Membrane modification : technology and applications 9781439866351, 143986635X

Citation preview

Membrane Modiication Technology and Applications

Membrane Modiication Technology and Applications

Edited by

Nidal Hilal Mohamed Khayet Chris J. Wright

Boca Raton London New York

CRC Press is an imprint of the Taylor & Francis Group, an informa business

CRC Press Taylor & Francis Group 6000 Broken Sound Parkway NW, Suite 300 Boca Raton, FL 33487-2742 © 2012 by Taylor & Francis Group, LLC CRC Press is an imprint of Taylor & Francis Group, an Informa business No claim to original U.S. Government works Version Date: 20120409 International Standard Book Number-13: 978-1-4398-6636-8 (eBook - PDF) This book contains information obtained from authentic and highly regarded sources. Reasonable efforts have been made to publish reliable data and information, but the author and publisher cannot assume responsibility for the validity of all materials or the consequences of their use. The authors and publishers have attempted to trace the copyright holders of all material reproduced in this publication and apologize to copyright holders if permission to publish in this form has not been obtained. If any copyright material has not been acknowledged please write and let us know so we may rectify in any future reprint. Except as permitted under U.S. Copyright Law, no part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any information storage or retrieval system, without written permission from the publishers. For permission to photocopy or use material electronically from this work, please access www.copyright. com (http://www.copyright.com/) or contact the Copyright Clearance Center, Inc. (CCC), 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400. CCC is a not-for-profit organization that provides licenses and registration for a variety of users. For organizations that have been granted a photocopy license by the CCC, a separate system of payment has been arranged. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Visit the Taylor & Francis Web site at http://www.taylorandfrancis.com and the CRC Press Web site at http://www.crcpress.com

Contents Preface......................................................................................................................vii Editors .................................................................................................................... xiii Contributors ...........................................................................................................xvii Chapter 1

Membranes in Nuclear Science and Technology: Membrane Modiication as a Tool for Performance Improvement ........................ 1 Grazyna Zakrzewska-Trznadel and Mohamed Khayet

Chapter 2

Use of Impedance Spectroscopy for Characterization of Modiied Membranes ..................................................................... 21 Juana Benavente

Chapter 3

Reduction of Membrane Fouling by Polymer Surface Modiication ................................................................................... 41 Victor Kochkodan

Chapter 4

Modiications of Polymeric Membranes by Incorporating Metal/Metal Oxide Nanoparticles...................................................... 77 Law Yong Ng, Abdul Wahab Mohammad, and Nidal Hilal

Chapter 5

Development of Antifouling Properties and Performance of Nanoiltration Membranes by Interfacial Polymerization and Photografting Techniques ................................................................ 119 Mazrul N. Abu Seman, Mohamed Khayet, and Nidal Hilal

Chapter 6

Integrating Hydrophobic Surface-Modifying Macromolecules into Hydrophilic Polymers to Produce Membranes for Membrane Distillation ..................................................................... 159 Mohammed Qtaishat, Mohamed Khayet, and Takeshi Matsuura

Chapter 7

Plasma Modiication of Polymer Membranes .................................. 179 Marek Bryjak and Irena Gancarz

v

vi

Chapter 8

Contents

Surface Modiication of Electrospun Nanoiber and Nanoibrous Membranes .................................................................. 215 Carmen García-Payo and Mohamed Khayet

Chapter 9

Development of Membranes for Pervaporation by Membrane Surface Modiication and Incorporation of Inorganic Particles ...... 259 Kailash Chandra Khulbe, Chaoyang Y. Feng, and Takeshi Matsuura

Chapter 10 Tailor-Made Polymeric Membranes for Advanced Crystallization of Biomolecules ....................................................... 333 Efrem Curcio, Enrica Fontananova, and Gianluca Di Proio Chapter 11 Chemical Cross-Linking Modiications of Polymeric Membranes for Gas Separation Applications....................................................... 363 Ahmad Fauzi Ismail and Farhana Aziz Chapter 12 Development of Fuel Cell Polymer Electrolyte Membranes by Radiation-Induced Grafting with Electron-Beam Irradiation ......385 Mohamed Mahmoud Nasef Chapter 13 Modiication of Sulfonated Poly(Ether Ether Ketone) for DMFC Application ...........................................................................409 Ahmad Fauzi Ismail, Muhammad Noorul Anam Mohd Norddin, Juhana Jaafar, and Takeshi Matsuura Chapter 14 Nanoiltration Membrane in Textile Efluent Treatment ..................449 Ahmad Fauzi Ismail and Woei Jye Lau Chapter 15 Future Prospects ............................................................................... 477 Nidal Hilal, Mohamed Khayet, and Chris J. Wright

Preface At the turn of the twentieth century, many chemists, physicists, and biologists were studying the barrier properties of membranes. However, it was not until 20 years later when the irst laboratory-scale membranes were manufactured and 50 years later when the most signiicant industrial breakthrough of asymmetric membranes led to greater application of membrane processes. Over the last 100 years, membrane processes have been established as an essential addition to conventional separation processes such as distillation and chemical extraction. Membrane processes have many advantages over other separation techniques that have led to their application in a wide range of industries including water treatment, biotechnology, food, pharmaceutical, and energy generation. The list continues to grow. The advantages membrane processes offer include highly selective separation, relatively low operating costs with low energy usage, and modular design with continuous and automatic operation. As more industrial applications exploit these advantages, within the second decade of the twenty-irst century, there has been a resurgence of activity looking to control the barrier properties of membranes, to improve selectivity, to reduce energy consumption, and to optimize membrane applications by means of membrane surface modiication. Hence, the rationale for this book, which presents an extremely timely review of membrane modiication, will hopefully act as an information resource and inspiration to scientists and engineers charged with improving membrane separation processes. To that end, we have split the book into two sections: irst, introducing membrane modiication techniques, followed by research areas focusing on the application of modiied membranes. The underlying process of membrane separation is a selective separation achieved due to differences in the physical and chemical properties between a membrane, a solvent, and solute(s). A driving force acting on the feed solution transports the solvent or the solute, or both, to a membrane surface. The solvent can pass through the membrane, and the solute can either be retained in the feed side of the membrane or pass through the membrane depending on its size, activity, partial pressure, or charge. The simplest example of membrane separation is arguably microiltration (MF), where suspended particles in the size range 0.05–10 μm, such as bacteria and dust particles, are separated in a liquid–liquid pressure-driven process. The separation mechanism of MF membranes is primarily controlled by steric rejection and by the sieving process of the membrane pores as the solvent passes through the membrane. Ultrailtration (UF) membranes have pores in the size range 5–100 nm and are used to separate colloids and macromolecules with molecular weight in the range 104 –106 Da. The separation mechanism of UF membranes is mainly steric; however, there is a growing acceptance that the charge differences between the solutes and the membrane surface also play a signiicant role. Nanoiltration (NF) and reverse osmosis (RO) are used to remove material existing as very small moieties, including synthetic dyes, agrochemicals, and aqueous salts. RO membranes ideally allow only water molecules to pass through; therefore, it is widely used on a large scale in the vii

viii

Preface

desalination industry. The separation mechanism for NF is governed by the steric and charge interactions between the solute/solvent and the membrane. In all types of membrane operations, it is the chemical and physical properties of the membrane surface that control the eficiency and speciicity of the separation process. Any new application of a membrane process must focus on the membrane surface, its chemical and physical characteristics, and its interaction with the components of the feed stream. This will inevitably lead to the following questions. Can the membrane surface be tailored to meet the speciic needs for a given new application? Can it be modiied to optimize a given separation process? An inherent problem associated with all membrane separation processes is that of membrane fouling. During normal process operation, some feed species “foulants” are deposited on the membrane surface, changing its chemical and physical properties. This foulant material may be removed partially by standard operating procedures such as inverted low operation and cleaning. Indeed, careful choice of the membrane material in terms of its chemical and physical interactions with the foulant(s) may reduce the buildup of foulant layers. However, if the foulant material builds up, then the fouling layers may alter the surface that interacts with the feed stream, modifying the surface roughness, hydrophobicity, charge and pore size. Membrane fouling can be reversible or irreversible, depending on the nature of the surface interaction. Both types can place economic constraints on the adoption of membrane separation processes. Thus, membrane fouling should be minimized, and again this can be achieved by modiication of the membrane, with the knowledge of the foulant materials guiding the choice of base polymer membrane and the modiication process. It is worth mentioning in this preface that there are two main classes of membranes, namely, organic and inorganic. Inorganic membranes are relatively new in industrial application, and typical materials for these membranes are ceramics and zeolites, including alumina, titanium, and zirconium. Inorganic membranes have good tensile strength, have greater resistance to chemical attack, and are more applicable under extremes of temperature and pressure compared with organic membranes. However, the main drawback with these membranes is that they are extremely expensive. On the other hand, organic membranes manufactured from polymeric materials are commonly used because of their excellent bulk physical and chemical properties. They are less robust than inorganic membranes but are inexpensive and easier to construct. However, the application of polymeric membranes is currently limited due to their surface properties such as hydrophobicity and lack of functional groups. To date, a number of polymers have been used in membrane fabrication, including cellulose acetate, polysulfone, polyester, and polyamide. Unfortunately, there are a limited number of suitable polymeric materials commercially available. In addition, there has been no large-scale production of brand-new polymers within the last decade and nor is there any expected in the near future. Thus, membrane modiication is one of the alternatives to increase the number and variety of membranes prepared for new and improved applications. There are two strategies for the development of novel membranes: one is to modify a bulk polymer system and subsequently prepare the membrane, and the second is to prepare the membrane from a standard polymer and then modify its surface. This last procedure can also be applied to inorganic membranes.

Preface

ix

In this book, we have assembled the contributions from experts in the various ields that exploit membranes and have developed modiication techniques to improve the eficiency of membrane operation. The book starts with an examination of the use of membrane modiication to optimize the performance of membranes used in the challenging industry of nuclear power with its process streams consisting of radioactive wastes and acids that can destroy membrane functionality. Chapter 1 details the use of both polymeric and inorganic membranes in radioactive wastes processing. Membranes have been used for the separation of isotopes from gases and for liquid radioactive waste treatment. Membrane distillation (MD) is particularly useful for the concentration of low-level radioactive wastes in a small volume, appropriate for fossilization and production of clean water streams for discharge. Liquid membranes are also very attractive for nuclear technology because of their good separation ability and selectivity toward speciic radionuclides. When considering membrane modiication, Chapter 1 discusses the combination of different organic and inorganic materials, to obtain hybrid membranes of desirable structure and properties applicable to nuclear processing. To improve the functionality of membranes within the nuclear industry, different chemical and physical methods can be used to change the membrane surface in order to meet the challenges of the extreme environments; these include oxidation, chemical functionalization, plasma treatment, and radiation-induced grafting. The chapter concludes with case studies describing key membrane modiication strategies that have beneited nuclear processing. Chapter 2 comprehensively describes the use of impedance spectroscopy (IS) in the characterization of commercial and novel membranes. This useful technique measures impedance plots in relevant electrolyte environments to determine electrical charge parameters such as resistance and capacitance of a membrane. From these results, electrical/geometrical parameters for membranes or individual layers, such as conductivity, porosity, thickness, and dielectric constant, can be obtained. Thus, the inluence of modiication processes can be monitored and used to guide further strategies for improving the membrane separation processes. We begin Chapter 3 with a discussion on the factors effecting membrane fouling. The chapter reviews recent studies in which the reduction of irreversible organic fouling and biofouling is attempted by the modiication of the membrane surface. The chapter introduces the modiication techniques of ultraviolet (UV)- and redoxinitiated surface grafting of hydrophilic polymers, low temperature plasma treatment, physical coating/adsorption of a thin hydrophilic layer on the membrane surface, chemical reactions on the membrane surface, and surface modiication of polymer membranes with nanoparticles. Chapter 4 then expands the discussion on the use of nanoparticles in membrane modiication processes. Materials in the form of nanoparticles have a large surface area to volume ratio, which infers many interesting properties on nanoparticulate systems due to the involved interfacial properties. As a consequence, nanoparticles are currently receiving a lot of interest in many industries, such as membrane technology where the control of interfacial interactions is important. Nanoparticles affect the permeability, selectivity, hydrophilicity, thermal and electrical conductivities, mechanical strength, thermal stability, and the antiviral and antibacterial properties of the polymeric membranes. Chapter 4 discusses important examples of

x

Preface

nanoparticles that have been incorporated into various types of polymeric membranes. The nanoparticles that are discussed include those based on silver, iron, zirconium, silica, aluminum, titanium, and magnesium, which have been applied to improve different types of membrane processes. A common problem in water treatment using membrane processes is fouling with natural organic matter (NOM). This is the topic of Chapter 5 that discusses the modiication of membranes to control fouling with NOM. The chapter describes how NF polyester thin-ilm composite membranes prepared by interfacial polymerization yield membranes with less fouling tendency for NOM. Another technique that can be used to improve the resistance of fouling by NOM is UV photografting. Chapter 5 discusses the results demonstrating that a membrane grafted with N-vinyl2-pyrrolidinone (NVP) has superior properties for irreversible fouling. Discussions and analysis are based on atomic force microscopy (AFM), illustrating the importance of AFM to membrane modiication characterization. AFM with its highresolution imaging capability combined with force measurement capability is an extremely useful tool for membrane science. MD is an emerging thermally driven membrane separation process that has been applied extensively to desalination, food processing, and removal of volatile organic compounds (VOCs) from water. It has several advantages compared with other separation processes, which include higher rejections of nonvolatile solutes, lower operating temperatures than conventional distillation, and lower operating pressures than conventional pressure-driven membrane separation processes. Chapter 6 describes how a signiicantly higher MD permeate lux could be achieved by using novel membranes fabricated by blending luorinated surface-modifying macromolecules (SMMs) into hydrophilic host polymers. During membrane formation by the phase inversion method, the SMMs migrate to the top polymer–air interface, changing the inal characteristics of the prepared membrane. Chapter 6 also demonstrates how a comprehensive characterization strategy beneits membrane modiication processes. The fabricated membranes were characterized by different techniques such as x-ray photoelectron spectroscopy (XPS), scanning electron microscopy (SEM), contact angle measurement, gas permeate test, and liquid entry pressure (LEP) of water. Chapter 7 examines the use of plasma treatment for the modiication of polymeric membranes. Plasma treatment is carried out at the membrane surface so that the beneicial properties of the bulk material remain unchanged. Surface properties such as roughness and functionality can be altered to improve the performance of the membrane. All these processes are very quick and the time taken for modiication is usually a few seconds up to a few minutes. The method uses chemicals in the gaseous form and produces very small amounts of wastes. Among all techniques of membrane surface modiication, plasma treatment seems to be the most versatile and environment-friendly. The authors of Chapter 7 discuss how these beneits impact on membrane modiication strategies. There is currently a renewed interest in the use of electrospinning techniques for the fabrication of membranes. Chapter 8 reviews the use of this versatile technique for the production of nanoiber webs or membranes. The chemical and physical properties of nanoiber membrane surfaces play an important role in their application to iltration, biomedical materials, tissue engineering scaffolds, drug delivery

Preface

xi

carriers, for reinforcement in composite materials and electronics, antibacterial materials, and within the food processing industry as membrane bioreactors. Chapter 8 gives an overview of various surface modiication techniques applied to enhance the bulk and surface properties of nanoiber membranes, rendering them suitable for speciic applications. Over the last 50 years, pervaporation (PV) has been developed as a technology with particular use in dehydration of solvents, alcohol/water separation, VOC removal from water, and separation of organic mixtures. Chapter 9 reviews the improvement of PV performance using membrane modiication with a discussion on different strategies for surface modiication of synthetic polymeric membranes and composite polymer/inorganic membranes. The focus of this book then changes to be more application-based and demonstrates how the modiied membranes have beneited different ields. An examination and a review of membrane modiication to improve membrane crystallization processes are reported in Chapter 10. In this innovative process, the crystallization of biomolecules is enhanced to produce crystals with controlled structure and polymorphism. The chapter describes the theoretical and experimental correlations between the physicochemical properties of membranes, kinetics of nucleation, and characteristics of the inal product. Membrane modiication techniques using copolymers or additives are discussed in the context of controlling nanoscale physical and chemical characters of membranes for improved crystallization. One of the most recent developments in the ield of membrane technology has been the successful commercialization of polymer membrane processes for gas separation/puriication. Chapter 11 argues that membrane modiication can meet the remaining challenges for this technique, which include achieving higher selectivity for the relevant application while maintaining equivalent productivity and preserving the acceptable membrane performance, thermal stability, and chemical resistance in the presence of aggressive feeds. The most widely used modiication strategies for the improvement of gas separation membranes are thermal treatment, polymer blending, and cross-linking. Cross-linking is achieved by a number of methods, including thermal, ion beam, and UV-irradiation treatment, or by chemical cross-linking. Another emerging area where membrane modiication is proving to be invaluable is that of polymer electrolyte membranes (PEMs) used for the construction of fuel cells, batteries, electrolyzers, sensors, and actuators. Chapter 12 discusses how membrane modiication can meet the strong demand for the design of PEMs with particular properties. PEMs can be fabricated using electron beams obtained from accelerators, which have the advantage to be applied using simultaneous and preirradiation techniques. Various types of high-energy radiation, such as γ-rays and electron beams, can be used in radiation-induced grafting, irrespective of the morphology of the starting polymer. Chapter 12 reviews the latest progress in using electron beam for the preparation of PEMs for proton exchange membrane fuel cell (PEMFC) and direct methanol fuel cell (DMFC) applications using novel radiation-induced grafting techniques. Chapter 13 continues the discussion on membrane modiication for DMFC improvements, with a special focus on polymeric membranes based on sulfonated poly (ether ether ketone) (SPEEK). The modiication strategies discussed for SPEEK include clay nanocomposite fabrication with

xii

Preface

compatibilizer, composite SPEEK and boron orthophosphate membrane, tungstosilicic acid supported on silica–aluminum oxide, and SPEEK membrane blended with charged SMMs. Chapter 14 covers the important area of the textile industry and how membrane modiication can improve the separation processes in this industrial sector, which is recognized as one of the worst global polluters of water sources. The ability of NF to separate small and neutrally charged molecules from textile wastewater streams has gained a lot of attention in recent years. Chapter 14 provides an overview with case studies on the latest development of NF fabrication technology and puts this in the context of problems associated with the use of other treatment processes in the textile industry, thus demonstrating the advantages of NF in removing dye components and dissolved salts. Membrane fouling within the textile industry is also reviewed with emphasis on unraveling the mechanisms involved in guiding techniques such as membrane modiication aimed at the reduction of fouling. Among the available practical solutions, the development of fouling-resistant NF membranes is the most sustainable one and is currently attracting considerable attention. Membrane modiication strategies include changing the chemical functionality or the physical structure of the membrane. Hydrophilic groups and/or charged functional groups can be immobilized onto membrane surfaces using coating, blending, or grafting techniques to reduce fouling during the processing of the highly variable process streams of the textile industry. In the inal chapter, we have pooled the thoughts of the contributors to suggest some of the ways in which membrane modiication may contribute to the development and application of membranes in the future. We thank all the authors who have contributed to the writing of this book. We very much appreciate their willingness to devote their valuable time and efforts to this task and for the delivery of the high-quality chapters in time. We believe that membrane scientists and technologists from different disciplines and research ields will ind this book stimulating and essential for their research and academic projects. Nidal Hilal, Mohamed Khayet, and Chris J. Wright United Kingdom and Spain

Editors Nidal Hilal holds a chair in nanomembranology and water technologies at Swansea University. He is also the founding director of the Centre for Water Advanced Technologies and Environmental Research (CWATER) at Swansea University in the United Kingdom. Currently, he is establishing a Center of Excellence in Membranes and Desalination at Masdar Institute of Science and Technology in Abu Dhabi. This institution is being founded in collaboration with MIT. He obtained a PhD in chemical engineering from the University of Wales in 1988. His research interests lie broadly in the identiication of innovative and cost-effective solutions to real-world process engineering problems within the ields of chemical engineering, nano-water, membrane technology, water treatment including desalination, colloid engineering, and nano-engineering applications of atomic force microscopy (AFM). He is now internationally recognized as a world leader in developing and applying the force measurement capability of AFM to the study of membrane separation and engineering processes at the nanoscale level. He has pioneered the application of AFM to chemical and process engineering problems and recently published a book in this area. His research has led to the use of AFM in the development of new membranes with optimized properties for dificult separations. He has published 5 books, 21 invited book chapters, and around 300 articles in the refereed scientiic literature. He has chaired international conferences and delivered numerous invited lectures around the world. In recognition of an outstanding research contribution, he was awarded senior doctorate (Doctor of Science) by the University of Wales in 2005. He was also awarded the Kuwait Prize for Water Resources Development in the Arab World for the year 2005. This prize is regarded as the “Arab Nobel Prize” and is one of the highest scientiic honors for intellectual achievement to be awarded in the Middle East. Professor Hilal is the editor-in-chief for the international journal Desalination, is on the editorial boards of a number of international journals, and is a member of the advisory boards of several multinational organizations. He is a registered European Engineer, a Chartered Engineer in the United Kingdom, and a fellow of the Institution of Chemical Engineers. Hilal has carried out extensive consultancy for industry, government departments, research councils, and universities on an international basis.

xiii

xiv

Editors

Mohamed Khayet was born in Marrakech, Morocco, in 1966, and graduated from the Faculty of Sciences, University Cadi Ayyad of Marrakech (Morocco), in 1990. Funded by the Spanish Agency of International Cooperation (AECI), he pursued his doctoral studies at the Faculty of Physical Sciences, Department of Atomic Molecular and Nuclear Physics, University Complutense of Madrid (UCM, Spain), and received his PhD degree from UCM in 1997. He joined the Department of Applied Physics I (UCM) in 1997, serving as an assistant and then as an associate professor. Funded by UCM, he undertook a postdoctoral visit at the Industrial Membrane Research Institute (IMRI) in Ottawa (Canada) during the period 2000–2001. Currently, he is the professor of thermodynamics, statistical physics, and renewable energy applications in the Department of Applied Physics I (UCM) and is director of the UCM research group Membranes and Renewable Energy. He has published over 100 scientiic papers in international refereed journals and has written several invited book chapters and 3 books. He has given over 70 presentations at scientiic conferences, workshops, and congresses. He is author of two patents in the ield of membrane science and technology and has supervised national and international projects as well as research and academic studies. He is referee of a number of international journals and a member of the editorial boards of the following international journals: Desalination, Applied Membrane Science and Technology, Membrane Water Treatment, Polymers and Membranes. His current research interests include future desalination technologies, renewable energy, membrane fabrication characterization, membrane processes, modeling, and transport phenomena. Chris J. Wright is a reader in bionanotechnology and membrane separation within the Multidisciplinary Nanotechnology Centre (MNC) at Swansea University. At Swansea, he is an executive member of the Centre for NanoHealth and is an associate director of the Centre for Complex Fluids Processing. He graduated from the University of Wales in 1996 with a PhD in biochemical engineering. In 2001, he was awarded a prestigious Advanced Research Fellowship by the Engineering and Physical Research Council (EPSRC), United Kingdom, in recognition of his innovative research applying AFM to the characterization of membrane and biological surfaces. This 5-year award allowed him to establish an internationally recognized research group exploiting the capabilities of AFM. His innovative research developing AFM force measurement capabilities to study biological interfaces has

Editors

xv

been adopted by many other researchers and industry. In 2006, he was appointed Portfolio Director for Process Engineering within the College of Engineering at Swansea University and is now director of PhD studies within the MNC. His research interests include the control of polymer surfaces for improved membrane separation and tissue engineering and the control of bioilms and the combination of AFM with advanced light microscopy methods. An underlying theme of this research is the application of nanotechnology to health care. His research has been sustained through major grants from government, charity, and industry. He is on the editorial board of the Journal of Nanoengineering and Nanosystems and is a member of the EPSRC College for the assessment of research grants. He has over 80 peer-reviewed international publications with 15 invited book chapters and review articles.

Contributors Farhana Aziz Faculty of Petroleum and Renewable Energy Engineering Universiti Teknologi Malaysia Johor, Malaysia

Irena Gancarz Department of Polymers and Carbon Materials Wroclaw University of Technology Wroclaw, Poland

Juana Benavente Grupo de Caracterización Electrocinética y de Transporte en Membranas e Interfases Departamento de Física Aplicada I Facultad de Ciencias Universidad de Málaga Málaga, Spain

Carmen García-Payo Department of Applied Physics I Faculty of Physics University Complutense of Madrid Madrid, Spain

Marek Bryjak Department of Polymers and Carbon Materials Wroclaw University of Technology Wroclaw, Poland Efrem Curcio Department of Chemical Engineering and Materials Institute on Membrane Technology University of Calabria Cosenza, Italy Chaoyang Y. Feng Industrial Membrane Research Lab Chemical and Biological Engineering Department University of Ottawa Ottawa, Canada Enrica Fontananova Institute on Membrane Technology University of Calabria Cosenza, Italy

Nidal Hilal Centre for Water Advanced Technologies and Environmental Research Multidisciplinary Nanotechnology Centre College of Engineering Swansea University Swansea, United Kingdom Ahmad Fauzi Ismail Advanced Membrane Technology Research Centre Universiti Teknologi Malaysia Johor, Malaysia Juhana Jaafar Faculty of Petroleum and Renewable Energy Engineering Universiti Teknologi Malaysia Johor, Malaysia Mohamed Khayet Department of Applied Physics I Faculty of Physics University Complutense of Madrid Madrid, Spain

xvii

xviii

Kailash Chandra Khulbe Industrial Membrane Research Lab Chemical and Biological Engineering Department University of Ottawa Ottawa, Canada Victor Kochkodan Institute of Colloid and Water Chemistry National Academy of Science of Ukraine Kyiv, Ukraine Woei Jye Lau Faculty of Petroleum and Renewable Energy Engineering Universiti Teknologi Malaysia Johor, Malaysia Takeshi Matsuura Industrial Membrane Research Laboratory Department of Chemical Engineering University of Ottawa Ottawa, Canada Abdul Wahab Mohammad Department of Chemical and Process Engineering Faculty of Engineering and Built Environment Universiti Kebangsaan Malaysia Bangi, Malaysia Mohamed Mahmoud Nasef Institute of Hydrogen Economy International City Campus Universiti Teknologi Malaysia Kuala Lumpur, Malaysia

Contributors

Law Yong Ng Department of Chemical and Process Engineering Faculty of Engineering and Built Environment Universiti Kebangsaan Malaysia Bangi, Malaysia Mohammad Noorul Anam Norddin Faculty of Petroleum and Renewable Energy Engineering Universiti Teknologi Malaysia Johor, Malaysia Gianluca Di Proio Institute on Membrane Technology University of Calabria Cosenza, Italy Mohammed Qtaishat Chemical Engineering Department University of Jordan Amman, Jordan Mazrul Nizam Abu Seman Faculty of Chemical and Natural Resources Engineering Universiti Malaysia Pahang Kuantan, Pahang, Malaysia Chris J. Wright Centre for Water Advanced Technologies and Environmental Research Multidisciplinary Nanotechnology Centre College of Engineering Swansea University Swansea, United Kingdom Grazyna Zakrzewska-Trznadel Institute of Nuclear Chemistry and Technology Warsaw, Poland

1

Membranes in Nuclear Science and Technology: Membrane Modification as a Tool for Performance Improvement Grazyna Zakrzewska-Trznadel and Mohamed Khayet

CONTENTS 1.1

Membranes and Membrane Processes Applied in Nuclear Science and Technology ................................................................................................. 2 1.1.1 Membranes for the Separation of Isotopes ........................................... 2 1.1.2 Membranes for Liquid Radioactive Waste Treatment .......................... 3 1.1.2.1 Properties of Membranes Selected for Radioactive Waste Treatment ....................................................................3 1.1.2.2 Pressure-Driven Membrane Processes .................................. 5 1.1.2.3 Electric Membrane Processes ................................................ 5 1.1.2.4 Thermal Processes: MD ........................................................6 1.1.2.5 Liquid Membranes ................................................................. 6 1.1.3 Other Applications of Membranes in Nuclear Technology ..................7 1.1.4 Preparation of Membranes ...................................................................7 1.1.5 Modiication of Membranes .................................................................7 1.2 Removal of Radionuclides with Surface-Modiied Membranes; Laboratory Experiments: UF and MD ............................................................. 9 1.2.1 Case Study: Application of Surface-Modiied Membranes Based on Polysulfone and Polyethersulfone in UF/Complexation Process for the Removal of 60 Co from Water Aqueous Solutions ........9 1.2.2 Case Study: Surface-Modiied Membranes Applied for the Treatment of Radioactive Wastes by MD ........................................... 12 References ................................................................................................................ 15

1

2

1.1 1.1.1

Membrane Modification: Technology and Applications

MEMBRANES AND MEMBRANE PROCESSES APPLIED IN NUCLEAR SCIENCE AND TECHNOLOGY MEMBRANES FOR THE SEPARATION OF ISOTOPES

The idea of applying membranes in nuclear technology came from the work of Graham, who, in the middle of the nineteenth century, reported the possibility of separating gaseous mixtures by molecular diffusion. Later, the process was studied for the separation of isotopes of light elements such as neon, hydrogen, nitrogen, or carbon. During the 1940s and 1950s, the method was employed in the United States for the enrichment of 235U, the issionable natural uranium isotope. Since the concentration of 235U in natural uranium is very low (0.711 wt%), the separation of this isotope is only possible by the application of a multistage process and a separation cascade with a large number of separating units. The crucial element of the system for isotopic enrichment is a membrane or porous barrier, the characteristics of which inluence the stage separation factor and the size of the stage required to handle the desired stage lows. The preparation of such a material with ine pores of size range 6–40 nm, and a high bulk and surface porosity, presents a dificult technological problem. Membranes for gaseous diffusion can be homogenic or composite, consisting of a porous, permeable support and a thin membrane layer with a small pore size. The information on the uranium technology applied during the Manhattan Project and the materials used for preparing the membranes is still being classiied; however, it is known that the French technology uses sintered and anodic alumina, gold and silver alloys, nickel, Telon, porcelain, or zirconium (Burggraaf and Cot 1996; Hsieh 1996). Composite membranes can be manufactured, for example, by sintering metallic powder with a ceramic material or by coating a metallic, porous support with a Telon emulsion. The development of membranes for uranium enrichment has led to the still growing market of ceramic membranes in France (Hsieh 1996). The characterization techniques of barriers suitable for the separation of isotopes by gaseous diffusion include adsorption methods, x-ray analysis, electron microscopy, mercury intrusion methods, permeability measurements with liquids or gases, and separation eficiency measurements. Various methods are used for preparing the membranes for gaseous diffusion, including etching with strong acids such as nitric acid for gold and silver alloys, sintering, pressing, or extrusion of the ine powders of metals or metal oxides such as nickel or alumina, as well as anodic oxidation (alumina). The use of metal oxides is beneicial because of their resistance to corrosive hydroluoric acid, which is generated from uranium hexaluoride used in gaseous diffusion technology. Typically, the membranes in gaseous diffusion are operated at elevated temperatures of 60°C–200°C and under a pressure of up to 3.3 bars. It was proved that isotopes other than uranium, such as those of Ar, Ne, O, and H, could also be separated by gaseous diffusion (Fain and Brown 1974). The membranes developed for this purpose are made from alumina, gold, or glass (Konishi et al. 1983; Evans et al. 1983). Attempts were made to use polymeric membranes for separating the isotopes. The membranes from polyethylene terephthalate (PET),

Membranes in Nuclear Science and Technology

3

polyethylene (PE), polytetraluoroethylene (PTFE), cellulose acetate (CA), and polyvinylidene chloride (PVC) were tested for the separation of hydrogen isotopes in a gaseous phase (Marcea 1983). For the separation of water isotopomers, which include heavy water (HDO), heavy oxygen water (H218O) and tritiated water, the membranes prepared from regenerated cellulose, CA, and porous barriers from PTFE were studied. The Institute of Nuclear Chemistry and Technology (Warsaw, Poland) applied PTFE membranes for the enrichment of HDO and H218O in the process of membrane distillation (MD). The isotope separation effect of HDO/H2O was also studied with cellulose membranes (Chmielewski et al. 1991; Zakrzewska-Trznadel et al. 1996). Polyphosphazene membranes for the separation of the water isotopomers were tested in the Paciic Northwest Laboratory (Nelson et al. 1996; Duncan and Nelson 1999). In the Korean Atomic Research Institute (Daejeon), an MD cascade was constructed for testing the enrichment of heavy oxygen water (H218O) for medical applications in positron emission tomography (Kim et al. 2004). Using various membranes and membrane processes, the isotopes of chlorine (Campbell 1985), carbon (Fritz et al. 1987), and lithium (Whitworth et al. 1994) were separated. Nanoiltration (NF) membranes with complexing agents were applied for the separation of the gadolinium and neodymium isotopes (Chitry et al. 2001).

1.1.2

MEMBRANES FOR LIQUID RADIOACTIVE WASTE TREATMENT

1.1.2.1

Properties of Membranes Selected for Radioactive Waste Treatment Radioactive wastes are generated in the power generation cycle and during the dismantling of nuclear installations. The wastes from nuclear power reactors include miscellaneous waste, secondary waste, and chemical and detergent wastes. The radioactive wastes are generated at other stages of the nuclear cycle, such as in fuel fabrication plants, in spent fuel reprocessing installations, and during uranium enrichment and environmental remediation. Institutional activities such as radiopharmaceutical production and the application of radioisotopes in medicine, industry, and research also produce liquid radioactive wastes. The treatment of liquid radioactive wastes involves the concentration of radioactive species in a small volume and the conversion of this concentrate into an inert matrix, such as concrete, ceramics, or glass. Several processes are adopted for the treatment of the liquid waste, such as precipitation, evaporation, ion exchange and, in recent years, membrane methods. The eficiency of the treatment is measured by evaluating two characteristics: the decontamination factor (DF) (Equation 1.1) and the volume reduction factor (VRF) (Equation 1.2): DF =

Ao , Af

(1.1)

VRF =

Vo , Vf

(1.2)

4

Membrane Modification: Technology and Applications

where Ao is the speciic activity of the waste before treatment, Af is the speciic activity of the waste after treatment, Vo is the initial volume of the waste, and Vf is the inal volume of the waste. Membranes employed for radioactive waste treatment have to fulill very strict requirements, appropriate to the target applications. In addition to their separation ability, good permeability, and long lifetime, the membrane modules should have suficient resistance to the ionizing radiation and a geometry enabling easy cleaning. Both types of materials, organic and inorganic, are applied for the preparation of membranes for liquid radioactive waste treatment (Zakrzewska-Trznadel 2008; Vienna Technical Reports Series 2004). The beneits of using polymeric membranes are their relatively low price and their wide commercial availability. Polymer membranes can be easily manufactured and arranged in various conigurations and can be modiied for speciic applications. The advantages of the inorganic materials come from their high resistance to the aggressive chemical environment, high temperature, and ionizing radiation. Polymeric membranes can be formed as lat sheets, tubes, or hollow ibers. The foils are also arranged as spiral-wound or pleated ilters. The modular design of the membrane apparatus permits easy development of the iltration area and the scale-up of the process. The coniguration of the membrane is very important from an operational point of view because it determines the low hydrodynamics in the membrane system. Hydrodynamic conditions promoting turbulence and good mixing reduce the boundary layers and enhance the mass transport through the membrane (Zakrzewska-Trznadel et al. 2009; Cojocaru et al. 2009a,b). The appropriate design of the iltration segment facilitates cleaning of the membranes, which is a common procedure within the iltration cycles. Flat sheet membranes in plate and frame modules with narrow channels and hollow iber membranes are susceptible to fouling. However, the former are easier to clean. In spiral-wound modules, the spacer promotes turbulence, thereby reducing the concentration and temperature polarization as well as the fouling of the membrane. The low in the tubular membrane modules is easily made turbulent, which is also beneicial to their application. In nuclear applications, membrane fouling is a particular problem. Frequent cleaning of the membranes creates secondary wastes that are radioactive and require additional treatment. Thus, the elimination of the phenomena that lead to membrane fouling is a key issue in the design of membrane installations. One of the critical characteristics of membranes is their radiation stability. Membrane materials exposed to ionizing radiation may undergo structural changes, leading to an alteration of their permeation and separation characteristics, faster aging, or polymer degradation. A number of publications have reported studies on the effects of irradiation on the polymers applied for the preparation of the membranes (Ramachandhran and Misra 1985, 1986; Chmielewski and Harasimowicz 1992, 1997). They showed that for most polymers, the threshold values of the radiation doses that cause the structural changes of the membrane are suficiently high; therefore polymeric membranes can be applied for low- and medium-level radioactive waste treatment for reasonable periods of time.

Membranes in Nuclear Science and Technology

5

1.1.2.2 Pressure-Driven Membrane Processes Membrane processes offer one-stage separation for all the dissolved components of radioactive waste, from suspended matter to ionic species. During the last decade, membrane technology has been gradually introduced into the nuclear industry. The pressure-driven membrane processes like reverse osmosis (RO), ultrailtration (UF), or microiltration (MF) have found application in radioactive waste treatment. RO involves a semipermeable, dense membrane that is able to reject all dissolved lowmolecular-weight organics and multivalent and monovalent ions. The applications reported in the nuclear industry are for laundry nuclear wastes, mixed laboratory wastes, and institutional wastes (Sen Gupta et al. 1996; Zakrzewska-Trznadel and Harasimowicz 2002, 2004; Zakrzewska-Trznadel 2003; Arnal et al. 2000, 2003a,b; IAEA-TECDOC-911 1996). NF was employed for boric acid recovery and recycling, for lanthanides/actinides separation, and for the treatment of uranium mill efluents (Macnaughton et al. 2002; Hwang et al. 2002). Laundry wastes and uranium containing efluents from mill and mine operations can be treated by UF (Karlin et al. 2001; Kryvoruchko et al. 2004). UF in combination with complexation or inorganic sorbents is an effective method for processing different waste streams from nuclear applications (Hooper 1997; Smyth et al. 1998, 1999; Ramachandhran and Misra 1998). Polymer-enhanced ultrailtration (PEU) and micellar-enhanced ultrailtration (MEUF) were tested for the removal of metallic impurities from model solutions (Xiarchos et al. 2008). They were also proven as methods for the removal of radionuclides from liquid radioactive wastes (Zakrzewska-Trznadel and Harasimowicz 2002, 2004). MF is employed in remediation activities and for the dewatering of precipitate from treatment procedures (Mann and Todd 2000; Brown et al. 1991). It should be noted that RO and NF involve polymer membranes. Inorganic membranes, in addition to polymeric barriers, are in common use in UF and MF. The modiication of the membranes used in the pressure-driven processes aims not only to improve the separation characteristics, but also to increase the wear resistivity and the chemical and radiation stability. 1.1.2.3 Electric Membrane Processes Electric membrane processes, including electroosmosis, electrodialysis, and membrane electrolysis, are studied for different applications along the power generation cycle (Andalaft et al. 1997; Hobbs 1999; Hegazy et al. 1999). The novel anion exchange membrane was applied to separate 125I and 36Cl ions by electrodialysis (Inoue et al. 2004). The membrane exhibited high selectivity for iodine ions over chlorine ions, and the ratio of electroconductive membrane permeabilities of 125I and 36Cl was 6.2, while the diffusion membrane permeabilities of the two components were almost the same. In electric membrane processes, anion- and cation-exchange membranes are used. Sodium super ion conducting ceramic membranes—NaSICON—were studied for the separation of sodium from radioactive wastes produced by the US Department of Energy (Fountain et al. 2008). Naion and NaSICON membranes were tested for the electrochemical separation and recycling of sodium hydroxide from high salt content radioactive wastes (Hobbs 1999; Kurath et al. 1997).

