [138]. To sum up, MOF coatings seem to be of interest, notably for anode materials, at least, but two questions remain open: – What is the exact state of these coatings at extreme potentials? Is the MOF structure (and the related microporosity) really maintained upon cycling? As the amount of MOF is usually very limited, its proper characterization is extremely challenging. – Ion transport might be facilitated by the microporosity, but, as mentioned many times, MOFs are generally electrical insulators. MOF coating could thus hamper proper charge transport; the MOF composition and thickness need probably to be optimized to truly maximize the performances of the active materials.
4.5
MOF-Based Separators
The separator in standard M-ion and M-metal batteries containing a liquid electrolyte is a critical component (see Fig. 4.2): it should obviously prevent electrical contacts between the anode and cathode while allowing the diffusion of ions from one electrode to the other one during the charge-discharge processes, and in a Li-metal battery, it could also affect dendrite growth. It ultimately prevents also the redox shuttling, i.e., the migration of redox-active species between the electrodes. These
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
139
species could be degradation products formed in the course of the cycling but also, in some emerging battery technologies, be an important component of the electrochemical storage process: this is the case of intermediate polysulfides in Li-S batteries, redox mediators in Li-O2 batteries, and the anolyte and catholyte in redox flow batteries. Standard separators are made of polymeric films or interwoven fibers. The aim here is to build MOF membrane filled with a liquid electrolyte and exploit the sieving properties of MOF materials (for a recent review on the topic, see reference [139]). Such membranes are generally prepared by the deposition of preformed MOF particles (or MOF-coated inorganic particles [140]) on standard membranes (polyethylene [141], polypropylene (e.g., Celgard) [142–145], or a glass fibers [146]). An organic binder is commonly added to both improve the adhesion and fill the void between the particles. Alternative preparation routes were also proposed, for example, (i) the in situ growth of the MOF on a standard organic [147] or inorganic [148] membrane, (ii) an electrospinning route [149], (iii) the sequential filtration of preformed MOF nanoparticles and graphene oxide layers [150], and (iv) the use of a blend of MOF and polymer (here PMMA) particles [151].
4.5.1
Separator for Li-Ion and Li-Metal Batteries
Liu et al. first reported the preparation of a commercial polypropylene separator coated with micron-sized MIL-125(Ti)-NH2 particles [152]. A homogenous ~20-μ m-thick coating was obtained. This separator was evaluated both in a symmetrical Li||Li cell and a Li||Cu cell, in which Li is alternatively deposited and removed from the Cu foil. Similar coulombic efficiencies (92–98%) were observed both with and without the MOF coating during the first cycles. Nevertheless, after ca. 40 cycles, dramatic variations were observed for the non-coated separator, indicative of the formation of short circuits. On the opposite, with the coated separator, the coulombic efficiency remained stable for more than 200 cycles. Such an improvement was observed at various current rates. Eventually, the comparison of separators coated with MIL-125(Ti) and MIL-125(Ti)-NH2 suggested that the amino groups have a beneficial impact on the stability upon long cycling, facilitating Li+ transport through NH2-Li+ interactions. Such a proposition is in line with the higher Li+ transference number measured for this separator when compared to the pure propylene- and MIL-125(Ti)-coated polypropylene membranes. Using a mixed membrane made of PVA and 60 wt% of UiO-66, Zhang et al. also observed an increase of the Li+ transference number and conductivity, leading to a better long-term cycling when used in a stainless steel||Li cell [149]. This membrane was also found to give some benefits in LFP||LTO and NMC||graphite full cells, especially at high rates (>1 C). Chen et al. further evaluated a ZIF-67-coated polypropylene mixed membrane in a NMC||Li cell [142]. An improvement of the capacity retention was observed at “high” temperature (55 C), that the authors attributed to a reduced thermal shrinkage of the coated membrane. They also noticed that the benefits depend on the nature
140
T. Devic
of the solvent used for the preparation of ZIF-67 particles, suggesting that the crystalline structure is not so critical, and the sieving properties might be better governed by the particle sizes and shapes, as well as their packing on the membrane. Finally, MOF-based membranes were also proposed to solve (to some extent) the matter of high potential (5 V) cathode materials through a dual-electrolyte strategy [153]. Qiao et al. built a full cell made of a 5 V lithium manganese nickel oxide (LMNO) at the cathode and graphite at the anode. Each electrode is in contact with a suitable electrolyte (an ionic liquid electrolyte stable at high potential for LMNO, a glyme-containing electrolyte favoring reversible Li+ interaction for graphite), and electrodes are separated by a HKUST-1-coated polypropylene membrane. Permeation tests revealed that such a membrane prevents the shuttling of critical electrolyte components, ultimately leading to improved long-term cycling reversibility and capacity. Luo et al. recently proposed to use a similar approach to build a Mg-metal battery combining two materials (Mg and a high-potential material) which request specific (and incompatible) electrolytes [148].
4.5.2
Separators for Emerging Battery Technologies
Many emerging battery technologies involve soluble redox-active species, and their shuttling between the anode and cathode should obviously be prevented. Most examples deal with Li-S batteries, whose problematics have been already discussed (see Sect. 4.3.2). The aim here is not to slow down the dissolution of polysulfides as discussed earlier but to prevent their migration to the anode (Li). Bai et al. first proposed the use of a self-supported membrane made of either Cu- or Zn-based HKUST-1 and graphene oxide [150, 154]. Permeation test carried out with Li2S6 revealed that such membranes indeed prevent the shuttling of polysulfides and ultimately lead to very good capacity retention event on longterm cycling (0.04% capacity decay per cycle at 1 C for 1000 cycles). IR spectroscopy further evidenced the formation of Zn-S bonds, confirming the capture of polysulfides by the coordinatively unsaturated metallic sites. Using the same MOF, it was possible to build a pouch (i.e., large) cell with a high sulfur loading (>5 mg cm2) which maintains a high gravimetric and areal capacity for at least 200 cycles (>900 mAh g1 and 5 mAh cm2, respectively) [145]. Following studies expanded this approach to other MOFs, notably UiO-66-NH2 [140], Mn-BTC [155], CPO-27 [141], and UiO-66(Ce) [144] but using various routes for the preparation of the membranes. Li et al. deeply investigated the effect of the nature of the MOF (Y-TFZB, ZIF-7, ZIF-8, HKUST-1) on the performances of the derived membranes, all being prepared by the same method [146]. They showed that the pore size is not the main criterion defining the capacity retention but rather the packing of the particles, as polysulfides could easily diffuse through the grain boundaries. The authors furthermore demonstrated that the chemical composition of the MOF is also a critical parameter, as some solids could degrade under the working conditions (e.g., HKUST-1(Cu) as ~1.5 V vs. Li+/Li).
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
141
MOF membranes were also evaluated in the field of Li-air (Li -O2) batteries. Such a system, which relies on the reversible formation of Li2O2, should theoretically lead to very high energy densities, but serious issues need to be solved first, such as (i) poisoning/degradation associated with the presence of CO2 and H2O, and (ii) sluggish redox reaction kinetics requesting the presence of catalysts for both oxygen reduction reaction (ORR: O2 + 2Li+ + 2e ! Li2O2) and oxygen evolution reaction (OER: Li2O2 ! O2 + 2Li+ + 2e). A first report dealt with the preparation of a mixed membrane able to permeate O2 while blocking CO2 and H2O diffusion [151]. This membrane is made of polydopamine (PDA)-coated CAU-1 particles, which were further mixed with poly(methyl methacrylate) (PMMA) particles. This membrane is deposited on a standard Ketjen black carbon positive electrode. Using a capacity limited method and a discharge cutoff voltage of 2 V, it was possible to cycle the battery made of the MOF-coated electrode up to 66 times, while only 6 for the MOF-free electrode. Nevertheless, additional experiments suggest that this improvement is not related solely to the MOF but also to the PDA which favors the formation of a hole-free coating and also interacts favorably with CO2. In another approach, Qiao et al. proposed to use an MOF membrane to prevent the diffusion of soluble redox mediators from the cathode to the anode [156]. These redox mediators are typically small redox-active molecules (here TTF and dibuthylbenzoquinone or DBBQ) and are mandatory to perform efficiently both ORR and OER at the cathode. A membrane made again of HKUST-1 particles deposited on a conventional Celgard separator was introduced between the cathodic (air) and the anodic (Li) compartment of the electrochemical cell. It significantly slowed down the diffusion of the redox mediator to the anode and hence allowed to lower the overpotential of the cell and to reach a higher performance than with the MOF-free separator over 100 cycles. The related approach was applied to a nonaqueous redox flow battery, with the idea to prevent the diffusion of the catholyte in the negative compartment (and vice versa) [147]. Here again, the efficiency of the sieving effect and the stability of the MOF in the working conditions are critical parameters which need to be addressed to rationalize the performances of such separators.
4.6
MOFs as Solid Electrolytes
Solid-state batteries use solid electrolytes instead of conventional liquid electrolytes. Such a change gave rise to great promises in terms of improved power density and enhanced safety (notably through the facilitated use of a Li anode) and focused a lot of research efforts [157]. Solid electrolytes should combine a high Li+ conductivity and transference number in a suitable temperature range (~room temperature), with good mechanical and chemical stabilities, and ideally be easily processable. Currently, best performing materials are sulfides and thiosulfates, but MOFs have been recently proposed as potential alternatives (for recent reviews on the topic, see references [158, 159]). A first approach relies on the preparation of composite electrolytes made of MOF particles and a polymer [160–164], typically those
142
T. Devic
found in conventional polymeric electrolytes such as polyethyleneoxide (PEO), with the aim at combining the benefits of the polymers (processability) with the ones of the MOF. Other approaches are based on merely pure MOFs and are discussed in detail below. The method of preparation of ion-conducting MOFs is strongly dependent on their chemical composition and structure. MOFs made of anionic frameworks could present decent intrinsic cationic conductivities and could, thus, be used as prepared, whereas neutral frameworks might request post-synthetic treatments to embed ionic species in their pores. In this last case, the aim here is to immobilize standard liquid electrolytes in the porosity of the MOFs, in order to ideally combine the high conductivity associated with the liquid state with the benefits of the solid state (see above). These systems are defined as quasi-solid electrolytes. In a recent example, Sun et al. reported the preparation of a solid electrolyte consisting of a mixed membrane made of PTFE and ZIF-8 nanoparticles filled with a LiPF6-containing liquid electrolyte [165]. The derived symmetrical Li||Li and LiCoO2||Li cells were found to behave slightly better than the one based on a liquid electrolyte at high temperature and high C-rate. Other reports focused mostly on the insertion of Li-containing ionic liquids. Wang et al. prepared an electrolyte made of the highly porous MOF-525(Cu) and LiTFSI in [EMIM][TFSI] [166]. The optimization of the MOF/ionic liquid ratio allowed to produce a solid made of 10, 55, and 35 wt% of LiTFSI, [EMIM][TFSI], and MOF, respectively. An ionic conductivity of ca. 104 S cm1 was reached, and the Li transference number was increased, likely because the confinement affects more deeply the motion of the bulkier ions. A decent capacity was finally obtained with a LiFePO4||Li cell on a large temperature range (20 to 150 C). The same approach was applied to UiO-66 [167]- and ZIF-67 [168]-derived electrolytes and in the latter case expanded to other electrode materials (nickel manganese cobalt oxide (NMC) and lithium titanate oxide (LTO)). It was also proposed to exploit the presence of coordinately unsaturated metal sites in certain MOFs to immobilize an alkaline (or alkaline-earth) salt in the porosity: the anions should bind to the metallic cations, whereas the alkaline (or alkaline-earth) cations remain free to diffuse. With this approach, a decent ionic conductivity and a high cation transference number are expected. Although not directly applied in electrochemical energy storage devices, first demonstrations were reported by the group of Long [169, 170]. Various Mg(II) salts solubilized in triglyme were introduced in the porosity of CPO-27-related MOFs [169, 169]. The amount of salt which could be inserted strongly depends on the nature of the anion (various substituted phenolates and TFSI were evaluated). With pperfluorobenzoate, it was even possible to reach a five times higher concentration in the pore than in bulk solution. After optimizing such a parameter, it was possible to reach a conductivity as high as 0.25 mS cm1. A similar approach was then applied to a Cu azolate MOF [171]. Shen et al. applied this method to encapsulate LiClO4 in PC in various MOFs (HKUST-1, MIL-100 s, UiO-66/67; MOF-5) and
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
143
again demonstrated the importance of the coordinately accessible metal site to reach a high conductivity (here ~104 S cm1) and a low activation energy (0.1–0.2 eV), i.e., with values similar to those found in polymeric and inorganic solid electrolytes [172]. Using a membrane made of LiClO4-PC-loaded UiO-67 and 10 wt% of PTFE, the authors are able to cycle a solid-state LiFePO4||Li cell at 1C with a capacity retention of 75% after 500 cycles. More recently, by using a solid electrolyte made of a LiClO4-loaded CPO-27(Cu), Yan et al. were able to prevent the dissolution of Mn (II) in a LiMn2O4||Li cell and hence to reach a higher capacity retention than with a conventional liquid electrolyte, especially above room temperature (55 C) [173]. Instead of using the metallic cation as a binding site, Zhu et al. proposed to chemically graft an anionic group (related to TFSI) on the UiO-66-NH2 network [174]. By doing so, Li+ was readily installed within the porosity, while the anionic part was obviously not mobile. Eventually, the last approach, which is probably the most natural one, relies on the use of intrinsically anionic frameworks. An example was reported by Ashraf et al., who synthesized an anionic framework made of octahedrally coordinated Ge(IV) ions and bis-catecholate ligands containing Li+ ions within the pores for charge compensation [175]. This solid presents a conductivity of about 106 S cm1 at room temperature, which could be increased to ~104 S cm1 after the further insertion of LiPF6 in EC/DEC. Eventually, Xu et al. described an MOF made of polyoxometalates inorganic building blocks and containing alkylammonium cations in its pristine state [176]. Upon suspending this solid in a LiTSFI acetonitrile solution, ammoniums were replaced by Li+ ions. This solid presents a conductivity of 104 S cm1 and a high Li+ transference number (0.87), and the derived solid-state electrolyte could again be used in a model LiFePO4||Li cell.
4.7
Conclusion and Prospects
This chapter gave a brief overview of the potential applications of MOFs in the field of electrochemical energy storage, not only as active materials, but also everywhere their tunable porosity could be advantageously exploited. As active materials, MOFs cannot yet be considered as serious competitors to conventional inorganic materials in terms of performances (volumetric and gravimetric). They might become competitive when compared to molecular organic materials, with which they share common features such as the potential use of renewable building blocks (for the organic part) and their likely easy recyclability (MOFs are rather easy to dissolve in either acidic or basic media) but with a lower solubility offering potentially better capacity retention. Nevertheless, most reports still focus on half cells, and their incorporation in full cells (e.g., as anodes [83, 94]), both in coin cells and more relevant pouch cells, remains scare. In other fields of applications (coating, separator, solid-state electrolyte), the relationship between the structure (not only of the
144
T. Devic
as-prepared materials but also at work) and the properties must be more thoroughly investigated. Some promising results have been reported, but the nature and dynamics of the interfaces between the MOF and the other components (active materials, especially metallic Li), which is known to be of critical importance in electrodes, is almost unexplored yet. In the field of solid electrolytes, reports focused on MOFs comprising ions but also solvent molecules in the pores; the question arises whether high conductivities could also be achieved in the unsolvated state, similarly to what happens in inorganic electrolytes. Eventually, MOFs might also offer interesting alternatives in emerging fields of electrochemical energy storage: – Besides alkaline (Li, Na, K)-based batteries, multivalent cations (Mg, Zn, Ca, Al) are also considered. MOFs will here face similar problems than conventional materials but will also face specific stability issues, as the exchange between the cations of the framework and the electrolyte is likely to occur if both wear the same charge [177]. Here, MOFs based on strong and/or inert ligand-metal bonds should be first investigated. – Aqueous batteries for stationary storage are considered with an ever-lasting interest [178, 179], but MOFs have been only scarcely explored in this field (Cu3(HHTP)2 used as a cathode material in a Zn aqueous battery [95]; see Sect. 4.2). Considering that serious issues still remain to be tackled [178], water-stable MOFs could be interesting here. – MOFs combining a high porosity with catalytic active sites could be of interest in gas-based batteries, in which the porosity might both favor gas pre-concentration close to the catalytic site and the storage of the solid reduction products; such approach has been recently proposed for Li-air [180, 181], Zn-air [182], and Li-CO2 [183] batteries. – MOFs could also be used as porous additives, for example, to act as scavengers for the degradation products (PF5, metallic cation) formed in the course of the cycling [184] or to buffer strong volume variations associated with alloying reactions [185]. – Semi-conductive MOFs (see Sect. 4.2) seem to be of interest as anode materials for either Li- or Na-ion capacitors (i.e., built up from a supercapacitive cathode and a faradic anode) [186, 187]. Acknowledgments Coworkers at the Institut des Matériaux Jean Rouxel, namely, Philippe Poizot, Stéven Renault, Joel Gaubicher, Bernard Lestriez, Nicolas Dupré, and Dominique Guyomard, are warmly thanked for fruitful discussions, as well as the PhD students involved in related research projects (Morgane Denis, Lucas Huet, Nassima Kana, Alia Jouhara). The Region Pays de la Loire (project PSR “MatHySE2”) and the Agence Nationale de la Recherche (project ANR-19-CE080029 “ThioMOFs”) are acknowledged for funding.
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
145
References 1. Li X, Cheng F, Zhang S, Chen J (2006) Shape-controlled synthesis and lithium-storage study of metal-organic frameworks Zn4O(1,3,5-benzenetribenzoate)2. J. Power Sources 160:542–547 2. Férey G, Millange F, Morcrette M et al (2007) Mixed-valence Li/Fe-based metal-organic frameworks with both reversible redox and sorption properties. Angew. Chem. Int. Ed. 46:3259–3263 3. Ke F-S, Wu Y-S, Deng H (2015) Metal-organic frameworks for lithium ion batteries and supercapacitors. J. Solid State Chem. 223:109–121 4. Wang L, Han Y, Feng X et al (2016) Metal-organic frameworks for energy storage: batteries and supercapacitors. Coord Chem Rev 307(Part 2):361–381. https://doi.org/10.1016/j.ccr. 2015.09.002 5. Zhou J, Wang B (2017) Emerging crystalline porous materials as a multifunctional platform for electrochemical energy storage. Chem. Soc. Rev. 46:6927–6945. https://doi.org/10.1039/ C7CS00283A 6. Zhang Y, Riduan SN, Wang J (2017) Redox active metal– and covalent organic frameworks for energy storage: balancing porosity and electrical conductivity. Chem. Eur. J. 23:16419–16431. https://doi.org/10.1002/chem.201702919 7. Wu HB, Lou XWD (2017) Metal-organic frameworks and their derived materials for electrochemical energy storage and conversion: promises and challenges. Sci Adv 3:eaap9252. https://doi.org/10.1126/sciadv.aap9252 8. Wu Z, Xie J, Xu ZJ et al (2019) Recent progress in metal–organic polymers as promising electrodes for lithium/sodium rechargeable batteries. J. Mater. Chem. A 7:4259–4290. https:// doi.org/10.1039/C8TA11994E 9. Calbo J, Golomb MJ, Walsh A (2019) Redox-active metal–organic frameworks for energy conversion and storage. J. Mater. Chem. A 7:16571–16597. https://doi.org/10.1039/ C9TA04680A 10. Wang Z, Tao H, Yue Y (2019) Metal-organic-framework-based cathodes for enhancing the electrochemical performances of batteries: a review. ChemElectroChem 6:5358–5374. https:// doi.org/10.1002/celc.201900843 11. Wang D-G, Liang Z, Gao S et al (2020) Metal-organic framework-based materials for hybrid supercapacitor application. Coord. Chem. Rev. 404:213093. https://doi.org/10.1016/j.ccr. 2019.213093 12. Guan BY, Yu XY, Wu HB, Lou XWD (2017) Complex nanostructures from materials based on metal–organic frameworks for electrochemical energy storage and conversion. Adv. Mater. 29:1703614. https://doi.org/10.1002/adma.201703614 13. Xie Z, Xu W, Cui X, Wang Y (2017) Recent progress in metal–organic frameworks and their derived nanostructures for energy and environmental applications. ChemSusChem 10:1645–1663. https://doi.org/10.1002/cssc.201601855 14. Tang H, Zheng M, Hu Q et al (2018) Derivatives of coordination compounds for rechargeable batteries. J. Mater. Chem. A 6:13999–14024. https://doi.org/10.1039/C8TA03644F 15. Xie X-C, Huang K-J, Wu X (2018) Metal–organic framework derived hollow materials for electrochemical energy storage. J. Mater. Chem. A 6:6754–6771. https://doi.org/10.1039/ C8TA00612A 16. Zou G, Hou H, Ge P et al (2018) Metal–organic framework-derived materials for sodium energy storage. Small 14:1702648. https://doi.org/10.1002/smll.201702648 17. Liang Z, Zhao R, Qiu T et al (2019) Metal-organic framework-derived materials for electrochemical energy applications. EnergyChem 1:100001. https://doi.org/10.1016/j.enchem.2019. 100001 18. Cai Z-X, Wang Z-L, Kim J, Yamauchi Y (2019) Hollow functional materials derived from metal–organic frameworks: synthetic strategies, conversion mechanisms, and electrochemical applications. Adv. Mater. 31:1804903. https://doi.org/10.1002/adma.201804903
146
T. Devic
19. Zhang L, Liu H, Shi W, Cheng P (2019) Synthesis strategies and potential applications of metal-organic frameworks for electrode materials for rechargeable lithium ion batteries. Coord. Chem. Rev. 388:293–309. https://doi.org/10.1016/j.ccr.2019.02.030 20. Simon P, Gogotsi Y, Dunn B (2014) Where do batteries end and supercapacitors begin? Science 343:1210–1211. https://doi.org/10.1126/science.1249625 21. Winter M, Brodd RJ (2004) What are batteries, fuel cells, and supercapacitors? Chem. Rev. 104:4245–4270. https://doi.org/10.1021/cr020730k 22. Whittingham MS (2004) Lithium batteries and cathode materials. Chem. Rev. 104:4271–4302. https://doi.org/10.1021/cr020731c 23. Poizot P, Laruelle S, Grugeon S et al (2000) Nano-sized transition-metal oxides as negativeelectrode materials for lithium-ion batteries. Nature 407:496–499. https://doi.org/10.1038/ 35035045 24. Wang F, Robert R, Chernova NA et al (2011) Conversion reaction mechanisms in lithium ion batteries: study of the binary metal fluoride electrodes. J. Am. Chem. Soc. 133:18828–18836. https://doi.org/10.1021/ja206268a 25. Obrovac MN, Chevrier VL (2014) Alloy negative electrodes for Li-ion batteries. Chem. Rev. 114:11444–11502. https://doi.org/10.1021/cr500207g 26. Shao Y, El-Kady MF, Sun J et al (2018) Design and mechanisms of asymmetric supercapacitors. Chem. Rev. 118:9233–9280. https://doi.org/10.1021/acs.chemrev.8b00252 27. Costentin C, Savéant J-M (2019) Energy storage: pseudocapacitance in prospect. Chem. Sci. 10:5656–5666. https://doi.org/10.1039/C9SC01662G 28. Hwang J-Y, Myung S-T, Sun Y-K (2017) Sodium-ion batteries: present and future. Chem. Soc. Rev. 46:3529–3614. https://doi.org/10.1039/C6CS00776G 29. Aurbach D, Suresh GS, Levi E et al (2007) Progress in rechargeable magnesium battery technology. Adv. Mater. 19:4260–4267. https://doi.org/10.1002/adma.200701495 30. Mao M, Gao T, Hou S, Wang C (2018) A critical review of cathodes for rechargeable Mg batteries. Chem. Soc. Rev. 47:8804–8841. https://doi.org/10.1039/C8CS00319J 31. Ambroz F, Macdonald TJ, Nann T (2017) Trends in aluminium-based intercalation batteries. Adv Energy Mater 7. https://doi.org/10.1002/aenm.201602093 32. Ding J, Hu W, Paek E, Mitlin D (2018) Review of hybrid ion capacitors: from aqueous to lithium to sodium. Chem. Rev. 118:6457–6498. https://doi.org/10.1021/acs.chemrev.8b00116 33. Aravindan V, Gnanaraj J, Lee Y-S, Madhavi S (2014) Insertion-type electrodes for nonaqueous Li-ion capacitors. Chem. Rev. 114:11619–11635. https://doi.org/10.1021/ cr5000915 34. Burtch NC, Jasuja H, Walton KS (2014) Water stability and adsorption in metal-organic frameworks. Chem. Rev. 114:10575–10612 35. Low JJ, Benin AI, Jakubczak P et al (2009) Virtual high throughput screening confirmed experimentally: porous coordination polymer hydration. J Am Chem Soc:15834–15842. https://doi.org/10.1021/ja9061344 36. Devic T, Serre C (2014) High valence 3p and transition metals based MOFs. Chem. Soc. Rev. 43:6097–6115 37. Sun L, Campbell MG, Dincă M (2016) Electrically conductive porous metal–organic frameworks. Angew. Chem. Int. Ed. 55:3566–3579. https://doi.org/10.1002/anie.201506219 38. Huang L-W, Yang C-J, Lin K-J (2002) Toward the design and synthesis of Lithium-ion intercalation into a coordination π–π framework host. Chem. Eur. J. 8:396–400. https://doi. org/10.1002/1521-3765(20020118)8:23.0.CO;2-Y 39. Masquelier C, Croguennec L (2013) Polyanionic (phosphates, silicates, Sulfates) frameworks as electrode materials for rechargeable Li (or Na) batteries. Chem. Rev. 113:6552–6591 40. Saouma CT, Tsou C-C, Richard S et al (2019) Sodium-coupled electron transfer reactivity of metal–organic frameworks containing titanium clusters: the importance of cations in redox chemistry. Chem. Sci. 10:1322–1331. https://doi.org/10.1039/C8SC04138E 41. Fateeva A, Horcajada P, Devic T et al (2010) Synthesis, structure, characterization, and redox properties of the porous MIL-68(Fe) solid. Eur. J. Inorg. Chem. 2010:3789–3794
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
147
42. Gallis DFS, Iii HDP, Anderson TM, Chapman KW (2016) Electrochemical activity of Fe-MIL-100 as a positive electrode for Na-ion batteries. J. Mater. Chem. A 4:13764–13770. https://doi.org/10.1039/C6TA03943J 43. Yamada T, Shiraishi K, Kitagawa H, Kimizuka N (2017) Applicability of MIL-101(Fe) as a cathode of lithium ion batteries. Chem. Commun. 53:8215–8218. https://doi.org/10.1039/ C7CC01712J 44. Shin J, Kim M, Cirera J et al (2015) MIL-101(Fe) as a lithium-ion battery electrode material: a relaxation and intercalation mechanism during lithium insertion. J. Mater. Chem. A 3:4738–4744 45. Schmidt S, Sheptyakov D, Jumas J-C et al (2015) Lithium Iron Methylenediphosphonate: a model material for new organic-inorganic hybrid positive electrode materials for Li ion batteries. Chem. Mater. 27:7889–7895 46. de Combarieu G, Hamelet S, Millange F et al (2009) In situ Fe XAFS of reversible lithium insertion in a flexible metal organic framework material. Electrochem. Commun. 11:1881–1884 47. Combelles C, Ben Yahia M, Pedesseau L, Doublet M-L (2010) Design of electrode materials for Lithium-ion batteries: the example of metal-organic frameworks. J. Phys. Chem. C 114:9518–9527 48. Barthelet K, Marrot J, Riou D, Férey G (2002) A breathing hybrid organic – inorganic solid with very large pores and high magnetic characteristics. Angew. Chem. Int. Ed. 41:281–284 49. Kaveevivitchai W, Jacobson AJ (2015) Exploration of vanadium benzenedicarboxylate as a cathode for rechargeable lithium batteries. J. Power Sources 278:265–273 50. Leclerc H, Devic T, Devautour-Vinot S et al (2011) Influence of the oxidation state of the metal center on the flexibility and adsorption properties of a porous metal organic framework: MIL-47(V). J. Phys. Chem. C 115:19828–19840 51. Deuchert K, Hünig S (1978) Multistage organic redox systems – a general structural principle. Angew. Chem. Int. Ed. 17:875–886. https://doi.org/10.1002/anie.197808753 52. Tian B, Ning G-H, Gao Q et al (2016) Crystal engineering of naphthalenediimide-based metal–organic frameworks: structure-dependent Lithium storage. ACS Appl. Mater. Interfaces 8:31067–31075. https://doi.org/10.1021/acsami.6b11772 53. Lou X, Geng F, Hu B et al (2019) Reversible high-voltage N-redox chemistry in metal– organic frameworks for high-rate anion-intercalation batteries. ACS Appl Energy Mater 2:413–419. https://doi.org/10.1021/acsaem.8b01428 54. Zhang L, Cheng F, Shi W et al (2018) Transition-metal-triggered high-efficiency lithium ion storage via coordination interactions with redox-active croconate in one-dimensional metal– organic anode materials. ACS Appl. Mater. Interfaces 10:6398–6406. https://doi.org/10.1021/ acsami.7b18758 55. Saravanan K, Nagarathinam M, Balaya P, Vittal JJ (2010) Lithium storage in a metal organic framework with diamondoid topology – a case study on metal formates. J. Mater. Chem. 20 56. Armand M, Grugeon S, Vezin H et al (2009) Conjugated dicarboxylate anodes for Li-ion batteries. Nat Mater 8:120–125 57. Lee HH, Park Y, Kim SH et al (2015) Mechanistic studies of transition metal-terephthalate coordination complexes upon electrochemical lithiation and delithiation. Adv. Funct. Mater. 25:4859–4866 58. Li C, Hu X, Tong W et al (2017) Ultrathin manganese-based metal–organic framework nanosheets: low-cost and energy-dense lithium storage anodes with the coexistence of metal and ligand redox activities. ACS Appl. Mater. Interfaces 9:29829–29838. https://doi.org/10. 1021/acsami.7b09363 59. Wang L, Zou J, Chen S et al (2017) Zinc terephthalates ZnC8H4O4 as anodes for lithium ion batteries. Electrochim. Acta 235:304–310. https://doi.org/10.1016/j.electacta.2017.03.095 60. Xia S-B, Yu S-W, Yao L-F et al (2019) Robust hexagonal nut-shaped titanium(IV) MOF with porous structure for ultra-high performance lithium storage. Electrochim. Acta 296:746–754. https://doi.org/10.1016/j.electacta.2018.11.135
148
T. Devic
61. Wu N, Wang W, Kou L-Q et al (2018) Enhanced Li storage stability induced by locating Sn in metal–organic frameworks. Chem. Eur. J. 24:6330–6333. https://doi.org/10.1002/chem. 201800215 62. Maiti S, Pramanik A, Manju U, Mahanty S (2016) Cu3(1,3,5-benzenetricarboxylate)2 metalorganic framework: a promising anode material for lithium-ion battery. Microporous Mesoporous Mater. 226:353–359 63. Ning Y, Lou X, Li C et al (2017) Ultrathin cobalt-based metal–organic framework nanosheets with both metal and ligand redox activities for superior lithium storage. Chem. Eur. J. 23:15984–15990. https://doi.org/10.1002/chem.201703077 64. Hu L, Lin X-M, Mo J-T et al (2017) Lead-based metal–organic framework with stable lithium anodic performance. Inorg. Chem. 56:4289–4295. https://doi.org/10.1021/acs.inorgchem. 6b02663 65. Schmidt S, Sallard S, Sheptyakov D et al (2017) Ligand influence in Li-ion battery hybrid active materials: Ni methylenediphosphonate vs. Ni dimethylamino methylenediphosphonate. Chem Commun 53:5420–5423. https://doi.org/10.1039/C7CC01982C 66. Dong C, Xu L (2017) Cobalt- and cadmium-based metal–organic frameworks as highperformance anodes for sodium ion batteries and lithium ion batteries. ACS Appl. Mater. Interfaces 9:7160–7168. https://doi.org/10.1021/acsami.6b15757 67. Fan C, Zhao M, Li C et al (2017) Investigating the electrochemical behavior of Cobalt (II) terephthalate (CoC8H4O4) as the organic anode in K-ion battery. Electrochim. Acta 253:333–338. https://doi.org/10.1016/j.electacta.2017.09.078 68. Li C, Hu X, Hu B (2017) Cobalt(II) dicarboxylate-based metal-organic framework for longcycling and high-rate potassium-ion battery anode. Electrochim. Acta 253:439–444. https:// doi.org/10.1016/j.electacta.2017.09.090 69. Grimaud A, Hong WT, Shao-Horn Y, Tarascon JM (2016) Anionic redox processes for electrochemical devices. Nat Mater 15:121–126 70. Kaim W (2011) Manifestations of noninnocent ligand behavior. Inorg. Chem. 50:9752–9765 71. Nguyen TLA, Devic T, Mialane P et al (2010) Reinvestigation of the MII (M ¼ Ni, Co)/ TetraThiafulvaleneTetraCarboxylate system using high-throughput methods: isolation of a molecular complex and its single-crystal-to-single-crystal transformation to a two-dimensional coordination polymer. Inorg. Chem. 49:10710–10717 72. Zhang Z, Yoshikawa H, Awaga K (2014) Monitoring the solid-state electrochemistry of Cu (2,7-AQDC) (AQDC ¼ Anthraquinone Dicarboxylate) in a Lithium battery: coexistence of metal and ligand redox activities in a metal-organic framework. J. Am. Chem. Soc. 136:16112–16115 73. Zhang Z, Yoshikawa H, Awaga K (2016) Discovery of a “bipolar charging” mechanism in the solid-state electrochemical process of a flexible metal–organic framework. Chem. Mater. 28:1298–1303. https://doi.org/10.1021/acs.chemmater.5b04075 74. Kambe T, Sakamoto R, Kusamoto T et al (2014) Redox control and high conductivity of Nickel Bis(dithiolene) complex-nanosheet: a potential organic two-dimensional topological insulator. J. Am. Chem. Soc. 136:14357–14360 75. Huang X, Zhang S, Liu L et al (2018) Superconductivity in a Copper(II)-based coordination polymer with perfect Kagome structure. Angew. Chem. Int. Ed. 57:146–150. https://doi.org/ 10.1002/anie.201707568 76. Heintz RA, Zhao H, Ouyang X et al (1999) New insight into the nature of Cu(TCNQ): solution routes to two distinct polymorphs and their relationship to crystalline films that display bistable switching behavior. Inorg. Chem. 38:144–156. https://doi.org/10.1021/ic9812095 77. Fang C, Huang Y, Yuan L et al (2017) A metal–organic compound as cathode material with superhigh capacity achieved by reversible cationic and anionic redox chemistry for highenergy sodium-ion batteries. Angew. Chem. Int. Ed. 56:6793–6797. https://doi.org/10.1002/ anie.201701213 78. Ma J, Zhou E, Fan C et al (2018) Endowing CuTCNQ with a new role: a high-capacity cathode for K-ion batteries. Chem. Commun. 54:5578–5581. https://doi.org/10.1039/C8CC00802G
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
149
79. Huang Y, Fang C, Zhang W et al (2019) Sustainable cycling enabled by a high-concentration electrolyte for lithium-organic batteries. Chem. Commun. 55:608–611. https://doi.org/10. 1039/C8CC09307E 80. Fang C, Ye Z, Wang Y et al (2019) Immobilizing an organic electrode material through π–π interaction for high-performance Li-organic batteries. J. Mater. Chem. A 7:22398–22404. https://doi.org/10.1039/C9TA07403A 81. Dühnen S, Nölle R, Wrogemann J et al (2019) Reversible anion storage in a metal-organic framework for dual-ion battery systems. J. Electrochem. Soc. 166:A5474–A5482. https://doi. org/10.1149/2.0681903jes 82. Meng C, Chen T, Fang C et al (2019) Multiple active sites: lithium storage mechanism of Cu-TCNQ as an anode material for lithium-ion batteries. Chem. Asian J. 14:4289–4295. https://doi.org/10.1002/asia.201901190 83. Chen Y, Tang M, Wu Y et al (2019) A one-dimensional π–d conjugated coordination polymer for sodium storage with catalytic activity in Negishi coupling. Angew. Chem. Int. Ed. 58:14731–14739. https://doi.org/10.1002/anie.201908274 84. Wu Y, Chen Y, Tang M et al (2019) A highly conductive conjugated coordination polymer for fast-charge sodium-ion batteries: reconsidering its structures. Chem. Commun. 55:10856–10859. https://doi.org/10.1039/C9CC05679C 85. Saines PJ, Yeung HHM, Hester JR et al (2011) Detailed investigations of phase transitions and magnetic structure in Fe(iii), Mn(ii), Co(ii) and Ni(ii) 3,4,5-trihydroxybenzoate (gallate) dihydrates by neutron and X-ray diffraction. Dalton Trans. 40:6401–6410 86. Cooper L, Hidalgo T, Gorman M et al (2015) A biocompatible porous Mg-gallate metalorganic framework as an antioxidant carrier. Chem. Commun. 51:5848–5851 87. Hidalgo T, Cooper L, Gorman M et al (2017) Crystal structure dependent in vitro antioxidant activity of biocompatible calcium gallate MOFs. J. Mater. Chem. B 5:2813–2822. https://doi. org/10.1039/C6TB03101C 88. Aubrey ML, Long JR (2015) A dual-ion battery cathode via oxidative insertion of anions in a metal-organic framework. J. Am. Chem. Soc. 137:13594–13602 89. Hmadeh M, Lu Z, Liu Z et al (2012) New porous crystals of extended metal-catecholates. Chem. Mater. 24:3511–3513 90. Gu S, Bai Z, Majumder S et al (2019) Conductive metal–organic framework with redox metal center as cathode for high rate performance lithium ion battery. J. Power Sources 429:22–29. https://doi.org/10.1016/j.jpowsour.2019.04.087 91. Wu H, Zhang W, Kandambeth S et al (2019) Conductive metal–organic frameworks selectively grown on laser-scribed graphene for electrochemical microsupercapacitors. Adv. Energy Mater. 9:1900482. https://doi.org/10.1002/aenm.201900482 92. Du X, Zhang J, Wang H et al (2020) Solid–solid interface growth of conductive metal–organic framework nanowire arrays and their supercapacitor application. Mater Chem Front 4:243–251. https://doi.org/10.1039/C9QM00527G 93. Liu J, Zhou Y, Xie Z et al (2020) Conjugated Copper–Catecholate framework electrodes for efficient energy storage. Angew. Chem. Int. Ed. 59:1081–1086. https://doi.org/10.1002/anie. 201912642 94. Guo L, Sun J, Zhang W et al (2019) Bottom-up fabrication of 1D Cu-based conductive metal– organic framework nanowires as a high-rate anode towards efficient Lithium storage. ChemSusChem 12:5051–5058. https://doi.org/10.1002/cssc.201902194 95. Nam KW, Park SS, dos Reis R et al (2019) Conductive 2D metal-organic framework for highperformance cathodes in aqueous rechargeable zinc batteries. Nat. Commun. 10:1–10. https:// doi.org/10.1038/s41467-019-12857-4 96. Taniguchi K, Chen J, Sekine Y, Miyasaka H (2017) Magnetic phase switching in a tetraoxolene-bridged honeycomb ferrimagnet using a lithium ion battery system. Chem. Mater. 29:10053–10059. https://doi.org/10.1021/acs.chemmater.7b03691 97. Nagatomi H, Yanai N, Yamada T et al (2018) Synthesis and electric properties of a two-dimensional metal-organic framework based on phthalocyanine. Chem. Eur. J. 24:1806–1810. https://doi.org/10.1002/chem.201705530
150
T. Devic
98. Chang C-H, Li A-C, Popovs I et al (2019) Elucidating metal and ligand redox activities of a copper-benzoquinoid coordination polymer as the cathode for lithium-ion batteries. J. Mater. Chem. A 7:23770–23774. https://doi.org/10.1039/C9TA05244E 99. Chen J, Taniguchi K, Sekine Y, Miyasaka H (2020) Electrochemical development of magnetic long-range correlations with Tc ¼ 128 K in a tetraoxolene-bridged Fe-based framework. J Magn Magn Mater 494:165818. https://doi.org/10.1016/j.jmmm.2019.165818 100. Wada K, Sakaushi K, Sasaki S, Nishihara H (2018) Multielectron-transfer-based rechargeable energy storage of two-dimensional coordination frameworks with non-innocent ligands. Angew. Chem. Int. Ed. 57:8886–8890. https://doi.org/10.1002/anie.201802521 101. Park J, Lee M, Feng D et al (2018) Stabilization of Hexaaminobenzene in a 2D conductive metal–organic framework for high power sodium storage. J. Am. Chem. Soc. 140:10315–10323. https://doi.org/10.1021/jacs.8b06020 102. Kapaev RR, Olthof S, Zhidkov IS et al (2019) Nickel(II) and Copper(II) coordination polymers derived from 1,2,4,5-Tetraaminobenzene for Lithium-ion batteries. Chem. Mater. 31:5197–5205. https://doi.org/10.1021/acs.chemmater.9b01366 103. de Combarieu G, Morcrette M, Millange F et al (2009) Influence of the benzoquinone sorption on the structure and electrochemical performance of the MIL-53(Fe) hybrid porous material in a Lithium-ion battery. Chem. Mater. 21:1602–1611 104. Xu J, Lawson T, Fan H et al (2018) Updated metal compounds (MOFs, -S, -OH, -N, -C) used as cathode materials for lithium–sulfur batteries. Adv. Energy Mater. 8:1702607. https://doi. org/10.1002/aenm.201702607 105. Zheng Y, Zheng S, Xue H, Pang H (2019) Metal–organic frameworks for lithium–sulfur batteries. J. Mater. Chem. A 7:3469–3491. https://doi.org/10.1039/C8TA11075A 106. Demir-Cakan R, Morcrette M, Nouar F et al (2011) Cathode composites for Li-S batteries via the use of oxygenated porous architectures. J. Am. Chem. Soc. 133:16154–16160 107. Zhou J, Li R, Fan X et al (2014) Rational design of a metal-organic framework host for sulfur storage in fast, long-cycle Li-S batteries. Energy Environ. Sci. 7:2715–2724 108. Wang Z, Li X, Cui Y et al (2013) A metal-organic framework with open metal sites for enhanced confinement of sulfur and lithium-sulfur battery of long cycling life. Cryst. Growth Des. 13:5116–5120 109. Zheng J, Tian J, Wu D et al (2014) Lewis acid-base interactions between polysulfides and metal organic framework in lithium sulfur batteries. Nano Lett. 14:2345–2352 110. Liu X-F, Guo X-Q, Wang R et al (2019) Manganese cluster-based MOF as efficient polysulfide-trapping platform for high-performance lithium–sulfur batteries. J. Mater. Chem. A 7:2838–2844. https://doi.org/10.1039/C8TA09973A 111. Liu X, Wang S, Wang A et al (2019) A new cathode material synthesized by a thiol-modified metal–organic framework (MOF) covalently connecting sulfur for superior long-cycling stability in lithium–sulfur batteries. J. Mater. Chem. A 7:24515–24523. https://doi.org/10. 1039/C9TA08043K 112. Baumann AE, Burns DA, Díaz JC, Thoi VS (2019) Lithiated defect sites in Zr metal–organic framework for enhanced sulfur utilization in Li–S batteries. ACS Appl. Mater. Interfaces 11:2159–2167. https://doi.org/10.1021/acsami.8b19034 113. Baumann AE, Han X, Butala MM, Thoi VS (2019) Lithium thiophosphate functionalized zirconium MOFs for Li–S batteries with enhanced rate capabilities. J. Am. Chem. Soc. 141:17891–17899. https://doi.org/10.1021/jacs.9b09538 114. Zhou J, Yu X, Fan X et al (2015) The impact of the particle size of a metal-organic framework for sulfur storage in Li-S batteries. J. Mater. Chem. A 3:8272–8275 115. Baumann AE, Aversa GE, Roy A et al (2018) Promoting sulfur adsorption using surface Cu sites in metal–organic frameworks for lithium sulfur batteries. J. Mater. Chem. A 6:4811–4821. https://doi.org/10.1039/C8TA01057A 116. Cai D, Lu M, Li L et al (2019) A highly conductive MOF of graphene analogue Ni3(HITP)2 as a sulfur host for high-performance lithium–sulfur batteries. Small 15:1902605. https://doi.org/ 10.1002/smll.201902605
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
151
117. Liu B, Baumann AE, Thoi VS (2019) Modulating charge transport in MOFs with zirconium oxide nodes and redox-active linkers for lithium sulfur batteries. Polyhedron 170:788–795. https://doi.org/10.1016/j.poly.2019.06.044 118. Zhao Z, Wang S, Liang R et al (2014) Graphene-wrapped chromium-MOF(MIL-101)/sulfur composite for performance improvement of high-rate rechargeable Li-S batteries. J. Mater. Chem. A 2:13509–13512 119. Wu Y, Jiang H, Ke F-S, Deng H (2019) Three-dimensional hierarchical constructs of MOFon-reduced graphene oxide for lithium–sulfur batteries. Chem. Asian J. 14:3577–3582. https:// doi.org/10.1002/asia.201900848 120. Mao Y, Li G, Guo Y et al (2017) Foldable interpenetrated metal-organic frameworks/carbon nanotubes thin film for lithium–sulfur batteries. Nat. Commun. 8:14628. https://doi.org/10. 1038/ncomms14628 121. Liu B, Bo R, Taheri M et al (2019) Metal–organic frameworks/conducting polymer hydrogel integrated three-dimensional free-standing monoliths as ultrahigh loading Li–S battery electrodes. Nano Lett. 19:4391–4399. https://doi.org/10.1021/acs.nanolett.9b01033 122. Jiang H, Liu X-C, Wu Y et al (2018) Metal–organic frameworks for high charge–discharge rates in lithium–sulfur batteries. Angew. Chem. Int. Ed. 57:3916–3921. https://doi.org/10. 1002/anie.201712872 123. Jin W-W, Li H-J, Zou J-Z et al (2018) Conducting polymer-coated MIL-101/S composite with scale-like shell structure for improving Li–S batteries. RSC Adv. 8:4786–4793. https://doi.org/ 10.1039/C7RA12800B 124. Xie Y, Chen S, Lin Z et al (2019) Enhanced electrochemical performance of Li-rich layered oxide, Li1.2Mn0.54Co0.13Ni0.13O2, by surface modification derived from a MOF-assisted treatment. Electrochem. Commun. 99:65–70. https://doi.org/10.1016/j.elecom.2019.01.005 125. Zhang J, Wan J, Wang J et al (2019) Hollow multi-shelled structure with metal–organicframework-derived coatings for enhanced Lithium storage. Angew. Chem. Int. Ed. 58:5266–5271. https://doi.org/10.1002/anie.201814563 126. Majeed MK, Ma G, Cao Y et al (2019) Metal–organic frameworks-derived mesoporous Si/SiOx@NC nanospheres as a long-lifespan anode material for Lithium-ion batteries. Chem. Eur. J. 25:11991–11997. https://doi.org/10.1002/chem.201903043 127. Li S, Fu X, Zhou J et al (2016) An effective approach to improve the electrochemical performance of LiNi0.6Co0.2Mn0.2O2 cathode by an MOF-derived coating. J. Mater. Chem. A 4:5823–5827. https://doi.org/10.1039/C5TA10773C 128. Yang SJ, Nam S, Kim T et al (2013) Preparation and exceptional Lithium anodic performance of porous carbon-coated ZnO quantum dots derived from a metal–organic framework. J. Am. Chem. Soc. 135:7394–7397. https://doi.org/10.1021/ja311550t 129. Devic T, Lestriez B, Roué L (2019) Silicon electrodes for Li-ion batteries. Addressing the challenges through coordination chemistry. ACS Energy Lett 4:550–557. https://doi.org/10. 1021/acsenergylett.8b02433 130. Qiao Q-Q, Li G-R, Wang Y-L, Gao X-P (2016) To enhance the capacity of Li-rich layered oxides by surface modification with metal–organic frameworks (MOFs) as cathodes for advanced lithium-ion batteries. J. Mater. Chem. A 4:4440–4447. https://doi.org/10.1039/ C6TA00882H 131. Yu Y, Yue C, Han Y et al (2017) Si nanorod arrays modified with metal–organic segments as anodes in lithium ion batteries. RSC Adv. 7:53680–53685. https://doi.org/10.1039/ C7RA10905A 132. Yu Y, Yue C, Lin X et al (2016) ZIF-8 cooperating in TiN/Ti/Si nanorods as efficient anodes in micro-Lithium-ion-batteries. ACS Appl. Mater. Interfaces 8:3992–3999. https://doi.org/10. 1021/acsami.5b11287 133. Han Y, Qi P, Zhou J et al (2015) Metal–organic frameworks (MOFs) as Sandwich coating cushion for silicon anode in lithium ion batteries. ACS Appl. Mater. Interfaces 7:26608–26613. https://doi.org/10.1021/acsami.5b08109
152
T. Devic
134. Jin J, Zheng Y, Huang S et al (2019) Directly anchoring 2D NiCo metal–organic frameworks on few-layer black phosphorus for advanced lithium-ion batteries. J. Mater. Chem. A 7:783–790. https://doi.org/10.1039/C8TA09327J 135. Han Y, Yu D, Zhou J et al (2017) A lithium ion highway by surface coordination polymerization: in situ growth of metal–organic framework thin layers on metal oxides for exceptional rate and cycling performance. Chem. Eur. J. 23:11513–11518. https://doi.org/10.1002/chem. 201703016 136. Fan L, Guo Z, Zhang Y et al (2019) Stable artificial solid electrolyte interphase films for lithium metal anode via metal–organic frameworks cemented by polyvinyl alcohol. J. Mater. Chem. A 8:251–258. https://doi.org/10.1039/C9TA10405D 137. Yuan S, Bao JL, Li C et al (2019) Dual lithiophilic structure for uniform Li deposition. ACS Appl. Mater. Interfaces 11:10616–10623. https://doi.org/10.1021/acsami.8b19654 138. Chu F, Hu J, Wu C et al (2019) Metal–organic frameworks as electrolyte additives to enable ultrastable plating/stripping of Li anode with dendrite inhibition. ACS Appl. Mater. Interfaces 11:3869–3879. https://doi.org/10.1021/acsami.8b17924 139. He Y, Qiao Y, Chang Z, Zhou H (2019) The potential of electrolyte filled MOF membranes as ionic sieves in rechargeable batteries. Energy Environ. Sci. 12:2327–2344. https://doi.org/10. 1039/C8EE03651A 140. Suriyakumar S, Stephan AM, Angulakshmi N et al (2018) Metal–organic framework@SiO2 as permselective separator for lithium–sulfur batteries. J. Mater. Chem. A 6:14623–14632. https://doi.org/10.1039/C8TA02259C 141. Lee DH, Ahn JH, Park M-S et al (2018) Metal-organic framework/carbon nanotube-coated polyethylene separator for improving the cycling performance of lithium-sulfur cells. Electrochim. Acta 283:1291–1299. https://doi.org/10.1016/j.electacta.2018.07.031 142. Chen P, Ren H, Yan L et al (2019) Metal–organic frameworks enabled high-performance separators for safety-reinforced lithium ion battery. ACS Sustain. Chem. Eng. 7:16612–16619. https://doi.org/10.1021/acssuschemeng.9b03854 143. Zang Y, Pei F, Huang J et al (2018) Large-area preparation of crack-free crystalline microporous conductive membrane to upgrade high energy lithium–sulfur batteries. Adv. Energy Mater. 8:1802052. https://doi.org/10.1002/aenm.201802052 144. Hong X-J, Song C-L, Yang Y et al (2019) Cerium based metal–organic frameworks as an efficient separator coating catalyzing the conversion of polysulfides for high performance lithium–sulfur batteries. ACS Nano 13:1923–1931. https://doi.org/10.1021/acsnano.8b08155 145. He Y, Chang Z, Wu S et al (2018) Simultaneously inhibiting lithium dendrites growth and polysulfides shuttle by a flexible MOF-based membrane in Li–S batteries. Adv. Energy Mater. 8:1802130. https://doi.org/10.1002/aenm.201802130 146. Li M, Wan Y, Huang J-K et al (2017) Metal–organic framework-based separators for enhancing Li–S battery stability: mechanism of mitigating polysulfide diffusion. ACS Energy Lett 2:2362–2367. https://doi.org/10.1021/acsenergylett.7b00692 147. Peng S, Zhang L, Zhang C et al (2018) Gradient-distributed metal–organic framework–based porous membranes for nonaqueous redox flow batteries. Adv. Energy Mater. 8:1802533. https://doi.org/10.1002/aenm.201802533 148. Luo J, Li Y, Zhang H et al (2019) A metal–organic framework thin film for selective Mg2+ transport. Angew. Chem. Int. Ed. 58:15313–15317. https://doi.org/10.1002/anie.201908706 149. Zhang C, Shen L, Shen J et al (2019) Anion-sorbent composite separators for high-rate lithium-ion batteries. Adv. Mater. 31:1808338. https://doi.org/10.1002/adma.201808338 150. Bai S, Zhu K, Wu S et al (2016) A long-life lithium–sulphur battery by integrating zinc– organic framework based separator. J. Mater. Chem. A 4:16812–16817. https://doi.org/10. 1039/C6TA07337A 151. Cao L, Lv F, Liu Y et al (2015) A high performance O2 selective membrane based on CAU-1NH2@polydopamine and the PMMA polymer for Li–air batteries. Chem. Commun. 51:4364–4367. https://doi.org/10.1039/C4CC09281C
4 The Potential of MOFs in the Field of Electrochemical Energy Storage
153
152. Liu W, Mi Y, Weng Z et al (2017) Functional metal–organic framework boosting lithium metal anode performance via chemical interactions. Chem. Sci. 8:4285–4291. https://doi.org/ 10.1039/C7SC00668C 153. Qiao Y, He Y, Jiang K et al (2018) High-voltage Li-ion full-cells with Ultralong term cycle life at elevated temperature. Adv. Energy Mater. 8:1802322. https://doi.org/10.1002/aenm. 201802322 154. Bai S, Liu X, Zhu K et al (2016) Metal–organic framework-based separator for lithium–sulfur batteries. Nat. Energy 1:16094. https://doi.org/10.1038/nenergy.2016.94 155. Suriyakumar S, Kanagaraj M, Kathiresan M et al (2018) Metal-organic frameworks based membrane as a permselective separator for lithium-sulfur batteries. Electrochim. Acta 265:151–159. https://doi.org/10.1016/j.electacta.2018.01.155 156. Qiao Y, He Y, Wu S et al (2018) MOF-based separator in an Li–O2 battery: an effective strategy to restrain the shuttling of dual redox mediators. ACS Energy Lett 3:463–468. https:// doi.org/10.1021/acsenergylett.8b00014 157. Janek J, Zeier WG (2016) A solid future for battery development. Nat. Energy 1:16141. https://doi.org/10.1038/nenergy.2016.141 158. Miner EM, Dincă M (2019) Metal- and covalent-organic frameworks as solid-state electrolytes for metal-ion batteries. Phil Trans R Soc A 377:20180225. https://doi.org/10.1098/rsta.2018. 0225 159. Fu X, Yu D, Zhou J et al (2016) Inorganic and organic hybrid solid electrolytes for lithium-ion batteries. CrystEngComm 18:4236–4258 160. Gerbaldi C, Nair JR, Kulandainathan MA et al (2014) Innovative high performing metal organic framework (MOF)-laden nanocomposite polymer electrolytes for all-solid-state lithium batteries. J. Mater. Chem. A 2:9948–9954 161. Suriyakumar S, Gopi S, Kathiresan M et al (2018) Metal organic framework laden poly (ethylene oxide) based composite electrolytes for all-solid-state Li-S and Li-metal polymer batteries. Electrochim. Acta 285:355–364. https://doi.org/10.1016/j.electacta.2018.08.012 162. Wang Z, Wang S, Wang A et al (2018) Covalently linked metal–organic framework (MOF)polymer all-solid-state electrolyte membranes for room temperature high performance lithium batteries. J. Mater. Chem. A 6:17227–17234. https://doi.org/10.1039/C8TA05642K 163. Gao L, Chan K-Y, Li C-YV et al (2019) Highly selective transport of alkali metal ions by nanochannels of polyelectrolyte threaded MIL-53 metal organic framework. Nano Lett. 19:4990–4996. https://doi.org/10.1021/acs.nanolett.9b01211 164. Wu J-F, Guo X (2019) MOF-derived nanoporous multifunctional fillers enhancing the performances of polymer electrolytes for solid-state lithium batteries. J. Mater. Chem. A 7:2653–2659. https://doi.org/10.1039/C8TA10124H 165. Sun C, Zhang J, Yuan X et al (2019) ZIF-8-based quasi-solid-state electrolyte for lithium batteries. ACS Appl. Mater. Interfaces 11:46671–46677. https://doi.org/10.1021/acsami. 9b13712 166. Wang Z, Tan R, Wang H et al (2018) A metal–organic-framework-based electrolyte with nanowetted interfaces for high-energy-density solid-state lithium battery. Adv. Mater. 30:1704436. https://doi.org/10.1002/adma.201704436 167. Wu J-F, Guo X (2019) Nanostructured metal–organic framework (MOF)-derived solid electrolytes realizing fast lithium ion transportation kinetics in solid-state batteries. Small 15:1804413. https://doi.org/10.1002/smll.201804413 168. Chen N, Li Y, Dai Y et al (2019) A Li+ conductive metal organic framework electrolyte boosts the high-temperature performance of dendrite-free lithium batteries. J. Mater. Chem. A 7:9530–9536. https://doi.org/10.1039/C8TA12539B 169. Aubrey ML, Ameloot R, Wiers BM, Long JR (2014) Metal-organic frameworks as solid magnesium electrolytes. Energy Environ. Sci. 7:667–671 170. Wiers BM, Foo M-L, Balsara NP, Long JR (2011) A solid Lithium electrolyte via addition of lithium isopropoxide to a metal-organic framework with open metal sites. J. Am. Chem. Soc. 133:14522–14525
154
T. Devic
171. Park SS, Tulchinsky Y, Dincă M (2017) Single-ion Li+, Na+, and Mg2+ solid electrolytes supported by a mesoporous anionic Cu–Azolate metal–organic framework. J. Am. Chem. Soc. 139:13260–13263. https://doi.org/10.1021/jacs.7b06197 172. Shen L, Wu HB, Liu F et al (2018) Creating lithium-ion electrolytes with biomimetic ionic channels in metal–organic frameworks. Adv. Mater. 30:1707476. https://doi.org/10.1002/ adma.201707476 173. Yuan S, Bao JL, Wei J et al (2019) A versatile single-ion electrolyte with a Grotthuss-like Li conduction mechanism for dendrite-free Li metal batteries. Energy Environ. Sci. 12:2741–2750. https://doi.org/10.1039/C9EE01473J 174. Zhu F, Bao H, Wu X et al (2019) High-performance metal–organic framework-based single ion conducting solid-state electrolytes for low-temperature lithium metal batteries. ACS Appl. Mater. Interfaces 11:43206–43213. https://doi.org/10.1021/acsami.9b15374 175. Ashraf S, Zuo Y, Li S et al (2019) Crystalline anionic germanate covalent organic framework for high CO2 selectivity and fast Li ion conduction. Chem. Eur. J. 25:13479–13483. https:// doi.org/10.1002/chem.201903011 176. Xu W, Pei X, Diercks CS et al (2019) A metal–organic framework of organic vertices and polyoxometalate linkers as a solid-state electrolyte. J. Am. Chem. Soc. 141:17522–17526. https://doi.org/10.1021/jacs.9b10418 177. Zhang L, Hu YH (2011) Strong effects of higher-valent cations on the structure of the zeolitic Zn(2-methylimidazole)2 framework (ZIF-8). J. Phys. Chem. C 115:7967–7971 178. Demir-Cakan R, Palacin MR, Croguennec L (2019) Rechargeable aqueous electrolyte batteries: from univalent to multivalent cation chemistry. J. Mater. Chem. A 7:20519–20539. https://doi.org/10.1039/C9TA04735B 179. Kim H, Hong J, Park K-Y et al (2014) Aqueous rechargeable Li and Na ion batteries. Chem. Rev. 114:11788–11827 180. Wu D, Guo Z, Yin X et al (2014) Metal–organic frameworks as cathode materials for Li–O2 batteries. Adv. Mater. 26:3258–3262. https://doi.org/10.1002/adma.201305492 181. Yan W, Guo Z, Xu H et al (2017) Downsizing metal–organic frameworks with distinct morphologies as cathode materials for high-capacity Li–O2 batteries. Mater Chem Front 1:1324–1330. https://doi.org/10.1039/C6QM00338A 182. Shinde SS, Lee CH, Jung J-Y et al (2019) Unveiling dual-linkage 3D hexaiminobenzene metal–organic frameworks towards long-lasting advanced reversible Zn–air batteries. Energy Environ. Sci. 12:727–738. https://doi.org/10.1039/C8EE02679C 183. Li S, Dong Y, Zhou J et al (2018) Carbon dioxide in the cage: manganese metal–organic frameworks for high performance CO2 electrodes in Li–CO2 batteries. Energy Environ. Sci. 11:1318–1325. https://doi.org/10.1039/C8EE00415C 184. Jia H, Billmann B, Onishi H et al (2019) LiPF6 stabilizer and transition-metal cation scavenger: a bifunctional bipyridine-based ligand for lithium-ion battery application. Chem. Mater. 31:4025–4033. https://doi.org/10.1021/acs.chemmater.9b00555 185. Malik R, Loveridge MJ, Williams LJ et al (2019) Porous metal–organic frameworks for enhanced performance silicon anodes in lithium-ion batteries. Chem. Mater. 31:4156–4165. https://doi.org/10.1021/acs.chemmater.9b00933 186. Zhou W, Lv S, Liu X et al (2019) A directly grown pristine Cu-CAT metal–organic framework as an anode material for high-energy sodium-ion capacitors. Chem. Commun. 55:11207–11210. https://doi.org/10.1039/C9CC06048K 187. Sun J, Guo L, Sun X et al (2019) Conductive Co-based metal–organic framework nanowires: a competitive high-rate anode towards advanced Li-ion capacitors. J. Mater. Chem. A 7:24788–24791. https://doi.org/10.1039/C9TA08788E
Chapter 5
Carbon Capture Using Metal–Organic Frameworks Ram R. R. Prasad, Qian Jia, and Paul A. Wright
5.1
Introduction
Anthropogenic emissions of the greenhouse gas CO2, largely but not exclusively from the combustion of fossil fuels, are widely acknowledged to be giving rise to an increase of the global level of CO2 (currently 411 ppm [1]), global warming and climate change, [2] so that the reduction of its production and release into the atmosphere is required with great urgency. Intensive efforts are therefore being made to decarbonise many of the processes emitting CO2, including the production of electricity and hydrogen from renewable sources and the development of automobiles powered by lithium batteries or fuel cells. However, there are many areas where, at least in the medium term, CO2 will continue to be produced and released on a great scale, and means are required to capture and store it [3–5]. Centralised power generation throughout the world, for example, in coal- and gas-powered power stations, is the most visible example of CO2 emission, but various industrial sources also contribute in a significant way. Coal-fired power stations still contribute 45% of the global output of energy-related emissions [5], but increasing use of natural gas suggests it will overtake coal for power generation by 2040 [6]. In many economies, such as in the USA, the rapid changeover from coal to gas means that the latter is already producing more CO2 than the former [7]. Furthermore, sources of natural gas (required for relatively clean energy generation) and biogas (generated from waste) also contain, in addition to methane, major amounts of CO2 [8, 9]. Upgrading of these fuels to enhance their calorific value requires CO2
R. R. R. Prasad Department of Chemistry, Dainton Building, The University of Sheffield, Sheffield, UK e-mail: r.ramprasad@sheffield.ac.uk Q. Jia · P. A. Wright (*) EaStCHEM School of Chemistry, University of St Andrews, St Andrews, UK e-mail: [email protected]; [email protected] © Springer Nature Switzerland AG 2021 P. Horcajada Cortés, S. Rojas Macías (eds.), Metal-Organic Frameworks in Biomedical and Environmental Field, https://doi.org/10.1007/978-3-030-63380-6_5
155
156
R. R. R. Prasad et al.
removal. Finally, capture of CO2 from the air (direct air capture) by solid adsorbents can become an important technology [10, 11]. It can be a solution to the niche problem of keeping CO2 levels low in confined spaces or in medical recirculating air, as well part of zero carbon strategies and of a future large-scale remediation process of reducing atmospheric CO2 levels. Each of these diverse processes produces a CO2-containing gas stream, from which the CO2 must be separated, concentrated and produced at high purity. It can then be used as a feedstock for conversion to more valuable chemical products or fuels or transported for storage underground in geological strata [3–5]. The use of solid adsorbents to perform the ‘carbon capture’ step is of great interest, because they have the potential to give easily handled, versatile and energy-efficient alternatives to existing aqueous amine-based chemisorption processes. However, the optimum type of adsorbent and chemical engineering solution will depend strongly on the gas composition and its temperature and total pressure. Many adsorbents have been investigated for these purposes, including porous carbons, zeolites and aminefunctionalised mesoporous silicas, and approaches to the rapid screening of these different materials have been developed [12–14]. Here we consider the potential of MOFs as adsorbents for carbon capture. MOFs have many strong attributes as solid adsorbents, which arise from their high surface areas and their chemical versatility; however, reservations remain concerning their widespread application due to concerns over long-term stability, particularly in the presence of water or corrosive impurity gases, and their cost, issues that are under continual consideration. One area where CO2 separation performance rather than cost is the deciding factor is in membrane separation, and this is why mixed matrix membranes are such an active area of research, as covered in Sect. 5.5. There are many excellent and thorough reviews and even books on the subject of carbon capture and of the application of MOFs in such processes [15–18]. Here we will illustrate key recent advances in this field, paying particular attention to the role of MOF structure in controlling carbon capture performance under different conditions. We will first summarise the compositions of CO2-containing gas streams that are the targets of carbon capture and then address some of the fundamentals of the use of solid adsorbents in processes to achieve this, including in pressure and temperature swing processes. The attributes and limitations of MOFs for these technologies are then discussed in the light of their requirements. Membrane-based methods offer an alternative approach to the separation, and these are addressed in a later part of this chapter. Bulk use of pure captured CO2, for example, in supplying greenhouses, in feeding algae or in drink carbonation, can be important, but the scale of CO2 emissions suggests that conversion to fine or even commodity chemicals will not have a major volumetric effect in their reduction, although it remains of fundamental and even commercial interest [17, 18]. Rather, conversion to fuels can be significant, where recycling carbon could reduce the rate of fossil fuel use. While application of MOFs in catalytic technologies to generate fuel can be of interest, they must compete with a range of other robust industrial catalysts well developed for small molecule conversion, and so we will deal only briefly with this.
5 Carbon Capture Using Metal–Organic Frameworks
5.2 5.2.1
157
Targets for Carbon Capture: CO2-Containing Gas Streams Power Generation
The main stationary sources of anthropogenic CO2 emissions are power generation, directly or indirectly from fossil fuels or, on a smaller scale, from biogas. Industrial generation of hydrogen also gives CO2 emissions, as do a variety of diverse industries. Power generation can be via one of three main routes, direct combustion of fossil fuel, generation of hydrogen for combustion by gasification and reforming of coal or gas or burning fuel in oxygen (oxy-fuel process). These are illustrated in Scheme 5.1, and representative compositions from post- and precombustion gases are given in Table 5.1. Direct combustion of coal is increasing globally, even as it is phased out in many European countries. The flue gas contains 10–15% CO2, mainly in N2 but with water and additional impurities, including sulphur and nitrogen oxides (SOx and NOx). By comparison, the combustion of natural gas (mainly methane, CH4) is a cleaner process, with the emissions containing around 3% CO2. In each case, the total pressure is close to atmospheric. The goal here would be to include carbon capture in the design of new power plants and retrofit existing ones. The cost associated with
Scheme 5.1 The roles of separation processes and carbon capture in power plants operating, directly or indirectly, on fossil fuel. (Figure adapted with permission from [16]. Copyright 2012 American Chemical Society)
158
R. R. R. Prasad et al.
Table 5.1 Representative compositions of post-combustion flue gases from various types of power plants and a typical precombustion flue gas after the water–gas shift reaction Post-combustion: concentration (%v) of each gas from different power plants Molecule Pulverised coal-fired Coal-fired IGCC (integrated gasification cycle) 76 66 N2 CO2 11 7 H2O 6 14 O2 6 12 SO2 10–200 ppmw NOx 500–800 ppmw 10–100 ppmw Precombustion Molecule Pulverised coal 3.9 N2 CO2 37.7 H2O 0.14 H2 55.5 CO 1.7 H2S 0.4 CH4 0
Gas-fired 76 3 6 14 10–300 Methane –– 15–25 70–80 1–3 –– 3–6
Data from [19, 20]
this process is variously estimated to add 20–25% to the total [5, 16]. In some current projects, part of the carbon capture cost is recovered by using the CO2 in enhanced oil recovery. The generation of H2 for combustion has the advantage that the combustion itself does not generate CO2. However, H2 is generated by gasification of fossil fuel and subsequent low temperature shift reactions by which H2 containing CO2 and CO is produced (as well as impurities stable in a reducing environment, such as H2S). In these ‘precombustion’ gases, the streams are at high pressure (5–40 bar) so that the partial pressure of CO2 is high, and additional pressurisation is not required. The related industrial production of hydrogen for processes such as ammonia synthesis, Fischer–Tropsch formation of hydrocarbons and the synthesis of methanol is well established, and CO2 adsorption technology is already an integral part of this. The oxy-fuel process, in which fuel is burnt in pure O2, generates a pure CO2 stream and therefore no CO2 separation step, but it does requires O2 separation and purification from air, and therefore adsorption and separation from air are very much part of this process. This air must be purified first by CO2 removal. The direct capture of CO2 from air (where the CO2 level is only ca. 400 ppm, with considerable H2O as well as O2, N2 and permanent gases) is also being investigated more widely to generate pure CO2 for applications and as a route to offset carbon emissions elsewhere in industries.