6

Membrane Modification: Technology and Applications

1.1.2.4 Thermal Processes: MD MD is one of the emerging nonisothermal membrane separation processes, known for about 50 years but still requiring development for its industrial implementation. MD refers to the thermally driven transport of vapor through porous hydrophobic membranes, the driving force being the vapor pressure difference between the two sides of the membrane pores. Simultaneous heat and mass transfer occurs in this process, and different MD conigurations include direct contact MD, sweeping gas MD, vacuum MD, and air gap MD. MD is an effective process for desalination, the concentration of salts and acidic solutions, and distilled water production; it can also be applied to water and wastewater treatment (Hogan et al. 1991; Gryta 2002; Coufin et al. 1998; Khayet 2011). The use of this process for low-level radioactive waste treatment leads to a concentration of radioactive species in a small volume, appropriate for fossilization and the production of clean water streams and for discharge (Zakrzewska-Trznadel 1998; Zakrzewska-Trznadel et al. 1999). MD units can be employed in the front end of liquid radioactive waste processing to improve the economy of the treatment by an initial concentration before evaporation or in the back end to obtain better separation of the radionuclides and produce clean efluents. The membranes used for MD processes are hydrophobic, porous barriers made from PTFE polypropylene (PP) or polyvinylidene luoride (PVDF). All modiications lead to an increase in the membrane hydrophobicity in order to avoid membrane wettability, thereby making the membrane useful for longtime operation. 1.1.2.5 Liquid Membranes Liquid membranes are very attractive in nuclear technology because of their good separation ability and their selectivity toward speciic radionuclides. Many applications of such membranes in the radioactive waste processing ield are reported. They are applied to different waste streams for the recovery of selected components of the waste and the separation of ission products from spent fuel reprocessing waste (Kocherginsky et al. 2002; Happel Streng et al. 2003; El-Reefy et al. 1996; El-Said et al. 2002). All types of liquid membranes, including bulk, emulsion, supported, and inclusion liquid membranes, have been tested for nuclear applications: Facilitated transport using an appropriate carrier enhances the separation ability and the selectivity of the membrane. It can also improve the kinetics of the transport through the membrane. Chemical modiication of the carrier makes the liquid membrane more selective to speciic substances. The organic liquids used for liquid membranes are as follows: dichloromethane, chloroform, kerosene, chlorobenzene, n-octane, n-decane, n-dodecane, and n-tetradecane. The most popular carriers are crown ethers, calixarenes, phosphoroorganic compounds (e.g., D2EHPA and CYANEX 302), hydroxyoximes (e.g., LIX 65N and SME 529), and amines (e.g., ALAMINA 336 and tri-n-octylamine). In general, liquid membranes are used mostly to decrease the radiotoxicity of the liquid waste, but not to reduce its volume. In radioactive waste management, liquid membranes are irst applied to remove 90Sr and 137Cs from acidic or alkaline solutions (Kocherginsky et al. 2002; Happel Streng et al. 2003).

Membranes in Nuclear Science and Technology

7

The removal of actinides from reprocessing acidic waste solutions is advantageous in terms of minimizing the radioactive discharge to the natural environment. The separation of plutonium using supported liquid membranes was extensively studied, as well as U(VI) and Pu(IV) selective transport over ission products and minor actinide contaminants (Lakshmi et al. 2004; Sriram et al. 2000; Kedari et al. 1999).

1.1.3

OTHER APPLICATIONS OF MEMBRANES IN NUCLEAR TECHNOLOGY

Membrane separation techniques are being considered as feasible methods for cleaning the exhaust gases discharged from nuclear facilities such as nuclear power plants and spent fuel processing plants. The exhaust gases contain radioactive aerosols, ission products, and radioactive corrosion products, which have to be separated from the gas streams before discharge. For the removal of noble gases, especially 85Kr and 133+135Xe, from reactor atmospheres, silicone rubber, siloxane rubber membranes, and different polyarylane–siloxane block copolymers were tested (Stern and Leone 1980; Ohno et al. 1977; Stern and Wang 1980). In recent years, the research has focused on the application of hollow iber membranes from polyimide, oriented PP, or lat sheet membranes from PET (Sasaki et al. 2003; Nörenberg et al. 2001). The use of membrane techniques for removing the tritium generated in nuclear reactors or spent fuel processing plants from the gas efluents has been reported (Le Digabel et al. 2002; Ishida et al. 2000; Hirata et al. 1995; Hayashi et al. 1995; Bridesell and Willms 1998). Some aspects of this research focused on the assurance of tritium for future fusion reactors (Violante et al. 1995; Tosti et al. 2000; Konishi et al. 1998; Heinze et al. 2003).

1.1.4

PREPARATION OF MEMBRANES

Membrane manufacturing techniques depend on the material applied and on the required membrane characteristics. Based on the preparation materials used, the basic types of membranes include polymeric, inorganic, and hybrid polymeric–inorganic membranes. Both the polymeric and the inorganic membranes can be applied to radioactive wastes processing. Inorganic membranes are of greatest interest because of their good stability and their resistance toward different types of radiation: α, β, and γ. In most cases, commercial membranes that were proved in other industrial applications are employed after testing their stability under ionizing radiation. Taking into consideration the separation principles and the structure, three types of membranes can be distinguished: porous, dense, and liquid membranes. The most common techniques available for preparing the membranes are sintering, stretching, extrusion, track etching, phase inversion, and coating. For composite membranes, methods such as plasma polymerization, interfacial polymerization, and dip coating can be employed (Mulder 1991).

1.1.5

MODIFICATION OF MEMBRANES

Membranes manufactured in primary processes can be modiied for different applications by changing the material’s chemical properties or by changing the pore size.

8

Membrane Modification: Technology and Applications

The combination of different materials, organic and inorganic, to obtain hybrid membranes of a desirable structure and properties applicable for nuclear purposes is also possible. Different chemical and physical methods are used for membrane modiication (Van der Bruggen 2009). The properties of membranes can be modiied by introducing ionic groups, which can alter the character of the membrane surface from hydrophobic to hydrophilic (Bolong et al. 2009; Rana et al. 2005; Wei et al. 2010; Zhou et al. 2009). This modiication is important since it can reduce fouling of the membranes. Hydrophilization can be attained by using methods such as chemical oxidation, plasma treatment, chemical functionalization, or radiation-induced surface grafting. Surfactant modiication and self-assembly of the hydrophilic nanoparticles are other methods for membrane surface modiication. Another approach is to modify the polymer before membrane formation, by incorporating the hydrophilic additives into the membrane matrix. This can be achieved by sulfonation, carboxylation, or nitration. Chemical oxidation is achieved by introducing oxygen-containing groups into the membrane surface, using oxidants such as potassium permanganate, nitric acid, and chromic acid, or redox initiators such as ferric chloride. They can oxidize the membrane surface to create active sites where surface graft polymerization is conducted. Chemical functionalization is achieved by attaching speciic functional groups that give new character to the membrane surface. Such groups are bonded with membrane surface groups by ionic or polar bonds and can cause the hydrophilization of the membrane. Plasma treatment in the presence of oxygen forms peroxides on the membrane surface that undergo further decomposition and form oxygen-containing radical groups, such as hydroxyls, carbonyls, or carboxyls. The technique can also be used for preparing membranes for surface graft polymerization. In this process, synthetic monomers are attached to the peroxide groups formed by the plasma treatment. The monomers then polymerize on the membrane surface, forming a thin layer with properties different from those of the initial material. Radiation-induced grafting is another modiication method that utilizes ultraviolet or ionizing radiation to produce active sites on the membrane surface for attaching different kinds of groups (IAEA-TECDOC-1465 2005). The polymerization of the monomer grafted to these active sites results in membranes with different properties. The most common monomers used for radiation-induced grafting are vinyl acetate, N-vinyl pyrrolidone, acrylic acid, metacrylic acid, or N-vinyl pyridine. Radiation-based technologies are employed for the development of a new class of materials, namely, stimuli-responsive polymers and membranes that can be applied in nuclear science and technology. It is foreseen that temperature-sensitive or pHsensitive membranes can be used in a variety of novel applications, including separation processes in the nuclear ield, for the recovery of uranium from seawater. Inorganic membranes can also be modiied according to their future applications. Ceramic membranes present many advantages, such as mechanical and chemical resistances, thermal stability, nonswelling properties, and easy cleaning, and are very

Membranes in Nuclear Science and Technology

9

attractive for nuclear technologies. However, separation based on size exclusion does not give suficient selectivity for a wide range of applications. The recent research on the modiication of ceramic membranes by alcohol adsorption demonstrated a simple method for modifying the membrane properties (Dainov et al. 2002). The chemisorbed alcohol causes hydrophobization of the surface imparted by the alkyl chains and, in consequence, decreases water transport. The ceramic membrane with the adsorbed layers presents a high stability against acids, down to pH = 1. This shows the potential for the use of such modiied membranes with highly acidic solutions, for example, in wastes originating from fuel reprocessing. The emerging ield of the study presents the application of nanostructured materials and nanoparticles embedded in the membrane matrix to produce functionalized membranes with controlled fouling resistance (Kim and Bruggen 2010).

1.2 1.2.1

REMOVAL OF RADIONUCLIDES WITH SURFACE-MODIFIED MEMBRANES; LABORATORY EXPERIMENTS: UF AND MD CASE STUDY: APPLICATION OF SURFACE-MODIFIED MEMBRANES BASED ON POLYSULFONE AND POLYETHERSULFONE IN UF/COMPLEXATION PROCESS FOR THE R EMOVAL OF 60 CO FROM WATER AQUEOUS S OLUTIONS

UF coupled with complexation by water-soluble polymers was tested to assess the membrane surface modiication for the removal of radioactive cobalt (60Co) from water solutions. The experiments were conducted with porous membranes prepared using PES and modiied PES membranes made using surface-modifying macromolecules (SMMs). Both the modiied and the unmodiied membranes were prepared by a simple phase inversion technique in a single casting step from a blend dope containing PES and an SMM. During the membrane formation, the SMM migrates toward the top air–polymer interface, rendering the membrane surface hydrophobic. SMMs are oligomeric luoropolymers synthesized by polyurethane chemistry and tailored with luorinated end groups (Khayet et al. 2003, 2006). The prepared membranes are named hereafter, SMM3/PES and SMM41/PS. A commercial PES membrane supplied by Millipore, having a 10 kDa molecular weight cutoff (MWCO), was used for comparison. A typical UF setup equipped with a stainless steel membrane cell of an effective iltration area of 14.52 × 10−4 m2 was employed in the experiments (Figure 1.1). To assess the permeability of each membrane, the permeate lux of the distilled water (Jw) was measured for the SMM-modiied membranes, the unmodiied PES membrane, and the commercial PES membrane. It was observed that the SMMmodiied membranes exhibited lower permeation rates than the commercial PES and the unmodiied membrane. The lux of the distilled water was highest for the PES commercial membrane (168.6 kg/m 2 h) and for the prepared unmodiied PES membrane (127.6 kg/m 2 h). For the modiied membranes, SMM3/PES and SMM41/PES, the permeate luxes were lower at 41.3 and 39.5 kg/m 2 h, respectively (Table 1.1).

10

Membrane Modification: Technology and Applications

4

7

1

5 8 6 9

3

2

FIGURE 1.1 Setup for ultrailtration experiments. 1: Membrane module; 2: feed tank; 3: permeate tank; 4: permeate low meter; 5: feed low meter; 6: pump; 7: feedback valve; 8: control valve; and 9: thermometer.

Before the UF experiments, to reduce the adsorption of the radioactive cobalt in the unit, the UF system was pretreated by circulating nonradioactive 0.1 g/dm3 Co(NO3)2 · 6H2O aqueous solution for 1 h. The electrical conductivity of this solution was about 100 μS/cm. With this solution for each tested membrane, the permeation lux (Js) and the retention factor (Rs) of the cobalt ions were determined. For all tested UF membranes, the retention of the cobalt ions was not high. The highest retention factor was 51.7%, which was achieved for the SMM3/PES membrane. To enhance the eficiency of the cobalt ions removal, a macromolecular complexing agent, allowing the formation of larger molecules with the cobalt ions, was added in further experiments. Polyethyleneimine (PEI) (1 g/dm3) was introduced for the complexation of the Co2+ ions. The pH of the solution was adjusted with 10% HNO3 at the optimal value (pH = 6). In these conditions, the permeation lux (JCo/PEI) and the retention factor (RCo/PEI) for the hybrid UF/complexation process were determined (Table 1.1). TABLE 1.1 Permeation Characteristics of Membranes Membrane Commercial PES Unmodiied PES Modiied SMM3/PES Modiied SMM4/PES

Jw (kg/m2 h)

Js (kg/m2 h)

JCo/PEI (kg/m2 h)

Rs (%)

RCo/PEI (%)

168.6 127.6 41.3 39.5

83.1 85.1 26.5 24.0

55.2 58.7 20.2 17.7

29.2 44.2 51.7 48.2

93.6 70.8 98.7 96.7

11

Membranes in Nuclear Science and Technology

TABLE 1.2 Decontamination Results for 0.1 g/dm3 Co(NO3)2 · 6H2O Aqueous Solution Containing 60Co Membrane Commercial PES Unmodiied PES Modiied SMM3/PES Modiied SMM41/PES

AF (kBq/dm3)

AR (kBq/dm3)

AP (kBq/dm3)

DF

Am (counts/100 sec)

6.94 6.77 6.75 6.89

1.70 2.84 2.82 3.65

0.103 0.067 0.022 0.034

44 75 223 163

2941 1532 401 600

Note: Cobalt ions complexed by 0.1 g/dm3 polyethyleneimine (PEI).

The SMM-modiied membranes exhibited the highest retention of the Co2+ ions. It was over 98% for the SMM3/PES membrane and 97% for the SMM41/PES membrane. Under the same conditions, the retention of the prepared unmodiied PES was only 71%. A radioactive solution containing 60Co of the total radioactivity of ca. 4000 counts/100 sec was employed as a radioactive feed. The respective radioactivity counting rate (pulse/sec), treated as the intensity of the γ-radiation of a 5 cm3 liquid sample containing 60Co, was measured by a γ-analyzer, LG-1B (INCT). The obtained values were compared with the radiation intensity of the 60Co standard sample. Therefore, the speciic activity of the feed (AF), the retentate (AR), and the permeate (AP) were calculated. The parameters (R), (DF), and (J) were determined when the radioactive solutions were treated. In all the experiments, the transmembrane hydrostatic pressure was maintained at 0.22 MPa and the feed low rate at 40 l/h. Substantially high DFs for radioactive 60Co were obtained for the SMM-modiied membranes used in the UF/complexation process: 223 for SMM3/PES and 163 for SMM41/PES (Table 1.2). For the unmodiied membranes, the DFs were lower: 44 for the commercial PES membrane and 75 for the prepared unmodiied PES membrane. Additionally, the SMM-modiied membranes showed smaller adsorption of the radioactive cobalt than the modiied membranes, which was beneicial, taking into account the considered applications. After a 2 h operation, the adsorption of 60 Co by the SMM-modiied membranes was four to ive times smaller than that of the unmodiied PES membrane. The washing tests showed that the adsorption of the radioactive substances occurred in all tested membranes. Subsequent cleaning cycles with water and chemicals (alternant acidic and base washing) did not result in the complete removal of radioactivity, owing to the deep adsorption inside the membrane pores and the strong binding of the radioactive species within the membrane matrix. The variation of the radioactivity of the membrane during the operation, interrupted by water washing cycles, is shown in Figure 1.2. It is apparent that the effect of the subsequent cleaning procedures decreases over time and that the radioactivity of the membrane stabilizes at the level corresponding to persistent sorption in the membrane material.

12

Membrane Modification: Technology and Applications

Am (counts/s)

30 25

Chemical cleaning

20 15

Comm. PES Unmod. PES SMM3 SMM41

10 5 0 0

50

100

150

200

250

300

t (min)

FIGURE 1.2 Changes of membrane radioactivity in polymer-enhanced UF removal of 60Co from a water solution after water washing cycles and chemical cleaning.

More radical chemical cleaning resulted in a decline in the membrane-bound activity; however, it remained signiicant. The lowest adsorption observed at the SMM3 and SMM41 membranes predestines these materials to nuclear applications, namely, radioactive waste treatment.

1.2.2 CASE STUDY: SURFACE-MODIFIED MEMBRANES APPLIED FOR THE TREATMENT OF R ADIOACTIVE WASTES BY MD Direct contact membrane distillation (DCMD) is an effective process applied for the concentration of dissolved matter in water solutions. The process can be considered a feasible method for liquid radioactive waste treatment and, in some cases, it can effectively serve in reducing the hazard connected with its toxicity. Four porous composite hydrophobic/hydrophilic membranes were tested for the removal of selected radionuclides from water solutions. The membranes were also prepared by the phase inversion method using SMMs. Two hydrophilic polymers, PES and PS, were used. Each polymer was dissolved in the solvent/nonsolvent mixture and the macromolecules SMM3 or SMM41 were then added to the polymer casting solutions, to construct the membranes denoted as SMM3/PES, SMM41/PES, SMM3/PS, and SMM41/PS. For comparison, the commercial PTFE membranes, supported by a PP net (TF200, Gelman: thickness 178 μm, average pore size of 0.20 μm and porosity of 80%), were used. DCMD experiments were conducted using the setup shown in Figure 1.3. The feed liquid was brought in direct contact with the hydrophobic side of the membrane, while the distilled water collecting the permeate was brought in direct contact with the hydrophilic side of the membrane. The effective membrane area was 7.55 × 10−4 m2, and the bulk feed and the permeate were kept at 49°C and 24°C, respectively. To circulate both the feed and the permeate, peristaltic pumps were used, and for measuring the temperatures inside the test cell, two Pt100 thermometers were applied. To determine the water vapor permeability (permeate lux Jw) in the DCMD process, distilled water was used as the feed. To avoid the adsorption of the radioactive species in the system before using the solutions containing 60Co, 137Cs, and 85Sr radioisotopes,

13

Membranes in Nuclear Science and Technology

9

17

4

8

12

6

10 11

14

16 3

5

15

7

1

13

2

FIGURE 1.3 Setup for membrane distillation experiments. 1: Membrane module (test-cell); 2: feed tank; 3, 17: control valves; 4: distillate tank; 5, 6: heat exchanger; 7, 8: peristaltic pumps; and 9, 10, 11, 12, 13, 14, 15, 16: thermometers.

the solutions of the same nonradioactive ions as salts, Co(NO3)2 · 6H2O, CsCl, and SrCl2 · 6H2O in concentrations equivalent to the conductivity of 1000 μS/cm, were irst circulated through the DCMD apparatus to saturate the unit. Finally, small volumes of radioactive solutions containing 60Co, 137Cs, or 85Sr in 0.1 M HCl were added, to establish the feed activity between 4000 and 4500 counts/100 sec for the 50 cm3 sample. For all the tested membranes, the initial conductivity of the permeate was 9–16 μS/cm and the activity was between 400 and 440 counts/100 sec. The results are shown in Table 1.3. It can be observed that the commercial TF200 membrane exhibited higher permeate lux (2.1 kg/m2 h) than the SMM-modiied membranes. This was because the SMM-modiied membranes were prepared with a high concentration of PES and PS polymers, leading to pore sizes of an order of magnitude lower than those of the membrane TF200. The DCMD results showed that the rejection of Co2+, Cs−, and Sr2+ ions, as well as the radioactive isotopes 60Co, 137Cs, and 85Sr, was almost complete. No changes in the electrical conductivity and the speciic radioactivity in the permeate samples collected at the beginning and the end of the experiments (after 4 h) were observed.

14

Membrane Modification: Technology and Applications

TABLE 1.3 Removal of Radionuclides of 60Co, 85Sr, and 137Cs by MD Process Jw (kg/m2 h)

ρF (μs/cm)

ρD (μs/cm)

R (%)

AF (kBq/dm3)

AD (Bq/dm3)

DF

TF200

2.10

SMM3/PES

0.86

SMM41/PES

0.72

SMM3/PS

0.44

SMM41/PS

0.35

982a 996b 1017c 977a 983b 1005c 1000a 982b 980c 978a 992b 984c 975a 994b 994c

2 3 14 4 14 15 46 21 20 7 3 20 16 3 10

99.18 99.70 98.62 99.59 98.57 98.51 95.40 97.86 97.96 99.28 99.70 97.97 98.31 99.70 98.99

7.15 2.85 3.66 7.03 2.90 2.69 7.02 2.91 3.70 7.07 2.87 3.68 7.01 2.88 3.69

3.9 7.2 1.2 3.7 7.2 6.2 4.5 14.5 9.1 2.5 8.5 2.0 1.4 4.5 1.0

1833 396 3050 1900 403 434 1560 201 407 2828 338 1840 5007 640 3690

Membrane

a

60

b

85

c

Co. Sr. 137Cs.

It was concluded that the SMM-modiied membranes could be used effectively for rejecting the radioactive compounds in water solutions. Moreover, the radioactivity of the SMMs membranes after 4 h DCMD experiments was substantially lower than that of the commercial membrane TF200 (Table 1.4), suggesting a smaller adsorption of the radionuclides on the SMM-modiied membrane, which can be regarded as an advantageous property and important for further application in nuclear technologies.

TABLE 1.4 Radioactivity of the Membrane after 4 h MD Experiments with 60Co, 85Sr, and 137Cs Solutions Am (counts/100 sec) Commercial Membrane Solution 60

Co 85Sr 137Cs

Modified Membranes

TF200

SMM3/PES

SMM41/PES

SMM3/PS

SMM41/PS

1799 584 917

487 282 358

679 152 447

574 101 251

691 54 334

Membranes in Nuclear Science and Technology

15

REFERENCES Andalaft, E., Vega, R., Correa, M., Araya, R. and Loyola, P. 1997. Zeta potential control in decontamination with inorganic membranes and inorganic adsorbents. In: Treatment Technologies for Low and Intermediate Level Waste from Nuclear Applications. Final Report of a Coordinated Research Programme 1991–1996. IAEA-TECDOC-929, pp. 15–32. IAEA: Vienna. Application of membrane technologies for liquid radioactive waste processing. 2004. Vienna Technical Reports Series No. 431. IAEA: Vienna. Arnal, J.M., Campayo Esteban, J.M., Lora Garcia, J., Sancho Fernandez, M., Iborra Clar, I. and Alcaina Miranda, I. 2000. Declassiication of radioactive waste solutions of iodine (I125) from radioimmune analysis (RIA) using membrane techniques. Desalination 129: 101–105. Arnal, J.M., Sancho, M., Verdu, G., Campayo, J.M. and Villaescusa, J.I. 2003a. Treatment of 137Cs liquid wastes by reverse osmosis. Part I. Preliminary tests. Desalination 154: 27–33. Arnal, J.M., Sancho, M., Verdu, G., Campayo, J.M. and Gozálvez, J.M. 2003b. Treatment of 137Cs liquid wastes by reverse osmosis. Part II. Real application. Desalination 154: 35–42. Bolong, N., Ismail, A.F., Salim, M.R., Rana, D. and Matsuura, T. 2009. Charged surface modifying macromolecule inluence to polyethersulfone hollow iber membrane with polyvinylpyrrolidone and water. J. Memb. Sci. 331: 40–49. Bridesell, S.A. and Willms, R.S. 1998. Tritium recovery from tritiated water with a two-stage palladium membrane reactor. Fusion Eng. Des. 39–40: 1041–1048. Brown, R.G., Crowe, M.H., Hebditch, D.J., Newman, R.N. and Smith, K.L. 1991. A comparative evaluation of cross-low microiltration membranes for radwaste dewatering. In: Effective Industrial Membrane Processes: Beneits and Opportunities, ed. M.K. Turner., pp. 103–113. Elsevier Applied Science: London. Burggraaf, A.J. and Cot, L. 1996. Membrane Science and Technology Series. Vol. 4. Fundamentals of Inorganic Membrane Science and Technology. Elsevier: Amsterdam. Campbell, D.J. 1985. Fractionation of stable chlorine isotopes during transport through semipermeable membranes. M.Sc. Thesis, University of Arizona. Chitry, F., Pellet-Rostaing, S., Vigneau, O. and Lemaire, M. 2001. Nanoiltration-complexation: A new method for isotopic separation of heavy metals. Chem. Lett. 770–771. Chmielewski, A.G. and Harasimowicz, M. 1992. Inluence of gamma and electron irradiation on transport properties of ultrailtration membranes. Nukleonika 37(4): 61–70. Chmielewski, A.G. and Harasimowicz, M. 1997. Inluence of gamma and electron irradiation on transport properties of nanoiltration and hyperiltration membranes. Nukleonika 42(4): 857–862. Chmielewski, A.G., Zakrzewska-Trznadel, G., Miljević, N. and Van Hook, W.A. 1991. Investigation of the separation factor between light and heavy water in the liquid/vapour membrane permeation process. J. Memb. Sci. 55: 257–262. Cojocaru, C., Zakrzewska-Trznadel, G. and Jaworska, A. 2009a. Removal of cobalt ions from aqueous solutions by polymer assisted ultrailtration using experimental design approach. Part 1: Optimization of complexation conditions. J. Hazard. Mater. 169: 599–609. Cojocaru, C., Zakrzewska-Trznadel, G. and Miskiewicz, A. 2009b. Removal of cobalt ions from aqueous solutions by polymer assisted ultrailtration using experimental design approach. Part 2: Optimization of hydrodynamic conditions for a cross-low ultrailtration module with rotating part. J. Hazard. Mater. 169: 610–620. Coufin, N., Cabassud, C. and Lahoussine-Turcaud, V. 1998. A new process to remove halogenated VOCs for drinking water production: Vacuum membrane distillation. Desalination 117: 233–245.

16

Membrane Modification: Technology and Applications

Dainov, A., Garcia-Valls, R. and Font, J. 2002. Modiication of ceramic membranes by alcohol adsorption. J. Memb. Sci. 196: 69–77. Duncan, J.B. and Nelson, D.A. 1999. The separation of tritiated water using supported polyphosphazene membranes. J. Memb. Sci. 157: 211–217. El-Reefy, S.A., El-Sherif, E.A. and Aly, H.F. 1996. Recovery of 234Th from natural uranium using liquid emulsion membrane based on HDEHP-HCl system. J. Radioanal. Nucl. Chem. 207(1): 129–136. El-Said, N., Rahman, N. and Borai, E.H. 2002. Modiication in PUREX process using supported liquid membrane separation of cerium(III) via oxidation to cerium(IV) from ission products from nitrate medium SLM. J. Memb. Sci. 198: 23–31. Evans, J., Harris, I.R. and Ross, D.K. 1983. A proposed method of hydrogen isotope separation using palladium alloy membranes. J. Less Common Met. 89: 407–414. Fain, D.E. and Brown, W.K. 1974. Neon isotope separation by gaseous diffusion transport in the transition low regime with regular geometries. U.S. Atomic Energy Commission Report. Fountain, M.S., Kurath, D.E., Sevigny, G.J. et al. 2008. Caustic recycle from Hanford tank waste using NaSICON ceramic membranes. Sep. Sci. Technol. 43: 2321–2342. Fritz, S.J., Hinz, D.L. and Grossman, E.L. 1987. Hyperiltration-induced fractionation of carbon isotopes. Geochim. Cosmochim. Acta 51: 1121–1134. Gryta, M. 2002. Concentration of NaCl solutions by membrane distillation integrated with crystallization. Sep. Sci. Technol. 37(15): 3535–3558. Happel Streng, R., Vater, P. and Ensinger, W. 2003. Sr/Y separation by supported liquid membranes based on nuclear track microilters. Radiat. Meas. 36: 761–766. Hayashi, T., Yamada, M., Suzuki, T., Matsuda, Y. and Okuno, K. 1995. Gas separation performance of hollow-ilament type polyimide membrane module for a compact tritium removal system. Fusion Technol. 28: 1503–1508. Hegazy, E.A., Abd El-Rehim, H.A., Ali, A.M.I., Nowier, H.G. and Aly, H.F. 1999. Characterization and application of radiation grafted membranes in treatment of intermediate active waste. Nucl. Instrum. Meth. Phys. Res. B 151: 393–398. Heinze, S., Bussiere, P. and Pelletier, T. 2003. French experience in tritiated water management. Fusion Eng. Des. 69: 67–70. Hirata, S., Kakuta, T., Ito, H. et al. 1995. Experimental and analytical study on membrane detritiation process. Fusion Technol. 28: 1521–1526. Hobbs, D.T. 1999. Caustic recovery from alkaline nuclear waste by an electrochemical separation process. Sep. Purif. Technol. 15: 239–253. Hogan, P.A., Sudjito, Fane, A.G. and Morrison, G.L. 1991. Desalination by solar heated membrane distillation. Desalination 81: 81–90. Hooper, E.W. 1997. Inorganic sorbents for removal of radioactivity from aqueous waste streams. In: Waste Treatment and Immobilization Technologies Involving Inorganic Sorbents. IAEA-TECDOC-947, pp. 221–235. IAEA: Vienna. Hsieh, H.P. 1996. Membrane Science and Technology Series. Vol. 3. Inorganic Membranes for Separation and Reaction. Elsevier: Amsterdam. Hwang, E.-D., Lee, K.-W., Choo, K.-H. et al. 2002. Effect of precipitation and complexation on nanoiltration of strontium-containing nuclear wastewater. Desalination 147: 289–294. IAEA-TECDOC-911. 1996. Processing of nuclear power plant waste streams containing boric acid. IAEA-TECDOC-911. IAEA: Vienna. IAEA-TECDOC-1465. 2005. Radiation synthesis of stimuli-responsive membranes, hydrogels and adsorbents for separation purposes. IAEA-TECDOC-1465. IAEA: Vienna. Inoue, H., Kagoshima, M., Yamasaki, M. and Honda, Y. 2004. Radioactive iodine waste treatment using electrodialysis with an anion exchange paper membrane. Appl. Radiat. Isot. 61: 1189–1193.

Membranes in Nuclear Science and Technology

17

Ishida, T., Hayashi, T., Mori, S., Suzuki, T. and Nishi, M. 2000. Design of a membrane atmosphere detritiation system using super high permeation module. Fusion Eng. Des. 49–50: 839–846. Karlin, Y., Gorbachev, D., Volkov, A. and Barinov, A. 2001. Advantageous technology treatment of laundry waters. In: Proceedings of the International Symposium on Technologies for the Management of Radioactive Waste for Nuclear Power Plants and Back End of Nuclear Fuel Cycle Activities, Daejon, 30 August–03 September, 1999. C&S Papers CD Series no. 6. IAEA: Vienna. Kedari, C.S., Pandit, S.S. and Ramnujam, A. 1999. Selective permeation of plutonium (IV) through supported liquid membrane containing 2-ethylhexyl 2-ethyl-hexyl phosfonic acid as ion carrier. J. Memb. Sci. 156: 187–196. Khayet, M. 2011. Membranes and theoretical modeling of membrane distillation: A review. Adv. Colloid Interface Sci. 164: 56–88. Khayet, M., Suk, D.E., Narbaitz, R.M., Santerre, J.P. and Matsuura, T. 2003. Study on surface modiication by surface-modifying macromolecule (SMMs) and its applications in membrane-separation processes. J. Appl. Polym. Sci. 89: 2902–2916. Khayet, M., Matsuura, T. and Mengual, J.I. 2006. Design of novel direct contact membrane distillation membranes. Desalination 192: 105–111. Kim, J. and Bruggen, B.V. 2010. The use of nanoparticles in polymeric and ceramic membrane structures: Review of manufacturing procedures and performance improvement for water treatment. Environ. Pollut. 158: 2335–2349. Kim, J., Park, S.E., Kim, T.-S., Leong, D-Y. and Ko, K.-H. 2004. Isotopic water separation using AGMD and VEMD. Nukleonika 49(4): 137–142. Kocherginsky, N.M., Zhang, Y.K. and Stucki, J.W. 2002. D2EHPA based strontium removal from strongly alkaline nuclear waste. Desalination 144: 267–272. Konishi, S., Yoshida, H. and Naruse, Y. 1983. A design study of a palladium diffuser for D-T fusion reactor fuel clean-up system. J. Less Common Met. 89: 457–464. Konishi, S., Maruyama, T., Okuno, K., Inoue, M. and Yamashita, A. 1998. Development of electrolytic reactor for processing of gaseous tritiated compounds. Fusion Eng. Des. 39–40: 1033–1039. Kryvoruchko, A.P., Yurlova, L.Yu., Atamanenko, I.D. and Kornilovich, B.Yu. 2004. Ultrailtration removal of U(VI) from contaminated water. Desalination 162: 229–236. Kurath, D.E., Brookes, K.P., Hollenberg, G.W., Sutija, D.P., Landro, T. and Balagopal, S. 1997. Caustic recycle from high-salt nuclear wastes using a ceramic-membrane saltsplitting process. Sep. Purif. Technol. 11: 185–198. Lakshmi, D.S., Mohapatra, P.K., Mohan, D. and Manchanda, V.K. 2004. Uranium transport using a PTFE lat-sheet membrane containing alamine 336 in toluene as the carrier. Desalination 163: 13–18. Le Digabel, M., Ducret, D., Laquerbe, C., Perriat P. and Niepce, J.-C. 2002. Application of gas separation membranes to detritiation systems. Desalination 148: 297–302. Macnaughton, S.J., McCulloch, J.K., Marshall, K. and Ring, R.J. 2002. Application of nanoiltration to the treatment of uranium mill efluents. In: Technologies for the Treatment of Efluents from Uranium Mines, Mills and Tailings. IAEA-TECDOC-1296, pp. 55–65. IAEA: Vienna. Mann, N.R. and Todd, T.A. 2000. Crosslow iltration testing of INEEL radioactive and nonradioactive waste slurries. Chem. Eng. J. 80: 237–244. Marcea, P. 1983. Permeation of H2 and D2 through polymers. Isotopenpraxis 19(5): 153–155. Mulder, M. 1991. Basic Principles of Membrane Technology. Kluwer Academic: Dordrecht. Nelson, D.A., Duncan, J.B., Jensen, G.A. and Burton, S.D. 1996. Isotopomeric water separations with supported polyphosphazene membranes. J. Memb. Sci. 112: 105–113.

18

Membrane Modification: Technology and Applications

Nörenberg, H., Burlakov, V.M., Kosmella, H.-J. et al. 2001. Pressure-dependent permeation of noble gases (He, Ne, Ar, Kr, Xe) through thin membranes of oriented polypropylene (OPP) studied by mass spectrometry. Polymer 42: 10021–10026. Ohno, M., Ozaki, O., Sato, A., Kimura, S. and Miauchi, T. 1977. Radioactive rare gas separation using a separation cell with two kinds of membrane differing in gas permeability tendency. J. Nucl. Sci. Technol. 14(8): 589–602. Ramachandhran, V. and Misra, B.M. 1985. Studies on the radiolytic degradation of cellulose acetate membranes. J. Appl. Polym. Sci. 30: 35–43. Ramachandhran, V. and Misra, B.M. 1986. Studies on the radiation stability of ion exchange membranes. J. Appl. Polym. Sci. 32: 5743–5747. Ramachandhran, V. and Misra, B.M. 1998. Radiocerium separation behaviour of ultrailtration membranes. J. Radioanal. Nucl. Chem. 237(1–2): 121–124. Rana, D., Matsuura, T., Narbaiz, R.N. and Feng, C. 2005. Development and characterization of novel hydrophilic surface modifying macromolecule for polymeric membranes. J. Memb. Sci. 249: 103–112. Sasaki, Sh., Tega, E., Shimada, A. et al. 2003. Basic characteristics of hollow-ilament polyimide membrane in gas separation and application to tritium monitors. J. Radioanal. Nucl. Chem. 255(1): 91–95. Sen Gupta, S.K., Buckley, L.P., Rimpelainen, S. and Tremblay, A.Y. 1996. Liquid rad-waste processing with spiral wound reverse osmosis. AECL-11542 Report, Chalk River, ON. Sriram, S., Mohapatra, P.K., Pandey, A.K., Manchanda, V.K. and Badheka, L.P. 2000. Facilitated transport of americium(III) from nitric acid media using dimethyl-dibutyltetradecyl-1,3malonamide. J. Memb. Sci. 177: 163–175. Smyth, B.F., Gibson, R.R., Jarvinen, G.D., Robinson, T.W., Schroeder, N.C. and Stalnaker, N. 1998. Preconcentration of low levels of americium and plutonium from waste waters by synthetic water-soluble metal-binding polymers with ultrailtration. J. Radioanal. Nucl. Chem. 234(1–2): 225–229. Smyth, B.M., Todd, P. and Bowman, N. 1999. Hyperbranched chelating polymers for the polymer-assisted ultrailtration of boric acid. Sep. Sci. Technol. 34(10): 1925–1945. Stern, S.A. and Leone, S.M. 1980. Separation of krypton and xenon by selective permeation. AIChE J. 26(6): 881–890. Stern, S.A. and Wang, S.C. 1980. Permeation cascades for the separation of krypton and xenon from nuclear reactor atmospheres. AIChE J. 26(6): 891–901. Tosti, S., Violante, V., Basile, A. et al. 2000. Catalytic membrane reactors for tritium recovery from tritiated water in the ITER fuel cycle. Fusion Eng. Des. 49–50: 953–958. Van der Bruggen, B. 2009. Chemical modiication of polyethersulfone nanoiltration membranes: A review. J. Appl. Polym. Sci. 114(1): 630–642. Violante, V., Basile, A. and Drioli, E. 1995. Composite catalytic membrane reactor analysis for water gas shift reaction in the tritium fusion fuel cycle. Fusion Eng. Des. 30: 217–223. Wei, X., Wang, Z., Chen, J., Wang, J. and Wang, S. 2010. A novel method of surface modiication on thin-composite reverse osmosis membrane by grafting hydantoin derivative. J. Memb. Sci. 346: 152–162. Whitworth, T.M., Mariñas, B.J. and Fritz, S.J. 1994. Isotopic fractionation and overall permeation of lithium by a thin-ilm composite polyamide reverse osmosis membrane. J. Memb. Sci. 88: 231–241. Xiarchos, I., Jaworska, A. and Zakrzewska-Trznadel, G. 2008. Response surface methodology for the modelling of copper removal from aqueous solutions using micellar-enhanced ultrailtration. J. Memb. Sci. 321: 222–231. Zakrzewska-Trznadel, G. 1998. Membrane distillation for radioactive waste treatment. Memb. Technol. (An International Newsletter) 103: 9–12. Zakrzewska-Trznadel, G. 2003. Radioactive solutions treatment by hybrid complexation–UF/NF process. J. Memb. Sci. 225: 25–39.

Membranes in Nuclear Science and Technology

19

Zakrzewska-Trznadel, G. 2008. Radioactive waste processing: Advancement in pressure driven processes and current world scenario. In: Handbook of Membrane Separations: Chemical, Pharmaceutical, and Biotechnological Applications, eds. A.K. Pabby, S.S.H. Rizvi and A.M. Sastre, pp. 843–882. CRC Press Taylor & Francis Group: Boca Raton, FL. Zakrzewska-Trznadel, G. and Harasimowicz, M. 2002. Removal of radionuclides by membrane permeation combined with complexation. Desalination 144: 207–212. Zakrzewska-Trznadel, G. and Harasimowicz, M. 2004. Application of ceramic membranes for hazardous wastes processing: Pilot plant experiments with radioactive solutions. Desalination 162: 191–199. Zakrzewska-Trznadel, G., Chmielewski, A.G. and Miljević, N. 1996. Separation of protium/ deuterium and oxygen-16/oxygen-18 by membrane distillation process. J. Memb. Sci. 113: 337–342. Zakrzewska-Trznadel, G., Harasimowicz, M. and Chmielewski, A.G. 1999. Concentration of radioactive components in liquid low-level radioactive waste by membrane distillation. J. Memb. Sci. 163: 257–264. Zakrzewska-Trznadel, G., Harasimowicz, M., Miśkiewicz, A., Jaworska, A., Dluska, E. and Wronski, S. 2009. Reducing fouling and boundary-layer by application of helical low in ultrailtration module employed for radioactive wastes processing. Desalination 240: 108–116. Zhou, Y., Yu, S., Gao, C. and Feng, X. 2009. Surface modiication of thin composite polyamide membranes by electrostatic self-deposition of polycations for improved fouling resistance. Sep. Purif. Technol. 66: 287–294.

2

Use of Impedance Spectroscopy for Characterization of Modified Membranes Juana Benavente

CONTENTS 2.1 2.2 2.3

Introduction .................................................................................................... 21 Theory............................................................................................................. 23 Application of Impedance Spectroscopy to Membranes ................................ 27 2.3.1 Membranes ......................................................................................... 27 2.3.2 Typical Impedance Spectroscopy Experimental System ...................28 2.4 Results and Discussion ................................................................................... 29 2.4.1 Reverse Osmosis Membrane............................................................... 29 2.4.2 Polysulfone–Polyamide/Polyethylene Glycol Membrane: Effect of PEG Content ........................................................................ 31 2.4.3 Study of Composite Structures: Effect of Sublayers Chemical Nature ................................................................................................. 32 2.4.4 Effect of Membrane Material Modiication ....................................... 33 2.4.5 Effect of Protein Fouling on EIS Plots ............................................... 35 2.5 Conclusion ...................................................................................................... 37 Acknowledgment ..................................................................................................... 38 References ................................................................................................................ 38

2.1

INTRODUCTION

It is well known that membrane charge can play an important role in the transport of electrolyte and charged particles across membranes and their electrical characterization has been of great importance since the beginning of membrane processes applications (Helfferich 1962; Laksminarayanaiah 1969; Compañ et al. 1997; Vallejo et al. 1999). Usually, the determination of an effective ixed charge concentration in the membranes (bulk) and the number of ions transported across them is made from the membrane potential (MP) measurements (Demish and Pusch 1980; Kimura et al. 1984; Benavente and Fernández-Pineda 1985; López et al. 2001), while the streaming potential (SP) gives information on the membrane–solution electrical interface, such as 21

22

Membrane Modification: Technology and Applications

the surface charge density and the isoelectric point; however, the characteristic adsorption parameters, such as Gibbs energy and the number of accessible sites, can also be determined by using the appropriate models (Benavente et al. 1993; Molina et al. 1999; de Lara and Benavente 2007). In particular, SP measurements are commonly used for determining membrane changes at both the pore wall and the membrane surface, associated with fouling, membrane material modiication, and deterioration or age (Childress and Elimelech 1996; Benavente and Jonsson 1998; Pointié 1999; Fievet et al. 2003). The membrane electrical resistance (or conductivity) is also an electrical parameter of major signiicance in the case of charged membranes, such as those used in electrodialysis, electrochemical devices, and fuel cells. The conductivity and the dielectric constant, both intrinsic and material parameters, are determined from the electrical resistance and the capacitance values obtained from impedance spectroscopy (IS) measurements. IS can be carried out with the membranes in dry and wet (solution embedded) states (Benavente et al. 2010a,b) or in “working conditions,” that is, in contact with the electrolyte solutions, which is termed electrochemical impedance spectroscopy (EIS) (Mälmgren-Hansen et al. 1989; Asaka 1990; Benavente et al. 1997; Oleinikova et al. 2000; Li and Zhao 2004; Freger and Bason 2007). IS is a nondestructive ac technique for the electrical characterization of solid and liquid systems. It is also used for determining the interfacial effects or for monitoring the system changes associated with the time evolution of the impedance (Mijovic and Bellucci 1996; Hamdy et al. 2006). IS has emerged with the development of instruments that are capable of measuring impedance in a wide range of frequencies (between 10−6 and 109 Hz), and it allows the determination of the electrical properties of heterogeneous systems formed by a series array of layers with different electrical and/or structural properties, such as membrane/electrolyte systems. This permits a separate evaluation of the electrical contribution of each layer by using the impedance plots and the equivalent circuits as models, where the different circuit elements are related to the structural/transport properties of the systems (Buck and Ciani 1975; Macdonald 1987). In particular, in the case of EIS measurements carried out with membranes in working conditions, separate contributions of the membranes and the solution are usually obtained, except for membranes with a high charge or solution content. In addition, if composite/asymmetric membranes are studied, the analysis of the impedance plots might allow the separate evaluation of the electrical contribution associated with dense and porous layers and even the estimation of the geometrical parameters (Coster et al. 1992; Cañas et al. 2001; Torras et al. 2007). The research described above, which has used IS, demonstrates that it can provide qualitative and quantitative information, obtained from impedance measurements, and that the technique can be applied to determine the modiication of different commercial and experimental membranes with diverse structures (porous, dense, and composites) and materials. These modiications are associated with membrane fouling and aging, as well as changes purposely made to optimize the membrane performance. Most of the impedance measurements were carried out with the membranes in contact with the electrolyte solutions at different concentrations (electrode/electrolyte (c)/membrane/electrolyte (c)/electrode system), and the impedance curves for the electrolytes alone, without any membrane in the cell system, were also considered

Use of Impedance Spectroscopy for Characterization of Modified Membranes 23

to verify the measurement. In addition, the results obtained without electrolyte solution (electrode/membrane/electrode system) are presented in some cases to provide complementary information. The itting of theoretical models to the experimental data allows the determination of the electrical parameters, although geometrical parameters for the membrane, or the layers forming a composite membrane, such as porosity and/or thickness, might also be estimated from these results. The differences found in the impedance plots, the equivalent circuits, and the parameters associated with the original and the modiied membranes give information on the effect of the modiication on the electrical and transport behavior of the membrane.

2.2

THEORY

When a linear system is perturbed by a small v(t) voltage, its response, the electric current i(t), is determined by a differential equation of nth order in i(t) or by a set of n differential equations of the irst order. If v(t) is a sine wave input (Equation 2.1): v ( t ) = Vo sin ωt ,

(2.1)

the current intensity i(t) is also a sine wave (Equation 2.2): i ( t ) = I o sin ( ωt + φ ),

(2.2)

where Vo and Io represent the maximum voltage and intensity, respectively, while ω = 2πf is the angular frequency. A transfer function, the admittance function, can be deined as Y*(ω) = |Y(ω)|ejϕ, where |Y(ω)| represents the amplitude and ϕ is the phase angle. The impedance function, Z(ω), is the inverse of the admittance function, Z(ω)=[Y*(ω)]−1, and since both the amplitude and the phase angle of the output may change with respect to the input values, the impedance is expressed as a complex number, Z = Z real + jZimg, where Z real is the real part and Zimg is the imaginary part. The overall admittance (Equation 2.3) for a parallel resistor–capacitor (RC) circuit is given by the sum of the conductance (1/R) and capacitance contributions, where the resistance (R) represents the dissipative component of the dielectric response, while the capacitance (C) describes the storage component. The impedance function for that circuit is

(1 Z *) = (1 R ) + ( jωC ).

(2.3)

These expressions correlate the impedance components with the electrical parameters of the system. Moreover, the impedance can be separated into the real and imaginary parts by algebraic manipulation and the impedance components are related to the electrical parameters of the system by the following expressions:

(

)

(

)

2 2 Z real = R 1 + ( ωRC )  ; Z img = − ωR 2C 1 + ( ωRC )  .    

(2.4)

Membrane Modification: Technology and Applications

24

10,000

100

Equivalent circuit: (RC) Arc tan (φ)

–Zimg (Ω)

7,500 ω 5,000

10–1

10–2

2,500

R/2

0 0

2,500

(a)

5,000 Zreal (Ω)

7,500

10,000

10,000

10–3 103 (b)

104

105 f (Hz)

106

107

104

103

–Zimg (Ω)

Zreal (Ω)

7,500

5,000

Interface electrode/solution

102

2,500 101

0 103 (c)

104

105 f (Hz)

106

107 (d)

102

103

5 104 10 f (Hz)

106

107

FIGURE 2.1 Different impedance plots for the electrode/electrolyte solution (c)/electrode system. (a) Nyquist plot, (b) phase angle ϕ vs. frequency, Bode plots: (c) Zreal vs. frequency, and (d) –Zimg vs. frequency. (♢) Experimental values and (♦) calculated values for a certain number of frequencies.

The analysis of the impedance data can be carried out by a complex plane method using the Nyquist plot (−Zimg vs. Z real). The equation for a parallel RC circuit gives rise to a semicircle in the Z*(ω) plane, as that shown in Figure 2.1a, which has intercepts on the Z real axis at R∞(ω → ∞) and R0(ω → 0), where (R0 − R∞) is the resistance of the system. The maximum of the semicircle equals 0.5 (R0 − R∞) and occurs at such a frequency that ωRC = 1; τ = RC is the relaxation time (Macdonald 1987). The impedance plot shown in Figure 2.1a (−Zimg vs. Z real, or the Nyquist plot) corresponds to an electrochemical cell (electrode/NaCl solution/electrode) and the equivalent circuit consists of a resistance (R) in parallel with a capacitor (C), which is represented as (RC), while Figure 2.1b shows the variation of the phase angle ϕ = arc tan(Zimg/Z real) with frequency (ϕ vs. f), but other typical impedance representations correspond to the variation of Z real and −Zimg with frequency (Bode plots), as indicated in Figure 2.1c and 2.1d. This latter representation allows the determination of the interval of frequency associated with a given relaxation process, between 104 and 107 Hz, with a maximum frequency around 2 × 106 Hz, for the NaCl solution

Use of Impedance Spectroscopy for Characterization of Modified Membranes 25

represented in Figure 2.1, while an almost constant value for a frequency lower than 106 Hz can be observed in Figure 2.1c. However, a slight variation seems to exist at the lowest frequencies, usually associated with the interfacial effects, which is more evident in Figure 2.1b. As can be seen, the different types of plots shown in Figure 2.1 provide complementary information on the sample studied, but all of them show a unique relaxation process; a semicircle in Figure 2.1a and a symmetric and well-deined peak in Figure 2.1d. The itting of the experimental points shown in the Nyquist plot using a nonlinear program has permitted the determination of the following electrical resistance and capacitance values: R = (22,500 ± 12,300) Ω and C = (5.2 ± 0.3) × 10−12 F. These results were used for the calculation of the Z real and Zimg values for certain frequencies using Equation 2.4, and are also represented in Figure 2.1 as dense symbols. The good agreement between the experimental and the calculated data can be considered a test of the correctness of the circuit selected and the values obtained. When a membrane is placed in the middle of the electrochemical cell, the total system, electrode/electrolyte (c)/membrane/electrolyte (c)/electrode, can be considered as a heterogeneous system formed of two different subsystems (electrolyte solution and membrane), and the membrane contribution may be obtained separately if its dielectric properties are suficiently different from those corresponding to the electrolyte solution. In these cases, two different contributions, associated with each of the relaxation processes in one of the subsystems, may be obtained. As an example, Figure 2.2a and 2.2b show the impedance plots for a dense and symmetric polyamide membrane in contact with an NaCl solution, and two subcircuits and two semicircles in the Nyquist plot, or two symmetric and narrow peaks in the Bode plot, one of them associated with the membrane (fmax at 5000 Hz) and the other associated with the electrolyte solution (fmax between 1 and 2 MHz), were obtained. In the case of the composite RO or NF membranes, three different contributions for the electrolyte, the porous sublayer, and the dense layer can be obtained. However, the impedance curves depend not only on the membrane structure, but also on the membrane material, and its hydrophilic/hydrophobic character might mask some processes. This point is clearly shown in Figure 2.2c and 2.2d, where the impedance plots for a regenerated cellulose membrane with a high water uptake (W) are presented, with W between 70% and 85% (Vázquez et al. 2008). Due to the electrolyte content in this membrane, a unique relaxation process for the membrane system was obtained, with the equivalent circuit represented as RsmCsm, which does not allow the separate determination of the membrane electrical parameters and those associated with the electrolyte solution between the electrodes and the membrane surfaces; in this case, Rm can only be determined if the electrolyte is independently measured (Re), taking into account the series association rule for electrical resistance (Rme = Re + Rm). On the other hand, for this system, the presence of the membrane is more evident, considering the Bode plot (Figure 2.2d), since the wider width of the curve and the shift of its maximum to a lower frequency are indications of a more dense contribution. In many cases, complex systems present a distribution of relaxation times and the resulting plot is a depressed semicircle, which is associated with a nonideal capacitor or a constant phase element (CPE), and its impedance is expressed by Q(ω) = Y0(jω)−n, where Y0 represents the admittance and n is an experimental parameter (0 ≤ n ≤ 1)

Membrane Modification: Technology and Applications

26 90,000

(ReCe)-------(RmCm)

60,000

–Zimg (Ω)

–Zimg (Ω)

30,000

Dense symmetric membrane 30,000

Electrolyte solution

20,000

10,000 m

m

e

0 0

30,000 60,000 Zreal (Ω)

(a)

0

90,000

0

e

103

(b)

105

107

105

107

f (Hz)

24,000 Electrolyte: (ReCe) Membrane system: (RsmCsm) –Zimg (Ω)

–Zimg (Ω)

18,000

104

12,000

6,000

103

102

0 0 (c)

6,000

101

12,000 18,000 24,000 Zreal (Ω)

(d)

103 f (Hz)

FIGURE 2.2 Nyquist and Bode plots for two different electrode/electrolyte/membrane/electrode systems. (a, b) Dense and symmetric polyamide membrane. (c, d) Dense and symmetric highly hydrophilic regenerated cellulose membrane.

(Macdonald 1987). In these cases, an equivalent capacitance (Ceq) can be determined (Jonscher 1983): (1 n)

C eq = ( RY0 )

R.

(2.5)

A particular case is obtained when n = 0.5, then the circuit element corresponds to a “Warburg impedance” (W), which is associated with a diffusion process according to Fick’s irst law. Once the membrane’s electrical resistance and capacitance are obtained by the it of the impedance plots to the circuit models, the membrane conductivity (σ) and the dielectric constant (ɛr) of the symmetric samples, or the dense active layer of the composite membranes, can be determined if the geometrical parameters are known, taking into account the expressions for the homogeneous conductors and the planeplate capacitors, respectively: σ = ∆xm Rm Sc,

(2.6)

Use of Impedance Spectroscopy for Characterization of Modified Membranes 27

C = ε o ε r Sa ∆xm,

(2.7)

where Δxm is the membrane thickness, ɛo and ɛr are the permittivity of the vacuum and membrane dielectric constant, respectively, while Sa corresponds to the surface where the charge is adsorbed (Sa ≈ Sm, for the dense membrane/layer), and Sc represents the cross section for the charge transport. These examples clearly show how the EIS measurements allow the estimation of the membrane electrical parameters (Rm and Cm); however, they also indicate the possibility of obtaining “qualitative” information on the membrane structure, which can also be of great interest. In any case, it should be pointed out that impedance is an extensive magnitude (it depends on the sample area), and for that reason, comparisons of the type of curves and the concentration dependence instead of particular values are usually made. In addition, IS measurements with the dry and wet membranes, but without an electrolyte solution between the electrode and the membrane surface, can also be performed and complementary information, mainly related to the membrane material itself or the interfacial (electrode/membrane) effects, can be obtained. In addition to the impedance, other derived quantities, such as the dielectric modulus (M), the complex dielectric constant (ɛ), or susceptibility (χ), can be calculated from the IS measurements; their interrelations have been tabulated elsewhere (Macdonald 1987). Complementary information on the dielectric response of a given system can be obtained from the different impedance plots and the related magnitudes. It is important to point out the different nature of these magnitudes: extensive or sample geometry dependent in the case of impedance or admittance, and intensive or characteristic of homogeneous materials in the case of conductivity and the dielectric constant.

2.3 2.3.1

APPLICATION OF IMPEDANCE SPECTROSCOPY TO MEMBRANES MEMBRANES

To show the potential of the impedance measurements for determining a membrane, modiications associated with both manufacture effects, in terms of changes in the material and/or structure, and working process effects, such as fouling or age, were considered. This chapter will describe the research on commercial and experimental membranes from different materials and with diverse structures (porous, dense, and composites) used in iltration processes, charged membranes, or those presenting working modiications. The membranes used to demonstrate the potentiality of IS are: 1. A commercial composite polyamide/polysulfone membrane for reverse osmosis (HR95) from DSS (Denmark). The characteristic membrane parameters, such as total thickness, hydraulic permeability, and rejection, are: Δxm = (165 ± 5) μm, L p = 8.5 × 10−12 m/(sec Pa), and σ = 99.5%, respectively (Jonsson and Benavente 1992).

Membrane Modification: Technology and Applications

28

2. Two experimental composite membranes consisting of a sulfonated-polysulfone support and a polyamide–polyethylene glycol top layer (PS/PA–PEG). Samples with two different PEG concentrations, 5 and 25 wt%, are characterized. These membranes (PS/PA–PEG5 and PS/PA–PEG25, respectively) were obtained and kindly submitted by Dr. X. Zhang and Professor R. García-Valls (Department Ingeniería Química, Universidad Rovira y Virgili, Tarragona, Spain), and the membrane preparation is extensively explained by Benavente et al. (2005). 3. A regenerated cellulose/polypropylene supported ultrailtration RC70PP membrane with 10 kDa cutoff (DSS) (Pelaez et al. 2010); a Tefzel ethylene tetraluoroethylene (ETFE) laser-perforated ilm (LZ200 from DuPont Fluoropolymers, Detroit, MI) with 8 × 8 holes of 150 μm pore radius and a center-to-center distance of 400 μm (González-Pérez et al. 2009); and a composite structure formed with the RC70PP membrane plus the ETFE perforated ilm. The interest of this system is double, because as a layered system it allows the possibility of individual layers and complete structure measurement to check the series association or to see possible interfacial/material effects that affect the total structure; moreover, the ETFE/RC70PP composite structure studied was considered as a possible encapsulation system for biomimetic membranes (Vogel et al. 2009). 4. Two experimental poly(ether ether ketone) (PEEK) membranes for methanol fuel cell application, one with a zirconium modiication to reduce the methanol crossover (samples PEEK-sn-0 and PEEK-sn-Zr, respectively), which were prepared and kindly submitted by Dr. S. Nunes (GKSS, Germany) (Nunes et al. 2002; Silva et al. 2006). To observe the electrical changes only due to membrane material modiication, impedance measurements with these membranes were performed both in contact with the NaCl solutions and in a dry state. 5. A lat, lexible, and symmetric commercial ceramic membrane by Degussa (Germany), with a composite structure formed by a ibrous stainless steel network covered by a sublayer of Al2O3 particles plus an external layer of ZrO2, with a pore size of 0.1 μm and 80–100 μm thickness, according to the supplier (Augustin et al. 2002). Changes due to membrane fouling as a result of protein (BSA) iltration were determined and correlated with a low reduction and the SP results (de Lara and Benavente 2009). These samples will be named Z100S (clean membrane) and Z100S+BSA (fouled membrane).

2.3.2

TYPICAL IMPEDANCE SPECTROSCOPY EXPERIMENTAL SYSTEM

We now discuss how IS measurements are made in the laboratory. IS measurements are performed with an impedance analyzer (Solartron 1260, United Kingdom), which is controlled by a computer and uses Ag/AgCl or Pt electrodes. The experimental data are corrected by software as well as the inluence of connecting cables

Use of Impedance Spectroscopy for Characterization of Modified Membranes 29

and other parasite capacitances. The measurements are carried out using 100 different frequencies in the range 10–107 Hz, at a maximum voltage of 0.01 V and a room temperature of (25 ± 2)°C. An electrochemical test cell similar to that described in Pelaez et al. (2010) is used for the IS measurements in “working conditions,” meaning that the membranes are placed between the two half-cells in contact with NaCl aqueous solutions at the same concentration (electrode/solution (c)/membrane/solution (c)/electrode system); different NaCl concentrations between 0.001 and 0.05 M are typically measured. Before use, the membranes are maintained in contact with a solution of the studied concentration over a certain time (10 ≤ t(h) ≤ 24), depending on the membrane structure. In the case of dry membranes, the test cell consists of a Telon structure on which two Pt electrodes are placed and screwed down (system: electrode/membrane/ electrode) (Ramos et al. 2010). We will now discuss, in turn, the IS of these different membranes.

2.4

RESULTS AND DISCUSSION

2.4.1

REVERSE OSMOSIS MEMBRANE

The impedance plots for the composite HR95 membrane are shown in Figure 2.3, where two different contributions associated with the membrane (m) and the electrolyte solution between the electrodes and the membrane surfaces (e) can clearly be observed. To check this assumption, the impedance data obtained with the electrolyte alone, without any membrane in the measuring cell, are also plotted in Figure 2.3. As can be observed, a parallel RC circuit with only a relaxation process and a maximum frequency of around 106 Hz (similar to that in Figure 2.1) was obtained for the electrolyte solution measured alone (ReCe). The circuit associated with the composite polyamide/polysulfone HR95 membrane shows two subcircuits,

80,000 Electrolyte alone: (ReCe) Electrolyte HR95 membrane (ReCe)----- (RpWp)-----(RdCd)

40,000

10,000 –Zimg (Ω)

–Zimg (Ω)

60,000

m

e

1,000

20,000 (ReCe)

0 0 (a)

20,000 40,000 60,000 80,000 Zreal (Ω)

100 (b)

101

103

105

107

f (Hz)

FIGURE 2.3 (a) Nyquist and (b) Bode plots for the electrode/0.002 M NaCl solution/HR95 membrane/0.002 M NaCl solution/electrode (•) and the electrode/0.002 M NaCl solution/ electrode (×) systems.

30

Membrane Modification: Technology and Applications

one for each layer, plus the electrolyte contribution, that is: (i) A parallel association of a resistance and a capacitor for the dense active layer (RdCd); (ii) A resistor in parallel with a Warburg impedance for the porous sublayer (RpWp); and (iii) the parallel RC circuit associated with the electrolyte solution placed between the membrane and the electrodes, which hardly differs from that obtained with the electrolyte alone. This point is also observed in Figure 2.3b, where the relaxation process associated with the membrane appears at a lower frequency ( fmax ≈ 5000 Hz), in agreement with the value indicated for the dense polyamide membrane (Figure 2.2b), but a certain asymmetry can be observed in Figure 2.3, which is attributed to the porous sublayer. The data shown in Figure 2.3a were itted using a nonlinear program to determine the values of the different circuit elements, and the variation of Rp and Rd with the NaCl concentration is shown in Figure 2.4a. The decrease in electrical resistance with the increase in salt concentration is due to the concentration dependence of the electrolyte embedded in the dense layer matrix (Rd) or illing the voids of the support (Rp). However, the capacitance for the dense membrane layer has little dependence on the salt concentration, as can be observed in Figure 2.4b, and an average value of Cd = (1.5 ± 0.3) × 10 −9 F for the whole range of concentrations was obtained, while the porous sublayer Warburg impedance (a diffusion-related parameter) slightly increases with the increase in the NaCl concentration. It should be noted for simplicity reasons that a two-layer model has been assumed for the HR95 reverse osmosis membrane, but a more complex structure, including an “intermediate layer” with gradual changes in the pore radii/porosity from one layer to another (three-layer model), could be more realistic (Zholkovskij 1995). In this context, the compaction or partial inclusion of the intermediate layer due to membrane aging determined by IS measurements for nanoiltration membranes shows the utility of this technique for membrane modiication characterization (Benavente and Vázquez 2004).

40,000

20,000

(a)

0 0.00 0.01 0.02 0.03 0.04 0.05 CNaCl (M)

1E – 7 C (F), W (Ω)

Ri (Ω)

60,000

1E – 8

1E – 9

1E – 10 (b)

0.00 0.01 0.02 0.03 0.04 0.05 CNaCl (M)

FIGURE 2.4 Concentration dependence for the (a) electrical resistance of the porous (Δ) and dense (▲) layers of the HR95 membrane; (b) dense layer capacitance (▲), porous layer Warburg impedance (◻), and porous layer equivalent capacitance (Δ) calculated using Equation 2.5.

Use of Impedance Spectroscopy for Characterization of Modified Membranes 31

2.4.2 POLYSULFONE–POLYAMIDE/POLYETHYLENE GLYCOL MEMBRANE: EFFECT OF PEG CONTENT The EIS characterization of the polysulfone–polyamide/PEG membranes was carried out not only to determine separately the contribution of the porous support and the PEG-modiied top layer, but also to correlate the electrical changes with the PEG content. Figure 2.5a shows an SEM micrograph of the cross section of the PS/PA– PEG membrane; the porous polysulfone structure and the dense polyamide top layer where the PEG is mainly located can be clearly observed in this igure (Benavente et al. 2005). Impedance plots (Nyquist and Bode plots) for the PS/PA–PEG5 and PS/PA– PEG25 membranes are shown in Figure 2.5b and 2.5c, where the effect of the asymmetric structure on the impedance (Nyquist) plot is indicated, but the differences depending on the PEG concentration are also evident. The equivalent circuit for the total membrane system, (ReCe) − (RmQm), is also indicated in Figure 2.5b; the depressed semicircle, attributed to the nonconstant phase circuit element (Qm), is due to the porous structure of these membranes and the mixture of the relaxation times associated with their electrical response (polymeric matrix and solution). The analysis of the impedance plots also allows the determination of Rm for the different concentrations studied. Figure 2.6a shows the decrease in the membrane electrical resistance with the increase in the salt concentration, where the differences Dense layer Porous sublayer

(a)

8,000

p.l. 4,000

d.l. 0 20,000 24,000 28,000 32,000

Zreal (Ω)

10,000

–Zimg (Ω)

20,000

–Zimg (Ω)

(ReCe)--(RmQm)

–Zimg (Ω)

12,000

12,000

30,000

8,000

e

4,000

e

0 0 (b)

m m

10,000 20,000 Zreal (Ω)

0

30,000 (c)

101

103

105

107

f (Hz)

FIGURE 2.5 (a) Cross section SEM micrograph of the polyethylene glycol–sulfonated polysulfone–supported membrane. (b) Nyquist and (c) Bode plots for the electrode/NaCl solution/membrane/NaCl solution/electrode system. (◻) PS/PA–PEG5 membrane and (Δ) PS/PA–PEG25 membrane.

32

Membrane Modification: Technology and Applications 105

10,000 Electrolyte concentration: 0.01 M

103 102

101 0.00 0.01 0.02 0.03 0.04 0.05 (a) c (M)

Rm Sm (Ω m2)

Rm (Ω)

104 1,000

100

10

0

(b)

10

20 30 40 50 60 PEG concentration (%)

70

FIGURE 2.6 (a) Membrane electrical resistance vs. NaCl concentration for (◻) PS/PA– PEG5 membrane and (Δ) PS/PA–PEG25 membrane. (b) Membrane electrical resistance as a function of the PEG concentration in the casting solution for c = 0.01 M NaCl.

depending on the PEG concentration in the membrane casting solution can be observed. A more complete study of the effect of PEG on the electrical response of the membrane was already performed (Benavente et al. 2005) by EIS measurements at other PEG concentrations (5, 40, and 60 wt%); the variation of the electrical resistance with the PEG concentration for a given NaCl solution (0.01 M) is shown in Figure 2.6b. As can be observed, the increase in the PEG content causes a decrease in the electrical resistance, and a PEG concentration of around 40 wt% seems to be the optimum content in the case of low electrical resistance and high conductivity application requirements.

2.4.3 STUDY OF COMPOSITE STRUCTURES: EFFECT OF SUBLAYERS CHEMICAL NATURE Recently, the interest in mass and charge/ions transport across composite structures has increased due to new applications in biosensors and drug release and biomimetic devices (Vogel et al. 2009). Multilayer structures formed by a series association of speciic elements are being developed; however, sublayer chemical compatibility and/or hydrophobicity can signiicantly modify the transport across the whole structure, which might be of signiicant importance depending on its particular application. As already indicated, the composite or multilayer structure studied consists of a regenerated cellulose/polypropylene ultrailtration membrane (RC70PP) and a laser-perforated hydrophobic ETFE ilm (sample RC70PP+ETFE) placed in the middle of two NaCl solutions of the same concentration (electrolyte/sample/electrolyte system). In addition, IS measurements of each individual layer, the ultrailtration RC70PP membrane and the hydrophobic ETFE ilm, were performed. Figure 2.7 shows the impedance plots for the different samples studied, as well as for the electrolyte solution alone. As can be observed, both the RC70PP membrane and the ETFE ilm show a unique depressed semicircle, a distribution of the relaxation times due to the electrolyte and the solid structure. However, two separated

Use of Impedance Spectroscopy for Characterization of Modified Membranes 33 20,000

40,000

Electrolyte/ETFE/electrolyte: (RfeQfe) Electrolyte/ETFE + RC70PP/electrolyte: (ReCe) – (RcsCcs) Electrolyte: (ReCe)

15,000 –Zimg (Ω)

–Zimg (Ω)

60,000

Electrolyte/RC70PP/electrolyte: (RmeQme)

20,000

5,000

0 0 (a)

10,000

20,000 40,000 Zreal (Ω)

0 100 101 102 103 104 105 106 107

60,000 (b)

f (Hz)

FIGURE 2.7 Impedance plots for the RC70PP+ETFE ilm (∙), the RC70 membrane (◻), the ETFE ilm (▿), and the electrolyte solution (×). (a) Nyquist plot (−Zimg vs. Zreal) and (b) Bode plot (−Zimg vs. frequency).

semicircles were obtained for the electrolyte/RC70PP+ETFE/electrolyte system, one associated with the composite structure and the other with the electrolyte between the electrodes and the sample surfaces, (ReCe) − (RcsCcs) circuit. The differences found when the individual layers or the composite structures were measured are attributed to the hydrophobic character of the ETFE ilm, which is more signiicant when the RC70PP membrane partially isolates one of its surfaces from the aqueous NaCl solution and, consequently, it allows its separation from the external solution. These results are a clear example of the strong inluence that the material layer may have on the electrical response of a composite layered structure. As already indicated, the electrical parameters for the different samples at the studied NaCl concentrations can be determined by nonlinear analysis of the data shown in Figure 2.7a; in the case of the membrane RC70PP and the ETFE ilm, these results also include the electrolyte contribution, but individual electrical resistance values, Rm or Rf, respectively, can be determined by subtracting those obtained for the electrolyte solution (Re) at the same concentration.

2.4.4

EFFECT OF MEMBRANE MATERIAL MODIFICATION

The effect of membrane material modiication on the impedance plots and the electrical parameters was studied by comparing the results obtained with the PEEK-sn-0 poly(ether ether ketone) membrane and the zirconium-modiied PEEK-sn-Zr sample. Figure 2.8 shows the impedance plots for both membranes in contact with an NaCl solution and the electrolyte solution alone; differences in the Nyquist plot (Figure 2.8a) as a result of the zirconium modiication can be observed, although a unique relaxation process for the electrolyte/membrane system was obtained with both samples. However, only slight differences in the interfacial region are observed in the Bode plot (Figure 2.8b) as a result of the membrane presence when compared with the electrolyte data (between 200 and 2000 Hz), indicating that no charge adsorption effect can be associated with the membrane samples.

34

Membrane Modification: Technology and Applications 30,000 10,000 –Zimg (Ω)

–Zimg (Ω)

(RsmCsm) 20,000

10,000

1,000

100 10 100

0 0 (a)

10000 20000 Zreal (Ω)

30000 (b)

102

104 f (Hz)

106

FIGURE 2.8 Nyquist (a) and Bode (b) plots for the membranes PEEK-sn-0 (⚬) and PEEKsn-Zr (•) in contact with an NaCl solution (×).

The analysis of the impedance data allows the determination of the electrical resistance, Rsm (membrane and solution contributions), of the membrane systems and their variation with the solution concentration is shown in Figure 2.9a, as well as that obtained for the NaCl solution. Differences between both membranes at low concentrations were obtained; however, at a high concentration, the Rsm values for both membranes hardly differ from the solution resistance, which seems to present the higher contribution to the total resistance of the system. As indicated above, the membrane electrical resistance was determined by subtracting the solution resistance (Rm = Rms − Rs), and the membrane conductivity (σm) was determined from Rm values using Equation 2.6. Figure 2.9b shows the variation of σm for both membranes as a function of the external solution concentration (embedded in the membrane structure). The values in Figure 2.9b show the strong effect of the inorganic iller on the conductivity of the membrane at moderate salt concentrations, which might be 0.005

50,000

0.004

30,000 20,000 10,000

(a)

0 0.000 0.002 0.004 0.006 0.008 0.010 cNaCl (M)

σm (Ω m)–1

Rsystem (Ω)

40,000

0.003 0.002 0.001

0.000 0.000 0.002 0.004 0.006 0.008 0.010 cNaCl (M) (b)

FIGURE 2.9 (a) Variation of the electrical resistance of the membrane system with salt concentration: (⚬) PEEK-sn-0, (•) PEEK-sn-Zr, and (×) NaCl solution. (b) Variation of membrane conductivity σm with the NaCl concentration: (o) PEEK-sn-0 and (•) PEEK-sn-Zr.

Use of Impedance Spectroscopy for Characterization of Modified Membranes 35 800,000 107 –Zimg(Ω)

–Zimg(Ω)

600,000 400,000

103

200,000

101

0 0 (a)

105

101

200,000 400,000 600,000 800,000 Zreal(Ω)

(b)

103

105

107

f (Hz)

FIGURE 2.10 Nyquist (a) and Bode (b) plots for dry PEEK-sn-0 (♢) and PEEK-sn-Zr (♦) membranes.

associated with a reduction of the PEEK membrane free volume, according to the lower water (or solution) content (W = (mwet − mdry)/mdry) of the inorganic-modiied membrane: WPEEK-sn-0 = 28% and WPEEK-sn-Zr = 19%. This undesired result for membrane application in direct methanol fuel cells (DMFCs) could be partially balanced by a reduction in the methanol permeability due to its denser structure, which is the main reason for this kind of membrane modiication. However, differences uniquely related to the membrane material are not always clearly relected when the solution/membrane systems are studied if high hydrophilic membranes are considered, as was previously stated. For that reason, measurements with dry PEEK membranes were also performed and the Nyquist and Bode plots obtained for each sample are shown in Figure 2.10, where signiicant differences in the electrical response of both membranes can be observed. The analysis of the impedance curves shown in Figure 2.10 allows the determination of the electrical resistance for each membrane, and the following values for the conductivity and the dielectric constant were determined using Equations 2.6 and 2.7: PEEK-sn-0 dry: σ m = 2.6 × 10 −6 (1 ohm m ) , ε m = 13.4, PEEK-sn-Zr dry: σ m = 2.1 × 10 −8 (1 ohm m ) , ε m = 4.0. According to these results, the zirconium inclusion reduces the membrane conductivity by two orders of magnitude, which could also be due to a decrease in the water adsorption of this hybrid material, with a reduction of the hydrophilic character, in agreement with its lower permittivity.