5 Carbon Capture Using Metal–Organic Frameworks
5.2.2
159
Natural Gas and Biogas Upgrading
Methane is a relatively clean fossil fuel, strongly preferred to coal for power plants, a potential replacement fuel for petrol and diesel for automobile engines and a widespread domestic heating fuel. It is mainly sourced from natural gas (associated with oil from the maturation of organic-rich source rocks, derived from deep burial of coal-bearing strata or derived from bacterial action on organic sediments at relatively shallow depths (300. In both cases, this enhancement was attributed to the attraction of CO2 to the basic NH2 sites at low pressures, but the exact species remained unclear.
176
R. R. R. Prasad et al.
Fig. 5.12 (a) Representation of mmen-Mg2(dobpdc) viewed down the channel, (b) mmenappended 1D SBU of Mg2(dobpdc) before and after cooperative insertion of the CO2, (c) CO2 adsorption isotherms from 25 C (blue) to 75 C (red), (d) the schematic representation of the multiple steps involved in the cooperative insertion and (e) schematic summary of 40 C CO2 step pressures and regeneration temperatures under 1 bar of CO2 for various alkyldiamine-appended Mg2(dobpdc) frameworks. (Figures reprinted or adapted with permission from [70, 73]. Copyright 2017 American Chemical Society)
To develop this approach of stable appending diamines in MOFs to achieve low pressure CO2 uptake, coordinatively unsaturated metal cations are required, and for high uptakes the density of these should be high. As described previously, MOF-74 has a very high density of coordinatively unsaturated sites, but the 1.1 nm pore size is insufficient to allow inclusion of one diamine per CUS without steric crowding. Long et al. therefore chose to prepare a larger, isoreticular, version of MOF-74 using the dioxybiphenyldicarboxylate (dobpdc) linker, giving a Mg2(dobpdc) MOF with a high CUS density and large enough channels to adsorb a range of diamines stoichiometrically at Mg2+ sites, including mmen [68]. This mmen-appended MOF showed remarkable properties, for not only is it much more stable to water than the parent MOF, but more significantly it shows stepped CO2 isotherms with a steep and low pressure CO2 uptake corresponding to one molecule per amine (and so per Mg2+) (Fig. 5.12). This reaches saturation at ca. 3.5 mmol g1 and achieves 2.5 mmol g1 at 400 ppm CO2 at 25 C: it has a heat of adsorption of 71 kJ mol1. The uptake at 0.4 mbar on mmen-Mg2(dobpdc) is 15 times higher than on Mg2(dobpdc) without amine. The adsorption step on mmen-Mg2(dobpdc) occurs
5 Carbon Capture Using Metal–Organic Frameworks
177
at progressively higher PCO2 as the temperature is raised, in line with the Clausius– Clapeyron relationship: ln p ¼
ΔH ads 1 þc R T
ð5:1Þ
This stepped shape is of particular interest for PSA, because the working capacity is large and achieved over a small difference between adsorption and desorption pressures. Furthermore, the stepwise equilibrium uptake at a constant CO2 pressure drops steeply with temperature, which would enable efficient TSA to be achieved over a temperature swing of a few degrees. A detailed study of the adsorption mechanism carried out on a series of mmenM2(dobpdc) MOFs (M ¼ Mg, Mn, Fe, Co, Ni and Zn) [69] revealed the formation of ammonium carbamate species via cooperative insertion of CO2 by the mmenappended frameworks. The adsorption process was further elucidated by a combination of diffraction, spectroscopy and modelling [70, 71]. There is an insertion of CO2 at a M–N bond to form a carbamate species and an ammonium species on a neighbouring diamine, simultaneously (Fig. 5.12). The uptake therefore corresponds to one CO2 molecule per M site. The insertion proceeds cooperatively along the internal surface of the MOF along the c-axis of the structure to give a chain of infinite ammonium carbamate species. The higher degree of cooperativity results in the stepped adsorption isotherm. The importance of the CO2 insertion at the M-N bond is illustrated by the dependence of the uptake behaviour on the identity of the divalent metal cation so that while Mg results in the step occurring at the lowest pCO2 of those analogues that do give the isotherm step (Mg, Co, Zn), the Ni-analogue exhibits no step at all. Remarkably, there are strong similarities between the mechanism of CO2 uptake on Mg2(dobpdc) and on the enzyme Rubisco, which is responsible for CO2 fixation in photosynthesis [72]. Again, the active site is Mg, coordinated to five O atoms and a reactive aliphatic amine, which adsorbs CO2 resulting in the formation of bound carbamates. Subsequently, the effect of varying the type of diamine appended was investigated [70, 71, 73–75], with the aim of controlling the pressure of the step relevant to CO2 capture from flue gas streams and improving the robustness of the amine–MOF complex to repeated thermal cycling. Diamines that contain both primary (1 ) and secondary (2 ) amines give steps at the lowest pCO2, with minimum hysteresis, and the largest are the most thermally stable. Due to stronger H-bonding interactions, diamines consisting of both 1 and 2 amine groups were found to form more stable ammonium carbamate chains than diamines with only 2 amine groups, with a similar mechanism to that determined by mmen. Systematic tuning of the diamine backbone and the primary, secondary and tertiary amine structure results in a wide range of step pressures relevant to CO2 capture from flue gas streams (Fig. 5.13). However, Mg2(dobpdc) MOFs containing larger diamines give two cooperative insertion steps (at 0.5 and 1 CO2 per Mg). This two-step adsorption results from unfavourable steric interactions encountered during the CO2 uptake by diamines
178
R. R. R. Prasad et al.
Fig. 5.13 (a) Mg2(dobpdc) and (b) Mg2(pc-dobpdc) viewed along the channel axis; associated linkers dobpdc4 and pc-dobpdc4. The CO2 adsorption isobars (at 1 atmosphere) are shown for a series of linear 1 and 2 -alkylethylenediamine-appended variants of (c) Mg2(dobpdc) and (d) Mg2(pc-dobpdc). (Figures reprinted with permission from [74], copyright 2018 Royal Society of Chemistry)
appended to Mg2+ cations on neighbouring chains in the a-b plane of Mg2(dobpdc). As a result of these steric effects, neighbouring amine sites interact sequentially, with the unwanted effect of increasing the size of the temperature (or pressure) swing required to achieve the full working capacity. The diamine aminomethylpiperidineappended Mg2(dobpdc) gives a good compromise [75]. The two adsorption steps are close together at a low pCO2, and the material is thermally robust and shows an excellent performance in moist CO2 streams. The steric effects described above result from a subtle detail of the crystal structure. The channels of Mg2(dobpdc) show threefold rather than sixfold symmetry in projection, with the walls bowed alternatively concave and convex to the channels (Fig. 5.13). This results in the appended amines being pushed together and constraints as CO2 is inserted at neighbouring cation sites. In an elegant example of structural control of properties made possible by the reticular chemistry of MOFs, a related version of the MOF can be prepared using the 3,30 -dioxidobiphenyl-4,40 -dicarboxylate (pc-dobpdc), Mg2(pc-dobpdc) [74]. This MOF possesses channels that are more regularly hexagonal in cross section. In this case, large amines can be included in a way that does not lead to steric constraints during CO2 insertion at metal sites to give ammonium carbamates, and single-step isotherms with high uptakes are again observed.
5 Carbon Capture Using Metal–Organic Frameworks
179
This work represents a major step forward in the development of sorbents for carbon capture in post-combustion streams from natural gas-fuelled power stations. Furthermore, the sorbent has already been prepared on a kg scale for subsequent pilot scale tests.
5.4.4
Air Capture: Ultramicroporous and Biomimetic MOFs
Some of the materials described in the post-combustion section show such strong uptake of CO2 that they might also be used to capture CO2 directly from the air, even without pressurisation, and in the presence of water vapour. Removal of CO2 from air before separation of O2 and N2 is currently performed over zeolites, after drying it first, but if this process is to be performed to reduce atmospheric CO2, improved adsorbents will be required to treat large volumes of moist air, without costly predrying. Current examples of air capture use existing amine solution absorption or even direct reaction with lithium hydroxide if only small quantities are to be removed. At the conditions of ambient air, 400 ppm and room temperature, for example, the MOFs that show the most promising uptakes are the ultramicroporous NbOFFiveNi-1, which interacts by strong physisorption, and ‘biomimetic’ MOFs that chemisorb CO2. These include CFA-1 and MFU-4ll that possess Zn-OH sites, in an analogous way to α-carbonic anhydrase, and diamine-appended Mg2(dobpdc). Salient information is included in Table 5.4, where their properties are compared to a recently reported hierarchically porous silica with poly(ethyleneimine) included at 2.6 g/g(SiO2) [76]. While the uptakes are lower in the MOFs than in the best amine-impregnated silicas, it is possible that the latter will suffer from similar amine loss and oxidation problems to those experienced by amine solutions.
Table 5.4 Performance of four leading candidates for direct capture of CO2 from air MOF NbOFFiveNi-1 CFA-1 mmen-mgdobpdc PEI-SiO2-H
Description Ni(pyrazine)2NbOF5 Zinc benzotriazolate Isoreticular diamine-appended mg (dobpdc) Poly(ethyleneimine)-loaded hierarchical porous silica (2.6 g/g)
*Reported for aqueous amine solution
Uptake/mmol g1 (0.4 mbar, 298 K) 1.3 (dry)
Q/kJ mol1 54
Ref. [60]
2.2 (dry only) 2.0 (dry)
45–72 70
[63] [75]
2.6 (dry) 3.4 (wet)
105*
[76]
180
5.4.5
R. R. R. Prasad et al.
Biogas and Natural Gas Upgrading
The removal of CO2 from natural gas and biogas primarily requires the separation of CO2 from methane. In natural gas, the CO2 is present at a few %, whereas in biogas the concentration is much higher, at ca. 40%. Therefore, CO2/CH4 separation is the required separation. A range of technologies have been commercialised, including amine absorption and adsorption on activated carbons, but an inexpensive MOF can also find application. The two molecules are readily discriminated on the basis of electrostatics (quadrupolar CO2 interacts more strongly with polar sites) and size (CH4 vs. CO2 sizes are ca. 3.8 Å vs. 3 Å). While activated carbons cannot perform equilibrium separations (the selectivity is only 2–4), careful control of pore size can introduce kinetic selectivity. By contrast, the usual approach for MOFs is to tailor the interaction strength, because the flexibility of MOFs makes difficult the accurate prediction of effective pore size from crystal structures. This approach is similar to those found to be effective for uptake of CO2 at relatively low partial pressures in post-combustion or even air capture applications. One example, among many, of a MOF possessing adsorption sites claimed to be suitable for the separation is the iron tetrachlorobenzene dicarboxylate LIFM-26, which has been examined for adsorption of a range of hydrocarbons and fuel-related gases. A selectivity of 36 was reported, close to that observed for zeolites, with a competitive uptake at 298 K and 0.4 bar of ca. 2 mmol g1 [77]. Investigation of the literature indicates many of the materials reported above will perform well here, but the separation is not demanding and many sorbents, including zeolites and engineered porous carbons, can also be interesting candidates, particularly if the gases are dried.
5.4.6
Summary of MOFs as Solid Adsorbents for Carbon Capture
It is clear from the literature that carbon capture by adsorption on MOFs has attracted the attention and inspired the research efforts of chemists and chemical engineers like few other recent challenges. The technology of PSA and TSA is already well developed for many applications using solid adsorbents, including hydrogen purification, so that it should in principle be possible for MOFs to give improved performance for this and many other CO2 separation processes without the need for redesigned processes. However, the cost and relative fragility of MOFs has inhibited their uptake as replacement sorbents. Nevertheless, recent breakthroughs do indicate that in some specific cases of carbon capture, such as post-combustion capture and even direct air capture, MOFs are beginning to find their place. Separation of CO2 via the use of membranes is an alternative approach to PSA and TSA. While less well established technologically than adsorption, it offers significant future advantages, and the development of MOF-containing membranes is described in the next section.
5 Carbon Capture Using Metal–Organic Frameworks
5.5
181
MOFs as Fillers for Mixed Matrix Membranes
5.5.1
Introduction
Membrane-based separations of CO2 offer an alternative technology to pressure (or temperature) swing adsorption or chemisorption in amine solutions [78]. If a practicable setup can be achieved, then membrane technology offers a considerable energy saving over the other methods, estimated at around 50%. Furthermore, membrane technologies have already been commercialised, either for air dehydration or natural gas treatment over polymeric membranes [79]. One of the main issues concerning porous polymer membranes is that, over time, ageing leads to plasticisation and loss of performance, especially for the more permeable polymers. Porous framework solids such as zeolites and more recently MOFs have been investigated in supported membranes, where they have been grown on functionalised porous substrates. Their well-defined porosity offers great scope for design and optimisation for specific gas separations via the molecular sieving principles well established for zeolites. However, for the separation of gases, the difficulty of achieving crack-free membranes is a serious limitation, for pinholes strongly reduce selectivity. There is therefore a strong incentive to develop mixed matrix membranes (MMMs) where nanoporous fillers improve the permeability and selectivity of a polymeric matrix that has excellent properties of workability. MOFs are preferred to zeolites as a filler, because the organic linkers are more compatible with the polymer, and both MOF and polymer can be tuned to enhance this compatibility. There is a fast-developing literature in this area, covered by some detailed and helpful reviews [80]. Here, we will discuss recent advances in MOF-containing mixed matrix membranes with an emphasis on material selection and chemical tuning and optimisation of performance. Before discussing these elements, we will briefly consider the theory of the gas transport mechanism and the practice of membrane evaluation. We will finish by summarising some recent advances and future directions in MOF-containing mixed matrix membranes.
5.5.2
Fundamentals of Gas Transport Through Membranes
5.5.2.1
Mechanism
Currently, the solution–diffusion mechanism is the most widely accepted theoretical framework to describe gas transport through polymeric membranes [81]. According to the theory, gas transport through dense polymer membranes is governed by the diffusion coefficient (Di) and solubility coefficient (Si), and the product of these two coefficients is defined as permeability (Pi), which can be expressed in Barrer (1 Barrer ¼ 1010 cm3 (STP) cm cm2 s1 cmHg1) [82]:
182
R. R. R. Prasad et al.
Pi ¼ Di Si
ð5:2Þ
Permeability can also be expressed as the product of gas flux and membrane thickness divided by transmembrane pressure difference. In practical application, gas permeability is a key factor to evaluate the efficiency of gas transport through membranes, because higher permeability means a smaller required area for a given gas flow and a lower cost for the membrane manufacture. Another important factor is selectivity (αi/j) for characterising membrane discrimination between gas pairs, and this is expressed as the ratio of permeabilities of a gas pair. The expression for selectivity can also be written:
αi=j
Di ¼ Dj
Si Sj
ð5:3Þ
Here, the ratio of two diffusion coefficients can be regarded as mobility selectivity and correlates with gas kinetic diameters, and the ratio of solubility coefficients represents the relative condensabilities of two gas species. As discussed above, membrane performance is a product of the diffusion coefficient and solubility coefficient, and both factors vary significantly in different materials. In polymeric materials, these factors depend largely on whether a polymer undergoes a glass phase transition at its inherent glass transition temperature (denoted as Tg) [83]. Below Tg, polymer chains are fixed to the backbone resulting in tough and rigid glassy polymers with high elasticity; above Tg, the thermal energy is sufficient to enable polymer chains to rotate, contributing to soft and elastic rubbery polymers. The diffusion coefficient is a dominant factor in glassy polymers, and it decreases dramatically with an increase of molecule size, while there is a slight decrease observed in rubbery polymers (Fig. 5.14) [84]. By contrast, the solubility
Fig. 5.14 (a) Diffusion coefficients as a function of molecule molar volume in rubbery polymer (polydimethysiloxane: PDMS) and a glassy polymer (polysulfone: PSF); (b) solubility coefficients for a typical glassy polymer (PSF) and a rubbery polymer (PDMS) as a function of critical temperature Tc of different gas species. (Figures adapted with permission from [85]. Copyright 1999 American Chemical Society)
5 Carbon Capture Using Metal–Organic Frameworks
183
coefficient outweighs the diffusion coefficient in rubbery polymers, and solubility coefficients change in line with the variation of molecular size because the larger molecules are more condensable than the smaller ones. To be noted, solubility coefficients for a specific gas species almost keep constant in both types of polymers (Fig. 5.14) [81, 85]. In general, glassy polymers are favourable for the permeation of small gas molecules, while rubbery polymers allow larger molecules to permeate preferably [84].
5.5.2.2
Robeson Plots
In 1991, Robeson summarised the gas permeation properties through various polymeric membranes and revealed most polymeric membranes suffer from a trade-off between selectivity and permeability, giving an upper bound and described by the following relation [86]: Pi ¼ kαni=j
ð5:4Þ
where k and n are invariant for specific gas pairs and are referred to as the front factor and the upper bound slope [79]. In 2008, the original correlation was revisited, and minor shifts were observed in the upper bound primarily due to a change in k [87]. Concerning CO2 capture applications, the upper bound relationships of three gas pairs of interest (i.e., CO2/CH4, CO2/N2 and CO2/H2) are shown in Fig. 5.15, and the performance of some representative polymeric membranes are highlighted. Details of these polymers are given later in Sect. 5.5.3.2. To overcome the upper bound limit, various approaches have been adopted, including surface modification, facilitated transport, polymer blending and mixed matrix membranes [79]. In mixed matrix membranes, both fillers and polymers contribute to gas permeation, and the conventional solution–diffusion mechanism cannot model it well. Instead, the Maxwell model is applied, and the overall gas permeability in two phases (i.e. effective permeability) can be described: Peff ¼ Pc
Pd þ 2Pc 2Φd ðPc Pd Þ Pd þ 2Pc þ Φd ðPc Pd Þ
ð5:5Þ
where Pc and Pd are signified as gas permeabilities in the continuous phase (polymer matrix) and disperse phase (filler particle) and Φd represents the volume fraction of filler. However, the Maxwell model is only valid with low filler content (usually below 30%), because when the loading ratio exceeds a certain value, the filler particles will be in contact and can form a continuous channel. Under these conditions, other models have been developed for membranes with broader range of Φd, including Lewis-Nielsen [88], Felske [89] and modified-Maxwell [90], and their predictions are in good agreement with experiments.
184
R. R. R. Prasad et al.
Fig. 5.15 (a) Upper bound correlation for CO2/CH4, (b) for CO2/N2, (c) for H2/CO2. The shaded parts represent the dataset of reported polymeric membranes for each gas pair separation, and positions of typical polymer membranes are marked in the plot according to their permeability and selectivity. (Figures adapted with permission from [87]. Copyright 2008 Elsevier)
5.5.2.3
Testing Membranes
Polymer-based membranes can be made in different forms. Most measurements are made on self-supporting thin membranes prepared by casting a polymer from a solution, but asymmetric membranes (supported on a porous substrate) are also of interest, and for commercial application, hollow fibre membranes (discussed later in Sect. 5.6) are often the most practical solution. Typically there are three routes to prepare a mixed matrix membrane on the laboratory scale. These are (i) dissolving a polymer in a solvent and adding filler particles into the polymeric solution [91], (ii) dispersing filler particles in a solvent before addition of polymer into the mixture [92], and (iii) dissolving polymer and dispersing filler particles in a solvent separately and then mixing them [93]. To obtain a well-mixed dispersion, the mixture is subjected to stirring and sonication, then cast on a flat plate and dried overnight [94]. In membrane characterisation, gas permeation is a major technique to determine membrane permeability and selectivity. Membrane permeability is characterised by pure gas permeation, which is operated either under constant pressure–variable
5 Carbon Capture Using Metal–Organic Frameworks
185
Fig. 5.16 Schematic diagrams of (a) constant pressure–variable volume system, (b) constant volume–variable pressure system and (c) mixed gas permeation system (R regulator, P pressure transducer, MFC mass-flow controller, GC gas chromatograph). (Figures adapted with permission from [95]. Copyright 2006 Springer Nature)
volume or constant volume–variable pressure (Fig. 5.16) [95]. Under constant pressure–variable volume conditions, the gas flux is measured at steady-state condition by a downstream flow meter. In a constant volume–variable pressure system, the permeation cell and connected pipes should be evacuated before the measurement. When the measurement starts, the rate of pressure rise in the downstream side is registered. As for membrane selectivity, the measurement requires mixed gas permeation, which resembles pure gas permeation except for the addition of a gas chromatograph to detect gas components in the permeate.
5.5.3
MOF-Based Mixed Matrix Membranes
5.5.3.1
Introduction
MOF-based mixed matrix membranes have been investigated intensively over the last two decades. For a typical MMM, the polymer phase determines the minimum membrane performance, and the aim of the inclusion of MOF filler particles is to increase the permeance, the selectivity or both [96]. Ideally, the surface MOF
186
R. R. R. Prasad et al.
Fig. 5.17 Schematic representation of gas transport through a MMM with ideal morphology for (a) CO2/N2 and CO2/CH4 separation, and (b) H2/CO2 separation
apertures should be accessible to gas molecules without any polymer chain blockage occurring around the particle. In CO2/N2 or CO2/CH4 separation, CO2 is the permeant species and should be able to diffuse through the filler. As filler particles are generally isolated within the polymer matrix, CO2 also needs to be able to travel through the polymer (Fig. 5.17). For H2/CO2 separation, by contrast, the fillers must facilitate H2 transport instead of CO2, and close packing of filler particles should enable H2 to pass through the membrane with reduced pathlength through the polymer matrix (Fig. 5.17), while the permeation rate of CO2 through the filler is low. To optimise both types of membrane separation, efforts have been made to select appropriate MOF/polymer pairs and to modify the MOF/polymer interface.
5.5.3.2
Polymer Choice
In membrane synthesis, five commonly used polymers include PBI, Matrimid®, Pebax-1657, PIM-1 and 6FDA-DAM (6FDA ¼ 4,40 -(hexafluoroisopropylidene)diphthalic anhydride, DAM ¼ 2,4,6-trimethyl-1,3-phyenylenediamine). In Table 5.5, we list their molecular structure and some related properties about membrane fabrication. According to Fig. 5.15, PBI- and Matrimid®-based membranes can achieve a high selectivity but relatively low permeability, while PIM-1 and 6FDA-DAM are favourable for gas transport but not selective to CO2. By comparison, Pebax-1657 features a modest selectivity as well as a desirable permeability, especially for CO2/CH4 and CO2/N2 separations.
5.5.3.3
Choice of MOF Fillers
The choice of MOF/polymer pairs for membranes therefore depends on the targeted separation and the chemistry and porosity of the polymers. The functionality of the MOF should be compatible with that of the polymer to ensure a good interfacial
5 Carbon Capture Using Metal–Organic Frameworks
187
Table 5.5 Molecular structure and properties of polymers frequently used for preparation of MMMs Polymer PBI
Molecular structure
Properties High thermal stability, good moisture and chemical resistance
Matrimid®
Thermoplastic polyimide; durable and tough polymer with excellent adhesion, chemically resistant and good thermal stability
Pebax1657
Block copolymer consisting polyether blocks and polyamide backbone; the rigid polyamide segment provides a good mechanical property; and flexible polymer chains facilitate gas transport
PIM-1
Polymer of intrinsic microporosity: rigid, sterically constrained backbone that can efficiently prevent chain packing. Often high permeance
6FDADAM
An aromatic polyimide with good thermal stability and chemical resistance, and its intrinsic structure prohibits the polymer chain packing which is favourable for gas transport
contact. Where CO2 is the permeant, the MOF should provide a fast and selective diffusion pathway in a polymer of low-to-moderate permeability, while in a highly permeable polymer such as a PIM it might act to enhance CO2 uptake and prevent aging. Where H2 is the permeant, the MOF should allow H2 but not CO2 through, and so a very narrow pore MOF in a poorly permeable MOF is a promising combination.
188
R. R. R. Prasad et al.
CO2 as Permeant. MOF structures can be tailored with different functional groups, which influence membrane separations, especially for CO2/N2 and CO2/CH4 mixtures. As CO2 has a larger quadrupole moment than N2 (13.4 1040 Cm2 cf. 4.72 1040 Cm2) (Table 5.2) and CH4 is nonpolar, polar functional groups such as -NH2, -OH, -NO2 and -CHO in MMMs enable membranes to discriminate CO2 from the other species by increasing the ‘solubility’ of CO2 in the filler. For example, zeolitic imidazolate frameworks (ZIFs), a family of MOFs comprising metal cations (commonly Zn2+) tetrahedrally coordinated by the N atoms of imidazolate linkers, can be functionalised by polar functional groups by incorporation of different imidazolate linkers. For example, whereas ZIF-8, prepared with methylimidazole, has low polarity, ZIF-94 (Fig. 5.18), which incorporates the methyl, aldehyde-functionalised imidazolate, interacts more strongly with CO2 and has been used successfully in mixed-membrane studies [97]. In addition, the incorporation of functional groups is an effective approach to enhance MOF/polymer interfacial adhesion and give rise to a defect-free membrane. MOFs functionalised with accessible amino groups exhibit good compatibility with polymers such as polyimide and polysulfone [98], which enables incorporation of a higher content of MOF particles and a resulting improvement in gas permeability [99]. Recent advances in the design of MOF functionality for use in MMMs will be discussed in Sect. 5.5.3.4.
Fig. 5.18 The imidazole linkers and zeolitic imidazolate structures of (a) ZIF-8, (b) ZIF-94, and (c) ZIF-7. H atoms omitted for clarity. The structure of ZIF-94 is courtesy of Prof. Shi Qi, Taiyuan University of Technology, China. Hydrogen atoms are omitted for clarity in the structures. Colour code: Zn blue, C black, N green and O red
5 Carbon Capture Using Metal–Organic Frameworks
189
The use of molecular sieving to effect separation of CO2 requires that it should be much smaller than admixed gases. The cross-sectional diameter of CO2 (3.0 Å) is slightly smaller than that of N2 (3.2 Å) and appreciably smaller than that of CH4 (3.8 Å), but few MOFs display sharp molecular size cutoffs [100]. Although the crystallographically determined pore size of ZIF-8 (3.4 Å) should show effects in excluding N2 or CH4, this does occur because the imidazole groups are mobile and can hinge and enlarge the aperture [101]. A different approach to enhance selectivity to CO2 by including ionic liquids into the pores, has been reported [102], but in general the introduction of chemical functionality into MOFs is more effective to separate CO2 from N2 or CH4. H2 as Permeant. Unlike CO2/CH4 and CO2/N2 separations, H2 is the preferable permeant in H2/CO2 separation. Gas separation can in principle be achieved by exploiting MOFs with a pore size localised between the two-centre Lennard-Jones diameters of H2 (2.6 Å) and CO2 (3.0 Å) [103]. For example, while ZIF-7 has the same topology as ZIF-8, the bulky benzimidazole groups in ZIF-7 (3.0 Å) reduce the window size (Fig. 5.18) resulting in more H2-selective membranes [104]. Along these lines, other MOFs have been prepared with bulky linkers to introduce molecular sieving effects and improve H2 selectivity [105]. Furthermore, where molecular sieving is a mechanism to enhance H2 selectivity, the inclusion of platy nanosheets aligned perpendicular to the direction of gas transport can improve selectivity, if they contain small pores through which H2 and not CO2 can pass. The CO2 must then find a much longer path through a relatively impermeable polymer matrix [106]. MOF Particle Size and Morphology. Regardless of the adsorption and separation properties of any MOF filler, inclusion in a MMM requires that it should be preparable as nanoparticles, ideally tens of nanometers in dimension. Indeed, the use of MOFs with small particle size enables the thickness of the membrane to be reduced and more MOF to be included, both outcomes that improve gas permeability [107]. Synthetic routes to the preparation of MOF nanoparticles are therefore of great use. They include the variation of metal/ ligand ratio and the use of modulating agents, as well as synthesis assisted by microwave heating and ultrasonic irradiation [108]. Not all promising MOF fillers have been prepared as nanoparticles, however. It is also not always true that the smaller particle size or the higher filler content can lead to higher membrane performance, which can be rationalised by poor compatibility between polymer and MOF particles in those cases. Zheng et al. compared a series of ZIF-8/Pebax MMMs with different sizes (40–110 nm) of particles and loading ratio (0–20%), and it was found that the membrane that included 5% ZIF-8 with 90 nm particles outperformed its counterparts [109].
190
5.5.3.4
R. R. R. Prasad et al.
Interface Engineering and Textural Optimisation
Based on the previous section, we can develop principles of the choice of MOF/polymer pair for MM membranes. In principle, knowing the bulk properties of the pure polymer and MOF and the size, morphology and distribution of MOF nanoparticles throughout the membrane, it should be possible to model the permeability and selectivity and their dependence on aspect ratio and particle orientation [110]. However, the nature of the MOF–polymer interface is crucial to these properties and remains difficult to optimise. Fundamental insight has been obtained by molecular modelling of the interface and by in situ IR microspectroscopy. For example, IR micro-imaging of a giant ZIF-8 crystal in 6FDA-DAM [111] shows that the MOF acts as a rapid ‘superhighway’ for CO2 to and from the polymer matrix and also reveals the strong presence of CO2 at the interface, which is attributed to the presence of microvoids. Surface modification of the MOF filler by altering linker functionality is a powerful tool to enhance MOF/polymer interaction via hydrogen bond formation between polymer and MOF surface. For example, Zr-based MOF UiO-66 was tuned by introducing amine functional groups (such as -NH2) and combined with PIM-1 to minimise nonselective voids around the UiO-66 [112]. The modification gave a high-performance membrane with a pronounced increase in CO2/N2 selectivity by 70% compared with pure PIM-1. The study also showed that inclusion of amine functional groups can mitigate polymer aging and stabilise the selectivity in longterm tests. Gas transport can also be facilitated by making a covalent bond between MOF particles and polymer matrix. Yu et al. achieved this by interweaving MOF (UiO-66CN) and polymer (PIM-1) via covalent bond formation by thermal treatment in order to construct pathways for CO2 transport, which contributed to an improvement in CO2 permeation and CO2/N2 selectivity [113]. Similarly, Kertik et al. used thermal treatment to cross-link between polymer (Matrimid) and organic linkers in ZIF-8, as confirmed by transmission electron microscopy (Fig. 5.19). The cross-linking resulted in ZIF-8 amorphisation which promoted membrane stability and plasticisation resistance as well as a very high CO2/CH4 selectivity (162) [114]. An alternative method to modify MOF/polymer interaction involves adding an interfacial layer between two phases. For instance, Jin et al. used the adhesive polydopamine to coat ZIF-8 nanoparticles via controllable self-polymerisation [115]. The as-synthesised fillers were successively embedded into polyimide (shown in Fig. 5.20), and a series of separation tests for different gas mixtures revealed enhancement for a range of gas separations. Another approach is the introduction of cross-linkers to bridge polymer matrix (6FDA-Durene) and MOF particles (ZIF-71) by treating the membrane in vapour agents (such as ethylenediamine, diethylenetriamine and tris-(2-aminoethyl) amine) [116]. The cross-linking appeared at the top surface of the membrane, which is highly selective to H2, especially at high temperature (150 C).
5 Carbon Capture Using Metal–Organic Frameworks
191
Fig. 5.19 (a) TEM electron diffraction image of MMMs with 40% loading of ZIF-8 confirming amorphous state of ZIF-8. HAADF-STEM image of a cross section of MMMs with 40% ZIF-8 in low magnification (b) and high magnification view (c). (d) EDX mapping across the area in (c) to detect the distribution of Zn. (e) Overlay between HAADF image and Zn map. (f) EDX map showing the homogeneous distribution of O which deny the possibility of forming ZnO under heat. (g) Average core energy-loss spectra for a polymer region and ZIF-8 regions marked in (h), and the sharper signals in ZIF reveal a better ordering of ZIF-8 units than polymer matrix. (h) HAADFSTEM overview image for STEM-SI map. (i) A map shows the ratio between C (in red) and N (in blue) contents calculated from spatially resolved core energy-loss spectra, which indicates the polymer region and amorphous ZIF region. (Figure and caption reprinted with permission from [114]. Copyright 2008 Royal Society of Chemistry)
Concerning particle morphology, MOFs with platelet or sheet morphology can facilitate selective gas transport [117]. The use of nanosheets enables efficient dispersion and a more efficient occupation of the membrane cross section compared to crystals, permitting more gas separation events during transport across the membrane. Additionally, the more efficient packing reduces the effects of plasticisation. Rodenas et al. successfully synthesised an ultrathin MOF–polymer composite
192
R. R. R. Prasad et al.
Fig. 5.20 Schematic representation of interfacial design for ZIF-8@PD-PI MMM. (Figure reprinted with permission from [115]. Copyright 2016 John Wiley and Company)
membrane which achieved outstanding CO2/CH4 separation performance [118]. They used 8 wt% of CuBDC nanosheets as a filler incorporated in polyimide which gave substantial enhancement in CO2/CH4 selectivity at 298 K (from 50–60 to 80–90). Kang et al. carried out a similar experiment by incorporating MOF [Cu2(ndc)2(dabco)]n (ndc ¼ naphthalenedicarboxylate, dabco ¼ diazabicyclooctane) in lamellar morphology in PBI, which is shown in Fig. 5.21 [106]. The membrane performance is superior to its counterparts (MMMs with MOF nanoparticles) in H2/ CO2 separation. As amine has a great affinity to CO2, the inclusion of amine function groups can also be used to improve membrane separation. Recently, Xin et al. fixed polyethyleneimine (PEI) in MIL-101(Cr), and the impregnated MOF was embedded into poly(ether ether ketone) (SPEEK) [119]. The immobilisation of PEI in MOF generates ammonium groups and HCO3 groups via intermediate carbamate in the presence of water, as illustrated in Scheme 5.2 of Sect. 5.3. The product HCO3 facilitates CO2 molecules ‘hopping’ from one ammonium site to another, giving considerable improvement in CO2/N2 and CO2/CH4 selectivity.
5.5.4
Summary of Mixed Matrix Membrane Performance
Table 5.6 summarises the permeability and selectivity of some recently reported mixed matrix membranes, and some of these are plotted on Robeson plots in Fig. 5.22. It is clear that MOF–polymer composite membranes can exceed the Robeson upper bound for permeability and selectivity in favourable cases. The extent to which these studies realised in working membranes is the next challenge.
5 Carbon Capture Using Metal–Organic Frameworks
193
Fig. 5.21 Field-emission SEM images of cross sections in (a) 10 wt% MOF nanosheet(ns)@PBI, (b) 20 wt% MOF ns@PBI, (c) 30 wt% MOF ns@PBI, and (d) 50 wt% MOF ns@PBI. (Figure reprinted with permission from [106]. Copyright 2013 Royal Society of Chemistry)
5.5.5
Towards Industrial Application: Hollow Fibre Membranes
Although MOF-based MMMs are maturing to the point that membrane separations can pass beyond the upper bound, there is still a huge gap between potential performance and real application. In the future, increasing emphasis will be placed on applying hollow fibre modules for large-scale applications, because it is not economical to employ flat membranes. With a high packing density and good pressure tolerance, hollow fibre geometry shows great merit compared to its counterparts and is widely applied in industrial gas separations. Hollow fibre membranes are normally prepared via phase inversion spinning, which gives an asymmetric membrane composed of two layers (Fig. 5.23). The outer dense thin layer plays a major part in gas separation, and the inner porous layer functions as a support without transport resistance [124]. This dual-layer membrane is usually produced by the co-extrusion of MOF–polymer dispersion and a bore fluid through a spinneret, and the mixture is subsequently precipitated in a nonsolvent bath via a phase inversion mechanism. Marti et al. [125] were able to prepare hollow fibre
194
R. R. R. Prasad et al.
Table 5.6 Examples of the performance of some MOF-based MMMs
MOF/polymer Polybenzimidazole ZIF-90/PBI Polybenzimidazole ZIF-7/PBI PIM-1 UiO-66-NH2/PIM1 PIM-1 UiO-66-CN/sPIM1 Matrimid® ZIF-8/Matrimid® TBDA2-6FDA-PI ZIF-8@PD-PI 6FDA-Durene ZIF-71/6FDADurene (TAEA) Matrimid 5218 CuBDC/Matrimid 5218 SPEEK PEI@MIL-101 (Cr)/SPEEK PPO-PEG Al-PCP-derived porous carbon/ PPO-PEG Polypropylene ZIF-7NH2@PANI-PP 6FDA-DAM Y-fum-fcu-MOF/ 6FDA-DAM
6FDA-DAM
Operation condition Single/mixed gas permeation at 308 K and 3.5 atm Single/mixed gas permeation at 308 K and 3.5 bar Single/mixed gas permeation (equimolar) at 298 K and 4 bar
CO2 permeability/ Barrer 4.1 (H2) 24.5 (H2)
Gas separation H2/CO2
Separation factor 8 4
0.4 1.8
H2/CO2
7.1 7.2
[104]
3054
CO2/CH4 CO2/N2 CO2/CH4 CO2/N2 CO2/N2 (1:1)
14.5 16.1 28.3 27.5 26.6 53.5
[112]
2869
Reference [103]
Single/mixed gas permeation at 298 K and 1.4 bar
3017.7 16121.3
Equimolar separation at 308 K and 10 bar Single/mixed gas permeation at 308 K and 1 bar
7.4 4.5
CO2/CH4
76.3 162
[114]
285
CO2/CH4 CO2/N2 CO2/CH4 CO2/N2 H2/CO2
35 24 27 21 0.09 50.9
[115]
702
[113]
Single/mixed gas permeation at 308 K and 3.5 atm
756 540
Equimolar separation at 298 K and 6 bar
5.48 3.59
CO2/CH4
52.6 87.1
[118]
Single/mixed gas permeation at 298 K and 1 bar
530
656 1425
25 40 71.8 80 19.2 41
[119]
Mixed gas separation at 303 K and 3 bar
CO2/CH4 CO2/N2 CO2/CH4 CO2/N2 CO2/CH4 (molar ratio 1:9)
Single gas permeation at 308 K and 1 bar
/ 16,478
H2/CO2
Single/mixed gas at 233 K and 1.38 bar (single gas) or 3.5 bar (mixed gas) Single/mixed gas at 308 K and
650 1050
CO2/CH4
85 130
[122]
730
CO2/CH4
25
[123]
2490
[116]
[120]
[121] 10.6
(continued)
5 Carbon Capture Using Metal–Organic Frameworks
195
Table 5.6 (continued)
MOF/polymer NbOFFIVE-1-Ni/ 6FDA-DAM
Operation condition 3.4 bar (single gas) and 6.9 bar (mixed gas)
CO2 permeability/ Barrer 1050
Gas separation
Separation factor 27
Reference
Fig. 5.22 Comparison of MMMs in (a) CO2/CH4, (b) CO2/N2 and H2/CO2 separations (the unfilled shape represents pure polymer performance, and filled shape represents MMM performance)
membranes by in situ growth of ZIF-8 from and aqueous solution of Zn2+ and methylimidazole in the micropores of hollow fibres of Torlon polymer and then as a dense layer on the outside of the fibres (Fig. 5.23b). One of the main challenges of hollow fibre MMM implementation is how to distribute fillers homogeneously in the spinning step. One promising solution to this problem is adapting a microfluidic method to produce the membrane with a controllable thickness, but despite this improvement, the selectivity of the reported membrane is far from industrial requirements [126]. Therefore, a major effort is still required to optimise hollow fibre MMMs.
196
R. R. R. Prasad et al.
Fig. 5.23 (a) and (b, left) Schematic illustrations of hollow fibre mixed matrix membranes. (Figure reprinted with permission from [124]. Copyright 2014 John Wiley and Sons). (b, right) SEM images of cross section of hollow fibre membranes. (Figure reprinted with permission from [125]. Copyright 2017 American Chemical Society). (c) Picture of a module of lab-synthesised hollow fibres of MMMs for CO2 separation prepared as part of the EC Consortium M4CO2
5 Carbon Capture Using Metal–Organic Frameworks
5.5.6
197
Summary
The investigation of mixed matrix membranes using MOFs as fillers for gas separation involving CO2 is now a very active area of research, and the number of literature publications is increasing exponentially. Much progress has been made in preparing functionalised nanoparticle MOFs and devising approaches to their incorporation into polymers with good interfacial contact and dispersion. More needs to be done to improve their stability in the presence of water and contaminants. On this note, Koros et al. have recently reported a MMM of NbOFFive-1-Ni in 6FDADAM, suitably modified, that can remove H2S and CO2 simultaneously from a simulated natural gas [122]. Major challenges remain in the preparation of MMMs in practicable form, such as hollow fibre membranes, and very thin (μm) polymer layers (and nanoparticulate MOFs) are required to raise the permeance and selectivity of membranes so that they become of commercial interest.
5.6
A Word on CO2 Utilisation
Once separated from a gas mixture, pure CO2 can be a C1 chemical feedstock, and routes involving MOF catalysts have been investigated that involve photochemistry and electrochemistry to give simple value-added chemicals including urea, methanol and formic acid, although these are in their infancy [18, 127, 128]. Approaches using CO2 as a reactant in organic chemistry are also being closely studied that are centred on the incorporation of CO2 into rings (such as epoxides) to give useful fine chemicals (such as cyclic organic carbonates) [18], which can be valuable – if unlikely ever to be a major CO2 sink. Finally, chemical reduction by hydrogen over metals supported on MOFs or in nanoparticle form on supports derived by MOF pyrolysis is also under intensive study to give CO or methanol, for example, which are used on a very large scale for fuels and plastics. These catalysts will be in direct competition with robust heterogeneous metal and metal oxide heterogeneous catalysts under wide investigation elsewhere to perform these reductions or ‘dry reforming’ of methane and CO2 to syngas, but reactions over MOFs or MOF-derived catalysts can at the very least aid understanding of these important processes.
5.7
Conclusions
The reduction in CO2 emissions from stationary point sources such as coal- and gas-fired power plants and industrial sources, together with the upgrading of natural gas and biogas, can become a major technology to reduce global warming, if the processes meet the economic constraints and performance requirements.
198
R. R. R. Prasad et al.
Additionally, carbon capture from air can develop from a prerequisite purification technology for air separation to a process to offset emissions and even to reduce atmospheric CO2 levels. Currently, amine absorption is successfully used in all these applications but has a high energy cost and involves corrosive solutions that themselves emit pollutants. There is therefore considerable incentive to extend clean, low-energy processes based on solid nanoporous adsorbents such as MOFs to expand from current processes, such as H2 purification, to a much wider range of carbon capture applications, but to do so they must surmount significant challenges. In adsorption processes, such as PSA and TSA, the chemical engineering is well established, and the risk lies with the metal–organic framework materials themselves and in particular their stability and performance in moist air, apart from their relatively high cost and mechanical fragility. Nevertheless, recent development of physisorption on ultramicroporous MOFs such as NbOFFive-Ni-1 suggests this approach can be important for flue gas and even air capture, if the fluorides can show long-term hydrolytic stability. Approaches have also been developed that mimic the uptake of CO2 by enzymes involved in respiration (α-carbonic anhydrase) and photosynthesis (Rubisco). Of these, the diamine-appended Mg-MOFs now compete with amine-loaded silicas as the most widely studied solid adsorbents, where moisture enhances rather than inhibits their action. In a sense, these return close to full circle to the well-established CO2-to-carbamate/bicarbonate chemistry of amine and alcohol solutions. Membrane separation using mixed matrix membranes comprising dispersions of MOFs in porous polymers has the potential to provide a lower energy, pressuredriven technology without moving parts. Unlike adsorption processes, mixed matrix membrane performance is less easy to model or predict, largely because of the properties of the interface, but like adsorption, the stability of the MOF, particularly in the presence of impurities, is an important issue. Additionally, the fabrication of high-quality, multicomponent membranes in hollow fibre form is a major practical challenge. Nevertheless, enough progress has been made, particularly in the light of measurements made on flat membranes under laboratory conditions, to encourage their development into real-world applications. Developing in parallel, these technologies can yield important advantages in carbon capture in the medium term over current methods and so encourage the continued development of MOFs with increased stability and with tailored adsorption sites or diffusion pathways. Acknowledgements PAW gratefully acknowledges funding for research programmes into carbon capture by adsorption and membrane separation from the EC (DeSANNS, STREP Program SES6CT-2005-020133; M4CO2, 7th Framework Program FP7/2007-2013 grant agreement 608490) and the UK EPSRC (EP/G062129/1; EP/J02077X/1 AMPgas; EP/N024613/1, FLEXICCS (Flexible Industrial Carbon Capture)), as well as many interesting discussions with the scientific coordinators of those projects, Philip Llewellyn (Marseilles), Stefano Brandani (Edinburgh University) and Freek Kapteijn (TU Delft). QJ acknowledges funding from the China Scholarship Council (CSC), studentship 201908140117
5 Carbon Capture Using Metal–Organic Frameworks
199
References 1. National Oceanic and Atmospheric Administration. Trends in atmospheric carbon dioxide. http://www.esrl.noaa.gov/gmd/ccgg/trends. Accessed 1 May 2020 2. Tollefson J (2018) Clock ticking on climate action. Nature 562:172–173 3. Gibbins J, Chalmers H (2008) Carbon capture and storage. Energy Policy 36:4317–4322 4. Bui M, Adjiman CS, Bardow A et al (2008) Carbon capture and storage (CCS): the way forward. Energy Environ. Sci. 11:1062–1076 5. Haszeldine RS (2009) Carbon capture and storage: how green can black be? Science 325:1647–1652 6. World Energy Outlook 2017; International Energy Agency (2017) doi: https://doi.org/10. 1787/weo-2017-en 7. Annual Energy Outlook 2018 with projections to 2050; US Energy Information Administration (2018) 8. Chaemchuen S, Alam Kabir N, Zhou K et al (2013) Metal-organic frameworks for upgrading biogas via CO2 adsorption to biogas green energy. Chem. Soc. Rev. 42:9304–9332 9. Bacsik Z, Cheung O, Vasiliev P et al (2016) Selective separation of CO2 and CH4 for biogas upgrading on NaKA and SAPO-56. Appl. Energy 162:613–621 10. Kumar A, Madden DG, Lusi M et al (2015) Direct air capture of CO2 by physisorbent materials. Angew. Chem. Int. Ed. 54:14372–14377 11. Sanz-Perez ES, Murdock CR, Didas SA et al (2016) Direct capture of CO2 from ambient air. Chem. Rev. 116:11840–11876 12. Pardakhti M, Jafari T, Tobin Z et al (2019) Trends in solid adsorbent materials development for CO2 capture. ACS Appl. Mater. Interfaces 11:34533–34559 13. Zhou D-D, Zhang X-W, Mo Z-W et al (2019) Adsorptive separation of carbon dioxide: from conventional porous materials to metal-organic frameworks. EnergyChem 1:100016 14. Gibson JAA, Mangano E, Shiko E et al (2016) Adsorption materials and processes for carbon capture from gas-fired power plants: AMPGas. Ind. Eng. Chem. Res. 55:3840–3851 15. Carbon capture and storage (2019) In: Bui M, MacDowell N (eds), RSC Publishing, Cambridge 16. Sumida K, Rogow DL, Mason JA et al (2012) Carbon dioxide capture in metal organic frameworks. Chem. Rev. 112:724–781 17. Tricket CA, Helal A, Al-Maythalony BA et al (2017) The chemistry of metal-organic frameworks for CO2 capture, regeneration and conversion. Nat Rev. Mater. 2:17045 18. Ding M, Robinson WF, Jiang H-L et al (2019) Carbon capture and conversion using metalorganic frameworks and MOF-based materials. Chem. Soc. Rev. 48:2783–2828 19. Zevenhoven R, Kilpinen P (2002) Control of pollutants in flue gases. Espro, Turku, 2nd Ed. Ch 2. p. 2–4 20. Tan Z (2014) In: ‘Air pollution and greenhouse gases: from basic concepts to engineering applications for air emission control’ Springer Singapore 1st Ed Ch. 12 p 357 21. Tissot BP, Welte DH (1989) Petroleum formation and occurrence. Springer-Verlag, Berlin Heidelberg Gmbh 22. Lide DR (2009) CRC handbook of chemistry and physics 89th Ed. CRC Press 23. Graham C, Imrie DA, Raab RE (1998) Measurement of the electric quadrupole moments of CO2, CO, N2, Cl2 and BF3. Mol. Phys. 93:49–56 24. Vrabec J, Stoll J, Hasse H (2001) A set of molecular models for symmetric quadrupolar fluids. J. Phys. Chem. B 105:12126–12133 25. Stoll J, Vrabec J, Hasse H (2003) A set of molecular models for carbon monoxide and halogenated hydrocarbons. J. Chem. Phys. 119:11396–11406 26. Brunauer S, Deming LS, Deming WE, Teller E (1940) Types of adsorption isotherms. J. Am. Chem. Soc. 62:142 27. Chang R (2005) Physical chemistry for the biosciences. University Science Books, pp 385–392
200
R. R. R. Prasad et al.
28. Miller SR, Wright PA, Devic T et al (2009) Single crystal X-ray diffraction studies of carbon dioxide and fuel-related gases adsorbed on the small pore scandium terephthalate metal organic framework, Sc2(O2CC6H4CO2)3. Langmuir 25:3618–3626 29. Wu H, Simmons JM, Srinivas G et al (2010) Adsorption sites and binding nature of CO2 in prototypical metal-organic frameworks: a combined neutron diffraction and first-principles study. J. Phys. Chem. Lett. 1:1946–1951 30. Valenzano L, Civalleri B, Chavan S et al (2010) Computational and experimental studies on the adsorption of CO, N2 and CO2 on Mg-MOF-74. J. Phys. Chem. C 114:11185–11191 31. Yang S, Sun J, Ramirez-Cuesta AJ et al (2012) Selectivity and direct visualisation of carbon dioxide and sulfur dioxide in a decorated porous host. Nat. Chem. 4:887–894 32. Llewellyn PL, Maurin G, Devic T et al (2008) Prediction of the conditions for breathing of metal organic frameworks using a combination of X-ray powder diffraction, microcalorimetry and molecular simulation. J. Am. Chem. Soc. 130:12808–12814 33. Liu B, Smit B (2010) Molecular simulation studies of separation of CO2/N2, CO2/CH4 and CH4/N2 by ZIFs. J. Phys. Chem. C 114:8515–8522 34. Yang Q, Jobic H, Salles F et al (2011) Probing the dynamics of CO2 and CH4 within the porous zirconium terephthalate UiO-66(Zr): a synergic combination of neutron scattering measurements and molecular simulation. Chem. Eur. J. 17:8882–8889 35. Dzubak AL, Lin L-C, Kim J et al (2012) Ab initio carbon capture in open-site metal-organic frameworks. Nat. Chem. 4:810–816 36. Chen L, Mowat JPS, Fairen-Jimenez D et al (2013) Elucidating the breathing of the metalorganic framework MIL-53(Sc) with ab initio molecular dynamics simulations and in situ X-ray powder diffraction experiments. J. Am. Chem. Soc. 135:15763–15773 37. Preston CK (2018) The carbon capture project at air products’ Port Arthur hydrogen production facility. 14th Greenhouse Gas Control Technologies Conference, Melbourne, Oct. 2018 https://papers.ssrn.com/sol3/papers (Accessed 2 May 2020) 38. Llewellyn PL, Bourrelly S, Serre C et al (2008) High uptakes of CO2 and CH4 in mesoporous metal-organic frameworks MIL-100 and MIL-101. Langmuir 24:7245–7250 39. Millward AR, Yaghi OM (2005) Metalorganic frameworks with exceptionally high capacity for storage of carbon dioxide at room temperature. J. Am. Chem. Soc. 127:17998–17999 40. Chui SS-Y, Lo SM-F, Charmant JPH et al (1999) A chemically functionalizable nanoporous material [Cu3(TMA)2(H2O)3]n. Science 283:1148–1150 41. Rosi NL, Kim J, Eddaoudi M et al (2005) Rod packings and metal-organic frameworks constructed from rod-shaped secondary building units. J. Am. Chem. Soc. 127:1505–1518 42. Dietzel PDC, Blom R, Fjellvag H (2008) Base-induced formation of two magnesium metalorganic framework compounds with a bifunctional tetratopic ligand. Eur. J. Inorg. Chem.:3624–3632 43. Yu D, Yazaydin AO, Lane JR et al (2013) A combined experimental and quantum chemical study of CO2 adsorption in the metal-organic framework CPO-27 with different metals. Chem. Sci. 4:3544–3556 44. Xiang S, He Y, Zhang Z et al (2012) Microporous metal-organic framework with potential for carbon dioxide capture at ambient conditions. Nat. Commun. 3:954 45. Grande CA, Blom R, Andreassen KA et al (2017) Experimental results of pressure swing adsorption (PSA) for pre-combustion CO2 capture with metal-organic frameworks. Energy Procedia 114:2265–2270 46. Chanut N, Bourelly S, Kuchta B et al (2017) Screening the effect of water vapour on gas adsorption performance: application to CO2 capture from flue gas in metal-organic frameworks. ChemSusChem 10:1543–1553 47. Yazaydin AÖ, Benin AI, Faheem SA et al (2009) Enhanced CO2 adsorption in metal-organic frameworks via occupation of open metal sites by coordinated molecules. Chem. Mater. 21:1425–1430 48. McEwen J, Hayman J-D, Yazaydin AÖ (2013) A comparative study of CO2, CH4 and N2 adsorption in ZIF-8, zeolite 13X and BPL activated carbon. Chem. Phys. 412:72–76
5 Carbon Capture Using Metal–Organic Frameworks
201
49. Mowat JPS, Miller SR, Griffin JM et al (2011) Structural chemistry, monoclinic-to-orthorhombic phase transition, and CO2 adsorption behaviour of the small pore scandium terephthalate, Sc2(O2CC6H4CO2)3, and its nitro- and amino-functionalized derivatives. Inorg. Chem. 50:10844–10858 50. Greenaway A, Gonzalez-Santiago B, Donaldson PM et al (2014) In situ Synchrotron IR microspectroscopy of CO2 adsorption on single crystals of the functionalized MOF Sc2(BDC-NH2)3. Angew. Chem. Int. Ed. 53:13483–13487 51. Pillai RS, Benoit V, Orsi A et al (2015) Highly selective CO2 capture by small pore scandiumbased metal–organic frameworks. J. Phys. Chem. C 119:23592–23598 52. Si X, Jiao C, Li F et al (2011) High and selective CO2 uptake, H2 storage and methanol sensing on the amine decorated 12-connected MOF CAU-1. Energy Environ. Sci. 4:4522–4527 53. Couck S, Denayer JFM, Baron GV et al (2009) An amine-functionalised MIL-53 metalorganic framework with large separation power for CO2 and CH4. J. Am. Chem. Soc. 131:6326–6327 54. An J, Geib SJ, Rosi NL (2010) High and selective CO2 uptake in a cobalt adeninate metalorganic framework exhibiting pyrimidine and amino-decorated pores. J. Am. Chem. Soc. 132:38–39 55. Xiong S, He Y, Krishna R et al (2013) Metal-organic framework with functional amide groups for highly selective gas separation. Cryst. Growth Des. 13:2670–2674 56. Serre C, Groves JA, Lightfoot P et al (2006) Synthesis, structure and properties of microporous N,N’-piperazine bismethylenephosphonates of aluminium and titanium. Chem. Mater. 18:1451–1457 57. Benoit V, Pillai RS, Orsi A et al (2016) MIL-91(Ti), a small pore metal-organic framework which fulfils several criteria: an upscaled green synthesis, excellent water stability, high CO2 selectivity and fast CO2 transport. J. Mater. Chem. A 4:1383–1389 58. Llewellyn PL, Garcia-Rates M, Gaberova L et al (2015) Structural origin of unusual CO2 adsorption behaviour of a small pore aluminium bisphosphonate MOF. J. Phys. Chem. C 119:4208–4216 59. Nugent P, Belmabkhout Y, Burd SD et al (2013) Porous materials with optimal adsorption thermodynamics and kinetics for CO2 separation. Nature 495:80–84 60. Bhatt PM, Belmabkhout Y, Cadiau A et al (2016) A fine-tuned fluorinated MOF addresses the needs for trace CO2 removal and air capture using physisorption. J. Am. Chem. Soc. 138:9301–9307 61. Liang W, Bhatt PM, Shkurenko A et al (2019) A tailor-made interpenetrated MOF with exceptional carbon capture performance from flue gas. Chem 5:950–963 62. Liang JY, Lipscomb WN (1990) Binding of substrate CO2 to the active site of human carbonic anhydrase II: a molecular dynamics study. Proc. Natl. Acad. Sci. U. S. A. 87:3675–3679 63. Bien CE, Chen KK, Chien S-C et al (2018) Bioinspired metal-organic framework for trace CO2 capture. J. Am. Chem. Soc. 140:12662–12666 64. Bien CE, Liu Q, Wade CR (2020) Assessing the role of metal identity on CO2 adsorption in MOFs containing M-OH groups. Chem. Mater. 32:489–497 65. Wright AM, Wu Z, Zhang G et al (2018) A structural mimic of carbonic anhydrase in a metalorganic framework. Chem 4:2894–2901 66. Hwang YK, Hong D-Y, Chang J-S et al (2008) Amine grafting on coordinatively unsaturated metal centers of MOFs: consequences for catalysis and metal encapsulation. Angew. Chem. Int. Ed. 47:4144–4148 67. Demessence A, D’Alessandro DM, Foo ML et al (2009) Strong CO2 binding in a water stable, triazolate-bridged meta-organic framework stabilised with ethylene diamine. J. Am. Chem. Soc. 131:8784–8785 68. McDonald TM, Lee WR, Mason JA et al (2012) Capture of carbon dioxide from air and flue gas in the alkylamine-appended MOF mmen-Mg2(dobpdc). J. Am. Chem. Soc. 134:7056–7065
202
R. R. R. Prasad et al.
69. McDonald TM, Mason JA, Kong X et al (2015) Cooperative insertion of CO2 in diamineappended metal-organic frameworks. Nature 519:303–308 70. Siegelman RL, McDonald TM, Gonzalez MI et al (2017) Controlling cooperative CO2 adsorption in diamine-appended Mg2(dobpdc) metal-organic frameworks. J. Am. Chem. Soc. 139:10526–10538 71. Forse AC, Milner PJ, Lee J-H et al (2018) Elucidating CO2 chemisorption in diamineappended metal-organic frameworks. J. Am. Chem. Soc. 140:18016–18031 72. Taylor TC, Andersson I (1996) Structural transitions during activation and ligand binding in hexadecameric rubisco inferred from the crystal structure of activated unliganded spinach enzyme. Nat Struct Biol 3:95–101 73. Milner PJ, Siegelman RL, Forse AC et al (2017) A diaminopropane-appended metal-organic framework enabling efficient CO2 capture from coal flue gas via a mixed adsorption mechanism. J. Am. Chem. Soc. 139:13541–13553 74. Milner PJ, Martell JD, Siegelmann RL et al (2018) Overcoming double-step CO2 adsorption and minimizing water co-adsorption in bulky diamine-appended variants of Mg2(dobpdc). Chem. Sci. 9:160–174 75. Siegelman RL, Milner PJ, Forse AC et al (2019) Water enables efficient CO2 capture from natural gas flue emissions in an oxidation-resistant diamine-appended metal-organic framework. J. Am. Chem. Soc. 141:13171–13186 76. Taek Kwan H, Sakwa-Novak MA, Pang SH et al (2019) Aminopolymer-impregnated hierarchical silica structures: unexpected equivalent CO2 uptake under simulated air capture and flue gas capture conditions. Chem. Mater. 31:5229–5237 77. Chen C-X, Zheng S-P, Wei Z-W et al (2017) A robust metal-organic framework containing open metal sites and polar groups for methane purification and CO2/fluorocarbon capture. Chem. Eur. J. 23:4060–4064 78. Seoane B, Coronas J, Gascon I et al (2015) Metal–organic framework based mixed matrix membranes: a solution for highly efficient CO2 capture? Chem. Soc. Rev. 44:2421–2454 79. Park HB, Kamcev J, Robeson LM et al (2017) Maximizing the right stuff: the trade-off between membrane permeability and selectivity. Science 356:eaab0530 80. Dechnik J, Sumby CJ, Janiak C (2017) Enhancing mixed-matrix membrane performance with metal–organic framework additives. Cryst. Growth Des. 17:4467–4488 81. Baker RW, Low BT (2014) Gas separation membrane materials: a perspective. Macromolecules 47:6999–7013 82. Freeman BD (1999) Basis of permeability/selectivity tradeoff relations in polymeric gas separation membranes. Macromolecules 32:375–380 83. Wesselingh JA, Krishna R (2000) Mass transfer in multicomponent mixtures. Delft University Press Delft 84. Baker RW (2012) Membrane technology and applications. John Wiley & Sons 85. Freeman BD, Pinnau I (1999) Polymeric materials for gas separations. In: ACS Symposium Series: Vol. 733. Polymer membranes for gas and vapor separation (p 1) 86. Robeson LM (1991) Correlation of separation factor versus permeability for polymeric membranes. J. Membr. Sci. 62:165–185 87. Robeson LM (2008) The upper bound revisited. J. Membr. Sci. 320:390–400 88. Lewis TB, Nielsen LE (1970) Dynamic mechanical properties of particulate-filled composites. J. Appl. Polym. Sci. 14:1449–1471 89. Felske JD (2004) Effective thermal conductivity of composite spheres in a continuous medium with contact resistance. Int. J. Heat Mass Transf. 47:3453–3461 90. Moore TT, Mahajan R, Vu DQ, Koros WJ (2004) Hybrid membrane materials comprising organic polymers with rigid dispersed phases. AICHE J. 50(2):311–321 91. Shen J, Liu G, Huang K et al (2016) UiO-66-polyether block amide mixed matrix membranes for CO2 separation. J. Membr. Sci. 513:155–165 92. Ahn J, Chung WJ, Pinnau I, Guiver MD (2008) Polysulfone/silica nanoparticle mixed-matrix membranes for gas separation. J. Membr. Sci. 314:123–133
5 Carbon Capture Using Metal–Organic Frameworks
203
93. Fang M, Wu C, Yang Z et al (2015) ZIF-8/PDMS mixed matrix membranes for propane/ nitrogen mixture separation: experimental result and permeation model validation. J. Membr. Sci. 474:103–113 94. Zornoza B, Tellez C, Coronas J et al (2013) Metal organic framework based mixed matrix membranes: an increasingly important field of research with a large application potential. Microporous Mesoporous Mater. 166:67–78 95. Czichos H, Saito T, Smith L (2006) Springer handbook of materials measurement methods, vol 978. Springer 96. Rezakazemi M, Amooghin A Ebadi et al T (2014) State-of-the-art membrane based CO2 separation using mixed matrix membranes (MMMs): an overview on current status and future directions. Prog. Polym. Sci. 39: 817–861 97. Etxeberria-Benavides M, David O, Johnson T et al (2018) High performance mixed matrix membranes (MMMs) composed of ZIF-94 filler and 6FDA-DAM polymer. J. Membr. Sci. 550:198–207 98. Ma L, Svec F, Lv Y, Tan T (2019) Engineering of the filler/polymer Interface in metal–organic framework-based mixed-matrix membranes to enhance gas separation. Chem – An Asian J 14:3502–3514 99. Su NC, Sun DT, Beavers CM et al (2016) Enhanced permeation arising from dual transport pathways in hybrid polymer–MOF membranes. Energy Environ. Sci. 9:922–931 100. Bae YS, Snurr RQ (2011) Development and evaluation of porous materials for carbon dioxide separation and capture. Angew. Chem. Int. Ed. 50:11586–11596 101. Fairen-Jimenez D, Moggach SA, Wharmby MT et al (2011) Opening the gate: framework flexibility in ZIF-8 explored by experiments and simulations. J. Am. Chem. Soc. 133:8900–8902 102. Li H, Tuo L, Yang K et al (2016) Simultaneous enhancement of mechanical properties and CO2 selectivity of ZIF-8 mixed matrix membranes: interfacial toughening effect of ionic liquid. J. Membr. Sci. 511:130–142 103. Yang T, Chung TS (2013) Room-temperature synthesis of ZIF-90 nanocrystals and the derived nano-composite membranes for hydrogen separation. J. Mater. Chem. A 1:6081–6090 104. Yang T, Xiao Y, Chung TS (2011) Poly-/metal-benzimidazole nano-composite membranes for hydrogen purification. Energy Environ. Sci. 4:4171–4180 105. Ghalei B, Wakimoto K, Wu CY et al (2019) Rational tuning of zirconium metal–organic framework membranes for hydrogen purification. Angew. Chem. Int. Ed. 58:19034–19040 106. Kang Z, Peng Y, Hu Z et al (2015) Mixed matrix membranes composed of two-dimensional metal–organic framework nanosheets for pre-combustion CO2 capture: a relationship study of filler morphology versus membrane performance. J. Mater. Chem. A 3:20801–20810 107. Chung TS, Jiang LY, Li Y, Kulprathipanja S (2007) Mixed matrix membranes (MMMs) comprising organic polymers with dispersed inorganic fillers for gas separation. Prog. Polym. Sci. 32:483–507 108. Stock N, Biswas S (2012) Synthesis of metal-organic frameworks (MOFs): routes to various MOF topologies, morphologies, and composites. Chem. Rev. 112:933–969 109. Zheng W, Ding R, Yang K et al (2019) ZIF-8 nanoparticles with tunable size for enhanced CO2 capture of Pebax based MMMs. Sep. Purif. Technol. 214:111–119 110. Scheider D, Kapteijn F, Valiullin R (2019) Transport properties of mixed-matrix membranes: a kinetic Monte Carlo study. Phys Rev Appl 12:044034 111. Hwang S, Semino R, Seoane B et al (2018) Revealing the transient concentration of CO2 in a mixed-matrix membrane by IR microimaging and molecular modelling. Angew. Chem. Int. Ed. 57:5156–5160 112. Ghalei B, Sakurai K, Kinoshita Y et al (2017) Enhanced selectivity in mixed matrix membranes for CO2 capture through efficient dispersion of amine-functionalized MOF nanoparticles. Nat. Energy 2:17086 113. Yu G, Zou X, Sun L et al (2019) Constructing connected paths between UiO-66 and PIM-1 to improve membrane CO2 separation with crystal-like gas selectivity. Adv. Mater. 31:1806853
204
R. R. R. Prasad et al.
114. Kertik A, Wee LH, Pfannmöller M et al (2017) Highly selective gas separation membrane using in situ amorphised metal–organic frameworks. Energy Environ. Sci. 10:2342–2351 115. Wang Z, Wang D, Zhang S et al (2016) Interfacial design of mixed matrix membranes for improved gas separation performance. Adv. Mater. 28:3399–3405 116. Japip S, Liao KS, Xiao Y, Chung TS (2016) Enhancement of molecular-sieving properties by constructing surface nano-metric layer via vapor cross-linking. J. Membr. Sci. 497:248–258 117. Deng J, Dai Z, Hou J, Deng L (2020) Morphologically Tunable MOF Nanosheets in Mixed Matrix Membranes for CO2 Separation. Chem Mater in press 118. Rodenas T, Luz I, Prieto G et al (2015) Metal–organic framework nanosheets in polymer composite materials for gas separation. Nat. Mater. 14:48–55 119. Xin Q, Ouyang J, Liu T et al (2015) Enhanced interfacial interaction and CO2 separation performance of mixed matrix membrane by incorporating polyethylenimine-decorated metal– organic frameworks. ACS Appl. Mater. Interfaces 7:1065–1077 120. Zhang W, Liu D, Guo X et al (2018) Fabrication of mixed-matrix membranes with MOF-derived porous carbon for CO2 separation. AICHE J. 64:3400–3409 121. Ma L, Svec F, Tan T, Lv Y (2019) In-situ growth of highly permeable zeolite imidazolate framework membranes on porous polymer substrate using metal chelated polyaniline as interface layer. J. Membr. Sci. 576:1–8 122. Liu Y, Liu G, Zhang C et al (2018) Enhanced CO2/CH4 separation performance of a mixed matrix membrane based on tailored MOF-polymer formulations. Adv. Sci. 5:1800982 123. Liu G, Cadiau A, Liu Y et al (2018) Enabling fluorinated MOF-based membranes for simultaneous removal of H2S and CO2 from natural gas. Angew. Chem. Int. Ed. 57:14811–14816 124. Zhang C, Zhang K, Xu L et al (2014) Highly scalable ZIF-based mixed-matrix hollow fiber membranes for advanced hydrocarbon separations. AICHE J. 60:2625–2635 125. Marti AM, Wickramanayake W, Dahe G et al (2017) Continuous flow processing of ZIF-8 membranes on polymeric porous hollow fiber supports for CO2 capture. ACS Appl. Mater. Interfaces 9:5678–5682 126. Cacho-Bailo F, Caro G, Etxeberría-Benavides M et al (2015) High selectivity ZIF-93 hollow fiber membranes for gas separation. Chem. Commun. 51:11283–11285 127. Alper E, Yuksel OO (2017) CO2 utilization: developments in conversion processes. Petroleum 3:109–126 128. Lei Z, Xue Y, Chen W et al (2018) MOFs-based heterogeneous catalysts: new opportunities for energy-related CO2 conversion. Adv. Energy Mater. 8:1801587
Chapter 6
Computational Screening of MOFs for CO2 Capture Cigdem Altintas, Ilknur Erucar, and Seda Keskin
6.1
Introduction
The demand for developing efficient and sustainable systems for carbon capture and storage (CCS) has been rapidly increasing because of significant carbon dioxide (CO2) emission (414.7 ppm in May 2019) [1] that contributes more than 60% of the global warming [2]. The increase in the average atmospheric temperature has serious effects on humans and our ecosystems such as rising of ocean levels. CCS technologies aim to reduce and control CO2 emission caused by fossil fuels. CCS mainly consists of three steps: CO2 separation from various emission sources and transportation of CO2 to a storage module and submarine storage which includes underground trapping of CO2 up to thousands of years [3]. Based on the combustion of fossil fuels, three processes including pre-combustion, oxy-fuel combustion, and post-combustion have been considered for CO2 capture. In pre-combustion, syngas composed of carbon monoxide (CO) and hydrogen (H2) is produced due to fuel combustion. CO then reacts with steam to yield CO2 and H2 at high pressures (5–40 bar) [3]. CO2 is separated from CO2/H2 mixture which has other minor components such as nitrogen (N2), water (H2O), CO, and hydrogen sulfide (H2S). In oxy-fuel combustion, oxygen (O2) is first separated from air and then used for the combustion of coal or gas to produce CO2 and H2O [4]. In post-combustion, CO2 is separated from the flue gas mixture mainly composed of N2 and CO2 with other minor gases such as H2O, CO, NOx, and SOx after combustion of fuel in air. The release of these toxic and hazardous pollutants into the atmosphere is a major
C. Altintas · S. Keskin (*) Department of Chemical and Biological Engineering, Koc University, Istanbul, Turkey e-mail: [email protected] I. Erucar Department of Natural and Mathematical Sciences, Faculty of Engineering, Ozyegin University, Istanbul, Turkey © Springer Nature Switzerland AG 2021 P. Horcajada Cortés, S. Rojas Macías (eds.), Metal-Organic Frameworks in Biomedical and Environmental Field, https://doi.org/10.1007/978-3-030-63380-6_6
205
206
C. Altintas et al.
concern for air pollution. Although CO2 can be nontoxic, the accumulation of CO2 in the atmosphere causes the greenhouse effect. Therefore, capturing the toxic and harmful gases which contain a large amount of CO2 is important for environmental reasons. Chemical absorption using liquid amines such as monoethanolamine (MEA), methyldiethanolamine (MDEA), and diethanolamine (DEA) is the most widely used method in the industry to capture CO2 [5]. However, this method has several drawbacks such as low absorption capacity, poor stability, corrosion of the equipment, loss of the solvent, and high regeneration cost of the large-scale operation [6]. Adsorption-based methods utilizing solid porous materials have been recently considered as alternative CO2 capture methods due to the reusability of solid adsorbents, lower energy demand for regeneration of the adsorbents, and comparably lower capital investment [7]. In an adsorption-based gas separation method, a porous solid adsorbent material is commonly used to adsorb CO2 by physisorption. Several adsorbent performance evaluation metrics such as adsorption capacity, selectivity, regenerability, working capacity, and enthalpies of adsorption as we will discuss in detail below are generally used to assess the performance of adsorbent materials. The performance of adsorbents is highly dependent on the identity and characteristic features of the material. Various types of adsorbents such as carbon-based materials including activated carbons, carbon fibers, graphene, zeolites, metal oxides, and porous crystalline solids including metal-organic frameworks (MOFs), zeolitic imidazolate frameworks (ZIFs), and covalent organic frameworks (COFs) have been examined for CO2 capture [1]. Figure 6.1 shows the number of publications on CO2 adsorption using various different adsorbent materials such as activated carbons, carbon fibers, graphene, zeolite, metal oxides, and MOFs. As can be seen from this figure, there has been a growing interest on studies focusing on MOFs. MOFs are a relatively new group of porous material family, and they have been synthesized based on the self-assembly of metal ions and organic ligands. MOFs generally have very large surface areas (up to 10000 m2/g), high pore volumes (up to 2.5 cm3/g), and permanent size-selective pores [8–12]. Unlike other adsorbents, chemical and physical properties of MOFs can be tuned during and after the synthesis which makes them superior compared to traditional adsorbents [13, 14]. MOFs have been shown to exceed the CO2 adsorption capacity of several traditional adsorbent materials. For example, Mg-MOF-74 was reported to have the highest CO2 uptake in 2010, 8.48 mol/kg at 1 bar, 298 K [15], which was higher than that of traditional zeolite 13X (4.7 mol/kg) and carbon Norbit RB2 (2.5 mol/kg) reported under the same conditions. Hu et al. [6] recently reported that Mg-MOF-74 still holds the highest CO2 uptake at 1 bar and 298 K, outperforming other promising MOFs such as UiO-66(Zr)-NH2 (~3 mol/kg) [16], SIFSIX-2-Cu-I (5.41 mol/kg) [17], PPN-SO3Li (3.7 mol/kg) [18], and widely studied zeolite, zeolite 13X (~5 mol/ kg) [19] at the same conditions. Several studies reviewed the CO2 adsorption properties of MOFs: Keskin et al. [15] reviewed experimental studies on CO2 adsorption in MOFs at various operating conditions; Sumida et al. [20] comprehensively reviewed experimental and computational studies on CO2 adsorption in
6 Computational Screening of MOFs for CO2 Capture
Number of publications (2008-2019)
700 600 500
207
Activated carbons Carbon fibers Graphene Zeolite Metal oxides Metal organic frameworks
400 300 200 100
20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16 20 17 20 18 20 19
0
Years Fig. 6.1 Number of publications featuring the term “CO2 capture” and “activated carbons,” “carbon fibers,” “graphene,” “zeolite,” “metal oxides,” and “metal-organic frameworks” in their topics. (Accessed: 31-12-2019 at Web of Science®)
MOFs at both high (10–300 bar) and low pressures (0–1 bar) considering a wide range of temperatures (273–319 K). Liu et al. [21] summarized the studies on selective CO2 adsorption in some MOFs; Sabouni et al. [22] reviewed the CCS technologies and listed both high-pressure and low-pressure CO2 adsorption properties of MOFs. Li et al. [23] focused on three properties including high porosity, optimum framework structure, and immobilized functional sites which make MOFs promising for CO2, H2, and CH4 storage. Belmabkhout et al. [24] compared the potential of MOF adsorbents with that of benchmark materials such as zeolites for low concentration CO2 capture. Yu et al. [25] reviewed both experimental and computational studies on CO2 adsorption and separation using MOFs. Li et al. [26] discussed the desired chemical properties of MOFs for gas storage and separation applications. Lin et al. [27] addressed several approaches such as introducing open metal sites and amine functionalization for improving CO2 adsorption capacities of MOFs at low pressures. Hu et al. [6] recently reviewed studies on CO2 separation from binary (CO2/H2 and CO2/N2) and ternary mixtures (CO2/N2/H2O) using MOFs and discussed the optimal process conditions to use MOFs in industrial CO2 capture operations. Bae and Snurr [28] reviewed the strategies for improving the performance of MOFs for CO2 capture and separation and defined five adsorbent performance evaluation criteria including CO2 uptake capacity, working capacity, regenerability, selectivity, and sorbent selection parameter to quickly assess the
208
C. Altintas et al.
performance of MOFs for natural gas purification, landfill gas separation, and flue gas separation. Due to the very large number of combinations of metal clusters and organic linkers, thousands of different MOF structures have been synthesized, and 99075 MOFs were deposited into the Cambridge Structural Database, CSD [29], according to the November 2019 version. Performing experimental measurements to assess the CO2 capture potential of every single MOF reported in the CSD is almost impossible due to the very large number of materials. Therefore, computational methods play a key role in mimicking the experiments and assessing CO2 adsorption and separation properties of large numbers of MOFs. Grand canonical Monte Carlo (GCMC) simulations as we will discuss in detail below have been widely used in the literature to compute CO2 adsorption in MOFs at a defined pressure and temperature. Due to the development of new algorithms and the increase in the computational power, large-scale molecular simulations have been recently performed to compute gas adsorption data of several thousands of MOFs in a time-efficient manner, and promising MOF candidates have been accurately identified based on the simulation results. Throughout this chapter, we aim to discuss the studies on large-scale computational screening of MOFs for CO2 capture and separation. We first introduced the computational tools used to compute CO2 adsorption properties of MOFs together with the details of molecular simulation techniques and calculation of adsorbent performance evaluation metrics used to assess the potential of MOFs. We then reviewed the molecular simulation studies on CO2 separation from CH4, N2, H2, and other gases by specifically focusing on studies combining molecular simulations with experimental testing and/or machine learning algorithms. Opportunities and challenges in the field of high-throughput computational screening of MOFs for CO2 separation were finally addressed together with future directions.