2.4.5

EFFECT OF PROTEIN FOULING ON EIS PLOTS

Membrane fouling is one of most important problems for membrane processes related to iltration for food applications, since it can dramatically reduce the low

36

Membrane Modification: Technology and Applications 25,000

10,000 Z100S: (RsmCsm) 7,500

Z100S + BSA: (ReCe) – (RmCm)

–Zimg(Ω)

–Zimg(Ω)

20,000 15,000 10,000

5,000 e

m 2,500

50,00

m m

0 0 (a)

0 100

5,000 10,000 15,000 20,000 25,000 Zreal(Ω)

102

(b)

104

106

f (Hz)

FIGURE 2.11 (a) Nyquist and (b) Bode plots for clean Z100S (∇) and protein-fouled Z100S+BSA (▼) membranes.

and, consequently, the beneit of the process. The EIS technique can also be used for membrane fouling characterization. Figure 2.11 shows a comparison of the impedance plots obtained for clean Z100S and protein-fouled Z100S+BSA membranes in contact with the same NaCl solution. As can be observed, a unique semicircle was obtained for the Z100S membrane with a maximum frequency at 2.5 MHz, which indicates that only one relaxation process for the whole membrane/electrolyte system exists. However, when itted to a parallel RC equivalent circuit (RsmCsm), which includes the contribution of both membrane and electrolyte solution, two different contributions (semicircles) can be observed for the Z100S+BSA samples, one associated with the fouled membrane and the other associated with the electrolyte between the electrodes and the membrane surface; the equivalent circuit for the whole system is (ReCe) − (RmCm). Figure 2.12a shows the variation of the membrane resistance with the NaCl concentration. The differences between both samples are mainly attributed to the 10.0

12,000

7.5 5.0

6,000

2.5 0.0

3,000 0 0.000 (a)

BSA isoelectric point

ζ (mV)

Rm (Ω)

9,000

–2.5 –5.0 0.005

0.010 0.015 cNaCl (M)

0.020

2 (b)

4

6 pH

8

10

FIGURE 2.12 (a) Membrane electrical resistance as a function of the salt concentration. (b) The ζ-potential as a function of the solution pH (c = 0.001 M NaCl). Membrane Z100S (⚬) and membrane Z100S+BSA (•).

Use of Impedance Spectroscopy for Characterization of Modified Membranes 37

internal protein fouling, which causes a pore reduction due to BSA deposition on the pore wall and, consequently, a decrease in the membrane free volume. This fact indicates an increase in the electrical resistance, as shown in Figure 2.12a, for the different NaCl solutions studied. These results show an average increase of (10 ± 1)% in the membrane electrical resistance (at a given NaCl concentration) associated with the pore wall reduction. BSA deposition might also exist, but it has a minimal effect on the electrical parameters (de Lara and Benavente 2009). To conirm the hypothesis of the presence of BSA on the pore wall, measurements of the SP through the pores of both the Z100S and Z100S+BSA membranes were performed with a 0.001 M NaCl solution at different pHs (de Lara and Benavente 2009); Figure 2.12b shows a comparison of the results obtained with both membranes. Positive ζ-potential values were obtained for the Z100S membrane in agreement with the electropositive character of both Al2O3 and ZrO2, but the ζ-potential/ pH dependence found for the Z100S+BSA-fouled membrane is completely different, with a zero ζ-potential value at a pH of ~5, that is, the isoelectric point obtained for the fouled membrane hardly differs from that for the protein (BSA isoelectric point ~4.9; Nyström and Järvinen 1991), which is a clear indication of protein deposition on the membrane pore wall and a conirmation of the EIS results. These results further conirm the use of EIS, a nondestructive technique, to study membrane modiications caused by fouling.

2.5

CONCLUSION

This chapter shows the possibility of using IS, a nondestructive ac technique, for determining membrane modiications. Changes in the membrane structure and/or material due to fouling, adsorption/deposition of particles, aging, or material alteration may cause modiications in the membrane electrical parameters and, consequently, in the impedance data and in the shape of the impedance curves. IS measurements have been made with the membranes in contact with the electrolyte solutions or membrane “working process condition” (NaCl solutions ranging between 0.001 and 0.05 M), and using 100 data points to cover a range of frequencies between 1 and 107 Hz, which provides information on the processes related to the interface, the external electrolyte solution layer, and the bulk membrane by analyzing the impedance plots and using equivalent circuits as models. The concentration dependence of the membrane electrical resistance was also studied. In addition, IS measurements for the dry samples were performed to obtain more information on the changes associated with membrane material modiication. IS measurements were performed to determine the membrane variations associated with: (i) Dense and porous layers of a commercial RO membrane; (ii) Different PEG concentrations in the top dense layer of a polyamide/polysulfone experimental membrane; (iii) Hydrophobic character of one layer in a composite or multilayer structure; (iv) Membrane matrix material modiication; and (v) Protein (BSA) fouling of a porous commercial membrane. The results obtained with other characterization techniques, such as morphological, chemical, and adsorption analyses, have validated the information obtained from the IS results.

38

Membrane Modification: Technology and Applications

ACKNOWLEDGMENT Many thanks to Dr. S. Nunes for providing the PEEK-sn-0 and PEEK-sn-Zr membranes, and the CICYT (Spain) for partial inancial support during the performance of some measurements (Project MAT2007/65065).

REFERENCES Asaka, K. 1990. Dielectric properties of cellulose acetate reverse osmosis membranes in aqueous solutions. J. Memb. Sci. 50: 71–84. Augustin, S., Hennige, V., Hörpel, G. and Hying, Ch. 2002. Ceramic but lexible: New ceramic membrane foils for fuel cells and batteries. Desalination 146: 23–28. Benavente, J. and Férnandez-Pineda, C. 1985. Electrokinetic phenomena in porous membranes: Determination of phenomenological coeficients and transport numbers. J. Memb. Sci. 23: 121–136. Benavente, J. and Jonsson, G. 1998. Effect of adsorbed protein on the hydraulic permeability, membrane and streaming potential values measured across a microporous membrane. Colloid Surface A 138: 255–264. Benavente, J. and Vázquez, M.I. 2004. Effect of age and chemical treatments on characteristic parameters for active and porous sublayers of polymeric composite membranes. J. Colloid Interface Sci. 273: 547–554. Benavente, J., Hernández, A. and Jonsson, G. 1993. Proper and adsorbed charge on the surfaces of the polysulphonic support of a composite membrane from electrokinetic phenomena. J. Memb. Sci. 80: 285–296. Benavente, J., de Abajo, J., de la Campa, J.G. and García, J.M. 1997. Electrical properties of modiied aromatic polyamide membranes. Sep. Sci. Technol. 32: 2189–2199. Benavente, J., Zhang, X. and García-Valls, R. 2005. Modiication of polysulfone membranes with polyethylene glycol and lignosulfate: Electrical characterization by impedance spectroscopy measurements. J. Colloids Interface Sci. 285: 273–280. Benavente, J., Silva, V., Prádanos, P., Palacio, L., Hernandez, A. and Jonsson, G. 2010a. A comparison of the volume charge density of nanoiltration membranes obtained from retention and conductivity experiments. Langmuir 26: 11841–11849. Benavente, J., Vázquez, M.I., Hierrezuelo, J., Rico, R., López-Romero, J.M. and LópezRamirez, M.R. 2010b. Modiication of a regenerated cellulose membrane with lipid nanoparticles and layers. Nanoparticle preparation, morphological and physicochemical characterization of nanoparticles and modiied membranes. J. Memb. Sci. 355: 45–52. Buck, R.P. and Ciani, S. 1975. Electroanalytical chemistry of membranes. Crit. Rev. Anal. Chem. 5: 323. Cañas, A., Ariza, M.J. and Benavente, J. 2001. Characterization of active and porous sublayers of a composite reverse osmosis membrane by impedance spectroscopy, streaming and membrane potentials, salt diffusion and X-ray photoelectron spectroscopy. J. Memb. Sci. 183: 135–146. Childress, A.E. and Elimelech, M. 1996. Effect of solution chemistry on the surface charge of polymeric reverse osmosis and nanoiltration membranes. J. Memb. Sci. 119: 253–268. Compañ, V., Sorensen, T.S., Andrio, A., López, L. and de Abajo, J. 1997. Transport numbers from initial time and stationary state measurements of EMF in non-ionic polysulphonic membranes. J. Memb. Sci. 123: 293–302. Coster, H.G.L., Kim, K.J., Dahlan, K., Smith, J.R. and Fell, C.J.D. 1992. Characterization of ultrailtration membranes by impedance spectroscopy. I. Determination of the separate electrical parameters and porosity of the skin and sublayers. J. Memb. Sci. 66: 19–26.

Use of Impedance Spectroscopy for Characterization of Modified Membranes 39 de Lara, R. and Benavente, J. 2007. Electrokinetic and surface chemical characterizations of an irradiated microiltration polysulfone membrane: Comparison of two irradiation doses. J. Colloid Interface Sci. 310: 519–525. de Lara, R., and Benavente, J. 2009. Use of hydrodynamic and electrical measurements to determine protein fouling mechanisms for microiltration membranes with different structures and materials. Sep. Purif. Technol. 66: 517–524. Demish, H.-U. and Pusch, W. 1980. Electrical and electrosomotic transport behaviour of asymmetric cellulose acetate membranes. II. Transport behaviour in hyperiltration experiments. J. Colloid Interface Sci. 76: 464–477. Fievet, P., Sbaï M., Szymczyk, A. and Vidonne, A. 2003. Determining the z-potential of plane membranes from tangential streaming potential measurements: Effect of the membrane body conductance. J. Memb. Sci. 226: 27–36. Freger, V. and Bason, S. 2007. Characterization of ion transport in thin ilms using impedance spectroscopy: I. Principles and theory. J. Memb. Sci. 302: 1–9. González-Pérez, A., Stibius, K.B., Vissing, T., Nielsen, C.H. and Mouritsen, O.G. 2009. Biomimetic triblock copolymer membrane arrays: A stable template for functional membrane proteins. Langmuir 25: 10447–10450. Hamdy, A.S., El-Shenawy, E. and El-Bitar, T. 2006. Electrochemical impedance spectroscopy study of the corrosion behaviour of some Niobium bearing stainless in 3.5% NaCl. Int. J. Electrochem. Sci. 1: 171–180. Helfferich, F.G. 1962. Ion Exchange. McGraw-Hill: New York. Jonscher, A.K. 1983. Dielectric Relaxation in Solid. Chelsea Dielectric Press: London. Jonsson, G. and Benavente, J. 1992. Determination of some transport coeficients for the skin and porous layer of a composite membrane. J. Memb. Sci. 69: 29–42. Kimura, Y., Lim, H.-J. and Iijima, T. 1984. Membrane potential of charged cellulosic membranes. J. Memb. Sci. 18: 285–293. Laksminarayanaiah, N. 1969. Transport Phenomena in Membranes. Academic Press: New York. Li, Y. and Zhao, K. 2004. Dielectric analysis of nanoiltration membranes in electrolyte solutions: Inluence of electrolyte concentration and species on membrane permeation. J. Colloid Interface Sci. 276: 68–76. López, M.L., Compañ, V., Garrido, J., Riande, E. and Acosta, J.L. 2001. Proton transport in membranes prepared from sulfonated polystyrene-poly (vinyliluoride) blends. J. Electrochem. Soc. 148: E372–E377. Macdonald, J.R. 1987. Impedance Spectroscopy. John Wiley: New York. Mälmgren-Hansen, B., Sörensen, T.S., Jensen, J.B. and Hennerberg, M. 1989. Electric impedance spectroscopy of cellulose acetate membranes and a composite membrane at different salt concentrations. J. Colloid Interface Sci. 130: 359–370. Mijovic, J. and Bellucci, F. 1996. Impedance spectroscopy of reactive polymers. Trends Polym. Sci. 4: 74–82. Molina, C., Victoria, L., Arenas, A. and Ibañez, J.A. 1999. Streaming potential and surface charge density of microporous membranes with pore diameter in the range of thickness. J. Memb. Sci. 154: 239–255. Nunes, S.P., Ruffmann, B., Rikowski, E., Vetter, S. and Richau, K. 2002. Inorganic modiication of proton conductive polymer membranes for direct methanol fuel cells. J. Memb. Sci. 203: 215–225. Nyström, M. and Järvinen, P. 1991. Modiication of polysulfone membranes with UV irradiation and hydrophilicity increasing agents. J. Memb. Sci. 60: 275–296. Oleinikova, M., Muñoz, M., Benavente, J. and Valiente, M. 2000. Evaluation of structural properties of novel activated composite membranes containing organophosphorous extractants as carriers. Langmuir 16: 716–721. Pelaez, L., Vázquez, M.I. and Benavente, J. 2010. Interfacial and fouling effects on diffusional permeability across a composite ceramic membrane. Ceram. Int. 36: 797–801.

40

Membrane Modification: Technology and Applications

Pointié, M. 1999. Effect of aging on UF membranes by a streaming potential (SP) method. J. Memb. Sci. 154: 213–220. Ramos, J.D., Milano, C., Romero, V., Escalera, S., Alba, M.C., Vázquez, M.I. and Benavente, J. 2010. Water effect on physical-chemical and elastic parameters for a dense cellulose regenerated membranes. Transport of different aqueous electrolyte solutions. J. Memb. Sci. 352: 153. Romero, V., Pelaez, L., Vázquez, M.I. and Benavente, J. 2011. Electrochemical characterization of an ultrailtration regenerated cellulose/polypropylene supported membrane. Desalination Water Treat. 27: 1–8. Silva, V.S., Ruffmann, B., Silva, H., Silva, V.B., Mendes, A., Madeira, L.M. and Nunes, S. 2006. Zirconium oxide hybrid membranes for direct methanol fuel cells—Evaluation of transport properties. J. Memb. Sci. 284: 137–144. Torras, C., Zhang, X., García-Valls, R. and Benavente, J. 2007. Morphological, chemical surface and electrical characterizations of lignosulfonate-modiied membranes. J. Memb. Sci. 297: 130–137. Vallejo, E., Pourcelly, G., Gavach, C., Mercier, R. and Pineri, M. 1999. Sulfonated polyimides as proton conductor exchange membranes. Physicochemical properties and separation H+/Mz+ by electrodialysis comparison with a perluorosulfonic membrane. J. Memb. Sci. 160: 127–137. Vázquez, M.I., de Lara, R. and Benavente, J. 2008. Chemical surface, diffusional, electrical and elastic characterizations of two different dense regenerated cellulose membranes. J. Colloid Interface Sci. 328: 331–339. Vogel, J., Perry, M.E., Hansen, J.S., Bollinger, P.-Y., Nielsen, C.H. and Geschke, O. 2009. Support structure for biomimetic applications. J. Micromech. Microeng. 19: 25–32. Zholkovskij, E. 1995. Irreversible thermodynamic and impedance spectroscopy of multilayer membranes. J. Colloid Interface Sci. 169: 267–283.

3

Reduction of Membrane Fouling by Polymer Surface Modification Victor Kochkodan

CONTENTS 3.1 3.2

Introduction .................................................................................................... 42 Surface Membrane Properties Affecting Membrane Fouling ........................ 43 3.2.1 Surface Hydrophilicity ....................................................................... 43 3.2.2 Surface Charge ...................................................................................44 3.2.3 Surface Roughness .............................................................................44 3.3 Development of Low-Fouling Polymer Membranes via Photoinitiated Grafting ............................................................................. 45 3.3.1 UV-Initiated “Grafting-To” the Membrane Surface ........................... 47 3.3.2 UV-Initiated “Grafting-From” the Membrane Surface ...................... 47 3.3.3 Membrane Modiication via “Grafting-From” Method Without the Use of a Photoinitiator .................................................... 47 3.3.4 Membrane Modiication via “Grafting-From” Approach with a Photoinitiator ........................................................................... 50 3.4 Miscellaneous Grafting Methods on the Membrane Surface ........................ 54 3.5 Plasma Treatment of Polymer Membranes ..................................................... 55 3.5.1 Plasma of Nonpolymerizable Gases ................................................... 55 3.5.2 Plasma of Polymerizable Molecules ................................................... 56 3.5.3 Plasma-Induced Grafting of Membrane Surface ................................ 57 3.6 Physical Coating/Adsorption on the Membrane Surface................................ 58 3.6.1 Coating via Casting ............................................................................ 58 3.6.2 Coating via Adsorption .......................................................................60 3.6.3 Coating via Filtration .......................................................................... 61 3.7 Chemical Reactions on the Membrane Surface for Fouling Reduction ......... 62 3.8 Surface Modiication of Polymer Membranes with Nanoparticles ................ 65 3.9 Conclusion ...................................................................................................... 69 References ................................................................................................................ 70

41

42

3.1

Membrane Modification: Technology and Applications

INTRODUCTION

Currently, the pressure-driven membrane processes, reverse osmosis (RO), nanoiltration (NF), ultrailtration (UF), and microiltration (MF), are widely used in water treatment, biotechnology, food industry, medicine, and other ields (Baker 2004). However, one of the main problems arising from the operation of the membrane units is membrane fouling, which seriously hampers the applications of membrane technologies (Scot and Hughes 1996). Membrane fouling is an extremely complex phenomenon that has not been deined precisely yet. In general, the term is used to describe the undesirable deposition of retained particles, colloids, macromolecules, and salts at the membrane surface or inside the pores. Depending on the membrane process and the chemical nature of the foulants, several types of fouling can occur in membrane systems, including inorganic fouling or scaling, particulate/colloidal fouling, organic fouling, and biofouling (Kimura et al. 2004; Flemming 1997). Inorganic fouling or scale formation at the membrane surface results from the increased concentration of one or more inorganic salts beyond their solubility limits and their ultimate precipitation on the membranes (Van de Lisdonk et al. 2000). Scaling usually refers to the formation of deposits of inverse solubility salts, such as CaCO3, CaSO4 · xH2O, and Ca3(PO4)2. The species with the greatest scaling potential in NF and RO are CaCO3 and CaSO4 · 2H2O, while the other potential scaling compounds are BaSO4, SrSO4, Ca3(PO4)2, and Fe(OH)3 (Hasson et al. 2001). Colloid fouling refers to membrane fouling with colloidal and suspended particles in the size range of a few nanometers to a few micrometers. Examples of common colloidal-sized foulants include clays, silica salts, and hydroxides of heavy metals (Potts et al. 1981). An increased concentration of the rejected ions at the front of the membrane surface facilitates the aggregation of dissolved organic substances, for example, natural organic matter (NOM), into colloidal-sized particles (Hong and Elimelech 1997). With organic fouling, organic compounds in water, such as proteins, humic substances, and polysaccharides, have been implicated as strong, irreversible membrane foulants in pressure-driven membrane processes (Escobar et al. 2005). Proteins are the main fouling components when membranes are used in medicine, biotechnology, and food industry. NOM is the main organic foulant in the membrane treatment of surface waters, brackish waters, and seawater (Roudman and DiGiano 2000; Al-Amoudi and Farooque 2005). It has been shown that the hydrophobic fraction of NOM, due to the strong adsorption on the membrane surface, was the major factor causing permeate lux decline, while the hydrophilic fraction of NOM had a relatively small effect on the membrane fouling (Nilson and DiGiano 1996). Biofouling is a term used to describe all instances of fouling where biologically active organisms are involved (Flemming et al. 1997). Biofouling is a dynamic process of microbial colonization and growth, which results in the formation of microbial bioilms on the membrane surface. Usually, membrane biofouling is initiated by the irreversible adhesion of one or more types of bacteria to the membrane surface followed by fast growth and multiplication of the sessile cells in the presence of feedwater nutrients. Under favorable conditions, the number of bacterial cells obtained

Reduction of Membrane Fouling by Polymer Surface Modification

43

after duplication of a single bacterium may reach 4.7 × 1021 in 24 h (Bharwadal et al. 2000). With time, the initial sessile microbial population can eventually form a conluent lawn of bacteria, a bioilm, on the membrane surface. Membrane biofouling is caused mainly by Corynebacterium, Pseudomonas, Bacillus, Arthrobacter, Flavobacterium, and Aeromonas bacterial species and to a lesser degree by fungi such as Penicillium, Trichoderma, and other eukaryotic microorganisms (Baker and Dudley 1998). Membrane fouling, as well as its prevention, has been the subject of many studies since the early 1960s when industrial membrane separation processes emerged. Membrane fouling can be controlled by the selection of an appropriate membrane, adjustment of the operating conditions in a membrane element, including hydrodynamics and operating pressure, and appropriate pretreatment of the feed solution (Ridgway et al. 1984; Vrouwenvelder et al. 1998; Sheikholeslami 1999). However, very often these actions are not suficient to cope with fouling. As fouling progresses, the membrane lux sharply declines; higher operating pressures and thus more energy must be expended to achieve the desired throughput. Usually, membrane cleaning is applied to remove the foulants and restore the membrane lux; however, in many cases, the fouling is irreversible and the membrane elements must be replaced. Currently, there is a consensus in research that membrane fouling with organic compounds and microorganisms is mainly determined by the foulants’ ability to adsorb on the membrane surface, inluenced by hydrophobic interactions, hydrogen bonding, van der Waals attractions, and electrostatic interactions (Escobar et al. 2005). Therefore, the main strategy toward reducing membrane fouling is the prevention of the undesired adsorption or adhesion interactions between a foulant and a membrane to inhibit or, at least, minimize the fouling process. This may be realized via surface membrane modiication because the active (top) membrane layer plays a vital role in both the fouling interactions and the membrane separation. This chapter surveys the latest studies in which the reduction of irreversible organic fouling and biofouling is attempted by the modiication of the membrane surface.

3.2

SURFACE MEMBRANE PROPERTIES AFFECTING MEMBRANE FOULING

Several of the surface characteristics of membranes, such as hydrophilicity, charge, and roughness, are known to be strongly related to fouling because they determine the interaction between the membrane and the foulants (Rana and Matsuura 2010).

3.2.1

SURFACE HYDROPHILICITY

The majority of the commercial membranes for pressure-driven processes are made from hydrophobic polymers with high thermal, chemical, and mechanical stabilities. Because of the hydrophobicity of these materials, they are prone to adsorption of the fouling substances. It has been well documented that membranes with hydrophilic surfaces are less susceptible to fouling (Fane and Fell 1987; Hilal et al. 2005).

44

Membrane Modification: Technology and Applications

Therefore, an increase in the hydrophilicity of the membrane surface is often a key goal to reducing membrane fouling by organic pollutants and microorganisms. Due to the formation of hydrogen bonds, a thin layer of bounded water exists on the surface of the hydrophilic membrane. This layer can prevent or reduce undesirable adsorption or adhesion of the foulants on the membrane surface.

3.2.2

SURFACE CHARGE

The charge of the membranes is an especially important consideration for reducing membrane fouling where the foulants are charged, which is often the case. Usually, it is appropriate to use a membrane carrying the same electrical charge as the foulants. When the surface and the foulant have a similar charge, the electrostatic repulsion forces between the solute and the membrane prevent the solute deposition on the membrane, thereby reducing the fouling (Al-Amoudi and Lovitt 2007; Van der Bruggen et al. 2008). Thus, there have been a number of attempts to reduce fouling by incorporating ionizable functional groups on the membrane surface. For example, a negative surface charge on the membrane will have a beneicial effect on the separation of the proteins around neutral pH, because most proteins also have a negative charge in such conditions (Ulbricht 2006). In addition, most of the colloidal particles, such as NOMs, that deposit on the membrane surface are negatively charged (Hong and Elimelech 1997). Similar to the negatively charged surface, the positively charged membrane surfaces exhibited electrochemical repulsion against positively charged solutes (Kato et al. 2003).

3.2.3

SURFACE ROUGHNESS

There is a strong correlation between fouling and the surface roughness of the RO and the NF membranes (Elimelech et al. 1997). A higher roughness increases the total surface area to which foulants can be attached, and the ridge-valley structure favors the accumulation of foulants at the surface. As a result, membranes with rougher surfaces are observed to be more favorable for foulants attachment, resulting in more extensive fouling and faster fouling rates (Vrijenhoek et al. 2001). A membrane with a smooth surface is, therefore, not as easily fouled. In contradiction, an increase in lux with an increase in the surface roughness of RO membranes, which was attributed to the increase of the area available for the membrane transport, was shown (Hirose et al. 1996). However, in general, membrane development is currently focused on the reduction of membrane surface roughness. Taking into account the importance of the surface membrane properties on the fouling processes, the main aims of the membrane surface modiications to prevent or reduce the membrane fouling consist of hydrophilization, smoothing, and the introduction of the charged/bactericide groups on the membrane surface to minimize undesired interactions with the potential foulants (Hilal et al. 2005; Al-Amoudi and Lovitt 2007; Rana and Matsuura 2010). These aims are easier to achieve with controllable modiication techniques, which will be discussed later in this chapter, such as ultraviolet (UV) and redox-initiated surface grafting of the hydrophilic compounds, low-temperature plasma treatment, physical coating/adsorption of a

45

Reduction of Membrane Fouling by Polymer Surface Modification

thin layer of hydrophilic polymer on the membrane surface, chemical reactions on the membrane surface, and surface modiication of the polymer membranes with nanoparticles.

3.3

DEVELOPMENT OF LOW-FOULING POLYMER MEMBRANES VIA PHOTOINITIATED GRAFTING

Currently, UV-initiated graft polymerization has been widely used for the surface modiication of polymer membranes in attempts to develop composite membranes with enhanced resistance toward organic fouling and biofouling. As can be seen in Table 3.1, different hydrophilic monomers, such as N-vinyl-2-pyrrolidone (NVP), N-vinylformamide (NVF), N-vinyl-caprolactam (NVC), 2-hydroxyethyl methacrylate (HEMA), acrylic acid (AA), acrylamide (AAm), 2-acrylamidoglycolic acid (AAG),

TABLE 3.1 Surface Modification of Polymer Membranes via UV-Initiated Grafting to Reduce Their Fouling Base Membrane

Photografted Monomer(s)

PES

NVP, NVF, and NVC

PES

PEGMA

PES

NVP

PES

NVP and 2-mercaptoethanol

PES and PVDF

AMPS, qDMAEM, and HEMA

PES and PVDF

qDMAEM and AMPS

Performance of Modified Membrane(s) UF of BSA solutions; modiied membranes, especially with NVP, showed higher lux and lower fouling compared with the unmodiied membrane (Pieracci et al. 1999) UF of sugarcane juice and BSA solutions; modiied membranes showed more resistance to fouling and a higher rejection than unmodiied membranes; relatively high monomer concentration (40 g/l) at medium irradiation times (1.5–3 min) were among the optimal modiication conditions to reduce membrane fouling (Susanto et al. 2007) NF of NOM solutions; fouling tendency was reduced for membrane irradiated for 60 sec compared with membrane irradiated for 180 sec; enlargement of pore structure was observed under long irradiation times (Kilduff et al. 2000) UF of BSA solutions; although protein rejection remained unchanged after modiication, the permeability of the membranes decreased with an increase in grafting degree (Pieracci et al. 2002) MF of E. coli suspensions; a neutral hydrophilic membrane surface was less susceptible to fouling than charged (positively or negatively) membranes (Kochkodan et al. 2006) MF of E. coli suspensions; the modiied membranes were more resistant to biofouling; the number of bacterial cells able to proliferate from countable colonies was reduced for qDMAEM-grafted samples compared with unmodiied membranes (Hilal et al. 2003, 2004) (continued)

46

Membrane Modification: Technology and Applications

TABLE 3.1 (Continued) Surface Modification of Polymer Membranes via UV-Initiated Grafting to Reduce Their Fouling Base Membrane

Photografted Monomer(s)

PES/PVDF blend

NVP

PS

MPDSAH inner salt

PS

NVP, HEMA, AA, AAG, SPMA, AMPS AA, HEMA, PEGMA

PAN

PP

AA, AAm

PP

GAMA

PP

HEMA

PVDF

AA, HEMA, PDA, EDA

PE

Poly (dimethylsiloxane), PEG

Performance of Modified Membrane(s) UF of BSA solutions; BSA adsorption on the membranes decreased from 159 ± 2 to 13 ± 2 μg/cm2 after 10 min of grafting, the fouling degree of the blend membrane with 7 min grafting was reduced by 66%, and the lux recovery after chemical cleaning increased by about 32% (Zhang et al. 2009) UF of BSA solutions; hydrophilicity and antifouling properties of modiied membranes were enhanced with the increase in grafting degree (Yu et al. 2009) UF of BSA solutions; membranes modiied by NVP, AMPS, and AA showed high protein retention, high solution lux, and low irreversible fouling (Taniguchi and Belfort 2004) UF of BSA solutions; protein–polymer surface interactions were diminished by membrane modiication: adsorption and fouling were reduced both for negatively and positively charged membranes (Ulbricht et al. 1996) MF of activated sludge; modiied membranes showed better iltration performances in a submerged membrane bioreactor than unmodiied membranes. AA-grafted membrane had the best antifouling characteristics (Yu et al. 2008a) MF of activated sludge; after continuous operation in a membrane bioreactor for about 70 h, the reduction of water lux was 90.7% for the unmodiied membrane and ranged from 80.8% to 87.2% for the modiied membranes, increasing with an increase in the length of the grafted chains (Gu et al. 2009) MF of BSA solutions; the modiied membrane showed better protein resistance as well as hemocompatibility due to the enhancement of surface hydrophilicity: static water contact angle of the membrane surface decreased from 145° to 42° with the grafting degree increasing from 0 to 35.67 wt% (Hu et al. 2006) UF of pasteurized and homogenized milk with 3.2% protein and 1.5% fat; antifouling properties: lux recovery, irreversible lux loss, total lux loss, and fouling resistance of modiied membranes were enhanced due to an increase in membrane hydrophilicity (Rahimpour et al. 2009) UF of P. aeruginosa suspensions; membrane fouling increases as the membrane surface becomes more rough and more hydrophobic or the membrane charge (whether positive or negative) increases (Pasmore et al. 2001)

Reduction of Membrane Fouling by Polymer Surface Modification

47

quaternized 2-(dimethylamino) ethyl methacrylate (gDMAEM), 2-acrylamido-methylpropane sulfonic acid (AMPS), 3-sulfopropyl methacrylate (SPMA), poly(ethylene glycol) (PEG), poly(ethylene glycol) methacrylate (PEGMA), d-gluconamidoethyl methacrylate (GAMA), [(methacryloylamino)propyl]-dimethyl (3-sulfopropyl) ammonium hydroxide inner salt (MPDSAH inner salt), 2,4-phenylenediamine (PDA), ethylene diamine (EDA), and poly(dimethylsiloxane) have been used for modiication of the MF, UF, NF, and RO base membranes of various chemical natures. These include polyvinylideneluoride (PVDF), polyethersulfone (PES), polysulfone (PS), polypropylene (PP), polyacrylonitrile (PAN), and polyethylene (PE). In general, modiication of the membrane surface by UV-graft polymerization can be performed via “grafting-from” and “grafting-to” approaches.

3.3.1

UV-INITIATED “GRAFTING-TO” THE MEMBRANE SURFACE

In the case of the “grafting-to” method, preformed polymer chains, carrying reactive groups at the end or on the side chains, are covalently coupled to the surface (Zhao and Brittain 2000). The “grafting-to” procedure allows precise control of the grafted chain structure; however, because of a low density of grafted polymer chains (Kato et al. 2003), this method is seldom used for membrane modiication. An example is functionalization of UF PAN membranes with low-molecular-weight aromatic azide derivatives composed of different hydrophilic and hydrophobic components (Ulbricht and Hicke 1993). The separation characteristics and the protein fouling tendency were essentially changed depending on the type of functional groups introduced. This was explained by the increased hydrophilicity and the charge of the active membrane layer.

3.3.2

UV-INITIATED “GRAFTING-FROM” THE MEMBRANE SURFACE

Compared with the “grafting-to” method, the “grafting-from” approach is widely used for the surface modiication of various types of polymer membranes (Yamagishi et al. 1995a,b; Ulbricht 1996; Ulbricht et al. 1996; Pieracci et al. 1999, 2000, 2002; Ma et al. 2000a,b; Kilduff et al. 2000; Kaeselev et al. 2001, 2002; Yang and Yang 2003; Hilal et al. 2003, 2004; Taniguchi and Belfort 2004; Hu et al. 2006; Kochkodan et al. 2006; Susanto et al. 2007; Gu et al. 2009; Rahimpour et al. 2009; Yu et al. 2009; Zhang et al 2009; Abu Seman et al. 2010). The majority of the studies have focused on antifouling modiication of the UF and MF membranes (Table 3.1), because a very high grafting density is required to use the grafted polymer layer as a selective barrier in RO. In general, membrane modiication via the “grafting-from” approach may be realized with or without using a photoinitiator.

3.3.3

MEMBRANE MODIFICATION VIA “GRAFTING-FROM” METHOD WITHOUT THE USE OF A PHOTOINITIATOR

This approach involves the direct generation of free radicals from the base membrane polymers under UV irradiation and requires a photosensitive base polymer, a

48

Membrane Modification: Technology and Applications

photoreactive side group or part of the polymer backbone, or the introduction of photosensitive groups onto the membrane surfaces prior to graft copolymerization (Kato et al. 2003). Yamagishi et al. (1995a,b) modiied the surface of the UF PS and PES membranes via UV grafting of the vinyl monomers using the intrinsic photosensitivity of the PS and PES polymers and their ability to generate free radicals on UV irradiation. The irst step of the modiication process involves the absorption of light by the phenoxyphenyl sulfone chromophores in the backbone of the polymer chain (Figure 3.1). This results in the homolytic cleavage of a carbon–sulfur bond at the sulfone linkage. The generated aryl and sulfonyl radicals act as starters for a graft polymerization. Additionally, the sulfonyl radical may lose sulfur dioxide to generate an additional aryl radical, which may also initiate copolymerization. The polymeric free radicals then serve as the sites for the grafting of vinyl monomers and their subsequent polymerization. Thus, grafting occurs only at the membrane surface and the formation of the homopolymer in the monomer solution is avoided. Using this approach, various vinyl monomers were grafted onto the surface of the PS and PES membranes to decrease their fouling by proteins (Yamagishi et al. 1995a). It was shown that HEMA was the most effective monomer to decrease both the static adsorption and the membrane fouling during BSA iltration. UF luxes for 0.1 wt% BSA solutions were considerably higher (20%–27%) for the membranes grafted with HEMA compared with the lux at the unmodiied membranes, while the BSA retentions were similar for both membranes. The adsorption isotherms showed O S

O

hυ

O n

O

SO2

O

O

SO2

254 nm (112 kcal/mole)

X

+

O

O

SO2

CH2

CH X

n

–SO2

O

X

O

CH2 CH X

n

FIGURE 3.1 The mechanism for the photochemical modiication of the PES membrane with vinyl monomers without a photoinitiator. (Reprinted from Journal of Membrane Science, 105, Yamagishi, H., et al., 237–248, Copyright (1995), with permission from Elsevier.)

Reduction of Membrane Fouling by Polymer Surface Modification

49

that the amount of BSA adsorbed onto the HEMA-modiied PES membrane was about 43% and 28% less than for the unmodiied PES membranes for 1 and 2 mg/l of BSA in solution, respectively. The antifouling resistance of the modiied membranes was improved due to an increase in their hydrophilicity. However, it was also shown that the membrane’s pore size increased on UV irradiation, obviously due to PES photodestruction. Using the intrinsic photosensitivity of PES, 10 kDa UF membranes were modiied via UV-induced surface grafting of NVP, NVF, and NVC (Pieracci et al. 1999). The modiied membranes have been shown to have increased hydrophilicity of the order of 12%–25%, as measured by the static contact angle. This resulted in a decrease in the BSA fouling compared with the unmodiied PES membrane and a higher relative iltration performance compared with the commercial low protein adsorbing membranes cast from regenerated cellulose. The best-performing NVPmodiied membrane showed a 25% increase in hydrophilicity, a 49% decrease in BSA fouling, and a 4% increase in BSA retention, compared with the unmodiied PES membrane. It should be noted that the process of surface modiication of photosensitive base membrane polymers depends on the UV wavelength and intensity. It was found that the grafting eficiency of NVP onto the PES membranes was signiicantly higher using a high-energy 254 nm UV lamp compared with using a 300 nm UV lamp (Pieracci et al. 2002a). The signiicant pore enlargement that leads to an increased lux and the loss of BSA rejection was also attributed to the high-energy wavelength. These side effects have been reduced by selecting longer-wavelength UV lamps of 350 nm or using an appropriate UV ilter for short-wavelength UV lamps. It was shown that the degree of modiication (DM) for a given UV lamp wavelength also depends on the photosensitivity of the base membrane. A much higher irradiation dose was required to achieve the desired DM when a PS membrane was used compared with a PES membrane, due to the higher sensitivity of PES to UV irradiation (Kaeselev et al. 2001). The UV irradiation of the PES membrane involves two parallel competitive processes, cross-linking and chain scission, which determine the transport properties of the modiied membrane. The cross-linking increased the hydrodynamic resistance of the modiied membrane, while the loss of the retention was affected by the chain scissions (Kaeselev et al. 2002). As a rule, two methods are used for membrane modiication via the “graftingfrom” approach without a photoinitiator: an immersion method when the membranes are irradiated while immersed in a monomer solution, and a dip method when the membranes are dipped in a monomer solution and then irradiated in nitrogen or another inert gas (Pieracci et al. 2000). UV-initiated graft polymerization of AA on the surface of a PES membrane via the immersion method using a 365 nm wavelength UV lamp was studied (Abu Seman et al. 2010). It was shown that irreversible membrane fouling with humic acids (HA) was reduced after modiication due to the increase in the hydrophilicity and the negative surface charge of the modiied membranes. In comparing the immersion and the dip methods, it should be noted that the immersion modiication usually requires a large amount of monomer and may be less adaptable to continuous processes on an industrial scale. Moreover, it was shown

50

Membrane Modification: Technology and Applications

that for the same unmodiied membrane, monomer concentration, and irradiation time, the achieved DM value using the dip method was two to three times higher than that by the immersion method (Pieracci et al. 2000). Using a dip-modiication method, very hydrophilic composite PES membranes were produced via grafting of the NVP, NVF, and NVC (Pieracci et al. 1999). However, membrane permeability was decreased after modiication. This loss of membrane lux has been linked to the blockage of the membrane pores by the grafted polymer chains. It was shown that a high grafted chain density and long grafted polymer chains promote lux loss. Thus, short grafting chains may be needed for modiication to preserve the lux of the modiied membrane. On the other hand, a high grafted chain density and long grafted chains may be essential to provide the necessary surface hydrophilicity to decrease membrane fouling. Therefore, the graft chain density and the chain length should be optimized to impart the necessary surface hydrophilicity but maintain the lux as high as possible. To control the degree of polymerization and chain length during UV-initiated surface grafting, chain transfer agents can be used. With 2-mercaptoethanol as a chain transfer agent, hydrophilic poly(NVP)-grafted membranes were prepared with three to ive times higher luxes than that of the unmodiied PES membrane and the dip-modiied membranes without the chain transfer agent (Pieracci et al. 2002).

3.3.4

MEMBRANE MODIFICATION VIA “GRAFTING-FROM” APPROACH WITH A PHOTOINITIATOR

For surface modiication with an added photoinitiator, the initiating radical sites should be generated on the membrane surface by the reaction of the photoinitiator with the base membrane polymer under UV irradiation. Benzophenon (BP) or its derivatives are most often used for the initiation of the UV-assisted graft polymerization of vinyl monomers on the surface of the polymer membranes. In this case, the key step in the initiation of generating radicals is hydrogen abstraction from the polymer backbone (Kato et al. 2003). The photoinitiator may be loaded on the membrane surface by adsorption or may be dissolved in the monomer solution. For example, PET nucleopore membranes were modiied using BP dissolved in the monomer solution following UV irradiation of the solution and the immersed membrane (Yang and Yang 2003). It was shown that photografting occurred mainly on the top membrane surface rather than in the membrane pores. This approach is relatively simple; however, its main drawback is a low local concentration of BP on the membrane surface because BP moves to the membrane surface only by diffusion. This results in a low grafting eficiency. High bulk BP concentration may cause a side reaction, such as homopolymerization. In addition, the use of monomers that do not have a common solvent with BP is limited; BP is almost insoluble in water. To improve the grafting process, preliminary adsorption of BP on the membrane surface was used (Ulbricht 1996). In this way, the local concentration of the photoinitiator on the membrane surface was increased, while the BP concentration in the bulk of the monomer solution was kept very low to reduce the homopolymerization

Reduction of Membrane Fouling by Polymer Surface Modification

51

1000 1 2 3

DM (µg/cm2)

800

600

400

200

0 0

2

4 t (min)

6

8

FIGURE 3.2 The degree of modiication (DM) of the MF PES membrane with qDMAEM (1), AMPS (2), and HEMA (3) vs. the UV-irradiation time.

process. Using this method, PAN membranes have been grafted with PEGMA to prepare a low-protein-adsorbing UF membrane with a relatively high permeability (Ulbricht et al. 1996). In our laboratory, in an attempt to enhance the membrane resistance to biofouling, MF PES membranes were photochemically modiied with a thin layer of a number of hydrophilic polymers: neutral poly(HEMA), negatively charged poly(AMPS), and positively charged and bactericide poly(qDMAEM). It was shown that the DM of a composite membrane, which was calculated from the difference between the weights of the membrane sample before and after modiication, depends on the amount of the photoinitiator loaded on the membrane surface and the polymerization time. As seen in Figure 3.2, the DM values increase with the polymerization time due to an increase in the molecular weight of the grafted macromolecules on the membrane surface. Higher DM values were obtained using HEMA compared with modiication with qDMAEM and AMPS. This is explained by the fact that qDMAEM and AMPS molecules carry positive and negative charges, respectively. Owing to the electrostatic repulsion of similarly charged molecules, the possibility of growing grafted chains for these monomers diminishes. As a result, grafted poly(AMPS) and poly(qDMAEM) chains are shorter than those for poly(HEMA). Note that the photochemical modiication of the PES membranes proceeds relatively fast and the DM values reach 300–500 μg/cm2 after a few minutes of UV irradiation. Membrane fouling during the iltration of Escherichia coli suspension is determined mainly by the hydrodynamic operation conditions and depends only to a minor extent on the chemical nature of the membrane surface (Ma et al. 2000a). At the same time, we have found that the chemical nature of the membrane surface signiicantly affects the ability of the membranes to recover their luxes after washing. As can be seen in Table 3.2, the membranes with positively or negatively charged surfaces are more inclined to fouling compared with the membranes with neutral

52

Membrane Modification: Technology and Applications

TABLE 3.2 Performance of Surface-Modified PES Membranes in the Filtration of an E. coli Suspension Membrane Initial PES PES + HEMA PES + AMPS PES + qDMAEM

Jc1a

Jc2a

Jc2/Jc1 (%)

J1/Job (%)

AAc (%)

900 690 650 680

270 240 235 225

30 35 36 33

51 72 63 51

0 0 25 97

DM of the composite membranes is 500 ± 20 μg/cm2. The concentration of the suspension is 1.58 × 104 CFU/cm3, the operating pressure is 0.5 bars, and the iltration time is 1 h. a J and J are the membrane luxes at the beginning and at the end of the iltration cycle, c1 c2 respectively. b J and J are the initial water lux and the water lux after iltration of the bacterial suspen0 1 sion and washing with distilled water. c Antibacterial activity is the ratio of the numbers of bacterial colonies that were grown on the unmodiied and modiied membranes at the identical conditions, respectively.

hydrophilic surfaces. Obviously, the charged surface regions serve as convenient sites for the attachment of microorganisms due to electrostatic attraction and the formation of ionic bonds (Pasmore et al. 2001). The highest value of lux recovery (J1/J0) for the membrane modiied by HEMA is explained by the absence of electrostatic attraction between the bacterial cell and the membrane surface. As a result, the cells are more easily removed from the membrane surface during washing, as compared with the membranes with charged surfaces. A distinct difference in the performances of the modiied membranes was found during treatment of the natural surface water in the noncontiguous regime, when the iltration cycle (1 h) alternates with stops and when the membranes were kept for 24 h at room temperature in the surface water being iltered. As seen in Figure 3.3, the average lux values for the initial PES membrane, as well as for the membranes modiied with HEMA and AMPS, are reduced with time more rapidly than for the membrane modiied with qDMAEM. This is explained by the high antimicrobial properties of the modiied poly(qDMAEM) membrane (Table 3.2). These antimicrobial properties inhibit the growth of bacteria on the surface, thereby minimizing the formation of the biofouling layer. To reduce the formation of an undesired homopolymer and a cross-linked or branched polymer in the “grafting-from” approach with an initiator, a photoinduced living graft polymerization method for surface membrane modiication was developed (Ma et al. 2000b). In the irst step, BP abstracts the hydrogen from the substrate to generate surface radicals and semipinacol radicals, which combine to form surface photoinitiators in the absence of monomer solutions (Figure 3.4). In the subsequent step, the monomer solutions are added onto the active substrate, and the surface initiators initiate the graft polymerization under UV irradiation. In this method, the graft density and the graft polymer chain length can be controlled independently,

53

Reduction of Membrane Fouling by Polymer Surface Modification 800

Jav (l/m2 h)

600

1

2

3

4

400

200

0

0

1

2

3

4

5

t (days)

FIGURE 3.3 The average luxes of the iltration cycle for the initial PES membrane (1) and the composite membranes modiied with HEMA (2), AMPS (3), and qDMAEM (4) vs. the time of the membranes exposure to surface water (river Nivka, Kyiv). The DM of the membranes is 500 ± 20 μg/cm2. The duration of the iltration cycle is 1 h at an operating pressure of 0.5 bar. The E. coli index of surface water is 3.52 × 103 CFU/l.

since the initiator formation and the graft polymerization occur independently in the successive steps. The control of the graft chain length and density is important to minimize protein adsorption and cell adhesion on the polymer surface (Fujimoto et al. 1993). When the graft layer is much larger in thickness than the protein radius and has a high water content, the protein molecules will readily migrate into the graft layer to be sorbed. Extensive grafting gives rise not only to an increase in the thickness of Surface initiator

UV 1st step

H

C

OH

C

OH

C O Substrate membrane Grafts

2nd step

C

OH

UV Monomer

C

OH Monomer

FIGURE 3.4 A schematic diagram of a photoinduced living graft polymerization on a membrane surface. (Reprinted with permission from Ma, H., et al., Macromolecules, 33, 331–335. Copyright 2000b American Chemical Society.)

54

Membrane Modification: Technology and Applications

the graft layer, but also to an increase in the volume fraction of the graft segments in the layer. The latter also has a signiicant inluence on protein adsorption. Therefore, the precise design of a graft layer and well-controlled graft architecture is a very important issue to minimize membrane fouling. In general, it should be noted that very detailed studies have been carried out on surface membrane modiication via UV-initiated graft polymerization, including the grafting mechanisms, the dependence of grafting eficiency on grafting time, the monomer type, and the UV wavelength and intensity. The attractive features of UV grafting are the easy and controllable introduction of graft chains with a high density and their exact localization to the membrane surface. Furthermore, covalent attachment of the graft chains onto a polymer surface avoids their delaminating and ensures the long-term chemical stability of the introduced chains in contrast to physically coated polymer layers. These advantages have meant that this modiication technique has gained interest from membrane manufacturers, for example, for the continuous modiication of PS hollow iber membranes with an anionic grafted polymer layer to obtain NF membranes (Bequet et al. 2002). The disadvantages of the modiied membranes are usually their reduced luxes compared with those of the unmodiied membranes because the grafting layer adds extra hydraulic membrane resistance. Grafting may also increase the manufacturing costs due to the additional use of organic solvents, monomers, and UV equipment.

3.4

MISCELLANEOUS GRAFTING METHODS ON THE MEMBRANE SURFACE

In contrast to UV-initiated graft polymerization, redox-initiated grafting gives the possibility of modifying the polymer membranes in situ, inside commercial wound membrane elements (Belfer et al. 1998a). A redox system composed of potassium persulfate and potassium metabisulite was used to generate the starting radicals for the graft polymerization of AA and other hydrophilic monomers on the surface of the CA membranes (Belfer et al. 1998b, 2001). It was shown that despite the gradual decrease of the lux, the surface-modiied membranes had a lower protein sorption and a better and more reversible lux recovery after cleaning. A similar approach was used for the in situ preparation of the NF PES membranes containing hydrophilic functional groups such as SO3H, COOH, or C(=O)NH2 (Reddy et al. 2005). Testing the modiied and unmodiied membranes over a period of 30 days demonstrated that the surface-modiied composite membranes have better fouling resistance characteristics. In the case of the unmodiied membranes, the lux decreased from 41.65 to 19.21 l/(m2 h), while for the surface-modiied membranes under similar conditions, the lux reduced from 46.75 to 31.62 l/(m2 h). However, in the case of the NF PA membranes, it was observed that polymerization could take place inside the pores of the base support membrane as a result of the penetration of the monomer through the active layer, particularly for high degrees of grafting (Freger et al. 2002). Gullinkala and Escobar (2010) used porcine pancreatic lipase to catalyze the polycondensation of PEG to the surface of the CA membranes. The main advantage of this proposed “green” approach, based on catalytic polymerization, is a low

Reduction of Membrane Fouling by Polymer Surface Modification

55

degree of homopolymerization because the reaction occurs between the two functional groups present in the monomers and the membrane surface. The unmodiied and the modiied membranes displayed comparable initial lux values, lux decline curves, and rejections of dextran and BSA. This is suggested to be due to high hydrophilicity and the similar charge characteristics of both membrane samples. However, the lux recovery after UF of NOM followed by backwashing was quite different; the modiied membrane regained nearly 97% of its initial lux value within 40 min of iltration, while the unmodiied membrane recovered only 85%. A lower cake accumulation on the membrane surface was also found for modiied membranes. These improvements are believed to be due to the high lexibility of the highly hydrophilic grafted PEG chains that prevents the membrane fouling. It should be noted, however, that the membrane modiication via the proposed “green” approach takes rather a long time, about 50 h.

3.5

PLASMA TREATMENT OF POLYMER MEMBRANES

Over the last two decades, the plasma treatment of the polymer membranes has been intensively studied in attempts to increase the hydrophilicity and induce lowfouling properties for membrane surfaces (Ulbricht and Belfort 1996; Bryjak et al. 1999; Chen and Belfort 1999; Gancarz et al. 1999a,b, 2002, 2003; Kim et al. 2002; Wavhal and Fisher 2002; Zhan et al. 2004; Dattatray et al. 2005; Kull et al. 2005; Zhao et al. 2005; Yu et al. 2005, 2006, 2007, 2008a,b; Tyszler et al. 2006; Pozniak et al. 2006; Dong et al. 2007; Yan et al. 2008; He et al. 2009). Usually, plasma treatment of the membranes can be carried out in three different modes: (i) with nonpolymerizable gas molecules; (ii) with polymerizable vapors; and (iii) with plasma-induced grafting of the polymer chains to the membrane surface, where the plasma treatment with nonpolymerizable gases and plasma-induced grafting are the most widely used.

3.5.1

PLASMA OF NONPOLYMERIZABLE GASES

Most studies in this ield are focused on the treatment of MF and UF PP, PS, and PES membranes in the plasmas of air, oxygen, nitrogen, ammonium, or carbon dioxide (Dattatray et al. 2005; Yu et al. 2005). After air plasma treatment, the water contact angles for the modiied PP membranes decreased from about 128.5° to 35.0° with an increase in the air plasma treatment time from 0 to 8 min (Yu et al. 2008b). It was found that after continuous operation in a submerged membrane bioreactor (SMBR) for about 110 h, the lux recoveries after water and caustic cleaning were 11.66% and 34.99% higher for the 4 and 2 min air plasma-treated membrane than those for the unmodiied membrane. The irreversible fouling exhibited a marked dependence on the surface membrane properties, while the reversible fouling was only weakly dependent on the membrane surface chemistry. The air and oxygen plasma treatments modify the membrane surface, making it more hydrophilic via the incorporation of the hydroxyl and peroxide groups (Figure 3.5). The higher hydrophilicity of the membranes treated with oxygen plasma for 20 sec meant that they were less fouled at UF of the gelatine solutions (Kim et al. 2002).

56

Membrane Modification: Technology and Applications

OOH O2 - plasma

O–O O–O

Aeration (O2)

OOH OOH OOH

FIGURE 3.5 Membrane surface modiication via O2 plasma treatment. (Reprinted from Desalination, 189, Tyszler, D., et al., 119–129, Copyright (2006), with permission from Elsevier.)

He et al. (2009) found that 30 sec CO2 plasma-treated membranes showed better regeneration behavior at UF of the BSA solutions. Flux recoveries after water and caustic cleaning were 33.1% and 27.5% higher, while a reduction in the pure water lux was 9.2% lower than that for the virgin membranes. In addition, the amount of protein adsorption decreased by over 50% for the modiied membranes compared with the unmodiied membranes. A potential disadvantage was shown with the prolonged exposition of the UF PS membrane to CO2 plasma, which resulted in severe material ablation and an increase in pore size (Gancarz et al. 1999a). Air and oxygen plasmas also etch polymer surfaces to a great extent (Dattatray et al. 2005; Yu et al. 2005), while polymer degradation may be minimized by using N2 plasma because of its weak effect. To generate the same hydrophilicity of a PS surface, 5 min for N2 plasma is needed compared with only 1 min for CO2 plasma (Bryjak et al. 1999). Treatment with N2 plasma introduces various chemical groups, such as amine, imine, amide, or nitrile, on the membrane surface, making it more hydrophilic and less liable to fouling (Kull et al. 2005). For example, after continuous operation in an SMBR for about 90 h, lux recoveries for plasma-treated PP membranes for 8 min were 62.9% and 67.8% higher than those for the virgin membrane after water and NaOH cleaning (Yu et al. 2007). The hydrophilicity of the NH3 plasma-treated PP membranes increased with the increase in plasma treatment time and decreased with the increase in storage time (Yan et al. 2008). The adsorption of BSA on the modiied membranes was lower than that on the unmodiied-membrane surface and the lux recoveries after water and caustic cleaning for the NH3 plasma-treated membranes for 1 min were 51.1% and 60.7% higher than those for the unmodiied membrane. However, the mechanical properties of the membranes decreased after prolonged plasma treatment, thus the optimal plasma treatment time for membrane modiication was taken as 1 min.

3.5.2

PLASMA OF POLYMERIZABLE MOLECULES

Plasma polymerization with easily polymerizable monomers, which carry multiple bounds, can be used for modifying the membrane surfaces by the deposition of new, thin polymer layers with thicknesses from several nanometers to one micrometer.

Reduction of Membrane Fouling by Polymer Surface Modification

57

Such layers are usually made of highly cross-linked material and show good adhesion to the substrate. Examples of such membrane surface treatment are plasma polymerization of allyl alcohol and allyl amine (Gancarz et al. 2002, 2003). It was shown that the membranes modiied with allyl amine do not foul as intensively during UF of the BSA solutions compared with the unmodiied membranes. Similar behavior was also shown for membranes modiied by the deposition of plasma-polymerized n-butylamine; however, in this case, the modiied layer deposited on the membrane surface was not as enriched in amines as the polymer formed from allyl amine (Gancarz et al. 2002).

3.5.3

PLASMA-INDUCED GRAFTING OF MEMBRANE SURFACE

Plasma can generate radicals on the polymer surface that are stable in vacuum but can react rapidly when exposed to a gas containing monomers. Additionally, when a plasma-activated membrane is exposed to oxygen or air, peroxide compounds are formed on the surface (Figure 3.5), which may be used for initiating plasmainduced graft copolymerization of the vinyl monomers (Tyszler et al. 2006). Grafting density and the length of the grafted chains can be controlled by the parameters of the plasma treatment, which include power, pressure, and treatment time, and the polymerization conditions, such as the kind of solvent, the monomer concentration, and the grafting time (Zhao et al. 2005; Pozniak et al. 2006). Gancarz et al. (1999b) compared three different approaches to modifying the PS membranes with AA via plasma-initiated graft polymerization. These were: (i) Grafting in solution: the plasma-treated polymer membrane was exposed to air for 5 min and dipped in a deaerated aqueous solution of monomer; (ii) Grafting in the vapor phase: when the Ar plasma treatment on the polymers was completed, a monomer vapor was introduced into the chamber; and (iii) Plasma polymerization of the monomer vapors in a plasma reactor. Comparing the above three methods, the PS membrane with AA plasma-initiated grafting in the vapor phase of the monomer was the most promising from the point of view of the iltration properties. A similar grafting process in the vapor phase was used for modifying the UF PES (Chen and Belfort 1999) and the PAN/PS membranes (Ulbricht and Belfort 1996). Note that the hydrophilicity of the membranes treated with the Ar plasmas was not permanent, in contrast to the membranes treated with the Ar plasma followed by AA grafting (Wavhal and Fisher 2002). The modiied membranes were easier to clean with NaOH to recover the lux after UF of the BSA solutions. Yu et al. (2006) used air plasma-initiated grafting of polyvinylpyrrolidone (PVP) in attempts to reduce the PP membrane fouling in an SMBR. It was shown that after continuous operation for about 50 h, the lux recovery, reduction of lux, and relative lux ratio for the modiied membranes were 53% higher, 17.9% lower, and 79% higher, respectively, than those for the unmodiied membranes. The water contact angle on the PVP-immobilized membrane showed a minimum value of 72.3°, approximately 57° lower than that on the unmodiied one. The grafting of a positively charged monomer (DMAEM) to the PS membrane reduced the adsorption of the positively charged lysozyme, while the grafting of a negatively charged monomer (AA) reduced the adsorption of a negatively charged

58

Membrane Modification: Technology and Applications

BSA due to an enhanced electrostatic repulsive force between the solutes and the modiied-membrane surfaces (Zhan et al. 2004). In addition, a reduction of about 96% in bioilm formation by Listeria monocytogenes on the modiied surfaces compared with the unmodiied PA and PE polymers was observed after SiCl4 plasma treatment followed by PEG grafting (Dong et al. 2007). In general, it can be concluded that the attractive features of membrane modiication via plasma treatment are the very short modiication time and the possibility to adjust the surface membrane properties to satisfy a particular request without affecting the bulk of the polymer. However, it should be noted that there are some problems with the repeatability of this modiication procedure and its scaling-up, especially for continued modiication.

3.6

PHYSICAL COATING/ADSORPTION ON THE MEMBRANE SURFACE

Coating a thin layer of water-soluble polymers or surfactants from solution by physical adsorption is a lexible technique to optimize the hydrophilicity, smoothness, and surface charge of the membrane surface (Li et al. 2000; Maartens et al. 2002; Reddy et al. 2003; Asatekin et al. 2006, 2009; Xie et al. 2007; Bruening et al. 2008; Boributh et al. 2009; Du et al. 2009; Sagle et al. 2009; Zhou et al. 2009; Ba et al. 2010). The irst studies in this ield dealt with the modiication of the UF membranes to reduce their fouling with proteins (Kim et al. 1988; Nystrom 1989; Brink and Romijn 1990). Kim et al. (1988) showed that the fouling of the UF membranes with proteins may be reduced by surface adsorption with watersoluble polymers, such as PVA, methylcellulose (MC), and PVP. The treatment provided an increase in the initial UF lux and a slower lux decline. MC was the most effective of the polymers tested in enhancing UF lux, showing an average lux advantage of 30%–40% for the irst usage. Nonionic, hydrophilic polymers were found to be the most effective in minimizing lactoglobulin adsorption as well as in decreasing membrane resistance during UF, while the application of surfactants and ionic polymers was less successful (Brink and Romijn 1990). On the other hand, it was shown that modifying the PS membrane with polyethylenimine (PEI) decreased the lux reduction during UF of the ovalbumin solutions due to increased hydrophilicity and electrostatic repulsion between the protein molecules and the modiied membrane (Nystrom 1989). Thereafter, many hydrophilic polymers, such as PVA, PAA, PEG-based hydrogels, and chitosan, have been coated on different MF, UF, NF, and RO membranes using casting (Asatekin et al. 2006, 2009; Du et al. 2009; Sagle et al. 2009), adsorption (Maartens et al. 2002; Ba et al. 2010), or iltration (Li et al. 2000; Reddy et al. 2003; Boributh et al. 2009) techniques.

3.6.1

COATING VIA CASTING

Asatekin et al. (2009) prepared novel composite NF membranes by casting the synthesized amphiphilic copolymer PAN-graft-poly(ethylene oxide) (PAN-g-PEO) onto

Reduction of Membrane Fouling by Polymer Surface Modification

59

UF PAN membranes. The coated membranes were immersed in isopropanol for 30 min and thereafter in a water bath. It was shown that during precipitation, the copolymer undergoes microphase separation, forming interpenetrating networks of PAN-rich and PEO-rich nanodomains. Transmission electron microscopy reveals that PEO domains act as water-permeable nanochannels and provide the size-based separation capability of the membrane. A small decline in lux (15%) was observed in a 24 h dead-end iltration experiment with 1 g/l BSA solution using the modiied membrane, while the base UF membrane lost 81% of its lux irreversibly in the same conditions. It was concluded that the PEO “brush” layer, formed on the membrane surface, acts as a steric barrier to protein adsorption, endowing these membranes with exceptional fouling resistance. Composite UF PVDF membranes modiied with a self-assembling graft copolymer PVDF-graft-poly(oxyethylene) methacrylate showed a good fouling resistance for BSA, HA, and sodium alginate at feed concentrations of 1000 mg/l and activated sludge (Asatekin et al. 2006). For example, dead-end iltration of the activated sludge with 1750 mg/l of volatile suspended solids resulted in a constant lux throughout the 16 h iltration period. Interfacial force measurements with an atomic force microscope showed the presence of steric foulant-membrane repulsive forces and a lack of adhesion forces between the foulant and the membrane. However, a possible ester bond linkage of the PEO side chain in acidic or basic media may restrict the application of the modiied membranes. Sagle et al. (2009) used a drawdown coating of cross-linked PEG-based hydrogels modiied with RO membranes to reduce their fouling. The cross-linked PEG-based hydrogels were synthesized via the photoinitated copolymerization of PEG diacrylate as the cross-linker and PEG acrylate, 2-hydroxyethyl acrylate, or AA as the comonomers. It was evaluated that the coatings deposited on the membrane surface were approximately 2 μm thick. It was shown that the water luxes of the coated membranes were smaller than those of the uncoated membranes, but the fouling of the modiied membranes with cationic dodecyltrimethyl ammonium bromide (DTAB), anionic sodium dodecyl sulfate (SDS) surfactants, and oil/water emulsions was essentially reduced. At the iltration of the oil/water emulsion made with DTAB, the lux of the base membrane after 24 h decreased to 26% of its initial value, while the water lux of a PEGDA-coated RO membrane was 73% of its initial value. It was shown that the membrane surface charge correlates with the fouling properties of the membranes; negatively charged membranes foul extensively in the presence of positively charged surfactants and experience minimal fouling in the presence of negatively charged surfactants. Commercial UF PVDF membranes with a cutoff of 120 kDa were modiied by surface coating with a PVA aqueous solution followed by solid–vapor interfacial cross-linking with glutaraldehyde (Du et al. 2009). Fouling tests using a 5 mg/l protein solution showed that a short period of coating and cross-linking improved the antifouling performance. Additionally, after UF of the surface water with a total organic carbon of approximately 7 mg/l during 18 h, the lux of the modiied membrane was twice as high as that of the unmodiied membrane. The cake-fouling layer could also be more easily removed from the PVA-modiied membrane by alkaline cleaning. The improved fouling resistance of the modiied membrane was related

60

Membrane Modification: Technology and Applications

to an increase in membrane smoothness and hydrophilicity after coating with the PVA layer.

3.6.2

COATING VIA ADSORPTION

Maartens et al. (2002) used the adsorption of the nonionic surfactants, Triton X-100 and Pluronic F108, to modify the tubular UF PES membranes, which were used for iltering the pulp and paper efluents. The adsorption of the surfactants onto the PES membranes is schematically presented in Figure 3.6. Triton X-100 is adsorbed to the membrane by hydrophobic interaction with the hydrophobic C6H4 groups, with the hydrophilic CH2CH2O groups facing toward the aqueous phase. Pluronic F108, on the other hand, is anchored onto the hydrophobic membrane surface by means of the hydrophobic poly(propylene oxide) center group. The two hydrophilic poly(ethylene oxide) groups at both ends of the molecule face toward the aqueous phase. It was shown that increasing the hydrophilic characteristics of the membranes due to the surfactants’ adsorption could reduce the amount of phenolic foulants adsorbed onto the membranes. Precoating of the PES membranes with Pluronic F108 drastically diminishes the foulants’ adsorption over a 90 h iltration time under cross-low conditions. It should be noted that the membrane modiication not only reduced fouling, but also improved the eficiency of cleaning to remove the foulant layers. The lux through the fouled membranes was successfully restored by cleaning with the nonionic detergent Triton X-100 and sponge balls. Pluronic F108-coated membranes were more easily cleaned. The antifouling effect induced by this agent was, however, lost after cleaning with Triton X-100, Hydrophilic tail Hydrophobic head

(a) OH OH OH OH OH OH

OH OH OH OH OH OH Hydrophilic buoy group

Hydrophobic group

OH

(b)

OH OH

OH

OH OH OH OH

FIGURE 3.6 A schematic representation of the adsorption of nonionic surfactants onto membranes: (a) Triton X-100 and (b) Pluronic F108. (Reprinted from Journal of Membrane Science, 209, Maartens, A., et al., 81–92, Copyright (2002), with permission from Elsevier.)

Reduction of Membrane Fouling by Polymer Surface Modification

61

thus the membranes had to be recoated with the surfactants after the cleaning procedure. The hydrophilicity of the MF PP membranes may be increased by increasing the amount of surfactant Tween-20 adsorbed onto the surface or in the pores of the membrane (Xie et al. 2007). The PP membrane modiied with a monolayer of the adsorbed surfactant showed higher lux and stronger antifouling ability than the unmodiied membrane after operating in an MBR for about 12 days. Charge reversal on a low-pressure RO PA membrane surface, due to the electrostatic self-assembly of PEI, was found to increase the fouling resistance of the modiied membrane to the cationic foulant DTAB, owing to the enhanced electrostatic repulsion and the increased surface hydrophilicity (Zhou et al. 2009). It was shown that the improved fouling resistance and the increased surface hydrophilicity compensated for the reduction in the membrane permeability due to the adsorption deposition of the PEI layer on the membrane surface. Ba et al. (2010) used the adsorption of water-soluble polymers, such as PVA, polyacrylic acid (PAA), and polyvinyl sulfate–potassium salt (PVS) on the surface of the positively charged P84-PEI membrane to form a protective coating layer to improve the membrane fouling resistance. PVA, PAA, and PVS as the coating materials represented neutral, partially charged, and highly charged polyelectrolytes, respectively. Surface coating experiments were carried out in a cross-low iltration cell with the circulation of 50 mg/l of a PVA, PAA, or PVS polyelectrolyte aqueous solution over the base membrane for 8–12 h. It was shown that by applying these coatings, the hydrophilicity, smoothness, and surface charge may be modiied and optimized. This reduced the membrane fouling with BSA, HA, and sodium alginate. Membrane surface charge was observed to play the most important role in foulant adsorption. The uncoated membrane had a strong positive charge so that foulants such as BSA, HA, and sodium alginate were adsorbed quickly and irmly. The PVA-coated membrane also had a positive charge, and fouling by negatively charged materials such as HA and sodium alginate was still high. The PVS-coated and PAA-coated membranes had a low surface charge and, as a result, the fouling with BSA and HA was diminished due to a reduction in the charge interactions.

3.6.3

COATING VIA FILTRATION

Li et al. (2000) prepared PVA-coated TFC membranes by iltrating aqueous solutions containing PVA and cross-linking agents through the porous membrane support, followed by heat treatment. As a result, a cross-linked PVA gel layer was formed on the surface and in the pores of the modiied membranes. The modiied membranes show higher antifouling characteristics compared with the unmodiied membranes during UF of the pepsin solutions. The UF PES membranes containing negatively charged sulfonic acid groups on the surface were obtained on iltration of an aqueous solution of poly(sodium 4-styrenesulfonate) (PSS) for about 100 min using a dead-end iltration cell (Reddy et al. 2003). It was shown that thin porous membranes are modiied only on the top surface because the PSS macromolecules are not able to enter the pores. However, for membranes with wider pores, PSS permeation results in the formation of charged

62

Membrane Modification: Technology and Applications

groups on both the surface and the pore walls of the membrane. The major difference between the modiied and the unmodiied membranes was found in their lux recovery after UF of the PEG and dextran solutions. Flux recovery ratios of >90% were obtained for the modiied membranes compared with 55% for the unmodiied membranes. Thus, the surface-modiied membranes have better “cleanability” and “antifouling” characteristics than the base membranes. Boributh et al. (2009) compared three different methods for modifying MF PVDF membranes with chitosan to reduce BSA fouling. These were: (i) An immersion method, when the membrane was immersed in a chitosan solution for a ixed time; (ii) A low-through method, when the chitosan solution was iltered through the membrane; and (iii) The combined low-through method and the surface low method. It was shown that the membranes modiied by a combined low-through method and a surface low method showed better antifouling properties compared with others. This is due to the deposition of the chitosan both on the surface and in the pores, resulting in the prevention of BSA adsorption. For a membrane modiied by immersion, the chitosan was deposited only on the membrane surface. Therefore, BSA could be adsorbed easily on the pore walls, which led to a high lux decline and irreversible fouling. At this point, it should be mentioned that depending on the adsorption afinity with the membrane surface, the adsorbed coating layer can be stable or removable. The thin-coated ilms prepared via the deposition of positively and negatively charged polyelectrolytes show good stability due to the electrostatic attraction between the membrane surface and the deposited layers (Bruening et al. 2008; Ba et al. 2010). On the other hand, for hydrogen-bonded modiied layers, the strength of the hydrogen bonding between the membrane surface and the deposited layer can be altered by changes in the solution pH, thus these layers can be removed and replaced (Sukhishvili and Granick 2002; Kharlampieva and Sukhishvili 2003). For example, the cleaning procedure for PVA-coated membranes included membrane treatment with HCl at a pH of 2 and stirring for 15–20 h (Ba et al. 2010). Thus, if membrane fouling occurs, the PVA layer and the attached foulants can be removed by acid cleaning to refresh the membrane. It may be much easier and more cost-effective to remove and replace the ilm instead of replacing the membrane. In general, the adsorbed coatings are relatively simple to apply and the process can be performed in commercial membrane elements. In addition, the type of coating can be tailored to the speciic application of interest. However, despite the lexibility of the coating and the adsorption methods to change the hydrophilicity, smoothness, and charge of the membrane surface, their main drawback is the limited stability of the modiied layer over time because of the possible desorption of the coated/ adsorbed polymers from the membrane surface into the bulk of the feed solutions.

3.7

CHEMICAL REACTIONS ON THE MEMBRANE SURFACE FOR FOULING REDUCTION

As discussed above, the introduction of charged groups on the membrane surface is a useful approach to reduce membrane fouling with charged organic compounds.

Reduction of Membrane Fouling by Polymer Surface Modification

63

In this context, various chemical reactions may be used for creating different functional groups, such as –SO3 (Duputell and Staude 1993) or –CO2H (Tremblay et al. 1992; Bryjak et al. 1998), on the membrane surface. The quantity of the introduced functional groups and the thickness of the modiied surface layer depend on the treatment time, temperature, and concentration of the modiication agent. For example, during the prolonged exposure of the PAN membrane to 1 M NaOH, the surface nitrile groups turned into carboxylic groups (Bryjak et al. 1998). The modiied membranes were less prone to fouling with BSA with a reduction in the average pore diameter of about 80% for the untreated membranes and 20% for the surface-modiied membranes. The blend UF PES/PAN membranes treated with aqueous NaOH solutions at room temperature for 24 h showed higher lux recovery ratios compared with the unmodiied membranes after the UF of the PEG, dextran, and PPS solutions. The increase in the fouling resistance is believed to be due to the higher hydrophilicity of the modiied-membrane surface (Reddy and Patel 2008). To improve the performance of the DS5DL (Osmonics) NF membrane, it was immersed for 14 days in 1% w/v hydroluoric acid (HF) (Gonzalez-Munoz et al. 2006). Such a prepared membrane was used for the puriication of industrial phosphoric acid (8 M) and for the removal of Na 2SO 4 from industrial wastewater. In both cases, the treated membrane showed an increase in lux and an improvement in the rejection of impurities as compared with the base membrane. The additional advantage was a reduced membrane fouling after treatment with HF (Gonzalez-Munoz et al. 2006). On the other hand, it was shown that despite the modiication of many membrane characteristics, such as charge, hydrophilicity, porosity, and pore size, by hydrolysis and oxidation of the CA membrane, neither treatment prevented HA adsorption on the modiied membranes (Combe et al. 1999). Liu et al. (2010) modiied blended chitosan (CS)/CA membranes via surface treatment with heparin or a quaternary ammonium to change the hydrophilicity and the membrane charge or via a reaction with AgNO3 to introduce a biocide on the membranes. The reaction of the heparin with the CS/CA base membrane was through the formation of a polycation–polyanion complex, where the –CH2SO3− and –NHSO3− groups in the heparin interacted with the –NH3+ groups in the CS (CS/CA-H membranes). The attachment of quaternary ammonium to the base membrane was realized via both the –CH 2OH and the –NH2 positions on the CS polymer chains (CS/CA-QN). Silver ions were loaded onto the membrane through surface complexation with the amine groups in CS and through physical adsorption (CS/CA-Ag or CS/CA-H/Ag membranes). It was shown that membranes modiied with heparin or the quaternary ammonium became much more hydrophilic, with a water contact angle for CS/CA-H of 42.3° and that for CS/CA-QN of 39.8°, compared with 69.6° for that of the base CS/CA membrane. The ζ-potential of the CS/CA membrane was relatively small, while the CS/CA-H membrane had negative ζ-potentials at around –10 mV and the CS/CA-Ag membrane had positive ζ-potentials at about +12 mV. As can be seen in Figure 3.7, the CS/CA and CS/CA-Ag membranes had large numbers of bacterial cells on the surfaces, in contrast to the CS/CA-H and CS/CA-H/Ag

64

Membrane Modification: Technology and Applications

4h

1d

7d (a)

(b)

(c)

(d)

FIGURE 3.7 Bacterial adhesion and growth on the (a) CS/CA, (b) CS/CA-H, (c) CS/CA-Ag, and (d) CS/CA-H/Ag membranes after immersion in an activated sludge bioreactor for a period of up to 7 days. (Reprinted from Journal of Membrance Science, 346, Liu, C.X., et al., 121–130, Copyright (2010), with permission from Elsevier.)

membranes, after immersion in an activated sludge bioreactor for several periods of time. The high hydrophobicity of the CS/CA membrane and the strong attractive electrochemical interactions between the CS/CA-Ag membrane and the bacteria probably promoted and contributed signiicantly to the adhesion of the bacterial cells on these membrane surfaces. After 1 week of immersion, a number of bacterial locs were also observed on the surface of the CS/CA-H membrane. Some of the initially adhered bacteria, even though at a very small number, may eventually grow and develop into the observed bacteria ilms because the membrane did not have an antibacterial function. However, the CS/CA-H/Ag membrane was still very clean after 1 week. Even if a small number of bacteria adhered on the surface, they were killed by the loaded silver ions, so could not grow on this membrane. Thus, the best performance for minimizing biofouling has been realized when the highly effective antiadhesion function of the CS/CA-H membrane was supplemented by the antibacterial properties of the CS/CA-H/Ag membrane.