6.2 6.2.1
Molecular Simulations of MOFs for CO2 Capture Identifying Structural Properties of MOFs
The large variety of physical and chemical properties of MOFs makes them promising materials for various applications. A general approach is to first calculate structural properties of MOFs such as the limiting and the largest pore diameters (PLD and LCD, respectively), pore volumes, and accessible surface areas using computational tools, such as Zeo++ [30], PoreBlazer [31], and RASPA [32], prior to molecular simulations of CO2 adsorption in MOFs. In this way, materials having guest-accessible pores with appropriate pore sizes can be identified. Generally, a probe molecule with a predetermined diameter is rolled inside the MOF for sampling the available space without overlapping with the MOF atoms. Conducting a GCMC simulation with helium atom is an alternative way to compute the pore volume of MOFs; however its dependency on the temperature and van der Waals parameters of
6 Computational Screening of MOFs for CO2 Capture
209
the MOF atoms should be noted [33]. Efficient calculation of structural properties of MOFs is important also for the establishment of structure-performance relations [30, 31, 33–36] which are very useful to guide the design and synthesis of new MOFs having the desired properties that can achieve high CO2 capture and separation performance.
6.2.2
Computing CO2 Adsorption in MOFs
Gas adsorption in MOFs is generally computed using GCMC simulations. A GCMC simulation computes the average number of adsorbed molecules in the system (gas uptakes) together with the isosteric heat of adsorption values. A series of moves, such as insertion, deletion, translation, rotation, and regrow, is used to create random configurations of gas molecules inside a MOF’s pores. These moves are accepted or rejected according to the Boltzmann probability of different configurations. As a result, configuration integrals containing potential energy functions over all positions and orientations of gas molecules inside the MOF are obtained. By solving these configuration integrals, GCMC simulation makes it possible to obtain the probability density of gas molecules at different adsorption sites in a structure considering the energy minimization of gas molecules at the thermodynamic equilibrium. In order to computationally predict gas adsorption in MOFs, GCMC simulations implemented in several software packages such as multipurpose simulation code (MUSIC) [37], RASPA [32], and Materials Studio [38] are widely used. For a GCMC simulation, the crystal structure of the MOF, gas model, gas-MOF interaction parameters, temperature (T), pressure (P), and composition of the gas feed are required. The crystal structure of a MOF contains the crystallographic information about the unit cell parameters, including cell lengths and angles, symmetry information, and fractional coordinates for all MOF atoms. Once a new MOF is synthesized, its structural information is derived via X-ray diffraction (XRD) and deposited into the CSD in the format of a crystallographic information file (cif). As long as the “cif” file is provided, one can visualize and use any MOF structure for further computational studies. MOFs can be synthesized using various techniques like solvothermal, hydrothermal, ionothermal, and microwave-assisted techniques. Details of MOF synthesis methods can be found in the literature [39–41]. Depending on the technique used for the MOF synthesis, the cif can also contain information about the residual solvent molecules, bound and unbound solvents in the pores. These solvent and guest molecules should be carefully removed from the pores before molecular simulations to imitate the experimental activation process of the MOF. It is important to note that both physisorbed and chemisorbed solvent molecules must be removed to mimic the activation process when it is successfully complete. However, some solvents, such as molecules bound on unsaturated metal centers, may be required to assure the structural stability. Therefore, after the removal of solvent process for a given structure, it is strongly suggested to check the stability information available in its synthesis paper [42]. A geometry
210
C. Altintas et al.
optimization process is also suggested especially after bound and unbound solvent removal to determine the final structure of MOFs before further computational investigations [43]. The number of experimentally synthesized MOFs in the CSD has quickly reached ~100000 in less than 30 years, and these structures are not completely computationready to be used in molecular simulations due to the practical problems such as the presence of disordered atoms, residual solvent molecules, missing hydrogen atoms, and/or extra framework ions in the cifs. Therefore, high-throughput computational screening of experimental MOFs mandates using a computation-ready experimental MOF structure database. To prepare such a database, primarily a MOF identification criterion and an algorithm to detect and clean the solvents from MOF’s pores are required. The first publicly available computation-ready experimental MOF database [44], CoRE MOF 2014, consisted of 5109 MOFs which are three-dimensional with pore sizes larger than 2.4 Å. The updated CoRE MOF 2019 version has 14,142 allsolvent-removed MOFs which included the updated MOFs in the CSD, reconstructed versions of some disordered MOFs, and some MOFs provided only with their synthesis papers and not deposited into the CSD [45]. In 2017, a CSD-integrated non-disordered MOF subset was provided using seven different MOF identification criteria together with a solvent removal script [46]. This database has been updated in accordance with the CSD, and it enables computational researchers to create their made-to-order MOF database which can also be nodeoriented or linker-oriented. These computation-ready MOF databases significantly facilitate the computational screening of MOFs, but automation of MOF identification and solvent removal is not trivial and might sometimes create problems. Keskin’s group recently compared CoRE MOF and non-disordered CSD MOF subset databases and identified the common MOFs reported with the same reference codes but with different structural properties [42]. Therefore, all computation-ready MOF structures, especially promising materials identified from the initial screening, are suggested to be carefully checked since a minor structural difference in MOFs might significantly affect the simulation results. Other than the experimentally synthesized MOFs, there are also hypothetical, computer-generated, MOF structures. Hypothetical MOFs can be designed with different MOF building algorithms. For example, in the “top-down” approach, nodular and linking building blocks are mapped on a selected topology, while in the “bottom-up” approach, the topology is identified at the end according to the connection of building blocks. The presence of various hypothetical MOF databases [47–49] helps the establishment of structure-performance relations and identification of promising structures using a very large set of MOFs. On the other hand, it is important to note that synthesizing exactly the same MOF with its hypothetical equivalence may not always be possible, and/or the synthesized hypothetical MOFs may have stability issues. Once the crystal structure of a MOF is ready, the potential energy of pairwise interactions is calculated to determine the Boltzmann distribution of particles [50]. This potential energy function includes both the nonbonded and bonded terms, and it is used to calculate the interaction energy between gas molecules and
6 Computational Screening of MOFs for CO2 Capture
211
Table 6.1 The representation of CO2 molecule by EPM2 and TraPPE force fields where ε is the energy parameter, σ is size parameter, and kB is the Boltzmann constant CO2 model C-C O-O
EPM2 ε/kB (K) 28.129 80.507
σ (Å) 2.757 3.033
q (e) 0.6512 0.3256
TraPPE ε/kB (K) 27.0 79.0
σ (Å) 2.80 3.05
q (e) 0.70 0.35
the framework. In order to compute CO2-MOF interactions, first a gas model is required. A gas model provides the physical properties and interaction parameters for a specific molecule. Due to the quadrupole moment of CO2, the potential energy function includes both the van der Waals and electrostatic interactions between MOF atoms and CO2 molecules. EPM2 is one of the preliminary force fields defining the potential energy of CO2 which is developed based on the pure component vaporliquid coexistence curve of CO2 molecule [51]. Transferable potentials for phase equilibria (TraPPE) is another mostly preferred force field for CO2 since it better estimates the binary mixture vapor-liquid equilibria properties of CO2 molecule [52]. According to these force fields, CO2 molecule is a three-site molecule consisting of two oxygen and one C atoms with a C-O bond length of ~1.16 Å and O-C-O bond angle of 180 . The representation of CO2 molecule defined by both force fields is given in Table 6.1 with the parameters. After defining the gas model to compute MOF-CO2 interactions, interatomic potential parameters for MOF atoms can be acquired from the quantum chemistry calculations or from the classical force fields. The nonbonded, intermolecular interactions between two particles are commonly known as van der Waals forces and long-range Coulombic interactions. Lennard-Jones (LJ) 12-6 potential is generally used to compute the van der Waals interactions in GCMC simulations. ULJ
" 6 # σ ij 12 σ ij ¼ 4εij r ij r ij
ð6:1Þ
In Eq. 6.1, ULJ is the intermolecular potential energy between two particles (i and j), r is the distance of separation from the center of one particle to the center of the other particle, εij is the well depth, and σ ij is the molecular length scale based on the particle diameter. The parameters, σ and ε, in the LJ potential can be taken from a force field. It is important to note that quantum chemistry-based calculations can provide a very accurate description of material properties at the expense of high computational resources. Classical force fields are mostly preferred due to their ease of applicability and transferability to different systems. Classical force fields which consider intermolecular interactions between large groups of molecules can be
212
C. Altintas et al.
developed based on fitting to experimental gas adsorption isotherms, vapor-liquid equilibrium curves, thermal expansion, or elasticity properties [53, 54]. Dreiding, universal force field (UFF), optimized potentials for liquid simulations (OPLS), UFF4MOF, MOF-FF, and QuickFF are the force fields utilized for simulations of MOFs. Dreiding [55] and UFF [56] are the two most used classical force fields. Dreiding force field considers general force constants and geometry parameters based on simple hybridization [55]. These force field parameters were mostly available for nonmetal elements and a few metals (e.g., Na, Ca, Zn, Fe). UFF parameters are derived based on the element; its hybridization, connectivity, and UFF can describe all the elements in the periodic table. More details about force fields can be found in a recent review [53]. Lorentz-Berthelot mixing rules which are the arithmetic average of parameters, (σ ij ¼ (σ i + σ j)/2), and geometric average sizep ffiffiffiffiffiffiffiffiffiffiffiffiffiffi of energy parameters, εij ¼ εi ε j , are commonly used to determine the pair interaction parameters between atoms. Electrostatic potential energy (Uelec) is commonly computed by using the Coulomb potential as shown in Eq. 6.2 where ε0, qi, and qj show the dielectric constant, partial atomic charges of i and j, respectively. Uelec ¼
1 qi q j 4πε0 r ij
ð6:2Þ
Partial atomic charges for each atom of a MOF should be calculated to compute the electrostatic interactions between MOF atoms and CO2 molecules since the latter have a quadrupolar moment. While partial charges for gas molecules can be obtained from the gas model, partial charges of MOF atoms are mostly calculated using partial charge assignment methods. Density functional theory (DFT)-based partial charge assignment methods can be utilized for assigning high-accuracy partial charges to MOF atoms, while approximate charge equilibration (Qeq) methods can be used in a faster manner for obtaining partial charges with a reasonable consistency. The ongoing modifications on Qeq methods to better estimate the electrostatic potential of MOFs have been recently addressed [57]. The only MOF database provided with the high-accuracy partial charges consists of 2932 density derived electrostatic and chemical (DDEC)-charged CoRE MOFs [58]. The DDEC method is a DFT-based method calculating the net atomic charges that can reproduce the electrostatic potential of a structure [59]. Furthermore, DFT-optimized DDEC-charged CoRE MOF containing 879 MOFs provides geometry-optimized and charged MOFs based on DFT calculations [43]. So far we reviewed the nonbonded intermolecular interactions which are considered when the rigid framework assumption is applied for MOFs. However, it is known that there are “stimuli-responsive” MOFs which have bonds that can behave flexibly upon different stimuli including temperature, pressure, and guest adsorption [60–64]. Modeling flexible MOFs requires a flexible force field which considers the intramolecular interactions including bond stretching, bending, and torsional potential terms besides the intermolecular interactions [53]. Due to the transferability
6 Computational Screening of MOFs for CO2 Capture
213
problem of the currently available flexible force fields among various MOFs and computational expense of flexible simulations, high-throughput computational screening studies are generally conducted using the rigid framework assumption. Conditions of GCMC simulations such as pressure (P), temperature (T ), and composition of the feed gas can be determined based on the real process conditions. The pressure of the bulk gas is converted to chemical potential (μ) using a gas equation of state like Peng Robinson in GCMC simulations. GCMC simulations are conducted in a μVT ensemble in which μ, V (volume of the MOF), and T are kept constant and the number of molecules, N, is allowed to change. A thermostat algorithm is used to keep the temperature constant. Different thermostat algorithms such as Nosé-Hoover [65] and Andersen [66] can be used as a heat bath coupled with the MOF adsorbent system. Detailed explanation of these methods can be seen elsewhere [50]. At the end of a GCMC simulation, Henry’s constants (KH), gas uptakes from single-component or mixture gas feed (Ni or N mix i , respectively), and heat of adsorption (Qst) values indicating the strength of gas adsorption can be obtained which can be further used to determine the adsorbent performance evaluation metrics of MOFs as we will discuss below. The Henry’s constant represents the affinity of the gas molecule towards the MOF adsorbent and can be calculated with GCMC simulations that are conducted at very low pressure to imitate infinite dilution conditions (one gas molecule interacting with the MOF adsorbent) or via using the Widom particle insertion method [67].
6.2.3
Calculating CO2 Separation Performances of MOFs
After computing the adsorbed CO2 amounts for a MOF from a GCMC simulation, this data is then processed to evaluate CO2 separation performances of MOFs. Several adsorbent performance evaluation metrics are used to accurately assess the CO2 separation performance of MOFs under real process conditions. For example, CO2 working capacity (ΔN CO2 (mol/kg)) shows the amount of CO2 that can be released when the pressure is decreased after adsorption. It is simply the difference between the CO2 uptakes at adsorption (N CO2 ,ads ) and desorption (N CO2 ,des ) pressures. ΔN CO2 ¼ N CO2 ,ads N CO2 ,des
ð6:3Þ
It is a better practice to calculate the working capacity with uptakes from gas mixture adsorption (N mix CO2 ) although this metric can be also computed using the single-component data [28, 68]. The gas uptake at desorption condition is usually obtained using feed composition of the gas mixture. It can also be taken as singlecomponent CO2 uptake due to the assumption that at the end of the depressurization step, the adsorption column will contain almost pure CO2 [69].
214
C. Altintas et al.
Vacuum swing adsorption (VSA), pressure swing adsorption (PSA), and temperature swing adsorption (TSA) are cyclic adsorption processes in which CO2 capture is followed by the CO2 release process where desorption conditions are set to release the vast majority of adsorbed CO2 for further processing and to regenerate the MOF for reuse [70]. Percent regenerability (R%) is a metric to evaluate the practicality of a MOF for repeated usage in cyclic gas adsorption processes. It can be calculated as follows: R% ¼ ðΔN CO2 =N CO2 ,ads Þ 100%
ð6:4Þ
A good MOF adsorbent for CO2 capture should have high R% values in addition to high N CO2 ,ads and high ΔN CO2 , so that it can capture a significant amount of CO2 between specified pressures for several times without a frequent need to replace the adsorbent. Adsorption selectivity (Sads) is another important metric which is the ratio of the uptake of the most strongly adsorbed gas component (CO2) to other gas species ( j) at adsorption pressure. It helps to determine to what extent the MOF adsorbent can separate the gas species from other impurities. Ideal selectivity can be defined as the ratio of Henry’s constants for different gas species (KH,i and KH,j) or the ratio of single-component gas uptakes obtained from GCMC simulations as follows where N CO2 is the adsorbed amount of CO2 and Nj is the adsorbed amount of less strongly adsorbed gas species j. Sads,ideal ¼ K H,CO2 =K H,j or Sads,ideal ¼ N CO2 =N j
ð6:5Þ
Computational determination of ideal selectivity is rather simple and requires less time compared to mixture selectivity. Therefore, it is frequently used as an initial screening criterion to identify the most promising MOF adsorbents based on their selectivities. However, under real operating conditions, in the presence of other gas components and impurities in the gas feed, collaborative and competitive gas adsorption can be observed [71, 72]. For example, electrostatic interactions or pore topology can strongly favor the adsorption of CO2 molecules, and the other molecules need to compete with CO2 for the available adsorption sites. Collaborative effects such as CO2-CO2 interactions can attract more CO2 molecules into the framework. In order to take these effects into account, mixture selectivity should be considered. It can be calculated using the ratio of adsorbed gas amounts obtained from mixture GCMC simulation normalized with the bulk composition of the gas mixture as shown in Eq. 6.6 (xCO2 represents the mole fraction of adsorbed CO2, xj represents the mole fraction of less strongly adsorbed gas j, yCO2 represents bulk feed composition of gas species CO2, and yj represents bulk feed composition of gas species j).
6 Computational Screening of MOFs for CO2 Capture
Sads,ðCO2 ,jÞ ¼
xCO2 =x j yCO2 =y j
215
ð6:6Þ
Experimental investigation of mixture gas adsorption for various compositions of a gas mixture is challenging due to the lack of proper instrumentation and expertise to determine the gas mixture adsorption isotherms and also lack of time and human and financial resources [73]. Determination of the adsorption loadings for each gas species through weight changes or pressure changes is also difficult. This requires using theoretical models that can predict mixture gas adsorption from the singlecomponent gas adsorption data. Ideal adsorbed solution theory (IAST) predicts mixture adsorption isotherms from the single-component gas uptakes based on an adsorption model such as the Langmuir model. It is based on the assumptions that all gas molecules have equal accessibility to the whole surface area of a homogeneous adsorbent, gases interact with the adsorbent similarly in strength, and there are no interactions between adsorbed gas molecules mimicking an ideal solution [74]. IAST method is shown to accurately work for gas adsorption in MOFs unless the MOF has an energetically heterogeneous surface [75]. A good adsorbent is expected to provide a combination of high ΔN CO2 and Sads,ðCO2 ,jÞ . However, there is generally a trade-off between these two metrics [76]. Porous materials with large pore sizes generally exhibit high gas working capacities, but due to the large pores, both gas molecules can be adsorbed within the adsorbent leading to a decrease in adsorption selectivity. Therefore, other metrics combining working capacity and selectivity are developed to provide a better estimation of MOFs’ performances. Adsorbent performance score (APS) [69] and PSA sorbent selection parameter (SSP) [68] can be defined as follows: APS ðmol=kgÞ ¼ ΔN CO2 Sads SSP ¼
ΔN CO2 Sads ΔN j
ð6:7Þ ð6:8Þ
Maximum achievable productivity or separation potential (ΔQ) for PSA is a different metric which combines volumetric gas uptake capacity (QCO2 and Q j ) and bulk fluid mixture mole fractions of each gas species to calculate the amount of weakly adsorbed component recovered during the adsorption phase of fixed-bed separations [77]. ΔQ ðmol=LÞ ¼ QCO2
yj Qj 1 yj
ð6:9Þ
It is important to note that instead of performing costly breakthrough simulations, this metric can be easily used to determine the achievable gas productivities in a fixed-bed unit [77]. As CO2 has a relatively high heat of adsorption, considering thermal effects during CO2 adsorption and desorption is important [78]. Adsorbent performance
216
C. Altintas et al.
indicator (API) [79] takes the heat generated during adsorption of the strongly adsorbed species (ΔH ads,CO2 ) into consideration besides ΔNi and Sads ðCO2 ,jÞ . API ¼
ðSads 1Þ ΔN CO2 jΔH ads,CO2 j
ð6:10Þ
It is very important to develop metrics that can reflect the cost of adsorption-based CO2 separation with MOFs. Parasitic energy [80] is a parameter to determine the energy load imposed on a power plant by CCS. In other words, it helps to estimate the difference between energy gained from burning fossil fuel and energy spent on capturing the CO2 produced by burning this fossil fuel. It is the total of the energy required to regenerate the adsorbent (Qthermal) and energy required to compress the gas to transport conditions (Wcomp). Heat is factored with Carnot efficiency (ηcarnot) and the efficiency of the gas turbine (75%, the general efficiency of a Carnot cycle). Eparasitic ¼ ð0:75 Qthermal ηcarnot Þ þ W comp
ð6:11Þ
To compare the performance of materials for CCS, the compression pressure of CO2 is commonly considered as 150 bar which is the standard requirement for transport and storage [81]. MOFs that exhibit lower parasitic energy compared to the energy used in the current MEA technology in CCS, 1060 kJ/(kg CO2), can be potential adsorbents to reduce the parasitic energy of CCS [81]. For example, covalent organic frameworks (COFs), covalently bonded organic structures, were geometrically optimized and recently deposited in a curated COF database [82], and they were examined to identify the best candidates having the lowest parasitic energy for CO2 capture. The performances of COFs were found to be below the best performing MOF, MOF-74 which has 0.705 MJ per kg of CO2, with a purity of 0.943 and a volumetric working capacity of 64.87 kg of CO2 per cubic meter of bed. A recent work by Sholl’s group [83] compared the predictions of adsorbent evaluation metrics (working capacity, adsorption selectivity, sorbent selection parameter, adsorbent performance score, regenerability) with the results of detailed process models (purity, recovery, productivity, energy) for dry CO2/N2 separation at sub-ambient temperatures at PSA conditions for 143 MOFs. Their results showed that CO2 working capacity and adsorbent performance score are the two parameters that can best predict the process-level rankings of MOFs for dry flue gas separation. From a different perspective, shape selectivity (γ ij), which considers the difference of activation energies required for the transport of unlike gas molecules through the pores of an adsorbent, was proposed to screen MOFs for CO2 separation [84]. This approach required the identification of all portals, channels, cages, and their connectivity in MOFs and the activation energy required for gas molecules to pass through these paths. Activation energy of molecules depends on their size and shape. Shape selectivity of different molecules can be calculated as follows for the minimum energy pathway (minp) in the MOF where Eip is the energy of pathway p for guest molecule i and R and T are the gas constant and temperature.
6 Computational Screening of MOFs for CO2 Capture
! ! min p E ip min p E pj exp γ ij ¼ exp RT RT
217
ð6:12Þ
A shape selectivity of zero means that molecules have similar activation energies, while a shape selectivity of unity indicates high selectivity. This approach was used for the separation of CO2/CH4, CO2/N2, and CO2/H2 and reconfirmed some of the previously identified promising MOFs [84].
6.3 6.3.1
Large-Scale Molecular Simulations of MOFs for CO2 Capture Refining MOF Databases
Before using high-throughput molecular simulations to screen MOFs, the MOF database is generally narrowed down to a smaller number of materials having the desired structural/chemical properties. For example, Watanabe and Sholl extracted MOFs from the CSD, narrowed down this initial MOF database to a smaller set of 3D and solvent-free MOFs, and finally performed molecular simulations only for the MOFs which have PLDs between 2.2 Å and 3.6 Å for CO2/N2 separation as shown in Fig. 6.2a [85]. In a similar manner, Keskin’s group [86] started with 54808 non-disordered MOFs and narrowed down the number of MOFs by eliminating materials with zero accessible surface area and PLD < 3.8 Å as shown in Fig. 6.2b. It is also possible to narrow down a MOF database by first excluding the materials that will have a higher affinity for a specific gas component in the mixture. Calculation of Henry’s constants is a rather easier simulation compared to single-component or multicomponent gas adsorption simulations under predefined operating conditions; therefore it is also used as a starting step for screening MOFs that can separate CO2 under humid conditions [87–89]. Such an approach was recently used to screen the CO2 capture performance of CoRE MOF database in the presence of water as shown in Fig. 6.2c [87]. In another work, Henry’s constants for water (KH2 O ) were calculated for many MOFs, and KH2 O of a known hydrophobic MOF was used as a threshold to identify the hydrophobic MOFs with KH2 O values lower than that threshold value [89]. This elimination made it possible to screen hydrophobic MOFs in which H2O, which is a strong competitor of CO2 in adsorption, would not be able to occupy available adsorption sites for CO2. Before discussing the details of studies on large-scale molecular simulations of MOFs for CO2 capture, we would like to note that every work has a different set of criteria such as the starting MOF database, criteria used to narrow down the initial MOF database, different parameters used in molecular simulations, and differences in the adsorbent performance evaluation metrics used to identify the most promising MOFs for a target gas separation.
218
C. Altintas et al.
Fig. 6.2 Refining initial MOF database according to different criteria. (Reproduced from (a) Watanabe and Sholl [85], (b) Altintas et al. [86] and (c) Li et al. [87]. Further permissions related to the material excerpted should be directed to the American Chemical Society)
6.3.2
Screening of MOFs
Initial computational materials screening studies on CO2 capture were based on zeolites [81], and a similar computational screening methodology was applied to MOFs. In this section, we focused on the recent large-scale (>100 MOFs) molecular simulation studies for CO2 adsorption and separation. Table 6.2 includes 30 computational studies which have screened a large number of experimentally synthesized and/or hypothetical MOFs for CO2 capture at various operating conditions considering several adsorbent performance evaluation metrics. Each study represented in Table 6.2 has a unique methodology to refine the initial MOF database such as narrowing down the number of MOFs based on the pore sizes, accessible surface areas, interpenetration of the frameworks that were generated by arbitrary concatenation of two or more frameworks, or isosteric heats of adsorption values. Early studies represented in Table 6.2 focused on experimental MOFs obtained directly from the synthesis papers and/or the CSD. For example, Wu et al. [90] studied 105 MOFs for post-combustion CO2 capture from CO2/N2 gas mixture at 1 bar, 298 K. Figure 6.3 shows the relation between the porosity of MOFs and
424
Hypothetical MOFs and experimental MOFs CoRE MOFs
4764
324500
Hypothetical MOFs
~83,000 non-interpenetrated MOFs, 58,000 randomly selected MOFs to train/calibrate the models Randomly selected 292,050 MOFs to form the calibration set used for QSPR models pcu, fcu, and ftw topologies, Zn4O and Zr6O4 (OH)4 organic clusters and 20 linkers. None of the MOFs have open metal sites Direct usage of the database
An algorithm to generate structures
~137000
~130000
Converged charges
Criteria for initial screening MOFs including IRMOFs, ZIFs, PCNs 3D network, with PLD in the range from 2.2 Å to 3.6 Å
489
359
Number of MOFs 105
Hypothetical MOFs
Ockwig et al.’s [77] list and CSD Hypothetical MOFs
Database Synthesis papers CSD
Single-component CO2 uptake at 0.2 bar and 1 bar, 298 K, singlecomponent CH4 uptake at 30 bar and 100 bar, 298 K CO2/N2: 15/85 mixture, 1 bar and 0.01 bar and 298 K, CO2/ CH4:50/50 mixture, 5 bar and 0.5 bar at 298 K
CO2/CH4:10/90 and CO2/ CH4:50/50 mixtures, 1 bar and 5 bar and 298 K, CO2/N2:10/90 mixture at 1 bar and 298 K Single-component CO2 uptake at 0.15 bar and 1 bar, 298 K
Single-component CO2, CH4, and N2 adsorption at 298 K
Conditions CO2/N2:15/85 mixture, 1 bar and 298 K Single-component CO2 and N2 adsorption from 106 to 100 bar at 303 K Single-component CO2, CH4, and N2 adsorption at 303 K
Table 6.2 Studies on high-throughput computational screening of MOFs for CO2 capture
[93]
[94]
[95]
N CO2 ,ads , ΔN CO2 , R%, Sads,ðCO2 ,N2 Þ
N CO2 ,ads and QSPR models
N CO2 ,ads and QSPR models
ΔN CO2 and R% (continued)
[97]
Qst for CO2, CH4 and N2, Sads,ðCO2 ,N2 Þ and Sads,ðCO2 ,CH4 Þ at infinite dilution,
[96]
CO2 and CH4 uptakes
, Sads,ðCO2 ,CH4 Þ , and SSP
[92]
[85]
References [90]
Sads,ðCO2 ,N2 Þ and Sads,ðCO2 ,CH4 Þ at infinite dilution
Adsorbent performance metrics Qst for CO2 and N2 at infinite dilution and Sads,ðCO2 ,N2 Þ at 1 bar, 298 K Sads,ðCO2 ,N2 Þ at infinite dilution and saturated CO2 uptakes at 100 bar
6 Computational Screening of MOFs for CO2 Capture 219
DDEC-charged CoRE MOF
CoRE MOFs
Hypothetical MOFs
Hypothetical MOFs Hypothetical MOFs
Hypothetical MOFs CoRE MOF and CSD
Database CoRE MOFs
1627
1.65 trillion structures 879
137953
~320000
100
55163
Number of MOFs 2054
Table 6.2 (continued)
MOFs with nonzero surface area
Geometrically optimized 879 MOFs and 465 DDECcharged MOFs
A set of 292,050 diverse MOF structures 81,679 non-interpenetrated MOFs, which also exhibit reasonable pore size >1.4 Å and surface area > 100 Å2 Sterically viable 96,156 structures
Criteria for initial screening MOFs having nonzero surface area and low adsorption energy of H2O (3.30 Å
606 hydrophobic MOFs
559
>400
427
54808
54808
6013
ToBaCCo
CSD non-disordered MOF subset
CSD non-disordered MOF subset CoRE MOF
MOFs with 16 different topologies with targeted porosity in the range of 0.50 to 0.85 3816 MOFs having nonzero surface areas and PLDs >3.8 Å
PLD > 2.4 Å indicating a sufficient window size to admit CO2
477
DFT-based DDEC-charged CoRE MOF Hypothetical MOFs ToBaCCo
H2O/N2 ternary mixture simulations at 298 K Single-component CO2 adsorption at 0.1 and 2 bar, 213, 228, 243, and 258 K Single-component CO2 adsorption at 10 bar and 298 K Single-component CO2 adsorption at 1 bar and 313 K, CO2/H2: 20/80 at 20 bar adsorption and pure CO2 at 1 bar desorption conditions; CO2/N2: 15/85 at 1 bar adsorption and CO2/N2: 90/10 at 0.1 desorption conditions Single-component CO2 adsorption isotherm up to 20 bar and 298 K CO2/N2: 15:85, CO2/CH4: 50:50, and CO2/N2/CH4: 10:70:20 mixtures, at 1 bar adsorption and 0.1 bar desorption conditions, 298 K Henry constant calculations for CO2 and H2. CO2/H2:15/85 mixture at 298 K, 0.1, 1, and 10 bar Henry constant calculations for H2O, CH4/C2H6/C3H8/H2S/CO2/ H2O mixture with the mole fraction of 0.70/0.10/0.05/0.05/0.10/ 0.0031 at 298 K and 10 bar [86]
N CO2 ,ads , ΔN CO2 , R%, Sads,ðCO2 ,N2 Þ
[89]
Qst, for H2S and CO2 at infinite dilution, N CO2 ,ads , N H2 S,ads , and Sads,ðH2 SþCO2 =C1 C3 Þ
(continued)
[106]
N CO2 ,ads , ΔN CO2 , R%, Sads,ðCO2 ,H2 Þ , APS, and separation potential
potential
, Sads,ðCO2 ,CH4 Þ , APS and separation
[105]
N CO2 ,ads , Henry constants for CO2, H2O, H2 and N2, and Qst, at 0.1 bar
[104]
[103]
N CO2 ,ads and Qst N CO2 ,ads and Sads,ðCO2 ,N2 Þ , Sads,ðCO2 ,H2 Þ
[102]
N CO2 ,ads , ΔN CO2, and Qst
6 Computational Screening of MOFs for CO2 Capture 221
100
358400
3399
ToBaCCo
CoRE MOF
PLD > 3 Å and sufficient ionization potential or electron affinity data in the EQeq
Randomly chosen
Randomly chosen
DFT-optimized 308 COFs
324
DFT-based DDEC-charged CoRE MOF
2338 MOFs considering def2qzvpp basis set
2338
DFT-based DDEC-charged CoRE MOF COF database
Criteria for initial screening 68 MOFs from CoRE MOF, and 32 MOFs including MOF-5, CuBTC, ZIFs ΔNCO2 > 2 mol/kg, CO2/N2:15/ 85 mixture selectivity >50 to determine MOFs to investigate CO2 binding sites
325000
Number of MOFs 100
ToBaCCo
Database CoRE MOF and CSD
Table 6.2 (continued)
Single-component CO2 and N2 adsorption isotherm up to 30 bar and 300 K CO2/CH4:50/50 and C2H6/ C2H4:50/50 mixtures at a total pressure of 20 bar, Xe/Kr mixtures at 40 bar, and C3H8/C3H6/ C4H10 equimolar mixtures at 10 bar CO2/H2:40/60 mixture at 1 bar and 40 bar, 313 K Equimolar CH4/N2, CH4/CO2, CH4/H2S, and CH4/NH3 mixtures at 298 K and 1 bar
Conditions CO2/H2:15/85, CO2/N2:15/85, CO2/CH4:50/50 at 0.1 bar, 1 bar and 298 K CO2/N2:15/85 mixture, 1 bar, 298 K adsorption, and 0.1 bar, 363 K desorption conditions Henry constant calculations for CO2, H2O and N2, CO2/H2O/N2 ternary mixture simulations Single-component CO2 adsorption at 0.2 bar and 298 K
[110] [111]
ΔN CO2 , Sads,ðCO2 ,H2 Þ , QSPR models N CO2 ,ads , ΔN CO2 , ΔN N2 , ΔN H2 S , ΔN NH3 , R%, Sads,ðN2 ,CH4 Þ , Sads,ðNH3 ,CH4 Þ , Sads,ðH2 S=CH4 Þ , and Sads,ðCO2 ,CH4 Þ
[109]
[82]
Parasitic energy
Sads,ðCO2 ,CH4 Þ at dilute and nondilute conditions
[57]
[108]
References [107]
N CO2 ,ads , Qst for CO2 at infinite dilution
Henry constants for CO2, H2O, H2, and N2, N CO2 ,ads , Sads,ðCO2 ,N2 Þ , and N mix CO2
, Sads,ðCO2 ,CH4 Þ , Sads,ðCO2 ,H2 Þ and SSP
Adsorbent performance metrics N CO2 ,ads , ΔN CO2 , R%, Sads,ðCO2 ,N2 Þ
222 C. Altintas et al.
6 Computational Screening of MOFs for CO2 Capture
223
Fig. 6.3 Effect of porosity and difference in isosteric heat on the selectivity of MOFs. φ represents porosity, ΔQ0st represents the difference of isosteric heats of adsorption between CO2 and N2 at infinite dilution, and color scale represents CO2/N2 selectivity at 1 bar, 298 K. (Reproduced from Ref. [90])
difference of isosteric heats of adsorption between CO2 and N2 at infinite dilution based on mixture selectivity calculated for CO2/N2:15/85 mixture at 1 bar, 298 K. As porosities of MOFs decrease, the difference of isosteric heats of adsorption between CO2 and N2 at infinite dilution increases, leading to high CO2/N2 selectivities. Watanabe and Sholl [85] first identified ~30000 MOFs in the CSD including ZIFs, which is a class of MOFs constructed by connecting tetrahedral metal ions such as Zn and Co with imidazolate linkers, and concentrated on 359 MOFs which have pore sizes in the range of 2.2 Å to 3.6 Å for flue gas separation. They computed CO2/N2 selectivities of these MOFs at infinite dilution and saturated CO2 uptakes at 100 bar. Several ZIFs, which have the highest surface areas (1232–1427 m2/g) among these materials such as ZIF-90, ZIF-8, and ZIF-65, exhibited the highest CO2 uptakes (10–12 mol/kg). Once the experimental CoRE MOF database was created [44], several research groups screened this database to identify the promising MOF adsorbents for CO2 capture. For example, Qiao et al. [97] examined the potential of 4764 CoRE MOFs for CO2 capture from natural gas and flue gas mixtures. They discussed that based on the computed metrics including CO2 working capacity, selectivity, and regenerability, lanthanide-based MOFs with open metal sites exhibit the highest CO2 separation performance whereas alkali-MOFs exhibit the lowest separation performance for CO2/CH4 and CO2/N2 mixtures. Sumer and Keskin [98] also examined the potential of 100 MOFs consisting of 68 CoRE MOFs and 32 widely studied MOFs such as MOF-5, Cu-BTC, ZIFs, and COFs for CO2 capture from CO2/CH4, CO2/N2, and CO2/
224
C. Altintas et al.
H2 mixtures. The top-performing MOFs identified for these separations using the ranking of MOFs based on either CO2 selectivity or CO2 working capacity were found to be similar, except the top MOFs ranked based on their regenerability values. Therefore, they suggested that initial large-scale screening of MOFs can be done by considering the regenerability criteria (should be at least 75% for an efficient adsorption-based CO2 separation) and then the remaining MOFs can be ranked based on their CO2 selectivities and working capacities. In a follow-up study, using the same MOF database, Dokur and Keskin [107] examined the effect of force field type on four adsorbent evaluation metrics of MOFs including selectivity, working capacity, sorbent selection parameter, and regenerability for CO2/H2, CO2/N2, and CO2/CH4 separations. Although gas uptakes obtained from GCMC simulations using UFF and Dreiding were different, the rankings of MOFs based on CO2 selectivity and regenerability were similar, indicating that both force fields can be used in highthroughput molecular simulations of MOFs to determine the potential adsorbents for CO2 capture. Similar results were also discussed by McDaniel et al. [96] where the ranking of 424 MOFs based on predicted CO2 uptake using different force fields did not change significantly. These computational studies provided a useful map to narrow down the material space to identify the best adsorbents for CO2 separation. Li et al. [87] examined the potential of CoRE MOFs for CO2 capture under 80% relative humidity at 298 K using a quick charge assignment method (EQeq) and a more accurate DFT-based charge assignment method (REPEAT). Henry’s law constants of H2O obtained from the GCMC simulations with EQeq charges were underestimated compared to the results obtained from the simulations with REPEAT charges. It was found that Coulombic interactions dominate H2O adsorption in MOFs, and therefore, selecting the charge assignment method for the atoms of MOFs is important to accurately describe the H2O-MOF interactions. Similar results were obtained in a follow-up study [88] where Henry’s constants of CO2, N2, and H2O were calculated using DDEC and EQeq charges. Results showed that CO2/H2O selectivities depend strongly on the partial charge assignment method. These works showed that MOFs with small pore sizes favored adsorption of CO2 and hindered H2O uptake at high humidity, but they might provide low CO2 working capacity. This idea was used in a recent study where 606 hydrophobic MOFs from the CoRE MOF database were initially identified and tested for H2S and CO2 separation from natural gas under humid conditions (including a gas mixture of CH4/C2H6/C3H8/ H2S/CO2/H2O) [89]. Results showed that MOFs with nitrogen (N)-rich organic linkers such as pyridine and azoles offer the best performance for H2S and CO2 separation, indicating that designing functionalized MOFs with these types of linkers will be a useful strategy for the efficient purification of natural gas. Multistage screening of CoRE MOFs was also used for shortlisting MOFs that can effectively purify CH4 from CO2, N2, H2S, and NH3 [111]. MOFs having a PLD larger than 3 Å and high CH4/CO2, CH4/N2, CH4/H2S, and CH4/NH3 selectivities were further studied for separation of five-component CH4/CO2/N2/H2S/NH3 mixture. Considering the combination of the highest values obtained for three criteria (working capacity, selectivity, and regenerability) for each component of the gas mixture, the top-performing MOFs were identified.
6 Computational Screening of MOFs for CO2 Capture
225
All these computational screening studies summarized above utilized different partial charge assignment methods. Especially for the MOFs where electrostatic interactions strongly influence the adsorbent-adsorbate interactions, the accuracy of the partial charges assigned to MOFs is important. The most accurate partial charges are the ones which can accurately describe the electrostatic potential surface of the MOF, but the trade-off between accuracy and cost of computation mandates researchers to use approximate partial charges for the screening of MOFs. The promising candidates identified with approximate methods can be further investigated using more accurate partial charge assignment methods. Haldoupis et al. [92] proposed the periodic Qeq method (PQeq) for the application of the Qeq method to fully periodic structures of MOFs and screened ~500 MOFs to obtain Henry’s constants for CO2 and N2. MOFs which exhibited the highest CO2/N2 and CO2/ CH4 selectivities with PQeq charges were found to be selective also when calculated with DDEC charges. Collins and Woo [103] developed a set of split-charge equilibration parameters, namely, SQE-MEPO (split-charge equilibration MOF electrostatic potential-optimized) for screening 559 MOFs for CO2 adsorption. They showed that CO2 uptake predictions using SQE-MEPO charges are in good agreement with the predictions using DFT derived charges, indicating that SQE-MEPO can be used for high-throughput screening of MOFs. In another study, Argueta et al. [105] developed a molecular building block-based (MBBB) charge assignment method to make it possible to have a library of MOF building blocks (nodes and edges) with high-accuracy partial charges for automated MOF generation. A comparison of CO2 uptakes obtained with MBBB, EQeq, CBAC, and REPEAT charges showed that partial charges assigned with different charge methods might create strong adsorption sites at different regions in a MOF and therefore can lead to significantly different CO2 uptakes. DDEC-charged and geometry-optimized CoRE MOFs were studied by Park et al. [102] to identify the promising MOFs that exhibit high CO2 swing capacity (working capacity) of a sub-ambient (from 213 to 258 K) PSA process. They reported that 20 MOFs among 477 MOFs provide a swing capacity over 10 mol/kg at all sub-ambient temperatures. This study discusses the importance of sub-ambient CO2 capture processes to improve the CO2 working capacity of MOFs. Nazarian et al. [43] created the database consisting of 879 geometrically optimized MOFs and assigned DDEC charges to 465 of them. CH4 adsorption in most MOFs was not sensitive to small structural changes, whereas CO2 adsorption in many MOFs changed by more than 5% after energy minimization of the structures. This showed the necessity of geometry optimization following solvent removal for MOFs in computation-ready databases. They also reported that performing flexible GCMC simulations for narrow-pored MOFs with pore sizes close to the adsorbates may affect the gas uptake predictions. In a recent study, the structural flexibility effect on single-component (CO2, CH4, ethane, ethene, propane, propene, butane, Xe, and Kr) and mixture (CO2/CH4, ethane/ethene, propane/propene/butane, and Xe/Kr) adsorption properties of 100 MOFs taken from CoRE MOF database with DDEC charges was examined [109]. Results showed that single-component gas adsorption at nondilute conditions is weakly affected by the framework flexibility whereas
226
C. Altintas et al.
Fig. 6.4 Relations between the ratio of Henry’s constants (KH, f/KH, r) of flexible and rigid MOFs as a function of Δd, the difference between the MOF’s LCD and the adsorbate’s kinetic diameter. (Reproduced from Ref. [109])
adsorption selectivities change at dilute and nondilute conditions. Figure 6.4 shows the ratios of Henry’s constants as a function of the difference between adsorbate’s kinetic diameter and the MOF’s pore size. As shown in this figure, the flexibility effect on Henry’s constants is much significant for the MOFs which have pore sizes close to the kinetic diameters of adsorbates. It is important to note that the flexibility effect on Henry’s constants is also important for the MOFs which have larger LCDs than the kinetic diameters of adsorbates when the electrostatic interactions are important as in the case of CO2 as shown in Fig. 6.4. After the creation of an updated and CSD-integrated MOF database [46], Keskin’s group screened 3816 MOFs for CO2/N2 and CO2/CH4 separation [86]. They concluded that MOFs having isosteric heat of adsorption>30 kJ/mol, 3.8 Å < PLD < 5 Å, 5 Å < LCD < 7.5 Å, 0.5 < porosity 2000, respectively. (Reproduced from Ref. [106]. Further permissions related to the material excerpted should be directed to the American Chemical Society)
separation, two arbitrary APS limits were defined. 3857 MOFs were compared with zeolites for separation of CO2/H2, and a high number of MOFs was reported to outperform zeolites in terms of CO2 working capacities and CO2 selectivities. The development of new algorithms facilitated the use of hypothetical MOFs to discover the desired properties of MOF adsorbents for efficient CO2 capture. For example, MOF properties affecting their CO2 capture performances were investigated for 426 hypothetical MOFs using the code topologically based crystal constructor (ToBaCCo) in a multistaged work [104]. GCMC simulations were used to compute CO2 capture metrics, CO2/N2 and CO2/H2 selectivities, and CO2 working capacities of materials. It is found that modification of chemical properties, such as functional groups on linkers, has a significant effect on the improvement of CO2 capture performance of MOFs. Figure 6.6 shows the probability density maps for average CO2 loadings obtained from GCMC simulations for IRMOF-1 analogues with different functionalization. As shown in this figure, CO2 molecules preferentially adsorb at the nodes of the framework in the presence of -(H)4, and -(CH3)2, whereas CO2 molecules preferentially adsorb at the sites that are close to the functional groups for the -(CN)4 case, indicating that adsorption mechanism can be different based on the functionalization of MOFs. Preserving the CO2 selectivity of MOFs under humid conditions is strongly desired [112]. In a recent computational and experimental collaborative work, 325,000 hypothetical MOFs were screened for CO2/N2 separation [108]. Among these MOFs, 8325 MOFs were found to outperform commercial zeolite 13X, in terms of CO2 working capacity and CO2/N2 selectivity at post-combustion dry flue gas separation conditions (CO2 working capacity >2 mol/kg and CO2/N2 selectivity >50). The term “adsorbaphore” was then described to identify the common pore geometry in MOFs having favorable adsorption sites for CO2 molecules but not for
228
C. Altintas et al.
Fig. 6.6 Probability density maps representing average CO2 uptake distribution for the three analogous structures of IRMOF-1 with different functionalization. The –(H)4 represents the non-functionalized linker to construct the parent MOFs. (Reproduced from Ref. [104])
H2O. With reverse searching, two MOFs containing the specified adsorbaphore among the set of ~8000 hypothetical MOFs were identified and experimentally synthesized. One of the MOFs outperformed zeolite 13X and activated carbon for wet flue gas separation in terms of CO2 working capacity.
6.4
Role of QSPR and Machine Learning in Screening of MOFs for CO2 Capture
Large-scale screening of MOFs using molecular simulations for CO2 adsorption has produced an extensive amount of data as discussed above. This data has been recently combined with machine learning methods where multiple structural and/or chemical descriptors are available to establish structure-performance relations. The first extensive computational screening of MOFs to develop structureperformance relationships for CO2/CH4 and CO2/N2 separations considering the operational conditions for several PSA and VSA processes was conducted by studying over 130,000 hypothetical MOFs [93]. Correlations of properties including surface area, heat of adsorption, the maximum pore diameter and helium void fraction, and chemical functionalities with five adsorbent evaluation metrics (CO2 uptake, working capacity, adsorption selectivity, regenerability, and sorbent selection parameter) were examined using single-component gas uptakes from GCMC simulations. Wang and coworkers [113] recently discussed the current opportunities of using machine learning methods in material simulation and design. Fernandez and Woo [114] reported a large-scale quantitative structure-property relationship (QSPR) analysis for the hypothetical MOF database [48]. They developed several models including multi-linear regression (MLR) models, decision trees (DTs), and nonlinear support vector machines (SVMs) that can estimate CH4 uptakes of MOFs
6 Computational Screening of MOFs for CO2 Capture
229
based on geometric features of MOFs, and simple descriptors such as pore size and void fraction were found to be the most strongly correlated with CH4 storage properties of MOFs. Fernandez et al. [94] later introduced a novel atomic property weighted radial distribution function (AP-RDF) descriptor tailored for large-scale QSPR predictions of CO2, CH4, and N2 adsorption of MOFs using 58,000 hypothetical structures and showed that AP-RDF scores yielded QSPR models of simulated uptake capacities that outperformed the combination of simple geometrical features with prediction accuracies ranging from 70% to 83%. Woo’s group [95] developed QSPR models for CO2 capture and identified the high-performing MOFs with enhanced CO2 adsorption capacity (>1 mol/kg at 0.15 bar and > 4 mol/ kg at 1 bar) at conditions relevant to post-combustion (0.15 bar) and landfill gas purification (1 bar). The models obtained using advanced machine learning algorithms were tested on hypothetical MOFs that were not part of the training set, and QSPR classifiers were shown to recover 945 of the top 1000 MOFs in the test set. Fernandez and Barnard [100] also developed QSPR models to predict CO2 and N2 uptake capacities of MOFs using 16,000 hypothetical structures derived from the prototypes and archetype frameworks. Uptake capacities were found to be poorly correlated to the void fraction, surface area, and pore size, but these structural properties were used to build binary classifier predictors that successfully identify high-performing MOFs for CO2 and N2 uptakes. Another important application of QSPR models is the computational identification of MOFs with high CO2/CH4 selectivity. Woo’s group [99] calculated selectivity of 320,000 hypothetical MOFs using GCMC simulations at conditions mimicking the natural gas purification and applied DT and SVM models. They concluded that MOFs having void fraction 3 mol/kg at 0.15 bar and 298 K. A multi-scale approach combining DFT, GCMC, and machine learning composed of six different algorithms, DT, SVM, random forests, neural networks, gradient boosting machines, and multiple linear regression, was used to understand the role of chemical and topological features of MOFs for improved CO2 capture performance [104]. It was shown that the improvement of CO2 capture metrics of parent MOFs depends more strongly on chemistry-related descriptors, whereas the absolute values of these metrics were found to depend more strongly on topology-related descriptors. Molecular simulations of a large number of materials require significant computational time and resources; therefore application of machine learning methods not only to search for the best MOFs but also to provide insights for the design of new materials has been very useful.
6 Computational Screening of MOFs for CO2 Capture
6.5
231
Conclusions and Outlook
Recent developments in algorithms and computational tools have facilitated the high-throughput computational screening of experimental and hypothetical MOFs for CO2 capture. Based on the encouraging results of many studies that we discussed throughout this chapter, MOFs can be considered as potential adsorbents for CO2 storage and separation. Several issues which require further experimental and/or computational investigation can be considered as the future directions of the field: • Large-scale molecular simulations are highly useful to direct the researchers to the top MOF candidates for CO2 capture. However, it is important to note that the top materials are identified based on several assumptions used in the molecular simulations. For example, high-throughput molecular simulations generally consider solvent-free, activated, rigid MOF structures. MOFs may not be perfectly crystalline, and they may have defects in their structures. After the solvent removal, some MOFs can lose their crystalline structure and collapse. Some MOFs can form into a different structure upon an external stimulus like temperature and pressure due to their flexibility. All these can affect their CO2 capture and separation performance under real conditions. Another important point is that a good MOF adsorbent should have a good thermal, mechanical, and chemical stability to find a place in real gas adsorption and separation processes. Especially, the stability of MOFs under a humid environment is very important for CO2 separation from flue gas since the presence of water vapor may adversely affect its CO2 selectivity and even the stability of the material. These issues, especially stability, are most likely to be examined by further experimental studies. The development of transferable flexible force fields will be also very useful for screening MOFs using a more realistic simulation approach. • Experimental investigation of multicomponent gas adsorption in MOFs is not straightforward as discussed previously. On the other hand, most of the molecular simulation studies have focused on either single-component CO2 uptake or binary gas mixtures including CO2. However, the presence of water and other gas-phase components such as O2, CO, SOx, and NOx in a CO2 capture process may significantly affect the predicted CO2 capture performances of MOFs. Therefore, considering the real CO2 capture process, multicomponent gas adsorption in MOFs, is an important future direction for MOF simulations. • In most high-throughput screening studies, hypothetical MOFs are used to identify the characteristics of the promising adsorbents for CO2 capture. Some hypothetical MOFs may not even be synthesized because of the toxic chemicals required for synthesis, lack of appropriate synthesis conditions, and/or stability problems under humid conditions. Recent studies on linking computational screening of hypothetical MOFs and experimental synthesis of one of the promising materials have been encouraging, and future studies in this direction will accelerate the data-driven design and development of new MOFs. • Another direction in the field is to combine quantum-level calculations, molecular simulations, and machine learning algorithms to estimate the detailed process-
232
C. Altintas et al.
level performance parameters of MOFs such as purity, recovery, productivity, and energy cost for CO2 capture. Among these, energy cost is of paramount importance since the future of MOFs in industrial applications is strongly dependent on a detailed cost analysis of the process. Finally, it is important to note that all the future directions that we briefly mentioned above need contributions from scientists with various backgrounds including chemistry, material science, data science, and chemical/process/energy engineering. We believe that the large-scale molecular simulations of MOFs will continue to provide a valuable source of data to accelerate the design and development of MOFs for CO2 capture.