Reduction of Membrane Fouling by Polymer Surface Modification

3.8

65

SURFACE MODIFICATION OF POLYMER MEMBRANES WITH NANOPARTICLES

The use of nanoparticles in preparing and modifying polymeric membranes has received much attention during the last few years in the attempts to enhance lux and reduce fouling (Zhao and Stevens 1998; Kwak et al. 2001; Kim et al. 2003; Yu et al. 2003; Son et al. 2004; Luo et al. 2005; Bae and Tak 2005; Bae et al. 2006; Chou et al. 2005; Yan et al. 2005, 2009; Lee et al. 2007, 2008; Li et al. 2007, 2009a,b; Kochkodan et al. 2008; Rahimpour et al. 2008; Mansourpanah et al. 2009; Zodrow et al. 2009; Kim and der Bruggen 2010). Two different methods are used for preparing nanoparticle-based membranes. One is the entrapment of the nanoparticles in a polymer matrix via a phase inversion method by the addition of the nanoparticles to a casting solution (Li et al. 2009a; Kim and der Bruggen 2010). The other is the deposition of the nanoparticles on the membrane surface via dipping the porous support in an aqueous suspension of nanoparticles (Kim et al. 2003). In this section, we will focus mainly on the preparation and antifouling properties of membranes modiied with deposited nanoparticles. Kwak et al. (2001) performed one of the irst studies in this ield. TiO2 nanoparticles of approximate size of 2 nm were immobilized via self-assembly with the terminal functional groups on the surface of the RO PA membrane. X-ray photoelectron spectroscopy demonstrated quantitatively that TiO2 particles were tightly self-assembled with a suficient bonding strength to the membrane, which meant that particles could withstand various washing procedures and RO operating conditions. The self-assembly mechanism of ixing TiO2 on the membrane surface with COOH functional groups may include bonding with the two oxygen atoms of the carboxylate group via a bidentate coordination to Ti4+ cations or through the formation of a hydrogen bond between a carbonyl group and the surface hydroxyl group of TiO2 (Kim et al. 2003). The self-assembly procedure was also used by Bae et al. (2006) for modifying sulfonated PES membranes and by Mansourpanah et al. (2009) for coating PES/PI blend membrane and the OH functionalized PES/PI membrane with TiO2 nanoparticles. Luo et al. (2005) also applied a similar approach to the deposition of TiO2 nanoparticles onto the PES membranes. For the PES membranes, self-assembly can be due to the coordination of the sulfone group and the ether bond to Ti4+, or by a hydrogen bond between the sulfone group and the ether bond and a surface hydroxyl group of TiO2 due to the strong electronegativity of oxygen in the ether bond and the sulfone group of the PES (Figure 3.8). The self-assembly of the TiO2 nanoparticles on a membrane surface is usually realized by dipping the porous membrane support in a colloidal suspension of TiO2. The concentration of the aqueous colloidal suspension of TiO2 may vary from 0.01 through 0.03 wt% (Mansourpanah et al. 2009) to 1 wt% (Luo et al. 2005; Li et al. 2009a), while the time of immersion of the porous supports in the suspension was suggested to be 1 min (Bae et al. 2006), 1 h (Luo et al. 2005), and 1 week (Li et al. 2009b). Rahimpour et al. (2008) studied the effect of dipping time in a 0.03 wt% TiO2 colloidal suspension by comparing 15, 30, and 60 min of dipping. They concluded that a 15 min immersion yielded the best performance in terms of permeability and hypothesized that longer dipping times led to more pore plugging.

66

Membrane Modification: Technology and Applications O

I

S

O

O Ti4+

Ti4+

O

O

IІ O

S

HO

O

Ti4+– O

HO 4+

Ti

FIGURE 3.8 The mechanism of self-assembly of TiO2 nanoparticles: (I) by coordination of a sulfone group and an ether bond to Ti4+ and (II) by a H-bond between the sulfone group and an ether bond and the surface hydroxyl group of TiO2. (Reprinted from Applied Surface Science, 249, Luo, M.J., et al., 76–84, Copyright (2005), with permission from Elsevier.)

Lee et al. (2008) applied an in situ interfacial polymerization procedure on the PES support for preparing composite nanoparticle-based membranes. In this procedure, commercial TiO2 nanoparticles of 30 nm were dispersed in an organic trimesoyl chloride (TMC) solution. The PES support was irst immersed in aqueous m-phenyl diamine with 0.05 wt% NaOH; the excess reagent was removed from the surface so that a controlled reaction was obtained on subsequent immersion in the solution of TMC in 1,1-dichloro-1-luoroethane. As a result, a thin modiied layer with immobilized nanoparticles was obtained on the surface of the PES support. Bae and Tak (2005) prepared two types of TiO2-immobilized UF membranes (TiO2 entrapped and deposited) and applied them to an activated sludge iltration, to evaluate their fouling mitigation effect. It was shown that the entrapment of the TiO2 nanoparticles in the membranes increased the hydrophilicity of their surfaces. The water contact angles were changed from 87.6° for neat PS to 73.1° for PS-TiO2, from 86.7° for PVDF to 81.1° for PVDF-TiO2, and from 45.5° for PAN to 43.1° for PAN-TiO2 membranes. Such hydrophilization leads to a reduction in membrane fouling during the iltration of activated sludge, which contains a great number of different organic and microbiological foulants. The TiO2-entrapped membrane showed a lower lux decline than the neat PS membrane. On the other hand, the TiO2-deposited membrane showed a greater fouling mitigation effect compared with that of the TiO2-entrapped membrane. Obviously, the degree of fouling mitigation is mainly affected by the surface area of the TiO2 nanoparticles, which are located

Reduction of Membrane Fouling by Polymer Surface Modification

67

on the membrane surface and are exposed to feed solution. In the case of the TiO2deposited membrane, the degree of surface modiication was higher than that for the TiO2-entrapped membrane, and the fouling mitigation effect signiicantly improved. The cake layer resistance of the modiied membrane, which is a major inluence on membrane fouling during the iltration of the activated sludge, was dramatically decreased compared with that of the initial PES membrane (Bae et al. 2006). As the introduction of nanoparticles increases the hydrophilicity of the polymeric membrane surfaces, the adsorbed foulants on the modiied membranes can be more readily dislodged by shear force than those on the unmodiied PES membranes. As a result, the hydrophilic modiication of the membrane surface by the introduction of the TiO2 nanoparticles inhibits the hydrophobic interactions between the organic foulants and the membrane surface. The increased hydrophilicity of the PES and PVDF membranes modiied with TiO2 results in improved permeability and antifouling ability compared with virgin membranes during iltration of the PEG-500 and BSA solutions, respectively (Li et al. 2009). It has been demonstrated that the antifouling potential of the TiO2-modiied membranes is much better realized with the application of UV irradiation (Mansourpanah et al. 2009). Rahimpour et al. (2008) compared TiO2-entrapped PES membranes and self-assembled TiO2-coated membranes with and without UV irradiation during iltration of nonskim milk. The initial pure water lux and the milk water permeation of the TiO2-entrapped membranes were low compared with the unmodiied PES membrane. However, the antifouling property and the longterm lux stability were signiicantly enhanced. UV illumination further improved the membrane performance and antifouling properties, and the UV-irradiated TiO2deposited membranes had increased lux and higher antifouling properties compared with the TiO2-entrapped membranes. The authors believed that the membranes with TiO2 nanoparticles on their surface and radiated by UV light obtained two main characteristics, namely, photocatalytic properties to decompose the organic compounds adsorbed on the membrane surface and superhydrophilicity that results in a decrease in the contact angle (Rahimpour et al. 2008). Therefore, foulants such as fats and proteins may be decomposed by photocatalysis and then removed from the surface by the feed low. Furthermore, with the increase in the membrane hydrophilicity, there is a competition between the adsorption of water and the foulant molecules, which leads to improved removal of the pollutants from the membrane surface. The photocatalytic properties and enhanced hydrophilicity of the TiO2-modiied NF PES/polyimide (PI) blend membrane and the OH-functionalized PES/PI blend membranes meant that they were also less fouled with BSA solutions (Mansourpanah et al. 2009). Al2O3 nanoparticles have also been used for reducing the organic fouling of polymeric membranes. Al2O3-PVDF UF membranes have been applied to the oil–wastewater treatment (Yan et al. 2005, 2009). The modiied membranes had an improved antifouling performance, and the lux recovery for these membranes reached up to 100% after washing with a 0.1% solution of OP-10 surfactant. The authors suggested that this is a result of hydrophilicity, with signiicantly decreased contact angles for the modiied membranes.

68

Membrane Modification: Technology and Applications

Quantity of P. putida cells, CFU/membrane

1.0E + 05

1.0E + 04

1 1.0E + 03

2 3 4

1.0E + 02 0

2

4

6

t (h)

FIGURE 3.9 The number of P. putida cells on the surface of the PS membranes vs. the time of black UV irradiation: (1) base membrane in the dark; (2) membrane with deposited TiO2 in the dark; (3) base membrane treated by UV; and (4) membrane with deposited TiO2 treated by UV. (Reprinted from Desalination, 220, Kochkodan, V., et al., 380–386, Copyright (2008), with permission from Elsevier.)

The photocatalytic bactericidal effect of the composite membranes with the deposited TiO2 was examined by determining the survival ratios of E. coli (Kim et al. 2003) and Pseudomonas putida cells (Kochkodan et al. 2008) with and without black UV illumination. As can be seen in Figure 3.9, a sharp drop in the number of P. putida cells on the membrane surface with the deposited TiO2 particles was found after UV irradiation. The mechanism of the bactericidal action of TiO2 under black UV light is based on the formation of OH•, O2−•, and HO2• radicals in water (Salih 2002). The adhesion of the bacterial cells to the TiO2 particles controlled by the hydrophobic and charge interactions allows the active oxygen-containing species to reach and damage the bacterial cell wall. Due to the strong photobactericidal properties under UV treatment, the modiied membranes are capable of inhibiting the growth of microorganisms on their surface and thus membrane biofouling is reduced. This reduction in biofouling was demonstrated when the membranes were used for surface water treatment as the luxes of the modiied membranes were 1.7–2.3 times higher compared with those for the control samples (Kochkodan et al. 2008). Membrane biofouling may also be reduced via surface membrane modiication with Ag nanoparticles. Silver-loading PAN hollow ibers were prepared via the dry jet–wet spinning technique from a dope containing 0.5 wt% silver nitrate (Yu et al. 2003). It was shown that at an Ag loading of 0.1 wt%, bacterial growth for both E. coli and Staphylococcus aureus was not observed on the membrane surface. The antibacterial activity of the modiied membrane is attributed to trace amounts of silver ions released from the iber (Zhao and Stevens 1998). The interaction between the silver ions and the bacteria can change the metabolic activity of the bacteria and prohibit the growth of bacteria. However, after lushing with water for 60 days,

Reduction of Membrane Fouling by Polymer Surface Modification

69

the silver content in the hollow ibers decreased from 0.19 to 0.1 wt%, while still keeping the antibacterial activity of E. coli and S. aureus. During the membrane process operation, the permeating water reduces the silver content of the hollow ibers, thus requiring periodical replenish. A dramatic effect on Pseudomonas fouling was observed when the silver nanoparticles were immobilized on a thin-ilm composite PA membrane (Lee et al. 2007). SEM measurements conirmed that all Pseudomonas cells were made inactive on the modiied-membrane surface, while water luxes and salt rejections remained unchanged. High antibacterial activity toward E. coli and S. aureus was also found with CA membranes modiied with Ag nanoparticles (Chou et al. 2005). However, a signiicant loss of silver was found as a result of water permeation, and the antibacterial activity of the membranes disappeared after 5 days (Son et al. 2004). The loss of the entrapped silver nanoparticles was also reported for modiied PS membranes, which have a high antimicrobial activity toward E. coli, P. mendocina, and the MS2 bacteriophage (Zodrow et al. 2009). In general, it may be concluded that despite the endeavor described earlier to develop low-fouling membranes via surface modiication with nanoparticles, further research is still needed to investigate the combined effects of the water chemistry, the nature of the nanoparticles, and the coating conditions on the modiied-membrane performance and fouling mitigation. Also, careful control and monitoring of the nanoparticles released from the modiied membranes are necessary to minimize potential environment (eco) toxicity effects (Tiede et al. 2009).

3.9

CONCLUSION

As can be seen from the presented data, various surface modiication methods, such as photochemical and redox grafting, plasma treatment, physical coating with hydrophilic polymers, chemical reactions on the membrane surface, and immobilization of nanoparticles, have been more or less successful over the last few decades in reducing organic membrane fouling and biofouling. It should be noted that each modiication method has advantages and disadvantages. Of these, the membrane surface modiication techniques, the UV/redox initiating graft polymerization, and the physical coating of the membranes with hydrophilic polymer layers have the advantages of simplicity, low cost of operation, mild reaction conditions, and a possibility of incorporation into the end stages of a membrane manufacturing process. Numerous studies have shown that increasing the membrane hydrophilicity or imposing surface charge and bactericide groups on the membrane surface may reduce membrane fouling with organic compounds and microorganisms. Unfortunately, the antifouling properties of the modiied membranes were evaluated mainly with model solutions of single organic compounds, such as BSA and HA, or model bacterial suspensions, such as E. coli, in short-time laboratory tests. There is a lack of studies on the estimation of the antifouling properties of the surface-modiied membranes with real polycomponent feed streams, especially for their long-term applications. Therefore, further research is still needed in this ield. It should also be mentioned that the development of absolutely nonfouling membranes seems to

70

Membrane Modification: Technology and Applications

be an extremely dificult task, if indeed possible, because apart from the surface properties of the membrane, fouling also depends on the operating conditions in the membrane elements and the feed stream characteristics, such as concentration and pH. Nevertheless, a combination of targeted tailoring of the membrane properties via surface modiication in combination with optimization of the operating conditions and feed stream pretreatment, undoubtedly, will reduce membrane fouling, thus providing an effective and long-term membrane performance.

REFERENCES Abu Seman, M.N., Khayet, M., Bin Ali, Z.L. and Hilal, N. 2010. Reduction of nanoiltration membrane fouling by UV-irradiated graft polymerization technique. J. Memb. Sci. 355: 133–141. Al-Amoudi, A.S. and Farooque, A.M. 2005. Performance, restoration and autopsy of NF membranes used in seawater pretreatment. Desalination 178: 261–271. Al-Amoudi, A.S. and Lovitt, R.W. 2007. Fouling strategies and the cleaning system of NF membranes and factors affecting cleaning eficiency. J. Memb. Sci. 303: 4–28. Asatekin, A., Menniti, A., Kang, S., Elimelech, M., Morgenroth, E. and Mayes, A.M. 2006. Antifouling nanoiltration membranes for membrane bioreactors from self-assembling graft copolymers. J. Memb. Sci. 285: 81–89. Asatekin, A., Olivetti, E.A. and Mayes, A.M. 2009. Fouling resistant, high lux nanoiltration membranes from polyacrylonitrile-graft-poly(ethylene oxide). J. Memb. Sci. 332: 6–12. Ba, C., Ladner, D.A. and Economy, J. 2010. Using polyelectrolyte coatings to improve fouling resistance of a positively charged nanoiltration membrane. J. Memb. Sci. 347: 250–259. Bae, T.H. and Tak, T.M. 2005. Effect of TiO2 nanoparticles on fouling mitigation of ultrailtration membranes for activated sludge iltration. J. Memb. Sci. 249: 1–8. Bae, T.H., Kim, I.C. and Tak, T.M. 2006. Preparation of fouling-resistant TiO2 self-assembled nanocomposite membranes. J. Memb. Sci. 275: 1–5. Baker, J.S. and Dudley, L.Y. 1998. Membrane biofouling. Desalination 118: 81–90. Baker, R.W. 2004. Membrane Technology and Applications. John Wiley: New York. Belfer, S., Purison, Y., Fanshtein, R., Radchenko, Y. and Kedem, O. 1998a. Surface modiication of commercial composite polyamide reverse osmosis membranes. J. Memb. Sci. 139: 175–181. Belfer, S., Purison, Y. and Kedem, O. 1998b. Surface modiication of commercial polyamide reverse osmosis membranes by radical grafting: An ATR-FT-IR study. Acta Polym. 49: 574–582. Belfer, S., Gilron, J., Purison, Y., et al. 2001. Effect of surface modiication of commercial SWRO membranes at Eilat seawater desalination plant. Desalination 139: 169–178. Bequet, S., Remigy, J.C., Rouch, J.C., Espenan, J.M., Clifton, M. and Aptel, M. 2002. From ultrailtration to nanoiltration hollow iber membranes: A continuous UV-photografting process. Desalination 144: 9–14. Bharwadal, U.J., Summerield, J.M., Coker, S.D. and Marsh, T.A. 2000. Winning the battle against biofouling of reverse osmosis membranes. Desalination Water Reuse 10: 53–54. Boributh, S., Chanachai, A. and Jiraratananon, R. 2009. Modiication of PVDF membrane by chitosan solution for reducing protein fouling. J. Memb. Sci. 342: 97–104. Brink, L.E.S. and Romijn, D.J. 1990. Reducing the protein fouling of polysulfone surfaces and polysulfone ultrailtration membranes: Optimization of the type of presorbed layer. Desalination 78: 209–233.

Reduction of Membrane Fouling by Polymer Surface Modification

71

Bruening, M.L., Dotzauer, D.M., Jain, P., Ouyang, L. and Baker, G.L. 2008. Creation of functional membranes using polyelectrolyte multilayers and polymer brushes. Langmuir 24: 7663–7673. Bryjak, M., Hodge, H. and Dach, B. 1998. Modiication of porous polyacrylonitrile membranes. Angew Makromol. Chem. 260: 25–29. Bryjak, M., Gancarz, I. and Pozniak, G. 1999. Surface evaluation of plasma-modiied polysulfone (Udel P-1700) ilms. Langmuir 15: 6400–6404. Chen, H. and Belfort, G. 1999. Surface modiication of poly(ether sulfone) ultrailtration membranes by low temperature plasma-induced graft polymerization. J. Appl. Polym. Sci. 72: 1699–1711. Chou, W.L., Yu, D.G. and Yang, M.C. 2005. The preparation and characterization of silverloading cellulose acetate hollow iber membrane for water treatment. Polym. Adv. Technol. 16(8): 600–607. Combe, C., Molis, E., Lucas, P., Riley, R. and Clark, M.M. 1999. The effect of CA membrane properties on adsorptive fouling by humic acid. J. Membr. Sci. 154: 73–87. Dattatray, S., Wavhal, D.S. and Fisher, E.R. 2005. Modiication of polysulfone ultrailtration membranes by CO2 plasma treatment. Desalination 172: 189–205. Dong, B., Jiang, H., Manolache, S., Lee Wong, A.C. and Denes, F.S. 2007. Plasma-mediated grafting of poly(ethylene glycol) on polyamide and polyester surfaces and evaluation of antifouling ability of modiied substrates. Langmuir 23: 7306–7313. Du, J.R., Peldszus, S., Huck, P.M. and Feng, X. 2009. Modiication of poly(vinylidene luoride) ultrailtration membranes with poly(vinyl alcohol) for fouling control in drinking water treatment. Water Res. 43: 4559–4568. Duputell, D. and Staude, E. 1993. Heterogeneous modiication of ultrailtration membranes made from poly(vinylidene luoride) and their characterization. J. Memb. Sci. 78: 45–51. Elimelech, M., Zhu, X., Childress, A.E. and Hong, S. 1997. Role of membrane surface morphology in colloidal fouling of cellulose acetate and composite aromatic polyamide reverse osmosis membranes. J. Memb. Sci. 127: 101–109. Escobar, I.C., Hoek, E.M., Gabelich, C.J., et al. 2005. Committee report: recent advances and research needs in membrane fouling. J. Am. Water Works Assoc. 97: 79–89. Fane, A.G. and Fell, C.J.D. 1987. A review of fouling and fouling control in ultrailtration. Desalination 62: 117–136. Flemming, H.C. 1997. Reverse osmosis membrane biofouling. Exp. Therm. Fluid Sci. 14: 382–391. Flemming, H.C., Schaule, G., Griebe, T., Schmitt, J. and Tamachkiarowa, A. 1997. Biofouling – The Achilles heel of membrane processes. Desalination 113: 215–225. Freger, V., Gilron, J. and Belfer, S. 2002. TFC polyamide membranes modiied by grafting of hydrophilic monomers: an FT-IR/AFM/TEM study. J. Memb. Sci. 209: 283–292. Fujimoto, K., Tadokoro, H., Ueda, Y. and Ikada, Y. 1993. Polyurethane surface modiication by graft polymerization of acrylamide for reduced protein adsorption and platelet adhesion. Biomaterials 14: 442–448. Gancarz, I., Bryjak, M. and Pozniak, G. 1999a. Modiication of polysulfone membrane. 1.CO2 plasma treatment. Eur. Polym. J. 35: 1419–1428. Gancarz, I., Pozniak, G., Bryjak, M. and Frankiewiez, A. 1999b. Modiication of polysulfone membranes. 2. Plasma grafting and plasma polymerization of acrylic acid. Acta Polym. 50: 317–326. Gancarz, I., Pozniak, G., Bryjak, M. and Tylus, W. 2002. Modiication of polysulfone membranes. 5. Effect of n-butylamine and allylamine plasma. Eur. Polym. J. 38: 1937–1946. Gancarz, I., Bryjak, J., Bryjak, M. and Pozniak, G. 2003. Plasma modiied polymers as a support for enzyme immobilization. 1. Allyl alcohol plasma. Eur. Polym. J. 39: 1615–1622. Gonzalez-Munoz, M.P., Navarro, R., Saucedo, I., et al. 2006. Hydroluoric acid treatment for improved performance of a nanoiltration membrane. Desalination 191: 273–278.

72

Membrane Modification: Technology and Applications

Gu, J.S., Yu, H.Y., Huang, L., et al. 2009. Chain-length dependence of the antifouling characteristics of the glycopolymer-modiied polypropylene membrane in a SMBR. J. Memb. Sci. 326: 145–152. Gullinkala, T. and Escobar, I. 2010. A green membrane functionalization method to decrease natural organic matter fouling. J. Memb. Sci. 360: 155–164. Hasson, D., Drak, A. and Semiat, R. 2001. Inception of CaSO4 scaling on RO membranes at various water recovery levels. Desalination 139: 73–81. He, X.C., Yu, H.Y., Tang, Z.Q., et al. 2009. Reducing protein fouling of a polypropylene microporous membrane by CO2 plasma surface modiication. Desalination 244: 80–89. Hilal, N., Al-Khatib, L., Atkin, B.P., Kochkodan, V. and Potapchenko, N. 2003. Photochemical modiication of membrane surfaces for (bio)fouling reduction: a nano-scale study using AFM. Desalination 158: 65–72. Hilal, N., Kochkodan, V., Al-Khatib, L. and Levadna, T. 2004. Surface modiied polymeric membranes to reduce (bio)fouling: a microbiological study using E. coli. Desalination 167: 293–300. Hilal, N., Ogunbiyi, O.O., Miles, N.J. and Nigmatullin, R. 2005. Methods employed for control of fouling in MF and UF membranes: A comprehensive review. Sep. Sci. Technol. 40: 1957–2005. Hirose, M., Ito, H. and Kamiyama, Y. 1996. Effect of skin layer surface structures on the lux behaviour of RO membranes. J. Memb. Sci. 121: 209–215. Hong, S. and Elimelech, M. 1997. Chemical and physical aspects of natural organic matter (NOM) fouling of nanoiltration membranes. J. Memb. Sci. 132: 159–181. Hu, M.X., Yang, Q. and Xu, Z.K. 2006. Enhancing the hydrophilicity of polypropylene microporous membranes by the grafting of 2-hydroxyethyl methacrylate via a synergistic effect of photoinitiators. J. Memb. Sci. 285: 196–205. Kaeselev, B., Pieracci, J. and Belfort, G. 2001. Photoinduced grafting of ultrailtration membranes: comparison of poly(ether sulfone) and poly(sulfone). J. Memb. Sci. 194: 245–261. Kaeselev, B., Kingshott, J. and Jonsson, G. 2002. Inluence of the surface structure on the iltration performance of UV-modiied PES membranes. Desalination 146: 265–271. Kato, K., Uchida, E., Kang, E.T., Uyama, Y. and Ikada, Y. 2003. Polymer surface with graft chains. Prog. Polym. Sci. 28: 209–259. Kharlampieva, E. and Sukhishvili, S.A. 2003. Ionization and pH stability of multilayers formed by self-assembly of weak polyelectrolytes. Langmuir 19: 1235–1243. Kilduff, J.E., Mattaraj, S., Pieracci, J.P. and Belfort, G. 2000. Photochemical modiication of poly(ether sulfone) and sulfonated poly(sulfone) nanoiltration membranes for control of fouling by natural organic matter. Desalination 132: 133–142. Kim, J. and Van der Bruggen, V.B. 2010. The use of nanoparticles in polymeric and ceramic membrane structures: Review of manufacturing procedures and performance improvement for water treatment. Environ. Pollut. 158: 2335–2349. Kim, K.J., Fane, A.G. and Fell, C.J.D. 1988. The performance of ultrailtration membranes pretreated by polymers. Desalination 70: 229–249. Kim, K.S., Lee, K.H., Cho, K. and Park, C.E. 2002. Surface modiication of polysulfone ultrailtration membrane by oxygen plasma treatment. J. Memb. Sci. 199: 135–145. Kim, S.H., Kwak, S.Y., Sohn, B.H. and Park, T.H. 2003. Design of TiO2 nanoparticle selfassembled aromatic polyamide thin-ilm-composite (TFC) membrane as an approach to solve biofouling problem. J. Memb. Sci. 211: 157–165. Kimura, K., Hane, Y., Watanabe, Y., Amy, G. and Ohkuma, N. 2004. Irreversible membrane fouling during ultrailtration of surface water. Water Res. 38: 3431–3441. Kochkodan, V.M., Hilal, N., Goncharuk, V.V., Al-Khatib L. and Levadna, T.I. 2006. Effect of the surface modiication of polymer membranes on their microbiological fouling. Colloid J. 68: 267–273.

Reduction of Membrane Fouling by Polymer Surface Modification

73

Kochkodan, V., Tsarenko, S., Potapchenko, N., Kosinova, V. and Goncharuk, V. 2008. Adhesion of microorganisms to polymer membranes: A photobactericidal effect of surface treatment with TiO2. Desalination 220: 380–385. Kull, K.R., Steen, M.L. and Fisher, E.R. 2005. Surface modiication with nitrogencontaining plasmas to produce hydrophilic, low-fouling membranes. J. Memb. Sci. 246: 203–215. Kwak, S.Y., Kim, S.H. and Kim, S.S. 2001. Hybrid organic/inorganic reverse osmosis (RO) membrane for bactericidal anti-fouling. 1. Preparation and characterization of TiO2 nanoparticle self-assembled aromatic polyamide thin-ilm-composite (TFC) membrane. Environ. Sci. Technol. 35: 2388–2394. Lee, H.S., Im, S.J., Kim, J.H., Kim, H.J., Kim, J.P. and Min, B.R. 2008. Polyamide thin-ilm nanoiltration membranes containing TiO2 nanoparticles. Desalination 219: 48–56. Lee, S.Y., Kim, H.J., Patel, R., Im, S.J., Kim, J.H. and Min, B.R. 2007. Silver nanoparticles immobilized on thin ilm composite polyamide membrane: characterization, nanoiltration, antifouling properties. Polym. Adv. Technol. 18(7): 562–568. Li, J.B., Zhu, J.W. and Zheng, M.S. 2007. Morphologies and properties of poly(phthala-zinone ether sulfone ketone) matrix ultrailtration membranes with entrapped TiO2 nanoparticles. J. Appl. Polym. Sci. 103: 3623–3629. Li, J.F., Xu, Z.L., Yang, H., Yu, L.Y. and Liu, M. 2009a. Effect of TiO2 nanoparticles on the surface morphology and performance of microporous PES membrane. Appl. Surf. Sci. 255: 4725–4732. Li, J.H., Xu, Y.Y., Zhu, L.P., Wang, J.H. and Du, C.H. 2009b. Fabrication and characterization of a novel TiO2 nanoparticle self-assembly membrane with improved fouling resistance. J. Membr. Sci. 326: 659–666. Li, N., Liu, Z. and Xu, S. 2000. Dynamically formed poly(vinyl alcohol) ultrailtration membranes with good anti-fouling characteristics. J. Memb. Sci. 169: 17–28. Liu, C.X., Zhang, D.R., He, Y., Zhao, X.S. and Bai, R. 2010. Modiication of membrane surface for anti-biofouling performance: Effect of anti-adhesion and anti-bacteria approaches. J. Memb. Sci. 346: 121–130. Luo, M.J., Zhao, J.Q., Tang, W. and Pu, C.S. 2005. Hydrophilic modiication of poly(ether sulfone) ultrailtration membrane surface by self-assembly of TiO2 nanoparticles. Appl. Surf. Sci. 249: 76–84. Ma, H., Bowman, C.N. and Davis, R.H. 2000a. Membrane fouling reduction by backpulsing and surface modiication. J. Memb. Sci. 173: 191–200. Ma, H., Davis, R.H. and Bowman, C.N. 2000b. A novel sequential photoinduced living graft polymerization. Macromolecules 33: 331–335. Maartens, A., Jacobs, E.P. and Swart, P. 2002. UF of pulp and paper efluent: Membrane fouling-prevention and cleaning. J. Memb. Sci. 209: 81–92. Mansourpanah, Y., Madaeni, S.S., Rahimpour, A., Farhadian, A. and Taheri, A.H. 2009. Formation of appropriate sites on nanoiltration membrane surface for binding TiO2 photo-catalyst: Performance, characterization, and fouling-resistant capability. J. Memb. Sci. 330: 297–306. Nilson, J.A. and DiGiano, F.A. 1996. Inluence of NOM composition on nanoiltration. J. Am. Water Works Assoc. 88: 53–66. Nystrom, M. 1989. Fouling of unmodiied and modiied polysulfone ultrailtration membranes by ovalbumin. J. Memb. Sci. 44: 183–196. Pasmore, M., Todd, P., Smith, S., et al. 2001. Effect of ultrailtration membranes surface properties on Pseudomonas aeruginosa bioilm initiation for the purpose of reducing biofouling. J. Memb. Sci. 194: 15–32. Pieracci, J., Crivello, J.V. and Belfort, G. 1999. Photochemical modiication of 10 kDa polyethersulfone ultrailtration membranes for reducing of biofouling. J. Memb. Sci. 156: 223–240.

74

Membrane Modification: Technology and Applications

Pieracci, J., Wood, D.W., Crivello, J.V. and Belfort, G. 2000. UV-assisted graft polymerization of N-vinyl-2-pyrrolidone onto poly(ether sulfone) ultrailtration membranes: Comparison of dip versus immersion modiication techniques. Chem. Mater. 12: 2123–2133. Pieracci, J., Crivello, J.V. and Belfort, G. 2002. UV-assisted graft polymerization of N-vinyl2-pyrrolidinone onto poly(ether sulfone) ultrailtration membranes using selective UV wavelengths. Chem. Mater. 14: 256–265. Potts, D.E., Ahlert, R.C. and Wang, S.S. 1981. A critical review of fouling of reverse osmosis membranes. Desalination 36: 235–264. Pozniak, G., Gancarz, I. and Tylus, W. 2006. Modiied poly(phenylene oxide) membranes in ultrailtration and micellar-enhanced ultrailtration of organic compounds. Desalination 198: 215–224. Rahimpour, A., Madaeni, S.S., Taheri, A.H. and Mansourpanah, Y. 2008. Coupling TiO2 nanoparticles with UV irradiation for modiication of polyethersulfone ultrailtration membranes. J. Memb. Sci. 313: 158–169. Rahimpour, A., Madaeni, S.S., Zereshki, S. and Mansourpanah, Y. 2009. Preparation and characterization of modiied nano-porous PVDF membrane with high antifouling property using UV photo-grafting. Appl. Surf. Sci. 255: 7455–7461. Rana, D. and Matsuura, T. 2010. Surface modiication for antifouling membranes. Chem. Rev. 110: 2448–2471. Reddy, A.V.R. and Patel, H.R. 2008. Chemically treated polyethersulfone/polyacrylonitrile blend ultrailtration membranes for better fouling resistance. Desalination 221: 318–323. Reddy, A.V.R., Mohan, D.J., Bhattacharya, A., Shah, V.J. and Ghosh, P.K. 2003. Surface modiication of ultrailtration membranes by preadsorption of a negatively charged polymer I. Permeation of water soluble polymers and inorganic salt solutions and fouling resistance properties. J. Memb. Sci. 214: 211–221. Reddy, A.V.R., Trivedi, J.J., Devmurari, C.V., et al. 2005. Fouling resistant membranes in desalination and water recovery. Desalination 183: 301–306. Ridgway, H.F., Rigby, M.G. and Argo, D.G. 1984. Adhesion of a Mycobacterium sp. to cellulose diacetate membranes used in reverse osmosis. Appl. Environ. Microbiol. 47: 202–218. Roudman, A.R. and DiGiano, F.A. 2000. Surface energy of experimental and commercial nanoiltration membranes: Effects of wetting and natural organic matter fouling. J. Memb. Sci. 175: 61–73. Sagle, A.C., Van Wagner, E.M., Ju, H., McCloskey, B.D., Freeman, B.D. and Sharma, M.M. 2009. PEG-coated reverse osmosis membranes: Desalination properties and fouling resistance. J. Memb. Sci. 340: 92–108. Salih, F.M. 2002. Enhancement of solar inactivation of Escherichia coli by titanium dioxide photocatalytic oxidation. J. Appl. Microbiol. 92: 923–929. Scot, K. and Hughes, R. 1996. Industrial Membrane Separation Technology. Blackie Academics & Professional: London. Sheikholeslami, R. 1999. Fouling mitigation in membrane processes. Desalination 123: 45–53. Son, W.K., Youk, J.H., Lee, T.S. and Park, W.H. 2004. Preparation of antimicrobial ultraine cellulose acetate ibers with silver nanoparticles. Macromol. Rapid Commun. 25(18): 1632–1637. Sukhishvili, S.A. and Granick, S. 2002. Layered, erasable polymer multilayers formed by hydrogen-bonded sequential self-assembly. Macromolecules 35: 301–310. Susanto, H., Balakrishnan, M. and Ulbricht, M. 2007. Via surface functionalization by photograft copolymerization to low-fouling polyethersulfone-based ultrailtration membranes. J. Memb. Sci. 288: 157–167.

Reduction of Membrane Fouling by Polymer Surface Modification

75

Taniguchi, M. and Belfort, G. 2004. Low protein fouling synthetic membranes by UV-assisted surface grafting modiication: varying monomer type. J. Memb. Sci. 231: 147–157. Tiede, K., Hassellov, M., Breitbarth, E., Chaudhry, Q. and Boxall, A.B.A. 2009. Considerations for environmental fate and ecotoxicity testing to support environmental risk assessments for engineered nanoparticles. J. Chromatogr. A 1216(3): 503–509. Tremblay, A.Y., Tam, C.M. and Guiver, M.D. 1992. Variations in the pore size of charged and noncharged hydrophilic polysulfone membranes. Ind. Eng. Chem. Res. 31: 834–838. Tyszler, D., Zytner, R.G., Batsch, A., et al. 2006. Reduced fouling tendencies of ultrailtration membranes in wastewater treatment by plasma modiication. Desalination 189: 119–129. Ulbricht, M. 1996. Photograft-polymer-modiied microporous membranes with environmentsensitive permeabilities. React. Funct. Polym. 31: 165–177. Ulbricht, M. 2006. Advanced functional polymer membranes. Polymer 47: 2217–2262. Ulbricht, M. and Belfort, G. 1996. Surface modiication of ultrailtration membranes by low temperature plasma. II Graft polymerization onto polyacrylonitrile and polysulfone. J. Memb. Sci. 111: 193–215. Ulbricht, M. and Hicke, H.G. 1993. Photomodiication of ultrailtration membranes. 1. Photochemical modiication of polyacrylonitrile ultrailtration membranes with aryl azides. Angew Makromol. Chem. 210: 69–95. Ulbricht, M., Matuschewski, H., Oechel, A. and Hicke, H.G. 1996. Photo-induced graft polymerization surface modiications for the preparation of hydrophilic and lowprotein-adsorbing ultrailtration membranes. J. Memb. Sci. 115: 31–47. Van de Lisdonk, C.A.C., Van Paassen, J.A.M. and Schippers, J.C. 2000. Monitoring scaling in nanoiltration and reverse osmosis membrane systems. Desalination 132: 101–108. Van der Bruggen, B., Mänttäri, M. and Nyström, M. 2008. Drawbacks of applying nanoiltration and how to avoid them: A review. Sep. Purif. Technol. 63: 251–263. Vrijenhoek, E.M., Hong, S. and Elimelech, M. 2001. Inluence of membrane surface properties on initial rate of colloidal fouling of reverse osmosis and nanoiltration membranes. J. Memb. Sci. 188: 115–128. Vrouwenvelder, H.S., van Paasen, J.A.M., Folmer, H.C., Hofman, J.A.M.H., Nederlof, M.M. and van der Kooij, D. 1998. Biofouling of membranes for drinking water production. Desalination 118: 157–166. Wavhal, D.S. and Fisher, E.R. 2002. Hydrophilic modiication of polyethersulfone membranes by low temperature plasma-induced graft polymerization. J. Memb. Sci. 209: 255–269. Xie, Y., Yu, H., Wang, S. and Xu, Z. 2007. Improvement of antifouling characteristics in a bioreactor of polypropylene microporous membrane by the adsorption of Tween 20. J. Environ. Sci. 19: 1461–1465. Yamagishi, H., Crivello, J.V. and Belfort, G. 1995a. Development of a novel photochemical technique for modifying poly(arylsulfone) ultrailtration membranes. J. Memb. Sci. 105: 237–248. Yamagishi, H., Crivello, J.V. and Belfort, G. 1995b. Evaluation of photochemically modiied poly(arylsulfone) ultrailtration membranes. J. Memb. Sci. 105: 249–259. Yan, L., Li, Y.S. and Xiang, C.B. 2005. Preparation of poly(vinylidene luoride) (PVDF) ultrailtration membrane modiied by nano-sized alumina (Al2O3) and its antifouling research. Polymer 46: 7701–7706. Yan, L., Hong, S., Li, M.L. and Li, Y.S. 2009. Application of the Al2O3-PVDF nano-composite tubular ultrailtration (UF) membrane for oily wastewater treatment and its antifouling research. Separ. Purif. Technol. 66: 347–352. Yan, M.G., Liu, L.Q., Tang, Z.Q., et al. 2008. Plasma surface modiication of polypropylene microiltration membranes and fouling by BSA dispersion. Chem. Eng. J. 145: 218–224. Yang, B. and Yang, W. 2003. Photografting modiication of PET nucleopore membranes. J. Macromol. Sci. A Pure Appl. Chem. A 40: 309–320.

76

Membrane Modification: Technology and Applications

Yu, D.G., Teng, M.Y., Chou, W.L. and Yang, M.C., 2003. Characterization and inhibitory effect of antibacterial PAN-based hollow iber loaded with silver nitrate. J. Memb. Sci. 225: 115–123. Yu, H., Cao, Y., Kang, G., Liu, J., Li, M. and Yuan, Q. 2009. Enhancing antifouling property of polysulfone ultrailtration membrane by grafting zwitterionic copolymer via UV-initiated polymerization. J. Memb. Sci. 342: 6–13. Yu, H.Y., Xie, Y.J., Hu, M.X. and Wang, J.L. 2005. Surface modiication of polypropylene microporous membrane to improve its antifouling property in a MBR: CO2 plasma treatment. J. Memb. Sci. 254: 219–227. Yu, H.Y., Xu, Z.K., Xie, Y.J., Liu, Z.M. and Wang, S.Y. 2006. Flux enhancement for polypropylene microporous membrane in a SMBR by the immobilization of poly(N-vinyl-2pyrrolidone) on the membrane surface. J. Memb. Sci. 279: 148–155. Yu, H.Y., He, X.C., Liu, L.Q., Gu, J.S. and Wei, X.W. 2007. Surface modiication of polypropylene microporous membrane to improve its antifouling characteristics in an SMBR: N2 plasma treatment. Water Res. 41: 4703–4709. Yu, H.Y., Liu, L.Q., Tang, Z.Q., Yan, M.G., Gu, J.S. and Wei, X.W. 2008a. Mitigated membrane fouling in an SMBR by surface modiication. J. Memb. Sci. 310: 409–417. Yu, H.Y., Liu, L.Q., Tang, Z.Q., Yan, M.G., Gu, J.S. and Wei, X.W. 2008b. Surface modiication of polypropylene microporous membrane to improve its antifouling characteristics in an SMBR: Air plasma treatment. J. Memb. Sci. 311: 216–224. Zhan, J., Liu, Z., Wang, B. and Ding, F. 2004. Modiication of a membrane surface charge by a low temperature plasma induced grafting reaction and its application to reduce membrane fouling. Sep. Sci. Technol. 39: 2977–2995. Zhang, M., Nguyen, Q.T. and Ping, Z. 2009. Hydrophilic modiication of poly(vinylidene luoride) microporous membrane. J. Memb. Sci. 327: 78–86. Zhao, B. and Brittain, W.J. 2000. Polymer brushes: surface-immobilized macromolecules. Prog. Polym. Sci. 25: 677–710. Zhao, J.G. and Stevens, E.S. 1998. Multiple parameters for the comprehensive evaluation of the susceptibility of Escherichia coli to the silver ion. Biometals 11: 27–35. Zhao, Z.P., Li, J., Wang, D. and Chen, C.X. 2005. Nanoiltration membrane prepared from polyacrylonitrile ultrailtration membrane by low-temperature plasma: 4. grafting of N-vinylpyrrolidone in aqueous solution. Desalination 184: 37–44. Zhou, Y., Yu, S., Gao, C. and Feng, X. 2009. Surface modiication of thin ilm composite polyamide membranes by electrostatic self-deposition of polycations for improved fouling resistance. Sep. Purif. Technol. 66: 287–294. Zodrow, K., Brunet, L., Mahendra, S., et al. 2009. Polysulfone ultrailtration membranes impregnated with silver nanoparticles show improved biofouling resistance and virus removal. Water Res. 43: 715–723.

4

Modifications of Polymeric Membranes by Incorporating Metal/Metal Oxide Nanoparticles Law Yong Ng, Abdul Wahab Mohammad, and Nidal Hilal

CONTENTS 4.1 4.2 4.3

Introduction .................................................................................................... 77 Introduction to Metal and Metal Oxide Nanoparticles................................... 78 Polymeric Membranes Impregnated with a Variety of Nanoparticles ...........80 4.3.1 Fuel Cell Application .......................................................................... 81 4.3.2 Microiltration and Ultrailtration....................................................... 87 4.3.3 Nanoiltration and Reverse Osmosis Membrane ................................97 4.3.4 Gas Separation .................................................................................. 100 4.4 Future Improvements in Fabrication of Polymeric Membrane Incorporated with Nanoparticles .................................................................. 107 4.5 Conclusion .................................................................................................... 107 References .............................................................................................................. 108

4.1

INTRODUCTION

The invention of membrane technology, including microiltration (MF), ultrailtration (UF), nanoiltration (NF), and reverse osmosis (RO), has caused a progressive development in various ields, such as water puriication, gas separation, biotechnology, and fuel cell technology. However, the most important limiting factor or disadvantage of membrane technology is fouling. In the membrane technology ield, fouling of the membrane refers to the accumulation of undesired materials on the membrane surface. Fouling materials are usually composed of living organisms or their derivatives (called biofouling) (Lee et al. 2007b; Kang et al. 2004; Subramani and Hoek 2010) and nonliving substances (called inorganic fouling). Fouling of membranes has been a key issue since the technology’s invention and application. 77

78

Membrane Modification: Technology and Applications

Membrane fouling causes an increase in expenditure in the industry due to shorter membrane life span, more maintenance works required, low eficiency in the separation performances, and higher pumping energy consumption. Thus, research has continued since the discovery of some of the methods used in overcoming fouling problems. Fouling problems can be overcome by pretreatment processes. Pretreatment processes are used to treat the feed solution before the solution enters the membrane separation processes. The common methods applied for feed solution pretreatment include adjusting the solution pH, the addition of complexing agents, heat treatment (Jin et al. 2009; Jawor and Hoek 2009), chlorination, chemical clariication, and iltration with the adsorption of the material onto activated carbon. However, not all the pretreatment processes are suitable for solving all the problems in different applications. For instance, chlorination is not suitable for polymeric membranes since most of the polymeric materials could be deteriorated by the chlorines. The pH adjustment in the pretreatment is used to precipitate certain proteins as the electrolytic environment reaches their isoelectric point; however, in protein mixtures, not all proteins will be precipitated and unwanted removal of materials may occur. Fouling can also be reduced by modifying the process conditions to reduce the effect of the concentration polarization. The process conditions could be modiied by increasing the low velocity or by using turbulence promoters; however, this is expensive. Flow velocity modiication is more suitable and easier to be applied in a wide range of applications with little increase in energy consumption in comparison with turbulence promoters. Membrane cleaning can also used to reduce membrane fouling. The methods used for membrane cleaning are varied and are dependent on the membrane application. However, some of the common membrane-cleaning processes are chemical cleaning using various chemicals such as biocides, mechanical cleaning, and hydraulic cleaning. The last method that should be discussed for controlling fouling, which is also the most attractive and widely investigated, is changing the membrane properties. For instance, when the hydrophilicity of a membrane is improved, its antifouling properties would be better in comparison with the hydrophobic membranes, depending on the type of application. The discussion in this chapter is mostly focused on membrane modiication by incorporating nanoparticles in polymeric membranes in order to modify the membrane properties.

4.2

INTRODUCTION TO METAL AND METAL OXIDE NANOPARTICLES

As the dimensions of a solid particle decrease to the scale of one millionth of a millimeter, the number of atoms constructing the particle is small and in the order of several hundreds or thousands. At this state, the fundamental physical properties, such as the melting point, can change drastically and ceramic materials may be sintered at a lower temperature. Also, as particles get smaller than the wavelength of visible light, they not only become transparent but also emit a special light by plasma absorption. They show completely different electromagnetic or physicochemical

Modifications of Polymeric Membranes

79

properties from their bulk counterparts, although they are made of the same materials. The deinition of nanoparticles differs depending on the materials and applications concerned. In the narrowest sense, they are regarded as particles smaller than 10–20 nm, where the physical properties of the solid materials themselves drastically change. On the other hand, particles in the range from 1 nm to 1 μm may also be called nanoparticles (Hosokawa et al. 2007). Particle size is the most important parameter in practical applications of powder particles. Usually, powder is constituted by particles of various sizes; therefore, it is necessary to obtain not only the mean particle size but also the size distribution. Recently, methods for particle size analysis have been greatly developed, especially analytical techniques with rapid response, high repeatability, and covering a wide range of particle sizes, as in the case of the laser scattering and diffraction methods. Various kinds of nanoparticles, produced by different methods, are applied as raw materials in different ields, including the production of cosmetics, medical supplies, catalysts, pigments, toner, and ink. Nanoparticle incorporation in the design and synthesis of new polymeric materials has been an area of intense research in recent years. Such composite materials have gained a widespread interest in many areas of science and technology due to their remarkable changes in properties, such as mechanical (Okada and Usuki 1995), thermal (Gilman 1999), and magnetic (Godovski 1995) as compared with virgin organic polymers. Thus, one of the latest applications includes the incorporation of nanoparticles into polymeric membranes in order to increase the performances of the membranes, such as the permeability, selectivity, strength, and hydrophilicity. For instance, poly(vinylidene luoride) (PVDF) membranes incorporated with silica nanoparticles can withstand a higher temperature and have higher selectivity and higher diffusivity (Yu et al. 2009); chitosan/zinc oxide nanoparticle composite membranes exhibited good mechanical properties and high antibacterial activities (Li et al. 2010); silica nanoparticles incorporated in polysulfone (PSF) membranes exhibited an enhanced gas permeability (Ahn et al. 2008); polyethersulfone/aluminum oxide membranes exhibited lower lux decline, higher porosity, and pseudo steady-state permeability (Maximous et al. 2009a); and polybenzimidazole/silica nanoparticle composite membranes showed increased permeability and selectivity in gas separation (Sadeghi et al. 2009). Some studies also indicated that the particles added into the polymer matrix might stabilize the polymeric membrane against the changes in the permselectivity with temperatures (Hu et al. 1997). Incorporating nanoparticles into the polymeric membranes has some drawbacks. One of the limiting factors is the dispersion of the nanoparticles in the polymers. The aggregation/dispersion behavior control, which is the irst process for the preparation of new functional materials incorporating nanoparticles, is very dificult for nanoparticles of less than 100 nm in diameter due to the surface interactions. Examples of the surface interaction between particles in the liquid phase are shown in Table 4.1. The surface interaction takes place when the nanoparticles in the polymeric solution fulill the conditions as stated. Although researchers have determined the surface interaction theories, the factors that contribute to enhance or further induce agglomerations remain unclear. This causes dificulty in dispersing the nanoparticles during

80

Membrane Modification: Technology and Applications

TABLE 4.1 Examples of Surface Interaction between Particles in Liquid Phase Surface Interaction Van der Waals interaction Overlap of electric double layer Steric interaction of adsorbed polymer Bridge force Hydration force Depletion

Generation Mechanism Short-range electromagnetic force between molecules and/ or atoms Electrical interaction by the overlap of electric double layer around the particles in solution Short-range interaction by the overlap of adsorbed polymer layer on the particles Formation of the bridge of polymer binder and/or surfactant between the particles Overlap of hydrogen-bonded water molecule on hydrophilic surface of a particle Negative adsorption of solute and polymer by having less afinity for the surface than the solvent

Source: Adapted from Hosokawa, M., Nogi, K., Naito, M., and Yokoyama, T. (eds), Nanoparticle Technology Handbook, 1st edn. Elsevier B.V., AE Amsterdam, 2007.

membrane fabrication. However, Yu et al. (2009a) suggested that an increase in the concentration of the nanoparticles could lead to an increase in nanoparticle agglomeration. In addition, Gilbert et al. (2009) suggested that the ionic strength and the pH of the solution also induce agglomeration of the nanoparticles.

4.3

POLYMERIC MEMBRANES IMPREGNATED WITH A VARIETY OF NANOPARTICLES

Membrane fouling has usually been explained by pore blocking, cake formation, ligand exchange reaction, charge interaction, or hydrophobic interaction (Maximous et al. 2009b). However, membrane material is frequently accepted as one of the predominant fouling modulators, with membrane fouling expected to be more severe with hydrophobic membranes compared with hydrophilic membranes (Yu et al. 2005; Sun et al. 2006). Thus, several strategies to alleviate membrane fouling have been investigated. One of these methods is the hydrophilic modiication of the membrane surfaces (Sheikholeslami 1999). There are various methods available for the hydrophilic modiication of the membrane surfaces (Singh et al. 2005; Nie et al. 2003; Wavhal and Fisher 2003; Ochoa et al. 2003; Nunes et al. 1995; Kim et al. 1988; Asatekin et al. 2006). Among these methods, the incorporation of various kinds of nanoparticles into the polymeric membranes has been the focus of numerous investigations in recent years (Ng et al. 2010). The common preparation methods for the incorporation of nanoparticles into the polymeric membranes can be simpliied as follows. First, casting solutions are prepared by dissolving a certain ratio of polymers and nanoparticles in the solvent (Yu et al. 2009b; Maximous et al. 2009a; Soroko and Livingston 2009; Bae and

Modifications of Polymeric Membranes

81

Tak 2005; Kang et al. 2008b; Kim et al. 2009). A dispersant might be required in order to well disperse the nanoparticles in the casting solutions (Park et al. 2009; Leeuwenburgh et al. 2010; Lee et al. 2007a; Wang et al. 2008). The immersion of the glass plate into a coagulation water bath at room temperature is required for the polymeric membranes prepared by the wet phase inversion methods (Li et al. 2009; Cao et al. 2006; Gao et al. 2009; Blanco et al. 2006; Ren et al. 2004; Zheng et al. 2006; Yang et al. 2006; Li et al. 2008c). The unique chemical and physical properties of nano-sized metals as compared with their bulk particles have led to increased studies of nanoparticle synthesis for speciic optical, magnetic, electronic, and catalytic purposes. Polymeric systems have been used in the preparation of nanoscale particles because of the presence of speciic functional groups on the backbone of the polymer chain. These groups are often ionic in nature or have lone-pair electrons that can serve as a chelating agent as well as impose a stabilizing effect on the synthesized nanoparticles. A typical nanoparticle containing atoms or molecules ranging from tens to tens of thousands has a size ranging between a few and a few tens of nanometers. The nanoparticles introduced to the polymer membranes include silica (Khayet et al. 2005), Fe3O4 (Du et al. 2004), ZrO2(Bottino et al. 2002), TiO2 (Bae and Tak 2005; Kim et al. 2003; Li et al. 2004; Losito et al. 2005), CdS (Trigo et al. 2004), and polymeric nanoparticles (Xu et al. 2002). Each of these nanoparticles can be incorporated with most of the polymeric materials available in order to produce membranes with speciic characteristics, as a result of the synergism of the properties between the polymeric materials and the nanoparticles. Some of the speciic properties of the nanoparticles listed above, which are incorporated into the polymeric membranes, will be discussed further in the following sections.

4.3.1

FUEL CELL APPLICATION

Growing concerns on the depletion of nonrenewable energy resources and climate change have caused fuel cell technologies to become a focus of attention in recent years owing to their high eficiencies and low emissions. Fuel cells, which are classiied according to the electrolytes employed, are electrochemical devices that directly convert chemical energy stored in fuels, such as hydrogen, to electrical energy. Their eficiency can reach as high as 60% in electrical energy conversion and overall 80% in the cogeneration of electrical and thermal energies with more than 90% reduction in major pollutants. Five categories of fuel cells have been the subject of major research efforts. These are: (1) Polymer electrolyte membrane (PEMEM) fuel cells or PEMFCs (also called PEFCs), (2) Solid oxide fuel cells (SOFCs), (3) Alkaline fuel cells (AFCs), (4) Phosphoric acid fuel cells (PAFCs), and (5) Molten carbonate fuel cells (MCFCs). PEMFCs are constructed using PEMEMs as proton conductors and platinum (Pt)-based materials as catalysts. Their noteworthy features include low operating temperature, high power density, and easy scale-up, making PEMFCs promising candidates for the next-generation power sources for transportation, stationary, and portable applications (Wang et al. 2011). Worldwide commercialization of PEMFCs has not yet been realized due to the limitations faced. The problems that have not been solved include durability, cost,

82

Membrane Modification: Technology and Applications

performance decay, membrane mechanical stability, membrane thermal stability, and membrane selectivity. Much research endeavor has concentrated on improving the polymeric membranes used in the fuel cells to overcome these problems. The membrane used in fuel cell applications refers to the layer of polymer electrolyte that conducts protons from the anode to the cathode. Desirable membrane materials are those that exhibit high ionic conductivity, while preventing electron transport and the crossover of hydrogen fuel from the anode and oxygen reactant from the cathode. In addition, they must be chemically stable in an environment with HO– and HOO radicals, thermally stable throughout the operating temperatures, and mechanically robust. The incorporation of various types of nanoparticles into the polymeric membranes is one of the new breakthroughs in improving the membrane performance of fuel cells. One of the polymeric membranes incorporated with iron compounds for improved fuel cells is Naion. Naion is a sulfonated tetraluoroethylene–based luoropolymer– copolymer. It is the irst of a class of synthetic polymers with ionic properties, which are called ionomers. Naion has received a considerable amount of attention as a proton conductor for proton exchange membrane (PEM) fuel cells because of its excellent thermal and mechanical stabilities. Naion-based ilms are also of interest as proton-conducting membranes in direct methanol fuel cells (DMFC). However, unmodiied Naion membranes generally possess a high methanol permeability, which makes them unsuitable for application in the current generation of commercial fuel cells (Gribov et al. 2007). Thus, incorporating nanoparticles of highly acidic inorganic materials into Naion membranes has been proved to be one of the effective approaches to reduce methanol permeability (Arico et al. 2003a,b; Lee et al. 2005; Baglio et al. 2005; Bauer and Willert-Porada 2004; Lin et al. 2006; Daiko et al. 2006). The high surface acidity of the incorporated nanomaterials allows high proton conductivity to be achieved in the composite membranes, while blockage of the pores by the particles reduces methanol transport. The sol–gel syntheses of the inorganic phases (SiO2, TiO2, and ZrO2) inside the pores of the Naion membranes had been reported as an effective modiication route to achieve high selectivities (Daiko et al. 2006; Xu et al. 2005b; Jalani et al. 2005; Ren et al. 2006). One of the modiiers for Naion membranes is zeolites. This is due to their inherent narrow pore size distributions, high surface acidity, and high water intake (Chen et al. 2006). In one study, the results of the methanol transport properties of several Naion–composite membranes incorporating micro-sized and nano-sized particles of zeolites (Fe-silicate-1), amorphous silica, and in situ crystallized Fe-silicalite-1 have been reported in comparison with an unmodiied commercial Naion-115 membrane (Gribov et al. 2007). In preparation of the composites, the supercritical carbon dioxide treatment of some of the membranes was used prior to incorporating the inorganic phase. Two methods of deposition of the zeolites were used: deposition from a colloid or a suspension solution and direct in situ synthesis inside the pores of the Naion membrane. Very low methanol permeability was achieved for composite membranes that were prepared using the colloidal intercalation route (from colloidal Fe-silicate-1 as well as the silica solution) and from the in situ synthesis of the Fe-silicate-1 inside the pores of the Naion membrane.

Modifications of Polymeric Membranes

83

Supercritical CO2 activation of a Naion membrane prior to zeolite deposition was used to modify its structure. The resultant Naion–zeolite composite membranes showed a dramatic decrease in methanol permeability (if the colloidal rather than the suspended Fe-silicalite-1 particles were used for the deposition) and a 19-fold higher selectivity compared with either the composite membranes prepared without previous supercritical treatment or the pure commercial Naion-115 membrane. The method of the in situ synthesis of a zeolite inside the membrane pores was found to be very effective for preparing the composites, giving a sixfold higher selectivity for the composite membrane compared with the pure Naion membranes (Gribov et al. 2007). Gao et al. (2005) also undertook the fabrication of ion-conducting membranes by the self-assembly of surface-charged nanoparticles. In this study, the researchers found that membranes made from closely packed nanoparticles had a signiicantly higher proton conductivity compared with the solution-cast ilms of a similar composition and ion-exchange capacity (IEC). However, there was a limitation on the maximum IEC, as high-IEC membranes exhibited excessive swelling in water, which hindered the testing for proton conductivity. Membranes with proton-conducting particles assembled in a suitable matrix can be designed to avoid such problems. The particles can be aligned to obtain the percolation needed for proton conduction, and the swelling in water or methanol can be controlled by choosing a water-resistant matrix. In a similar work, the synthesis of a composite with iron oxide particles and sulfonated cross-linked polystyrene (SXLPS) for application in the PEMs for fuel cells was described (Brijmohan and Shaw 2007). The technique used for the polymerization was similar to the miniemulsion polymerization (Ramirez and Landfester 2003). However, some modiication to the procedure was required to make the crosslinked and functional polymer–iron oxide composites. Also reported was the membrane fabrication process, which includes the alignment of synthesized particles in a high-performance sulfonated poly(etherketoneketone) (SPEKK) matrix (Gasa et al. 2006), and the properties of such PEMs for fuel cell applications. The inal properties of the membrane depend on various factors, such as the IEC, the matrix, and the size of the particles. However, the main emphasis of the research was to demonstrate a useful membrane-fabrication technique that can be utilized to enhance the conductivity of the PEMs. Composite ion-conducting nanoparticles have been synthesized by emulsion polymerization (Brijmohan and Shaw 2007). The polymeric component of the particles was composed of SXLPS and the inorganic component was γ-Fe2O3. A particle synthesis mechanism was proposed to explain the unusual morphologies of the composite particles. The average diameters of the synthesized particles with different feed compositions ranged from 230 to 340 nm and the distribution breadth increased with the sulfonated content (Figure 4.1). Using thermogravimetric methods, it was determined that the particles contained 75–80 wt% of polymer. This was also conirmed by measuring the magnetic susceptibility. The synthesized particles could be easily aligned in SPEKK for PEM application. The proton conductivity of the membranes with the particles was less than that of the unmodiied SPEKK membranes due to the low IEC of the particles. However, the conductivity appeared

84

Membrane Modification: Technology and Applications XLPS

γ-Fe2O3

500 nm 200 nm

NaSXLPS-0.08

NaSXLPS-0.15

500 nm

500 nm

NaSXLPS-0.23

500 nm

FIGURE 4.1 TEM images of the starting material γ-Fe2O3 and the synthesized cross-linked polystyrene γ-Fe2O3 particles. (Adapted from Brijmohan, S.B. and Shaw, M.T., J. Membrane Sci., 303, 64–71, 2007.)

to be improved with alignment. ZrO2 was also used in sulfonated polyetherketone (SPEK) membranes to reduce the water and methanol permeability. These properties make the ZrO2 and polyetherketone combination an attractive choice for use in fuel cell applications. This is particularly true for polyelectrolyte membrane fuel cells (PEFC). PEFCs are interesting for mobile applications and are currently an important research topic in all leading automobile industries. However, storage and delivery still provide complex problems. An alternative is the use of reformers to generate hydrogen from liquid fuels, such as methanol or gasoline. The reformer

85

Modifications of Polymeric Membranes 12,000

Flux (gh m2)

9,000

1,200

800

6,000 Methanol

Flux (gh m2)

SPEK/SiO2

Water

400

3,000

0 0

5

10

0 15

% SiO2

FIGURE 4.2 Methanol and water lux in pervaporation experiments at 55°C by modiication of a SPEK membrane with hydrolyzed TEOS. (Adapted from Nunes, S.P., et al., J. Membrane Sci., 203, 215–225, 2002.)

could be eliminated, gaining in technical simplicity and saving space, if methanol could directly feed the fuel cell. For a breakthrough in DMFC technology, suitable membranes with high proton conductivity and low water and methanol permeability are required (Nunes et al. 2002). Naion membranes, as discussed, have been intensively used for fuel cells because they show high proton conductivity and chemical stability, but their methanol permeability is too high. However, the critical aspect of Naion is still its high cost. Several nonluorinated membranes, with potentially lower costs, have been tested for fuel cells. Sulfonated PSF, sulfonated poly(ether ether ketone) (SPEEK), sulfonated polyphosphazene, and sulfonated polyamides (PA) with good performance for hydrogen fuel cells are described in several reports (Savadogo 1998; Zaidi et al. 2000; Guo et al. 1999; Vallejo et al. 1999). However, the methanol permeability in many cases is still relatively high. In the research reported by Nunes et al. (2002), the polymers chosen for the membrane preparation were SPEK and SPEEK. Inorganic networks were generated in the organic polymer matrix by the hydrolysis of tetraethoxy silane (TEOS) and 1-(3-triethoxysilyl propyl)-4,5-dihydroimidazole (I-silane). The inorganic modiication with SiO2 led to a considerable decrease in permeability, as shown in Figure 4.2. These workers reported that there was a reduction in the water and methanol permeability after the modiication with TiO2 and ZrO2. Figure 4.3a illustrates the considerable reduction in the water and methanol permeability after the modiication with TiO2. With this oxide, only a little improvement was achieved by the pretreatment with 1,1′-carbonyl-diimidazole (CDI) and aminopropylsilane (AS). The distribution of ZrO2 in the SPEK membranes strongly reduced the permeability of the methanol and water as well (Figure 4.3b). No additional improvement was observed using the pretreatment with CDI and AS (Nunes et al. 2002). In addition, the distribution of the inorganic phase in the membrane was very homogeneous. A scanning electron

86

Membrane Modification: Technology and Applications 12,000 SPEK/TiO2

Flux (gh m2)

800

6,000 Water

3,000

Flux (gh m2)

1,200

9,000

400 Methanol

0

0

5

(a)

10

15

0 20

% TiO2 400

2,500 SPEK/ZrO2

1,500 200

Water 1,000

Flux (gh m2)

Flux (gh m2)

2,000

500 Methanol 0 (b)

0

10

20 % ZrO2

30

0 40

FIGURE 4.3 Methanol and water lux in pervaporation experiments at 55°C after modiication of a SPEK membrane with (a) TiO2 and (b) ZrO2. (Adapted from Nunes, S.P., et al., J. Membrane Sci., 203, 215–225, 2002.)

microscope (SEM) image of a SPEK membrane modiied with 22 wt% TiO2 is shown in Figure 4.4. The good distribution of the nanoparticles reported was attributed to the method itself, where the inorganic networks were generated in the polymeric matrix by the hydrolysis of TEOS and I-silane. Modiication of the Naion membrane by incorporating hygroscopic inorganic nanoparticles such as SiO2 and ZrO2 has been demonstrated to improve the water retention and enhance the proton conductivity (Antonucci et al. 1999; Miyake et al. 2001; Jalani et al. 2005; Adjemian et al. 2006; Sacca et al. 2006; Di Noto et al. 2006; Shao et al. 2006; Zhai et al. 2006; Tang et al. 2007; Park et al. 2008). More recently, hybrid membranes of Naion and mesoporous silica containing sulfuric acid groups, formed using a sol–gel process, were synthesized (Pereira et al.

87

Modifications of Polymeric Membranes

1 µm

FIGURE 4.4 The SEM of a SPEK membrane modiied with 22 wt% TiO2. (Adapted from Nunes, S.P., et al., J. Membrane Sci. 203, 215–225, 2002.)

2008). Compared with the standard Naion 112 membrane, the hybrid membranes show improved proton conductivity at 95°C and 120°C and over the whole range of relative humidity. Other studies that were previously carried out to improve the properties of the fuel cell membrane include the incorporation of the silica nanoparticles in a sulfonated poly(arylene ether sulfone) membrane (Lee et al. 2007a). Interestingly, the SiO2 particles with a high surface area and a small particle size showed the best results: high proton conductivity, long membrane lifetime under oxidative conditions, good dimensional stability, outstanding single cell performance, and reduced methanol crossover. Moreover, the SiO2 content plays an important role in the membrane microstructures and membrane properties, such as proton conductivity and methanol barrier behavior. However, an excessive SiO2 content caused a large aggregation of the SiO2 particles, leading to the deterioration of the mechanical properties in nanocomposite membranes.

4.3.2

MICROFILTRATION AND ULTRAFILTRATION

MF is a well-known technique for the removal of microparticles and microorganisms from luid streams and has, therefore, many industrial applications in food and bioprocess engineering. Conventional MF processes are mostly based on the principle of size exclusion. MF membranes are typically used for removing particles in the range of 0.1–10 μm from a suspension. Larger particles cannot enter the pores in the membrane and accumulate on the surface of the membrane. Blocking of the iltration membrane is prevented by the application of a cross-low. However, the accumulation of the retained material on the membrane surface and membrane fouling remain the limiting factors in many applications. Thus, in certain circumstances, membrane engineers have opted to use UF membranes with much smaller pore sizes (5 g/l), the water permeability increases proportionally to the UV-irradiation time. Kilduff et al. (2000, 2005) observed a decline in the water permeability by 58% and 40% for a UV-irradiation time of 10 and 60 sec, respectively; while for a longer irradiation time of 180 sec, the water permeability increased up to three times that of the unmodiied NFPES10 membrane. In our case, although the long irradiation time of 300 sec was applied, none of the grafted membranes showed higher water permeability than the unmodiied membrane. On the other hand, a decrease in the water permeability with an increase in the NVP monomer concentration can be observed. For each irradiation time, a higher monomer concentration produces grafted membranes with a lower permeability. For a lower NVP concentration, once the membrane surface is exposed to UV light, this can easily penetrate the monomer solution, conserving enough energy to attack the polymer backbone of the PES membrane surface and form radicals ready to be bonded with the NVP monomer. When the monomer concentration is low, a short irradiation time is not enough for NVP to be grafted onto the membrane surface sites. A longer irradiation time is necessary for grafting. In addition, the 20 18 16 Pm (l/m2 h bar)

14 12 10 5 g/l NVP

8

15 g/l NVP

6 4

30 g/l NVP

2

50 g/l NVP

0 0

1

2

3

4

5

UV-irradiation time (min)

FIGURE 5.17 The effects of UV-irradiation time on pure water permeability (Pm) at different NVP concentrations. (From Desalination, 287, Abu Seman, M.N., Khayet, M., and Hilal, N., Comparison of two different UV-grafted nanoiltration membranes prepared for reduction of humic acid fouling using acrylic acid and N-vinylpyrrolidone, 19–29, Copyright (2012), with permission from Elsevier.)

140

Membrane Modification: Technology and Applications

observed increase in the water permeability with the UV-irradiation time at higher NVP concentrations (>5 g/l) or after 3 min of irradiation time for modiication with 5 g/l may be explained by the change in the hydrophilicity of the membrane, the possibility of the grafted polymer being detached from the membrane surface and trunk polymer scission, or (pore enlargement) when longer irradiation time is applied. It is worth stating that similar results were observed by Pieracci et al. (1999) and Taniguchi et al. (2003b). Pieracci et al. (1999) studied the effect of the variation of irradiation time (0.5–10 min) on the degree of grafting during the modiication of the PES membrane using 0.5 wt% NVP. They found that the degree of grafting increased linearly with the irradiation time up to 5 min, but started to decrease after 7 min of irradiation time and dropped signiicantly (approximately three times lower than the maximum) at 10 min of irradiation time. They reported that at the longer irradiation time, some grafted NVP may be lost from the membrane surface, leading to a lower degree of grafting. Taniguchi et al. (2003b) explained this phenomenon by a relationship between the irradiation time (energy dose) and the grafting mechanism. Once the saturated membrane with the NVP monomer is exposed to UV irradiation, the number of formed radical sites increases proportionally with the irradiation energy (or irradiation time) because there is no obstacle for UV light to reach the membrane. This simply suggests that the relationship between the degree of grafting and the irradiation energy (E) should be linear at low irradiation energy and that the degree of grafting should also be proportional to the monomer concentration. However, when a long irradiation time is applied, the high energy may disrupt the NVP polymer that is already formed, reducing the degree of grafting. They suggested that below a critical energy dose (E < 4 kJ/m2 in their case), both the chain scission and the effective grafting occurred, while the undesired polymer loss took place when the energy applied was above the critical energy. In the case of a constant irradiation time, with a similar amount of energy, the grafted membrane had a low permeability at a higher monomer concentration because of the high degree of grafting and the formation of a dense and thick PVP layer. Moreover, pore constriction or pore reduction during the grafting process could also contribute to the reduction in the membrane permeability. It should be mentioned that although both monomers (AA and NVP) were used applying the same modiication conditions, the permeability of the grafted membranes is different (Table 5.2). Most of the AA-grafted membranes showed higher water permeability than the NVP-grafted membranes except for low monomer concentrations (5 and 15 g/l) with a UV-irradiation time of 5 min. The lower water permeability of the NVP membranes could be explained by the morphology and the hydrophilicity of the polymer. According to Taniguchi and Belfort (2004), AA is more reactive than NVP due to the smaller size of AA, which increases its diffusivity and its intrinsic reactivity. Taniguchi and Belfort (2004) used a UF membrane with a 50 kDa molecular weight cutoff. In our case, the NF membrane with a smaller pore size was used. Therefore, the monomer diffusivity was not a dominant factor for grafting. It is believed that the structure of the PVP polymer formed on the top of the membrane surface exerts a signiicant effect on the water permeability. Moreover, as the charged anionic AA monomer may generate a repulsive force with the charged membrane surface due to the functional groups such as the sulfonyl group, less AA

Development of Antifouling Properties and Performance of NF Membranes

141

will be grafted onto the membrane surface. The neutral NVP monomer may be easily grafted onto the membrane surface without any constraint to forming a thicker layer, which increases the hydrodynamic resistance of the membrane. Kaeselev et al. (2002) investigated the performance of the UF membrane modiied with NVP and charged carboxylic acid monomers AAG and 2-acrylamido-2-methyl-1-propanesulfonic acid (AAP). They found that the permeate lux of both modiied membranes modiied with the carboxylic acid monomers remained constant at 40 l/m2 h, whereas the modiied membranes with NVP exhibited a lower permeate lux (10 l/m2 h) especially when the irradiation energy increased from 30 to 95 mJ/cm2. Kaeselev et al. (2002) suggested that the surface structure of the grafted polymer was responsible for the change in the membrane performance. NVP has a stronger tendency to form a cross-linking chain net compared with AAG and AMPS, which corresponds to the linear PVP species that are more lexible than the dendric polymers formed by both carboxylic acid monomers. The membranes that exhibit higher permeate luxes than the unmodiied membrane might have experienced pore enlargement due to UV irradiation. Therefore, the membranes having a lower or a similar water permeability to the unmodiied membrane were considered for further NF tests using a 15 mg/l HA feed aqueous solution at a pH of 7 and 3, a hydrostatic pressure of 6 × 105 Pa, and a feed low rate of 0.4 l/min. These membranes are 5AA-5, 15AA-5, 30AA-3, and 50AA-1 for the AA-grafted membranes, and 5NVP-3, 15NVP-1, 30NVP-1, and 50NVP-1 for the NVP-grafted membranes. Figures 5.18 and 5.19 show the normalized permeate lux, Jt/J0, during 4 h of NF experiments. 1.2

Normalized flux (Jt/J0)

1.1 1.0 0.9 0.8

NFPES10 5AA-5 15AA-5 30AA-3 50AA-1

0.7 0.6 0.5 0

50

100 150 Time (min)

200

250

FIGURE 5.18 The normalized permeate lux of the unmodiied (NFPES10) membrane and the UV-grafted membranes by AA (15 mg/l HA, pH 7, 6 × 105 Pa transmembrane pressure, 0.4 l/min feed low rate). (From Journal of Membrane Science, 355, Abu Seman, M.N., Khayet, M., Bin Ali, Z.I., and Hilal, N., Reduction of nanoiltration membrane fouling by UV-initiated graft polymerization technique, 133–141, Copyright (2010), with permission from Elsevier.)

142

Membrane Modification: Technology and Applications 1

Normalized flux (Jt/J0)

0.9

0.8

0.7 NFPES10 5AA-5 15AA-5 30AA-3 50AA-1

0.6

0.5 0

50

100 150 Time (min)

200

250

FIGURE 5.19 The normalized permeate lux of the unmodiied (NFPES10) membrane and the UV-grafted membranes by AA (15 mg/l HA, pH 3, 6 × 105 Pa transmembrane pressure, 0.4 l/min feed low rate). (From Journal of Membrane Science, 355, Abu Seman, M.N., Khayet, M., Bin Ali, Z.I., and Hilal, N., Reduction of nanoiltration membrane fouling by UV-initiated graft polymerization technique, 133–141, Copyright (2010), with permission from Elsevier.)

As can be seen, after grafting, the antifouling abilities of the membrane were improved. At both pH values, all of the polyacrylic acid (PAA) membranes exhibit a lower lux decline than the unmodiied membrane. Under a neutral pH environment when the pH value is higher than the pKa of PAA (pH 4.75), PAA is negatively charged (Turan and Caykara 2007). In this case, the negatively charged HA molecules might have high electrostatic repulsion, preventing the HA molecules from being attached to the UV-grafted membrane surface. It must be mentioned that during the NF test at a pH of 7, the UV-initiated grafting membranes showed slight permeate lux enhancements. This may be attributed to the increase in the hydrophilicity of the membrane surface caused by the hydrophilic HA molecules concentrated at the membrane surface. When the pH of the HA solution was changed to be acid (pH = 3), as illustrated in Figure 5.19, both the unmodiied membrane and the grafted membranes showed a rapid permeate lux decline. This is attributed to the charge effect that becomes insigniicant at this pH value. The HA lost its charge and the membrane surface became less charged. This increased the interaction between the HA molecules and the membrane surface, leading to a decrease in the permeate lux. However, it can be seen that all the modiied membranes still showed a lower permeate lux decline than the unmodiied membrane. At a pH of 3, some PVP membranes exhibit a high permeate lux decline, similar to the unmodiied membrane (5NVP-3 and 50NVP-1).

Development of Antifouling Properties and Performance of NF Membranes

143

For both pH conditions (pH 7 and 3), the lowest permeate lux decline was observed for the 15AA-5 and 30AA-3 membranes (Figures 5.18 and 5.19) and for the 15NVP-1 membrane (Figures 5.20 and 5.21). Based on the results presented in Figures 5.20 and 5.21, one can speculate that only a moderate monomer concentration is required to produce a membrane with a high NF performance and fouling resistance. As mentioned earlier, membrane surface chemistry may affect the permeate lux decline behavior during iltration. Irreversible fouling is strongly inluenced by the membrane surface chemistry (Taniguchi et al. 2003a). As stated earlier, FRw can be used as an indicator to estimate the degree of irreversible fouling of both the unmodiied membrane and the grafted membranes. The results are shown in Figure 5.22. Under a neutral pH, a moderate concentration of 30 g/l of AA and 15 g/l of NVP can produce a membrane with low fouling. Interestingly, almost all of the NVP-grafted membranes, except the 50NVP-1 membrane, possess FRw values even smaller than those of the AA-grafted membranes. For NVP-grafted membranes, it seems that the FRw increases with the increase in the NVP monomer concentration. The opposite behavior was detected for the AA-grafted membranes. In a previous study, it was shown that AA had the tendency to reduce NOM fouling more than NVP when a UF membrane was used as support due to the higher density charge of the grafted PAA layer (Taniguchi et al. 2003a). In our study, the degree of AA grafting could be less than NVP due to the repulsion force between the 1.1

Normalized flux (Jt/J0)

1

0.9

0.8 NFPES10 5NVP-3

0.7

15NVP-1 30NVP-1

0.6

50NVP-1 0.5 0

50

150 100 Time (min)

200

250

FIGURE 5.20 The normalized permeate lux of the unmodiied (NFPES10) membrane and the UV-grafted membranes by NVP (15 mg/l HA, pH 7, 6 × 105 Pa transmembrane pressure, 0.4 l/min feed low rate). (From Desalination, 287, Abu Seman, M.N., Khayet, M., and Hilal, N., Comparison of two different UV-grafted nanoiltration membranes prepared for reduction of humic acid fouling using acrylic acid and N-vinylpyrrolidone, 19–29, Copyright (2012), with permission from Elsevier.)