References 1. Pardakhti M, Jafari T, Tobin Z et al (2019) Trends in solid adsorbent materials development for CO2 capture. ACS Appl Mater Interfaces 11:34533–34559. https://doi.org/10.1021/ acsami.9b08487 2. Yu CH, Huang CH, Tan CS (2012) A review of CO2 capture by absorption and adsorption. Aerosol Air Qual Res 12:745–769. https://doi.org/10.4209/aaqr.2012.05.0132 3. Zou L, Sun Y, Che S et al (2017) Porous organic polymers for post-combustion carbon capture. Adv Mater 29:1–35. https://doi.org/10.1002/adma.201700229 4. Wang WJ, Zhouab M, Yuan DQ (2017) Carbon dioxide capture in amorphous porous organic polymers. J Mater Chem A 5:1334–1347. https://doi.org/10.1039/c6ta09234a 5. Azmi AA, Aziz MAA (2019) Mesoporous adsorbent for CO2 capture application under mild condition: a review. J Environ Chem Eng 7:1–13. https://doi.org/10.1016/j.jece.2019.103022 6. Hu ZG, Wang YX, Shah BB et al (2019) CO2 capture in metal-organic framework adsorbents: an engineering perspective. Adv Sustain Syst 3:1–21. https://doi.org/10.1002/adsu.201800080 7. Zhao RK, Deng S, Wang SP et al (2018) Thermodynamic research of adsorbent materials on energy efficiency of vacuum-pressure swing adsorption cycle for CO2 capture. Appl Therm Eng 128:818–829. https://doi.org/10.1016/j.applthermaleng.2017.09.074 8. Furukawa S, Reboul J, Diring S et al (2014) Structuring of metal-organic frameworks at the mesoscopic/macroscopic scale. Chem Soc Rev 43:5700–5734. https://doi.org/10.1039/ c4cs00106k 9. Li H, Li L, Lin R-B et al (2019) Porous metal-organic frameworks for gas storage and separation: status and challenges. EnergyChem 1:1–39. https://doi.org/10.1016/j.enchem. 2019.100006 10. Silva P, Vilela SM, Tome JP et al (2015) Multifunctional metal-organic frameworks: from academia to industrial applications. Chem Soc Rev 44:6774–6803. https://doi.org/10.1039/ c5cs00307e 11. Sinha P, Datar A, Jeong C et al (2019) Surface area determination of porous materials using the Brunauer–Emmett–Teller (BET) method: limitations and improvements. J Phys Chem C 123:20195–20209. https://doi.org/10.1021/acs.jpcc.9b02116 12. Zhou HC, Long JR, Yaghi OM (2012) Introduction to metal-organic frameworks. Chem Rev 112:673–674. https://doi.org/10.1021/cr300014x 13. Eddaoudi M, Kim J, Rosi N et al (2002) Systematic design of pore size and functionality in isoreticular MOFs and their application in methane storage. Science 295:469–472. https://doi. org/10.1126/science.1067208
6 Computational Screening of MOFs for CO2 Capture
233
14. Li H, Eddaoudi M, O'Keeffe M et al (1999) Design and synthesis of an exceptionally stable and highly porous metal-organic framework. Nature 402:276–279. https://doi.org/10.1038/ 46248 15. Keskin S, van Heest TM, Sholl DS (2010) Can metal–organic framework materials play a useful role in large-scale carbon dioxide separations? ChemSusChem 3:879–891. https://doi. org/10.1002/cssc.201000114 16. Cmarik GE, Kim M, Cohen SM et al (2012) Tuning the adsorption properties of UiO-66 via ligand functionalization. Langmuir 28:15606–15613. https://doi.org/10.1021/la3035352 17. Nugent P, Belmabkhout Y, Burd SD et al (2013) Porous materials with optimal adsorption thermodynamics and kinetics for CO2 separation. Nature 495:80–84. https://doi.org/10.1038/ nature11893 18. Lu WG, Yuan DQ, Sculley JL et al (2011) Sulfonate-grafted porous polymer networks for preferential CO2 adsorption at low pressure. J Am Chem Soc 133:18126–18129. https://doi. org/10.1021/ja2087773 19. Cavenati S, Grande CA, Rodrigues AE (2004) Adsorption equilibrium of methane, carbon dioxide, and nitrogen on zeolite 13X at high pressures. J Chem Eng Data 49:1095–1101. https://doi.org/10.1021/je0498917 20. Sumida K, Rogow DL, Mason JA et al (2012) Carbon dioxide capture in metal-organic frameworks. Chem Rev 112:724–781. https://doi.org/10.1021/cr2003272 21. Liu J, Thallapally PK, McGrail BP et al (2012) Progress in adsorption-based CO2 capture by metal–organic frameworks. Chem Soc Rev 41:2308–2322. https://doi.org/10.1039/ c1cs15221a 22. Sabouni R, Kazemian H, Rohani S (2014) Carbon dioxide capturing technologies: a review focusing on metal organic framework materials (MOFs). Environ Sci Pollut Res Int 21:5427–5449. https://doi.org/10.1007/s11356-013-2406-2 23. Li B, Wen H-M, Zhou W et al (2014a) Porous metal–organic frameworks for gas storage and separation: what, how, and why? J Phys Chem Lett 5:3468–3479. https://doi.org/10.1016/j. enchem.2019.100006 24. Belmabkhout Y, Guillerm V, Eddaoudi M (2016) Low concentration CO2 capture using physical adsorbents: are metal–organic frameworks becoming the new benchmark materials? Chem Eng J 296:386–397. https://doi.org/10.1016/j.cej.2016.03.124 25. Yu JM, Xie LH, Li JR et al (2017) CO2 capture and separations using MOFs: computational and experimental studies. Chem Rev 117:9674–9754. https://doi.org/10.1021/acs.chemrev. 6b00626 26. Li H, Wang KC, Sun YJ et al (2018) Recent advances in gas storage and separation using metal-organic frameworks. Mater Today 21:108–121. https://doi.org/10.1016/j.mattod.2017. 07.006 27. Lin YC, Kong CL, Zhang QJ et al (2017) Metal-organic frameworks for carbon dioxide capture and methane storage. Adv Energy Mater 7:1–29. https://doi.org/10.1002/aenm. 201601296 28. Bae YS, Snurr RQ (2011) Development and evaluation of porous materials for carbon dioxide separation and capture. Angew Chem Int Ed Eng 50:11586–11596. https://doi.org/10.1002/ anie.201101891 29. Allen FH (2002) The Cambridge structural database: a quarter of a million crystal structures and rising. Acta Crystallogr Sect B: Struct Sci Cryst Eng Mater 58:380–388. https://doi.org/10. 1107/s0108768102003890 30. Willems TF, Rycroft C, Kazi M et al (2012) Algorithms and tools for high-throughput geometry-based analysis of crystalline porous materials. Microporous Mesoporous Mater 149:134–141. https://doi.org/10.1016/j.micromeso.2011.08.020 31. Sarkisov L, Harrison A (2011) Computational structure characterisation tools in application to ordered and disordered porous materials. Mol Simul 37:1248–1257. https://doi.org/10.1080/ 08927022.2011.592832
234
C. Altintas et al.
32. Dubbeldam D, Calero S, Ellis DE et al (2016) RASPA: molecular simulation software for adsorption and diffusion in flexible nanoporous materials. Mol Simul 42:81–101. https://doi. org/10.1080/08927022.2015.1010082 33. Ongari D, Boyd PG, Barthel S et al (2017) Accurate characterization of the pore volume in microporous crystalline materials. Langmuir 33:14529–14538. https://doi.org/10.1021/acs. langmuir.7b01682 34. Barrett EP, Joyner LG, Halenda PP (1951) The determination of pore volume and area distributions in porous substances. I. Computations from nitrogen isotherms. J Am Chem Soc 73:373–380. https://doi.org/10.1021/ja01145a126 35. Düren T, Millange F, Férey G et al (2007) Calculating geometric surface areas as a characterization tool for metalorganic frameworks. J Phys Chem C 111:15350–15356. https://doi. org/10.1021/jp074723h 36. Gelb LD, Gubbins KE (1999) Pore size distributions in porous glasses: a computer simulation study. Langmuir 15:305–308. https://doi.org/10.1021/la9808418 37. Gupta A, Chempath S, Sanborn MJ et al (2003) Object-oriented programming paradigms for molecular modeling. Mol Simul 29:29–46. https://doi.org/10.1080/0892702031000065719 38. Akkermans RL, Spenley NA, Robertson SH (2013) Monte Carlo methods in materials studio. Mol Simul 39:1153–1164. https://doi.org/10.1080/08927022.2013.843775 39. Evans JD, Garai B, Reinsch H et al (2019) Metal–organic frameworks in Germany: from synthesis to function. Coord Chem Rev 380:378–418. https://doi.org/10.1016/j.ccr.2018.10. 002 40. Lee Y-R, Kim J, Ahn W-S (2013) Synthesis of metal-organic frameworks: a mini review. Korean J Chem Eng 30:1667–1680. https://doi.org/10.1007/s11814-013-0140-6 41. Stock N, Biswas S (2011) Synthesis of metal-organic frameworks (MOFs): routes to various mof topologies, morphologies, and composites. Chem Rev 112:933–969. https://doi.org/10. 1021/cr200304e 42. Altintas C, Avci G, Daglar H et al (2019) An extensive comparative analysis of two MOF databases: high-throughput screening of computation-ready MOFs for CH4 and H2 adsorption. J Mater Chem A 7:9593–9608. https://doi.org/10.1039/c9ta01378d 43. Nazarian D, Camp JS, Chung YG et al (2017) Large-scale refinement of metal-organic framework structures using density functional theory. Chem Mater 29:2521–2528. https:// doi.org/10.1021/acs.chemmater.6b04226 44. Chung YG, Camp J, Haranczyk M et al (2014) Computation-ready, experimental metalorganic frameworks: a tool to enable high-throughput screening of nanoporous crystals. Chem Mater 26:6185–6192. https://doi.org/10.1021/cm502594j 45. Chung YG, Haldoupis E, Bucior BJ et al (2019) Advances, updates, and analytics for the computation-ready, experimental metal–organic framework database: CoRE MOF 2019. J Chem Eng Data 64:5985–5998. https://doi.org/10.1021/acs.jced.9b00835 46. Moghadam PZ, Li A, Wiggin SB et al (2017) Development of a Cambridge structural database subset: a collection of metal-organic frameworks for past, present, and future. Chem Mater 29:2618–2625. https://doi.org/10.1021/acs.chemmater.7b00441 47. Colón YJ, Gómez-Gualdrón DA, Snurr RQ (2017) Topologically guided, automated construction of metal–organic frameworks and their evaluation for energy-related applications. Cryst Growth Des 17:5801–5810. https://doi.org/10.1021/acs.cgd.7b00848 48. Wilmer CE, Leaf M, Lee CY et al (2011) Large-scale screening of hypothetical metal-organic frameworks. Nat Chem 4:83–89. https://doi.org/10.1038/nchem.1192 49. Yang L, Shi C, Li L et al (2019) High-throughput model-building and screening of zeolitic imidazolate frameworks for CO2 capture from flue gas. Chin Chem Lett 31:227–230. https:// doi.org/10.1016/j.cclet.2019.04.025 50. Frenkel D, Smit B (2002) Understanding molecular simulation: from algorithms to applications, vol 1. Elsevier (formerly published by Academic Press)
6 Computational Screening of MOFs for CO2 Capture
235
51. Harris JG, Yung KH (1995) Carbon dioxide’s liquid-vapor coexistence curve and critical properties as predicted by a simple molecular model. J Phys Chem 99:12021–12024. https:// doi.org/10.1021/j100031a034 52. Potoff JJ, Siepmann JI (2001) Vapor-liquid equilibria of mixtures containing alkanes, carbon dioxide, and nitrogen. AICHE J 47:1676–1682. https://doi.org/10.1002/aic.690470719 53. Dubbeldam D, Walton KS, Vlugt TJH et al (2019) Design, parameterization, and implementation of atomic force fields for adsorption in nanoporous materials. Adv Theor Simul 2:1–62. https://doi.org/10.1002/adts.201900135 54. Sturluson A, Huynh MT, Kaija AR et al (2019) The role of molecular modelling and simulation in the discovery and deployment of metal-organic frameworks for gas storage and separation. Mol Simul 45:1082–1121. https://doi.org/10.1080/08927022.2019.1648809 55. Mayo SL, Olafson BD, Goddard WA (1990) Dreiding – a generic force-field for molecular simulations. J Phys Chem 94:8897–8909. https://doi.org/10.1021/j100389a010 56. Rappe AK, Casewit CJ, Colwell KS et al (1992) UFF, a full periodic-table force-field for molecular mechanics and molecular-dynamics simulations. J Am Chem Soc 114:10024–10035. https://doi.org/10.1021/ja00051a040 57. Ongari D, Boyd PG, Kadioglu O et al (2019a) Evaluating charge equilibration methods to generate electrostatic fields in nanoporous materials. J Chem Theory Comput 15:382–401. https://doi.org/10.1021/acs.jctc.8b00669 58. Nazarian D, Camp JS, Sholl DS (2016) A comprehensive set of high-quality point charges for simulations of metal-organic frameworks. Chem Mater 28:785–793. https://doi.org/10.1021/ acs.chemmater.5b03836 59. Manz TA, Sholl DS (2010) Chemically meaningful atomic charges that reproduce the electrostatic potential in periodic and nonperiodic materials. J Chem Theory Comput 6:2455–2468. https://doi.org/10.1021/ct100125x 60. Bahr DF, Reid JA, Mook WM et al (2007) Mechanical properties of cubic zinc carboxylate IRMOF-1 metal-organic framework crystals. Phys Rev B 76:1–7. https://doi.org/10.1103/ PhysRevB.76.184106 61. Chang Z, Yang DH, Xu J et al (2015) Flexible metal-organic frameworks: recent advances and potential applications. Adv Mater 27:5432–5441. https://doi.org/10.1002/adma.201501523 62. Coudert FX (2015) Responsive metal-organic frameworks and framework materials: under pressure, taking the heat, in the spotlight, with friends. Chem Mater 27:1905–1916. https://doi. org/10.1021/acs.chemmater.5b00046 63. Li W, Henke S, Cheetham AK (2014b) Research update: mechanical properties of metalorganic frameworks – influence of structure and chemical bonding. Appl Mater 2:1–10. https:// doi.org/10.1063/1.4904966 64. Moghadam PZ, Rogge SMJ, Li A et al (2019) Structure-mechanical stability relations of metal-organic frameworks via machine learning. Matter 1:219–234. https://doi.org/10.1016/j. matt.2019.03.002 65. Evans DJ, Holian BL (1985) The Nosé–Hoover thermostat. J Chem Phys 83:4069–4074. https://doi.org/10.1063/1.449071 66. Andersen HC (1980) Molecular dynamics simulations at constant pressure and/or temperature. J Chem Phys 72:2384–2393. https://doi.org/10.1063/1.439486 67. Widom B (1963) Some topics in the theory of fluids. J Chem Phys 39:2808–2812. https://doi. org/10.1063/1.1734110 68. Rege SU, Yang RT (2001) A simple parameter for selecting an adsorbent for gas separation by pressure swing adsorption. Sep Sci Technol 36:3355–3365. https://doi.org/10.1081/Ss100107907 69. Chung YG, Gomez-Gualdron DA, Li P et al (2016) In silico discovery of metal-organic frameworks for precombustion CO2 capture using a genetic algorithm. Sci Adv 2:1–9. https:// doi.org/10.1126/sciadv.1600909
236
C. Altintas et al.
70. Trickett CA, Helal A, Al-Maythalony BA et al (2017) The chemistry of metal–organic frameworks for CO2 capture, regeneration and conversion. Nat Rev Mater 2:1–16. https:// doi.org/10.1038/natrevmats.2017.45 71. Cui X, Bustin RM, Dipple G (2004) Selective transport of CO2, CH4, and N2 in coals: insights from modeling of experimental gas adsorption data. Fuel 83:293–303. https://doi.org/10.1016/ j.fuel.2003.09.001 72. García EJ, Pérez-Pellitero J, Pirngruber GD et al (2017) Sketching a portrait of the optimal adsorbent for CO2 separation by pressure swing adsorption. Ind Eng Chem Res 56:4818–4829. https://doi.org/10.1021/acs.iecr.6b04877 73. Walton KS (2019) 110th anniversary: commentary: perspectives on adsorption of complex mixtures. Ind Eng Chem Res 58:17100–17105. https://doi.org/10.1021/acs.iecr.9b04243 74. Myers AL, Prausnitz JM (1965) Thermodynamics of mixed-gas adsorption. AICHE J 11:121–127. https://doi.org/10.1002/aic.690110125 75. Walton KS, Sholl DS (2015) Predicting multicomponent adsorption: 50 years of the ideal adsorbed solution theory. AICHE J 61:2757–2762. https://doi.org/10.1002/aic.14878 76. García EJ, Pérez-Pellitero J, Pirngruber GD et al (2014) Tuning the adsorption properties of zeolites as adsorbents for CO2 separation: best compromise between the working capacity and selectivity. Ind Eng Chem Res 53:9860–9874. https://doi.org/10.1021/ie500207s 77. Krishna R (2018) Methodologies for screening and selection of crystalline microporous materials in mixture separations. Sep Purif Technol 194:281–300. https://doi.org/10.1016/j. seppur.2017.11.056 78. Maring BJ, Webley PA (2013) A new simplified pressure/vacuum swing adsorption model for rapid adsorbent screening for CO2 capture applications. Int J Greenhouse Gas Control 15:16–31. https://doi.org/10.1016/j.ijggc.2013.01.009 79. Wiersum AD, Chang JS, Serre C et al (2013) An adsorbent performance indicator as a first step evaluation of novel sorbents for gas separations: application to metal-organic frameworks. Langmuir 29:3301–3309. https://doi.org/10.1021/la3044329 80. Huck JM, Lin LC, Berger AH et al (2014) Evaluating different classes of porous materials for carbon capture. Energy Environ Sci 7:4132–4146. https://doi.org/10.1039/c4ee02636e 81. Lin LC, Berger AH, Martin RL et al (2012) In silico screening of carbon-capture materials. Nat Mater 11:633–641. https://doi.org/10.1038/nmat3336 82. Ongari D, Yakutovich AV, Talirz L et al (2019b) Building a consistent and reproducible database for adsorption evaluation in covalent-organic frameworks. ACS Cent Sci 5:1663–1675. https://doi.org/10.1021/acscentsci.9b00619 83. Park J, Rubiera Landa HO, Kawajiri Y et al (2019) How well do approximate models of adsorption-based CO2 capture processes predict results of detailed process models? Ind Eng Chem Res 2019:1–12. https://doi.org/10.1021/acs.iecr.9b05363 84. First EL, Gounaris CE, Floudas CA (2013) Predictive framework for shape-selective separations in three-dimensional zeolites and metal–organic frameworks. Langmuir 29:5599–5608. https://doi.org/10.1021/la400547a 85. Watanabe T, Sholl DS (2012) Accelerating applications of metal-organic frameworks for gas adsorption and separation by computational screening of materials. Langmuir 28:14114–14128. https://doi.org/10.1021/la301915s 86. Altintas C, Avci G, Daglar H et al (2018) Database for CO2 separation performances of MOFs based on computational materials screening. ACS Appl Mater Interfaces 10:17257–17268. https://doi.org/10.1021/acsami.8b04600 87. Li S, Chung YG, Snurr RQ (2016) High-throughput screening of metal-organic frameworks for CO2 capture in the presence of water. Langmuir 32:10368–10376. https://doi.org/10.1021/ acs.langmuir.6b02803 88. Li W, Rao ZZ, Chung YG et al (2017) The role of partial atomic charge assignment methods on the computational screening of metal-organic frameworks for CO2 capture under humid conditions. Chemistry 2:9458–9465. https://doi.org/10.1002/slct.201701934
6 Computational Screening of MOFs for CO2 Capture
237
89. Qiao Z, Xu Q, Jiang J (2018) Computational screening of hydrophobic metal–organic frameworks for the separation of H2S and CO2 from natural gas. J Mater Chem A 6:18898–18905. https://doi.org/10.1039/c8ta04939d 90. Wu D, Yang Q, Zhong C et al (2012) Revealing the structure-property relationships of metalorganic frameworks for CO2 capture from flue gas. Langmuir 28:12094–12049. https://doi. org/10.1021/la302223m 91. Ockwig NW, Delgado-Friedrichs O, O'Keeffe M et al (2005) Reticular chemistry: occurrence and taxonomy of nets and grammar for the design of frameworks. Acc Chem Res 38:176–182. https://doi.org/10.1021/ar020022l 92. Haldoupis E, Nair S, Sholl DS (2012) Finding MOFs for highly selective CO2/N2 adsorption using materials screening based on efficient assignment of atomic point charges. J Am Chem Soc 134:4313–4323. https://doi.org/10.1021/ja2108239 93. Wilmer CE, Farha OK, Bae YS et al (2012) Structure-property relationships of porous materials for carbon dioxide separation and capture. Energy Environ Sci 5:9849–9856. https://doi.org/10.1039/c2ee23201d 94. Fernandez M, Trefiak NR, Woo TK (2013a) Atomic property weighted radial distribution functions descriptors of metal-organic frameworks for the prediction of gas uptake capacity. J Phys Chem C 117:14095–14105. https://doi.org/10.1021/jp404287t 95. Fernandez M, Boyd PG, Daff TD et al (2014) Rapid and accurate machine learning recognition of high performing metal organic frameworks for CO2 capture. J Phys Chem Lett 5:3056–3060. https://doi.org/10.1021/jz501331m 96. McDaniel JG, Li S, Tylianakis E et al (2015) Evaluation of force field performance for highthroughput screening of gas uptake in metal–organic frameworks. J Phys Chem C 119:3143–3152. https://doi.org/10.1021/jp511674w 97. Qiao ZW, Zhang K, Jiang JW (2016) In silico screening of 4764 computation-ready, experimental metal-organic frameworks for CO2 separation. J Mater Chem A 4:2105–2114. https:// doi.org/10.1039/c5ta08984k 98. Sumer Z, Keskin S (2016) Ranking of MOF adsorbents for CO2 separations: a molecular simulation study. Ind Eng Chem Res 55:10404–10419. https://doi.org/10.1021/acs.iecr. 6b02585 99. Aghaji MZ, Fernandez M, Boyd PG et al (2016) Quantitative structure-property relationship models for recognizing metal organic frameworks (MOFs) with high CO2 working capacity and CO2/CH4 selectivity for methane purification. Eur J Inorg Chem 2016:4505–4511. https:// doi.org/10.1002/ejic.201600365 100. Fernandez M, Barnard AS (2016) Geometrical properties can predict CO2 and N2 adsorption performance of metal-organic frameworks (MOFs) at low pressure. ACS Comb Sci 18:243–252. https://doi.org/10.1021/acscombsci.5b00188 101. Collins SP, Daff TD, Piotrkowski SS et al (2016) Materials design by evolutionary optimization of functional groups in metal-organic frameworks. Sci Adv 2:1–6. https://doi.org/10. 1126/sciadv.1600954 102. Park J, Lively RP, Sholl DS (2017) Establishing upper bounds on CO2 swing capacity in sub-ambient pressure swing adsorption via molecular simulation of metal-organic frameworks. J Mater Chem A 5:12258–12265. https://doi.org/10.1039/c7ta02916k 103. Collins SP, Woo TK (2017) Split-charge equilibration parameters for generating rapid partial atomic charges in metal–organic frameworks and porous polymer networks for highthroughput screening. J Phys Chem C 121:903–910. https://doi.org/10.1021/acs.jpcc.6b10804 104. Anderson R, Rodgers J, Argueta E et al (2018) Role of pore chemistry and topology in the CO2 capture capabilities of MOFs: from molecular simulation to machine learning. Chem Mater 30:6325–6337. https://doi.org/10.1021/acs.chemmater.8b02257 105. Argueta E, Shaji J, Gopalan A et al (2018) Molecular building block-based electronic charges for high-throughput screening of metal–organic frameworks for adsorption applications. J Chem Theory Comput 14:365–376. https://doi.org/10.1021/acs.jctc.7b00841
238
C. Altintas et al.
106. Avci G, Velioglu S, Keskin S (2018) High-throughput screening of MOF adsorbents and membranes for H2 purification and CO2 capture. ACS Appl Mater Interfaces 10:33693–33706. https://doi.org/10.1021/acsami.8b12746 107. Dokur D, Keskin S (2018) Effects of force field selection on the computational ranking of MOFs for CO2 separations. Ind Eng Chem Res 57:2298–2309. https://doi.org/10.1021/acs. iecr.7b04792 108. Boyd PG, Chidambaram A, Garcia-Diez E et al (2019) Data-driven design of metal-organic frameworks for wet flue gas CO2 capture. Nature 576:253–256. https://doi.org/10.1038/ s41586-019-1798-7 109. Agrawal M, Sholl DS (2019) Effects of intrinsic flexibility on adsorption properties of metal– organic frameworks at dilute and nondilute loadings. ACS Appl Mater Interfaces 11:31060–31068. https://doi.org/10.1021/acsami.9b10622 110. Dureckova H, Krykunov M, Aghaji MZ et al (2019) Robust machine learning models for predicting high CO2 working capacity and CO2/H2 selectivity of gas adsorption in metal organic frameworks for precombustion carbon capture. J Phys Chem C 123:4133–4139. https://doi.org/10.1021/acs.jpcc.8b10644 111. Demir H, Cramer CJ, Siepmann JI (2019) Computational screening of metal–organic frameworks for biogas purification. Mol Syst Des Eng 4:1125–1135. https://doi.org/10.1039/ c9me00095j 112. Keskin S (2019) Screening for selectivity. Nat Energy:8–9. https://doi.org/10.1038/s41560019-0514-z 113. Wang H, Ji Y, Li Y (2019) Simulation and design of energy materials accelerated by machine learning. Wiley Interdiscip Rev Comput Mol Sci:1–18. https://doi.org/10.1002/wcms.1421 114. Fernandez M, Woo TK, Wilmer CE et al (2013b) Large-scale quantitative structure-property relationship (QSPR) analysis of methane storage in metal-organic frameworks. J Phys Chem C 117:7681–7689. https://doi.org/10.1021/jp4006422
Chapter 7
Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks (MOFs) Sanjit Nayak
7.1 7.1.1
Heavy Metals Sources of Heavy Metals in Water
It has been estimated that the availability of clean water will be one of the major challenges in the coming decades and will have a wide global impact on food production, maintaining the ecosystem, and supply of drinking water [1]. The scale of the problem demands an urgent development of technologies for wastewater treatment and the removal of hazardous pollutants effectively [2–6]. Heavy metal pollution is ubiquitous and is of greatest concern in developing and developed countries alike [7–16]. The term heavy metals is often used to refer the metals with a specific density over 5 g/cm3 [17]. However, with much debate and no clear definition, the term “heavy metal” has been used widely in literature to indicate a group of metals and semimetals that can cause potential toxicity or ecotoxicity even at very low concentrations [18, 19]. The release of heavy metals in the environment occurs due to either natural or anthropogenic actions. Among natural activities, geological erosion of rocks contributes largely to the release of heavy metals which are then carried away by the groundwater and adsorbed in clay minerals and on iron and manganese oxyhydroxides. The aquatic species also absorb the heavy metals in water and leads to bioaccumulation which further leads to the entrance of the heavy metals into the food web (Scheme 7.1) [20]. Heavy metals are also released into the environment due to several anthropogenic actions, such as mining, agriculture, production of nuclear waste, solid waste incineration, and so on. Heavy metals are released in the atmosphere by different mechanisms, including
S. Nayak (*) Assistant Professor, School of Chemistry and Biosciences, University of Bradford, Richmond Rd, Bradford BD7 1DP, UK e-mail: [email protected] © Springer Nature Switzerland AG 2021 P. Horcajada Cortés, S. Rojas Macías (eds.), Metal-Organic Frameworks in Biomedical and Environmental Field, https://doi.org/10.1007/978-3-030-63380-6_7
239
240
S. Nayak
Scheme 7.1 A flowchart showing the natural and anthropogenic origin of heavy metals, their accumulation in the aquatic life, and introduction to the food web. (Reproduced with permission from [24])
mineral dusts, sea salt particles, extraterrestrial matter, volcanic ashes, forest fire ashes, emissions from aviation and transport, coal combustion, and so on [21]. There is a clear correlation between anthropogenic actions such as mining with a dramatic increase in the emission of heavy metals into the atmosphere [22]. For example, the plot in Fig. 7.1 shows a dramatic increase in heavy metal emission in the twentieth century as a consequence of the industrial revolution and modernization [23].
7.1.2
Effects on Health
The presence of heavy metals on the outer surface of the Earth, and particularly in air and water, is a major concern due to their impact on the health of humans and animals. Accumulation of heavy metals in the cells can cause long-term damages leading to different diseases, including cancer [25]. It is to be noted that there are several metals that fit in the definition of heavy metal, but not all of them are toxic. Some of the heavy metals are required at certain concentrations for maintaining the normal activities in the human body. For example, manganese, iron, nickel, copper, or zinc are all essential elements for a healthy human body as they play key important roles in the active center of proteins and enzymes to carry out important redox processes [26] involving electron transfer or to maintain certain structural
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . .
241
Fig. 7.1 A graph showing a steep rise in global production and emission of heavy metals in the twentieth century. The graph was plotted using the data in reference (23)
features [27] of these proteins and enzymes through specific geometry of the coordination bonds around these metal ions. However, the presence of these essential metals in excess or the other heavy metals can potentially form unintended complexes with biological molecules containing ligands rich in oxygen, nitrogen, and sulfur. These unintended changes in chemical bonds lead to changes in protein structures and inhibition of enzyme activities. This disruptive mechanism leads to toxicity affecting the nervous system (Hg, Pb, As), kidneys and liver (Cu, Cd, Hg, Pb), and skin, bones, and teeth (Ni, Cu, Cd, Cr), and some of these can lead to cancer [28]. Based on the exposure and toxicity, three heavy metals (mercury, lead, and cadmium) are of major concern and widely studied. The adverse health effects of these three heavy metals are summarized below. Mercury Mercury has been in use since the prehistoric time when people used red cinnabar (HgS) for painting on the cave walls [13]. Mercury had popular medicinal usage in ancient Greece and was used as a cosmetic as well. At present mercury amalgam is still widely used as dental filling material, in thermometer and barometer, and as electrodes. Exposure to mercury in general occurs in two forms, inorganic mercury and methyl mercury. Dentists have suffered from large exposure to inorganic mercury due to the use of mercury amalgam filling. However, the wider
242
S. Nayak
population is more exposed to methyl mercury that accumulates in the food chain, particularly in certain kinds of fishes [29], such as mackerel, swordfish, and tuna. Both forms of mercury have severe health effects. Inorganic mercury causes lung damage, neurological and psychological disorders, anxiety, restlessness, and depression. Metallic mercury does not affect the central nervous system due to the bloodbrain barrier. Methyl mercury is highly toxic to humans, and it affects the nervous system. Early symptoms of organic mercury poisoning include numbness of hands and feet, leading to coordination difficulties, and visual impairment at a later stage. High doses can also lead to death. Lead Human exposures to lead occur both from air and food [13]. Airborne lead emission originates from the combustion of leaded petrol in automobiles, mines, smelters, and welding of lead-painted metals. However, the lead emission from automobiles has reduced significantly since many countries introduced a ban on leaded petrol. Lead enters the food chain through the deposition of airborne lead in soil and water and from water pipes [30]. Lead was widely used in water pipes in the past and is still in use in many old water supply networks. Lead accumulates in the lung, blood, and bone and has a very long half-life ranging up to 20–30 years. The toxicity of lead results in restlessness, headache, abdominal pain, learning and concentration difficulties, and renal tubular damage. Inorganic lead does not cross the blood-brain barrier in adults, but children are prone to it, and it may lead to permanent brain damage [31]. Organic lead compounds (tetramethyl lead and tetraethyl lead) are more toxic and can cross the blood-brain barrier for adults too [13]. Lead has been also identified as a potential carcinogen [32], specially for lung cancer, stomach cancer, and gliomas. Cadmium Cadmium is used in pigment, phosphate fertilizer, PVC-stabilizer, anticorrosion agent, and Ni-Cd batteries. Cadmium is released in the environment by poor recycling of Cd-containing products, which are disposed with household products. Natural and anthropogenic release of cadmium accumulates into water and soil and enters the food chain through the uptake of vegetables and crops. Smoking is also one of the sources of cadmium intake. However, the majority of the cadmium intake in the general population occurs through consumption of food that contains cadmium. Exposure to cadmium can cause severe kidney damage and renal failure [33]. Long-term cadmium exposure can also lead to itai-itai disease which occurs due to poor bone conditions. Cadmium has also been identified as a potential human carcinogen, causing lung cancer, kidney cancer, and prostate cancer [34].
7.2
MOFs for Removal of Heavy Metals from Water
For the removal of heavy metals from water, there are several existing technologies in use currently, such as chemical precipitation (as hydroxides, sulfides, or by chelate complexes), ion exchange, coagulation-flocculation, floatation, membrane filtration,
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . .
243
Table 7.1 The maximum allowed concentration of Cd, Hg, and Pb in drinking water as recommended by the World Health Organization (WHO) and their adverse effects on human health Heavy metal Cd
Pb
Hg
Provisional tolerable weekly/monthly intake (PTWI/PTMI) PTMI: 25 μg/kg body weight
Previously established PTWI (0.025 μg/kg body weight) withdrawn due to no possible establishment of a new PTWI that would be considered safe PTWI: 4 μg/kg body weight
Common sources Electronic waste, nickelcadmium batteries, anticorrosive metal coatings, alloys, plastic stabilizers, coal combustion, pigments Mining, smelting, battery manufacturing, use of leaded petrol, and food contamination through packaging and handling
Health effects Renal tubular dysfunction
Fish, seafood, amalgamation, catalysts, dental fillings, Hg vapor lamps, solders, X-ray tubes, pharmaceuticals, fungicides
Weight changes, kidney damage, and progressive nephropathy
Neurobehavioral development of infants and children and systolic blood pressure for adults
and adsorption [35]. Due to recyclability, cost-effectiveness, and generation of fewer by-products/sludge, adsorption has been recognized as a very effective process for wastewater treatment and removal of heavy metals. A large number of adsorbents have been studied for the separation and removal of heavy metals, including activated carbon, carbon nanotubes, fly ash, and zeolites [36]. Apart from these well-studied materials, a relatively new class of materials, metal-organic frameworks or MOFs, has caught great attention recently with their high efficiency and selectivity for the removal of heavy metals from water. This is particularly an advantage over the other materials in use, particularly due to their ultrahigh surface area and functional properties that can be engineered by synthetic modifications. Due to the extreme global concern of water contamination, the past 10 years has seen a rapid growth in research on the use of MOFs for the efficient removal of heavy metals from water. Based on relevance and toxicity, the removal of three metals, namely, mercury, lead, and cadmium, has been studied widely. Table 7.1 shows the maximum allowed concentration of these three heavy metals in drinking water (as recommended by the World Health Organization (WHO)) and their adverse effects on human health. Toxicities of some other metals (Fe, Co, Ni, Cu, Zn, Cr, and Ag) are also known but less studied due to their relevance and occurrence, and not covered in this chapter. A summary of the studied MOFs with their reported efficiency is given in Table 7.2. In the following sections, we will discuss the adsorption and removal of three heavy metals (mercury, lead, and cadmium) by different MOFs as they have been studied in the past few years.
Heavy metal Hg(II) Uptake: 714 mg/g, equilibrium in 120 min 40–60% adsorption
Soft-soft interaction
Interaction between NH and metal ion Interaction between anionic MOF and cationic heavy metals Soft-soft interaction
Soft-soft interaction Soft-soft interaction
[Co2(TATAB)(OH) (H2O)2] [(NH2Me2)2][Zn3(L)2] (AMOF-10 ) [Ni(3-bpd)2(NCS)2]n
UiO-66-NHC(S)NHMe [CaIICuII6[(S,S) methox]3(OH)2(H2O)] 16H2O Thiol-functionalized Fe3O4@HKUST-1 magnetic microspheres MOF-A/S (hybrid mixedlinker MOF) MOF-808-EDTA MIL-100(Fe)/PDA (MOF-polymer composite) TMU-40 348.43 mg/g RE > 90% 78.8 mg/g 592 mg/g 1634 mg/g
269 mg/g
Soft-soft interaction
Soft-soft interaction
Chelation via EDTA Interaction with catechol groups of PDA
Interaction with open nitrogen sites
Uptake: 78 mg/g; RE ¼ 98.7% in 24 h 713 mg/g; up to 94.30% removal efficiency (RE) 769 mg/g 900 mg/g (for HgCl2); 166 mg/g (for CH3HgCl)
Uptake/removal efficiency (RE) Up to 94% removal
Mechanism Soft-soft interaction
MOF Thioether modified MOF-5 Thiol-HKUST-1
– 7
– 30 to 120
10
120 99% 98.18 mg/g
Lewis acid-base interaction Soft-soft interaction
TMU-46S Few-layered CoCNSP nanosheets PCN-221 HKUST-1-Fe3O4-DHz magnetic composite HKUST-1MW@H3PW12O40 MIL-53-Fe3O4 magnetic nanocomposite [(Fe3O4– ethylenediamine)/MIL101(Fe)] [(NH2Me2)2][Zn3(L)2] (AMOF-10 ) [Dy(BTC)](H2O) (DMF)1.1 UiO-66-NHC(S)NHMe HS-mSi@MOF-5
171.5 mg/g Distribution coefficient of 1.99 x 107 mL/g 714 mg/g 716 mg/g
Lewis acid-base interaction Soft-soft interaction
Soft-soft interaction
Metal-sulfide bonding
Thiol-modified UiO-66
Sx-MOFs
[49]
(continued)
[61]
[44]
[60]
[42] [59]
[58]
[40]
[57]
[56]
[55]
[53] [54]
[51] [52]
[50]
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . . 245
Heavy metal
MIL-88B(Fe)–Ag/TiO2 nanotubes/Ti plates MIL-101(Fe)/GO BC@ZIF-8 PAN/chitosan/UiO-66NH2 nanofibers Ca-MOF MIL-101-NH2 UiO-66-EDTA Zn3(BTC)2MOF fabricated using graphene nanoscrolled
[(Me2NH2) Eu5(L)4(DMA)4 (H2O)6]n Zr-phosphate-based MOF
CMC-MOF/cloth composite membranes TMU-6, TMU-21, TMU-23, and TMU-24
MOF [Zn3L3(BPE)1.5]n MOF-808-EDTA MIL-100(Fe)/PDA
Table 7.2 (continued)
< 15 2880 60 3 ~120 < 30 –
522 26 mg/g 1.1 mM/g 357.9 mg/g –
Exchange and adsorption Adsorption Coordination and adsorption Adsorption
120
128.6 mg/g RE ¼ 81% 441.2 mg/g
–
–
Adsorption Adsorption Adsorption
–
–
7 0.02 5 7 –
6 5.5 0.5 6
7
8
5
Optimal pH 6 – 7
10
150
113 mg/g
RE > 95%
230, 221, 267, and 256 mg/g for TMU-6, TMU-21, TMU-23, and TMU-24, respectively RE ¼ 99.5%
862.44 mg/g
Uptake/removal efficiency (RE) 616.64 mg/g 313 mg/g 394 mg/g
Soft and tight binding through metal coordination Adsorption
Adsorption and ionic interactions
Adsorption
Mechanism Adsorption Coordination by EDTA Adsorption facilitated by the catechol groups in polydopamine Adsorption
Time to adsorption equilibrium (min) 180 120 99% 155 mg/g
253.8 mg/g
Adsorption
180 20
5.02–8.87
7
5.1–5.4
Soft-soft interaction
102.03 mg/g 536.22 mg/g
Coordination Adsorption
< 60
20
–
534 mg/g
–
98.5 mg/g
6π-η5-thiophene-Pb2+ coordination Ion exchange (inner-sphere complexation) and electrostatic interaction (outer-sphere complexation) Soft-soft interaction
(continued)
[81] [82]
[60]
[42] [59]
[40]
[80] [39]
[57]
[54]
[79]
[77] [78]
[52]
[76]
[75]
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . . 247
Heavy metal
60 3 20
415.6 mg/g 220 12 mg/g 693 mg/g
Adsorption
Adsorption
Exchange and adsorption Adsorption
MIL-88B(Fe)–Ag/TiO2 nanotubes/Ti plates PAN/chitosan/UiO-66NH2 nanofibers Ca-MOF nFe3O4@MIL-88A(Fe)/ APTMS
120
Uptake/removal efficiency (RE) 47, 49, 55, and 51 mg/g for TMU-6, TMU-21, TMU-23, and TMU-24, respectively 138 mg/g
Mechanism Adsorption
Time to adsorption equilibrium (min) 10
MOF TMU-6, TMU-21, TMU-23, and TMU-24
Table 7.2 (continued)
7 0.02 7
6
7
Optimal pH 8
[71] [78]
[70]
[67]
References [64]
248 S. Nayak
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . .
7.2.1
249
Mercury
In 2011, He et al. reported an analogue of MOF-5 with a thioether side chain on the linker to facilitate soft-soft interactions between sulfur and Hg(II). The MOF showed efficient separation of HgCl2 from a highly dilute solution of 84 mg/L [37]. In the same year, Ke et al. reported a post-synthetically modified HKUST-1, ThiolHKUST-1 (Fig. 7.2) [38]. The thiol functionalization was added by post-synthetic treatment of the MOF with dithioglycol. While the pristine HKUST-1 exhibited no affinity toward Hg(II) ions, Thiol-HKUST-1 exhibited excellent adsorption efficiencies of 714 mg/g, with almost 100% removal from water, reaching an equilibrium within 120 min. The high uptake was attributed to the large specific surface area and thiol-functionalized channels of the MOF. Abbasi et al. have reported a Co(II)-based MOF, [Co2(TATAB)(OH)(H2O)2] (where TATAB ¼ 4,40 ,400 -s-triazine-1,3,5triyltri-p-aminobenzoate), with 40–60% adsorption capacity of Hg(II) depending on the pH of the medium [39]. Higher efficiency for the nanostructured MOFs has been also reported in this work in comparison to bigger size crystals. Chakraborty et al. have reported a 3D flexible MOF [(NH2Me2)2][Zn3(L)2] (AMOF-10 ; L ¼ 5,50 -(1,4-phenylenebis(methylene))bis(oxy)diisophthalic acid) which is able to capture and remove Hg(II) from water with a detection limit in the ppm level (Fig. 7.3) [40]. Time-dependent measurements of desolvated AMOF-1 showed almost complete (98.7%) removal of Hg(II) from water in 24 h. Halder et al. have reported a Ni-based MOF [Ni(3-bpd)2(NCS)2]n (where 3-bpd is 1,4-bis(3-pyridyl)-2,3-diaza-1,3-butadiene) for selective visual detection and removal of Hg(II) ion in aqueous medium [41]. The presence of sulfur atoms on the coordinated thiocyanate group plays an important role in efficiently capturing Hg(II) by soft-soft interactions. The uptake
Fig. 7.2 An illustration of the functionalization process of Thiol-HKUST-1 with thiol groups being grafted onto the unsaturated metal centers (denoted as UMCs). (Reproduced with permission from [38])
250
S. Nayak
Fig. 7.3 Schematic representation of AMOF-1 and its applications to capture toxic heavy metal ions and Lewis acidic catalytic activity through PSMet. (Reproduced with permission from [40])
capacity of Hg(II) for this MOF is compared to the previously reported MOF by Ke et al. [38]. In 2016, Saleem et al. have reported a post-synthetically modified UiO-66-NH2, UiO-66-NHC(S)NHMe, which shows efficient removal of Hg (II) with a maximum uptake capacity of 769 mg/g [42]. The uptake capacity for UiO-66-NHC(S)NHMe was found to be much higher compared to the original UiO-66 and other three MOFs (UiO-66-NH2, UiO-66-NCS, and UiO-66-NCO) tested in this study. The authors proposed that the less sterically demanding methylthiourea group is responsible for the high uptake capacity for UiO-66-NHC (S)NHMe. In the same year, Mon et al. reported a BioMOF, [CaIICuII6[(S,S) methox]3(OH)2(H2O)]16H2O, with selective capture of CH3Hg+ and Hg(II) from water with excellent Hg(II) uptake capacity of 900 mg/g for HgCl2 and 166 mg/g for CH3HgCl [43]. The MOF was able to reduce the concentrations of [Hg2+] and [CH3Hg+] in water from a hazardous level of 10 ppm to safer values of 6 and 27 ppb, respectively. The high affinity of the MOF was attributed to the binding of the cations by thioether arms as shown in Fig. 7.4. In 2017, Ke et al. reported the Hg(II) removal from wastewater using thiolfunctionalized Fe3O4@HKUST-1 magnetic microspheres which were synthesized successfully by post-synthetic strategy (Fig. 7.5) [44]. The magnetically separable microspheres are able to uptake up to 348.43 mg/g of Hg(II) with an efficiency >90% even at a very low concentration of Hg(II). Easy recovery and recyclability of the material from an aqueous solution by using a permanent magnet make it very promising for future applications. Xiao et al. reported a MOF constructed by pillared
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . .
251
Fig. 7.4 Perspective views along the crystallographic c-axis of the porous structures of 3HgCl2@1 (a), 5HgCl2@1 (b), and CH3HgCl@1 (c). Cu and Ca atoms are represented by cyan and blue polyhedra, respectively. Hg and Cl atoms are depicted as purple and green spheres, whereas sulfur and carbon atoms from the methionine residues and the methyl groups, respectively, are shown as yellow and gray spheres. The remaining carbon, nitrogen, and oxygen atoms from the ligand are shown as sticks. Free water solvent molecules are omitted for clarity. The blue insets highlight the coordination environment of the captured mercury species. Reproduced with permission from [43]
Fig. 7.5 Schematic representation of the preparation of the thiol-functionalized Fe3O4@HKUST-1 core-shell magnetic microspheres and its use for selective heavy metal ion removal. Reproduced with permission from [44]
252
S. Nayak
(by pipyridine units) sheets of 1,4-dibromo-2,3,5,6-tetrakis(4-carboxyphenyl) benzene carboxylates (abbreviated as TCPBBr) and square paddle-wheel [Zn2(OOC)4] units [45]. The bipyridine units were then partly substituted with thioether 4-((pyridin-4-ylthio)methyl)pyridine) to produce the S-functionalized hybrid MOF-A/S. It was reported that the direct use of the thioether linkers did not produce the MOF. MOF-A/S was tested for Hg(II) capture, and it showed an uptake capacity of 78.8 mg/g which is much lower than some comparable MOFs. The authors suggested that the partial substitution of the linkers is responsible for this low uptake. In 2018, Peng et al. reported a broad-spectrum heavy metal ion trap (BS-HMT), a post-synthetically modified version of MOF-808. In this work, formate groups of MOF-808 have been strategically replaced by a chelating ligand, EDTA, which is equipped with four “hard” carboxylic groups and two relatively softer tertiary amine groups to produce MOF-808-EDTA (Fig. 7.6) [46]. The presence of these two different types of groups allows the MOF to capture a range of metal ions with different extent of Lewis acid strengths. MOF-808-EDTA shows an excellent uptake of 592 mg/g for Hg(II). Sun et al. have reported a MOF-polymer composite, MIL-100(Fe)/PDA, that exhibits rapid, selective removal of Hg(II) with an excellent uptake of 1634 mg/g [47]. The composite was prepared by treating MIL-100(Fe) dopamine, which polymerizes to polydopamine (PDA) the pores of MIL-100(Fe) via Fe(III) sites, creating a porous PDA attached to the internal MOF surface. The high efficiency was attributed to the heavy metal scavenging catechol groups inside the PDA composite (Fig. 7.7). The composite has also shown excellent selectivity over other common inorganic cations present in water. Rouhani et al. reported a Zn-based MOF, [Zn(BDC)(L*)]DMF (TMU-40) (where BDC is benzene-1,4-dicarboxylic acid, and L* is 5,6-di(pyridin-4-yl)-1,2,3,4-tetrahydropyrazine with Hg(II) uptake capacity of 269 mg/g) [48]. Ding et al. have reported a thiol-functionalized UiO-66 type Zr-based MOF that shows an uptake of 171.5 mg/g, approximately 9 times that of the pristine UiO-66 [49]. The work has also proposed that the protons are released from the thiol groups following the binding of the Hg(II) ions. Yazdi et al. have reported a family of polysulfide inserted HKUST-1 MOFs for heavy metal capture, showing an excellent distribution coefficient of 1.99 x 107 mL/g [50]. The strong affinity was attributed to the direct metal-sulfur chemical bonding as shown in Fig. 7.8. Esrafili et al. have reported a post-synthetically modified copper-based amide-functionalized MOF, TMU-46S, showing an uptake of 714 mg/g [51]. Li et al. have reported the heavy metal uptake by an exfoliated 2D nanosheet made of a cobalt-based MOF, [Co(CNS)2(pyz)2] (pyz ¼ pyrazine) or CoCNSP [52]. The few-layered CoCNSP nanosheets showed excellent Hg(II) uptake of 716 mg/g and achieving equilibrium at a very short time (20 min). The high affinity was attributed to the interaction between the heavy metals and the free S atom on coordinated thiocyanate groups which was further supported by DFT calculations. Hasankola et al. have reported a Zr-based MOF using 5,10,15,20-tetrakis(4-carboxyphenyl) porphyrin (H2TCPP) as a linker, and the resulting MOF PCN-221 has shown an uptake of 277 mg/g for Hg(II) [53]. The adsorption efficiency was attributed to the interaction between the Hg(II) and the electron donor N-sites on the linkers.
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . .
253
Fig. 7.6 Characterization and illustration of MOF-808 and MOF-808-EDTA. (a, b) Schematic illustration of the structures of the two MOFs. (c) PXRD patterns. (d) N2 adsorption-desorption isotherms. (e) Pore size distributions. (f, g) SEM images (scale bar, 500 nm). (h) 1H NMR spectra of (A) alkaline-digested MOF-808-EDTA, (B) alkaline-digested MOF-808, (C) EDTA-2Na, and (D) H3BTC in KOH/D2O solution. (Reproduced with permission from [46])
7.2.2
Lead
In 2013, Taghizadeh et al. reported a composite of Cu-BTC, or HKUST-1, (BTC ¼ benzene-1,3,5-tricarboxylate) MOF and dithizone-modified Fe3O4 nanoparticles for extraction of several heavy metals including lead (Fig. 7.9) [54]. The group reported very efficient extraction and recovery (up to >99%) of lead from different samples including distilled water, tap water, river water, seawater, soil, and so on. The easy recovery of the magnetic nanocomposite is an advantage of this material for heavy metal separation. In the same year, Zou et al. reported microwave-assisted synthesis of HKUST-1 (HKUST-1-MW) and a
254
S. Nayak
Fig. 7.7 Characterization of Fe-BTC/PDA. (Left) Polyhedral view of a large cage in the Fe-BTC with PDA embedded inside the channels. The purple sphere represents the void space inside of the cage prior to the dopamine addition. (Top) The simulated XRD pattern of MIL-100(Fe) (black) compared to the synchrotron XRD data (λ ¼ 0.50084 Å) of MIL-100(Fe) (red) and MIL-100(Fe)/ PDA-19 (blue). (Bottom) X-ray photoelectron spectroscopy data obtained from the N 1 s spectrum of Fe-BTC/PDA-19. (Reproduced with permission from [47])
polyoxometalate-containing HKUST-1 (HKUST-1-MW@H3PW12O40) with enhanced stability in water [55]. It was found that HKUST-1-MW@H3PW12O40 shows very efficient adsorption of Pb(II), reaching an equilibrium within 10 min with a removal efficiency of 98.18 mg/g. Interestingly this material shows high selectivity towards adsorption of Pb(II) over the other three heavy metals which were tested (Cd(II), Hg(II), and Cr(III)). In 2015, Ricco et al. reported a magnetic nanocomposite of Fe3O4 and aluminum-based MOF-53 with a varied ratio of terephthalic acid/aminoterephthalic acid as a linker to tune the Pb(II) uptake capacity [56]. The group also studied the sequestration behavior in a range of solvents (methanol, DMSO, and water), pH [2, 7, 12], and concentration of Pb(II) (10–8000ppm). A higher amount of aminoterephthalic acid resulted in better sequestration, with the superior adsorption observed when aminoterephthalic acid was used as the only linker. The magnetic nanocomposite showed an excellent uptake of 492.4 mg/g for Pb(II) with an additional advantage of easy separation of the material using a
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . .
255
Fig. 7.8 (a) Schematic illustration of the synthesis of metal-organic framework functionalized with polysulfide (Sx-MOF) for adsorption of heavy metal ions. (b) Structure of H3BTC and HKUST-1 metal-organic framework. (H3BTC ¼ 1,3,5-benzene tricarboxylic acid). (Reproduced with permission from [50])
Fig. 7.9 (a) A schematic diagram of dithizone-functionalized Fe3O4 synthesis. (b) The schematic illustration of synthesized magnetic MOF-DHz nanocomposite. (Reproduced with permission from [54])
256
S. Nayak
Fig. 7.10 Schematic representation of the interaction between the heavy metal ions and HS-mSi@MOF-5 with increasing pH. (Reproduced with permission from [59])
magnetic field. Babazadeh et al. reported another magnetic MOF composite of [(Fe3O4–ethylenediamine)/MIL-101(Fe)] which has shown efficient adsorption capacity for Pb(II) of 198 mg/g from agricultural samples [57]. Chakrabarty et al. reported desolvated anionic AMOF-1 which showed almost complete removal of Pb (II) (97.6%) from water in 24 h with an uptake capacity of 71 mg/g (Fig. 7.3) [40]. In 2016, Jamali et al. reported a family of isostructural lanthanide MOFs, [Ln(BTC)] (H2O)(DMF)1.1 (Ln ¼ Tb, Dy, Er, and Yb), using benzene tricarboxylic acid (BTC) as a linker [58]. The Dy-based MOF was found most efficient in adsorption of Pb (II) ions from water samples from different sources and showed an uptake capacity of 3.050 mg/g. The affinity was attributed to the Lewis acid-base interaction between the Pb(II) ions and the carboxylic oxygen in the channels along with surface adsorption. Saleem et al. reported a number of post-synthetically modified UiO-66-NH2 (UiO-66-NHC(S)NHMe, UiO-66-NHC(S)NHPh, UiO-66-NCS, and UiO-66-NCO) with their adsorptive removal of Pb(II) from aqueous solution [42]. A significant adsorption capacity of 232 mg/g was reported using UiO-66-NHC (S)NHMe. Zhang et al. reported a silica-treated thiol-functionalized MOF-5 (HS-mSi@MOF-5) that has shown a high capacity of 312 mg/g for Pb (II) [59]. The value is much higher than the untreated MOF-5 (211 mg/g). The material also shows superior stability compared to untreated MOF-5. The study also reported the pH dependence of the heavy metal adsorption with an increase in efficiency at higher pH. At lower pH there is a competition between the cationic metal ions and the H3O+ ions in an aqueous medium leading to the lower efficiency of heavy metal adsorption. In addition, at lower pH, the surface of the HS-mSi@MOF-5 becomes positive leading to electrostatic repulsion between the metal ions and the material and resulting in low efficiency (Fig. 7.10). Mao et al. reported a ZIF-8 and reduced grapheme oxide composite aerogel that efficiently separates several pollutants including heavy metals [60]. The team used a one-step method to fabricate the composite using self-assembly and a synergistic
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . .
257
Fig. 7.11 Schematic showing the mechanism of adsorption mechanism for Pb(II) ions in the MOF by the interaction with the free O groups in the linkers. (Reproduced with permission from [62])
chemical reduction and cross-linking by the metal ions. The material has shown an adsorption capacity of 281.5 mg/g for Pb(II), which is higher than the Zn(II)/reduced graphene with a removal capacity of 240 mg/g. The team proposed electrostatic interaction between the -OH groups of the reduced graphene and the cationic heavy metals as the driving force behind the adsorption mechanism. The thiolfunctionalized Fe3O4@HKUST-1 magnetic microspheres reported by Ke et al. display an adsorption capacity of 215.05 mg/g for Pb(II) [44]. The adsorption was facilitated by the soft-soft interaction between the heavy metal and the thiol groups. The material possesses an advantage of magnetic separation due to the integrated Fe3O4 nanoparticles. Wang et al. reported a composite magnetic cellulose nanocrystals (MCNC)/MOF material (MCNC@Zn-BTC) that shows an excellent adsorption capacity of 558.66 mg/g with an equilibrium reaching only in 30 min [61]. The material also showed an excellent recyclability with removal capacity retained over 80% after five cycles. Wu et al. reported a novel porous MOF, [Zn3L3(BPE)1.5]n (BPE ¼ bis(4-pyridyl)ethylene), with free -O groups at its channel which efficiently coordinates and adsorbs Pb(II) at an exceptional removal capacity of 616.64 mg/g (Fig. 7.11) [62]. The adsorption efficiency becomes higher at higher pH, and the MOF shows excellent selectivity over other metal ions, such as Ca(II) and Mg(II), and shows very good stability in water. The BS-HMT, MOF-808-EDTA, shows an uptake capacity of 313 mg/g for Pb(II) ions (Fig. 7.6) [46]. Sun et al. reported the MIL-100(Fe)/PDA, a MOF-polydopamine composite (Fig. 7.7) which shows an excellent removal and separation of Pb(II) with an uptake capacity of 394 mg/g [47]. The composite exhibits a selective removal of Pb(II) even in presence of interferants such as sodium ions in an excess of 14,000 times that of Pb(II) ions. Yang et al. reported a unique
258
S. Nayak
Fig. 7.12 Schematic representation of the preparation of CMC/cloth and CMC-MOF/cloth composite membranes. (Reproduced with permission from [63])
CMC-MOF/cloth composite membrane (where MOF is a Co/Zn-ZIF and CMC is carboxymethylcellulose sodium), which was fabricated on cloth by taking advantage of the hydroxyl groups on cellulose (Fig. 7.12) [63]. The composite showed an excellent uptake capacity of 862.44 mg/g for Pb(II) and has great features including its low-cost, easy preparation and handling, and environment-friendly aspects. Esrafili et al. reported a series of mechanochemically synthesized isoreticular pillared MOFs with and without imine groups present on the linkers [64]. The MOFs, namely, [Zn(oba)(4-bpmb)0.5] (DMF)1.5 (TMU-6) (bpmb ¼ N,N0 -bis-(4-pyridylmethylene)-1,4-benzenediamine), [Zn(oba)(bpmn)0.5] (DMF)1.5 (TMU-21), (bpmn ¼ N,N0 -bis-(4-pyridylmethylene)-1,5-naphthalenediamine), TMU-23 ([Zn2(oba)2(bpfb)] (DMF)5), (H2oba ¼ 4,40 -oxybis(benzoic acid), bpfb ¼ N,N0 -bis-(4-pyridylformamide)-1,4-benzenediamine), and TMU-24 ([Zn2(oba)2(bpfn)] (DMF)2) (bpfn ¼ N,N0 -bis(4-pyridylformamide)-1,5naphthalenediamine), were reported with Pb(II) adsorption capacities of 230, 221, 267, and 256 mg/g, respectively. The study reports that the MOFs with amidedecorated linkers (TMU-23 and TMU-24) showed greater adsorption capacities for Pb(II) compared to the MOFs with imine-decorated linkers (TMU-6 and TMU-21). Lu et al. reported an anionic MOF, [(Me2NH2)Eu5(L)4(DMA)4 (H2O)6]n (where H4L ¼ 1,3-bis-[3,5-bis(carboxy)phenoxy]propane) [65]. The MOF showed good water stability and a removal efficiency of 99.5% for Pb(II) ions. In a recent
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . .
259
Fig. 7.13 Schematic showing the fabrication process of BC@MOFs composite aerogels on the bacterial cellulose. (Reproduced with permission from [69])
computational study by Gu et al., it was proposed that the ultra-stable porous Zr-phosphate MOFs show high efficiency (>95%) for the removal of heavy metal ions by using two different mechanisms [66]. The study proposed two different mechanisms based on the mode of the interactions: the “loose mode” of interaction that results from a coordination of the heavy metals into the pores by solvent molecules (such as water) which are weakly held by hydrogen bond interactions with the coordinating ligands in the MOF and “tight mode” of interaction which results from direct interaction between the heavy metals and the ligands in the MOF, both mechanisms leading to trapping the heavy metals in the pore of the MOF filters. Mohaghegh et al. reported a multifunctional MIL-88B(Fe)-coated Ag/TiO2 nanotubes/Ti plate that shows an adsorption capacity of 113 mg/g for Pb(II) ions [67]. In order to improve the stability in an aqueous environment, Lu et al. developed a layered composite of Fe-based MOF MIL-101(Fe) and graphene oxide (GO) that shows an efficiency of 128.6 mg/g for removal of Pb(II) from an aqueous solution with a fast adsorption equilibrium time reaching in 15 min [68]. Ma et al. reported a flexible macroscopic aerogel of ZIF-8 deposited on bacterial cellulose (BC@ZIF-8) that shows a removal efficiency of 81% for Pb(II) ions, 1.2% greater compared to pure ZIF-8 [69]. The hydroxyl groups on the bacterial cellulose walls attract metal ions that facilitate the formation of the ZIF-8 on the surface, as shown in Fig. 7.13. Jamshidifard et al. reported a composite of UiO-66-NH2 grown on PAN/chitosan nanofiber (PAN ¼ polyacrylonitrile), which was fabricated by using an electrospinning process [70]. The nanofiber achieved a removal efficiency of 441.2 mg/g for Pb(II) ions with optimized condition using 10% MOF content and at pH 6 at 25 C. The adsorbent also showed good recyclability. Pournara et al. reported a calcium-based MOF, [Ca(H4L)(DMA)2] 2DMA (Ca-MOF) (H6L ¼ N, N0 -bis(2,4-dicarboxyphenyl)-oxalamide), which shows an excellent adsorption capacity of 522 26 mg/g for Pb(II) ions [71]. When a silica column was loaded
260
S. Nayak
with 1% Ca-MOF, the column was able to remove traces of Pb(II) from a wastewater simulant solution containing a large excess of competitive metal ions. Lv et al. reported an amine-functionalized Fe-based MOF, MIL-101-NH2, which shows an uptake capacity of 1.1 mM/g for Pb(II) ions [72]. The MOF also acts as a sensor for the heavy metal ions due to the fluorescence originating from the linker. Wu et al. reported a modified UiO-66 which contains attached EDTA on the Zr(IV)-ions [73]. The attached EDTA coordinates to a range of metal ions and thus facilitates their removal with high efficiency. The MOF showed an adsorption capacity of 357.9 mg/g for the Pb(II) ions. Samantaray et al. reported a rodlike morphology of Zn3(BTC)2MOF fabricated using nanoscrolled graphene [74]. The MOF showed efficient remediation of tap water samples spiked with 0.1 ppm of Pb(II) ion, for a period over 15 days. In a recent study, Geisse et al. reported the remediation of Pb (II) from water using a thiophene-containing MOF, DUT-67, with a loading efficiency of 98.5 mg/g [75]. The mechanism was studied in detail using the 207Pb NMR analysis which indicates the interaction between the Pb(II) ions and the MOF is via 6π-η5-thiophene-Pb2+ coordination bonds. From a comparison study, the work also showed that similar MOFs, without S-containing heterocycles, are less effective for Pb(II) adsorption. Zhang et al. reported an insight into the mechanism for heavy metal adsorption by iron-containing MOF, MIL-100(Fe), and amorphous Fe-BTC [76]. Both materials achieve equilibrium very quickly with a high kinetic rate constant, with MIL-100(Fe) showing superior performance. The performance was dependent on MOF dosage, solution pH, and temperature. All adsorptions were endothermic, with an observation showing better removal efficiency for Pb (II) compared to Cd(II) for both MOFs. This is ascribed to more negative hydration enthalpy and larger ionic radius of Cd(II). The adsorption kinetics was found as of second order. The adsorption of the heavy metals was driven by two mechanisms, namely, ion exchange (inner-sphere complexation) and electrostatic interaction (outer-sphere complexation). The exfoliated 2D nanosheet made of cobalt-based MOF, [Co(CNS)2(pyz)2] (pyz ¼ pyrazine) or CoCNSP, shows a high uptake of 534 mg/g with very fast capture dynamics (99% recovery with an adsorption capacity of 188 mg/g for Cd(II) ions from different sources including distilled water, tap water, river water, seawater, soil, and so on (Fig. 7.9) [54]. Another magnetic MOF composite of [(Fe3O4–ethylenediamine)/MIL-101(Fe)] reported by Babazadeh et al. showed efficient recovery (>90% in all cases) and high adsorption capacity (155 mg/g) for Cd(II) from several agricultural samples [57]. The authors also reported that the adsorption efficiency of the material was not affected in presence of other common metal ions. Wang et al. reported a post-synthetically modified HKUST-1 with channels functionalized with –SO3H groups [80]. The MOF (HKUST-1-SO3H) shows a very good adsorption capacity of 88.7 mg/g for Cd(II).
262
S. Nayak
The material shows a superior adsorption capacity for Cd(II) compared to the parent HKUST-1 and HKUST-1-SH which shows an uptake capacity of 67.8 mg/g and 74.5 mg/g, respectively. The lower adsorption capacity for the parent HKUST-1 is attributed to the complexation of Cd(II) with the residual carboxylic groups. The adsorption capacity improves with the introduction of the -SH groups in HKUST-1SH. With the introduction of the –SO3H groups, the adsorption capacity is further increased due to the possible complexation by multiple binding sites of the sulfanilic group. The MOF reaches adsorption equilibrium, within 10 minutes, and shows the best performance at pH 6. The Co(II)-based MOF, [Co2(TATAB)(OH)(H2O)2], reported by Abbasi et al. shows around 60% adsorption capacity of Hg(II) in the pH range of 2–6 with slightly improved efficiency at higher pH [39]. The nanostructured MOF shows slightly better performance than the bigger-sized crystalline material. The interaction between the metal ions and the MOF was through the NH groups on the linkers and was evident from the shifting of the NH stretching band in the infrared spectra. The AMOF-10 reported by Chakraborty et al. showed an uptake capacity of 41 mg/g for Cd(II) ions when immersed in contaminated water for 7 days [40]. The MOF using UiO-66-NHC(S)NHMe. reported by Saleem et al. showed an uptake capacity of 49 mg/g for Cd(II) [42]. A silica-treated thiol-functionalized MOF-5 (HS-mSi@MOF-5) showed an uptake capacity of 65.2 mg/g for Pb (II) [59]. The value is much higher than the untreated MOF-5 (4.2 mg/g). The ZIF-8 and reduced graphene oxide composite aerogel reported by Mao et al. efficiently separates several pollutants including heavy metals [60]. The material showed an adsorption capacity of 101.1 mg/g for Cd(II), which is higher than the Zn(II)/reduced graphene with a removal capacity of 72.4 mg/g. Electrostatic interaction between the -OH groups of the reduced graphene and the cationic heavy metals was indicated as the driving force behind the adsorption mechanism. Fan et al. reported a hybrid Fe3O4/MOF/L-cysteine using the Cu(II) and 5-aminoterephthalic acid [81]. The hybrid material showed very efficient removal of Cd(II) with an uptake capacity of 248.24 mg/g. The hybrid also acts as a fluorescent sensor for “turnoff” detection of Cd(II), with a detection limit of 0.94 ng/mL. The BS-HMT, MOF-808-EDTA (Fig. 7.6), reported by Peng et al. has shown a Cd(II) uptake capacity of 417 mg/g [46]. The chelating EDTA groups play the key role in efficient sequestration of the metal ions by formation of chelate complex. Efome et al. reported an electrospun composite membrane, PAN/MOF808, made of polyacrylonitrile (PAN) and Zr-based MOF-808 [82]. The membrane with an optimized MOF content of 20% showed an uptake capacity of 225.055 mg/g for Cd(II) ions. The high uptake was attributed to the high surface charge (ζ potential ¼ 36 mV at pH 4.5) of the composite fibers and large surface area (936 m2/g). The composite also showed a higher selectivity towards Cd(II) in presence of other common interfering metal ions (Na+, Ca2+, and Mg2+), and the superior adsorption was attributed to the higher electronegativity of Cd(II) compared to the other metal ions tested. The previously described TMU MOFs were also tested for the adsorption of Cd(II) ions, and their maximum adsorption capacities were found as 47, 49, 55, and 51 mg/g for TMU-6, TMU-21, TMU-23, and TMU-24, respectively [64]. The multifunctional MIL-88B(Fe)-coated Ag/TiO2 nanotubes/Ti
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . .
263
plate reported by Mohaghegh et al. showed an adsorption capacity of 138 mg/g for Cd(II) ions [67]. The adsorption capacity of the MOF-coated material was approximately 1.6 times higher than that of pure TiO2 nanotubes/Ti. A composite of UiO-66-NH2 grown on electrospun PAN/chitosan nanofiber (PAN ¼ polyacrylonitrile) reported by Jamshidifard et al. showed an uptake capacity of 415.6 mg/g for Cd (II) ions [70]. The calcium-based MOF, [Ca(H4L)(DMA)2] 2DMA (Ca-MOF) (H6L ¼ N,N0 -bis(2,4-dicarboxyphenyl)-oxalamide), shows an excellent adsorption capacity of 220 12 mg/g, for Cd(II) ions [71]. The sorption capacities of Ca-MOF for Cd(II) were found to be 98% for a spring water sample containing competitive metal ions, such as 93.1 ppm of Ca2+, 1.9 ppm of Mg2+, 2.6 ppm of Na+, and 0.7 ppm of K+. Recently reported magnetic hybrid composite of nFe3O4@MIL-88A (Fe) and 3-aminopropyltrimethoxysilane (APTMS) showed one of the highest uptake capacities of 693 mg/g of Cd(II) ions [78]. The material also showed very good recyclability.
7.3
Summary
Besides natural phenomena, anthropogenic actions, such as the industrial revolution and development of sophisticated technologies, have led to a global rise in the pollution level of groundwater. With the current rate of pollution, the data suggests that the availability of drinkable water will be one of the major crises in the near future due to the level of contamination with hazardous materials. Heavy metals constitute a large part of the contaminants of groundwater and possess several health risks when present even in very low concentration. With the current scale of the problem, urgent development is needed in innovative technologies that can help removing the heavy metals from water efficiently. The majority of the existing materials in use produce a large amount of side products/sludge or economically not viable because of their one-time usage or high price. As a relatively new class of materials, MOFs have shown promising properties for use in heavy metal removal by adsorption mechanism. The ultrahigh surface area of this class of materials plays the key role in the large uptake capacity of the heavy metals. The capacity is further enhanced by careful engineering of the pores with the incorporation of functional groups that can contribute to the enhanced affinity of the MOFs towards the heavy metal ions. Particularly several groups have used the thiol groups, which enhance affinity towards the heavy metal ions by preferential soft acid-soft base interaction. This approach also helps in enhancing the selectivity towards the heavy metal ions in presence of the common essential metal ions present in drinking water. The use of additional free hydroxyl groups or carboxylic groups on the linkers has been also used to enhance the adsorption capacities by coordination and electrostatic interactions. Additionally, some anionic MOFs have been used to capture and remove the heavy metal ions by substitution of the counterions. Theoretical studies also indicated that there is more than one kind of interactions between the MOFs and the guest metal ions in the pores. For example, some guest metal ions are coordinated by
264
S. Nayak
the ligands that are part of the MOF and are named as “tight mode” of interaction, while some metals are coordinated by the solvent molecules that are weakly held by hydrogen bond interactions with the coordinating ligands in the MOF and named as “loose mode” of interactions. Besides the pure MOFs, recent developments have seen a large number of MOF-based composite materials for studying heavy metal removal from water. There are mainly two motivations behind using the hybrid materials: (i) improved stability and (ii) easy handling. Water stability of MOFs is a well-known problem, and only a handful of MOFs (particularly Zr(IV)-based MOFs and Al(III) based MOFs) have shown significant stability in water, and a large number of MOFs studied for heavy metal removal from water are based on these type of MOFs and their composites. In some studies, MOFs have been deposited on porous polymeric fibers, such as polyacrylonitrile (PAN). Taking advantage of the polyhydroxy groups of cellulose, MOFs have been also deposited on cloth for heavy metal removal. In recent studies, MOF-graphene nanocomposites have been also prepared for heavy metal removal. The uses of composite materials also have broadened the scope of their applications by giving the means of easy handling and manipulation for use in practical devices. Besides different MOF-based materials, there are other parameters which play very important roles, such as pH of the medium and recyclability. It has been observed that in majority of cases, the removal efficiency is best at around pH 6–7. At lower pH there is a competition between the protonated water (H3O+) and the heavy metals, and at higher pH a majority of the MOFs decompose, and the heavy metals form insoluble hydroxides. The reusability of the MOFs or MOF composites is one of the major issues to address before these materials can be used in a large scale. A majority of these materials lose their efficiency gradually within a few cycles of use. The other issue to address is the cost of preparing these MOFs on an industrial scale for wide applications, particularly MOFs that show promising uptake capacity for heavy metal ions. In summary, MOFs have shown some excellent uptake capacities for the heavy metal ions and have several advantages over the traditional materials that are in use currently for heavy metal removal during wastewater treatment. However, there are some limitations (such as upscaling and stability) currently with MOFs, which need to be addressed before they can be used on an industrial scale for heavy metal removal from water.
References 1. Jury WA, Vaux HJ (2007) The emerging global water crisis: Managing scarcity and conflict between water users. In: Advances in Agronomy, Vol 95. Advances in Agronomy, vol 95. Elsevier Academic Press Inc, San Diego, pp 1–76 2. Fu FL, Wang Q (2011) Removal of heavy metal ions from wastewaters: A review. J. Environ. Manag. 92:407–418 3. Chong MN, Jin B, Chow CWK et al (2010) Recent developments in photocatalytic water treatment technology: A review. Water Res. 44:2997–3027
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . .
265
4. Gupta VK, Suhas (2009) Application of low-cost adsorbents for dye removal – a review. J. Environ. Manag. 90:2313–2342 5. Mohan D, Pittman CU (2007) Arsenic removal from water/wastewater using adsorbents – a critical review. J. Hazard. Mater. 142:1–53 6. Chen GH (2004) Electrochemical technologies in wastewater treatment. Sep. Purif. Technol. 38:11–41 7. Islam MS, Ahmed MK, Raknuzzaman M et al (2015) Heavy metal pollution in surface water and sediment: A preliminary assessment of an urban river in a developing country. Ecol. Indic. 48:282–291 8. Robinson BH (2009) E-waste: An assessment of global production and environmental impacts. Sci. Total Environ. 408:183–191 9. Kaushik A, Kansal A, Santosh et al (2009) Heavy metal contamination of river Yamuna, Haryana, India: Assessment by Metal Enrichment Factor of the Sediments. J. Hazard Mater. 164:265–270 10. Lim HS, Lee JS, Chon HT et al (2008) Heavy metal contamination and health risk assessment in the vicinity of the abandoned Songcheon Au-Ag mine in Korea. J. Geochem. Explor. 96:223–230 11. Choong TSY, Chuah TG, Robiah Y et al (2007) Arsenic toxicity, health hazards and removal techniques from water: An overview. Desalination 217:139–166 12. Razo I, Carrizales L, Castro J et al (2004) Arsenic and heavymetal pollution of soil, water and sediments in a semi-arid climate mining area in Mexico. Water Air Soil Pollut. 152:129–152 13. Jarup L (2003) Hazards of heavy metal contamination. Br. Med. Bull. 68:167–182 14. Ikem A, Egiebor NO, Nyavor K (2003) Trace elements in water, fish and sediment from Tuskegee Lake, Southeastern USA. Water Air Soil Pollut. 149:51–75 15. Alam MGM, Snow ET, Tanaka A (2003) Arsenic and heavy metal contamination of vegetables grown in Samta village, Bangladesh. Sci. Total Environ. 308:83–96 16. Capel PD, Giger W, Reichert P et al (1988) Accidental input of pesticides into the Rhine river. Environ. Sci. Technol. 22:992–997 17. Tchounwou PB, Yedjou CG, Patlolla AK et al (2012) Heavy metal toxicity and the environment. Exp. Suppl. 101:133–164 18. Duffus JH (2002) “Heavy metals” a meaningless term? (IUPAC technical report). Pure Appl. Chem. 74:793 19. Fergusson JE (1990) The Heavy Elements: Chemistry, Environmental Impact and Health Effects. Pergamon, Oxford 20. Bradl HB (2005) Heavy Metals in the Environment: Origin, Interaction and Remediation, 1st edn. Elsevier Academic Press, London; 283 p 21. Kouimtzis T, Samara C (1995) Airborne Particulate Matter. Springer, New York/Berlin/ Heidelberg 22. Nriagu JO (1996) A history of global metal pollution. Science 272:223 23. Nriagu JO (1979) Global inventory of natural and anthropogenic emissions of trace metals to the atmosphere. Nature 279:409–411 24. Kobielska PA, Howarth AJ, Farha OK et al (2018) Metal–organic frameworks for heavy metal removal from water. Coord. Chem. Rev. 358:92–107 25. Järup L (2003) Hazards of heavy metal contamination. Br. Med. Bull. 68:167–182 26. Ochia E-I (1987) Oxidation—Reduction and enzymes and proteins. In: General Principles of Biochemistry of the Elements. Biochemistry of the Elements. Springer, Boston, MA 27. Valasatava Y, Rosato A, Furnham N et al (2018) To what extent do structural changes in catalytic metal sites affect enzyme function? J. Inorg. Biochem. 179:40–53 28. Aragay G, Pons J, Merkoçi A (2011) Recent trends in macro-, micro-, and nanomaterial-based tools and strategies for heavy-metal detection. Chem. Rev. 111:3433–3458 29. Mozaffarian D, Rimm EB (2006) Fish intake, contaminants, and human health – evaluating the risks and the benefits. J. Am. Med. Assoc. 296:1885–1899
266
S. Nayak
30. Schock MR, Hyland RN, Welch MM (2008) Occurrence of contaminant accumulation in lead pipe scales from domestic drinking-water distribution systems. Environ. Sci. Technol. 42:4285–4291 31. Lidsky TI, Schneider JS (2003) Lead neurotoxicity in children: Basic mechanisms and clinical correlates. Brain 126:5–19 32. Steenland K, Boffetta P (2000) Lead and cancer in humans: Where are we now? Am. J. Ind. Med. 38:295–299 33. Jarup L, Akesson A (2009) Current status of cadmium as an environmental health problem. Toxicol. Appl. Pharmacol. 238:201–208 34. Huff J, Lunn RM, Waalkes MP et al (2007) Cadmium-induced cancers in animals and in humans. Int. J. Occup. Environ. Health 13:202–212 35. Fu F, Wang Q (2011) Removal of heavy metal ions from wastewaters: A review. J. Environ. Manag. 92:407–418 36. Fu F, Wang Q (2011) Removal of heavy metal ions from wastewaters: A review. J. Environ. Manag. 92:407–418 37. He J, Yee KK, Xu ZT et al (2011) Thioether side chains improve the stability, fluorescence, and metal uptake of a metal-organic framework. Chem. Mater. 23:2940–2947 38. Ke F, Qiu LG, Yuan YP et al (2011) Thiol-functionalization of metal-organic framework by a facile coordination-based postsynthetic strategy and enhanced removal of Hg2+ from water. J. Hazard. Mater. 196:36–43 39. Abbasi A, Moradpour T, Van Hecke K (2015) A new 3D cobalt (II) metal-organic framework nanostructure for heavy metal adsorption. Inorg. Chim. Acta 430:261–267 40. Chakraborty A, Bhattacharyya S, Hazra A et al (2016) Post-synthetic metalation in an anionic MOF for efficient catalytic activity and removal of heavy metal ions from aqueous solution. Chem. Commun. 52:2831–2834 41. Halder S, Mondal J, Ortega-Castro J et al (2017) A Ni-based MOF for selective detection and removal of Hg2+ in aqueous medium: A facile strategy. Dalton Trans. 46:1943–1950 42. Saleem H, Rafique U, Davies RP (2016) Investigations on post-synthetically modified UiO-66NH2 for the adsorptive removal of heavy metal ions from aqueous solution. Microporous Mesoporous Mater. 221:238–244 43. Mon M, Lloret F, Ferrando-Soria J et al (2016) Selective and efficient removal of mercury from aqueous media with the highly flexible arms of a BioMOF. Angew. Chem. Int. Ed. 55:11167–11172 44. Ke F, Jiang J, Li YZ et al (2017) Highly selective removal of Hg2+ and Pb2+ by thiolfunctionalized Fe3O4@metal-organic framework core-shell magnetic microspheres. Appl. Surf. Sci. 413:266–274 45. Xiao GW, Chen TF, Sun XZ et al (2017) A route to robust thioether-functionalized MOF solid materials displaying heavy metal uptake and the ability to be further oxidized. Dalton Trans. 46:12036–12040 46. Peng YG, Huang HL, Zhang YX et al (2018) A versatile MOF-based trap for heavy metal ion capture and dispersion. Nat. Commun. 9:187 47. Sun DT, Peng L, Reeder WS et al (2018) Rapid, selective heavy metal removal from water by a metal-organic framework/Polydopamine composite. ACS Cent. Sci. 4:349–356 48. Rouhani F, Morsali A (2018) Fast and selective heavy metal removal by a novel metal-organic framework designed with in-situ ligand building block fabrication bearing free nitrogen. Chem. Eur. J. 24:5529–5537 49. Ding L, Luo XB, Shao PH et al (2018) Thiol-functionalized Zr-based metal-organic framework for capture of hg(II) through a proton exchange reaction. ACS Sustain. Chem. Eng. 6:8494–8502 50. Yazdi MN, Yamini Y, Asiabi H et al (2018) A metal organic framework prepared from benzene-1,3,5-tricarboxylic acid and copper(II), and functionalized with various polysulfides as a sorbent for selective sorption of trace amounts of heavy metal ions. Microchim. Acta 185:1–8
7 Water Purification: Removal of Heavy Metals Using Metal-Organic Frameworks. . .
267
51. Esrafili L, Gharib M, Morsali A (2019) Selective detection and removal of mercury ions by dual-functionalized metal-organic frameworks: Design-for-purpose. New J. Chem. 43:18079–18091 52. Li J, Duan Q, Wu Z et al (2020) Few-layered metal-organic framework nanosheets as a highly selective and efficient scavenger for heavy metal pollution treatment. Chem. Eng. J. 383:123189 53. Seyfi Hasankola Z, Rahimi R, Shayegan H et al (2020) Removal of Hg2+ heavy metal ion using a highly stable mesoporous porphyrinic zirconium metal-organic framework. Inorg. Chim. Acta 501:119264 54. Taghizadeh M, Asgharinezhad AA, Pooladi M et al (2013) A novel magnetic metal organic framework nanocomposite for extraction and preconcentration of heavy metal ions, and its optimization via experimental design methodology. Microchim. Acta 180:1073–1084 55. Zou F, Yu RH, Li RG et al (2013) Microwave-assisted synthesis of HKUST-1 and functionalized HKUST-1-@H3PW12O40: Selective adsorption of heavy metal ions in water analyzed with synchrotron radiation. ChemPhysChem 14:2825–2832 56. Ricco R, Konstas K, Styles MJ et al (2015) Lead(II) uptake by aluminium based magnetic framework composites (MFCs) in water. J. Mater. Chem. A 3:19822–19831 57. Babazadeh M, Hosseinzadeh-Khanmiri R, Abolhasani J et al (2015) Solid phase extraction of heavy metal ions from agricultural samples with the aid of a novel functionalized magnetic metal-organic framework. RSC Adv. 5:19884–19892 58. Jamali A, Tehrani AA, Shemirani F et al (2016) Lanthanide metal-organic frameworks as selective microporous materials for adsorption of heavy metal ions. Dalton Trans. 45:9193–9200 59. Zhang JM, Xiong ZH, Li C et al (2016) Exploring a thiol-functionalized MOF for elimination of lead and cadmium from aqueous solution. J. Mol. Liq. 221:43–50 60. Mao JJ, Ge MZ, Huang JY et al (2017) Constructing multifunctional MOF@rGO hydro/ aerogels by the self-assembly process for customized water remediation. J. Mater. Chem. A 5:11873–11881 61. Wang N, Ouyang XK, Yang LY et al (2017) Fabrication of a magnetic cellulose Nanocrystal/ metal-organic framework composite for removal of Pb(II) from water. ACS Sustain. Chem. Eng. 5:10447–10458 62. Yu CX, Shao ZC, Hou HW (2017) A functionalized metal-organic framework decorated with O- groups showing excellent performance for lead(II) removal from aqueous solution. Chem. Sci. 8:7611–7619 63. Yang WX, Wang J, Yang QF et al (2018) Facile fabrication of robust MOF membranes on cloth via a CMC macromolecule bridge for highly efficient Pb(II) removal. Chem. Eng. J. 339:230–239 64. Esrafili L, Safarifard V, Tahmasebi E et al (2018) Functional group effect of isoreticular metalorganic frameworks on heavy metal ion adsorption. New J. Chem. 42:8864–8873 65. Lu SQ, Liu YY, Duan ZM et al (2018) Improving water-stability and porosity of lanthanide metal-organic frameworks by stepwise synthesis for sensing and removal of heavy metal ions. Cryst. Growth Des. 18:4602–4610 66. Gu ZL, Song W, Yang ZX et al (2018) Metal-organic framework as an efficient filter for the removal of heavy metal cations in water. Phys. Chem. Chem. Phys. 20:30384–30391 67. Mohaghegh N, Faraji M, Abedini A (2019) Highly efficient multifunctional Ag/TiO2 nanotubes/Ti plate coated with MIL-88B(Fe) as a photocatalyst, adsorbent, and disinfectant in water treatment. Appl. Phys. A Mater. Sci. Process. 125. https://doi.org/10.1007/s00339-0182324-8 68. Lu MJ, Li L, Shen SQ et al (2019) Highly efficient removal of Pb2+ by a sandwich structure of metal-organic framework/GO composite with enhanced stability. New J. Chem. 43:1032–1037 69. Ma XT, Lou Y, Chen XB et al (2019) Multifunctional flexible composite aerogels constructed through in-situ growth of metal-organic framework nanoparticles on bacterial cellulose. Chem. Eng. J. 356:227–235
268
S. Nayak
70. Jamshidifard S, Koushkbaghi S, Hosseini S et al (2019) Incorporation of UiO-66-NH2 MOF into the PAN/chitosan nanofibers for adsorption and membrane filtration of Pb(II), Cd(II) and Cr(VI) ions from aqueous solutions. J. Hazard. Mater. 368:10–20 71. Pournara AD, Margariti A, Tarlas GD et al (2019) A Ca2+ MOF combining highly efficient sorption and capability for voltammetric determination of heavy metal ions in aqueous media. J. Mater. Chem. A 7:15432–15443 72. Lv SW, Liu JM, Li CY et al (2019) A novel and universal metal-organic frameworks sensing platform for selective detection and efficient removal of heavy metal ions. Chem. Eng. J. 375:122111 73. Wu J, Zhou J, Zhang SW et al (2019) Efficient removal of metal contaminants by EDTA modified MOF from aqueous solutions. J. Colloid Interface Sci. 555:403–412 74. Samantaray PK, Baloda S, Madras G et al (2019) Nanodelivery in scrolls-based Nanocarriers: Efficient constructs for sustainable scavenging of heavy metal ions and inactivate bacteria. ACS Sustain. Chem. Eng. 7:18775–18784 75. Geisse AR, Ngule CM, Genna DT (2020) Removal of lead ions from water using thiophenefunctionalized metal-organic frameworks. Chem. Commun. 56:237–240 76. Zhang BL, Qiu W, Wang PP et al (2020) Mechanism study about the adsorption of Pb(II) and Cd(II) with iron-trimesic metal-organic frameworks. Chem. Eng. J. 385:123507 77. Liu Q, Li S, Yu H et al (2020) Covalently crosslinked zirconium-based metal-organic framework aerogel monolith with ultralow-density and highly efficient Pb(II) removal. J. Colloid Interface Sci. 561:211–219 78. Mahmoud ME, Amira MF, Seleim SM et al (2020) Amino-decorated magnetic metal-organic framework as a potential novel platform for selective removal of chromium (Vl), cadmium (II) and lead (II). J. Hazard. Mater. 381:120979 79. Xu R, Jian M, Ji Q et al (2020) 2D water-stable zinc-benzimidazole framework nanosheets for ultrafast and selective removal of heavy metals. Chem. Eng. J. 382:122658 80. Wang Y, Ye GQ, Chen HH et al (2015) Functionalized metal-organic framework as a new platform for efficient and selective removal of cadmium(II) from aqueous solution. J. Mater. Chem. A 3:15292–15298 81. Fan L, Deng M, Lin CX et al (2018) A multifunctional composite Fe3O4/MOF/L-cysteine for removal, magnetic solid phase extraction and fluorescence sensing of Cd(II). RSC Adv. 8:10561–10572 82. Efome JE, Rana D, Matsuura T et al (2018) Insight studies on metal-organic framework Nanofibrous membrane adsorption and activation for heavy metal ions removal from aqueous solution. ACS Appl. Mater. Interfaces 10:18619–18629
Chapter 8
Adsorptive Purification of Water Contaminated with Hazardous Organics by Using Functionalized Metal-Organic Frameworks Dong Kyu Yoo, Biswa Nath Bhadra, and Sung Hwa Jhung
8.1
Introduction
With increasing population and economic growth, contamination of water resources has been very severe worldwide. Not only surface water but also groundwater has been contaminated with various hazardous organic materials such as agrochemicals, dyes, and pharmaceutical and personal care products (PPCPs) [1–5]. Moreover, the requirement of freshwater increases rapidly with increasing population and economic growth. Therefore, both prevention of water contamination and facile purification of polluted water are very important for our life with safe water [1–5]. Various efforts, including oxidation, photocatalysis, and extraction, have been exerted in order to purify water contaminated with organics [1–5]. Adsorptive removal of such contaminants has been interesting to researchers both in academia and industry because of possible application in large-scale and low operation cost/ investment [6–9]. However, an adsorptive removal technology can be successful or competitive only when feasible/cheap adsorbents with high capacity and ready reusability are available [6–9]. Porous materials [10, 11], which can be useful/competitive adsorbents, have been advanced recently mainly because of the introduction of metal-organic frameworks (MOFs) [12–15], composed of both organic and metallic species. MOFs have been very popular, especially in adsorption (including adsorptive purification of contaminated water), because of huge porosity, designable porous structure, and ready
Dong Kyu Yoo and Biswa Nath Bhadra contributed equally to this work. D. K. Yoo · B. N. Bhadra · S. H. Jhung (*) Department of Chemistry and Green-Nano Materials Research Center, Kyungpook National University, Daegu, South Korea e-mail: [email protected] © Springer Nature Switzerland AG 2021 P. Horcajada Cortés, S. Rojas Macías (eds.), Metal-Organic Frameworks in Biomedical and Environmental Field, https://doi.org/10.1007/978-3-030-63380-6_8
269
270
D. K. Yoo et al.
Number of publication
10000
1200
MOF MOF/adsorption MOF/adsorption/water
8000 900 6000 600 4000 300 2000
0
0
2015
2016
2017
2018
2019
Year Fig. 8.1 Number of papers (with year) on MOF, MOF/adsorption, and MOF/adsorption/water, searched with Web of Science
functionalization [7–9]. As summarized in Fig. 8.1, studies on MOFs have been steadily increased, especially MOFs for adsorption and water. In this review, advancements of adsorptive purification of contaminated water (with especially organics, chemical structures of some described organics in this work are shown as Fig. 8.2) by using MOFs will be discussed. More importantly, the contribution of functional groups introduced onto isostructural or analogous MOFs on adsorptive purification will be the main topic of this review. To the best of our knowledge, there has been no review on the purification of water that focused on the role of functional groups of adsorbents on MOFs even though there have been several review articles, including ours, on possible purification of water via adsorption with MOFs during last few years [7–9, 16–19]. Moreover, a method on how to functionalize MOFs, for a special adsorption/purification, will also be discussed very briefly. Finally, there will be a prospect on the field, especially a direction for new research and for potential application. Before a detailed description, some introduction will be shown as below. Firstly, MOFs will be introduced. Earlier, porous materials were composed of mainly inorganics or carbon [10, 11]. Or, silica, alumina, carbon, zeolite, aluminophosphate, and mesoporous material have been typical or traditional porous materials that were usually applied in adsorption. However, because of the development of MOFs, porous materials composed of both inorganic and organic components have been advanced. In this review, MOFs itself will not be discussed since there are several good review articles on MOFs [12–15]. However, it should be emphasized that MOFs are versatile in modification [20, 21], compared with inorganic porous materials, which is understandable considering the fact that organic materials have
8 Adsorptive Purification of Water Contaminated with Hazardous Organics by. . .
271
N SO3-Na+
N N N+Cl-
S
N
N
Methyl orange
Methylene blue O
Cl OH
O
O H2N
HO
As
N+
O -
OH
p-Nitrophenol
p-Arsanilic acid
Cl
2,4-Dichlorophenoxyacetic acid
-
OH
O
+
O Na
O
O
Cl H N
HO O
OH
Cl
O
O Cl
Clofibric acid
Diclofenac sodium
Naproxen
Fig. 8.2 Chemical structures of some adsorbates that were frequently mentioned in this work. The first row shows dye molecules; the second row illustrates herbicide (2,4-D), feed additive (p-arsanilic acid), and industrial chemical (p-nitrophenol). The third row shows PPCPs such as clofibric acid (also can be an herbicide), naproxen, and diclofenac sodium
been widely/facilely modified because of possible functionalization or well-known organic chemistry. Secondly, the adsorptive purification of water will be introduced. Purification of contaminated water has been investigated for years, and adsorption is regarded as a competitive method. Adsorptive purification has been advanced very rapidly, partly because of the development of MOFs. This might be owing to huge porosity and tunable pore size and functionality. Considering the informative review articles [6– 9], a detailed description of the previous arts on adsorptive purification of water with MOFs will not be done in this review. Again, the purpose of the review is described below. In this review, adsorptive purification of water, contaminated with hazardous organics, by using MOFs will be discussed. In more detail, we will discuss (i) functionalized MOFs with a very brief description of preparation methods, (ii) adsorptive purification with functionalized MOFs (especially -NH2, -OH, -COOH, and -SO3H groups), and (iii) perspectives of the relevant field. Especially, in this work, the effect of functional groups on the adsorption will be dealt in order to develop an effective adsorbent for water purification.
272
8.2 8.2.1
D. K. Yoo et al.
Discussion Introduction to Functionalized MOFs
Some of the MOFs are isomorphous, isostructural, or isoreticular in topologies having nearly similar or analogous framework structures [22, 23]. Therefore, MOFs, especially analogous MOFs [23], have been very effective to study the effect of central metal ions and linkers (including different lengths and flexibility) on physical properties and adsorption. Similarly, the effect of functional groups can be estimated by using MOFs with or without a functional group [17]. Because of the possible synthesis of analogous MOFs with various linkers and facile modification of MOFs (especially because of the presence of organic components in MOFs), MOFs having various functional groups can be obtained facilely via either direct synthesis [22] or post-synthetic modification (PSM) [20]. For example, MOFs such as UiO-66(Zr) [24] and MIL-53(Fe) [25] can be prepared from anions of functionalized terephthalic acids or benzenedicarboxylic acids (TPAs or BDCAs), which are shown in Fig. 8.3. Direct synthesis of functionalized MOFs has the advantage of simple process; however, functionalized linkers are usually expensive, and porosity of functionalized MOFs is generally lower than that of the pristine MOFs without a functional group on linkers. Moreover, many MOFs cannot be synthesized with high crystallinity when functionalized linkers were applied in direct synthesis (on the contrary, only functionalized MOF can be easily/facilely synthesized in some cases like CAU-1(Al) [26]). In this review, methods to introduce some functional groups like -NH2, -OH, and -COOH will be very briefly discussed, in each section, before applications in adsorption.
8.2.2
Mechanism of Adsorptive Purification
Recently, studies on adsorption mechanism have been rapidly advanced owing to fundamental researches with various adsorbents/adsorbates in wide conditions. Understanding the plausible mechanisms for adsorption will be very fruitful not
Fig. 8.3 Functionalized BDC linkers that can be used in the synthesis of analogous MOFs like UiO-66(Zr) and MIL-53(Fe)s
8 Adsorptive Purification of Water Contaminated with Hazardous Organics by. . .
273
only for effective adsorption/removal of hazardous materials but also for selective interactions for controlled release such as drug delivery system. Potential mechanisms for adsorption, especially in aqueous solution, were reviewed well in open literatures [17, 18]; therefore, in this work, typical mechanisms will be discussed very briefly just for ready understanding of this review.
8.2.2.1
Electrostatic Interaction
The status of both adsorbent and adsorbate can be changed much with the pH of the solution applied for adsorption. For example, organics with -NH2 and -COOH can be protonated (under low pH) and deprotonated (under high pH), respectively. Therefore, there should be electrostatic interaction (ESI) (attraction between opposite charges or repulsion between the same charges) depending on the surface charge of adsorbent and status of adsorbate. Conclusively, several adsorptions to remove organics from water have been frequently explained with ESI mechanism [27, 28].
8.2.2.2
H-Bonding Interaction
Hazardous organics generally have ample polar functional groups such as -NH2, -COOH, -C¼O, -OH, and so on; therefore, it can make H-bonding when an adsorbent also has functional groups that have H-donor or H-acceptor sites. Or, several adsorbates interacted with MOFs via H-bonding mechanism [29]. However, it can be emphasized that the direction of H-bonding (H-donor and H-acceptor) is important to fully understand an adsorption mechanism of H-bonding.
8.2.2.3
Π-Interactions
Species with ample π-electrons (π-systems, such as aromatic rings) can interact with cation and another π-electron-containing species via cation-π and π-π interactions, respectively. Or, cation-π interaction is an interactive process between a cation and a π-system; and π-π interaction is an interaction between electron-rich π-system and electron-lean π-system. Both cation-π interaction [30, 31] and π-π interaction [32– 35] mechanisms have been used to explain adsorptions in a wide range of conditions.
8.2.2.4
Other Mechanisms
Other mechanisms such as hydrophobic interaction, coordination on open metal site (OMS) or coordinatively unsaturated site (CUS) (or, Lewis acid-base interaction), contribution of breathing effect, size exclusion (pore size effect), and influence of framework metals have been utilized to explain some adsorptive removals with MOFs [17, 18]. However, these mechanisms are not discussed here since those
274
D. K. Yoo et al.
adsorptions are based on MOF itself, and those are not very closely dependent on functional groups (especially organic ones).
8.2.3
Contribution of Functional Groups on Adsorption
As mentioned earlier, functional groups can be introduced onto MOFs via direct synthesis and PSM. Considering ample functional groups on adsorbates or hazardous organics in water, functional groups on MOFs might have preferential interactions with hazardous adsorbates. Therefore, such interactions will be effective in selective adsorption, removal, or storage/delivery. In this chapter, the contribution of functional groups (such as -NH2, -OH, and -COOH) loaded onto MOFs in the adsorptive removal of organics from contaminated water will be discussed, based on the type of functional groups.
8.2.3.1
Functional Group of -NH2 or -NH-
Amino groups such as -NH2 and -NH- are one of the most common functional groups that can be introduced onto MOFs. Various MOFs such as CAU-1(Al), IRMOF-3(Zn), MIL-53(Al, Ce, Fe)-NH2, MIL-88(Al, Fe, or Sc)-NH2, MIL-101 (Al)-NH2, MIL-125(Ti)-NH2, UiO-66(Zr)-NH2, and UMCM-1(Zn)-NH2 have been synthesized from amine-containing linker (or, amino-benzenedicarboxylate or BDC-NH2) in order to have amine functionality on the MOFs [36]. CuBTC-NH2 (BTC: 1,3,5-benzenetricarboxylate) could be also prepared from 2-amino-1,3,5benzenetricarboxylate (BTC-NH2) [37]. Bio-MOF-1(Zn), Zn8(adenine)4(biphenyldicarboxylate)6O, was synthesized from adenine and biphenyl-4,40 -dicarboxylic acid [38]. Moreover, amino functionality could be introduced onto MOFs via grafting diamines like ethylenediamine (EDA), diethylenetriamine (DETA), triethylenetetramine (TETA), tetraethylenepentamine (TEPA), etc., when MOFs have open metal sites (OMSs) or coordinatively unsaturated sites [39, 40]. Amino-functionalized MOFs have been actively applied in water purification or adsorption of organics from water via H-bonding or ESI interactions. This is understandable considering ample sites for H-bonding (N: H-acceptor and H: H-donor) and facile protonation of -NH2 to form -NH3+ even though the condition for protonation is dependent on the basicity of amino groups. Li et al. [41], as shown in Fig. 8.4, illustrated that amino-MIL-53(Al) was very effective, compared with pristine MIL-53(Al), in the adsorption of cationic dyes (methylene blue (MB) and malachite green (MG)) with -N(CH3)2 group. They could suggest H-bonding, between H of NH2 (of MOF) and N of -N(CH3)2 (in dyes) and π-π interaction, for the effective adsorption. Similarly, UiO-66(Zr)-NH2 was applied to remove methyl orange (MO) and MB from water [42]. The uptake was dependent much on the pH of the solution (Fig. 8.5). Firstly, the favorable adsorption of MO, especially under low pH, could be explained with H-bonding where protonated NH2 of MOF and MO
8 Adsorptive Purification of Water Contaminated with Hazardous Organics by. . .
275
Fig. 8.4 Effect of the contact time on the adsorption of MB (a) and MG (b) on MIL-53(Al)-NH2 and MIL-53(Al) at 20 C. (Reproduced with permission from Ref. [41]. Copyright 2015 American Chemical Society)
Fig. 8.5 Effect of solution pH on the MB and MO adsorption over UiO-66-NH2@SiO2. (Reproduced with permission from Ref. [42]. Copyright 2019 American Chemical Society)
were H-donor and H-acceptor, respectively. The pH dependence suggested the contribution of ESI, based on the charge of dyes (anionic MO and cationic MB). Additionally, it was suggested that the chemistry of MOF nanocage could be controlled easily because of the presence of -NH2 group (or depending on protonation to form -NH3+). Three cationic dyes (basic red 46 (BR46), basic blue 41 (BB41), and MB) were adsorbed over MIL-125(Ti)s with different contents of -NH2 group [43]. MIL-125(Ti)-NH2 had higher adsorption capacity than pristine MIL-125(Ti), and the adsorption could also be explained with H-bonding and ESI
276
D. K. Yoo et al.
Fig. 8.6 Adsorption isotherms of pNP over MOFs and activated carbon at 303 K. Inset shows a plausible mechanism (H-bonding) to explain the favorable pNP adsorption on the amino group. (Reproduced with permission from Ref. [49]. Copyright 2014 American Chemical Society)
mechanisms. MIL-125(Ti)-NH2 showed lower electron density than MIL-125(Ti); therefore, it could be more effective in interaction with cationic dyes. UiO-66(Zr)-NH2 had similarly higher adsorption for cationic dyes (like MB) than pristine UiO-66(Zr); and this observation also could be explained with negative zeta potential (the zeta potential of UiO-66(Zr) and UiO-66(Zr)-NH2 are about 3.65 and 4.91 mV, respectively) or favorable ESI [44]. The authors suggested negative zeta potential (or negative surface charge) with the fact that the inside of MOF is positive because of -NH3+ and the surface is negative for charge balance. Adsorption of four dyes (MB, MG, MO, and indigo blue (IB)) were also done with MIL-125(Ti) and MIL-125(Ti)-NH2 [45]. Because of H-bonding, MIL-125 (Ti)-NH2 showed higher adsorption compared with pristine MIL-125(Ti). MIL-101 (Al)-NH2 similarly showed relatively high adsorption of MG and indigo carmine because of ESI and H-bonding, together with π-π stacking [46]. In another study, UiO-66(Zr)-NH2 [47] and MIL-101(Al)-NH2 [48] showed better performance than pristine MOFs for MB removal even though there was little detailed explanation on the contribution of -NH2 group on adsorption. Liu et al. used three MOFs such as MIL-100(Fe, Cr) and MIL-101(Al)-NH2 in water purification or removal of p-nitrophenol and phenol [49]. They found that the three MOFs showed comparable adsorption capacity for phenol; however, interestingly, only MIL-101(Al)-NH2 was very effective in adsorption of pNP (Fig. 8.6). The authors could explain the extraordinary performance of the amino-MOF in pNP adsorption with H-bonding (Fig. 8.6, inset).
8 Adsorptive Purification of Water Contaminated with Hazardous Organics by. . .
277
Removal of naphthalenesulfonic acids such as 1,5-naphthalenedisulfonic acid (NDS) and 2-naphthalenesulfonic acid (NSA) was attempted with MIL-101(Cr), MIL-101(Cr)-NH2, and MIL-101(Cr)-SO3H [50]. Because of -NH2 or H-bonding between H of -NH2 (of MOF) and O of -SO3 (of naphthalenesulfonic acids), adsorption could be improved by using MIL-101(Cr)-NH2. Moreover, removal via ESI could also be increased with -NH2 introduction (which increased zeta potential). However, MIL-101(Cr)-SO3H had poor performance owing to ES repulsion between adsorbates and MIL-101(Cr)-SO3H (even with the favorable contribution of H-bonding). Similar mechanisms (H-bonding and ESI) could be utilized in the explanation of favorable adsorption of diclofenac sodium (DCFS) over UiO-66(Zr)NH2 [51]. UiO-66(Zr)-NH2, compared with pristine UiO-66(Zr), could be more efficient for ESI (with DCF anion) because of higher pHpzc (~7.6), compared with that of UiO-66(Zr) (7.4). However, the change in surface charge of these two studies [50, 51] might be different from those in earlier studies [43, 44], which needs further study. Moreover, higher DCFS adsorption over UiO-66(Zr)-NH2 also could be supported with H-bonding mechanism, where UiO-66(Zr)-NH2 and DCFS were H-donor and H-acceptor, respectively. MIL-68(In)-NH2 was applied in the adsorption of p-arsanilic acid (p-ASA) [52]. Compared with pristine MIL-68(In), the functionalized MOF showed increased adsorption because of H-bonding in a few interactions like (i) N of NH2 from MOF and H of NH2 of p-ASA and (ii) H of NH2 from MOF and O of p-ASA. Adsorption of p-ASA was similarly carried out with UiO-67(Zr)s, prepared from biphenyl4,40 -dicarboxylic acids (BPDCs) with 0–2 NH2 groups [53]. Even though aminated UiO-67(Zr)s were not very effective in adsorption of p-ASA (because of low porosity of UiO-67(Zr) with amino group), the maximum adsorption capacity, if estimated by unit surface area of UiO-67(Zr)s, followed the order: UiO-67 (Zr) < UiO-67(Zr)-NH2 < UiO-67(Zr)-(NH2)2. As-O-Zr coordination, π-π stacking, and H-bonding (NH∙∙∙O) could be suggested as plausible mechanisms of ASA adsorption (Fig. 8.7). Similar to other works, -NH2 group increased the adsorption because of H-bonding. Zhao et al. [54] prepared several UiO-66(Zr)s, with various functional groups such as -Br, -CH3, -COOH, -OH, -NH2, -NO2, and -SO3H, by direct synthesis using functionalized linkers. They applied the MOFs in adsorption/removal of 2,4-dichlorophenoxyacetic acid (2,4-D), clofibric acid, and DCFS. The improved adsorptions over UiO-66(Zr)-NH2 could be explained with H-bonding (because of extra -NH2 group) and other mechanisms like ESI and π-π interaction. Seo et al. converted MIL-101(Cr) into MIL-101(Cr)-NH2 via oxidation (to introduce NO2 on the phenyl ring of the linker) and subsequent reduction and showed that -NH2 was effective (even less effective than –(OH)2, vide infra) in naproxen adsorption [55]. The contribution of -NH2 group in naproxen adsorption could be explained with H-bonding, where -NH2 was H-donor and naproxen was H-acceptor. Functionalized MIL-101(Cr) with amino group, introduced via grafting, has been also used in adsorptive purification of water contaminated with naproxen and clofibric acid [56]. Compared with pristine MIL-101(Cr), MIL-101(Cr)-NH2 had
278
D. K. Yoo et al.
Fig. 8.7 p-ASA adsorption mechanism over UiO-67-NH2. (Reproduced with permission from Ref. [53]. Copyright 2018 American Chemical Society)
slightly improved adsorption; however, MIL-101(Cr)-SO3H, prepared by grafting amino methanesulfonic acid on OMS of MIL-101(Cr), showed very poor adsorption of the two adsorbates. Similar amino-functionalized MOF (or, EDA-grafted MIL-101(Cr)) was applied in capturing antibiotics oxytetracycline (OTC) from water [57]. The functionalized MIL-101(Cr) showed much higher adsorption capacity than the pristine MIL-101(Cr) because of favorable ESI especially at pH 4–10. The effect of pH on adsorbed quantity could be explained with the zeta potential of the MOF and status of OTC (protonated, zwitterionic, and deprotonated OTCs). Not only ethylenediamine but also 1,2-bis(3-aminopropylamino)ethane was coordinated on MIL-101(Fe) to check the effect of number of amino groups on the release of naproxen; and the release rate of naproxen could be controlled by the number of amino groups under pH 7.4, different from the release under pH 2 (where there was no effect of amino groups) [58]. Even though the adsorption capacity of naproxen was not reported in the paper, the MIL-101(Fe)s might have considerable capacity since the huge porosity was decreased very much after naproxen loading [58]. So far, MOFs with a basic site on the linker itself or ring of the linker were not very common. Razavi et al. synthesized TMU-34(Zn) by using 3,6-di(pyridin-4-yl)1,4-dihydro-1,2,4,5-tetrazine as a pillar; and the MOF was applied in the removal of rose-bengal B (RB-B) dye [59]. Because of host-guest interaction or H-bonding shown in Fig. 8.8 (MOF was H-donor and anion form of RB-R was H-acceptor), TMU-34(Zn) was highly effective in RB-B removal than isostructural TMU-4(Zn) (without H-donor site). Zhu et al. applied {[Tb-(TATMA)(H2O)2H2O}n (TATMA: 4,40 ,400 -s-triazine-1,3,5- triyl tri-m-aminobenzoate) in luminescence detection of nitrofurantoin (NFT) and nitrofurazone (NFZ) [60]. Weak intermolecular H-bonding interactions between the MOF and NFT or NFZ were suggested because of the presence of various H-bonding donors and acceptors.
8 Adsorptive Purification of Water Contaminated with Hazardous Organics by. . .
279
Fig. 8.8 Possible hostguest interactions between TMU-34(Zn) and RB-B. (Reproduced with permission from Ref. [59]. Copyright 2018 American Chemical Society)
Fig. 8.9 Plausible mechanism (H-bonding) to explain favorable adsorption of artificial sweeteners over –NH2 (of the amide of MOF). (Reproduced with permission from Ref. [61]. Copyright 2016 American Chemical Society)
Functional groups of -NH2 in amide, obtained via grafting urea and melamine onto MIL-101(Cr), were also effective in the removal of artificial sweeteners [61] and nitroimidazole antibiotics [62] from water even though such amino group is not basic. The remarkable performances are mainly because of H-bonding (especially via stable 6-membered rings, as shown in Fig. 8.9, between artificial sweeteners and the -NH2 group). Finally, basicity of amino groups should be considered in adsorption, especially in aqueous phase solutions since the pH where protonation of basic sites occurs is dependent on the basicity of amino groups. The pKaH values of protonated aliphatic amines (like triethylamine) and aromatic amines (such as aniline) are generally 10–11 and 4–5, respectively. Therefore, aliphatic and aromatic amines can be protonated under pH < 10–11 and < 4–5, respectively, and might be effective in ES attraction with organics having anionic charge (such as anionic dyes like MO and acid red 1) under such condition.
280
8.2.3.2
D. K. Yoo et al.
Functional Group of -OH
Similar to amino group, hydroxyl group can be introduced onto MOFs via a few methods like direct synthesis (by using linkers having hydroxyl group) and PSM (including method of grafting aminoalcohols on OMSs). For example, MOFs with hydroxyl groups (such as MOF-74 s, UiO-66(Zr)-(OH)2 [63–65], MIL-53(Al)(OH)2 [66], and CAU-1(Al)-(OH)2 [26]), especially phenolic hydroxyl groups, can be prepared from functionalized linker like 2,5-dihydroxyterephthalic acid. Amines with hydroxyl group can be grafted onto OMSs of MOFs; or, grafting might be an effective way to load hydroxyl group, especially with no acidity, on MOFs with OMSs [55, 67]. Moreover, hydroxyl groups, especially bridged one, are frequently observed in some MOFs. For example, MOFs, such as MIL-53 s, CAU-1(Al), PCN-222(Zr), UiO-66(Zr), MOF-808(Zr), and NU-1000(Zr) (different from MOF-5(Zn), MOF-177(Zn) and Cu-BTC), have hydroxyl groups on the MOFs. Introduced hydroxyl groups can be applied in water purification, via adsorption, mainly because of polarity of -OH group. Hydroxyl group, similar to amino group, can be very useful especially for H-bonding (as both H-acceptor and H-donor (excluding phenolic hydroxyl group at very high pH, where phenolate is usually observed)). Actually, Song and Jhung showed that PPCP adsorption increased linearly with an increasing number of H-acceptor in PPCPs and with an increasing number of -OH groups (these -OH groups acted as H-donor) on MOF or MIL-101 (Cr) [67]. Seo et al. showed similar results by using MIL-101(Cr) having OH and -NH2 groups as H-donors [55]. Likewise, organic arsenic acids (OAAs) such as phenylarsonic acid (PAA) and p-ASA could be removed effectively from water by using MIL-101(Cr) having three hydroxyl groups (MIL-101(Cr)-(OH)3) [68]. For the adsorption of OAAs, the adsorbent acted as H-donor because of ample -OH groups. MOFs with phenolic hydroxylic group might be also very useful as adsorbent considering the attractive -OH groups on the benzene ring, for example. Recently, Ma et al. utilized a Tb-MOF, with hydroxyl group on the benzene ring, in adsorptive removal of catechol [69]. The competitive adsorption could be explained with appropriate pore diameter and H-bonding (between -OH groups on the MOF and catechol). Zhao et al. [54] applied UiO-66(Zr)-OH (similar to UiO-66(Zr)-NH2) in adsorption/removal of 2,4-D, clofibric acid, and DCFS. Even though the adsorption over UiO-66(Zr)-OH is not very remarkable, compared with UiO-66(Zr)-NH2, H-bonding (because of extra -OH group) was also suggested as a mechanism for interaction between the adsorbates and UiO-66(Zr)-OH. Curiously, however, there has been not much application of MOFs with phenolic hydroxylic group on linker or benzene ring (such as MOF-74 s, UiO-66(Zr)-(OH)2, MIL-53(Al)-(OH)2, and CAU-1(Al)-(OH)2)) in water purification even though there are ample H-acceptor or H-donor sites on MOFs and H-bonding has been very effective in adsorption/removal of organics from water. This might be due to (i) relatively unstable structure of, for example, MOF-74 s [70, 71] in water, and (ii) generally not high porosity of MOFs with functional groups of -(OH)2.
8 Adsorptive Purification of Water Contaminated with Hazardous Organics by. . .
281
The hydroxyl groups that are originally present on MOFs also have been utilized in purification of water contaminated with organics such as bisphenol-A [72], dimetridazole [73], chloramphenicol [74], organophosphorus pesticides [75], and tinidazole [76]; and the observed high adsorption capacities were explained with mainly H-bonding. MOFs with ample hydroxyl groups, if stable in water, will be very attractive in water purification, based on no need of expensive linkers or of tedious PSMs. Finally, the acidity of -OH should be considered especially when such hydroxyl group is applied as a H-donor. Aliphatic alcohols do not have any acidity; therefore, very stable under a wide range of pH conditions. However, aromatic alcohols such as phenolic hydroxyl groups can be deprotonated at high pH (> 10) because of weak acidity. Therefore, such -OH can be active as H-donor only when pH ~ 4.5), both from direct synthesis and PSM [86, 87], should be considered since property of carboxylate is very much different
282
D. K. Yoo et al.
UiO-66(Zr)-NH2
UiO-66(Zr)-NH-CO-COOH
Fig. 8.10 Scheme to prepare UiO-66(Zr)-NH-(C¼O)-COOH via a reaction between UiO-66(Zr)NH2 and oxalyl chloride. (Redrawn to follow a reported work [81])
from that of free -COOH group. Moreover, at pH > ~4.5, -COOH is converted into -COO which is very effective in ESI and H-bonding (as H-acceptor).
8.2.3.4
Functional Group of -SO3H
MOFs functionalized with -SO3H could be obtained similarly via direct synthesis [88] or PSM [89, 90]. The obtained MOFs with -SO3H were applied mainly in catalysis because of the strong acid site of -SO3H [89, 90]. However, MOFs with -SO3H group have also been applied in liquid phase adsorption to purify water, for example. MIL-101(Cr)-SO3H and MIL-101(Cr) showed preferential adsorption for cationic (safranine T) and anionic (fluorescein sodium) dyes, respectively [91]. The sharp difference in adsorption selectivity was because of ES interactions. Or, because of -SO3, MIL-101(Cr)-SO3H was effective in capturing cationic dye. On the other hand, fluorescein sodium could be adsorbed effectively over pristine MIL-101(Cr) because of a positive charge. Similarly, MIL-101(Cr)-SO3H was effective in capturing cationic dyes such as MB and MG because of ESI between cationic dyes and -SO3 group of the MOF [92]. Hasan et al. showed that UiO-66 (Zr)-SO3H, synthesized directly from TPA-SO3Na, could adsorb effectively DCFS, compared with UiO-66(Zr) and UiO-66(Zr)-NH2 [93]. UiO-66(Zr)-SO3H [54] and MIL-101(Cr)-SO3H [50], obtained from functionalized linkers, were applied in adsorption of pharmaceutical wastes and naphthalenesulfonic acid, respectively. However, the performances were poor compared with other analogous MOFs including pristine ones, because of, for example, ES repulsion. Considering the strong acidity of sulfonic acid, MOF-SO3H exists usually in a deprotonated form in aqueous solutions. Therefore, ESI should be considered at first, and such -SO3 will not be effective as H-donor (even though such group is very good as H-acceptor).
8 Adsorptive Purification of Water Contaminated with Hazardous Organics by. . .
8.2.3.5
283
Other Functional Groups
Other functional groups such as -NO2, alkyl, phenyl, or aromatic groups might be also useful, if introduced onto MOFs, in adsorptive purification of water. The functional group of -NO2 can be good H-acceptor site; and alkyl (or, phenyl or aromatic) group might be effective in hydrophobic interaction, especially when long alkyl chain or polyaromatic group is present [94, 95]. Moreover, phenyl or aromatic rings might be useful for π-π and hydrophobic interactions. So far, curiously, there has been little application of such functionalized MOFs in water purification. Finally, functional groups with charge (like -NH3+) have been applied in the removal of perfluorooctanoic acid (PFOA) [96] and MO [97] (also adsorption of water vapor [98]), suggesting the favorable contribution of the charged group in adsorptions. Cationic sites might be useful in removal of organics having aromatic rings or anionic charges via cation-π and electrostatic interactions, respectively.
8.3
Conclusions and Perspective
In this review, we discussed the contribution of functional groups loaded onto MOFs in adsorption of hazardous organics from water; and we can conclude that adsorptive purification of water by using MOFs will be a very competitive mean especially when MOFs are functionalized adequately. H-bonding and ESI have been the most frequently observed mechanism to explain the adsorption/removal of organics from water over functionalized MOFs. Based on the discussion, the following can be suggested as prospects in the relevant field. First, the idea that applied in water purification will be applied in related fields such as fuel purification [99], storage/delivery (drugs or perfumes) [100], and enrichment of analytes (for analytical chemistry) [101]. Second, considering the expensive MOFs, especially functionalized linkers, utilization of cheap raw materials in synthesis of functionalized MOFs will be very important. Moreover, recovery and reuse of both spent MOFs and removed organics should be considered since (i) spent MOFs may cause secondary pollution especially when MOFs are composed of heavy/toxic metallic species like Cr and Cd and (ii) removed organics can be useful resources if treated/purified well. Third, the condition of adsorption should be relevant to industrial conditions. For example, the concentration of contaminants in water is usually very low; therefore, the adsorption should be done in such concentration if considered in industrial application (and the maximum adsorption capacity is not very meaningful in some cases, even though the maximum capacity is useful to understand an adsorption or storage). Effects of competing species such as humic acid and ionic species should be also considered, based on the possible competitive adsorption of such materials. Fourth, the importance of solution pH also should be considered since the pH of solution has remarkable effect on the adsorption. Even though many studies were done in natural condition (e.g., by just
284
D. K. Yoo et al.
dissolution of organics in water or water/alcohol), such condition might be far from a usual industrial condition. If there is no detailed guideline on solution pH, neutral condition or pH ~7 might be one option, based on the general pH of river or rainwater [102]. Finally, MOFs with various functionalized linkers are available, especially in recent days [103]; therefore, studies to utilize such functionalities will be very meaningful not only for basic researches but also for possible applications. Acknowledgement This work was supported by Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Science, ICT and future Planning (grant number: 2017R1A2B2008774).
References 1. Gao Q, Xu J, Bu XH (2019) Recent advances about metal–organic frameworks in the removal of pollutants from wastewater. Coord. Chem. Rev. 378:17–31. https://doi.org/10.1016/j.ccr. 2018.03.015 2. Li X, Wang B, Cao Y, Zhao S, Wang H, Feng X, Zhou J, Ma X (2019) Water contaminant elimination based on metal-organic frameworks and perspective on their industrial applications. ACS Sustain. Chem. Eng. 7:4548–4563. https://doi.org/10.1021/acssuschemeng. 8b05751 3. Li J, Wang H, Yuan X, Zhang J, Chew JW (2020) Metal-organic framework membranes for wastewater treatment and water regeneration. Coord. Chem. Rev. 404:213116. https://doi.org/ 10.1016/j.ccr.2019.213116 4. Mon M, Bruno R, Ferrando-Soria J, Armentano D, Pardo E (2018) Metal-organic framework technologies for water remediation: towards a sustainable ecosystem. J. Mater. Chem. A 6:4912–4947. https://doi.org/10.1039/C8TA00264A 5. Rojas S, Horcajada P (2020) Metal–organic frameworks for the removal of emerging organic contaminants in water. Chem. Rev. 120:8378–8415. https://doi.org/10.1021/acs.chemrev. 9b00797 6. Jiang D, Chen M, Wang H, Zeng G, Huang D, Cheng M, Liu Y, Xue W, Wang Z (2019) The application of different typological and structural MOFs-based materials for the dyes adsorption. Coord. Chem. Rev. 380:471–483. https://doi.org/10.1016/j.ccr.2018.11.002 7. Dhaka S, Kumar R, Deep A, Kurade MB, Ji SW, Jeon BH (2019) Metal–organic frameworks (MOFs) for the removal of emerging contaminants from aquatic environments. Coord. Chem. Rev. 380:330–352. https://doi.org/10.1016/j.ccr.2018.10.003 8. Gusain R, Kumar N, Sinha Ray S (2020) Recent advances in carbon nanomaterial-based adsorbents for water purification. Coord. Chem. Rev. 405:213111. https://doi.org/10.1016/j. ccr.2019.213111 9. Drout RJ, Robison L, Chen Z, Islamoglu T, Farha OK (2019) Zirconium metal–organic frameworks for organic pollutant adsorption. Trends Chem 1:304–317. https://doi.org/10. 1016/j.trechm.2019.03.010 10. Nugent P, Belmabkhout Y, Burd SD, Cairns AJ, Luebke R, Forrest K, Pham T, Ma S, Space B, Wojtas L, Eddaoudi M, Zaworotko MJ (2013) Porous materials with optimal adsorption thermodynamics and kinetics for CO2 separation. Nature 495:80–84. https://doi.org/10. 1038/nature11893 11. Slater AG, Cooper AI (2015) Function-led design of new porous materials. Science 348:988–1008. https://doi.org/10.1126/science.aaa8075
8 Adsorptive Purification of Water Contaminated with Hazardous Organics by. . .
285
12. Kirchon A, Feng L, Drake HF, Joseph EA, Zhou HC (2018) From fundamentals to applications: a toolbox for robust and multifunctional MOF materials. Chem. Soc. Rev. 47:8611–8638. https://doi.org/10.1039/C8CS00688A 13. Adil K, Belmabkhout Y, Pillai RS, Cadiau A, Bhatt PM, Assen AH, Maurin G, Eddaoudi M (2017) Gas/vapour separation using ultra-microporous metal–organic frameworks: insights into the structure/separation relationship. Chem. Soc. Rev. 46:3402–3430. https://doi.org/10. 1039/C7CS00153C 14. Silva P, Vilela SMF, Tome JPC, Paz FAA (2015) Multifunctional metal–organic frameworks: from academia to industrial applications. Chem. Soc. Rev. 44:6774–6803. https://doi.org/10. 1039/C5CS00307E 15. Li J, Wang X, Zhao G, Chen C, Chai Z, Alsaedi A, Hayat T, Wang X (2018) Metal–organic framework-based materials: superior adsorbents for the capture of toxic and radioactive metal ions. Chem. Soc. Rev. 47:2322–2356. https://doi.org/10.1039/C7CS00543A 16. Khan NA, Hasan Z, Jhung SH (2013) Adsorptive removal of hazardous materials using metalorganic frameworks (MOFs): a review. J. Hazard. Mater. 244–245:444–456. https://doi.org/ 10.1016/j.jhazmat.2012.11.011 17. Hasan Z, Jhung SH (2015) Removal of hazardous organics from water using metal-organic frameworks (MOFs): plausible mechanisms for selective adsorptions. J. Hazard. Mater. 283:329–339. https://doi.org/10.1016/j.jhazmat.2014.09.046 18. Kim S, Chu KH, Al-Hamadani YAJ, Park CM, Jang M, Kim DH, Yu M, Heo J, Yoon Y (2018) Removal of contaminants of emerging concern by membranes in water and wastewater: a review. Chem. Eng. J. 335:896–914. https://doi.org/10.1016/j.cej.2017.11.044 19. Dias EM, Petit C (2015) Towards the use of metal–organic frameworks for water reuse: a review of the recent advances in the field of organic pollutants removal and degradation and the next steps in the field. J. Mater. Chem. A 3:22484–22506. https://doi.org/10.1039/ C5TA05440K 20. Cohen SM (2012) Postsynthetic methods for the functionalization of metal–organic frameworks. Chem. Rev. 112:970–1000. https://doi.org/10.1021/cr200179u 21. Yin Z, Wan S, Yang J, Kurmoo M, Zeng MH (2019) Recent advances in post-synthetic modification of metal–organic frameworks: new types and tandem reactions. Coord. Chem. Rev. 378:500–512. https://doi.org/10.1016/j.ccr.2017.11.015 22. Eddaoudi M, Kim J, Rosi N, Vodak D, Wachter J, O’Keeffe M, Yaghi OM (2002) Systematic design of pore size and functionality in isoreticular MOFs and their application in methane storage. Science 295:469–472. https://doi.org/10.1126/science.1067208 23. Jhung SH, Khan NA, Hasan Z (2012) Analogous porous metal–organic frameworks: synthesis, stability and application in adsorption. CrystEngComm 14:7099–7109. https://doi.org/10. 1039/C2CE25760B 24. Hu Z, Peng Y, Kang Z, Qian Y, Zhao D (2015) A modulated hydrothermal (MHT) approach for the facile synthesis of UiO-66-type MOFs. Inorg. Chem. 54:4862–4868. https://doi.org/10. 1021/acs.inorgchem.5b00435 25. Devic T, Horcajada P, Serre C, Salles F, Maurin G, Moulin B, Heurtaux D, Clet G, Vimont A, Greneche JM, Le-Ouay B, Moreau F, Magnier E, Filinchuk Y, Marrot J, Lavalley JC, Daturi M, Ferey G (2010) Functionalization in flexible porous solids: effects on the pore opening and the hostguest interactions. J. Am. Chem. Soc. 132:1127–1136. https://doi.org/ 10.1021/ja9092715 26. Ahnfeldt T, Stock N (2012) Synthesis of isoreticular CAU-1 compounds: effects of linker and heating methods on the kinetics of the synthesis. CrystEngComm 14:505–511. https://doi.org/ 10.1039/C1CE05956D 27. Khan NA, Jung BK, Hasan Z, Jhung SH (2015) Adsorption and removal of phthalic acid and diethyl phthalate from water with zeolitic imidazolate and metal–organic frameworks. J. Hazard. Mater. 282:194–200. https://doi.org/10.1016/j.jhazmat.2014.03.047
286
D. K. Yoo et al.
28. Seo YS, Khan NA, Jhung SH (2015) Adsorptive removal of methylchlorophenoxypropionic acid from water with a metal-organic framework. Chem. Eng. J. 270:22–27. https://doi.org/10. 1016/j.cej.2015.02.007 29. Ahmed I, Jhung SH (2017) Applications of metal-organic frameworks in adsorption/separation processes via hydrogen bonding interactions. Chem. Eng. J. 310:197–215. https://doi.org/ 10.1016/j.cej.2016.10.115 30. Ahmed I, Khan NA, Yoon JW, Chang JS, Jhung SH (2017) Protonated MIL-125-NH2: remarkable adsorbent for the removal of quinoline and indole from liquid fuel. ACS Appl. Mater. Interfaces 9:20938–20946. https://doi.org/10.1021/acsami.7b01899 31. Sarker M, Song JY, Jeong AR, Min KS, Jhung SH (2018) Adsorptive removal of indole and quinoline from model fuel using adenine-grafted metal-organic frameworks. J. Hazard. Mater. 344:593–601. https://doi.org/10.1016/j.jhazmat.2017.10.041 32. Akpinar I, Drout RJ, Islamoglu T, Kato S, Lyu J, Farha OK (2019) Exploiting π–π interactions to design an efficient sorbent for atrazine removal from water. ACS Appl. Mater. Interfaces 11:6097–6103. https://doi.org/10.1021/acsami.8b20355 33. Lin KYA, Chang HA (2015) Ultra-high adsorption capacity of zeolitic imidazoleframework67 (ZIF-67) for removal of malachite green from water. Chemosphere 139:624–631. https:// doi.org/10.1016/j.chemosphere.2015.01.041 34. Sarker M, Bhadra BN, Seo PW, Jhung SH Adsorption of benzotriazole and benzimidazole from water over a co-based metal azolate framework MAF-5(Co). J. Hazard. Mater. 324:131–138. https://doi.org/10.1016/j.jhazmat.2016.10.042 35. Kim TK, Lee JH, Moon D, Moon HR (2013) Luminescent Li-based metal–organic framework tailored for the selective detection of explosive nitroaromatic compounds: direct observation of interaction sites. Inorg. Chem. 52:589–595. https://doi.org/10.1021/ic3011458 36. Lin Y, Kong C, Chen L (2016) Amine-functionalized metal–organic frameworks: structure, synthesis and applications. RSC Adv. 6:32598–32614. https://doi.org/10.1039/C6RA01536K 37. Peikert K, Hoffmann F, Froba M (2012) Amino substituted Cu3(btc)2: a new metal–organic framework with a versatile functionality. Chem. Commun. 48:11196–11198. https://doi.org/ 10.1039/C2CC36220A 38. An J, Geib SJ, Rosi NL (2009) Cation-triggered drug release from a porous zincadeninate metalorganic framework. J. Am. Chem. Soc. 131:8376–8377. https://doi.org/10.1021/ ja902972w 39. Hwang YK, Hong DY, Chang JS, Jhung SH, Seo YK, Kim J, Vimont A, Daturi M, Serre C, Férey G (2008) Amine grafting on coordinatively unsaturated metal centers of MOFs: consequences for catalysis and metal encapsulation. Angew. Chem. Int. Ed. 47:4144–4148. https://doi.org/10.1002/anie.200705998 40. Hong DY, Hwang YK, Serre C, Férey G, Chang JS (2009) Porous chromium terephthalate MIL-101 with coordinatively unsaturated sites: surface functionalization, encapsulation, sorption and catalysis. Adv. Funct. Mater. 19:1537–1552. https://doi.org/10.1002/adfm. 200801130 41. Li C, Xiong Z, Zhang J, Wu C (2015) The strengthening role of the amino group in metal– organic framework MIL-53 (Al) for methylene blue and malachite green dye adsorption. J. Chem. Eng. Data 60:3414–3422. https://doi.org/10.1021/acs.jced.5b00692 42. Ibrahim AH, El-Mehalmey WA, Haikal RR, Safy MEA, Amin M, Shatla HR, Karakalos SG, Alkordi MH (2019) Tuning the chemical environment within the UiO-66-NH2 nanocages for charge-dependent contaminant uptake and selectivity. Inorg. Chem. 58:15078–15087. https:// doi.org/10.1021/acs.inorgchem.9b01611 43. Oveisi M, Asli MA, Mahmoodi NM (2018) MIL-Ti metal-organic frameworks (MOFs) nanomaterials as superior adsorbents: synthesis and ultrasound-aided dye adsorption from multicomponent wastewater systems. J. Hazard. Mater. 347:123–140. https://doi.org/10.1016/ j.jhazmat.2017.12.057
8 Adsorptive Purification of Water Contaminated with Hazardous Organics by. . .
287
44. Chen Q, He Q, Lv M, Xu Y, Yang H, Liu X, Wei F (2015) Selective adsorption of cationic dyes by UiO-66-NH2. Appl. Surf. Sci. 327:77–85. https://doi.org/10.1016/j.apsusc.2014.11. 103 45. Fan YH, Zhang SW, Qin SB, Li XS, Qia SH (2018) An enhanced adsorption of organic dyes onto NH2 functionalization titanium-based metal-organic frameworks and the mechanism investigation. Micro Meso Mater 263:120–127. https://doi.org/10.1016/j.micromeso.2017. 12.016 46. Liu H, Chen L, Ding J (2016) Adsorption behavior of magnetic aminofunctionalized metal– organic framework for cationic and anionic dyes from aqueous solution. RSC Adv. 6:48884–48895. https://doi.org/10.1039/c6ra07567c 47. Huang L, He M, Chen B, Hu B (2018) Magnetic Zr-MOFs nanocomposites for rapid removal of heavy metal ions and dyes from water. Chemosphere 199:435–444. https://doi.org/10.1016/ j.chemosphere.2018.02.019 48. Haque E, Lo V, Minett AI, Harris AT, Church TL (2014) Dichotomous adsorption behaviour of dyes on an amino-functionalised metal–organic framework, amino-MIL-101(Al). J. Mater. Chem. A 2:193–203. https://doi.org/10.1039/C3TA13589F 49. Liu B, Yang F, Zou Y, Peng Y (2014) Adsorption of phenol and p-nitrophenol from aqueous solutions on metal–organic frameworks: effect of hydrogen bonding. J. Chem. Eng. Data 59:1476–1482. https://doi.org/10.1021/je4010239 50. Zhao H, Zhao X, Gao Z, Liu D (2018) Effective removal of naphthalenesulfonic acid from water using functionalized metal–organic frameworks. J. Chem. Eng. Data 63:3061–3067. https://doi.org/10.1021/acs.jced.8b00318 51. Zhuang S, Cheng R, Wang J (2019) Adsorption of diclofenac from aqueous solution using UiO-66-type metal-organic frameworks. Chem. Eng. J. 359:354–362. https://doi.org/10.1016/ j.cej.2018.11.150 52. Lv Y, Zhang R, Zeng S, Liu K, Huang S, Liu Y, Xu P, Lin C, Cheng Y, Liu M (2018) Removal of p-arsanilic acid by an amino-functionalized indium-based metal–organic framework: adsorption behavior and synergetic mechanism. Chem. Eng. J. 339:359–368. https://doi.org/ 10.1016/j.cej.2018.01.139 53. Tian C, Zhao J, Ou X, Wan J, Cai Y, Lin Z, Dang Z, Xing B (2018) Enhanced adsorption of p-arsanilic acid from water by amine-modified UiO-67 as examined using extended x-ray absorption fine structure, x-ray photoelectron spectroscopy, and density functional theory calculations. Environ. Sci. Technol. 52:3466–3475. https://doi.org/10.1021/acs.est.7b05761 54. Zhao P, Liu N, Jin C, Chen H, Zhang Z, Zhao L, Cheng P, Chen Y (2019) UiO-66: an advanced platform for investigating the influence of functionalization in the adsorption removal of pharmaceutical waste. Inorg. Chem. 58:8787–8792. https://doi.org/10.1021/acs. inorgchem.9b01172 55. Seo PW, Bhadra BN, Ahmed I, Khan NA, Jhung SH (2016) Adsorptive removal of pharmaceuticals and personal care products from water with functionalized metal-organic frameworks: remarkable adsorbents with hydrogen-bonding abilities. Sci. Rep. 6:34462. https://doi. org/10.1038/srep34462 56. Hasan Z, Choi EJ, Jhung SH (2013) Adsorption of naproxen and clofibric acid over a metal– organic framework MIL-101 functionalized with acidic and basic groups. Chem. Eng. J. 219:537–544. https://doi.org/10.1016/j.cej.2013.01.002 57. Hu T, Jia Q, He S, Shan S, Su H, Zhi Y, He L (2017) Novel functionalized metal-organic framework MIL-101 adsorbent for capturing oxytetracycline. J. Alloys Compd. 727:114–122. https://doi.org/10.1016/j.jallcom.2017.08.116 58. Almáši M, Zeleňák V, Palotai P, Beňová E, Zeleňáková A (2018) Metal-organic framework MIL-101(Fe)-NH2 functionalized with different long-chain polyamines as drug delivery system. Inorg. Chem. Commun. 93:115–120. https://doi.org/10.1016/j.inoche.2018.05.007 59. Razavi SAA, Masoomi MY, Morsali A (2018) Host–guest interaction optimization through cavity functionalization for ultra-fast and efficient water purification by a metal–organic framework. Inorg. Chem. 57:11578–11587. https://doi.org/10.1021/acs.inorgchem.8b01611
288
D. K. Yoo et al.
60. Zhu QQ, He H, Yan Y, Yuan J, Lu DQ, Zhang DY, Sun F, Zhu G (2019) An exceptionally stable TbIII-based metal–organic framework for selectively and sensitively detecting antibiotics in aqueous solution. Inorg. Chem. 58:7746–7753. https://doi.org/10.1021/acs. inorgchem.9b00147 61. Seo PW, Khan NA, Hasan Z, Jhung SH (2016) Adsorptive removal of artificial sweeteners from water using metal–organic frameworks functionalized with urea or melamine. ACS Appl. Mater. Interfaces 8:29799–29807. https://doi.org/10.1021/acsami.6b11115 62. Seo PW, Khan NA, Jhung SH (2017) Removal of nitroimidazole antibiotics from water by adsorption over metal–organic frameworks modified with urea or melamine. Chem. Eng. J. 315:92–100. https://doi.org/10.1016/j.cej.2017.01.021 63. Rada ZH, Abid HR, Sun H, Wang S (2015) Bifunctionalized metal organic frameworks, UiO-66-NO2-N (N ¼ -NH2, (OH)2, (COOH)2), for enhanced adsorption and selectivity of CO2 and N2. J. Chem. Eng. Data 60:2152–2161. https://doi.org/10.1021/acs.jced.5b00229 64. Noh J, Kim Y, Park H, Lee J, Yoon M, Park MH, Kim Y, Kim M (2018) Functional group effects on a metal-organic framework catalyst for CO2 cycloaddition. J. Ind. Eng. Chem. 64:478–483. https://doi.org/10.1016/j.jiec.2018.04.010 65. Chen S, Liu J, Xu Y, Li Z, Wang T, Xu J, Wang Z (2018) Hydrogen storage properties of the novel crosslinked UiO-66-(OH)2. Int. J. Hydrog. Energy 43:15370–15377. https://doi.org/10. 1016/j.ijhydene.2018.06.106 66. Biswas S, Ahnfeldt T, Stock N (2011) New functionalized flexible Al-MIL-53-X (X ¼ -cl, -Br, -CH3, -NO2, (OH)2) solids: syntheses, characterization, sorption, and breathing behavior. Inorg. Chem. 50:9518–9526. https://doi.org/10.1021/ic201219g 67. Song JY, Jhung SH (2017) Adsorption of pharmaceuticals and personal care products over metal-organic frameworks functionalized with hydroxyl groups: quantitative analyses of H-bonding in adsorption. Chem. Eng. J. 322:366–374. https://doi.org/10.1016/j.cej.2017.04. 036 68. Sarker M, Song JY, Jhung SH (2017) Adsorption of organic arsenic acids from water over functionalized metal-organic frameworks. J. Hazard. Mater. 335:162–169. https://doi.org/10. 1016/j.jhazmat.2017.04.044 69. Ma DY, Zhang SY, Zhan SH, Feng LT, Zeng SG, Lin QQ, Pan Y (2019) Adsorptive removal of catechol from aqueous solution with a water-stable and hydroxyl-functionalized terbium– organic framework. Ind. Eng. Chem. Res. 58:20090–20098. https://doi.org/10.1021/acs.iecr. 9b05067 70. DeCoste JB, Peterson GW, Schindler BJ, Killops KL, Browe MA, Mahle JJ (2013) The effect of water adsorption on the structure of the carboxylate containing metal–organic frameworks cu-BTC, mg-MOF-74, and UiO-66. J. Mater. Chem. A 1:11922–11932. https://doi.org/10. 1039/C3TA12497E 71. Zuluaga S, Fuentes-Fernandez EMA, Tan K, Xu F, Li J, Chabal YJ, Thonhauser T (2016) Understanding and controlling water stability of MOF-74. J. Mater. Chem. A 4:5176–5183. https://doi.org/10.1039/C5TA10416E 72. Park EY, Hasan Z, Khan NA, Jhung SH (2013) Adsorptive removal of bisphenol-a from water with a metal-organic framework, a porous chromium-benzenedicarboxylate. J. Nanosci. Nanotechnol. 13:2789–2794. https://doi.org/10.1166/jnn.2013.7411 73. Peng Y, Zhang Y, Huang H, Zhong C (2018) Flexibility induced high-performance MOF-based adsorbent for nitroimidazole antibiotics capture. Chem. Eng. J. 333:678–685. https://doi.org/10.1016/j.cej.2017.09.138 74. Zhao X, Zhao H, Dai W, Wei Y, Wang Y, Zhang Y, Zhi L, Huang H, Gao Z (2018) A metalorganic framework with large 1-D channels and rich OH sites for high-efficiency chloramphenicol removal from water. J. Colloid Interface Sci. 526:28–34. https://doi.org/10.1016/j. jcis.2018.04.095 75. Zhu X, Li B, Yang J, Li Y, Zhao W, Shi J, Gu J (2015) Effective adsorption and enhanced removal of organophosphorus pesticides from aqueous solution by Zr-based MOFs of UiO-67. ACS Appl. Mater. Interfaces 7:223–231. https://doi.org/10.1021/am5059074
8 Adsorptive Purification of Water Contaminated with Hazardous Organics by. . .
289
76. Zhao H, Hou S, Zhao X, Liu D (2019) Adsorption and pH-responsive release of tinidazole on metal–organic framework CAU-1. J. Chem. Eng. Data 64:1851–1858. https://doi.org/10. 1021/acs.jced.9b00106 77. Yang Q, Vaesen S, Ragon F, Wiersum AD, Wu D, Lago A, Devic T, Martineau C, Taulelle F, Llewellyn PL, Jobic H, Zhong C, Serre C, Weireld GD, Maurin G (2013) A water stable metal–organic framework with optimal features for CO2 capture. Angew. Chem. Int. Ed. 52:10316–10320. https://doi.org/10.1002/anie.201302682 78. Ragon F, Campo B, Yang Q, Martineau C, Wiersum AD, Lago A, Guillerm V, Hemsley C, Eubank JF, Vishnuvarthan M, Taulelle F, Horcajada P, Vimont A, Llewellyn PL, Daturi M, Devautour-Vinot S, Maurin G, Serre C, Devic T, Clet G (2015) Acid-functionalized UiO-66 (Zr) MOFs and their evolution after intra-framework cross-linking: structural features and sorption properties. J. Mater. Chem. A 3:3294–3309. https://doi.org/10.1039/C4TA03992K 79. Biswas S, Zhang J, Li Z, Liu YY, Grzywa M, Sun L, Volkmerc D, Voort PVD (2013) Enhanced selectivity of CO2 over CH4 in sulphonate-, carboxylate- and iodo-functionalized UiO-66 frameworks. Dalton Trans. 42:4730–4737. https://doi.org/10.1039/C3DT32288B 80. Hu Z, Faucher S, Zhuo Y, Sun Y, Wang S, Zhao D (2015) Combination of optimization and metalated-ligand exchange: an effective approach to functionalize UiO-66(Zr) MOFs for CO2 separation. Chem. Eur. J. 21:17246–17255. https://doi.org/10.1002/chem.201503078 81. Sarker M, Song JY, Jhung SH (2018) Carboxylic-acid-functionalized UiO-66-NH2: a promising adsorbent for both aqueous- and non-aqueous-phase adsorptions. Chem. Eng. J. 331:124–131. https://doi.org/10.1016/j.cej.2017.08.017 82. Sarker M, Shin S, Jhung SH (2019) Functionalized mesoporous metal-organic framework PCN-100: an efficient carrier for vitamin E storage and delivery. J. Ind. Eng. Chem. 74:158–163. https://doi.org/10.1016/j.jiec.2019.02.022 83. DeCoste JB, Demasky TJ, Katz MJ, Farha OK, Hupp JT (2015) A UiO-66 analogue with uncoordinated carboxylic acids for the broad-spectrum removal of toxic chemicals. New J. Chem. 39:2396–2399. https://doi.org/10.1039/C4NJ02093F 84. Song JY, Ahmed I, Seo PW, Jhung SH (2016) UiO-66-type metal–organic framework with free carboxylic acid: versatile adsorbents via H-bond for both aqueous and nonaqueous phases. ACS Appl. Mater. Interfaces 8:27394–27402. https://doi.org/10.1021/acsami.6b10098 85. Luo Z, Fan S, Liu J, Liu W, Shen X, Wu C, Huang Y, Huang G, Huang H, Zheng M (2018) A 3D stable metal–organic framework for highly efficient adsorption and removal of drug contaminants from water. Polymers 10:209–222. https://doi.org/10.3390/polym10020209 86. Sarker M, Jhung SH (2019) Zr-MOF with free carboxylic acid for storage and controlled release of caffeine. J. Mol. Liq. 296:112060. https://doi.org/10.1016/j.molliq.2019.112060 87. Sarker M, Shin S, Jhung SH (2019) Synthesis and functionalization of porous Zr-diaminostilbenedicarboxylate metal–organic framework for storage and stable delivery of ibuprofen. ACS Omega 4:9860–9867. https://doi.org/10.1021/acsomega.9b01139 88. Foo ML, Horike S, Fukushima T, Hijikata Y, Kubota Y, Takata M, Kitagawa S (2012) Ligandbased solid solution approach to stabilisation of sulphonic acid groups in porous coordination polymer Zr6O4(OH)4(BDC)6 (UiO-66). Dalton Trans. 41:13791–13794. https://doi.org/10. 1039/C2DT31195J 89. Chen J, Li K, Chen L, Liu R, Huang X, Ye D (2014) Conversion of fructose into 5-hydroxymethylfurfural catalyzed by recyclable sulfonic acid-functionalized metal–organic frameworks. Green Chem. 16:2490–2499. https://doi.org/10.1039/C3GC42414F 90. Chung YM, Kim HY, Ahn WS (2014) Friedel–crafts acylation of p-xylene over sulfonated zirconium terephthalates. Catal Lett 144:817–824. https://doi.org/10.1007/ s10562-014-1242-4 91. Zhao X, Wang K, Gao Z, Gao H, Xie Z, Du X, Huang H (2017) Reversing the dye adsorption and separation performance of metalorganic frameworks via introduction of SO3H groups. Ind. Eng. Chem. Res. 56:4496–4501. https://doi.org/10.1021/acs.iecr.7b00128
290
D. K. Yoo et al.
92. Luo XP, Fu SY, Du YM, Guo JZ, Li B (2017) Adsorption of methylene blue and malachite green from aqueous solution by sulfonic acid group modified MIL-101. Micro Meso Mater 237:268–274. https://doi.org/10.1016/j.micromeso.2016.09.032 93. Hassan Z, Khan NA, Jhung SH (2016) Adsorptive removal of diclofenac sodium from water with Zr-based metal–organic frameworks. Chem. Eng. J. 284:1406–1413. https://doi.org/10. 1016/j.cej.2015.08.087 94. Nguyen JG, Cohen SM (2010) Moisture-resistant and superhydrophobic metalorganic frameworks obtained via postsynthetic modification. J. Am. Chem. Soc. 132:4560–4561. https://doi.org/10.1021/ja100900c 95. Garibay SJ, Cohen SM (2010) Isoreticular synthesis and modification of frameworks with the UiO-66 topology. Chem. Commun. 46:7700–7702. https://doi.org/10.1039/C0CC02990D 96. Liu K, Zhang S, Hu X, Zhang K, Roy A, Yu G (2015) Understanding the adsorption of PFOA on MIL-101(Cr)-based anionic-exchange metal–organic frameworks: comparing DFT calculations with aqueous sorption experiments. Environ. Sci. Technol. 49:8657–8665. https://doi. org/10.1021/acs.est.5b00802 97. Haque E, Lee JE, Jang IT, Hwang YK, Chang JS, Jegal J, Jhung SH (2010) Adsorptive removal of methyl orange from aqueous solution with metal-organic frameworks, porous chromium-benzenedicarboxylates. J. Hazard. Mater. 181:535–542. https://doi.org/10.1016/j. jhazmat.2010.05.047 98. An HJ, Sarker M, Yoo DK, Jhung SH (2019) Water adsorption/desorption over metal-organic frameworks with ammonium group for possible application in adsorption heat transformation. Chem. Eng. J. 373:1064–1071. https://doi.org/10.1016/j.cej.2019.05.121 99. Ahmed I, Jhung SH (2016) Adsorptive desulfurization and denitrogenation using metalorganic frameworks. J. Hazard. Mater. 301:259–276. https://doi.org/10.1016/j.jhazmat.2015. 08.045 100. Rojas S, Arenas-Vivo A, Horcajada P (2019) Metal-organic frameworks: a novel platform for combined advanced therapies. Coord. Chem. Rev. 388:202–226. https://doi.org/10.1016/j.ccr. 2019.02.032 101. Li X, Ma W, Li H, Bai Y, Liu H (2019) Metal-organic frameworks as advanced sorbents in sample preparation for small organic analytes. Coord. Chem. Rev. 97:1–13. https://doi.org/10. 1016/j.ccr.2019.06.014 102. Song JY, Bhadra BN, Jhung SH (2017) Contribution of H-bond in adsorptive removal of pharmaceutical and personal care products from water using oxidized activated carbon. Micro Meso Mater 243:221–228. https://doi.org/10.1016/j.micromeso.2017.02.024 103. Razavi SAA, Morsali A (2019) Linker functionalized metal-organic frameworks. Coord. Chem. Rev. 399:213023. https://doi.org/10.1016/j.ccr.2019.213023
Chapter 9
MOFs Constructed from Biomolecular Building Blocks Zachary M. Schulte and Nathaniel L. Rosi
9.1
Introduction
Metal-organic frameworks (MOFs) are ultimate synthetic examples of ordered 3-D molecular space [28]. The development of their chemistry over the last several decades has led to rational placement of metals and molecules in specific positions with respect to one another to enable unique properties and applications. It is naturally enticing to compare MOFs and their 3-D molecular complexity to biological molecules such as proteins, for which the placement and 3-D organization of amino acids determine properties and function. Further, researchers are inspired to consider using coordination-driven assembly and MOF design principles (i.e., reticular chemistry; [41, 93]) to organize natural biological building blocks (e.g., nucleobases, nucleosides, and nucleotides; amino acids, sugars, etc.) and larger biomolecules (e.g., peptides and proteins) into biomolecule-containing MOFs. Such MOFs having biomolecules integral to their structure are defined in the literature as “bio-MOFs” [4] or “MbioFs” [40]. They are attractive materials due to their potential biological and environmental compatibility and nontoxicity, which is important for applications in biomedicine as well as the personal care and food industries [33]. From a practical synthetic standpoint, biomolecules are ideal MOF building blocks due to their accessibility and ability to coordinate metal ions. This chapter will focus on the construction of MOFs using biomolecule linkers. The content is not intended to be exhaustive; rather, major advances and milestones in the use of biomolecular linkers in MOF chemistry are highlighted.
Z. M. Schulte · N. L. Rosi (*) Department of Chemistry, Kenneth P. Dietrich School of Arts and Sciences, University of Pittsburgh, Pittsburgh, PA, USA e-mail: [email protected]; [email protected] © Springer Nature Switzerland AG 2021 P. Horcajada Cortés, S. Rojas Macías (eds.), Metal-Organic Frameworks in Biomedical and Environmental Field, https://doi.org/10.1007/978-3-030-63380-6_9
291
292
9.2
Z. M. Schulte and N. L. Rosi
Nucleobases
Purines and pyrimidines (Fig. 9.1) are the simplest building blocks of nucleic acids. They contain multiple nitrogen donor atoms that make them attractive ligands for coordination-driven assembly of discrete complexes and polyhedra as well as extended frameworks. Their rigid ring systems and multiple possible metal coordination modes underline their potential as building blocks for a diverse range of robust materials.
9.2.1
Discrete Complexes and 1-D Polymers
Metal-nucleobase complexes have been targeted and studied since the 1950s. At that time it was well-established that silver ions and purine nucleobases formed precipitates when mixed in solution, yet their structural determination was challenging. In 1964, Davidson and coworkers hypothesized that they consisted of one-dimensional chains of alternating Ag+ and adeninate ions linked through the N7 and N9 imidazolate nitrogens [32]. However, they found that precipitates still formed when functional groups were added to the N9 position of adenine, preventing coordination at that site. Additionally, they determined that the ratio of Ag+ to purine was 3:2, suggesting adeninate may bind to two or more Ag+ in different coordination modes. Much later Navarro and Lippert demonstrated how certain coordination sites on various purine and pyrimidine ligands could be used to construct complexes with specific shapes and geometries [67]. They prepared various macrocycles using nucleobase-Pt2+ coordination as well as hydrogen bonding interactions. Discrete complexes were targeted by N-methylating certain nucleobase N-donor sites or blocking coordination sites on the Pt2+ with non-bridging ligands such as methylamine. Furthermore, they noted that purine and pyrimidine complexes led to N-Pt-N bond angles near 90 and 120 , respectively. They hypothesized that these binding motifs could potentially be used to “dial-in” specific vertices in the construction of macrocycles or polyhedra. These seminal studies laid the groundwork for developing extended multidimensional metal-nucleobase materials.
Fig. 9.1 Chemical structures of natural nucleobases
9 MOFs Constructed from Biomolecular Building Blocks
9.2.2
293
Purine-Based Bio-MOFs
Prior to 2000, most of the research on metal-nucleobase coordination assemblies focused on synthesizing metal-nucleobase macrocycles with precious metals and functionalized purines such as 9-methylguanadine or 9-methyl-6-dimethylamino adenine. At this time, MOFs were emerging from their infancy, and researchers began to investigate the possibility of constructing MOFs with transition metals and nucleobases. Unfunctionalized nucleobases such as adenine, with a wide variety of coordination modes due to the number of N-donor sites available for coordination and their relative positions to one another, were considered ideal candidates for bridging multiple metal ions (Fig. 9.2). García-Terán et al. combined adenine with a Cu-oxalate complex to yield the first adeninate-based MOF: {[Cu2(μAd)4(H2O)2]-[Cu(ox)(H2O)]2• ~ 14H2O}n (Ad ¼ adeninate) [29]. In this structure, two Cu2+ ions are bridged by four distinct adeninate ions through coordination at N3 and N9, forming a pseudo paddlewheel motif. Further coordination of a second Cu2+ ion at the N7 site leads to the formation of a three-dimensional extended structure. This second Cu2+ site is four-coordinated with two monodentate adeninates and a chelating bidentate oxalate. Therefore, the adeninate is the only bridging linker and acts as a three-connected linker. The framework consists of one-dimensional channels with a diameter of ~13 Å. Led by early work from Rosi and coworkers, metal-adeninate MOF chemistry saw dramatic developments between 2009 and 2013. An and Rosi reported bio-MOF-11, [Co2(Ad)2(COOCH3)2] • 2DMF, 0.5H2O (DMF ¼ dimethylformamide), a 3-D MOF composed of Co2+, adeninate, and acetate ions [5]. Co2+ dimers are bridged by two adeninates via N3 and N9 and two acetates (Fig. 9.3a). The Co2+ sites are then apically coordinated to N7 of adjacent adeninates, which serve to connect neighboring clusters together. Narrow channels with a diameter of 5.2 Å run throughout the MOF and are lined with Lewis basic primary amines and uncoordinated pyrimidine N1 from the adeninate (Fig. 9.3b). The highly polarized pore space leads to selective adsorption of CO2
Fig. 9.2 Possible coordination modes of adenine and adeninate (Ad) when binding to at least two metal ions
294
Z. M. Schulte and N. L. Rosi
Fig. 9.3 Structure of bio-MOF-11. (a) Paddlewheel SBUs are bridged by alternating aliphatic carboxylates and bridging adeninates. Two additional adeninates coordinate at the apical sites. All four adeninates link to adjacent Co2+ paddlewheels to form (b) a permanently porous 3-D MOF having channels lined with Lewis basic sites. Pale blue tetrahedra and black, blue, maroon, and white spheres represent Co2+, carbon, nitrogen, oxygen, and hydrogen atoms, respectively
over N2. Román and coworkers expanded upon this work in 2011 by creating a family of Cu-ad MOFs isostructural to bio-MOF-11 using acetate derivatives with varying alky tail lengths [71]. Extending the alkyl tail length from a methyl group to a pentyl group decreased the BET surface area from 505 to 9 m2g1. Similarly, Rosi and coworkers prepared a family of bio-MOF-11 analogues by replacing the acetate with other aliphatic monocarboxylates [45–47]. They demonstrated that the alkyl chain length decorating the pore space could be adjusted to tune the porosity and water stability of the framework. For example, bio-MOF-11 (acetate) has more pore space but essentially dissolves in water, while bio-MOF14 (valerate) has limited pore space but maintains its structure and porosity after soaking in water up to 1 month. Based on this research, the Rosi group designed a core-shell MOF material comprising a bio-MOF-11 core and a bio-MOF-14 shell [47]. This material featured the higher porosity of bio-MOF-11 yet the water stability of bio-MOF-14. It should be noted that in order to build the core-shell material, the core must be prepared with a mixture of acetate and valerate so that the unit cell of the core would be close enough to the unit cell of the shell to allow for epitaxy between core and shell. Rosi and Carreon later developed membranes with these cobalt-adeninate MOFs and explored their application in CO2/CH4 separations [92]. A sub-category of MOFs, zeolitic imidazolate frameworks (ZIFs), can also be constructed with purine nucleobases. The imidazolate portion of the purine ring can serve to bridge the metal ions together to form ZIFs. Hayashi et al. developed a series of four MOFs composed of Co2+ or Zn2+ and purine derivatives with varying positions and numbers of nitrogen atoms on the six-membered ring [39]. Only one of the purines, 4-azabenzimidazole, formed the dia net (ZIF-23), while the other two purines, purinate and 5-azabenzimidazole, formed structures with the lta topology. It was proposed that the presence of dipole-dipole interactions between the nitrogen
9 MOFs Constructed from Biomolecular Building Blocks
295
atoms at positions 5 and 6 of adjacent purines promote formation of the lta topology. In addition to changing the topology, different purines incorporated into the ZIF also alter the adsorption properties of CO2. ZIF-20 (Zn-purinate) adsorbs five times more CO2 than CH4 at 1 atm. Paddlewheel motifs, dimers, or single-metal tetrahedral geometries are common in simple nucleobase bio-MOFs. However, in 2019, Zhou and coworkers synthesized NbU-1, a Cu-adeninate bio-MOF with a novel secondary building unit (SBU) based on Cu7(OH)6 clusters in a planar geometry [49]. Each Cu7 cluster includes two types of Cu ion: a central Cu in an octahedral geometry bound to six μ 3-OH groups and six tetrahedral Cu ions bound to two μ 3-OH and two nitrogen sites from two distinct adeninates at N3 or N9. At the periphery of the cluster is a third type of Cu bound to the N7 positions of four neighboring adeninates in a tetrahedral arrangement. The resulting geometry is a pseudo-planar hexagon with six points of extension. The framework adopts the gar topology and has both discrete pores and continuous channels. Interesting aspects of this unusual SBU include the number of open metal sites per Cu7 cluster and the exposed Lewis basic amines, which could be useful for molecular separations, as demonstrated through acetylene/ethylene breakthrough experiments.
9.2.3
Purine-Based Bio-MOFs with Secondary Linkers
Major advances in purine-based MOF construction coincided with the introduction of a second linker into bio-MOF syntheses. Mixing purines with metal ions and carboxylated ligands has resulted in a broad and ever-developing class of new multicomponent bio-MOFs. It should be noted that the first examples of these structures can be considered some of the earliest multicomponent MOF materials. The Rosi group was the first to demonstrate this concept by introducing 4,40 -biphenyldicarboxylic acid (H2-BPDC) to a mixture of Zn2+ and adenine to form bio-MOF-1 [4]. Bio-MOF-1, Zn8(Ad)4(BPDC)6(O) • 2DMA, 8DMF, 11H2O (DMA ¼ dimethylammonium), consists of Zn2+-adeninate columns interconnected by BPDC linkers (Fig. 9.4a). The columns consist of corner-fused Zn2+-adeninate octahedral cages formulated as Zn8(Ad)4(O)2 (Fig. 9.4b). An adeninate resides at every other face of the octahedron, while Zn2+ sit at four equatorial corner positions as well as two positions adjacent to either apical O2. Bio-MOF-1 is permanently porous and has an overall negative charge that is balanced by DMA ions within the pores (Fig. 9.4c). It was demonstrated that the nature of the counteraction can significantly affect the stability of bio-MOF-1 in water [81]. When DMA is replaced with more hydrophobic alkylammonium cations, water stability improves. The anionic nature of bio-MOF-1 has been exploited in multiple different studies. In the original paper, the dimethylammonium cations were replaced with the cationic drug procainamide, and the drug storage and release properties of the material were investigated. Rosi and coworkers also examined how the identity of the organic cation impacted properties such as CO2 uptake [3, 44]. Lanthanide (Ln3+) cations
296
Z. M. Schulte and N. L. Rosi
Fig. 9.4 Structure of bio-MOF-1. (a) Zn8Ad4 SBUs share μ4-O to form (b) infinite-rod SBUs. (c) Rod-like SBUs are interconnected by BPDC ligands to form an anionic 3-D framework with 1-D channels (charge-balancing dimethylammonium cations not shown). Dark blue tetrahedra and black, light blue, maroon, and white spheres indicate Zn2+, carbon, nitrogen, oxygen, and hydrogen atoms
Fig. 9.5 Structure and connectivity of bio-MOF-100. (a) Zn8Ad4 SBUs are (b) interconnected bundles of three BPDC ligands in a tetrahedral arrangement. (c) Discrete cages and (d) large one-dimensional channels within the 3-D framework. Dark blue tetrahedra and black, light blue, maroon, and white spheres indicate Zn2+, carbon, nitrogen, oxygen, and hydrogen atoms. The large yellow sphere designates void space
can also be exchanged into bio-MOF-1, and the framework can be used as both an antenna for the Ln3+ and a protective matrix [6]. The resulting Ln3+@bio-MOF-1 can be used as a sensor for O2 [6, 90]. In the presence of O2, the luminescence is quenched until the material is reactivated via purging with N2. Bio-MOF-1 has also been used for applications such as sequestration of carbon monoxide complexes and detection of explosive molecules [18, 89]. The Rosi group also discovered bio-MOF-100, which has the molecular formula Zn8(Ad)4(BPDC)6(O)2 • 4DMA, 49DMF, 31H2O. Bio-MOF-100 is constructed from the same zinc-adeninate SBU observed in bio-MOF-1, yet now as separated clusters (Fig. 9.5a) [7]. Bundles of three BPDC linkers coordinate to every other face of the discrete Zn8Ad4(O)2 octahedral clusters, forming a pseudotetrahedron (Fig. 9.5b). These tetrahedra are linked together into the 3-D bio-MOF100 structure, which can be simplified as an augmented lcs net. Large mesoporous channels with a diameter of 28 Å run along [110], [101], and [011], resulting in a
9 MOFs Constructed from Biomolecular Building Blocks
297
highly open, porous structure (Fig. 9.5c, d). As with bio-MOF-1, bio-MOF-100 is also anionic and permanently porous, and it has been investigated for drug delivery and cation capture applications [14, 20, 75]. The large open channels and unique 3-linker bundles have led to interesting postsynthetic chemistry on the bio-MOF-100 system. When bio-MOF-100 is prepared with N3-BPDC instead of BPDC, strain-promoted “click” chemistry can be used to append large molecules to the walls of the framework [51]. Tandem post-synthetic modification (PSM) reactions were used to conjugate diphenylalanine to the framework, indicating that small biomolecules could be anchored to bio-MOF-100. Crucial to these demonstrations was the large accessible pore space that allowed for facile diffusion of large molecules. Individual BPDC ligands within bio-MOF100 are highly labile because each BPDC in the three-linker bundles coordinates to Zn2+ through one monodentate carboxylate oxygen. Therefore, the temporary removal of one BPDC does not compromise the crystalline integrity of the MOF. Soaking bio-MOF-100 crystals in solutions of functionalized BPDC ligands, such as 2-azido-, 2-amino-, and 2-formyl-BPDC, results in nearly quantitative ligand exchange depending on the chosen conditions [52]. Sequential linker exchange reactions were designed to construct bio-MOF-100 variants with up to three different functionalized linkers. It is envisaged that such functionalized versions of bio-MOF100 could act as a 3-D scaffold for organizing large, complex molecules in 3-D space, which would be selectively appended to the functionalized MOF. In addition to exchange with functionalized ligands, Li et al. determined that isoreticular analogues of bio-MOF-100 could be prepared by replacing shorter dicarboxylate linkers with longer dicarboxylate linkers [45–47]. BPDC can be quantitatively replaced by azobenzene-4,40 -dicarboxylate (ABDC) to yield bio-MOF-102, increasing the unit cell by 6.34 Å. Then, ABDC in bio-MOF-102 can be replaced by 20 -amino-1,10 :4,100 -terphenyl-4,400 -dicarboxylate (NH2-TPDC) to yield bio-MOF-103 with unit cell dimensions a ¼ b ¼ c ¼ 82.25 Å. This was the first demonstration that short linkers could be exchanged with longer linkers in MOFs to systematically increase the MOF unit cell and, consequently, the porosity. Rosi and coworkers later demonstrated that this sort of linker exchange could be halted at intermediate stages of exchange to yield hierarchical “gradient MOFs” having a gradual transition in linker composition from BPDC in the core of the MOF crystal to ABDC in the shell [53]. Gradient MOFs with up to three layers were synthesized and have unique potential for applications in separations and sequestration of specific analytes [50]. Other allotropes of bio-MOF-100 have been discovered through slight modification of the originally reported synthetic conditions. Platero-Prats and coworkers showed that introduction of stirring to an otherwise identical synthetic procedure for bio-MOF-100 led to the formation of a dia-c network composed of Zn8Ad4(O)2 SBUs and BPDC ligands [34]. Although the components and SBUs are identical to that of bio-MOF-100, the product, dia-c-bio-MOF-100, is doubly interpenetrated and has lower symmetry. They further demonstrated complete conversion from the dia-c topology to the larger, more porous lcs topology at room temperature using solvent-assisted ligand exchange with an Ir-bipyridinedicarboxylate complex. In
298
Z. M. Schulte and N. L. Rosi
another work, Ma et al. developed a 3-D MOF based on infinite rods of Zn6Ad4 clusters interconnected by BPDC ligands [58]. These clusters are similar to those found in bio-MOF-1 and bio-MOF-100 but are missing two Zn2+, decreasing the connectivity of each SBU to eight. The resulting MOF, JXNU-4, is anionic and has large 1-D channels with a diameter of 9.8 Å running parallel to the columns of Zn6Ad4 clusters. This MOF shows high uptake and selectivity of propane over smaller paraffins ethane and methane due to strong interactions between more polarizable analytes and the anionic framework. The authors also highlight the relatively high stability of JXNU-4 in aqueous solutions from pH 4 to 11. Linkers other than BPDC have been used as secondary linkers to create bio-MOFs with nucleobases. Zhang and coworkers synthesized a MOF with the general formula Zn(Ad)(AIN)•DMF, where AIN is the secondary linker 2-aminoisonicotinate [94]. Both ligands only coordinate to two tetrahedral Zn2+. Adeninate ligands bind through N3 and N9 positions, while the AIN ligands coordinate in a monodentate fashion with the carboxylate and the pyridyl moiety. The resulting MOF adopts the dmp topology with rhombic pores measuring 4.2 x 5.3 Å in dimension lined by primary amines from both the adeninate and AIN ligands. The additional Lewis basic sites in the pores provided by the aminofunctionalized AIN increase the isosteric heat of adsorption (Qst) by 9 kJ/mol and improve the CO2/N2 selectivity from 90 to 707 at 273 K compared to an unfunctionalized isoreticular MOF [88]. Using the same approach, Zhang replaced AIN with another linear linker, 4-pyrazolecarboxylic acid (4-pca), to yield an anionic MOF with the general formula [NH2(CH3)2][Zn3(4-pca)3(Ad)] •10DMF•8H2O [27]. Each tetrahedral Zn2+ is coordinated by three distinct 4-pca ligands and one adeninate. Two Zn2+ form a dimer bridged through the pyrazolate moieties of two 4-pca ligands. The Zn2+ dimers are interconnected by a third 4-pca via monodentate coordination to the carboxylate (Fig. 9.6a), resulting in planar [Zn2(4-pca)2]n motifs that are then pillared by adeninate ligands coordinating through N3, N9, and N7 to form a porous 3-D framework (Fig. 9.6b and c). Changes in the crystalline structure were observed by variable temperature PXRD where notable peaks shifted to higher 2 θ angles with increasing temperatures, suggesting a transformation to a less porous structure (Fig. 9.6d). In addition to linear ditopic ligands, nucleobase bio-MOFs incorporating adenine have been synthesized with tritopic and tetratopic secondary linkers. Perhaps the simplest and most common tritopic linker is 1,3,5-benzenetricarboxylic acid (BTC). An anionic 3-D Zn-BTC MOF with the formula [Zn3(BTC)2(Ad)(H2O)• (CH3)2NH2•xDMF•yH2O] was synthesized by Cai et al. [17]. In this MOF, both the adenine and the BTC are tritopic and thus coordinate to three distinct Zn2+. The structure contains one-dimensional channels with dimensions 8 11 Å lined with uncoordinated pyrimidine and NH2 moieties from adenine. Infrared and Raman spectroscopies in addition to X-ray diffraction studies were used to indirectly observe the interaction of these sites with the complementary nucleobase thymine after soaking crystals in thymine solutions.
9 MOFs Constructed from Biomolecular Building Blocks
299
Fig. 9.6 Structure of Zn-pca. (a) [Zn2(4-pca)2] dimers formed by bridging pyrazolate moieties of two 4-pca ligands. Two additional 4-pca ligands coordinate in a monodentate fashion. The tetrahedral geometry around each Zn2+ is completed by a bridging adeninate coordinating through N3 and N7. (b) Wave-like planes of [Zn2(4-pca)2] are pillared by adeninate to form (c) a 3-D structure with 1-D channels. (d) Topological distortions from heating alter the pore dimensions. Dark blue, black, light blue, and maroon spheres represent Zn2+, carbon, nitrogen, and oxygen atoms. Hydrogen atoms are omitted for clarity
The tetratopic linker, 1,3,6,8-tetrakis( p-benzoic acid) pyrene, or PTBA, was added to solutions of adenine and Zn(NO3)2 to yield SION-19 [9]. Chains of Zn6Ad4 clusters are interconnected by monodentate coordination of PTBA at the Zn2+ sites (Fig. 9.7a, b). The resulting framework has two types of spiraling channels with slightly different dimensions: a basic channel lined with free primary and pyrimidyl amines and an acidic channel decorated with uncoordinated oxygen atoms from the carboxylates (Fig. 9.7c, d). The functionality of the basic channel was investigated for selective adsorption and dimerization of thymine via WatsonCrick base pairing. Thymine uptakes of 40–45% within SION-19 led to dimerization of thymine in approximately 58% yield upon irradiation with 240 nm UV light (Fig. 9.7e). Altering thymine loading within the channels led to a drastic drop in the yield of the dimerized product. Therefore, SION-19 serves as an excellent example of how the unique structural and functional properties of nucleobase bio-MOFs are ideally suitable to the study of complex processes in controlled environments.
300
Z. M. Schulte and N. L. Rosi
Fig. 9.7 Crystal structure and thymine dimerization in SION-19. (a) Infinite rods of Zn6Ad4 SBUs (b) interconnected by PTBA at both axial (gray) or equatorial (green) positions. (c) Representative Connolly surface areas of the acidic (red) and basic (blue) channels. (d) Magnified image of the exposed Watson-Crick faces projecting into the basic pore space. (e) Schematic of thymine loading sites signified by red dots within the pore by molecular dynamic simulations and formation of dimerized thymine after UV irradiation. Pink, gray, blue, red, and pale yellow spheres indicate Zn2+, carbon, nitrogen, oxygen, and hydrogen atoms
9.3
Amino Acids, Peptides, and Proteins
Naturally occurring amino acids contain an amine and a carboxylic acid, two functional groups that can coordinate metal ions, making them natural ditopic linkers (Fig. 9.8). Amino acids and peptides are highly versatile ligands for constructing MOFs: (i) increasing the length of a peptide can potentially affect the crystallographic dimensions and pore metrics; (ii) pore functionality can be imparted by including natural or unnatural side chains, such as hydroxyl, thiol, or aliphatic moieties; (iii) side chain or dangling functional groups are amenable to further chemical functionalization, e.g., dimerization and oxidation/reduction; and (iv) the inherent chirality of α-amino acids can lead to chiral pores or channels. Modified peptides or secondary linkers could also be used to impart additional stability or higher degrees of connectivity to frameworks otherwise solely composed of flexible peptides. In addition to their use as organic ligands for bio-MOF syntheses, amino acids have also been explored as modulators to control the size or morphology of MOF crystals and as capping agents to functionalize either defect sites or the external surface; however, that research will not be included in this chapter [15, 38, 60, 78]. The main focus of this section will be the incorporation of amino acids, peptides, and proteins as bridging organic ligands in bio-MOFs and their properties and applications.
9 MOFs Constructed from Biomolecular Building Blocks
301
Fig. 9.8 Common amino acid linkers used in bio-MOF syntheses
9.3.1
Amino Acids
In many proteins, transition metals anchored to specific amino acid residues play key roles in catalysis and molecular transport. Decades of research have been devoted to understanding the interactions between metals and amino acids. In 1966, Gramaccioli determined the structure of two metal-glutamate hydrate crystalline systems based on Cu and Zn ions [36, 37]. In the Cu framework, one deprotonated glutamic acid bridges two Cu2+ via a monodentate carboxylate bond from the side chain moiety and bidentate chelation through the C-terminal carboxylate oxygen and the α-amine. Each Cu2+ is a distorted octahedron where three of the four equatorial sites are occupied by two distinct glutamates and the fourth by a water molecule. Two additional oxygen atoms, one each from the second and third coordinated glutamate, occupy the axial positions through monodentate carboxylate coordination. The Zn2+ framework is nearly identical. Although the field of MOFs had not yet emerged, these examples can be considered early examples of bio-MOFs incorporating amino acid ligands. Amino acids can exist in a number of different zwitterionic forms based on environmental pH. Therefore, in order to prepare multidimensional amino acid bio-MOFs, one must carefully control the pH of reaction conditions. Zheng and coworkers demonstrated the synthesis of cuboid [Ln4(μ 3-OH)4]8+ clusters with Sm3 + , Er3+, and Nd3+ [87]. They found that at near physiological pH, discrete lanthanideamino acid clusters formed with glycine, alanine, and valine. However, with glutamic acid, a 3-D porous framework with parallelogram-shaped channels was observed. Each Ln4 cluster was bound to six glutamates via bidentate chelating carboxylates. The concept of controlling pH to direct the assembly of ligands was
302
Z. M. Schulte and N. L. Rosi
further developed by Anokhina and Jacobson [10]. They synthesized both homochiral and achiral versions of a 1-D nickel-aspartic acid (NiAsp) framework. Each Asp binds to five Ni2+ through all four carboxylate oxygen atoms and the αamine to form infinite 1-D helices of NiO2 octahedra. These rods are held together by hydrogen bonding through guest water molecules. Upon increasing the pH of the synthetic mixture, additional [NiAsp2]2 octahedra form bridges between adjacent helices and thus extend the dimensionality to a 3-D network [11]. The latter structure exhibited stability and permanent porosity. Cysteine, another naturally occurring amino acid, has a thiolated side chain. Disulfide bridges between two cysteine (Cys) residues produce the dimerized form, cystine (CYS), a motif that is important for protein folding. Researchers adapted this feature to amino acid bio-MOFs by mixing Zn(CO3)2 with a solution of L-cysteine to yield a Zn(CYS)2 structure containing octahedral Zn2+ sites coordinated by four different cystines (Fig. 9.9a) [23]. Each Zn2+ coordinates two bidentate chelating cystines via a monodentate carboxylate oxygen and the αamine and two monodentate carboxylate oxygens from two other cystines. Therefore, the cystines serve as a flexible ditopic linker with carboxylate moieties acting as bridges between adjacent Zn2+ to form a 3-D bio-MOF (Fig. 9.9b, c). An isostructural compound, Mg(CYS)2, was designed and synthesized for oxidation control in cells [91]. Substituting Mg2+ for Zn2+, they investigated the ability of Mg (CYS)2 to regulate reactive oxygen species in cells by slow decomposition of the MOF in solution. They determined that the MOF was nontoxic and released Mg2+ over time, which assists in the synthesis of glutathione, a natural antioxidant formed intracellularly. Another sulfur-containing amino acid, methionine (Met), was used to develop a heterobimetallic MOF, [{Ag3Cu3(L-Met)6(NO3)3(H2O)3}•7H2O]n [56]. The structure contains alternating 1-D rods formed either via interactions between the Cu2+ and carboxylate and α-amino groups or “soft” interactions between Ag+ trimers and the methyl sulfide side chain. The MOF exhibits large homochiral helical channels filled with chains of hydrogen-bonded water molecules.
Fig. 9.9 Structure of Zn(CYS)2. (a) Hexacoordinated Zn2+ from four distinct CYS ligands. (b) The connectivity of Zn2+ octahedra via bridging carboxylates and bidentate chelation at both termini. (c) Extended 3-D framework with open pores. Dark blue octahedra and black, light blue, maroon, and yellow spheres correspond to Zn2+, carbon, nitrogen, oxygen and sulfur atoms. Hydrogen atoms are omitted for clarity
9 MOFs Constructed from Biomolecular Building Blocks
9.3.2
303
Small Peptides and Secondary Linkers
Bio-MOFs constructed using single amino acids are rarely linked in more than two dimensions and are generally dense structures. Expanding the channels and accessible pore volume of amino acid-based MOFs can be critical to realizing the unique structure and functionality the amino acids may impart. One approach to increasing the accessible porosity as well as the potential functional complexity involves using di- or tripeptides as linkers. Takayama et al. used glycine-glycine (GlyGly) to synthesize three coordination polymers [84]. Two isostructural compounds were synthesized with Zn2+ and Cd2+ at pH 6 where each metal ion is a distorted octahedron bound to four distinct ligands (Fig. 9.10a). The dipeptide GlyGly has a C-terminal carboxylate and an N-terminal primary amine; glycine does not possess a side chain that can coordinate metal ions. Two ligands chelate the metal ions at the equatorial positions via the carbonyl oxygen from the amide group and the N-terminal amino group. Carboxylate oxygens from the C-termini of two additional GlyGly coordinate in a monodentate fashion at the axial sites, resulting in two-dimensional sheets that are then interconnected by hydrogen bonding with guest water molecules (Fig. 9.10b, c). When NaOH was added to the synthesis of
Fig. 9.10 Different frameworks of Cd-GlyGly formed at (a–c) pH 6 or d-f) pH 9. (a) Cd(GlyGly)4 SBU with two equatorial bidentate ligands and two axial monodentate ligands. (b) The resulting 2-D sheets stack into (c) 3-D structures via water-induced hydrogen bonding. (d) Cd(GlyGly)6 SBUs form (e) 2-D wave-like sheets resulting in a unique hydrogen-bonded framework. Cd2+, carbon, nitrogen, and oxygen atoms are represented by purple octahedra and black, blue, and maroon spheres. Hydrogen atoms and coordinated water molecules are omitted for clarity
304
Z. M. Schulte and N. L. Rosi
the Cd2+ compound to adjust the pH to 9, a new connectivity was discovered. The number of ligands at each metal site increased to six, and the metal ion coordination geometry shifted slightly to a regular octahedron (Fig. 9.10d). Two ligands bind through primary amines, while the remaining four bridge two adjacent metal centers through bridging μ2-carboxylates. The resulting MOF consists of stacked 2-D zigzag sheets (Fig. 9.10e, f). Tiliakos et al. used an unnatural dipeptide H-Aib-Aib-OH, where Aib is αaminoisobutyric acid, to form two extended coordination complexes with identical formulae but unique connectivities [85]. The first compound was synthesized with CuCl2 and consisted of 1-D chains of square planar Cu2+ each linked to two peptides. When Cu(COOCH3)2 was used as the precursor, a 3-D connected bio-MOF resulted, in which pentacoordinate square pyramidal Cu2+ are bound by three peptide linkers. In both compounds, each Cu2+ is chelated by a tridentate dipeptide via the N-terminal amine, the amide nitrogen, and one of the carboxylate oxygens. The remaining carboxylate oxygen coordinates a second Cu2+. The extra coordination in the 3-D bio-MOF arises from a third linker which binds through the carbonyl of the amide, extending the connectivity into three dimensions. Isostructural 3-D compounds have also been reported with other dipeptides [13, 21, 26, 59]. Rosseinsky and coworkers also developed a series of 3-D dipeptide bio-MOFs and investigated how slight modifications to the amino acid residues affected MOF properties. The first structure, prepared using GlyAla and Zn(NO3)2, consists of 2-D sheets of Zn2+ tetrahedra where each Zn2+ is coordinated by monodentate carboxylates from two linkers and N-terminal amines from an additional two linkers (Fig. 9.11b) [73]. These sheets stack directly on top of one another via hydrogen bonding between adjacent amide groups (Fig. 9.11d). Methanol guest molecules can be removed through activation, resulting in a sharp loss of crystallinity and a shift of major diffraction peaks, indicating structural changes brought about by conformational changes in the peptide linkers. Substituting serine (Ser) for Ala led to an isostructural framework that, because of the hydroxyl side chain, is selective to polar analytes and demonstrates stepwise pore closing due to intraframework hydrogen bond formation upon removal of guest molecules (Fig. 9.11e) [62]. Yet another framework was prepared by replacing Ala with threonine (Thr), where the additional hydroxy group on the side chain now resides in the pores [61]. Small changes in the coordination around Zn2+ ions resulted in a drastic change in the gas adsorption properties. Like the GlyAla MOF, four peptides coordinate to each metal ion; however, the N-termini chelate through both the primary amine and the carbonyl oxygen of the amide, increasing the overall coordination number to six (Fig. 9.11c). Additional hydrogen bonding through the hydroxy group of Thr allows for stronger interactions between the 2-D [Zn(GlyThr)2] layers (Fig. 9.11f). As a result, the material maintains crystallinity after solvent removal and exhibits permanent porosity to CO2. The flexibility observed in Zn-GlyAla allows for the management of mechanical stress and prevents complete collapse of the framework. However, materials that are too flexible risk losing porosity as more possible conformational
9 MOFs Constructed from Biomolecular Building Blocks
305
Fig. 9.11 Dipeptide MOFs formed with (a) GlyAla, GlySer, or GlyThr. (b) Tetrahedral Zn2+ results from GlyAla and GlySer, and (c) octahedral coordination is observed in the GlyThr framework. Note that both coordination complexes include only four peptide ligands. Resulting frameworks of (d) GlyAla, (e) GlySer, and (f) GlyThr demonstrating the effect of side chain functionalities on the pore dimensions and relative hydrophobicity. Dark blue polygons and black, light blue, maroon, and white spheres represent Zn2+, carbon, nitrogen, oxygen, and hydrogen atoms. Hydrogen atoms are omitted from b-c) for clarity
arrangements decrease periodicity. Replacing tyrosine (Tyr) for Thr led to a compound with an intermediate stability where long-range order and crystallinity were completely maintained at pressures up to 4 GPa [68]. Further alteration of the GlyAla peptide led to a significantly more stable structure. The dipeptide carnosine (Car) is composed of L-histidine (His) and β-Ala, which has an additional methylene unit between the C- and N-termini. In a Zn-Car MOF, each Car coordinates four distinct tetrahedral Zn2+ centers via a monodentate carboxylate, N-terminal amine, or the imidazolate nitrogens on the His side chain [42]. The structure exhibits 1-D channels that have minimum and maximum apertures of 3.78 and 5.18 Å, respectively. This microporous 3-D bio-MOF undergoes slight structural changes upon solvent removal and redispersion in solvent. Density functional theory (DFT) studies revealed that DMF and methanol guests engage in similar hydrogen bonding interactions with the framework that cause local changes in the structure without affecting long-range order. The most notable structural deviation occurs upon soaking Zn-Car crystals in water. In this case, Car linkers rotate into two different conformations, resulting in two distinct pore types. Reactivation causes rotation of the Car ligands back to the original conformations in the activated framework. The activated material displays permanent porosity and has a BET surface area of 448 m2g1 as determined by CO2 adsorption. Both CO2 and CH4 have high affinities and strong interactions with ZnCar with experimentally
306
Z. M. Schulte and N. L. Rosi
determined Qst values of 49 and 27 kJmol1, respectively. The inherent flexibility and responsiveness of “peptide” MOFs to different environmental conditions point toward the potential of building MOFs with adaptive cavities for highly specific analyte recognition. Bio-MOFs with tripeptide linkers are much rarer than those constructed from single amino acid linkers or dipeptides. Lee et al. reported one of the first examples in which Cd2+ was linked with tripeptides GlyGlyGly (Gly3) and AlaAlaAla (Ala3) [43]. The former is isostructural to the dipeptide synthesized earlier by Takayama and crystallizes as 2-D sheets of [Cd(Gly)3] which are then interconnected by hydrogen bonding. The three methyl groups of Ala3 led to a loss of dimensionality in the latter compound. The additional steric bulk provided by the Ala residues prevents C-terminal carboxylates from bridging the metal centers, reducing the number of ligands at each Cd2+ from six to four. Hydrogen bonding interactions are only possible in one dimension due to the conformation of Ala3, resulting in 1-D chains that are held together by a combination of hydrogen bonding and hydrophobic interactions. Lee and coworkers were not able to synthesize 3-D tripeptide bio-MOFs, but their work led to a greater understanding of the roles and extents side chain functionality and peptide length play on the formation of extended crystalline systems with metal ions. Rosseinsky and coworkers rationalized that longer linkers would increase the pore metrics and that the additional degrees of freedom provided by the extra amino acid could be used to create even more dynamic porosity than in his previous dipeptide systems [63]. Two isostructural tripeptide MOFs comprising GlyHisGly or GlyHisLys (Lys ¼ lysine) and Cu(II) salts were synthesized, and the crystal structures were determined, revealing a novel form with the former. In both crystals, the Cu2+ adopts a distorted square pyramidal geometry and is coordinated by three distinct peptides, one of which is tridentate through one His nitrogen of imidazole, the adjacent nitrogen from the amide, and the terminal α-amine. The remaining two ligands each bridge two Cu2+ metals via monodentate C-terminal carboxylates. Therefore, [Cu2(GlyHisX)4] dimers serve as the metal nodes stabilized by π-π stacking of adjacent imidazole rings. These dimers are connected into fourfold helical coils which are interconnected by μ2-carboxylate bridges, extending the framework into three dimensions. The resulting network of interconnected 1-D pores is either occupied by solvent in the case of GlyHisGly or by the side chain of the Lys residue. The solvent-accessible volumes in GlyHisGly and GlyHisLys are 60.2% and 47.2%, respectively, both of which are significantly higher than that of Zn-GlyAla at 28%. Although both MOFs experience structural collapse upon solvent removal, the original structure can be recovered by exposing to water vapor, demonstrating a sponge-like behavior. Finally, they successfully performed PSM on the Lys side chain by reacting with ethyl isocyanate to introduce urea groups within the MOF pores. Another strategy to impart additional stability and increase porosity is to add a secondary linker. A common secondary linker, 4,40 -bipyridine, or bipy, is considerably more rigid than individual amino acids or peptides due to its extended π system. In 2010, Rosseinsky and coworkers attempted to incorporate bipy as a secondary
9 MOFs Constructed from Biomolecular Building Blocks
307
Fig. 9.12 NiAsp MOFs with bridged bipy ligands. (a) Octahedral coordination of Ni2+ showing three Asp ligands in either monodentate (forward) or tridentate (rear) modes. One slightly twisted bipy ligand coordinates at an axial site. (b) 2-D sheets are formed from bridging carboxylates which are then interconnected by bipy ligands to make (c) a 3-D framework with one-dimensional channels. Green octahedra and black, blue, and maroon spheres represent Ni2+, carbon, nitrogen, and oxygen atoms. Hydrogen atoms have been excluded for clarity
linker into a Zn-Asp framework [35]. However, in this case, only a Zn-Asp bio-MOF was formed, and bipy was not incorporated into the framework. To overcome this problem, they later performed syntheses with bipy and Ni(L-Asp)3 salts as starting materials and formed [Ni2(L-Asp)2(bipy)] crystals [86]. The chirality of the resulting crystals can be controlled through the ratio of enantiomers used in the synthesis. In both cases, octahedral Ni2+ centers are chelated by a tridentate Asp via one oxygen from both carboxylates and the α-amine; two additional monodentate carboxylate Asp ligands and one pyridyl group from bipy complete the coordination sphere (Fig. 9.12a). The Ni2+ octahedra form 1-D chiral helices that are then pillared by bipy at each Ni2+ site (Fig. 9.12b, c). The BET surface area was measured to be 247 m2g1, which is higher than that of the NiAsp MOF without bipy. Supplemental work by Perez Barrio and others investigated the ability to optimize the pore metrics of the bipy-pillared NiAsp with other secondary linkers, including 4,40 -azobipyridine (apy) and bis(4-pyridyl)ethylene (bpe) [70]. Both of these linkers formed isostructural frameworks with slightly enhanced surface areas. Other secondary linkers such as 1,3,-benzenedicarboxylic acid and azides have been used to prepare 3-D bio-MOFs with interesting magnetic properties [19, 57].
9.3.3
Functionalized Peptides
Ring systems such as triazoles and tetrazoles can coordinate multiple metals in close proximity, resulting in stable SBUs and robust frameworks. Qu et al. developed a method for constructing 3-D homochiral MOFs using amino acids having tetrazole functional groups [72]. Tetrazole was formed in situ during MOF synthesis by reacting (S)-3-cyanophenylalanine with sodium azide and either ZnCl2 or CdCl2. In the resulting frameworks, each ligand links four distinct metal ions together, and chains of metal centers are stabilized by inter-ligand hydrogen bonding. Naik and
308
Z. M. Schulte and N. L. Rosi
Fig. 9.13 (a) 2-D and 3-D renderings of triazole-functionalized glycine. (b) Triangular [Cu3(μ 3-O) (triz-Gly)6(H2O)3] SBUs interlink to form (c) cage-like motifs that further pack into (d) an extended structure with spherical cavities. Red, black, blue, maroon, and white spheres represent copper, carbon, nitrogen, oxygen, and hydrogen atoms, respectively. Hydrogen atoms and BF4 anions are omitted from b-d for clarity. Large yellow spheres indicate void space
coworkers incorporated triazole functionalities at the N-termini of amino acids via transamination (Fig. 9.13a) [64, 65]. In their original work, a hydrogen-bonded coordination network was developed with triazole-functionalized methionine and Zn(II) salts. A 3-D framework with spherical cages was later synthesized utilizing Cu(BF4)2 and triazole-functionalized Gly (Fig. 9.13b–d) [66]. Sawada et al. functionalized the tripeptide GlyProPro at both termini with pyridyl groups [76]. The ligands adopt a polyproline II (PPII) helix, and each terminal pyridyl group coordinates to Ag+. Six Ag-GlyProPro-Ag units arrange in hexagonal coils and result in a porous hexagonal crystalline system with two distinct homochiral channels with different sizes, the larger having a diameter of 2 nm. The framework is cationic, and charge balance is achieved by BF4 anions residing in the pores. The chirality of the channels is defined by the handedness of the amino acids. The homochiral frameworks preferentially adsorb enantiomers of 1,10 -bi-2naphthol with like chirality from racemic mixtures. Mono- and oligosaccharides were also observed to adsorb into the material via hydrogen bonding between the guests and the exposed amide units of the peptide.
9.3.4
Proteins
Proteins are the epitome of biological molecular complexity and exhibit unique properties and functions that depend intrinsically on the sequence and 3-D assembly of their constituent amino acids. Ensembles of different proteins can perform even more complex functions. Therefore, developing rational methods for organizing proteins in 3-D space is an attractive, yet highly challenging, endeavor. Indeed, few examples of proteins as bridging ligands in bio-MOFs exist. Douglas and coworkers used the dodecameric Dps protein from the hypothermophilic archaeon Sulfolobus solfataricus, a single-celled organism found often near volcanoes and hot
9 MOFs Constructed from Biomolecular Building Blocks
309
springs, as a ligand [16]. They incorporated amino acid substitutions and other modifications at very specific sites to allow for metal binding in a controlled fashion that resulted in extended structures. By substituting Ser for Cys at site 126, they effectively disrupted a disulfide linkage between this site and site 101, leaving an exposed Cys at site 101. Further modification of 101Cys with a phenanthroline group yielded 12 metal binding sites. When the modified protein was mixed with Fe2+ in aqueous media, pink aggregates formed. The distribution of aggregate sizes was multimodal and varied by more than 1 μm. Although not crystalline, low amounts of CO2 adsorption into the protein cavities was observed at high pressures, suggesting potential porosity. Tezcan and coworkers have made major contributions to metal coordinationdirected protein assembly. They functionalized the hemeprotein cytochrome cb562 with phenanthroline at specific locations and then used either Zn2+ or Ni2+ to bridge two proteins together [74]. The protein dimers stack into a tetramer through π-π stacking interactions. The tetramers of cb562 assemble into hexagonal rings through helix-helix interactions between the proteins. The resulting extended structure features large channels with an interior diameter of 6 nm. Building off these early developments, they continued to use highly symmetric proteins as substitutes for metal nodes or SBUs with similar geometries [80]. They selected the octahedral, cage-like human heavy-chain ferritin protein and modified Thr for His at the 122 sites (T112Hferritin), forming tridentate “pockets” where embedded Zn2+ could preferentially bind (Fig. 9.14a). Additional modifications of all Cys residues and a specific K86Q mutation were performed to facilitate the crystallization of a facecentered cubic Zn-T112Hferritin structure. From this structure, they were able to determine that the Zn2+ were nearly ideal tetrahedra with a single coordinated water molecule protruding from the surface of the protein (Fig. 9.14b, c). In order to connect these protein SBUs, they incorporated ligand-mediated self-assembly by introducing benzene-1,4-dihydroxamic acid (BDH) to solutions of Zn2+ and T112H ferritin, which yielded a body-centered cubic protein-based bio-MOF (Fig. 9.14d–f). The incorporation of BDH linkers increases the inter-protein distances to >6 Å where only very slight protein-protein interactions are possible. Therefore, the majority of the structural stability arises from the strong coordination bonds between Zn2+, 122His, and BDH. The native structure of the cage-like ferritin results in a highly porous structure with a solvent content of 67%. Translating MOF design and construction strategies to protein-MOFs, Tezcan produced a family of M2+-ferritin bio-MOFs with Ni2+, Zn2+, and Co2+ and a variety of secondary linkers [12]. A total of 15 different analogues were synthesized using ditopic ligands with varying flexibility and angles between the coordinating hydroxamate groups. Similar to Zn-T112Hferritin, Co2+ adopted a tetrahedral coordination geometry and formed a cubic framework with BDH. The Ni2+ analogue, however, crystallized with a tetragonal symmetry due to the octahedral geometry around the metal center. Ligands with parallel but slightly offset hydroxamate moieties (E-ethylenedihydroxamate and naphthalene-2,6-dihydroxamate) formed
310
Z. M. Schulte and N. L. Rosi
Fig. 9.14 (a) Schematic showing the different products of ferritin-Zn with and without a secondary ligand. (b) Dodecahedral structure of ferritin and (c) enlarged Zn2+ coordination sites where Zn2+ and exposed histidine residues are represented by blue spheres and pink lines, respectively. (d) Connectivity of Zn-T112Hferritin with BDH linkers with (e) inter-protein distances and (f) specific coordination bond lengths
tetragonal structures regardless of the metal ion. Benzene-1,3-dihydroxamate is bent at approximately 120 and follows the reverse pattern of lattice symmetry to that of BDH. The last linker they utilized, xylene-1,4-dihydroxamate, is nearly identical to BDH but with methylene units between the coordinating groups and the benzene ring, providing access to different molecular conformations due to rotation. This linker yielded a mixture of tetragonal structures with Co2+, while Zn2+ crystallized as either cubic or tetragonal frameworks. Only cubic frameworks were formed with Ni2 + and xylene-1,4-dihydroxamate.
9.4
Saccharides
Highly functionalized with hydroxyl groups and terminating carboxylates, saccharides are amenable to metal coordination (Fig. 9.15). Edible, cheap, flexible, and biodegradable, saccharides are ideal candidates for organic linkers in bio-MOFs. Numerous MOF structures have been synthesized with flexible, unfunctionalized linear dicarboxylic acids such as succinic or fumaric acid, but these will not be discussed here. Instead, this section highlights bio-MOFs with oxidized monosaccharides or cyclic oligosaccharides as organic ligands.
9 MOFs Constructed from Biomolecular Building Blocks
311
Fig. 9.15 Examples of naturally occurring hexoses and their oxidized products. Deprotonation of glucaric acid (top right) and mucic acid (bottom right) yield glucarate and mucate, respectively. All structures are shown as D-isomers
9.4.1
Bio-MOFs Constructed from Simple Sugars
Discrete complexes and lower dimensional assemblies of metal ions and sugars were investigated using crystallographic methods in the late 1970s. It was well known that certain saccharides, namely, glucaric acid, or saccharic acid, interacted strongly in solution with alkaline ions. Taga and coworkers determined the crystal structures of calcium and potassium complexes with glucaric acid using X-ray diffraction and spectroscopic techniques [82, 83]. In both monoclinic structures, the carboxylate moieties at the ends of the glucarate ion chelate to the metal centers along with αhydroxyl groups. The crystal structure extends into three dimensions through hydrogen bonding between other hydroxyl groups on the glucarate ions and coordinated water molecules. Ferrier and coworkers expanded upon the work of Taga by substituting a transition metal for alkaline metals [24]. This led to the discovery of a triclinic Cu-glucarate crystal system that was similar but structurally unique to the Ca+ and K+ systems. Each Cu2+ is octahedral and coordinated to two glucarates, forming one-dimensional chains. Both α-hydroxy and carboxylate groups coordinate at the equatorial sites, and two additional coordinated water molecules occupy the apical sites. Although all structures were non-porous, the aforementioned studies helped researchers understand how sugars could be used to develop functional MOFs and provided early evidence of how the inherent flexibility of saccharides could be harnessed to form unique extended three-dimensional structures when coordinated to metal ions.
312
Z. M. Schulte and N. L. Rosi
Robson and coworkers prepared one of the earliest examples of a porous transition metal-saccharide bio-MOF by mixing potassium hydrogen saccharate with Zn (COOCH3)2 in an aqueous solution, resulting in single crystals of Zn-saccharate [1, 2]. Each Zn2+ center is coordinated by four distinct saccharate ions through the carboxylate moieties and through one α-hydroxy group, resulting in four connected SBUs. This Zn-saccharate MOF crystallizes in a tetragonal space group and possesses two distinct one-dimensional channels due to the chirality of the D-saccharate ligand. One pore environment is functionalized with hydroxyl groups along the channels, and the other, conversely, is lined with aliphatic moieties, creating a checkerboard pattern of alternating hydrophilic and hydrophobic channels. Differences between the two channels were highlighted by exposing the MOF crystals to molten azobenzene, I2 vapor, and other guest species and then repeating single crystal diffraction studies. In all cases, hydrophobic guest molecules reside in the hydrophobic channels, while the hydrophilic channels contain coordinated water molecules. Two new metal-saccharide MOFs were synthesized by Robson using mucic acid and Ln(NO3)3 (Ln ¼ La, Ce, Pr, Nd) [1, 2]. Four isostructural compounds having the formula [Ln2(muc)3•8H2O] were synthesized, where Ln ¼ La3+, Ce3+, Pr3+, or Nd3+. Each Ln3+ coordinates to six mucate ligands through the carboxylate groups. Each carboxylate on the mucate bridges two Ln3+, forming rectangular “panels” where each corner is a Ln3+ ion. These panels stack in one-dimensional infinite rods along the c-axis, resulting in the formation of long hexagonal channels. Another crystalline system with the formula [Ln2(H2O)4(muc)3•10H2O] was discovered when Nd3+ or Eu3+ were mixed with potassium mucate. The orthorhombic crystals consist of 2-D sheets of hexacoordinate Ln3+ centers interconnected by three mucate ligands. Two ligands coordinate in a bidentate fashion through one carboxylate oxygen and one αhydroxyl group, while the third ligand is tridentate with additional coordination from a β-hydroxyl group. Therefore, the geometry around each metal center is simplified to T-shaped. Each two-dimensional sheet is stacked in an ABAB fashion via hydrogen bonding between free hydroxyl groups of the mucate ligands and water molecules. The resulting structure has pseudo-square channels of similar dimensions to that of Zn-saccharate crystals.
9.4.2
Cyclodextrin Bio-MOFs
All of the saccharide MOFs listed above were synthesized using simple carbohydrates containing a backbone of five or six carbons and terminating in carboxylate moieties. However, saccharides found in nature can be much more complex and include multiple saccharide units, e.g., oligosaccharides. Cyclodextrins are examples of cyclic oligosaccharides composed of multiple alpha-D-glucopyranose units (Fig. 9.16). These biomolecules form a toroidal, or doughnut, shape with a hydrophilic exterior and a relatively hydrophobic interior (Fig. 9.17a). Natural cyclodextrins can be found in three different sizes, α-, β-, and γ-cyclodextrins, which have
9 MOFs Constructed from Biomolecular Building Blocks
313
Fig. 9.16 Structures of cyclodextrins. The primary faces are projecting out of the plane of the page towards the reader, and the secondary face is projecting through the back of the same plane
Fig. 9.17 Structure of CD-MOF-1. (a) Schematic of CD tori showing the primary and secondary faces and corresponding hydroxyl groups available for coordination. (b) Gray and red sticks represent carbon and oxygen atoms, respectively, and K+ is represented by purple spheres. (c) Cubic SBU of CD-MOF-1 where each different color represents a distinct CD and (d) stacking of cubes to generate the extended cubic framework, shown as (e) tiling and (f) the underlying topological net
6, 7, and 8 monosaccharide units, respectively. Cyclodextrins have been used in the food, cosmetics, and pharmaceutical industries. A detailed review published by Norkus describes the diverse coordination chemistry of cyclodextrins [69]. However, in most of these examples the metal ions formed inclusive complexes within the center of the cyclodextrin tori, thus inhibiting the formation of extended networks through coordinative bonding.
314
Z. M. Schulte and N. L. Rosi
Stoddart and coworkers developed a series of cyclodextrin-MOFs (CD-MOFs) [79]. CD-MOF-1 forms as colorless cubic crystals with the empirical formula [(CD) (KOH)2] upon mixing γ-cyclodextrin and an aqueous solution of KOH. Each γ-CD coordinates to eight K+, four of which are on the primary and secondary faces, respectively. Six γ-CD orient to form a cube (γ-CD)6 with the primary faces facing inward (Fig. 9.17b, c). The secondary faces of the CD lie at the six faces of the cube, projecting outward and serving as points of extensions to other (γ-CD)6 units. The extended structure is generated by repeated stacking of multiple cubes in three dimensions, yielding a face-centered cubic crystalline system with topology rra (Fig. 9.17d–f). CD-MOF-1 can be prepared from entirely food-grade starting materials and solvents, which is important for the development of consumable “foodgrade” MOFs. To date, CD-MOF-1 remains a widely investigated bio-MOF for a variety of applications including drug delivery, CO2 capture, and controllable release of fragrances [30, 48, 95]. After the initial report of CD-MOF-1, numerous other CD-MOFs have been developed [25]. The use of different metal cations and different CD ring systems has led to significant structural diversity. For example, combining Rb+ with α-CD yielded large needles with visibly hollow centers, termed a CD-MCF (MCF ¼ metal coordination framework) [31]. Structural analysis by single-crystal XRD revealed long, adjacent coils of α-CD tori connected by Rb+ coordinated on both the top and bottom edges. The overlap of coils within CD-MCF produces a series of channels that are uniformly chiral, thus yielding a material with small, left-handed, porous channels. Sun and coworkers combined sodium oxalate and β-CD to form a material bearing some structural similarities to CD-MCF [55]. Coiling stacks of β-CD are connected by shared Na+ coordination resulting in helical chains with a pitch of 10.353 Å. Each helical chain is then surrounded by four equivalent chains, resulting in porous channels. Sha et al. developed an α-CD-MOF with chiral channels utilizing KOH as the metal source [77]. Dimeric stacks of α-CD tori fused in a tail-to-tail arrangement are coordinated by shared K+. Adjacent dimers are bridged by K+ resulting in two-dimensional sheets. Individual sheets then stack in a staggered ABAB pattern, resulting in a 3-D structure. The uptake of the anti-cancer drug molecule 5-fluorouracil (5-FU) was investigated for both the β-CD and α-CD bio-MOFs described above [55, 77]. A template-design approach was employed by Liu et al. where two distinct β-CDMOFs, here named β-CD-MOF-2 and β-CD-MOF-3, could be selectively synthesized by changing the templating molecule [54]. When 1,2,3-triazole-4,5-dicarboxylic acid (H3tzdc) was mixed with Cs salts and β-CD, β-CD-MOF-2 was synthesized. By using comparatively larger templating agents ibuprofen or methyl benzene sulfonic acid (TsOH), β-CD-MOF-3 resulted as the sole product. Both MOFs are helical but differ greatly in terms of their connectivity. Both MOFs were studied for drug loading and delivery with 5-FU and compared to solvated βCD matrices. The more porous β-CD-MOF-3 had the highest loadings of 1.51 g g1, which is higher than mesoporous silica and other previously reported MOFs.
9 MOFs Constructed from Biomolecular Building Blocks
9.5
315
Conclusions and Future Outlook
As MOFs progress further toward real-world applications and more widespread use, consideration of their composition, and more specifically their environmental compatibility, will become increasingly important. For applications where the MOF is intended to be disposable, such as one-use sensors, food additives, etc., bio-MOFs may become particularly relevant. In the biomedical field, the advantage of using a biocompatible bio-MOF is clear: upon degradation in the body, the MOF components would likely be benign and not have unintended deleterious effects. More exploration of their use in drug delivery is anticipated. While structure design in MOF chemistry, in general, is well-established, designing bio-MOFs has proven more challenging because of the diverse and often unpredictable coordination modes between biomolecules and metal ions. The rational design of bio-MOF materials is certainly a growth area and will mature as more and more bio-MOF examples are reported. Perhaps the most exciting opportunity within the bio-MOF field is the construction of evermore complex bio-MOF materials, such as bio-MOFs that incorporate multiple different biomolecular linkers (i.e., multivariate bio-MOFs) [22]. With the discovery and development of such materials, we may potentially begin to realize 3-D periodic biological systems whose properties depend on the sequence of biomolecules within the structure.
References 1. Abrahams BF, Moylan M, Orchard SD, Robson R (2003a) Zinc Saccharate: A robust, 3D coordination network with two types of isolated, parallel channels, one hydrophilic and the other hydrophobic. Angew. Chem. Int. Ed. 42:1848–1851 2. Abrahams BF, Moylan M, Orchard SD, Robson R (2003b) Channel-containing lanthanide Mucate structures. CrystEngComm 5:313–317 3. An J, Rosi NL (2010) Tuning MOF CO2 adsorption properties via cation exchange. J. Am. Chem. Soc. 132:5578–5579 4. An J, Geib SJ, Rosi NL (2009) Cation-triggered drug release from a porous zincAdeninate metalorganic framework. J. Am. Chem. Soc. 131:8376–8377 5. An J, Geib SJ, Rosi NL (2010) High and selective CO2 uptake in a cobalt Adeninate metalorganic framework exhibiting pyrimidine- and amino-decorated pores. J. Am. Chem. Soc. 132:38–39 6. An J, Shade CM, Chengelis-Czegan DA, Petoud S, Rosi NL (2011) Zinc-Adeninate metalorganic framework for aqueous encapsulation and sensitization of near-infrared and visible emitting lanthanide cations. J. Am. Chem. Soc. 133:1220–1223 7. An J, Farha OK, Hupp JT, Pohl E, Yeh JI, Rosi NL (2012) Metal-adeninate vertices for the construction of an exceptionally porous metal-organic framework. Nat. Commun. 3:604 8. Anderson SL, Stylianou KC (2017) Biologically derived metal organic frameworks. Coord. Chem. Rev. 349:102–128 9. Anderson SL, Boyd PG, Gładysiak A, Nguyen TN, Palgrave RG, Kubicki D, Emsley L, Bradshaw D, Rosseinsky MJ, Smit B, Stylianou KC (2019) Nucleobase pairing and photodimerization in a biologically derived metal-organic framework nanoreactor. Nat. Commun. 10:1612
316
Z. M. Schulte and N. L. Rosi
10. Anokhina EV, Jacobson AJ (2004) [Ni2O(l-Asp)(H2O)2]4H2O: A Homochiral 1D helical chain hybrid compound with extended NiONi bonding. J. Am. Chem. Soc. 126:3044–3045 11. Anokhina EV, Go YB, Lee Y, Vogt T, Jacobson AJ (2006) Chiral three-dimensional microporous nickel aspartate with extended NiONi bonding. J. Am. Chem. Soc. 128:9957–9962 12. Bailey JB, Zhang L, Chiong JA, Ahn S, Tezcan FA (2017) Synthetic modularity of protein– metal–organic frameworks. J. Am. Chem. Soc. 139:8160–8166 13. Bear CA, Freeman HC (1976) Crystallographic studies of metal-peptide complexes. VIII. Glycyl-l-methioninatocopper(II). Acta Crystallogr. Sect. B Struct. Sci. 32:2534–2536 14. Bernini MC, Fairen-Jimenez D, Pasinetti M, Ramirez-Pastor AJ, Snurr RQ (2014) Screening of bio-compatible metal–organic frameworks as potential drug carriers using Monte Carlo simulations. J. Mater. Chem. B 2:766–774 15. Bonnefoy J, Legrand A, Quadrelli EA, Canivet J, Farrusseng D (2015) Enantiopure peptidefunctionalized metal–organic frameworks. J. Am. Chem. Soc. 137:9409–9416 16. Broomell CC, Birkedal H, Oliveira CLP, Pedersen JS, Gertenbach J-A, Young M, Douglas T (2010) Protein cage nanoparticles as secondary building units for the synthesis of 3-dimensional coordination polymers. Soft Matter 6:3167–3171 17. Cai H, Li M, Lin X-R, Chen W, Chen G-H, Huang X-C, Li D (2015) Spatial, hysteretic, and adaptive host–guest chemistry in a metal–organic framework with open Watson–crick sites. Angew. Chem. Int. Ed. 54:10454–10459 18. Carmona FJ, Rojas S, Sánchez P, Jeremias H, Marques AR, Romão CC, Choquesillo-Lazarte D, Navarro JAR, Maldonado CR, Barea E (2016) Cation exchange strategy for the encapsulation of a photoactive CO-releasing organometallic molecule into anionic porous frameworks. Inorg. Chem. 55:6525–6531 19. Chen Z-L, Jiang C-F, Yan W-H, Liang F-P, Batten SR (2009) Three-dimensional metal Azide coordination polymers with amino carboxylate Coligands: Synthesis, structure, and magnetic properties. Inorg. Chem. 48:4674–4684 20. Crawford SE, Gan XY, Lemaire PCK, Millstone JE, Baltrus JP, Ohodnicki PR (2019) ZincAdeninate metal–organic framework: A versatile Photoluminescent sensor for rare earth elements in aqueous systems. ACS Sens. 4:1986–1991 21. Dehand J, Jordanov J, Keck F, Mosset A, Bonnet JJ, Galy J (1979) Synthesis, crystal structure, and electronic properties of (L-methionylglycinato)copper(II). Inorg. Chem. 18:1543–1549 22. Deng H, Doonan CJ, Furukawa H, Ferreira RB, Towne J, Knobler CB, Wang B, Yaghi OM (2010) Multiple functional groups of varying ratios in metal-organic frameworks. Science 327:846 23. Ferrer P, da Silva I, Rubio-Zuazo J, Castro GR (2014) Synthesis and crystal structure of the novel metal organic framework Zn(C3H5NO2S)2. Powder Diffract. 29:366–370 24. Ferrier F, Avezou A, Terzian G, Benlian D (1998) Synthesis and crystal structure of copper (II) d-glucarate, tetra-hydrate. J. Mol. Struct. 442:281–284 25. Forgan RS, Smaldone RA, Gassensmith JJ, Furukawa H, Cordes DB, Li Q, Wilmer CE, Botros YY, Snurr RQ, Slawin AMZ, Stoddart JF (2012) Nanoporous carbohydrate metal–organic frameworks. J. Am. Chem. Soc. 134:406–417 26. Freeman HC, Healy MJ, Scudder ML (1977) A crystallographic study of the structure of glycylL-alaninatocopper(II) hydrate and the conformations of copper(II) – dipeptide complexes. J. Biol. Chem. 252:8840–8847 27. Fu H-R, Zhang J (2015) Structural transformation and hysteretic sorption of light hydrocarbons in a flexible Zn–Pyrazole–adenine framework. Chem. Eur. J. 21:5700–5703 28. Furukawa H, Cordova KE, O’Keeffe M, Yaghi OM (2013) The chemistry and applications of metal-organic frameworks. Science 341:1230444 29. García-Terán JP, Castillo O, Luque A, García-Couceiro U, Román P, Lezama L (2004) An unusual 3D coordination polymer based on bridging interactions of the nucleobase adenine. Inorg. Chem. 43:4549–4551
9 MOFs Constructed from Biomolecular Building Blocks
317
30. Gassensmith JJ, Furukawa H, Smaldone RA, Forgan RS, Botros YY, Yaghi OM, Stoddart JF (2011) Strong and reversible binding of carbon dioxide in a green metal–organic framework. J. Am. Chem. Soc. 133:15312–15315 31. Gassensmith JJ, Smaldone RA, Forgan RS, Wilmer CE, Cordes DB, Botros YY, Slawin AMZ, Snurr RQ, Stoddart JF (2012) Polyporous metal-coordination frameworks. Org. Lett. 14:1460–1463 32. Gillen K, Jensen R, Davidson N (1964) Binding of silver ion by adenine and substituted adenines. J. Am. Chem. Soc. 86:2792–2796 33. Giménez-Marqués M, Hidalgo T, Serre C, Horcajada P (2016) Nanostructured metal–organic frameworks and their bio-related applications. Coord. Chem. Rev. 307:342–360 34. González Miera G, Bermejo Gómez A, Chupas PJ, Martín-Matute B, Chapman KW, PlateroPrats AE (2017) Topological transformation of a metal–organic framework triggered by ligand exchange. Inorg. Chem. 56:4576–4583 35. Gould JA, Jones JTA, Bacsa J, Khimyak YZ, Rosseinsky MJ (2010) A homochiral threedimensional zinc aspartate framework that displays multiple coordination modes and geometries. Chem. Commun. 46:2793–2795 36. Gramaccioli CM (1966a) Acta Cryst 21:600–605 37. Gramaccioli CM (1966b) Acta Cryst 21:594–600 38. Gutov OV, Molina S, Escudero-Adán EC, Shafir A (2016) Modulation by amino acids: Toward superior control in the synthesis of zirconium metal–organic frameworks. Chem. Eur. J. 22:13582–13587 39. Hayashi H, Côté AP, Furukawa H, O’Keeffe M, Yaghi OM (2007) Zeolite A imidazolate frameworks. Nat. Mater. 6:501–506 40. Imaz I, Rubio-Martinez M, An J, Sole-Font I, Rosi NL, Maspoch D (2011) Metal-biomolecule frameworks (MBioFs). Chem. Commun. 47:7287–7302 41. Kalmutzki MJ, Hanikel N, Yaghi OM (2018) Secondary building units as the turning point in the development of the reticular chemistry of MOFs. Sci. Adv. 4:eaat9180 42. Katsoulidis AP, Park KS, Antypov D, Martí-Gastaldo C, Miller GJ, Warren JE, Robertson CM, Blanc F, Darling GR, Berry NG, Purton JA, Adams DJ, Rosseinsky MJ (2014) Guest-adaptable and water-stable peptide-based porous materials by imidazolate side chain control. Angew. Chem. Int. Ed. Engl. 53:193–198 43. Lee H-Y, Kampf JW, Park KS, Marsh ENG (2008) Covalent metalpeptide framework compounds that extend in one and two dimensions. Cryst. Growth Des. 8:296–303 44. Li T, Rosi NL (2013) Screening and evaluating aminated cationic functional moieties for potential CO2 capture applications using an anionic MOF scaffold. Chem. Commun. 49:11385–11387 45. Li T, Chen D-L, Sullivan JE, Kozlowski MT, Johnson JK, Rosi NL (2013a) Systematic modulation and enhancement of CO2: N2 selectivity and water stability in an isoreticular series of bio-MOF-11 analogues. Chem. Sci. 4:1746–1755 46. Li T, Kozlowski MT, Doud EA, Blakely MN, Rosi NL (2013b) Stepwise ligand exchange for the preparation of a family of mesoporous MOFs. J. Am. Chem. Soc. 135:11688–11691 47. Li T, Sullivan JE, Rosi NL (2013c) Design and preparation of a Core–Shell metal–organic framework for selective CO2 capture. J. Am. Chem. Soc. 135:9984–9987 48. Li X, Guo T, Lachmanski L, Manoli F, Menendez-Miranda M, Manet I, Guo Z, Wu L, Zhang J, Gref R (2017) Cyclodextrin-based metal-organic frameworks particles as efficient carriers for lansoprazole: Study of morphology and chemical composition of individual particles. Int. J. Pharm. 531:424–432 49. Li J, Jiang L, Chen S, Kirchon A, Li B, Li Y, Zhou H-C (2019) Metal–organic framework containing planar metal-binding sites: Efficiently and cost-effectively enhancing the kinetic separation of C2H2/C2H4. J. Am. Chem. Soc. 141:3807–3811 50. Liu C, Rosi NL (2017) Ternary gradient metal-organic frameworks. Faraday Discuss. 201:163–174
318
Z. M. Schulte and N. L. Rosi
51. Liu C, Li T, Rosi NL (2012) Strain-promoted “click” modification of a mesoporous metal– organic framework. J. Am. Chem. Soc. 134:18886–18888 52. Liu C, Luo T-Y, Feura ES, Zhang C, Rosi NL (2015) Orthogonal ternary functionalization of a mesoporous metal–organic framework via sequential Postsynthetic ligand exchange. J. Am. Chem. Soc. 137:10508–10511 53. Liu C, Zeng C, Luo T-Y, Merg AD, Jin R, Rosi NL (2016) Establishing porosity gradients within metal–organic frameworks using partial Postsynthetic ligand exchange. J. Am. Chem. Soc. 138:12045–12048 54. Liu J, Bao T-Y, Yang X-Y, Zhu P-P, Wu L-H, Sha J-Q, Zhang L, Dong L-Z, Cao X-L, Lan Y-Q (2017) Controllable porosity conversion of metal-organic frameworks composed of natural ingredients for drug delivery. Chem. Commun. 53:7804–7807 55. Lu H, Yang X, Li S, Zhang Y, Sha J, Li C, Sun J (2015) Study on a new cyclodextrin based metal–organic framework with chiral helices. Inorg. Chem. Commun. 61:48–52 56. Luo T-T, Hsu L-Y, Su C-C, Ueng C-H, Tsai T-C, Lu K-L (2007) Deliberate design of a 3D Homochiral CuII/l-met/AgI coordination network based on the distinct softhard recognition principle. Inorg. Chem. 46:1532–1534 57. Luo F, Yang Y-t, Che Y-x, Zheng J-m (2008) Construction of Cu(II)–Gd(III) metal–organic framework by the introduction of a small amino acid molecule: Hydrothermal synthesis, structure, thermostability, and magnetic studies. CrystEngComm 10:1613–1616 58. Ma H-F, Liu Q-Y, Wang Y-L, Yin S-G (2017) A water-stable anionic metal–organic framework constructed from columnar zinc-Adeninate units for highly selective light hydrocarbon separation and efficient separation of organic dyes. Inorg. Chem. 56:2919–2925 59. Maccarrone G, Nardin G, Randaccio L, Tabbi G, Rosi M, Sgamellotti A, Rizzarelli E, Zangrando E (1996) Structure of copper(II) complexes with L-leucyl-D- or L-leucyl-L-phenylalanine and molecular orbital calculations on their stabilization. J. Am. Chem. Soc. Dalton Trans. 16:3449–3453 60. Marshall RJ, Hobday CL, Murphie CF, Griffin SL, Morrison CA, Moggach SA, Forgan RS (2016) Amino acids as highly efficient modulators for single crystals of zirconium and hafnium metal–organic frameworks. J. Mater. Chem. A 4:6955–6963 61. Martí-Gastaldo C, Warren JE, Stylianou KC, Flack NLO, Rosseinsky MJ (2012) Enhanced stability in rigid peptide-based porous materials. Angew. Chem. Int. Ed. 51:11044–11048 62. Martí-Gastaldo C, Antypov D, Warren JE, Briggs ME, Chater PA, Wiper PV, Miller GJ, Khimyak YZ, Darling GR, Berry NG, Rosseinsky MJ (2014) Side-chain control of porosity closure in single- and multiple-peptide-based porous materials by cooperative folding. Nat. Chem. 6:343 63. Martí-Gastaldo C, Warren JE, Briggs ME, Armstrong JA, Thomas KM, Rosseinsky MJ (2015) Sponge-like behaviour in Isoreticular cu(Gly-his-X) peptide-based porous materials. Chem. Eur. J. 21:16027–16034 64. Naik AD, Dîrtu MM, Léonard A, Tinant B, Marchand-Brynaert J, Su B-L, Garcia Y (2010) Engineering three-dimensional chains of porous nanoballs from a 1,2,4-Triazole-carboxylate supramolecular synthon. Cryst. Growth Des. 10:1798–1807 65. Naik AD, Beck J, Dîrtu MM, Bebrone C, Tinant B, Robeyns K, Marchand-Brynaert J, Garcia Y (2011a) Zinc complexes with 1,2,4-triazole functionalized amino acid derivatives: Synthesis, structure and β-lactamase assay. Inorg. Chim. Acta 368:21–28 66. Naik AD, Dîrtu MM, Railliet AP, Marchand-Brynaert J, Garcia Y (2011b) Coordination polymers and metal organic frameworks derived from 1,2,4-Triazole amino acid linkers. Polymers 3:1750–1775 67. Navarro JAR, Lippert B (1999) Molecular architecture with metal ions, nucleobases and other heterocycles. Coord. Chem. Rev. 185–186:653–667 68. Navarro-Sánchez J, Mullor-Ruíz I, Popescu C, Santamaría-Pérez D, Segura A, Errandonea D, González-Platas J, Martí-Gastaldo C (2018) Peptide metal–organic frameworks under pressure: Flexible linkers for cooperative compression. Dalton Trans. 47:10654–10659
9 MOFs Constructed from Biomolecular Building Blocks
319
69. Norkus E (2009) Metal ion complexes with native cyclodextrins. An overview. J. Incl. Phenom. Macrocycl. Chem. 65:237 70. Perez Barrio J, Rebilly J-N, Carter B, Bradshaw D, Bacsa J, Ganin AY, Park H, Trewin A, Vaidhyanathan R, Cooper AI, Warren JE, Rosseinsky MJ (2008) Control of porosity geometry in amino acid derived Nanoporous materials. Chem. Eur. J. 14:4521–4532 71. Pérez-Yáñez S, Beobide G, Castillo O, Cepeda J, Luque A, Aguayo AT, Román P (2011) Openframework copper Adeninate compounds with three-dimensional microchannels tailored by aliphatic monocarboxylic acids. Inorg. Chem. 50:5330–5332 72. Qu Z-R, Zhao H, Wang X-S, Li Y-H, Song Y-M, Y-j L, Ye Q, Xiong R-G, Abrahams BF, Xue Z-L, You X-Z (2003) Homochiral Zn and cd coordination polymers containing amino acidTetrazole ligands. Inorg. Chem. 42:7710–7712 73. Rabone J, Yue YF, Chong SY, Stylianou KC, Bacsa J, Bradshaw D, Darling GR, Berry NG, Khimyak YZ, Ganin AY, Wiper P, Claridge JB, Rosseinsky MJ (2010) An adaptable peptidebased porous material. Science 329:1053 74. Radford RJ, Lawrenz M, Nguyen PC, McCammon JA, Tezcan FA (2011) Porous protein frameworks with unsaturated metal centers in sterically encumbered coordination sites. Chem. Commun. 47:313–315 75. Ren J, Lan PC, Chen M, Zhang W, Ma S (2019) Heterogenization of Trinuclear palladium complex into an anionic metal–organic framework through Postsynthetic cation exchange. Organometallics 38:3460–3465 76. Sawada T, Matsumoto A, Fujita M (2014) Coordination-driven folding and assembly of a short peptide into a protein-like two-nanometer-Sized Channel. Angew. Chem. Int. Ed. 53:7228–7232 77. Sha J-Q, Zhong X-H, Wu L-H, Liu G-D, Sheng N (2016) Nontoxic and renewable metal– organic framework based on α-cyclodextrin with efficient drug delivery. RSC Adv. 6:82977–82983 78. Shearer GC, Vitillo JG, Bordiga S, Svelle S, Olsbye U, Lillerud KP (2016) Functionalizing the defects: Postsynthetic ligand exchange in the metal organic framework UiO-66. Chem. Mater. 28:7190–7193 79. Smaldone RA, Forgan RS, Furukawa H, Gassensmith JJ, Slawin AMZ, Yaghi OM, Stoddart JF (2010) Metal–organic frameworks from edible natural products. Angew. Chem. Int. Ed. 49:8630–8634 80. Sontz PA, Bailey JB, Ahn S, Tezcan FA (2015) A metal organic framework with spherical protein nodes: Rational chemical design of 3D protein crystals. J. Am. Chem. Soc. 137:11598–11601 81. Spore AB, Rosi NL (2017) Effect of countercation on the water stability of an anionic metal– organic framework. CrystEngComm 19:5417–5421 82. Taga T, Osaki K (1976) Interactions of calcium ions with carbohydrates: Crystal structure of calcium D-Glucarate Tetrahydrate. Bull. Chem. Soc. Jpn. 49:1517–1520 83. Taga T, Kuroda Y, Osaki K (1977) Interactions of calcium ions with carbohydrates: X-ray diffraction and NMR spectroscopic studies on the potassium salt and the calcium salt of D-Clucaric acid. Bull. Chem. Soc. Jpn. 50:3079–3083 84. Takayama T, Ohuchida S, Koike Y, Watanabe M, Hashizume D, Ohashi Y (1996) Structural analysis of cadmium–Glycylglycine complexes studied by X-ray diffraction and high resolution 113Cd and 13C solid state NMR. Bull. Chem. Soc. Jpn. 69:1579–1586 85. Tiliakos M, Raptis D, Terzis A, Raptopoulou CP, Cordopatis P, Manessi-Zoupa E (2002) Metal complexes of dipeptides containing the α-aminoisobutyric residue (Aib): Preparation, characterization and crystal structures of copper(II) complexes with H–Aib–Aib–OH. Polyhedron 21:229–238 86. Vaidhyanathan R, Bradshaw D, Rebilly J-N, Barrio JP, Gould JA, Berry NG, Rosseinsky MJ (2006) A family of Nanoporous materials based on an amino acid backbone. Angew. Chem. Int. Ed. 45:6495–6499
320
Z. M. Schulte and N. L. Rosi
87. Wang R, Liu H, Carducci MD, Jin T, Zheng C, Zheng Z (2001) Lanthanide coordination with α-amino acids under near physiological pH conditions: Polymetallic complexes containing the Cubane-like [Ln4(μ3-OH)4]8+ cluster Core. Inorg. Chem. 40:2743–2750 88. Wang F, Tan Y-X, Yang H, Zhang H-X, Kang Y, Zhang J (2011) A new approach towards tetrahedral imidazolate frameworks for high and selective CO2 uptake. Chem. Commun. 47:5828–5830 89. Wang C, Tian L, Zhu W, Wang S, Wang P, Liang Y, Zhang W, Zhao H, Li G (2017) Dye@bioMOF-1 composite as a dual-emitting platform for enhanced detection of a wide range of explosive molecules. ACS Appl. Mater. Interfaces 9:20076–20085 90. Weng H, Xu X-Y, Yan B (2017) Novel multi-component photofunctional nanohybrids for ratio-dependent oxygen sensing. J. Colloid Interface Sci. 502:8–15 91. Wiśniewski M, Bieniek A, Roszek K, Czarnecka J, Bolibok P, Ferrer P, da Silva I, Terzyk AP (2018) Cystine-based MBioF for maintaining the antioxidant–oxidant balance in airway diseases. ACS Med. Chem. Lett. 9:1280–1284 92. Xie Z, Li T, Rosi NL, Carreon MA (2014) Alumina-supported cobalt-adeninate MOF membranes for CO2/CH4 separation. J. Mater. Chem. A 2:1239–1241 93. Yaghi OM, O’Keeffe M, Ockwig NW, Chae HK, Eddaoudi M, Kim J (2003) Reticular synthesis and the design of new materials. Nature 423:705–714 94. Yang E, Li H-Y, Wang F, Yang H, Zhang J (2013) Enhancing CO2 adsorption enthalpy and selectivity via amino functionalization of a tetrahedral framework material. CrystEngComm 15:658–661 95. Zhang B, Huang J, Liu K, Zhou Z, Jiang L, Shen Y, Zhao D (2019) Biocompatible Cyclodextrin-based metal–organic frameworks for long-term sustained release of fragrances. Ind. Eng. Chem. Res. 58:19767–19777
Chapter 10
Natural Polymer-Based MOF Composites Tanay Kundu, Bikash Garai, and Stefan Kaskel
10.1
Introduction
Natural polymers are one of the most important constituents in modern life. The most widely available polymers are cellulose, silk fibroin, and chitin [1]. Cellulose is a major component in the textile, paper, and fabric industry, in the form of pulp and cotton. They have many advantages for being inexpensive, pliable in desired shape to suit application [2], large production backed with fabrication technology, and established protocol for functionalization/composite stabilization. MOF particles can incorporate porosity, crystallinity, color, functionality, and selectivity in the composite with natural polymers by acting as filler materials, when added as powder or grown internally [3, 4]. However, the difference in physical properties of MOFs and natural polymers can lead to phase separation and reduced activity in composites. Functionalization, order of component addition, and nucleation growth points are the most common strategies to introduce homogeneous distribution of MOF particles in the polymer matrix [5]. These aspects prevent MOF leaching and ensure long-term durability of the composites. This can also afford larger percentage of MOF loading in the composite for better performances. Various fabrication methods are used to prepare composites of diverse shapes, sizes, and content, aimed at suitability in specific applications. Deposition methods also play an important role in determining the accessible surface area and the effectiveness of the composites. For example, deposition of the MOF as thin films on natural polymers enables the complete use of pore volume, while encapsulation into a matrix reduces the active surface. Finally, due to the flexible, durable, and non-toxic nature
T. Kundu · B. Garai · S. Kaskel (*) Technische Universität Dresden, Dresden, Germany e-mail: [email protected]; [email protected]; [email protected] © Springer Nature Switzerland AG 2021 P. Horcajada Cortés, S. Rojas Macías (eds.), Metal-Organic Frameworks in Biomedical and Environmental Field, https://doi.org/10.1007/978-3-030-63380-6_10
321
322
T. Kundu et al.
of the composites, applications in both liquid and gaseous state as well as in physiological media can be envisaged [6]. In this chapter, MOF-natural polymer composites with three major types of polymers, viz., cellulose, silk, and chitin, and their subclasses are discussed. Detailed synthesis, processing, and applications are highlighted, with special emphasis on the advantages of the composite over individual components. In general, pliability, durability, and recyclability of the MOFs are enhanced, while polymers are endowed with porosity, diffusivity, and activity in the composite. Further advancement can be anticipated in terms of selectivity and loading amount of the MOFs, as well as the polymer-MOF compatibility.
10.2
Processing Methodologies
10.2.1 Electrospinning Electrospinning is a widely used method for fabrication of nanofibers from solution, where electric potential is used to draw a very thin (usually in the range of few hundreds of nm) fiber from the solution or molten state of the fiber-forming material. The device has two major parts: a reservoir for the liquid substance and a collector. The liquid is passed through a hollow needle, and a high electric potential is applied to the needle. Under the high electric potential, the liquid gets charged, and the electrostatic repulsion overrules the capillary force to form a typical Taylor cone [7]. This stream of the liquid is then finally deposited on the electrically grounded collector plate (Fig. 10.1). By controlling the flow rate of the liquid, electric
Fig. 10.1 Schematic representation and working principle for electrospinning method for nanofiber fabrication. (Adapted with permission from Ref. [7])
10
Natural Polymer-Based MOF Composites
323
potential, and distance to the detector, it is possible to finely tune the diameter and deposition pattern of the forming fiber. While this technology is frequently applied in academia, only few industrial companies use it. The productivity is low, and a stable support is required. However, for fine-particle filtration, such coatings are produced, for example, by Neenah Gessner, for niche applications. A general challenge is needle clogging by particles. Elmarco has developed an alternative technology using continuous feeding modes based on rotating cylinders decorated with metallic spikes. The first paper on electrospinning of MOFs was published by Rose et al. [8]. Soon, researchers were able to cast electrospun MOF membranes for interesting applications such as explosive detection and oxygen reduction reaction [9, 10, 11]. For the natural fiber-based MOF composites, silk fibroin is typically processed via electrospinning to obtain a fabric of uniform fiber diameter and coating thickness. For this purpose, purified silk is first dissolved in suitable ionic liquid, and then the solution is subjected to electrospinning [12].
10.2.2 Hot-Pressing Method Considering the large-scale industrial application of a material, it is utmost important to have an easy and reproducible bulk-scale production method. Hot-pressing method represents a unique way for the formation of MOF-based composites on surfaces of natural-based polymers. Conventional syntheses of MOFs are often restricted to smaller-scale synthesis and require large amount of solvent, adding to the environmental and economic impacts. Also, as mentioned previously, pure MOF crystals are rarely processable and so can’t be used directly for industrial application purposes. Hot-pressing method is a well-known industrial process for making various products under the simultaneous influence of heat and pressure. For chemical productions, this process is a continuous scale synthesis and also eliminates the requirement for large quantity of solvent. Furthermore, by introducing suitable substrate in the synthesis setup, composite of the target material can be achieved. And since it involves a direct synthesis of composite materials, it does not require any additional binder; this gives a higher content of the active material and better accessibility. Hot-pressing methods for natural polymer-based MOF composite formulation have been achieved by taking the components of the respective MOF on the desired support surface with or without solvent. PEG has been tested as a solvent of choice for better dispersion of the components which were then covered with Al foil and pressed at higher temperature (Fig. 10.2), achieved through stamping device or electric iron [13].
324
T. Kundu et al.
Fig. 10.2 Schematics for a hot-pressing method to produce MOF-based composites in a continuous scale method. (Adapted with permission from Ref. [13]) Fig. 10.3 Schematic representation for biomimetic biomineralization approach towards MOF composite formation with silk nanofibers. (Adapted with permission from Ref. [12])
10.2.3 Biomimetic Biomineralization Biomineralization is a biological process where living organisms produce a hard shell of minerals around a soft tissue to protect it from harsh environment. This concept of bio-inorganic composite formation has been translated into synthetic chemistry to produce strong and efficient hybrid composites and generally termed as “biomimetic biomineralization.” The presence of biological entities benefits the process in two methods: (i) by triggering and controlling the growth of composite on the inner support core and (ii) by enhancing the interaction between the support and coating to make a stronger composite (Fig. 10.3). From the class of natural-based polymers, silk has remained a choice of material because of its processability into nanofibers and chemical composition. Silk is constructed from polymeric chains of amino acids, which are connected through amide chains. These amide bonds can be easily hydrolyzed to create open carboxylic acid groups on the surface of the fibers. Such carboxylic acid groups are available to coordinate with metal ions as anchoring points to create a strong interaction between the fiber and the coating. Another example is chitin backbone, which provides hierarchical pores and support for MOF growth (e.g., HKUST-1) to enable effective air filtration properties [14].
10
Natural Polymer-Based MOF Composites
325
10.2.4 Layer-by-Layer Deposition Layer-by-layer deposition is another interesting way to grow MOF crystals on supports such as natural polymeric fibers. As the name suggested, MOF coating in this process is developed on the suitable surface through sequential growing. For the deposition of the initial layer, suitable nucleation sites are required which help the growth of MOF nanocrystals on the surface. Such nucleation sites are typically obtained by chemical functionalization of the nanofibers, producing free hydroxyl or carboxyl groups. These fiber textiles are then first dipped into metal salt solution to anchor the metal ions on the decorated sites of the surface. In the next step, after removal of excess metal ions through washing, the textile is subjected to treatment with linker solution where the MOF growth started occurring. After the growth of this layer, the newly formed surface possesses more anchoring sites because of incomplete coordination of linkers. These sites help in growing the second layer through successive treatment of metal ion and linker. By repeating this process, a uniform coating of MOF on the surface is obtained (Fig. 10.4). Because of the presence of definite anchoring sites in each layer, they are held very tightly to make a strong composite, which is again attached coordinatively to the base surface.
10.3
Natural Polymer-MOF Composites
10.3.1 Cellulose-Based Composites Cellulose is a polysaccharide made of glucose units condense through β-glycosidic bonds (Fig. 10.5). It is the most abundant naturally occurring organic compound and can be processed to produce papers, fibers, and aerogels or chemically modified to make plastics, celluloids, adhesives, etc. Carboxymethylation and esterification are the most commonly used reactions as chemical modifications of cellulose, which
Fig. 10.4 Schematics for the layer-by-layer growth of MOFs on silk fibers. (Adapted with permission from Ref. [15])
326
T. Kundu et al.
Fig. 10.5 Structural units for cellulose
also help in composite stabilization [16]. Cellulose-based MOF materials are mainly classified in two important classes, (1) cellulose nanofiber and (2) cellulose aerogel.
10.3.1.1
Cellulose Nanofiber-Based Composites
Cellulose nanofibers consist of straight-chain polymers thanks to extensive interchain hydrogen bonding via hydroxyl groups to form microfibrils with high tensile strength, depending on the degree of polymerization [17]. Rodriguez et al. first introduced cellulosic nanofiber-MOF composite in 2014, mainly studying antibacterial properties. Cu-BTC (aka MOF-199 or HKUST-1) was synthesized in situ on carboxymethylated cellulosic substrates using Cu(OAc)2, 1,3,5-benzenetricarboxylic acid and trimethylamine solutions [18]. A strong antimicrobial activity against Escherichia coli on agar plates and liquid cultures (Fig. 10.6). The anionic cellulosic fibers allowed strong attachment with the MOF particles, allowing the modified textile to be washed and reused. A year later, Laurila et al. optimized in situ growth of HKUST-1 on electrospun cellulose nanofibers, dependent on the type of anionic pretreatment using layer-bylayer approach (Fig. 10.7a–c) [19]. Two different methods for introducing carboxyl groups on the nanofiber surface were used, viz., carboxymethylation using sodium chloroacetate and adsorption of carboxymethyl cellulose (CMC). The electrospun nanofibers of unmodified cellulose and CMC-adsorbed cellulose displayed inhomogeneous loading of HKUST-1, while the carboxymethylated cellulose nanofibers were homogeneously coated. With the increase in HKUST-1 content (Fig. 10.7d), the BET surface area could be increased from 44 to 440 m2g1. In 2018, Su et al. prepared cellulose-based air filter (CFs-ZIF-8 filter) by growing ZIF-8 nanocrystals in situ on the surface of cellulose fibers [20]. The ZIF-8 nanocrystals increased the specific surface area of composite filter, strengthened the interactions between filter and the analytes, and provided pores to trap pollutants (Fig. 10.8a–b). The filtration efficiency of CFs@ZIF-8 filter for PM0.3 could reach up to quantitative level (Fig. 10.8c). Later, another cellulose-based air filter was fabricated by in situ generation of double-component MOFs and reinforcement of cellulose nanofiber, viz., Ag-MOFs@CNF@ZIF-8. The filtration efficiency of the composite filter could reach 94.3% for PM2.5. The composite exhibited excellent antibacterial activity against Escherichia coli, and the compressive strength of the
10
Natural Polymer-Based MOF Composites
327
Fig. 10.6 Antibacterial activity results according to ASTM E2149-13a. Bottom: The celluloseMOF fabric (anionic cotton1MOF-199) shows complete inhibition of bacterial growth on the contact area. The absence of halo indicates lack of diffusion of the antimicrobial compound around the contact area. In sharp contrast, no inhibition zone was formed when only anionic cellulose (anionic cotton) was placed on the LB plates seeded with confluent E. coli. Top: No inhibition halos were observed when cells were grown in contact with solutions of copper acetate, 1,3,5benzenetricarboxylic, and DMF-ethanol-water, indicating that substances involved in the MOF-199 synthesis are not individually able to kill E. coli at the reported experimental conditions. (Adapted with permission from Ref. [18])
328
T. Kundu et al.
Fig. 10.7 SEM images of (a) cellulose nanofibers after 32 synthesis cycles, (b) CMC-adsorbed nanofibers after 32 cycles, (c) carboxymethylated nanofibers after 32 cycles. (Adapted with permission from Ref. [19])
composite filter reaches up to 501 kPa, which is 3.8 times higher than pure cellulose filter. Another approach of incorporating MOFs into cellulose matrix is by acid-base interaction. To demonstrate this, a Zr-based MOF, viz., UiO-66-NH2, with free basic amino groups was wrapped by the densely packed cellulose nanofibrils (CNF-COOH) via vacuum filtration method (Fig. 10.9a) [21]. In the composite membrane, the UiO-66-NH2 was well-dispersed in the CNF matrix (Fig. 10.9b) to strengthen the CO2 diffusion process, increasing both the permeation flux and separation factor. The optimum separation performance is observed for CM-1 membrane with a CO2 permeability of 139 Barrer and a CO2/N2 selectivity ratio of 46 (Fig. 10.9c).
10.3.1.2
Cellulose Aerogel-Based Composites
Cellulose aerogels are attractive due to their renewable nature, biocompatibility, and biodegradability along with low density, high porosity, and large specific surface area. MOFs can be added on these properties, making it a promising candidate for adsorption and separation, insulation, and biomedical applications [22].
10
Natural Polymer-Based MOF Composites
329
Fig. 10.8 Schematic diagram for the preparation of CFs@ZIF-8 and CFs@ZIF-8 filter, (b) the SEM image of CFs@ZIF-8, (c) the air filtration performance of CFs@ZIF-8 filter (blank represents filtration efficiency; black represents pressure drop). (Adapted with permission from Ref. [20])
In 2016, Zhu et al. first designed cellulosic aerogel-based MOF composite for separation application [23]. The flexible and porous aerogel with hierarchical structure could hold large loading of MOF particles (up to 50 wt%) without chemical modifications (Fig. 10.10a–b). Moreover, the celluloses individually form colloidally stable suspensions but could entrap MOFs by covalent crosslinking (Fig. 10.10c). The freeze-dried composite gave hybrid aerogels with hierarchical pores that remain intact even after compression (Fig. 10.10d). The MOFs retain their crystallinity and pore accessibility in the aerogel backbone. In 2018, Ren et al. introduced MOFs inside cellulose aerogel to circumvent the difficulty of separating powdery MOF particles after catalytic reactions [24]. As a proof of concept, they studied the catalytic activation of peroxymonosulfate (PMS) by cellulosic aerogels of ZIF-9 and ZIF-12 (Fig. 10.11a–c) to remove recalcitrant organic contaminants like Rhodamine-B (RB), tetracycline hydrochloride (TC), and p-nitrophenol (PNP) via oxidative degradation (Fig. 10.11d). The hybrid aerogels/
330
T. Kundu et al.
Fig. 10.9 (a) Optical view of CM-1 membrane and (b) its bendable behavior. (c) Table describing the gas permeation test parameters of CM-1 membrane. (Adapted with permission from Ref. [21])
Fig. 10.10 Photographs of (a) CNC–CMC-based hybrid aerogels (CNC:CMC:MOF ¼ 1:1:1 by weight) and (b) all-CNC-based hybrid aerogels (CNC:CNC:MOF ¼ 1:1:1 by weight); aerogels are about 7 mm in diameter and 5 mm in height. Photographs showing that a wet hybrid aerogel (50 wt % UiO-66) can be incorporated into a syringe and compressed fully by the piston (top left), also shown from the bottom view of the syringe (top middle). When removed from the syringe, the compressed aerogel in air maintains the shape of the container it was compressed in (top right) but recovers its original shape completely when placed in solution again (bottom). (Adapted with permission from Ref. [23])
PMS system exhibit good pH tolerance and could remove PNP up to 90% in 1 h from water. The hybrid can be recycled at least for three times without sign of degradation. Later, two MOF-based cellulosic aerogels were prepared with two different MOFs, viz., ZIF-8 and UiO-66 [25]. The MOFs showed varied loading and were successfully applied to adsorptive removal of Cr (IV) and Pb(II)/ Cu(II) from water.
10
Natural Polymer-Based MOF Composites
331
Fig. 10.11 Photos of (a) pure aerogels and (b), (c) hybrid aerogels. (d) Degradation performance of RB and TC using ZIF-9@GEL and ZIF-12@GEL. (Adapted with permission from Ref. [24])
Surprisingly, the aerogel performance superseded that of the individual MOFs, highlighting the advantage of such systems.
10.3.2 Cotton-Based Composites Cotton is a naturally abundant, inexpensive cellulose, consisting of 1,4-dglucopyranose units linked together to form a linear polymeric chain. The high natural abundance makes it an important member of the naturally occurring polymer, and the scope for chemical modification has made it interesting towards biocompatible and flexible composites of materials such as MOFs. The first cotton composite of MOF was reported to be prepared by Silva Pinto et al. in 2012 [26]. The neutral structure of cotton was not suitable for effective interaction with the compositeforming MOF, and the interaction was enhanced creating anionic sites. The anionic form was prepared through carboxymethylation of the –OH groups on the surface, and these carboxylate groups acted as anchoring sites for the metal ions. MOF-199 (also popularly known as HKUST-1 and Cu-BTC) was deposited on the anionic cotton fibers to form the final composite. They attempted to prepare the composite by direct reaction of the metal precursor, organic linker, and anionic cotton in suitable solvent. However, the order in which the reactants are added to the reaction system played a key role in forming the final composite. When the linker was added to the
332
T. Kundu et al.
Fig. 10.12 Optical images for the MOF-199-cotton composite fabricated from (a) one-pot addition and (b) stepwise addition synthetic approaches. (Adapted with permission from Ref. [26])
reaction mixture at the very beginning (one-pot addition approach), MOF crystals were formed as separate entities, detached from the cotton fabric (Fig. 10.12a). However, given sufficient time for the metal salt to interact with the carboxylic functionalities on the cotton fiber surface (stepwise addition approach), a strong and homogeneous composite of MOF on the cotton was obtained (Fig. 10.12b). The role of nucleation sites and kinetics on the cotton surface thus played a major impact on the formation of composite, which has been utilized in later reports for development of interesting composites [27, 28]. Reynold’s group has shown some interesting application of using cotton as naturally occurring polymer for forming composite with MOFs. In one study, Cu-based MOF (HKUST-1) has been utilized as a stimulator for generating therapeutic bio-agent (NO) from RSNOs. Transformation of bulk MOF powder into composite showed a better interaction with RSNOs, making the NO generation more efficient. This concept was later transferred for chitosan-based composite, as discussed in Sect. 10.3.5. In this case the composite was formed using carboxymethylated cotton fibers by LBL deposition method (Fig. 10.13a–e), which overcame the synthetic difficulty observed in the previous example [29]. The LBL deposition method deposited the MOF crystals uniformly throughout the surface, which created a platform for efficient interaction with RSNOs. CysamNO was chosen as the source molecule for this study that undergoes decomposition to produce NO (Fig. 10.13f). The NO release study was monitored using chemoluminescence property, which showed the highest amount for total NO generation from CuBTC-cotton composite, surpassing that from free Cu acetate and cotton-grafted Cu-acetate (Fig. 10.13g). Another Cu-based MOF (Cu-BTC-NH2) was grown on cotton surface using similar LBL deposition technique for application as antibacterial agent [30]. NH2 functionality on the BTC linker has a great potential for functionalization through post-synthetic treatment, which can alter both physical and chemical properties of
10
Natural Polymer-Based MOF Composites
333
Fig. 10.13 (a) Scheme for introducing carboxyl functionalities on cotton fibers through carboxymethylation of glucopyranose moieties, (b) scheme showing carboxylic groups as anchoring points in the LBL deposition method for composite formation. Optical images of (c) cotton fabric, (d) as-made and (e) dried HKUST-1-cotton composite. (f) Chemical reaction and (g) reaction kinetics of the catalytic reaction. (Adapted with permission from Ref. [29])
Fig. 10.14 (a) Post-synthetic modification of MOF-cotton composite to obtain AM-5-modified MOF cotton. Optical images for (b) cotton cloth and (c) MOF-cotton composite. (d) Cu-generation and (e) antibacterial properties of the composites. (Adapted with permission from Ref. [30])
the MOF. By LBL deposition on a carboxymethylated cotton fiber, they have made a uniform composite of MOF microcrystals on the cotton surface. The composite was characterized by the typical green color from the MOF (Fig.10.14b–c) and crystalline peak. Like the previous cases, SEM images showed a complete coverage of the surface with grown MOF crystals. However, it is noteworthy that Cu-BTC-NH2 MOF is not water stable and undergoes slow decomposition in aqueous medium, releasing Cu(II) to the medium. The generated Cu(II) ions have been utilized as an
334
T. Kundu et al.
Fig. 10.15 Schematics for synthesis of MOF-cotton composite for air purification application through PM removal. (Adapted with permission from Ref. [31])
antibacterial agent towards Escherichia coli. Post-synthetic modification of the composite (MOF-cotton) was successful in converting some of the amine groups of the framework into amide groups (AM-5-modified MOF cotton), through reaction with valeric anhydride (Fig. 10.14a). Post-synthetic modification showed a change in the framework property, and overall rate for hydrolysis is changed. This lowering in hydrolysis rate effectively controlled the Cu ion concentration in the solution. MOF-cotton composite swatches showed increased death rate for E. coli bacteria, over that of natural cotton or pure linker. But the Cu ion generated from the MOF decomposition triggered the antibacterial effect. Since the bulk crystals of the MOFs have different decomposition rate, swatches from MOF-cotton and AM-5 modified MOF-cotton also differed in producing Cu flux to the solution (Fig. 10.14d). This resulted in a slow killing of E. coli bacteria by the Cu ions, where the activity of AM-5-modified MOF-cotton is more controlled because of the slower hydrolysis kinetics (Fig. 10.14e). Cotton-based MOF-based composites have also been prepared efficiently through hot-pressing method. Using cotton as the textile backbone (Fig. 10.15), four classical MOFs (viz., MIL-53(Al) (Fig. 10.15 row a1), ZIF-8 (Fig. 10.15 row a2), UiO-66 (Fig. 10.15 row a3), and MFM-300(In) (Fig. 10.15 row a4)) were synthesized on its fibers through hot-pressing method [31]. For the preparation of the composite, a commercial hot-press machine was used to form the MOF crystals from its constituents, without use of any solvent. A uniform nature of the nanocrystals coating was observed from SEM imaging of the composites. The MOF nanocrystals were supposedly attached to the fabric through formation of coordination interaction with the hydroxyl groups of the cotton, as suggested from IR and Raman spectra. However, these composites suffered from low porosity compared to respective bulk MOF, which is a potential drawback from the hot-pressing method. However, the nanoparticle-sized MOFs were capable of polarizing surface of polluting particulate matters (PM) to enhance the electrostatic interaction. This property made the composites suitable for use in air purification application.
10
Natural Polymer-Based MOF Composites
335
The performance of the MOF fabrics for air purification through PM removal was tested by measuring the PM level before and after passing the polluted air through the composite membrane. The polluted air was generated by burning incense smoke and contained small-sized solid particles with an average PM concentration of PM2.5 > 280 μg m3 and PM10 > 360 μg m3. Presence of MOF nanoparticles on the cotton surface showed a significant enhancement of PM removal efficiency, with an increase to >90% for all the MOF-loaded cotton composites from 100 nm) were not detected in the brain, whereas the smaller MIL-88B_4CH3 (around 40 nm) seemed to bypass the blood-brain barrier (BBB) and were detected in brain tissue, although no toxicity was noted. Finally, and interestingly, the fumarate linker of the MOF was only detected in trace form in the urine of the animals, likely due to it being incorporated into the Krebs cycle. However, further research into how this might impact cellular function and proliferation would be important to determine the downstream effect on cells. Another early study from Ruyra et al., characterized the toxicity of 16 MOFs in both the liver (HepG2 cells) and breast (MCF-7 cells) before transitioning to zebrafish embryo models [60]. Table 11.2 shows the comparison of the different MOF toxicity in the different systems from this study. The panel of MOFs was chosen with different metals, including Zr, Co, Zn, Cu, and Fe. While the linkers tested showed good compatibility, interestingly, Zn, Zr, and Mg metals gave cell viabilities greater than 75%. However, Fe, Cu, and Mn showed moderate to high cytotoxicity, respectively. Interestingly, the Fe-based MIL-101 showed toxicity in the zebrafish model but had good biocompatibility in previous in vitro studies [57], most likely due to differences in the assays or degradation profiles. This highlights the importance of fully characterizing nanoMOF systems in different models in order to understand how best to select the optimal candidates to take forward.
11
Metal-Organic Frameworks as Delivery Systems of Small Drugs and Biological Gases
359
In order to translate MOFs to the real application in drug delivery, and despite some early interesting studies, further work is required to characterize fully the MOF cytotoxicity profiles. Although some work has been done in the Fe-based MOFs from the MIL series, this is especially relevant for other commonly used nanoMOFs. As such, it is essential to develop new elegant study designs, evaluating different particle concentrations with exposure over longer periods of time, using more biologically relevant MOF concentrations (i.e., in the μg scale, not mg) and in relevant models such as immunogenicity assays. When analyzing the performance of MOFs in drug delivery, the composition and size of the nanoparticles are as important as the aggregation. Aggregation has a significant impact on cell membrane interactions, particle internalization, and behavior. MOFs, especially those which are not functionalized, are very likely to form aggregates in solution, and therefore care must be taken around compatibility conclusions drawn on non-functionalized materials. In this sense, a detailed characterization of external-surface modified materials is required to establish the biocompatibility of these systems.
11.4
The Control of Small Molecule Drug Release
When considering MOFs for biomedical applications, the rate of cargo release, both in vitro and in vivo, is important to determine. One of the advantages of using DDSs is to improve the pharmacokinetic and pharmacodynamic (PKPD) properties of encapsulated cargo. However, in many formulations, a phenomenon termed “the burst release effect” occurs, whereby the cargo dissociates rapidly from its carrier immediately upon treatment, thus impacting downstream efficacy and potentially causing off-target toxicity [61]. Indeed, the “burst release effect” is also one of the main limitations when considering MOFs as DDSs, with some reporting the rapid desorption of cargo occurring in less than 48 h [62]. Typical MOFs show drug release within the first 48 h, and a big question is if the drugs are loaded in the porosity or the external surface – something that is particularly difficult to assess in a MOF. Now, given the versatility of MOF chemistry, it is possible, in principle, for the rational synthesis and selection of nanoMOFs which will circumvent this effect [20, 40, 43, 63–65]. Orellana-Tavra et al. studied the Bi-MOF CAU-7 [66] to study the release of two drugs, DCA and α-CHC, demonstrating a progressive drug release profile of 17 days and 31 days, respectively [20]. There are different approaches in the literature describing the extension of the kinetics of cargo release from MOFs after adsorption of small drugs in the porosity [20, 40, 64, 65, 67]. For example, Orellana-Tavra et al. reported the mechanical amorphization of the Zr-MOF UiO-66 through a ball-milling process, which extended the release of calcein – a hydrophilic fluorescent molecule with a similar structure to the cancer drug doxorubicin – from 2 days in the crystalline UiO-66 to more than 30 days [40]. Similarly, the temperature treatment of the mesoporous Zr-MOFs NU-1000 and NU-901 partially collapsed their porosities during water removal due to the high surface tension of the solvent, slowing down the initial
360
E. Linnane and D. Fairen-Jimenez
Fig. 11.4 SEM images of NU-1000 (a) prior to calcein loading, (b) post-loading, (c) post-loading, and temperature treatment. (b) Calcein release from crystalline NU-1000 and t.t.NU-1000. (Adapted from Templensky et al. [65])
release rate of calcein from the MOF structures (Fig. 11.4) [65]; the thermal amorphization of the Bi-MOF CAU-7 caused, again, a 32% slower release rate of the encapsulated cancer drug α-CHC [20]. The concept of developing multifunctional DDSs is garnering interest in the field of nanomedicines, especially with the rise of personalized medicine. Combining therapeutic, diagnostic, and imaging functions (“theranostics”) is an important development for nanomedicines as a platform. MOF composites are an emerging class of DDS, with capabilities such as imaging and stimulus-responsive release, coupled with controlled cargo delivery. For example, the use of organic silica to create MOF hybrid materials has been reported [14, 68, 69], where incorporation of a silica shell around a nanoMOF core was shown to increase particle stability, while also helping to control the guest-molecule release [14]. Nanoparticle-MOF composites are also showing promise in this area; for example, iron oxide-MOF (Fe3O4@UiO-66) core-shell composites, synthesized with doxorubicin, showed a sustained drug release and exhibit long-lasting and efficient anticancer therapeutic efficacy [70]. Another way to circumvent issues with the premature release of cargo is through the synthesis of stimuli-responsive materials, sometimes known as “smart MOF carriers” [71, 72]. Stimulus-induced release of cargo can be achieved through
11
Metal-Organic Frameworks as Delivery Systems of Small Drugs and Biological Gases
361
various methods, including response to light, pH, enzymes, adenosine triphosphate (ATP), and heat. For example, Chen et al. synthesized zirconium nanoMOFs loaded with doxorubicin and then modified them with nucleic acid binding strands including the ATP-aptamer. Doxorubicin was unlocked from the MOF complex in the presence of ATP via the formation of ATP-aptamer complexes. These particles were shown to have an ATP-responsive drug release in MDA-MB-231 breast cancer cells [73]. Liu et al. synthesized core-shell ZIF-8@MnO2 hybrid particles to reduce intracellular glutathione (GSH) and trigger an enhanced photodynamic cancer therapy [74]. Zn-based MOF ZIF-8 is pH sensitive, reported to be degraded at pH < 7, and it is, therefore, often selected as an ideal candidate for controlled cargo release, since it degrades in the acidic lysosome environment, releasing its contents into the cell. Yang et al. synthesized nanocomposite ZIF-8 material with encapsulated Pd@Au nanoparticles and doxorubicin (DOX) (DOX/Pd@Au@ZIF-8) and showed pH-responsive controlled release behavior using ultrapure water under different pH values (7.4, 6.0, 5.0, and 4.0) [75]. Further investigation into harnessing natural cellular processes for controlled drug release using the relevant cellular systems is needed to fully elucidate the effect on drug PK profiles. Although this is an emerging area, the remarkable tunability of MOF chemistry with the option of creating multifunctional hybrid materials has the potential to impact greatly the PKPD properties of guest molecules. To date, there has been significant progress in synthesis methods to control the release of drugs from MOF structures. However, finding the optimal rate of cargo release and of MOF degradation is still both model- and context-dependent. Indeed, prudent selection of relevant disease and cell models is crucial, and investigation of drug release using both ex vivo and, eventually, in vivo systems should also be probed. The continuing evolution of these materials will open up more biological questions regarding the application and versatility of MOFs as controlled drug delivery systems.
11.5
Metal-Organic Frameworks as Delivery Systems for Biological Gases
In addition to small molecule delivery systems, MOFs have shown promise in the transport of gasotransmitters – gaseous signaling molecules – that are used to transmit chemical signals that induce biochemical changes in the organism, tissue, or cell. Examples of such gases are nitric oxide (NO), carbon monoxide (CO), and hydrogen sulfide (H2S) [50] although, to date, there is a limited number of studies on the latter two gases. Hydrogen sulfide, while being a very toxic colorless gas, is also reported to be important for signaling in the cardiovascular and nervous systems of mammals [76, 77]. Morris and co-workers demonstrated the binding and storage of hydrogen sulfide in MOFs with the CPO-27/MOF-74 structure (2,5-dihydroxy-1,4benzene dicarboxylic acid) with various metals including Ni and Zn. Their preliminary work showed delivery of biologically active gas in vasodilation experiments
362
E. Linnane and D. Fairen-Jimenez
using porcine arterial tissue [78]. Carbon monoxide, also well known for its toxic properties, has also been reported to have antiproliferative and proapoptotic effects on cancer models [79]. However, controlled and trackable gas release is crucial; therefore establishing ways for effective delivery is a subject of much interest. Jin et al. established a carbon monoxide-releasing nanomaterial comprising of manganese carbonyl [MnBr(CO)5] and a Ti-based MOF for an intratumoral hydrogen peroxide-triggered gas release and real-time carbon monoxide monitoring by fluorescence imaging [80]. However, limited biological work in this area means that its translation is still a challenge. NO is a biological messenger molecule produced by different cell types, with functions including muscle relaxant [81] and mediating wound healing, vasodilation, platelet adhesion, and aggregation. With an intravascular half-life of 2 ms and an extravascular half-life of 0.09–2 s, [82] NO has also an anti-inflammatory role and is reported to be involved in a variety of diseases, including septic shock and asthma [83]. The biological effects of NO are contextdependent, and the concentration of NO is crucial for its function as any imbalance can lead to undesirable effects. For example, the vasodilator effects of NO are seen at very low concentrations, whereas inflammatory cells generate high local concentrations of NO. As such, delivery and release of NO need to be precise and well controlled, meaning that applications for NO-related delivery systems are attracting much interest. One of the most reported functions of NO is in wound healing [84], with numerous studies showing the role of NO in decreasing the time it takes for a wound to heal in diabetic [85, 86] and corneal [87, 88] injury sites. A number of novel NO delivery systems are being developed for biomedical applications. Given the high porosity and tunable properties of MOFs, these have the potential to be ideal candidates for storing, delivering, and releasing NO. Pioneering work in the field by Morris and co-workers investigated the use of zeolites for storing and releasing NO, demonstrating anti-thrombosis activity on platelet-rich plasma [50]. Following on from this work, they turned their attention to MOFs, which have shown great promise for the storage and release of NO due to their versatile properties. The open metal sites found in some MOFs such as HKUST-1 and CPO-27/MOF-74 can provide active sites for coordination bonds to form, and the ligand moieties contain amine functionality that can form NONOate or nitroso groups [11]. As such, in order to exploit this route, the nature of the metals is critical. Morris and co-workers characterized first the porous copper MOF, HKUST-1, reporting a loading on 3 mmol of NO per gram [12]. Platelet aggregation studies showed that NO-loaded HKUST-1 sample completely inhibits aggregation compared to the control (Fig. 11.5). Following this, Cattaneo et al. used Mg-, Zn-, and Ni-based CPO-27 MOFs to study NO storage and release in the CPO-27/MOF-74 series [11]. These MOFs were prepared through the coordination of 2,5-dihydroxyterephthalic acid with different metals and with pore sizes of 11–12 Å. The NO release was profiled using porcine coronary artery relaxation tests [11]. However, the selection of Cu and Ni metals from these studies presents questions over toxicity and biocompatibility. For example, after a short exposure of HKUST-1 to platelet-rich plasma, they observed dissolution of copper from the MOF [12].
11
Metal-Organic Frameworks as Delivery Systems of Small Drugs and Biological Gases
363
Fig. 11.5 Platelet aggregation experiments show NO-HKUST-1 (pink) completely inhibits platelet aggregation compared to the control (blue). Aggregation was initiated by the addition of collagen marked by the arrow. (Adapted from Xiao et al. [12])
Clearly, for the rational selection of MOFs and linkers for NO delivery, biocompatibility is an important consideration, and, therefore, the building blocks used in the MOF construction need to be carefully considered. Changing the linker has also impact in the toxicological profile of the MOFs, whereas the possibility of getting NO stable compounds in the presence of, e.g., water is challenging. Furukawa and co-workers proposed a new approach for developing PCPs capable of releasing NO only on light irradiation. In particular, they used two zeolitic imidazolate frameworks (ZIF) based on Zn and 2-nitroimidazole and 5-methyl-4-nitroimidazole. By avoiding the use of toxic metals, they demonstrated the biological release of NO in HEK293TRPC5 cells using two-photon near-infrared laser irradiation [89]. Kim et al. also demonstrated the controlled release of NO from their synthesized PCPs by light irradiation through N-nitrosamine functional groups which acted as photoactive NO donors [90]. However, despite this progress, further work to optimize methods of synthesis for biologically compatible models is needed for the translation of these ideas to biological systems.
11.6
The Intracellular Fate of MOFs
While there have been numerous studies of MOF behavior in vitro, there is limited information regarding intracellular uptake. Indeed, for the application of MOFs as DDSs, it is important to understand and characterize intracellular mechanisms. Firstly, in order to comprehend how the MOFs behave and, secondly, in the context of model selection and productive delivery applications [91]. Endocytosis is an energy-dependent process of active transport in which molecules outside the cell are engulfed and internalized into the cell. Endocytosis occurs at the cell membrane surface and can happen through a number of different mechanisms, depending on the cargo being transported. Figure 11.6 shows the main endocytosis routes. The most well-established mechanisms are clathrin-mediated uptake, caveolin-mediated uptake, macropinocytosis, and phagocytosis, with the clathrin- and caveolinindependent pathways less characterized. Endocytosis is characterized by transport
364
E. Linnane and D. Fairen-Jimenez
Fig. 11.6 Endocytic pathways of mammal cells. Clathrin-dependent, caveolin-dependent, and clathrin- and caveolin-independent, macropinocytosis, and phagocytosis. (Adapted from De Souza at al. [94])
of vesicular structures along these different pathways and is coordinated by Rab proteins from the Rab GTPase family [92]. In their active form, Rab GTPases recruit effector proteins onto membranes and act to regulate vesicle formation, actin- and tubulin-dependent vesicle movement, and membrane fusion events needed for membrane trafficking [93]. Clathrin-mediated uptake is a type of receptor-mediated endocytosis, whereby materials are internalized into clathrin-coated pits that bud off to form vesicles. These vesicles fuse with early endosomes and are then transported via the late endosome-lysosome pathway, for degradation, or recycled back to the cell membrane via the recycling endosome [95]. Caveolae-mediated endocytosis is the internalization via the formation of lipid-raft-enriched membrane invaginations. Particles that internalize via this pathway can also end up transported to the lysosomes but can be delivered to the caveosome as well, therefore avoiding degradation by the acidic environment and the presence of enzymes in the lysosomal compartment. Macropinocytosis, known as “cell drinking,” is also an energy-dependent process, whereby larger particles are engulfed by the actin ruffles on the cell membrane, forming macropinosomes, which are heterogeneous in size, ranging from 0.2 to 5 μm [96]. These different membrane trafficking pathways are cell-type dependent and, in
11
Metal-Organic Frameworks as Delivery Systems of Small Drugs and Biological Gases
365
the complex tumor microenvironment, are often driven by dysregulation of cellular metabolic processes. Why is it then so important to understand the route of nanomaterial entry? It is critical because the route of entry for a MOF confers its activity and downstream efficacy. On top of that, the importance of a productive route (i.e., one that avoids or escapes the acidic lysosomes) is also often cargo-dependent. For example, a MOF transporting antisense oligonucleotides would require transport to the nucleus for a productive knockdown of the target. Indeed, even for small molecules, transport to the lysosome can mean a reduction in efficacy or MOF being recycled out the cell before it can effectively release its cargo. To develop optimal and efficient nanoMOF DDS, therefore, it is essential to understand the biology surrounding their intracellular trafficking. Only a few studies have assessed the intracellular distribution and pathways of MOF uptake in vitro, where the particle size and shape, composition, surface chemistry, and zeta potential play an important role in the transport across the membrane and delivery efficiency. At the same time, the same MOF system might show different intracellular trafficking mechanisms in different cell lines. The first study on MOF trafficking was from Orellana-Tavra, Mercado, and Fairen-Jimenez, who used different pharmacological inhibitors (sucrose, inhibitor of clathrinmediated endocytosis and other endocytic routes; chlorpromazine, inhibitor of clathrin-mediated endocytosis; nystatin, inhibitor of caveolae-/lipid-mediated endocytosis; and rottlerin, inhibitor of micropinocytosis) and confocal co-localization studies to characterize the endocytosis mechanisms of Zr-MOF UiO-66 of different sizes [64]. They found that the active trafficking of 150 nm UiO-66 nanoparticles was – almost exclusively – through the clathrin-mediated pathway, whereas the larger 260 nm UiO-66 nanoparticles used both the clathrin and caveolin pathways. As it is hypothesized that the caveolin pathway can often avoid the lysosomes and therefore bypass its degradation effects on the cargo [97], this is often said to be the preferred endocytic route for drug delivery. In this case, it seemed that particle size was an important consideration for productive cargo delivery. Following this, Orellana-Tavra et al. continued reporting on the effect of linker functionalization in UiO-66 on MOF internalization [64]. They found that longer ligands’ naphthalene-2,6-dicarboxylic and 4,40 -biphenyldicarboxylic acids promoted entry through the caveolin pathway and avoided lysosomal degradation, thus increasing productive uptake of the particles and small molecule cargo (Fig. 11.7). Other MOFs that have been analyzed in terms of trafficking include Bi-based CAU-7 and Zr-based NU-901 and NU-1000. Confocal imaging of calcein-loaded CAU-7 in HeLa cells following 24 h of treatment showed a distribution of the MOF particles in the cell cytoplasm [20]; in this case, CAU-7 was taken up through clathrin- and caveolin-mediated endocytosis, with reduction of fluorescence signal of 23% and 47% using clathrin inhibitor and 58% using the caveolin inhibitor [50]. A similar pattern was observed for MOFs NU-901 and NU-1000, whereby uptake was reduced upon inhibition of the caveolin pathway, suggesting this was the main route of cellular entry [65].
366
E. Linnane and D. Fairen-Jimenez
Fig. 11.7 (a) Organic linkers used in the series of Zr-based MOFs and (b) the effect of pharmacological endocytosis inhibitors on their uptake, measured by flow cytometry. (c) Co-localization of the Zr-based MOFs, with different linkers (green), with lysosome (red) in HeLa cells. (Adapted from Orellana-Tavra et al. [64])
When looking at the surface chemistry of DDSs, one needs to bear in mind the broad possibilities that nanoMOFs offer in terms of post-synthetic surface modifications. The grafting of different groups on the external surface can confer a specific route of uptake, especially in the context of receptor-targeting. Indeed, the addition of a polyethylene glycol (PEG) coating on UiO-66 was shown to be internalized differently than UiO-66 alone, escaping lysosomal degradation through trafficking through caveolin-mediated routes, indicating that either due to size, hydrophilicity, or charge, PEG has an effect on what pathway the MOF takes [63]. Most in vitro studies have characterized endocytic uptake using cancer models, given that this is one of the most prevalent applications for MOFs as DDS. However, Durymanov et al. used a fluorescent-labeled MIL-88B-NH2 MOF to analyze the cellular uptake and intracellular trafficking in stellate macrophages, Kupffer liver cells. They showed that phagocytosis was the major endocytic pathway involved in MIL-88B-NH2 internalization, for these cells. They also showed that MIL-88B-NH2 MOFs stayed intact for up to 15 min after internalization and that accumulated in acidic cellular compartments, degrading to ca. 10–15% over 24 h [98]. Further work and in-depth analysis of MOF uptake routes – and context – are required to aid the development of new nanoMOF systems and in order to provide a further understanding of their behavior in vitro, for a translation in vivo.
11
Metal-Organic Frameworks as Delivery Systems of Small Drugs and Biological Gases
11.7
367
External Surface Chemistry
As mentioned above, the rational design of MOFs for biomedical applications extends beyond the control of particle size, pore volume, and the selection of biocompatible metals and organic linkers. Tuning the chemistry of the external surface through the grafting of different moieties can enhance the material properties. It can also enable the control of the cargo release or can include targeting capabilities. Functionalizing the MOF external surface can occur through either postsynthetic modifications or directly during synthesis [99]. For the former, functional groups can be attached through non-covalent interactions, conjugation, or coordination to free metal sites. The latter can be achieved through the addition of modulators which act as capping agents [63, 100]. Chemical modulators can be used to control MOF particle size and to maintain colloidal stability and control dispersion [101, 102]. For example, Morris et al. demonstrated that, by increasing the concentration of acetic acid when synthesizing UiO-66, the net value of the zeta potential was increased, leading to the repulsion between neighboring particles. The resulting nanocrystals were colloidally stable in water and monodisperse, although highly defective [102]. Surface functionalization is crucial, therefore, for the successful application of nanoMOFs in small molecule drug delivery as well as for the delivery of biological gases. Firstly, they can offer a solution to overcome some challenges inherent in MOF synthesis relating to material aggregation and colloidal stability. Secondly, they can improve the in vivo nanoMOF bio-distribution and blood circulation. There is an emerging number of studies to develop nanoMOF coatings to enhance the material performance in vitro and in vivo while not hindering drug activity or release by blocking the porosity [14, 63, 103–107]. Indeed, despite a large number of studies characterizing MOF pharmacology in vitro, challenges with particle stability and aggregation in solution, due to the drying process, have posed a challenge for scaling-up synthesis for in vivo translation. Particle surface charge – linked to the z-potential – hydrophilicity and stability are key factors in maintaining an effective bio-distribution, as well as a subsequent successful delivery and uptake to the desired location [108]. Indeed, one of the main challenges in the field of drug delivery is the rapid clearance of nanoparticles from blood circulation, which in turn can impact drug availability and efficacy [108, 109]. Macrophage cells, components of the mononuclear phagocytic system (MPS), play an important role in the body’s immune defense system through phagocytosis of undesirable or unwanted components (e.g., pathogens). Macrophages identify nanoparticles as foreign bodies due to opsonization of their surface by serum proteins [110]. As depicted in Fig. 11.8, grafting of PEG chains onto the nanoparticle external surface can prevent the uptake by macrophages and therefore increase their stability and half-life [109, 111]. Doxil® was the first FDA-approved PEGylated liposomal nanoparticle, increasing the drug doxorubicin’s bioavailability, circulation, and half-life [112, 113]. Since then, PEGylation has become widely used in the field due to its advantageous properties. PEG chains are hydrophilic and
368
E. Linnane and D. Fairen-Jimenez
Fig. 11.8 Surface modifications of nanoMOFs can impact cellular uptake and clearance from circulation. (Adapted from Rattan et al. [115])
act to reduce serum protein binding through steric hindrance. They can shield the particle surface from aggregation, whereas their length can also influence the behavior and performance of nanoparticles [114]. To improve and enhance the capabilities of MOFs for drug delivery, Abanades Lazaro et al. used click modulation to modify the external surface of UiO-66 [63]. By covalently attaching PEG chains of different lengths, PEG550 and PEG2000, to the external surface of UiO-66, they showed a reduction in the formation of aggregates in aqueous solution, from over 2 μm (UiO-66 alone) to 1 μm with UiO-66-PEG550 and around 400 nm with UiO-66-PEG2000. The analysis of these particles after exposure to phosphate buffer PBS, using PXRD, indicated that the crystallinity was retained for longer in samples with PEGylation than without. This showed enhanced stability and protection from phosphates present in solution, which can attack the MOF and, eventually, destroy the structure. Using the cancer drug DCA, they also reported enhanced cytotoxicity of PEGylated UiO-66 particles (DCA@UiO-66-L1-PEG2000) compared to UiO-66 alone (DCA@UiO-66) in HeLa cells, linked to the different cellular uptake, as described in the previous section. Further work to investigate macrophage internalization of these particles would be important to determine if these modifications offer any protection from phagocytosis. Another example in the field has combined the properties of PEGylation, to reduce particle aggregation, with targeting capabilities, to allow for the selective uptake of small molecule-loaded nanoparticles into cancer cells. Shi et al. modified
11
Metal-Organic Frameworks as Delivery Systems of Small Drugs and Biological Gases
369
the external surface of ZIF-8 to study selective targeting including folic acid (FA), whose receptor is overexpressed in certain cancer cells, such as HeLa cells [34]. First, ZIF-8 was loaded with an autophagy inhibitor, chloroquine diphosphate (CQ), known to have anti-tumor and anti-metastatic activity in some cancer models [116]. Then, the external surface was functionalized through direct adsorption of PEG and FA on CQ@ZIF-8; chloroquine diphosphate drug release was shown to be pH-responsive. Dong et al. demonstrated selective targeting of cancer cells through coordinating the terminal carboxylates of the folic acid to the Zr6 clusters of the Zr-MOFs MOF-808 and UiO-66-NH2 [117]. By combining modifications such as PEG and targeting capabilities on the MOF surface, cell-specific uptake can be facilitated (Fig. 11.8). Post-synthetic modifications include hyaluronic acid (HA), a linear anionic polymer that has been extensively studied for cancer therapies due to its promising characteristics in terms of biocompatibility and immune-toxicity profile [118– 120]. HA can provide a negative surface charge and an affinity for receptors expressed on cancer cells, allowing for cell-specific internalization [121]. Shu et al. coated HA into a ZIF-8-doxorubicin complex (DOX@ZIF-HA) and reported that the material efficiently enhanced intracellular uptake, delivering more doxorubicin into PC-3 prostate cancer cells [122]. Cai et al. engineered indocyanine green (ICG) MOFs using MIL-100(Fe), conjugated to HA for imaging-guided, anticancer photothermal therapy (PTT). They demonstrated both in vitro, using MCF-7 breast cancer cells, and also in vivo, using tumor xenografts, that the MOF@HA@ICG nanoparticles exhibited greater cellular uptake and enhanced tumor accumulation compared to MOFs which were not conjugated to HA [123]. Bellido et al. modified the external surface of MIL-100(Fe) with heparin through a biocompatible one-pot method [48]. They reported a 12.5 wt.% heparin coating distributed along the external MOF surface, increasing the particle size from 141 to 173 nm (measured in water) and improved colloidal stability of the MOF through steric hindrance. In addition, the heparin coating provided stealth qualities to the MOF without disrupting the porosity, showing a reduced uptake by the macrophage cell line compared to the uncoated MOF [48]. Other work that linked hydrated dextran shells to ZIF-8 through coordination bonds was shown to improve the dispersity of the nanoMOF particles [124]. In addition to more standard functional groups, including biologically derived coatings onto the external surface of MOFs offers the advantage of stealth properties, avoiding undesirable immune toxicity issues. A study by Illes et al. showed that exosomes might be a beneficial capping system [125]. Exosomes are small membrane-bound vesicles produced and secreted by some cell types. They have multiple properties that make them an interesting proposition for coating nanoparticles; they have been proposed as a feasible option for lipid-like coating of MOFs, due to their cell-like properties and stealth capabilities. Indeed, the incubation of exosome-coated MIL-88A(Fe) in HeLa cells demonstrated successful delivery of cargo (Fig. 11.9). The incorporation of a targeting moiety onto the external surface of a nanoMOF offers the opportunity for cell-specific delivery of payloads to a specific area.
370
E. Linnane and D. Fairen-Jimenez
Fig. 11.9 (a) HeLa cells (stained in red) with exosome-coated calceinloaded particles (stained in green) after 2 days of incubation; (b) HeLa cells (red) with exosome-coated calcein-loaded MIL-88A nanoparticles (green) after 4 days of incubation. (Image from Illes et al. 2017 [125])
Cherkasov et al. conjugated the antibody trastuzumab to a composite material, Fe3O4 nanoparticles with a MIL-100(Fe) shell (Fe3O4@MIL-100), which they also coated with carboxymethyl-dextran and loaded with doxorubicin/daunorubicin (Fig. 11.10). In addition to a sustained release, the system showed selective targeting of HER2/neu-positive BT-474 breast cancer cells [25]. The preliminary in vitro results for this method of antibody grafting seem promising and open up many possibilities for cell-specific targeting payloads.
11.8
Current Challenges
The field of MOFs for use as delivery systems for small molecule and biological gases has expanded rapidly over the past decade, with significant progress in material design. In vitro studies have progressed, with the inclusion of different
11
Metal-Organic Frameworks as Delivery Systems of Small Drugs and Biological Gases
371
Fig. 11.10 Antibody targeting of MOF-NP composites. (Adapted from Cherkasov et al. [25])
models and experimental approaches to improve the understanding of nanoMOFs as DDSs. There is, however, much scope for progress in this area. Throughout many studies involving MOFs, there are multiple limitations regarding the biological assay approach and experimental design. Thoughtful selection of material concentration and dosing schedule needs to be considered when evaluating the behavior of MOFs in vitro in order to allow in vivo translation. For example, using biologically relevant concentrations of material (ng-μg instead of mg) doses will prevent misinterpretation of phenotypes associated with general off-target effects due to the high concentration of nanoparticles. While surface functionalization techniques have led to an improvement in MOF stability, uptake, and in vitro performance, appropriate controls need to be used to account for non-specific effects of modifications that might impact cell behavior in order to draw the correct biological conclusions. In particular, the design of studies to elucidate endocytic entry of nanoMOFs remains a challenge, since most endocytic inhibitors commonly used in the field are often cell- and contextdependent, produce limited sensitivity, and have off-target effects on cells [126, 127]. It is critical to understand and characterize intracellular mechanisms of nanoMOF uptake, to establish how the materials behave, and also to select the most appropriate models to analyze activity. Therefore the use of techniques such a gene knockout, pathway analysis, and use of clustered regularly interspaced short palindromic repeats (CRISPR)/CRISPR-associated system (Cas) cell lines are more robust and would provide more specific mechanistic insights into nanoMOF uptake and trafficking. The selection of downstream assays to evaluate nanoMOF efficacy and behavior is also limited, with assays such as MTT (3-(4,5-dimethylthiazol-2-Yl)-2,5diphenyltetrazolium bromide)/MTS (3-(4,5-dimethylthiazol-2-yl)-5(3-carboxymethoxyphenyl)-2-(4-sulfophenyl)-2H-tetrazolium) being commonly used, however lacking in sensitivity and mechanistic insight. The MTT assay is a colorimetric assay used to assess cell metabolic activity and can reflect the number of viable cells present. However this assay is reported to be less accurate in detecting
372
E. Linnane and D. Fairen-Jimenez
changes in cell number and is not very robust compared to other methods [128]. Additionally, the XTT (2,3-bis-(2-methoxy-4-nitro-5-sulfophenyl)-2H-tetrazolium-5-carboxanilide) viability assay, used by many to assess MOF cytotoxicity, has also been reported to have issues with robustness [129]. By including a broader selection of assays such as qPCR, western blotting, cell cycle assays, and reporterbased systems to assess MOF downstream impact, it will be possible to establish a mechanistic understanding around on-target effects beyond that of simple cell death or senescence evaluation. Despite some progress described earlier, the main barrier for MOF development in biomedical applications is, arguably, the lack of understanding of MOF toxicity. Studies to date have not assessed long-term impact of MOF exposure, with a limited investigation into DNA damage and reactive oxygen species formation within cells. Indeed, the analysis across a broader selection of models to assess immunogenicity as well as toxicity is required to build confidence in these materials as DDSs. It is important to pay attention to types of media used, serum concentration, and PKPD profiles. For example, the use of “pulse-chase” experiments rather than constant exposure could model PKPD more effectively. Finally, the progression from 2D to 3D cell cultures and organoid systems would increase confidence and provide interesting mechanistic insights into behavior and understanding of nanoMOFs before advancing into in vivo systems.
11.9
Outlook
Over the past few decades, there has been much development in different carrier systems for controlled drug release, enhancing the therapeutic benefit and reducing off-target effects. MOFs have been proposed as an exciting new delivery system, with properties such as high loading capacity and tunable pores making them attractive candidates for such an application. To date, there have been numerous studies demonstrating successful encapsulation of small molecules and biological gases into MOF structures. There has also been some progress in the characterization of MOF toxicity, behavior, and uptake in vitro. However, there is still a need to improve and develop an understanding of MOFs in biological systems. The evolution of methods for particle synthesis, surface functionalization techniques, and the scope for developing new composite-hybrid materials has produced nanomaterials with potential for enhanced capabilities such as stimulus-responsive drug release, imaging, cell-specific targeting, and theranostic capabilities. Combining these developments in synthesis, together with a rational selection of biological models, mechanistic characterization and appropriate selection of in vitro assays will offer exciting prospects for the future of this field.
11
Metal-Organic Frameworks as Delivery Systems of Small Drugs and Biological Gases
373
References 1. Helfand WH, Cowen DL (1983) Evolution of pharmaceutical oral dosage forms. Pharm Hist 25:3–18 2. Park K (2014) Controlled drug delivery systems: past forward and future back. J Control Release 190:3–8 3. Li C, Wang J, Wang Y et al (2019) Recent progress in drug delivery. Acta Pharm Sin B 9:1145–1162 4. Horcajada P, Gref R, Baati T et al (2012) Metal-organic frameworks in biomedicine. Chem Rev 112:1232–1268 5. Moghadam PZ, Li A, Wiggin SB et al (2017) Development of a Cambridge structural database subset: a collection of metal–organic frameworks for past, present, and future. Chem Mater 29:2618–2625 6. Horcajada P, Serre C, Vallet-Regi M et al (2006) Metal-organic frameworks as efficient materials for drug delivery. Angew Chem Int Ed Eng 45:5974–5978 7. Bernini MC, Fairen-Jimenez D, Pasinetti M et al (2014) Screening of bio-compatible metal– organic frameworks as potential drug carriers using Monte Carlo simulations. J Mater Chem B 2:766–774 8. Lu K, He C, Lin W (2014) Nanoscale metal-organic framework for highly effective photodynamic therapy of resistant head and neck cancer. J Am Chem Soc 136:16712–16715 9. He C, Lu K, Lin W (2014) Nanoscale metal-organic frameworks for real-time intracellular pH sensing in live cells. J Am Chem Soc 136:12253–12256 10. Liu D, Poon C, Lu K et al (2014) Self-assembled nanoscale coordination polymers with trigger release properties for effective anticancer therapy. Nat Commun 5:4182 11. Cattaneo D, Warrender SJ, Duncan MJ et al (2016) Tuning the nitric oxide release from CPO-27 MOFs. RSC Adv 6:14059–14067 12. Xiao B, Wheatley PS, Zhao X et al (2007) High-capacity hydrogen and nitric oxide adsorption and storage in a metal-organic framework. J Am Chem Soc 129:1203–1209 13. Zhou HC, Long JR, Yaghi OM (2012) Introduction to metal-organic frameworks. Chem Rev 112:673–674 14. Rieter WJ, Taylor KM, Lin W (2007) Surface modification and functionalization of nanoscale metal-organic frameworks for controlled release and luminescence sensing. J Am Chem Soc 129:9852–9853 15. Lin SX, Pan WL, Niu RJ et al (2019) Effective loading of cisplatin into a nanoscale UiO-66 metal-organic framework with preformed defects. Dalton Trans 48:5308–5314 16. He C, Lu K, Liu D et al (2014) Nanoscale metal-organic frameworks for the co-delivery of cisplatin and pooled siRNAs to enhance therapeutic efficacy in drug-resistant ovarian cancer cells. J Am Chem Soc 136:5181–5184 17. Farboudi A, Mahboobnia K, Chogan F et al (2020) UiO-66 metal organic framework nanoparticles loaded carboxymethyl chitosan/poly ethylene oxide/polyurethane core-shell nanofibers for controlled release of doxorubicin and folic acid. Int J Biol Macromol 150:178 18. Ray Chowdhuri A, Bhattacharya D, Sahu SK (2016) Magnetic nanoscale metal organic frameworks for potential targeted anticancer drug delivery, imaging and as an MRI contrast agent. Dalton Trans 45:2963–2973 19. Gao X, Cui R, Song L et al (2019) Hollow structural metal–organic frameworks exhibit high drug loading capacity, targeted delivery and magnetic resonance/optical multimodal imaging. Dalton Trans 48:17291–17297 20. Orellana-Tavra C, Koppen M, Li A et al (2020) Biocompatible, crystalline and amorphous bismuth-based metal-organic frameworks for drug delivery. ACS Appl Mater Interfaces 12:5633 21. Yang B, Ding L, Yao H et al (2020) A metal-organic framework (MOF) fenton nanoagentenabled nanocatalytic cancer therapy in synergy with autophagy inhibition. Adv Mater 32: e1907152
374
E. Linnane and D. Fairen-Jimenez
22. Zhang L, Gao Y, Sun S et al (2020) pH-responsive metal-organic framework encapsulated gold nanoclusters with modulated release to enhance photodynamic therapy/chemotherapy in breast cancer. J Mater Chem B 8:1739 23. Lu L, Ma M, Gao C et al (2020) Metal organic framework@polysilsesequioxane core/shellstructured nanoplatform for drug delivery. Pharmaceutics 12:98 24. Wyszogrodzka-Gawel G, Dorozynski P, Giovagnoli S et al (2019) An inhalable theranostic system for local tuberculosis treatment containing an isoniazid loaded metal organic framework Fe-MIL-101-NH2-from raw MOF to drug delivery system. Pharmaceutics 11:687 25. Cherkasov VR, Mochalova EN, Babenyshev AV et al (2020) Antibody-directed metal-organic framework nanoparticles for targeted drug delivery. Acta Biomater 103:223–236 26. Javanbakht S, Hemmati A, Namazi H et al (2019) Carboxymethylcellulose-coated 5-fluorouracil@MOF-5 nano-hybrid as a bio-nanocomposite carrier for the anticancer oral delivery. Int J Biol Macromol 155:876–882 27. Chen WH, Karmi O, Willner B et al (2019) Thrombin aptamer-modified metal-organic framework nanoparticles: functional nanostructures for sensing thrombin and the triggered controlled release of anti-blood clotting drugs. Sensors (Basel) 19. https://doi.org/10.3390/ s19235260 28. Pan YB, Wang S, He X et al (2019) A combination of glioma in vivo imaging and in vivo drug delivery by metal-organic framework based composite nanoparticles. J Mater Chem B 7:7683–7689 29. Chen J, Liu J, Hu Y et al (2019) Metal-organic framework-coated magnetite nanoparticles for synergistic magnetic hyperthermia and chemotherapy with pH-triggered drug release. Sci Technol Adv Mater 20:1043–1054 30. Abazari R, Ataei F, Morsali A et al (2019) A luminescent amine-functionalized metal-organic framework conjugated with folic acid as a targeted biocompatible pH-responsive nanocarrier for apoptosis induction in breast cancer cells. ACS Appl Mater Interfaces 11:45442–45454 31. Chen G, Luo J, Cai M et al (2019) Investigation of metal-organic framework-5 (MOF-5) as an antitumor drug oridonin sustained release carrier. Molecules 24:3369 32. Xue T, Xu C, Wang Y et al (2019) Doxorubicin-loaded nanoscale metal-organic framework for tumor-targeting combined chemotherapy and chemodynamic therapy. Biomater Sci 7:4615–4623 33. Miri B, Motakef-Kazemi N, Shojaosadati SA et al (2018) Application of a nanoporous metal organic framework based on iron carboxylate as drug delivery system. Iran J Pharm Res 17:1164–1171 34. Shi Z, Chen X, Zhang L et al (2018) FA-PEG decorated MOF nanoparticles as a targeted drug delivery system for controlled release of an autophagy inhibitor. Biomater Sci 6:2582–2590 35. Zhang FM, Dong H, Zhang X et al (2017) Postsynthetic modification of ZIF-90 for potential targeted codelivery of two anticancer drugs. ACS Appl Mater Interfaces 9:27332–27337 36. Li Y, Li X, Guan Q et al (2017) Strategy for chemotherapeutic delivery using a nanosized porous metal-organic framework with a central composite design. Int J Nanomedicine 12:1465–1474 37. Gao X, Zhai M, Guan W et al (2017) Controllable synthesis of a smart multifunctional nanoscale metal-organic framework for magnetic resonance/optical imaging and targeted drug delivery. ACS Appl Mater Interfaces 9:3455–3462 38. Zheng H, Zhang Y, Liu L et al (2016) One-pot synthesis of metal-organic frameworks with encapsulated target molecules and their applications for controlled drug delivery. J Am Chem Soc 138:962–968 39. Abanades Lazaro I, Haddad S, Rodrigo-Munoz JM et al (2018) Surface-functionalization of Zr-fumarate MOF for selective cytotoxicity and immune system compatibility in nanoscale drug delivery. ACS Appl Mater Interfaces 10:31146–31157 40. Orellana-Tavra C, Baxter EF, Tian T et al (2015) Amorphous metal-organic frameworks for drug delivery. Chem Commun (Camb) 51:13878–13881
11
Metal-Organic Frameworks as Delivery Systems of Small Drugs and Biological Gases
375
41. Unamuno X, Imbuluzqueta E, Salles F et al (2018) Biocompatible porous metal-organic framework nanoparticles based on Fe or Zr for gentamicin vectorization. Eur J Pharm Biopharm 132:11–18 42. Abazari R, Mahjoub AR, Ataei F et al (2018) Chitosan immobilization on bio-MOF nanostructures: a biocompatible pH-responsive nanocarrier for doxorubicin release on MCF-7 cell lines of human breast cancer. Inorg Chem 57:13364–13379 43. Horcajada P, Serre C, Maurin G et al (2008) Flexible porous metal-organic frameworks for a controlled drug delivery. J Am Chem Soc 130:6774–6780 44. Zhuang J, Kuo CH, Chou LY et al (2014) Optimized metal-organic-framework nanospheres for drug delivery: evaluation of small-molecule encapsulation. ACS Nano 8:2812–2819 45. Hu Q, Yu J, Liu M et al (2014) A low cytotoxic cationic metal-organic framework carrier for controllable drug release. J Med Chem 57:5679–5685 46. Rodriguez-Ruiz V, Maksimenko A, Anand R et al (2015) Efficient “green” encapsulation of a highly hydrophilic anticancer drug in metal-organic framework nanoparticles. J Drug Target 23:759–767 47. Wang XG, Dong ZY, Cheng H et al (2015) A multifunctional metal-organic framework based tumor targeting drug delivery system for cancer therapy. Nanoscale 7:16061–16070 48. Bellido E, Hidalgo T, Lozano MV et al (2015) Heparin-engineered mesoporous iron metalorganic framework nanoparticles: toward stealth drug nanocarriers. Adv Healthc Mater 4:1246–1257 49. Miller SR, Heurtaux D, Baati T et al (2010) Biodegradable therapeutic MOFs for the delivery of bioactive molecules. Chem Commun (Camb) 46:4526–4528 50. Chen W, Wu C (2018) Synthesis, functionalization, and applications of metal-organic frameworks in biomedicine. Dalton Trans 47:2114–2133 51. Stock N, Biswas S (2012) Synthesis of metal-organic frameworks (MOFs): routes to various MOF topologies, morphologies, and composites. Chem Rev 112:933–969 52. Tanabe KK, Cohen SM (2011) Postsynthetic modification of metal-organic frameworks--a progress report. Chem Soc Rev 40:498–519 53. Reinsch H (2016) “Green” synthesis of metal-organic frameworks. Eur J Inorg Chem 2016:4290–4299 54. Yang Q, Vaesen S, Ragon F et al (2013) A water stable metal-organic framework with optimal features for CO2 capture. Angew Chem Int Ed Eng 52:10316–10320 55. Hou S, Wu YN, Feng L et al (2018) Green synthesis and evaluation of an iron-based metalorganic framework MIL-88B for efficient decontamination of arsenate from water. Dalton Trans 47:2222–2231 56. Howarth AJ, Liu Y, Li P et al (2016) Chemical, thermal and mechanical stabilities of metal– organic frameworks. Nat Rev Mater 1:15018 57. Tamames-Tabar C, Cunha D, Imbuluzqueta E et al (2014) Cytotoxicity of nanoscaled metal– organic frameworks. J Mater Chem B 2:262–271 58. Wagner A, Liu Q, Rose OL et al (2019) Toxicity screening of two prevalent metal organic frameworks for therapeutic use in human lung epithelial cells. Int J Nanomedicine 14:7583–7591 59. Baati T, Njim L, Neffati F et al (2013) In depth analysis of the in vivo toxicity of nanoparticles of porous iron(iii) metal–organic frameworks. Chem Sci 4:1597–1607 60. Ruyra A, Yazdi A, Espin J et al (2015) Synthesis, culture medium stability, and in vitro and in vivo zebrafish embryo toxicity of metal-organic framework nanoparticles. Chemistry 21:2508–2518 61. Huang X, Brazel CS (2001) On the importance and mechanisms of burst release in matrixcontrolled drug delivery systems. J Control Release 73:121–136 62. Horcajada P, Chalati T, Serre C et al (2010) Porous metal-organic-framework nanoscale carriers as a potential platform for drug delivery and imaging. Nat Mater 9:172–178 63. Abanades Lazaro I, Haddad S, Sacca S et al (2017) Selective surface PEGylation of UiO-66 nanoparticles for enhanced stability, cell uptake, and pH-responsive drug delivery. Chem 2:561–578
376
E. Linnane and D. Fairen-Jimenez
64. Orellana-Tavra C, Haddad S, Marshall RJ et al (2017) Tuning the endocytosis mechanism of Zr-based metal-organic frameworks through linker functionalization. ACS Appl Mater Interfaces 9:35516–35525 65. Teplensky MH, Fantham M, Li P et al (2017) Temperature treatment of highly porous zirconium-containing metal-organic frameworks extends drug delivery release. J Am Chem Soc 139:7522–7532 66. Feyand M, Mugnaioli E, Vermoortele F et al (2012) Automated diffraction tomography for the structure elucidation of twinned, sub-micrometer crystals of a highly porous, catalytically active bismuth metal-organic framework. Angew Chem Int Ed Eng 51:10373–10376 67. Orellana-Tavra C, Marshall RJ, Baxter EF et al (2016) Drug delivery and controlled release from biocompatible metal–organic frameworks using mechanical amorphization. J Mater Chem B 4:7697–7707 68. Jia X, Yang Z, Wang Y et al (2018) Hollow mesoporous silica@metal-organic framework and applications for pH-responsive drug delivery. ChemMedChem 13:400–405 69. Pan QS, Chen TT, Nie CP et al (2018) In situ synthesis of ultrathin ZIF-8 film-coated MSNs for codelivering Bcl 2 siRNA and doxorubicin to enhance chemotherapeutic efficacy in drugresistant cancer cells. ACS Appl Mater Interfaces 10:33070–33077 70. Zhao HX, Zou Q, Sun SK et al (2016) Theranostic metal-organic framework core-shell composites for magnetic resonance imaging and drug delivery. Chem Sci 7:5294–5301 71. Wang Y, Yan J, Wen N et al (2020) Metal-organic frameworks for stimuli-responsive drug delivery. Biomaterials 230:119619 72. Liu Y, Zhao Y, Chen X (2019) Bioengineering of metal-organic frameworks for nanomedicine. Theranostics 9:3122–3133 73. Chen W-H, Yu X, Liao W-C et al (2017) ATP-responsive aptamer-based metal–organic framework nanoparticles (NMOFs) for the controlled release of loads and drugs. Adv Funct Mater 27:1702102 74. Liu Y, Gong CS, Lin L et al (2019) Core-shell metal-organic frameworks with fluorescence switch to trigger an enhanced photodynamic therapy. Theranostics 9:2791–2799 75. Yang X, Li L, He D et al (2017) A metal–organic framework based nanocomposite with co-encapsulation of Pd@Au nanoparticles and doxorubicin for pH- and NIR-triggered synergistic chemo-photothermal treatment of cancer cells. J Mater Chem B 5:4648–4659 76. Tan BH, Wong PT, Bian JS (2010) Hydrogen sulfide: a novel signaling molecule in the central nervous system. Neurochem Int 56:3–10 77. Szabo C (2007) Hydrogen sulphide and its therapeutic potential. Nat Rev Drug Discov 6:917–935 78. Allan PK, Wheatley PS, Aldous D et al (2012) Metal-organic frameworks for the storage and delivery of biologically active hydrogen sulfide. Dalton Trans 41:4060–4066 79. Heinemann SH, Hoshi T, Westerhausen M et al (2014) Carbon monoxide--physiology, detection and controlled release. Chem Commun (Camb) 50:3644–3660 80. Jin Z, Zhao P, Zhang J et al (2018) Intelligent metal carbonyl metal-organic framework nanocomplex for fluorescent traceable H2 O2 -triggered CO delivery. Chemistry 24:11667–11674 81. Palmer RM, Ferrige AG, Moncada S (1987) Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature 327:524–526 82. Thomas DD, Liu X, Kantrow SP et al (2001) The biological lifetime of nitric oxide: implications for the perivascular dynamics of NO and O2. Proc Natl Acad Sci U S A 98:355–360 83. Kannan MS, Guiang S, Johnson DE (1998) Nitric oxide: biological role and clinical uses. Indian J Pediatr 65:333–345 84. Rizk M, Witte MB, Barbul A (2004) Nitric oxide and wound healing. World J Surg 28:301–306 85. Malone-Povolny MJ, Maloney SE, Schoenfisch MH (2019) Nitric oxide therapy for diabetic wound healing. Adv Healthc Mater 8:e1801210
11
Metal-Organic Frameworks as Delivery Systems of Small Drugs and Biological Gases
377
86. Boykin JV Jr (2010) Wound nitric oxide bioactivity: a promising diagnostic indicator for diabetic foot ulcer management. J Wound Ostomy Continence Nurs 37:25–32. quiz 33-4 87. Park JH, Kim JY, Kim DJ et al (2017) Effect of nitric oxide on human corneal epithelial cell viability and corneal wound healing. Sci Rep 7:8093 88. Bonfiglio V, Camillieri G, Avitabile T et al (2006) Effects of the COOH-terminal tripeptide alpha-MSH(11-13) on corneal epithelial wound healing: role of nitric oxide. Exp Eye Res 83:1366–1372 89. Diring S, Wang DO, Kim C et al (2013) Localized cell stimulation by nitric oxide using a photoactive porous coordination polymer platform. Nat Commun 4:2684 90. Kim C, Diring S, Furukawa S et al (2015) Light-induced nitric oxide release from physiologically stable porous coordination polymers. Dalton Trans 44:15324–15333 91. Zaki NM, Tirelli N (2010) Gateways for the intracellular access of nanocarriers: a review of receptor-mediated endocytosis mechanisms and of strategies in receptor targeting. Expert Opin Drug Deliv 7:895–913 92. Somsel Rodman J, Wandinger-Ness A (2000) Rab GTPases coordinate endocytosis. J Cell Sci 113(Pt 2):183–192 93. Stenmark H, Olkkonen VM (2001) The Rab GTPase family. Genome Biol 2:REVIEWS3007 94. de Souza W, Sant’Anna C, Cunha-e-Silva NL (2009) Electron microscopy and cytochemistry analysis of the endocytic pathway of pathogenic protozoa. Prog Histochem Cytochem 44:67–124 95. Kaksonen M, Roux A (2018) Mechanisms of clathrin-mediated endocytosis. Nat Rev Mol Cell Biol 19:313–326 96. Recouvreux MV, Commisso C (2017) Macropinocytosis: a metabolic adaptation to nutrient stress in cancer. Front Endocrinol (Lausanne) 8:261 97. Bathori G, Cervenak L, Karadi I (2004) Caveolae--an alternative endocytotic pathway for targeted drug delivery. Crit Rev Ther Drug Carrier Syst 21:67–95 98. Durymanov M, Permyakova A, Sene S et al (2019) Cellular uptake, intracellular trafficking, and stability of biocompatible metal-organic framework (MOF) particles in Kupffer cells. Mol Pharm 16:2315–2325 99. McGuire CV, Forgan RS (2015) The surface chemistry of metal-organic frameworks. Chem Commun (Camb) 51:5199–5217 100. Forgan RS (2019) The surface chemistry of metal-organic frameworks and their applications. Dalton Trans 48:9037–9042 101. Schaate A, Roy P, Godt A et al (2011) Modulated synthesis of Zr-based metal-organic frameworks: from nano to single crystals. Chemistry 17:6643–6651 102. Morris W, Wang S, Cho D et al (2017) Role of modulators in controlling the colloidal stability and polydispersity of the UiO-66 metal-organic framework. ACS Appl Mater Interfaces 9:33413–33418 103. Wuttke S, Braig S, Preiss T et al (2015) MOF nanoparticles coated by lipid bilayers and their uptake by cancer cells. Chem Commun (Camb) 51:15752–15755 104. Taylor-Pashow KM, Della Rocca J, Xie Z et al (2009) Postsynthetic modifications of ironcarboxylate nanoscale metal-organic frameworks for imaging and drug delivery. J Am Chem Soc 131:14261–14263 105. Fan G, Dundas CM, Zhang C et al (2018) Sequence-dependent peptide surface functionalization of metal-organic frameworks. ACS Appl Mater Interfaces 10:18601–18609 106. Rijnaarts T, Mejia-Ariza R, Egberink RJ et al (2015) Metal-organic frameworks (MOFs) as multivalent materials: size control and surface functionalization by monovalent capping ligands. Chemistry 21:10296–10301 107. Wang S, Morris W, Liu Y et al (2015) Surface-specific functionalization of nanoscale metalorganic frameworks. Angew Chem Int Ed Eng 54:14738–14742 108. Ernsting MJ, Murakami M, Roy A et al (2013) Factors controlling the pharmacokinetics, biodistribution and intratumoral penetration of nanoparticles. J Control Release 172:782–794
378
E. Linnane and D. Fairen-Jimenez
109. Qie Y, Yuan H, von Roemeling CA et al (2016) Surface modification of nanoparticles enables selective evasion of phagocytic clearance by distinct macrophage phenotypes. Sci Rep 6:26269 110. Nagayama S, Ogawara K, Fukuoka Y et al (2007) Time-dependent changes in opsonin amount associated on nanoparticles alter their hepatic uptake characteristics. Int J Pharm 342:215–221 111. Harris JM, Chess RB (2003) Effect of pegylation on pharmaceuticals. Nat Rev Drug Discov 2:214–221 112. Gabizon A, Shmeeda H, Barenholz Y (2003) Pharmacokinetics of pegylated liposomal doxorubicin: review of animal and human studies. Clin Pharmacokinet 42:419–436 113. Anders CK, Adamo B, Karginova O et al (2013) Pharmacokinetics and efficacy of PEGylated liposomal doxorubicin in an intracranial model of breast cancer. PLoS One 8:e61359 114. Suk JS, Xu Q, Kim N et al (2016) PEGylation as a strategy for improving nanoparticle-based drug and gene delivery. Adv Drug Deliv Rev 99:28–51 115. Rattan R, Bhattacharjee S, Zong H et al (2017) Nanoparticle-macrophage interactions: a balance between clearance and cell-specific targeting. Bioorg Med Chem 25:4487–4496 116. Jiang PD, Zhao YL, Deng XQ et al (2010) Antitumor and antimetastatic activities of chloroquine diphosphate in a murine model of breast cancer. Biomed Pharmacother 64:609–614 117. Dong H, Yang GX, Zhang X et al (2018) Folic acid functionalized zirconium-based metalorganic frameworks as drug carriers for active tumor-targeted drug delivery. Chemistry 24:17148–17154 118. Vasvani S, Kulkarni P, Rawtani D (2019) Hyaluronic acid: a review on its biology, aspects of drug delivery, route of administrations and a special emphasis on its approved marketed products and recent clinical studies. Int J Biol Macromol 151:1012 119. Kim K, Choi H, Choi ES et al (2019) Hyaluronic acid-coated nanomedicine for targeted cancer therapy. Pharmaceutics 11:301 120. Rao NV, Yoon HY, Han HS et al (2016) Recent developments in hyaluronic acid-based nanomedicine for targeted cancer treatment. Expert Opin Drug Deliv 13:239–252 121. Huang G, Huang H (2018) Application of hyaluronic acid as carriers in drug delivery. Drug Deliv 25:766–772 122. Shu F, Lv D, Song X-L et al (2018) Fabrication of a hyaluronic acid conjugated metal organic framework for targeted drug delivery and magnetic resonance imaging. RSC Adv 8:6581–6589 123. Cai W, Gao H, Chu C et al (2017) Engineering phototheranostic nanoscale metal–organic frameworks for multimodal imaging-guided cancer therapy. ACS Appl Mater Interfaces 9:2040–2051 124. Lai X, Liu H, Zheng Y et al (2019) Drug loaded nanoparticles of metal-organic frameworks with high colloidal stability for anticancer application. J Biomed Nanotechnol 15:1754–1763 125. Illes B, Hirschle P, Barnert S et al (2017) Exosome-coated metal–organic framework nanoparticles: an efficient drug delivery platform. Chem Mater 29:8042–8046 126. Canton J (2018) Macropinocytosis: new insights into its underappreciated role in innate immune cell surveillance. Front Immunol 9:2286 127. Vercauteren D, Vandenbroucke RE, Jones AT et al (2010) The use of inhibitors to study endocytic pathways of gene carriers: optimization and pitfalls. Mol Ther 18:561–569 128. van Tonder A, Joubert AM, Cromarty AD (2015) Limitations of the 3-(4,5-dimethylthiazol-2yl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay when compared to three commonly used cell enumeration assays. BMC Res Notes 8:47 129. Wang S, Yu H, Wickliffe JK (2011) Limitation of the MTT and XTT assays for measuring cell viability due to superoxide formation induced by nano-scale TiO2. Toxicol in Vitro 25:2147–2151
Chapter 12
MOFs and Biomacromolecules for Biomedical Applications Francesco Carraro, Miriam de J. Velásquez-Hernández, Mercedes Linares Moreau, Efwita Astria, Christopher Sumby, Christian Doonan, and Paolo Falcaro
12.1
Introduction
Biomacromolecules play a key role in natural biological systems but are increasingly finding greater application as therapeutics, and as biomarkers for disease prognosis and monitoring ongoing treatment. Indeed, with the rapid development of new medical technologies and treatments, especially for age-related and emerging diseases, the need to cost effectively and reliably store or transport biomacromolecules (biopreservation) and improve detection (biosensing) continues to grow. A significant challenge in this area is that biomacromolecules are typically unstable when removed from their finely tuned biological environment and lose their functionality when exposed to elevated temperatures, nonaqueous media, or non-native pH. Low-temperature storage and lyophilization (freeze-drying) are employed to protect biomacromolecules for storage and transport; however, these forms do not typically allow biomacromolecules to be directly used as therapeutics or sensing. Other strategies to improve biomolecule stability for applications in biomedicine include using more stable homologs from extremophiles, genetic engineering, and posttranslational (chemical) modification of biomacromolecules to provide access to more stable variants; however, these are not universal approaches. To this end, researchers have focused on developing more general approaches, for enhancing biomacromolecule stability and providing protection that also facilitate practical use of the biomolecule. Among these, porous materials have been actively researched for
F. Carraro · M. d. J. Velásquez-Hernández · M. Linares Moreau · E. Astria · P. Falcaro (*) Institute of Physical and Theoretical Chemistry, Graz University of Technology, Graz, Austria e-mail: [email protected] C. Sumby · C. Doonan Department of Chemistry and Centre for Advanced Nanomaterials, The University of Adelaide, Adelaide, Australia © Springer Nature Switzerland AG 2021 P. Horcajada Cortés, S. Rojas Macías (eds.), Metal-Organic Frameworks in Biomedical and Environmental Field, https://doi.org/10.1007/978-3-030-63380-6_12
379
380
F. Carraro et al.
stabilizing biomacromolecules either by adsorbing or grafting them to the surface of the material, via pore infiltration or via sol-gel encapsulation [1–3]. Materials explored for this purpose range from soft hydrogels to hard inorganic materials like silica, metallophosphates, and metal oxides [1–4]. While a vast array of biocomposites have been prepared that confer stability to protein-based therapeutics and biomacromolecule-based sensing platforms, there is still a need to develop new strategies to stabilize and protect biomacromolecules. Although inorganic materials (e.g., silica) have been synthesize from biocompatible precursors (e.g., sodium silicates in water), some intrinsic problems, such as large degrees of shrinkage (up to 80%) or limited range of pore size tuneability, morphology and polarity, result in a limited range of biomacromolecules that can be immobilized [5]. The building block synthetic approach, pre- and post-synthetic chemical mutability, and intrinsic porosity of metal-organic frameworks (MOFs) provide opportunities for application to biomacromolecule protection that solve or minimize some of the existing challenges of other materials. This includes tuneable biocompatibility, access to the biomacromolecule through a robust and regular pore network, and differing framework chemistry that can provide sustained and targeted biologically relevant release. MOF biocomposites are obtained by integrating biomacromolecules with MOFs, which has provided an emerging class of materials for biomedicine, biopreservation, biosensing, and biocatalysis [6–8]. Depending on the synthesis protocol used and the spatial localization of the biomacromolecules in or on the MOF particle, different types of MOF biocomposites (Fig. 12.1) can be identified: 1. Biomacromolecule-on-MOF: In this configuration, preformed MOF particles are surface-decorated with biomacromolecules. The preparation process is named surface immobilization and involves (a) adsorption via noncovalent interactions or (b) grafting (covalent bonding, also termed bioconjugation) of biomacromolecules on MOFs.
Fig. 12.1 Schematic view of the general strategies for the synthesis of MOF biocomposites
12
MOFs and Biomacromolecules for Biomedical Applications
381
2. Biomacromolecule@MOF: In this composite, biomacromolecules are embedded within the MOF particles. These composites can be obtained by (a) infiltration or (b) one-pot encapsulation methods. In this book chapter, we first discuss the different preparation methods used to prepare MOF biocomposites, including the merits and challenges of each strategy. We then focus on the different MOF biocomposites by examining each single class of biomacromolecules (proteins, fatty acids, carbohydrates, and nucleic acids) and highlight the applications of these materials for biomedical applications.
12.2
Synthesis Methods
12.2.1 Surface Immobilization Biomacromolecule-on-MOF composites are obtained by immobilizing biomacromolecules onto the surface of MOF particles by surface adsorption or grafting. These are processes heavily influenced by strategies for biomacromolecule immobilization on other materials and, while not directly dependent on the permanent porosity of MOFs, are enabled by the diverse chemical structures of MOFs and the mutability of their surface chemistry.
12.2.1.1
Adsorption of Biomacromolecules on MOFs
The easiest approach to prepare biomacromolecule-on-MOF composites is the surface adsorption of biomacromolecules on preformed MOF particles. This immobilization method depends on noncovalent interactions (e.g., electrostatic interactions, hydrophilic/hydrophobic interactions, or hydrogen bonds) between the biomacromolecules and the MOF surface. Electrostatic interactions are often exploited to decorate surfaces with biomacromolecules [9, 10]. For example, in the case of proteins, surface-exposed amino acids determine the overall charge of the protein and the electrostatic properties of this system depend on the pH of the solution and on the pK values of the ionizable groups [11]. Thus, by selecting the appropriate biomolecule, adsorption conditions, and an MOF material with suitable functional groups, introduced pre- or postsynthetically, it is possible to control the attractive or repulsive forces. To favor attraction and adsorption, the pH and ionic strength of the solution should be tuned to induce opposite net charges on the MOF and on the biomacromolecule [12]. For example, Li et al. adsorbed pectinase on polymethacrylic acid (PMMA) decorated UiO-66-NH2 particles and showed that if the pH of the solution was significantly higher than the pectinase isoelectric point (pH 3.8), both the protein and the PMMA-
382
F. Carraro et al.
decorated MOF surfaces were negatively charged and the immobilization was not effective [12]. Conversely, close to the isoelectric point, successful adsorption was achieved, allowing the biomacromolecule to form hydrogen bonds with the functional groups exposed on the PMMA decorated MOF surface [13]. These weak, but multipoint, interactions typically impede leaching (i.e., release of the biomacromolecules in solution) and stabilize the biomacromolecule but can be contingent on maintenance of the conditions [13]. Additionally, however, when the interaction with a surface is sufficiently strong, changes in the biomacromolecule structure (e.g., protein unfolding) and in its bioactivity can result [14]. Hydrophilic/hydrophobic surfaces can also be used to favor the adsorption of biomacromolecules. In general, proteins possess a high affinity for hydrophobic surfaces [15]; however, such interactions can perturb the proteins tertiary structure and result in loss of activity [16]. For example, Doonan and coworkers demonstrated that enzymes adsorbed on zeolitic imidazolate frameworks (ZIFs) of varied hydrophilicity/phobicity showed different enzymatic activities [17]. When catalase (CAT) was adsorbed on hydrophilic MAF-7 (synthesized from Zn2+ and 3-methyl-1,2,4triazolate) or ZIF-90 (synthesized from Zn2+ and 2-imidazolatecarboxaldehyde), its enzymatic activity was largely maintained. Conversely, when CAT was adsorbed on hydrophobic ZIF-8 (synthesized from Zn2+ and 2-methyilmidazole (HmIM)), the enzymatic activity was inhibited. However, different biomacromolecules have different conformational sensitivities and the potential activity loss should be assessed case by case. For example, some antibodies and enzymes can be supported on different hydrophobic surfaces, without showing significant unfolding of the protein structure [18–20]., Furthermore, the interaction between an enzyme and the MOF surface can influence the orientation of the biomacromolecule on the surface, as determined by Pan et al. in the case of recombinant T4 phage lysozyme partially encapsulated into ZIF-8 particles [21]. Thus, for biomacromolecules-on-MOF biocomposites, understanding the interaction between the MOF surface and biomacromolecules is a fundamental research topic that needs to be further understood to allow for their development.
12.2.1.2
Grafting of Biomacromolecules on MOFs
A more robust method of biomacromolecule immobilization onto MOF surfaces is to form a covalent bond between specific functional groups on the surfaces of the biomolecule and MOF. This strategy, termed grafting, takes advantage of the vast library of covalent bond forming protocols available to combine proteins and MOFs. A widely used grafting procedure reacts carboxylic and amino groups via the N, N0 -dicyclohexylcarbodiimide (DCC)-mediated coupling reaction or the 1-ethyl-3(3-dimethylaminopropyl)carbodiimide/N-hydroxysuccinimide (EDC/NHS) method [22]. These protocols are widely used for the permanent immobilization of proteins [23], nucleic acid [23], carbohydrates [24], and cells [25] on different materials. As DCC is not soluble in water, the DCC coupling can be only be applied if the selected biomacromolecule is stable in the organic solvent used for the reaction. Conversely,
12
MOFs and Biomacromolecules for Biomedical Applications
383
Fig. 12.2 Schematic view of the trypsin immobilization onto DCC-activated MOFs. (Adapted with permission of Wiley from Ref. [26], © 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)
the EDC/NHS method can be performed in water and buffer solutions, and this is the reason for the preferential use of this coupling reaction in the preparation of biomacromlecules-on-MOFs via grafting. Carbodiimide coupling techniques were exploited to prepare biomacromoleculeson-MOF biocomposites with proteins [26, 27], carbohydrates [28–30] and nucleic acids [31]. As an example, Huang et al. [26] reported the DCC-mediated coupling of trypsin (a digestive enzyme) to the surface of MIL-88-NH2(Cr) (Fig. 12.2). In this case, due to the approach used, the authors suggested that the covalent bond was formed on the uncoordinated carboxylic acid group of the MOF linker. To use the amino-functionality in NH2-MIL-53(Al) for the immobilization of β-glucosidase, Falcaro and coworkers used a glutaraldehyde-mediated grafting procedure [27]. These examples demonstrate the versatility of the “crystals as molecules” reactivity of MOFs for grafting that could be further extended via postsynthetic modification (PSM) processes [32].
12.2.1.3
General Considerations for Biomacromolecules-On-MOF Composites
This method is extremely versatile, and a variety of biomacromolecules-on-MOF composites can be prepared. The flexibility of this system derives from: (1) the large number of available MOFs; (2) postsynthetic modification methods for fine-tuning of the chemical functional groups on the MOF surface; (3) number of conditions/ protocols available for adsorption/bioconjugation; and, (4) the possibility to perform the bioconjugation under biocompatible conditions (e.g., in water or buffer solution). The final, solid, biocomposite can be easily recovered and recycled, thus facilitating the use of biomacromolecules mostly for biocatalytic, biosensing, and drug delivery applications [6, 27, 33]. It should be noted that not all the reports pioneering biomacromolecules-on-MOF composites exploit the porous properties of MOFs and, in many cases, only the external surface of the porous crystal is used, thus providing little difference to nonporous nanoparticles (aside from contributing slightly less mass to the biocomposite) [6]. However, it has been recently shown that surface functionalization methods can significantly improve the properties of
384
F. Carraro et al.
MOF-based drug delivery systems where MOF pores are infiltrated with therapeutics. Indeed, several therapeutic@MOF systems have shown an improved colloidal stability, blood circulation time, and cellular uptake when their surface is bioconjugated with selected biomacromolecules [31, 34–36].
12.2.2 Embedding of Biomacromolecules in MOFs Homogenous distribution of biomacromolecules within MOFs has been achieved either via infiltration or one-pot encapsulation. These biocomposites are named biomacromolecule@MOF.
12.2.2.1
Infiltration
Infiltration consists of the insertion of biomacromolecules into the pores of preformed MOF particles. For this approach, it is crucial that the pore-size distribution and connectivity can accommodate the selected biomacromolecules and, moreover, allow for the diffusion of the cofactors/substrates through the material once the biomacromolecule has been infiltrated. Because of the typical size of biomacromolecules, infiltration methods are usually limited to mesoporous MOFs. For example, successful infiltrations were performed with Aspergillus saitoi (2.85 nm) [37], organophosphorus acid anhydrolase (Fig. 12.3, 4.4 nm) [38], green fluorescent protein (4.5 nm) [39], microperoxidase-11 (3.3 nm) [40, 41] in MIL-101(Al)-NH2 (up to 3.6 nm) [37], NU-1003 (Fig. 12.3, up to 4.5 nm) [38], IRMOFs (up to 5 nm) [39], Tb-TATB (up to 4.7 nm) [40], respectively. Although MOFs with pores larger than biomacromolecules are desired, they are not strictly necessary. For example, it was observed that certain biomacromolecules with dimensions slightly bigger than the MOF pore aperture can undergo conformational
Fig. 12.3 (a) Schematic illustration of the infiltration of organophosphorus acid anhydrolase (OPAA) in the mesoporous channels of the MOF (NU-1003). (b) Time-resolved confocal laser scanning microscopy images of a single crystal of OPAA@NU-1003 collected during the infiltration process (scale bar is 10 μm). (Adapted with permission of ACS from Ref. [38], https://pubs.acs. org/doi/10.1021/acsnano.6b04996, further permissions related to the material excerpted should be directed to the ACS)
12
MOFs and Biomacromolecules for Biomedical Applications
385
changes that allow access to the MOF structure [42]. So far, the infiltration strategy has been applied to immobilize proteins [38, 39, 42, 43] and nucleic acids [44, 45] into MOFs. Experimentally, the infiltration protocol involves exposing the MOF crystals to a solution containing the biomacromolecule. To maximize the loading of the biomacromolecules via infiltration, the diffusion within the MOF pores should be facilitated; however, diffusion depends on: (1) adsorption–desorption equilibria process that is governed by the specific biomacromolecules–MOF affinity (e.g., electrostatic and hydrophobic/philic interactions) and by MOF particle size; and, (2) the size of the biomacromolecules with respect to the MOF pore size [44]. In addition, aspects such as the MOF structural stability and biomacromolecule structural preservation should be assessed for each protocol. For example, Hidalgo et al. infiltrated RNA molecules in MIL-100 and MIL-101-NH2 MOFs that possess two types of mesocages (25 & 29 or 29 & 34 Å, respectively) accessible through microporous windows (5 & 8.6 Å or 12 & 16 Å, respectively) [44]. In this case, to favor the infiltration, the anionic nature of the nucleic acids (negatively charged biomacromolecules) was paired with a cationic MOF nanoparticles (positively charged porous matrix). By changing the pH, the surface charge of the MOF particles was tuned, and it was found that for pH50% wt) obtained by washing the sample with water. TD-EtOH (b) represents the main phases (>50% wt) obtained by washing the sample first with water and then with ethanol. (c) BSA release profiles from micrometric BSA@ZIF particles with different phases. (Adapted from Ref. [54] – Published by The Royal Society of Chemistry. https://creativecommons. org/licenses/by/3.0/)
12
MOFs and Biomacromolecules for Biomedical Applications
391
different phases encountered for the encapsulation of biomacromolecules are at its infancy, but encouraging pioneering work suggests that engineering of ZIF crystalline phases will impel the progress of biomacromolecule@ZIF biocomposites. For example, Wu et al. demonstrated that the catalytic activity of the enzyme (e.g., glucose oxidase, GOx) encapsulated in amorphous ZIF was up to 20 times higher compared to the same enzyme encapsulated in ZIF-8 with sod topology [71]. The authors associated the higher performance of enzyme@amorphousZIF biocomposites with the presence of coordination defects and mesopores in the amorphous MOF particles that facilitated the reagents diffusion.
Fig. 12.7 (a) (i) Schematic representation of a microfluidic setup used for the preparation of protein@ZIF-8 composites; the residence time prior quenching can be varied by changing the length of the reactor or the flow rate. (ii) Average crystallite size of BSA@ZIF-8 obtained, versus the ethanol flow rate employed. The red line is the fitted exponential decay (crystallite size¼a +bex/t, with a¼533, b¼22030, τ¼0.60.1, x¼flow rate ratio, R2¼0.98). (iii) Average particle size obtained from AFM topography as a function of the residence time, including a power law fit of the experimental data (particle size¼a+bxc, with a¼453, b¼31, c¼0.60.1, x¼residence time, R2¼0.97). (Adapted from Ref. [72], Published by Wiley-VCH Verlag GmbH & Co. KGaA. https:// creativecommons.org/licenses/by/4.0/) (b) Top: schematic representation of the synthesis of enzyme@MOF biocomposites in a microfluidic laminar flow that lead to defective MOF particles, as reported in ref. [73]. Bottom: schematic representation of enzyme@MOF without defects obtained in bulk solution. (c) Schematic representation of the mechanochemical synthesis of enzyme@MOF biocomposites and of their biocatalytic activity and protective properties. (Adapted from Ref. [75], Copyright © 2019, Springer Nature https://creativecommons.org/licenses/by/4.0/)
392
F. Carraro et al.
Recent Developments of Encapsulation Synthetic Protocols So far, all the discussed strategies for the synthesis of MOF biocomposites were solution-based syntheses performed in batch by mixing different reagents in a vial. Recently, in the field of direct encapsulation, two novelties were introduced: the syntheses using flow reactors and the mechanochemical syntheses (Fig. 12.7). Carraro et al. and Hu et al. simultaneously explored two different flow chemistry approaches for the synthesis of protein@ZIF-8 materials [72, 73]. Carraro et al. reported that the continuous flow synthesis of protein@ZIF-8 biocomposites could provide control over particle size (Fig. 12.7a) [72]. It was found that the synthesis of the protein@ZIF-8 biocomposite started with the formation of amorphous protein@ZIF-8 particles and that ethanol triggered the crystallization of the MOF. By using a simple flow setup (e.g., Y and T mixers, 1/16” PFA tubes), it was possible to control the residence time of the growing amorphous protein@ZIF8 particles prior the introduction of an ethanol flow: this triggered the MOF crystallization and stopped the growth of the particle size. This strategy was employed to encapsulate a protein therapeutic (α1-antitrypsin) in ZIF-8. Hu et al. controlled the protein encapsulation in a microfluidic laminar flow system (PDMS chip) by tuning the residence time of the growing ZIF-8 particles prior to introducing a flow of the enzyme solution (Fig. 12.7b) [73]. This method yielded GOx@ZIF-8 that showed 98% of the activity of the native GOx, whereas bulk synthesized GOx@ZIF-8 showed less than 15% activity of the native GOx. The authors explained the enhanced activity by the presence of defects in the laminar flow synthesized MOF framework (e.g., mesopores due to Zn coordination defects) that were associated with an easier diffusion of reagents through the MOF matrix. We posit that the syntheses in flow could be applied to the highthroughput preparation of biomacromolecules@MOF composites with controlled properties (e.g., therapeutic dose and release profile). In all these solution-based one-pot encapsulation strategies, the choice of the MOF matrix is limited to the compatibility of the MOF synthesis conditions and the stability of the biomacromolecules. This is particularly important in the case of proteins and nucleic acids that can irreversibly degrade if exposed to high temperatures, organic solvents, denaturing agents (e.g., urea), or extreme pHs [49]. Therefore, these wet approaches are generally only employed for MOFs that can assemble in mild conditions, like ZIFs [54]. To the best of our knowledge, there are no published examples of the one-pot encapsulation of biomacromolecules via solution-based synthesis in MOFs that are typically synthesized in harsh conditions (e.g., high temperature and organic solvents), such as UiOs [74]. As an alternative to solvent-based encapsulation strategies, mechanochemical synthesis was proposed for the direct encapsulation of biomacromolecules in MOFs (Fig. 12.7c) [75]. Mechanochemical processes (e.g., ball milling) are industrially scalable solvent-free methods and are commonly employed for the processing of different materials. Wei et al. reported the synthesis of enzymes@MOFs (e.g., β-glucosidase, invertase, catalase in ZIF-8, UiO-66-NH2, and Zn-MOF-74) via ball milling [75]. The powdered MOF precursors were added into a zirconia grinding
12
MOFs and Biomacromolecules for Biomedical Applications
393
jar containing lyophilized enzyme and the mixture was ground to obtain the enzyme@UiO-66-NH2 biocomposites. Once encapsulated, the enzymes maintained their enzymatic activity and showed increased resistance to proteases. Based on this result, ball milling processes are an attractive method to expand the choice of the MOF matrixes for the encapsulation of biomacromolecules. For each case, the compatibility of the ball-milling process with the stability of the biomacromolecule should be assessed.
General Considerations on Biomacromolecules@MOF Composites Obtained Via Encapsulation When compared with surface immobilization of biomacromolecules on the MOF surface, encapsulation methods provide a high degree of protection against harsh environments (e.g., temperature, organic solvents, and proteolytic agents). For example, enzymes and antibodies encapsulated in MOFs are usually not affected by proteolytic agents, since the porous framework acts as a molecular sieve and blocks the access of these digestive agents [17, 76]. In comparison to infiltration methods, encapsulation has the advantage of being MOF pore size-independent, as the MOF grows around the biomacromolecules [49]. In fact, even micrometric bioentities, including virus and cells, can be encapsulated in MOF shells following one-pot encapsulation methods [52, 53, 55, 67]. In the case of smaller bioentities, their spatial distribution within MOF crystals is not defined a priori as in the case of infiltration strategies, but depends on the bioentity nature and synthesis conditions. The recyclability of the encapsulated biomacromolecule is usually improved when compared to biocomposites prepared via surface immobilization [33]. In fact, repeated washings of a biomacromolecule@MOF biocomposite typically do not show significant leaching [17, 48]. Conversely, in the case of protein adsorption, repeated washings, especially under conditions that weaken the non-covalent interactions, can lead to the removal of the adsorbed biomacromolecule [17, 48].
12.2.3 General Properties of MOFs Biocomposites Prior to discussing the different class of MOF biocomposites and their applications, we introduce some fundamental concepts such as controlled release, biocompatibility, and particle size. These properties are often used to assess the performances of biocomposites for biomedicine and other biotechnological applications.
12.2.3.1
Controlled MOF Degradation and Cargo Release
Pharmacokinetics describes the fate of an administrated substance and includes the uptake by the body, its transformation, the biodistribution in the tissues, and finally, its removal from the organism [77]. To understand the therapeutic properties of a
394
F. Carraro et al.
drug, pharmacokinetics studies are needed. The biodistribution is related to the transfer and accumulation of the drug/carrier within the body [78]. Studies of the localization of the drug delivery system provide information on the biodistribution that describes the capability of the system to target organs [79]. The use of nanocarriers for drug delivery can allow precise control over different aspects of pharmacokinetics and biodistribution, by modifying the physicochemical properties of the carrier. An additional relevant aspect in drug delivery is the release profile that describes the amount of drug that is released from the carrier into the surrounding environment as a function of time [80]. The release kinetics can drastically influence the therapeutic effect of a drug and the efficacy of an administration method [80]. For example, a fast release is usually preferred for analgesics and anticoagulants [81]. Conversely, a slow release would be preferred for prolonged treatments that could replace frequent administration via parenteral route [82–84]. For example, treatments that require frequent injections and, consequently, pain and discomfort for the patient are protein-based treatments such as insulin, growth hormones, or oxytocin [84, 85]. Typically, the drug release profile from a carrier is characterized by a typical unwanted initial burst and followed by a slower sustained release [86]. In the case of MOF biocomposites, by tuning the carrier structure (e.g., different MOF topologies [8]), composition (e.g., different MOFs [8]), or the drug spatial localization (e.g., different drug immobilization methods as previously discussed) within the carrier, the burst effect could be minimized and a steady sustained drug release could be obtained [86]. Recently, research into drug delivery has moved from regular drug delivery systems (DDS) [87] that exploit nonspecific diffusion to active-targeting and stimuli-responsive materials that can control the carrier localization, release time, and dosage [87]. Internal stimuli (e.g., pH, chemical environment, and temperature) that are related to the local environment of the target cells/tissues could trigger the carrier decomposition and the drug release. Alternatively, the release could be regulated via external controls like light, magnetic field, or temperature [88]. In this context, MOFs possess properties that can be exploited for their use as carriers. By selecting the appropriate building blocks, it is possible to synthesize MOFs with different stabilities to chemical or physical stimuli and impart either regular DDS properties or triggered-release responses. For example, MOF-based systems have been shown to change their structure or to decompose under specific conditions including acidic pH, presence of certain anions, and irradiation with light [88]. An exemplary case of a responsive MOF for drug delivery is ZIF-8. The widespread interest in ZIF-8 is due to several reasons: (i) the encapsulation of drugs/biomacromolecules can be performed in aqueous media; (ii) the drug/ biomacromolecule loading and release efficiency can reach 100%; (iii) ZIF-8 matrix can protect the cargo against harsh conditions; and (iv) the cargo release can be controlled either by exposing the ZIF biocomposite to pH below 6.5 or to chelating agents (e.g., ethylenediaminotetraacetic acid, EDTA) [89]. Since the cargo is released via the decomposition of ZIF-8, it is important to study the effect of different chemical environments on the MOF stability and the degradation
12
MOFs and Biomacromolecules for Biomedical Applications
395
Fig. 12.8 (a) EDX elemental maps of fresh ZIF-8 powder. (b) EDX elemental maps of the powder recovered after 24 h of incubation in PBS 1x pH 7.4. (c) 31P NMR of PBS prepared in D2O before (lowest trace) and after adding ZIF-8 particles (0.5 mg mL1, 1 and 24 h, middle and upper trace, resp.). (d) Quantitative determination of 2-methylimidazole released after the incubation process in PBS (1 h, 3 h, 6 h and 24 h). (Reproduced with permission from Ref. [89] – Published by The Royal Society of Chemistry. https://creativecommons.org/licenses/by/3.0/)
mechanism. Only a profound understanding of these aspects will permit to design a ZIF-8 drug delivery system with precise controlled release properties. For example, Luzuriaga et al. showed that ZIF-8 particles are degraded in several buffer solutions that are commonly used to mimic the physical conditions [90]. Phosphate buffer solution 1X (PBS 1X) is commonly employed because it closely mimics the pH, osmolarity, and ion concentrations of the human body. In a detailed study, Velásquez-Hernández et al. investigated the mechanism of ZIF-8 particle degradation in PBS 1X (Fig. 12.8) [89]. It was found that the coordination equilibrium between Zn2+ and HmIM in solution is changed by the presence of a phosphate buffer. Due to having a high affinity for Lewis metal centers, the phosphates induce the formation of insoluble zinc phosphate by-products and favor the release of HmIM from the composite into solution. The pH of the buffer (pH ¼ 7.4) may also favor this process, since under these conditions, the ligand can be protonated (pKa1 ¼ 7.85; pKa2 ¼ 15.1) and Zn2+ coordinating ability is compromised. These investigations into the stability of MOFs in buffer solutions and bodily fluids are fundamental to anticipate side effects for MOF-based DDS [7] and furthermore to assess biocatalytic and biosensing activity data for MOF biocomposites (e.g., enzyme@MOF) that are often tested or stored in different buffers and pHs conditions [91].
12.2.3.2
MOF Biocompatibility
When an MOF biocomposite is used for drug delivery, the MOF is degraded and releases both the drug and the MOF building blocks (i.e., cations and ligands) in the body. Therefore, a fundamental step for the development of MOF biocomposites for biomedical applications is the assessment of the toxicity of MOF constituents. Horcajada, Serre, and coworkers suggested the use of MOFs made of nature-derived or biocompatible building blocks and named them bioMOFs (Fig. 12.9) [7]. Endogenous molecules (amino acids, peptides, proteins, nucleobases, carbohydrates, and
396
F. Carraro et al.
Fig. 12.9 Schematic view of the building blocks used for the synthesis of BioMOFs, the concept is explained in detail in Ref. [7]
porphyrins) or exogenous bioactive ingredients (nicotinic acid, curcumin, olsalazine, and some dicarboxylic acid including fumaric acid) were selected as ligand candidates [7]. For the metal nodes, cations that are part of the daily requirement of the human body would be the best choice [92]. Nevertheless, each metal has its own toxicity which is quantified by the median lethal dose (LD50). LD50 is defined as the amount of compound required to kill 50% of a tested population within a selected time [92]. Based on this, the metal cations with low LD50 values that can be used for the synthesis of biocompatible MOFs are Mg2+ (LD50 MgSO4 ¼ 5000) > Ca2+ (LD50 CaCl2 ¼ 1940) > Fe3+ (LD50 FeCl3 ¼ 450) > Fe2+ (LD50 FeCl2 ¼ 984) > Zn2+ (LD50 Zn(OAc)2 ¼ 100–600) [7, 93]. Referring to ZIF-8, it has been determined that an excessive concentration of this MOF has a cytotoxic effect on different cell lines (i.e., HEK-293, MDA-MB-231, HaCaT, RAW 264.7, NIH/3T3, and MG-63) [94]. The reason proposed was that the released Zn2+ cations activate apoptotic pathways in cells; however, it was found that a concentration of up to 30 μg mL1 only causes a small reduction of cell viability to approximately 80% (i.e., IC20) compared to the control, and thus any value below this threshold would be suitable for drug delivery applications [94]. These values are useful to perform a preliminary assessment of the amount of an MOF that could be administrated in one dose. However, MOFs that target clinical biomedical applications would need to be studied both in vitro and in vivo. In vitro studies provide fundamental information on some aspects of cytotoxicity, but the biocompatibility of a new material cannot be fully assessed without in vivo studies. In fact, inside a living system, there are several important aspects (e.g., interferences, permanence in the circulatory system, accumulation in organs, and immune response) that could influence the MOF toxicity or show side effects that are not predictable from in vitro studies [95].
12
MOFs and Biomacromolecules for Biomedical Applications
397
Biocomposite Particle Size Biocomposite-based DDS are appealing for different administration routes, including parenteral injection and inhalation [96]. The particle size and shape of biocomposites play a crucial role for blood circulation time, biodistribution, and cellular internalization [97, 98]. In the case of cellular internalization, different mechanisms, including phagocytosis, micropinocytosis, or caveolar-mediated endocytosis, are particle size-dependent [99, 100]. Small particles (80% at 60 C). A similar study was performed by Chen and coworkers [76] who tested the stability of polyclonal antibodies including human immunoglobulin G (IgG), polyclonal antibody (H-IgG), and goat anti BSA IgG (G-IgG) encapsulated within two different MOF matrices (ZIF-8 and ZIF-90). To evaluate the protection effect of the MOF matrix on G-IgG@ZIF-90 and G-IgG@ZIF-8 biocomposites, the samples were exposed to a series of environments that would typically lead to denaturation of proteins (i.e., high temperatures, organic solvents, and mechanical pressure). Subsequently, the bioactivity of the encapsulated and free G-IgG was assessed by enzyme-linked immunosorbent assay (ELISA) test. The results revealed that the free G-IgG antibody, stored at 75 C, lost its initial binding activity (< 10%) and presented severe aggregation (88%). By contrast, the G-IgG released from the MOF matrix retained its binding capability (>90%), and showed low aggregation (13–25%) after being exposed to 75 C for 20 min (Fig. 12.12d). These results highlight that MOF matrices can protect antibodies from thermal decomposition. In summary, the preparation of biomacromolecules@MOFs was found to be an effective strategy for biospecimen preservation due to the unprecedented protection
12
MOFs and Biomacromolecules for Biomedical Applications
405
properties and on-demand degradability of the MOF matrices. Hormones, enzymes, biomarkers, vaccines, and antibodies have all been encapsulated within ZIF-8 and protected from temperature, solvents, and mechanical stress. After their release from the ZIF matrix, their activity was found always superior to the free biomacromolecules exposed to identical conditions. This represents an emerging attractive technology alternative to cold-chain transportation and storage of biotherapeutics, potentially reducing their shipping costs and enhancing their use.
12.3.3 Protein-On-MOFs and Proteins@MOFs Biocomposites in Assays A biosensor is a self-contained integrated device capable of providing selective quantitative or semiquantitative analytical information [131]. The biosensor is constructed by placing a biological recognition element in direct spatial contact with a signal transducer, allowing it to convert a biological response mediated by enzymes, immunosystems, or cells into a quantified processable signal. The biological recognition unit acts as a chemical receptor that responds selectively to a target analyte, and this response is transformed by the transducer into an electrochemical, colorimetric, or optical signal [132–136]. Although different sensing, transduction, and integration methods are available, sensitivity and reproducibility remain the major challenges in current diagnostic technologies to facilitate early diagnoses and prompt treatments. In this sense, protein-based MOFs biocomposites are emerging materials for the design of new, highly sensitive, and cost-effective biosensors [137– 141]. In such systems, the protein acts as a biorecognition element and it can be either embedded (protein@MOFs) in or bioconjugated to MOFs (protein-on-MOFs) [138]. The use of MOF composites as detection probes permits the colocalization of the biorecognition element and a large number of signaling elements in one single particle, thereby improving considerably the detection threshold of the system. So far, MOF biocomposites have been extensively studied for sensing a wide variety of analytes ranging from small molecules (glucose, H2O2, phenol, etc.) – generally exploiting the catalytic activity of supported enzymes – to large biomolecules such as antigens, biomarker, infectious agents and exosomes – generally exploiting the targeting capabilities of supported antibodies [137–141].
12.3.3.1
Applications of Protein@MOF Biocomposites for Small Molecule Detection
In biochemistry, an analyte with molecular weight below 1000 Da is classified as small molecule. On this basis, most of the reports about the use of protein@MOF composites for small molecules sensing are focused on the detection of H2O2 and
406
F. Carraro et al.
Fig. 12.13 (a) Schematic representation of the synthesis of Cyt c@ZIF-8 biocomposites and TEM image of the Cyt c@ZIF-8 composite. (b) Fluorometric detection of H2O2 using the enzymatic activity of Cyt c. The graph shows the relative peroxidase activity of Cyt c, Cyt c@ZIF-8 composite, PVP/Cyt c mixture, Cyt c/zinc ion mixture, Cyt c/2-methylimidazole mixture, and Cyt c/ZIF8 mixture. (Adapted with permission from Ref. [47], Copyright 2014 American Chemical Society)
glucose, and are based on the catalytic activity of encapsulated enzymes. These proof of concepts are discussed below.
Protein@MOF as H2O2 Sensors In biology, hydrogen peroxide (H2O2) is an important reactive oxygen species obtained as by-product of numerous metabolic reactions. Although H2O2 plays an important role in the transmission of cellular signals, H2O2 can decompose to hydroxyl radicals, which are strong oxidants capable of reacting with biological molecules and causing damage to cells and tissues. Therefore, it is important to develop new biosensing technologies for the detection of H2O2 in living organisms (e.g., determination of absolute rates of H2O2 production and steady-state concentrations in cells) [142]. A pioneering report by Ge and Liu et al. [47] in 2014 suggested the use of Cyt c encapsulated within ZIF-8 as fluorometric sensor to detect H2O2, methyl ethyl ketone peroxide (MEKP), and tert-butyl hydroperoxide (TBHP) in solution. The authors used N-acetyl-3,7-dihydroxyphenoxazine (Amplex Red, fluorogenic probe) as a signal molecule, since in the presence of the target peroxides Cyt c catalyzes the oxidation of Amplex Red to yield a fluorescent phenoxazine (i.e., resorufin) (Fig. 12.13) [143]. This work inspired the development of other protein@MOF biosensors for the detection of H2O2 [144, 145]. For instance, Yang et al. [145] designed a colorimetric biosensor encapsulating hemoglobin (BHb) in ZIF-8 particles; while H2O2 was detected by using 4-aminoantipyrine (AAP) as signal molecules, the peroxidase-like activity of BHb@ZIF-8 was used to perform the catalytic co-oxidation of phenol and AAP in the presence of H2O2 [146]. The catalytic activity of this system was 423% higher than that observed in the free BHb. Additionally, the BHb@ZIF-8 sensor showed a faster catalytic response (4 min) than the free enzyme (15 min), and a wide linear range (0–800 μM) for H2O2 with a limit of detection (LOD) of 1 μM.
12
MOFs and Biomacromolecules for Biomedical Applications
407
Protein@MOF as Glucose Sensors The relevance of glucose detections relies on its relationship to diabetes. This disease results in abnormal levels of insulin in the body, due to either a malfunction of the pancreas (diabetes type 1) or the ineffective use of insulin by cells (diabetes type 2). Insulin is the hormone that regulates the level of glucose in the blood, and thus, its deficiency in diabetic patients can cause hypoglycemic or hyperglycemic conditions, leading to severe health issues including tissue damage, kidney failure, and blindness, among others [147]. As a consequence, regular glucose monitoring in diabetic patients can prevent further health complications. The use of MOF-based biocomposites for the enzymatic detection of glucose has been extensively explored mostly as colorimetric or electrochemical sensors [141]. Liu and coworkers reported the first example of a colorimetric glucose biosensor based on the coencapsulation of multiple enzymes (GOx and HRP) in ZIF-8 particles [148]. This multi-enzyme system (GOx&HRP@ZIF-8) operates via a biocatalytic cascade process: (1) GOx in the presence of O2 catalyzes the oxidation of glucose to yield gluconic acid and H2O2; (2) HRP consumes H2O2 for the oxidation of ABTS (2,20 -azino-bis (3-ethylbenzothiazoline-6-sulphonic acid) into ABTS•+. The latter is a chromogenic agent that can be monitored by UV-vis spectroscopy at 420 nm. The reported limit of detection (LOD) was 0.5 μM, demonstrating a sensitivity higher than the most common colorimetric glucose sensors. Additionally, irrespective of interfering compounds (e.g., like fructose, maltose), GOx&HRP@ZIF-8 showed specificity toward glucose detection. It is worth mentioning that the two enzymes are randomly distributed through the MOF particle, but their close spatial location in the porous microenvironment facilitates molecular diffusion and enhances the efficiency of the enzymatic cascade reaction. Recent studies further supported the importance of the spatial distribution of the enzymes within the MOF composites for enhanced multi-enzyme cascade catalysis [149]. For instance, Jiang et al. [150] demonstrated that the compartmentalization of GOx/HRP multicatalytic system within ZIF-8 is an effective strategy to improve the sensitivity and increase the linear range of colorimetric biosensors for glucose detection. The compartmentalization of the enzymes was achieved by mixing sodium deoxycholate (NaDC), HRP, and the Zn2+ precursor. This strategy permits the embedding of HRP in a hydrogel coating. Then, a second solution containing both HmIM and GOx was added to this mixture. The authors suggested that the hydrogel allowed for the spatial separation between enzymes and served as a soft template to form hollow ZIF-8 spheres. Thus, the HRP is located within the central cavity of the hollow MOF capsules, while the GOx is supported onto the outer region of the particle. The spatially controlled localization of enzymes promotes the efficient diffusion of products from GOx to HRP pulling the equilibrium toward the product formation. The previous systems were all based on ZIF-8; however, recent reports have demonstrated that the enzymatic detection of glucose can be prepared by immobilization of a biorecognition element inside different MOFs (e.g., MAF-2 [151]) and on the outer surface of the MOF. The bioconjugation strategy allows for the use of a
408
F. Carraro et al.
Fig. 12.14 (a) Schematic representation of the synthesis of a GOx-on-Fe-MIL-88B-NH2 biocomposite. (b) UV-vis spectra of the enzymatic-cascade reaction at different concentrations of glucose (1–500 uM). (c) Selectivity studies of the glucose biosensor in the presence of other carbohydrates. (Reprinted with permission from Ref. [153], Copyright 2019 American Chemical Society)
variety of preformed MOF materials with peroxidase-mimicking activity [152]. This strategy could exploit the enzymatic-like activity of the MOF to reduce issues related to the production of intermediates during multienzyme cascade reactions. This hypothesis was probed by Zhu et al. [153] who reported the fabrication of colorimetric glucose biosensor based on grafting (covalent immobilization) of GOx onto Fe-MIL-88B-NH2, an MOF that showed a peroxidase-like activity (Fig. 12.14a). In this catalytic process, first GOx catalyzes the glucose oxidation to yield gluconic acid and H2O2, then Fe-MIL-88B-NH2 consumes H2O2 to produce •OH, which oxidizes the chromogenic substrate 3,30 ,5,50 - tetramethylbenzidine (TMB), into a green-blue colored ox-TMB intermediate (λmax 652 nm) (Fig. 12.14b). Accordingly, when GOx-on-Fe-MIL-88B-NH2 was used as a glucose biosensor, it presented high selectivity and displayed a linear response range of 1–500 μM, with an LOD of 0.478 μM (Fig. 12.14c). Furthermore, GOx-on-Fe-MIL-88B-NH2 showed higher tolerance to temperature and pH changes in comparison with the free enzyme system, and its reusability was tested up to five cycles. Based on these results, the authors tested this material for the detection of glucose in human serum. The GOx-on-Fe-MIL-88BNH2 results were in good agreement with the results obtained with a commercial glucometer (