144

Membrane Modification: Technology and Applications 1

NFPES10 5NVP-3 15NVP-1

Normalized flux (Jt/J0)

0.9

30NVP-1 50NVP-1

0.8

0.7

0.6

0.5 0

50

100 150 Time (min)

200

250

FIGURE 5.21 The normalized permeate lux of the unmodiied (NFPES10) membrane and the UV-grafted membranes by NVP (15 mg/l HA, pH 3, 6 × 105 Pa transmembrane pressure, 0.4 l/min feed low rate). (From Desalination, 287, Abu Seman, M.N., Khayet, M., and Hilal, N., Comparison of two different UV-grafted nanoiltration membranes prepared for reduction of humic acid fouling using acrylic acid and N-vinylpyrrolidone, 19–29, Copyright (2012), with permission from Elsevier.)

80 pH 7

70

pH 3

FRW (%)

60 50 40 30 20 10 0 10

ES

FP

N

5N

-3 VP 1

5N

-1 VP 3

0N

-1 VP

-3 -1 -1 -5 -5 VP 5AA AA AA AA 0 0 5 1 3 5 5 0N

FIGURE 5.22 The irreversible fouling factor (FRw) of the unmodiied membrane (NFPES10) and the modiied membranes used for the treatment of 15 mg/l HA at pH 7 and 3 under a transmembrane pressure of 6 × 105 Pa and a feed low rate of 0.4 l/min. (From Desalination, 287, Abu Seman, M.N., Khayet, M., and Hilal, N., Comparison of two different UV-grafted nanoiltration membranes prepared for reduction of humic acid fouling using acrylic acid and N-vinylpyrrolidone, 19–29, Copyright (2012), with permission from Elsevier.)

Development of Antifouling Properties and Performance of NF Membranes

145

anionic AA monomer and the negatively charged NF membrane, whereas because of the neutral characteristic of the NVP monomer, it is easily grafted onto the NF membrane surface. At a low pH value, as can be seen in Figure 5.22, the membranes grafted with low monomer concentrations (5 g/l) show a high fouling tendency. Higher monomer concentrations (>15 g/l) are required for low fouling caused by the changes in the HA molecular structure and coniguration, as explained in detail previously.

5.4

ATOMIC FORCE MICROSCOPY ANALYSIS OF MODIFIED MEMBRANES

Atomic force microscopy (AFM) images of the unmodiied membrane NFPES10 and all of the surface-modiied membranes were taken in an air environment at room temperature and at different locations across each membrane sample. Figures 5.23 and 5.24 show examples of the three-dimensional AFM images obtained. Several studies have reported changes in the surface morphology (roughness) after modifying the membrane surfaces (Khayet 2004; Hilal et al. 2005). The surface topography of the membranes modiied by IP and UV grafting is quite different from that of the unmodiied membrane surface. The morphology of the membranes was compared based on the surface roughness parameters. For this purpose, the same AFM cantilever tip was used and all the AFM images were treated in the same way since the parameters of the surface morphology depend on the treatment of the captured surface data, such as the iltering, leveling, and tip convolution removal, as well as on the curvature size of the AFM tip. As may be expected, the formation of a new polymer layer on the membrane surface signiicantly changes the membrane surface morphology. Tables 5.3 through 5.5 show the roughness parameters of the unmodiied membrane NFPES10 and the modiied membranes. It is clear that the membrane surface roughness increases after IP with BPA. For a short reaction time (10 sec), the roughness increases with a BPA concentration of up to 1% w/v BPA and then decreases for higher BPA concentrations. For longer reaction times, 30 and 60 sec, the roughness is slightly decreased with an increasing BPA concentration and this was obvious for a 60 sec polymerization time. The reaction polymerization is almost completed at 60 sec and the higher monomer concentration generates a more compact and denser layer with a low surface roughness. However, there is no clear trend between the surface roughness and the reaction time except for the BPA-modiied membrane with a concentration of 2% w/v. A rougher surface was generated after a longer reaction time. The membrane roughness is one of the factors that may contribute to fouling, thus the effect of the surface roughness on the irreversible fouling of HA iltration was investigated. It was reported that the surface roughness might affect the degree of fouling when the size of the solute molecules and the relative scale of the roughness are similar (Kilduff et al. 2005). The HA molecule size may be in the range between 1.7 and 3.5 nm, depending on its molecular weight, 1 and 10 kDa, respectively (Schäfer et al. 1998). In this study, the molecular weight of the HA used is 4.1 kDa (Chin et al. 1994). Therefore, the molecule size can be between 1.7 and 3.5 nm. Recent studies by Lee et al. (2008b) show that Aldrich HA contains three

146

Membrane Modification: Technology and Applications Digital instruments Nanoscope Scan size 1.000 µm 1.969 Hz Scan rate 512 Number of samples Height Image data 10.00 nm Data scale

0.2

Digital instruments Nanoscope 1.000 µm Scan size Scan rate 1.969 Hz Number of samples 512 Image data Height Data scale 25.00 m

0.2 0.4

0.4 X 0.200 µm/div Z 10.000 nm/div

0.6 0.8

X 0.200 µm/div Z 10.000 nm/div

0.6 0.8

µm

µm Mazrul membrane 0 NP-s probe

Mazrul membrane 3 NP-s probe

(a)

(b) Digital instruments Nanoscope Scan size 1.000 µm 1.969 Hz Scan rate 512 Number of samples Height Image data 25.00 nm Data scale

Digital instruments Nanoscope Scan size 1.000 µm 1.969 Hz Scan rate 512 Number of samples Height Image data 15.00 nm Data scale

0.2

0.2

0.4

0.4 X 0.200 µm/div Z 25.000 nm/div

0.6 0.8 µm

0.6

X 0.200 µm/div Z 15.000 nm/div

0.8 µm

Mazrul membrane 4 NP-s probe

(c)

Mazrul membrane 7 NP-s probe

(d) Digital instruments Nanoscope Scan size 1.000 µm 1.969 Hz Scan rate 512 Number of samples Height Image data 25.00 nm Data scale

0.2 0.4 X 0.200 µm/div Z 25.000 nm/div

0.6 0.8

(e)

µm

Mazrul membrane 10 NP-s probe

FIGURE 5.23 Three-dimensional AFM images of the unmodiied membrane (NFPES10) and the modiied BPA membranes by IP: (a) unmodiied, (b) 1BPA-6, (c) 5BPA-1, (d) 10BPA-1, and (e) 20BPA-1. (From Desalination, 273, Abu Seman, M.N., Khayet, M., and Hilal, N., Development of antifouling properties and performance of nanoiltration membranes modiied by interfacial polymerization, 36–47, Copyright (2011), with permission from Elsevier.)

Development of Antifouling Properties and Performance of NF Membranes Digital instruments Nanoscope Scan size 1.000 µm 1.969 Hz Scan rate 512 Number of samples Height Image data 25.00 nm Data scale

0.2

147

Digital instruments Nanoscope Scan size 1.000 µm 1.969 Hz Scan rate 512 Number of samples Height Image data 25.00 nm Data scale

0.2 0.4

0.4

0.6

0.6 0.8

0.8

X 0.200 µm/div Z 25.000 nm/div

µm

µm

Final set Mazrul membranes, number 0

(a)

X 0.200 µm/div Z 25.000 nm/div

Final set Mazrul membranes, number 2

(b)

Digital instruments Nanoscope Scan size 1.000 µm 1.969 Hz Scan rate 512 Number of samples Height Image data 25.00 nm Data scale

Digital instruments Nanoscope Scan size 1.000 µm 1.001 Hz Scan rate 512 Number of samples Height Image data 20.00 nm Data scale

0.2

0.2

0.4

0.4

0.6

0.6 0.8

X 0.200 µm/div Z 25.000 nm/div

0.8

µm Final set Mazrul membranes, number 4

(c)

X 0.200 µm/div Z 20.000 nm/div µm Membrane number 7

(d) Digital instruments Nanoscope Scan size 1.000 µm 1.969 Hz Scan rate 512 Number of samples Height Image data 50.00 nm Data scale

0.2 0.4 0.6 0.8

X 0.200 µm/div Z 50.000 nm/div µm Membrane number 10

(e)

FIGURE 5.24 Three-dimensional AFM images of the unmodiied membrane (NFPES10) (a) and the UV-grafted membranes with NVP, (b) 5NVP-3, (c) 15NVP-1, (d) 30NVP-1, and (e) 50NVP-1. (From Desalination, 287, Abu Seman, M.N., Khayet, M., and Hilal, N., Comparison of two different UV-grafted nanoiltration membranes prepared for reduction of humic acid fouling using acrylic acid and N-vinylpyrrolidone, 19–29, Copyright (2012), with permission from Elsevier.)

main groups of molecular weights, G1 (3 up to >20 kDa), G2 (0.5–3 kDa), and G3 ( NFPES10 > 15NVP-1 ≈ 30NVP-1 ≈ 50NVP-1 at a pH of 3 of the HA solution (Figure 5.26). Based on the AFM analysis of both the unmodiied membrane and the UV-grafted membranes, the 5NVP-3 membrane exhibits the highest roughness. For the other NVP-grafted membranes, which have similar roughness parameters compared with the unmodiied membrane,

Roughness, RMS (nm)

4

50 45 40 35 30 25 20 15 10 5 0

RMS FRw

3.5 3 2.5 2 1.5 1 0.5 0 NFPES10

5NVP-3

15NVP-1 30NVP-1 Membrane

151

Irreversible fouling, FRw (%)

Development of Antifouling Properties and Performance of NF Membranes

50NVP-1

FIGURE 5.26 The membrane roughness (RMS) and the irreversible fouling factor (FRw) of the unmodiied membrane (NFPES10) and the UV-grafted membranes with NVP. (From Desalination, 287, Abu Seman, M.N., Khayet, M., and Hilal, N., Comparison of two different UV-grafted nanoiltration membranes prepared for reduction of humic acid fouling using acrylic acid and N-vinylpyrrolidone, 19–29, Copyright (2012), with permission from Elsevier.)

the irreversible fouling is lower than that of the unmodiied membrane. This indicates that other factors, such as the hydrophilicity of the membrane surface due to the hydrophilic PVP polymer formed, reduce the interaction with the more hydrophobic HA at a pH of 3, and this should be taken into consideration. The interaction of the solutes with the membrane surface is normally characterized by the membrane hydrophilicity/hydrophobicity and a ixed charge using a contact angle measurement and a streaming potential, respectively. In this study, AFM was used to determine the force of adhesion between a HA-coated silica probe (illustrated in Figure 5.27) and the membrane surface. A typical force measurement

Sig HV Mag WD Spot HFW SE 20.0 kV 25000x 7.8 mm 2.0 10.82 µm

5.0µm

FIGURE 5.27 An SEM picture of a HA-coated silica probe.

152

Membrane Modification: Technology and Applications 15 Approach

Force (nN)

10

Retract

5 0 –100

0

100

200

300

400

–5 –10 –15

Separation (nm)

FIGURE 5.28 An example of force versus distance between a silica probe and a membrane surface. (The attractive forces are negative and the magnitude of the maximum attractive force reached on the retract curve [.......] gives the adhesion value.)

between the colloid probe and the membrane surface is presented in Figure 5.28. The force measurements were carried out at ive different locations on the membrane surface to minimize the effects of localized surface geometry. The AFM force–distance curves are described in detail elsewhere (Hilal et al. 2003). In order to understand the behavior of the foulant (HA molecules in this case) and the membrane surface properties, the relationship between fouling and the adhesion force was investigated. Three AA-modiied membranes with different UV-irradiation times of 1, 3, and 5 min, respectively (5AA-1, 5AA-3, and 5AA-5) were selected to understand the correlation between the degree of modiication, the adhesion force, and the irreversible fouling factor. The estimated adhesion forces of the three AA-grafted membranes and the unmodiied membrane (NFPES10) are given in Figure 5.29. The measurement with a HA-coated silica probe showed a very good agreement with the experimental hypothesis. At a loading force of 15 nN, when the HA-coated 14 Adhesion force (nN)

12 10 8 6 4 2 0 NFPES10

5AA-1 5AA-3 Membrane

5AA-5

FIGURE 5.29 The adhesion force of the unmodiied membrane (NFPES10) and the UV-grafted membranes with AA at 15 nN loading force in pure water.

Development of Antifouling Properties and Performance of NF Membranes

153

silica colloid probe was retracted from the membrane surface, the adhesion force was much higher for the unmodiied membrane compared with all the grafted membranes. This trend was also observed previously and the authors found a higher adhesion force for a UF-unmodiied PES membrane compared with a modiied membrane (Hilal et al. 2003). The grafted membranes 5AA-3 and 5AA-5, corresponding to 3 and 5 min irradiation time, respectively, exhibited the lowest adhesion forces, indicating that these membranes have a higher resistance to fouling than the unmodiied membrane. This could be due to the more negatively charged surface on the grafted membrane being strong enough to repel/exclude the negative charge of HA. This is consistent with the previously performed experimental NF data.

5.5

CONCLUSION

In this chapter, the surface modiication of an NF membrane by means of two different methods, namely, IP and UV photografting, is described and discussed in detail by experimental examples. The results show that the modiied membranes exhibit superior antifouling characteristics to that of the unmodiied membrane, especially at a neutral pH. However, at acidic environment (pH 3), some modiied membranes fouled even more than the unmodiied membrane. The membrane morphology, especially the surface roughness, is expected to contribute to fouling. Moreover, the membrane chemistry, such as the hydrophilicity, also affects to a certain degree the irreversible fouling because the foulant chemistry also changes at different environmental conditions. The UV-photografting technique seems better than IP for membrane surface modiication because no substantial reduction in the membrane permeability was detected. Among the UV-photografting modiied membranes, the grafted membranes with NVP at moderate concentrations and for a short irradiation time exhibit a better NF performance with a higher permeability and a low irreversible fouling at both neutral and acidic pH conditions.

REFERENCES Ahmad, A.L. and Ooi, B.S. 2005. Properties–performance of thin ilm composites membrane: Study on trimesoyl chloride content and polymerization time. J. Memb. Sci. 255: 67–77. Ahmad, A.L., Ooi, B.S. and Choudhury, J.P. 2003. Preparation and characterization of copolyamide thin ilm composite membrane from piperazine and 3,5-diaminobenzoic acid. Desalination 158: 101–108. Al-Amoudi, A. and Lovitt, R.W. 2007. Fouling strategies and the cleaning system of NF membranes and factors affecting cleaning eficiency. J. Memb. Sci. 303: 4–28. Bhattacharya, A. and Misra, B.N. 2004. Grafting: A versatile means to modify polymers: Techniques, factors and applications. Prog. Polym. Sci. 29(8): 767–814. Boussu, K. 2007. Inluence of membrane characteristics on lux decline and retention in nanoiltration. Ph.D. Dissertation, Katholieke Universiteit Leuven. Childress, A.E. and Elimelech, M. 1996. Effect of solution chemistry on surface charge of polymeric reverse osmosis and nanoiltration membranes. J. Memb. Sci. 119: 253–268. Chin, Y.P., Aiken, G. and O’Loughlin, E. 1994. Molecular weight, polydispersity and spectroscopic properties of aquatic humic substances. Environ. Sci. Technol. 28: 1853–1858.

154

Membrane Modification: Technology and Applications

Chng, M.L., Xiao, Y., Chung, T.S., Toriida, M. and Tamai, S. 2007. The effects of chemical structure on gas transport properties of poly(aryl ether ketone) random copolymers. Polymer 48: 311–317. Chu, L.Y., Wang, S. and Chen, W.M. 2005. Surface modiication of ceramic-supported polyethersulfone membranes by interfacial polymerization for reduced membrane fouling. Macromol. Chem. Phys. 207: 1934–1940. Chu, L.Q., Tan, W.J., Mao, H.Q. and Knoll, W. 2006. Characterization of UV-induced graft polymerization of poly(acrylic acid) using optical waveguide spectroscopy. Macromolecules 39: 8742–8746. Ernst, M., Bismarck, A., Springer, J. and Jekel, M. 2000. Zeta-potential and rejection rates of a polyethersulfone nanoiltration membrane in single salt solutions. J. Memb. Sci. 165: 251–259. Ghosh, K. and Schnitzer, M. 1980. Macromolecular structures of humic substances. Soil Sci. 129(5): 266–276. Guimarães, D.H., Brioude, M.d.M., Fiúza, R.d.P., Prado, L.A.S., Boaventura, J.S. and José, N.M. 2007. Synthesis and characterization of polyesters derived from glycerol and phthalic acid. Mater. Res. 10: 257–260. He, D. 2008. Surface-selective and controllable photo-grafting for synthesis of tailored macroporous membrane adsorbers. Ph.D. Dissertation, Universitat Duisburg-Essen, Germany. Hilal, N. and Kochkodan, V. 2003. Surface modiied microiltration membranes with molecularly recognising properties. J. Memb. Sci. 213: 97–113. Hilal, N., Al-Khatib, L., Atkin, B.P., Kochkodan, V. and Potapchenko, N. 2003. Photochemical modiication of membrane surfaces for (bio)fouling reduction: A nano-scale study using AFM. Desalination 158: 65–72. Hilal, N., Al-Khatib, L., Al-Zhoubi, H. and Nigmatulin, R. 2005. Atomic force microscopy study of membranes modiied by surface grafting of cationic polyelectrolyte. Desalination 184: 45–55. Hong, S. and Elimelech, M. 1997. Chemical and physical aspects of natural organic matter (NOM) fouling of nanoiltration membranes. J. Memb. Sci. 132: 159–181. http://en.wikipedia.org/wiki/Methyl_group (accessed May 20, 2010). Huang, S.H., Li, C.L., Hu, C.C., Tsai, H.A., Lee, K.R. and Lai, J.L. 2006. Polyamide thin-ilm composite membranes prepared by interfacial polymerization for pervaporation separation. Desalination 200: 387–389. Hwang, T.S. and Park, J.W. 2003. UV-induced graft polymerization of polypropylene-g-glycidyl methacrylate membrane in the vapour phase. Macromol. Res. 11(6): 495–500. Jayarani, M.M. and Kulkarni, S.S. 2000. Thin-ilm composite poly(esteramide)-based membranes. Desalination 130: 17–30. Jho, J.Y. 1990. Cooperative molecular motion on bisphenol-A polycarbonate (molecular motion). Ph.D. Dissertation, The University of Michigan. Ji, J. 1996. Fabrication and photochemical surface modiication of photoreactive thin-ilm composite membranes and model development for thin ilm formation by interfacial polymerization. Ph.D. Dissertation, McMaster University. Ji, J. and Mehta, M. 2001. Mathematical model for the formation of thin-ilm composite hollow iber and tubular membranes by interfacial polymerization. J. Memb. Sci. 192: 41–54. Kaeselev, B., Kingshott, P. and Jonsson, G. 2002. Inluence of the surface structure on the iltration performance of UV-modiied PES membranes. Desalination 146: 265–271. Kang, G., Liu, M., Lin, B., Cao, Y. and Yuan, Q. 2007. A novel method of surface modiication on thin ilm composite reverse osmosis membrane by grafting poly(ethylene glycol). Polymer 48: 1165–1170. Khayet, M. 2004. Membrane surface modiication and characterization by X-ray photoelectron spectroscopy, atomic force microscopy and contact angle measurements. Appl. Surf. Sci. 238(1–4): 269–272.

Development of Antifouling Properties and Performance of NF Membranes

155

Kilduff, J.E., Mattaraj, S., Pieracci, J.P. and Belfort, G. 2000. Photochemical modiication of poly(ether sulfone) and sulfonated poly(sulfone) nanoiltration membranes for control of fouling by natural organic matter. Desalination 132: 133–142. Kilduff, J.E., Belfort, G., Nayak, A. and Zhou, M. 2004. Progress report: Graft polymerization as a route to control nanoiltration membrane surface properties to manage EPA candidate contaminants and reduce NOM fouling. http://cfpub.epa.gov/ncerabstracts/index.cfm/ fuseaction/display.abstractDetail/abstract/6236/report/2004 (accessed April 23, 2009). Kilduff, J.E., Mattaraj, S., Zhou, M. and Belfort, G. 2005. Kinetics of membrane lux decline: The role of natural colloids and mitigation via membrane surface modiication. J. Nanopart. Res. 7: 525–544. Kim, S. and Liu, J. 2001. Molecular modelling of bisphenol-A polycarbonate and tetramethyl bisphenol-A polycarbonate. Korea Polym. J. 9: 129–142. Korikov, A.P., Kosaraju, P.B. and Sirkar, K.K. 2006. Interfacially polymerized hydrophilic microporous thin ilm composite membranes on porous polypropylene hollow ibers and lat ilms. J. Memb. Sci. 279: 588–600. Kuroda, S., Mita, I., Obata, K. and Tanaka, S. 1990. Degradation of aromatic polymers: Part IV: Effect of temperature and light intensity on the photodegradation of polyethersulfone. Polym. Degrad. Stab. 27: 257–270. Kwak, S.Y., Yeom, M.O., Roh, I.J., Kim, D.Y. and Kim, J.J. 1997. Correlations of chemical structure, atomic force microscopy (AFM) morphology and reverse osmosis (RO) characteristics in aromatic polyester high-lux RO membranes. J. Memb. Sci. 132: 183–191. Lee, D.H., Kim, H.I. and Kim, S.S. 2004. Surface modiication of polymeric membranes by UV grafting. In: Advanced Material for Membrane Separations, Chapter 19, pp. 281–299. ACS Symposium Series 876. American Chemical Society: Washington, DC. Lee, H.S., Im, S.J., Kim, J.H., Kim, H.J., Kim, J.P. and Min, B.R. 2008a. Polyamide thin-ilm nanoiltration membranes containing TiO2 nanoparticles. Desalination 219: 48–56. Lee, H.-H., Weng, Y.-H. and Li, K.C. 2008b. Electro-ultrailtration study on Aldrich humic substances with different molecular weights. Sep. Purif. Technol. 63: 23–29. Liang, L., Rieke, P.C., Fryxell, G.E., Liu, J., Engehard, M.H. and Alford, K.L. 2000. Temperature-sensitive surfaces prepared by UV photografting reaction of photosensitizer and N-isopropylacrylamide. J. Phys. Chem. B 104: 11667–11673. Li-hua, L., Shan, D., Ye, T., Jin-huan, T. and Chang-ren, Z. 2008. Surface modiication of poly-D, L-lactic acid ilm with plasma graft polymerization. Tissue Eng. Res. Clin. Recovery 12: 1–10. Mänttäri, M., Puro, L., Nuortila-Jokinen, J. and Nystrom, M. 2000. Fouling effects of polysaccharides and humic acid in nanoiltration. J. Memb. Sci. 165(1): 1–17. Mohammad, A.W., Hilal, N. and Abu Seman, M.N. 2003. A study on producing composite nanoiltration membranes with optimized properties. Desalination 158: 73–78. Mulder, M. 2000. Basic Principles of Membrane Technology, 2nd edn. Kluwer Academic: Dordrecht. Petersen, R.J. 1993. Composite reverse osmosis and nanoiltration membranes. J. Memb. Sci. 83: 81–150. Pieracci, J., Wood, D.W., Crivello, J.V. and Belfort, G. 1999. Increasing membrane permeability of UV-modiied poly(ether sulfone) ultrailtration membranes. J. Memb. Sci. 156: 223–240. Pieracci, J., Wood, D.W., Crivello, J.V. and Belfort, G. 2000. UV-assisted graft polymerization of N-vinyl-2-pyrrolidinone onto poly(ether sulfone) ultrailtration membranes: Comparison of dip versus immersion modiication techniques. Chem. Mater. 12: 2123–2133. Pieracci, J., Wood, D.W., Crivello, J.V. and Belfort, G. 2002. Increasing membrane permeability of UV-modiied poly(ethersulfone) ultrailtration membranes. J. Memb. Sci. 202: 1–16.

156

Membrane Modification: Technology and Applications

Pinnau, I. and Freeman, B.D. 2000. Formation and modiication of polymeric membranes: Overview. In: Membrane Formation and Modiication, pp. 1–22. ACS Symposium Series 744. American Chemical Society: Washington, DC. Puro, L., Manttari, M., Pihlajamaki, A. and Nystrom, M. 2006. Characterization of modiied nanoiltration membrane by octanoic acid permeation and FTIR analysis. Trans. IChemE A Chem. Eng. Res. Des. 84(A2): 87–96. Qiu, C., Xu, F., Nguyen, Q.T. and Ping, Z. 2005. Nanoiltration membrane prepared from cardo polyetherketone ultrailtration membrane by UV-induced grafting method. J. Memb. Sci. 255: 107–115. Qiu, C., Nguyen, Q.T. and Ping, Z. 2007. Surface modiication of cardo polyetherketone ultrailtration membrane by photo-grafted copolymers to obtain nanoiltration membranes. J. Memb. Sci. 295: 88–94. Rao, A.P., Desai, N.V. and Rangarajan, R. 1997. Interfacially synthesized thin ilm composite RO membranes for seawater desalination. J. Memb. Sci. 124: 263–272. Razdan, U. and Kulkarni, S.S. 2004. Nanoiltration thin-ilm-composite polyesteramide membranes based on bulky diols. Desalination 161: 25–32. Schäfer, A.I., Fane, A.G. and Waite, T.D. 1998. Nanoiltration of natural organic matter: Removal, fouling and inluence of multivalent ions. Desalination 118: 109–122. Schäfer, A.I., Fane, A.G. and Waite, T.D. 2000. Fouling effects on rejection in the membrane iltration of natural waters. Desalination 131: 215–224. Solak, E.K. and Sanli, O. 2009. Permeation and separation characteristics of dimethylformamide/water mixtures by vapour permeation and vapour permeation with temperature difference methods through a sodium alginate-g-n-vinyl-2-pyrrolidone membrane. Desalin. Water Treat. 2: 148–155. Song, Y., Sun, P., Henry, L.L. and Sun, B. 2005. Mechanisms of structure and performance controlled thin ilm composite membrane formation via interfacial polymerization process. J. Memb. Sci. 251: 67–79. Susanto, H. and Ulbricht, M. 2008. High-performance thin-layer hydrogel composite membranes for ultrailtration of natural organic matter. Water Res. 42: 2827–2835. Tang, B., Huo, Z. and Wu, P. 2008. Study on a novel polyester composite nanoiltration membrane by interfacial polymerization of triethanolamine (TEOA) and trimesoyl chloride (TMC) I. Preparation, characterization and nanoiltration properties test of membrane. J. Memb. Sci. 320: 198–205. Taniguchi, M. and Belfort, G. 2004. Low protein fouling synthetic membranes by UV-assisted surface grafting modiication: Varying monomer type. J. Memb. Sci. 231: 147–157. Taniguchi, M., Kilduff, J.E. and Belfort, G. 2003a. Low fouling synthetic membranes by UV-assisted graft polymerization: Monomer selection to mitigate fouling by natural organic matter. J. Memb. Sci. 222(1–2): 59–70. Taniguchi, M., Pieracci, J., Samsonoff, W.A. and Belfort, G. 2003b. UV-assisted graft polymerization of synthetic membranes: Mechanistic studies. Chem. Mater. 15: 3805–3812. Tarboush, B.J.A., Rana, D., Matsuura, T., Arafat, H.A. and Narbaitz, R.M. 2008. Preparation of thin-ilm-composite polyamide membranes for desalination using novel hydrophilic surface modifying macromolecules. J. Memb. Sci. 325: 166–175. Turan, E. and Caykara, T. 2007. Swelling and network parameters of pH-sensitive poly(acrylamide-co-acrylic acid) hydrogels. J. Appl. Polym. Sci. 106: 2000–2007. Uyama, Y., Kato, K. and Ikada, Y. 1998. Surface modiication of polymers by grafting. Adv. Polym. Sci. 137: 1–3. Veríssimo, S., Peinemann, K.V. and Bordado, J. 2006. Inluence of the diamine structure on the nanoiltration performance, surface morphology and surface charge of the composite polyamide membranes. J. Memb. Sci. 279: 266–275. Wang, K.Y. and Chung, T.S. 2005. The characterization of lat composite nanoiltration membranes and their applications in the separation of cephalexin. J. Memb. Sci. 247: 37–50.

Development of Antifouling Properties and Performance of NF Membranes

157

Xi, W., Rong, W., Zhansheng, L. and Fane, A.G. 2006. Development of a novel electrophoresisUV grafting technique to modify PES UF membranes for NOM removal. J. Memb. Sci. 273: 47–57. Yamagishi, H., Crivello, J.V. and Belfort, G. 1995a. Development of a novel photochemical technique for modifying poly(arylsulfone) ultrailtration membranes. J. Memb. Sci. 105: 237–247. Yamagishi, H., Crivello, J.V. and Belfort, G. 1995b. Evaluation of photochemically modiied poly(arylsulfone) ultrailtration membranes. J. Memb. Sci. 105: 249–259. Yu, H.Y., He, J.M., Liu, L.Q., He, X.C., Gu, J.S. and Wei, X.W. 2007. Photoinduced graft polymerization to improve antifouling characteristics of an SMBR. J. Memb. Sci. 302: 235–242. Yusof, A.H.M. and Ulbricht, M. 2008. Polypropylene-based membrane adsorbers via photoinitiated graft copolymerization: Optimizing separation performance by preparation conditions. J. Memb. Sci. 311: 294–305.

6

Integrating Hydrophobic Surface-Modifying Macromolecules into Hydrophilic Polymers to Produce Membranes for Membrane Distillation Mohammed Qtaishat, Mohamed Khayet, and Takeshi Matsuura

CONTENTS 6.1 6.2

Introduction .................................................................................................. 159 Characteristics Required for MD Membranes ............................................. 160 6.2.1 High Liquid Entry Pressure .............................................................. 161 6.2.2 High Permeability ............................................................................. 161 6.2.3 Low Thermal Conductivity .............................................................. 162 6.3 Surface-Modifying Macromolecules ............................................................ 162 6.4 SMM-Modiied Membranes for DCMD ...................................................... 165 6.5 Case Study: Fabrication, Characterization, and Application of SMMModiied Membranes .................................................................................... 166 6.6 Conclusion .................................................................................................... 176 References .............................................................................................................. 177

6.1

INTRODUCTION

Membrane distillation (MD) is an emerging, thermally driven membrane separation process that has been widely investigated for many applications, including seawater desalination, food processing, and removal of volatile organic compounds from water (Lawson and Lloyd 1997; Alklaibi and Lior 2000; Burgoyne and Vahdati 2000; Curcio and Drioli 2005; El-Bourawi et al. 2006; Khayet 2011). The porous membrane in MD acts as a physical support that separates a hot feed solution from a cooling chamber containing either a liquid or a gas, depending on the MD 159

160

Membrane Modification: Technology and Applications

coniguration. There are four possible MD conigurations: direct contact membrane distillation (DCMD), air gap membrane distillation (AGMD), sweeping gas membrane distillation (SGMD), and vacuum membrane distillation (VMD) (Khayet 2011). In DCMD, an aqueous solution colder than the feed solution to be treated is maintained in direct contact with the permeate side. Both the feed and the permeate aqueous solutions are circulated tangentially to the membrane surfaces by means of circulating pumps or are stirred inside the membrane cell by means of a magnetic stirrer. In this case, the transmembrane temperature difference induces a vapor pressure difference. Consequently, volatile molecules evaporate at the hot liquid–vapor interface, cross the membrane pores in the vapor phase, and condense in the cold liquid–vapor interface inside the membrane module (Khayet 2011). As the process is nonisothermal, vapor molecules will migrate through the membrane pores from the high to the low vapor pressure side, from the warmer to the cooler side, where they will be condensed. The main requirement of the MD membrane is that the pores must not be wetted and only vapor is present. This limits the choice of MD membrane to those made of hydrophobic materials, such as polytetraluoroethylene (PTFE), polypropylene (PP), and polyvinylidene luoride (PVDF). Although these membranes were manufactured for microiltration (MF) and ultrailtration (UF) purposes, they have been used in MD research for many decades due to their hydrophobic nature (Lawson and Lloyd 1997; Alklaibi and Lior 2000; Burgoyne and Vahdati 2000; Curcio and Drioli 2005; El-Bourawi et al. 2006; Khayet 2011). MD technology exhibits several advantages compared with other separation processes. These advantages are high rejections of nonvolatile solutes with values near 100%, lower operating temperatures than conventional distillation, lower operating pressures than conventional pressure-driven membrane separation processes, and reduced vapor spaces compared with the conventional distillation processes, such as multi-effect distillation and multistage lash. Despite all of these advantages, the MD process has not yet been commercialized for large-scale plants. Some of the reasons are the relatively lower MD permeate lux and membrane wetting, which diminish the durability of the MD membranes. In many cases, these problems arise from the inadequate design of the MD membranes. The objectives of this chapter are to describe the fabrication of novel composite hydrophobic/hydrophilic membranes for DCMD using different surface-modifying macromolecules (SMMs) and a hydrophilic polymer polyetherimide (PEI). The membrane characteristics are related to the DCMD performance.

6.2

CHARACTERISTICS REQUIRED FOR MD MEMBRANES

In one of our recent publications (Khayet et al. 2006), we have speciied the desired characteristics for DCMD membranes. As it is well known, an MD membrane must be porous and hydrophobic, with good thermal stability and excellent chemical resistance to feed solutions. The characteristics needed for DCMD membranes are detailed in the following sections.

Integrating Hydrophobic SMMs into Hydrophilic Polymers for MD

6.2.1

161

HIGH LIQUID ENTRY PRESSURE

This is the minimum hydrostatic pressure that must be applied to the liquid feed solution before it overcomes the hydrophobic forces of the membrane and penetrates into the membrane pores. Liquid entry pressure (LEP) is a characteristic of each membrane and prevents wetting of the membrane pores when it is high. A high LEP may be achieved using a membrane material with high hydrophobicity (i.e., a large water contact angle, CA) and a small maximum pore size. However, as the maximum pore size decreases, the mean pore size of the membrane decreases and the permeability of the membrane becomes low.

6.2.2

HIGH PERMEABILITY

The DCMD lux will “increase” with an increase in the membrane pore size and porosity and with a decrease in the membrane thickness and pore tortuosity. In other words, to obtain a high DCMD permeability, the surface layer that governs the membrane transport must be as thin as possible and its surface porosity as well as its pore size must be as large as possible. However, it must be mentioned here that there exists a critical pore size equal to the mean free path of the water vapor molecules for the given experimental DCMD conditions. In the DCMD process, air is always trapped within the membrane pores with pressure values close to the atmospheric pressure. Therefore, if the pore size is comparable to the mean free path of the water vapor molecules, the molecules of the water vapor collide with one another and diffuse among the air molecules. In this case, the vapor transport takes place via the combined Knudsen/molecular diffusion low. On the other hand, if the pore size is smaller than the mean free path of the water vapor molecules, the molecule–pore wall collisions become dominant and the Knudsen type of low will be responsible for the mass transport in DCMD. It should be noted that for the given experimental conditions, the calculated DCMD lux based on the Knudsen mechanism is higher than that based on the combined Knudsen/molecular diffusion mechanism. For example, for the commercial PTFE membrane TF200 (a mean pore size of 0.2 μm and a porosity of 80%) and a bulk temperature difference of 10 K, when considering the Knudsen lux, the calculated DCMD lux is 2.3 (for a feed temperature of 45°C) to 2.6 (for a feed temperature of 20°C) times higher than the corresponding DCMD lux calculated from the combined Knudsen/molecular diffusion lux. For the membrane TF450 having a higher pore size (i.e., a mean pore size of 0.45 μm and a porosity of 80%), the difference between the calculated DCMD luxes is even greater. The calculated lux using the Knudsen type of low is 3.7 (for a feed temperature of 45°C) to 4.3 (for a feed temperature of 20°C) times higher than that using the combined Knudsen/molecular diffusion type of low. This fact indicates that when the pore size of a given membrane is near the critical pore size (i.e., the mean free path of the water vapor molecules), the DCMD lux will not necessarily increase with the increase in the pore size. Under a certain operating condition, it would be better to use membranes with pore sizes lower than the value of the corresponding mean free path of the water vapor molecules so that the Knudsen type of low will take place, leading to a higher DCMD lux compared with that of the

162

Membrane Modification: Technology and Applications

membranes with larger pore sizes where the combined Knudsen/molecular diffusion lux is responsible for mass transfer. Therefore, care must be taken to choose the appropriate membrane pore size, taking into account the value of the mean free path of the water vapor molecules so that the membrane can work under the Knudsen type of low.

6.2.3

LOW THERMAL CONDUCTIVITY

In DCMD, heat loss by conduction occurs through both the pores and the matrix of the membrane. The conductive heat loss is greater for thinner membranes. Various possibilities may be applied to diminish the conductive heat loss by using 1. Membrane materials with low thermal conductivities. This does not necessarily guarantee the improvement of the DCMD process because most hydrophobic polymers have similar heat conductivities, at least in the same order of magnitude. 2. Membranes with a high porosity, since the conductive heat transfer coeficient of the gas entrapped within the membrane pores is an order of magnitude smaller than that of the membrane matrix. This possibility is parallel to the need for high DCMD permeability as the available surface area of evaporation is enhanced. 3. Thicker membranes. However, there is a conlict between the requirements of a high mass transfer associated with thinner membranes and a low conductive heat transfer through the membrane obtained by using thicker membranes. 4. Composite porous hydrophobic/hydrophilic membranes, having a very thin hydrophobic layer that is responsible for the mass transfer and a thick hydrophilic layer, the pores of which are illed with water, to prevent heat loss through the overall membrane. This seems to be a relatively simple solution that fulills all of the above conditions for achieving high permeability and low thermal conductivity. The composite hydrophobic/hydrophilic membranes could be made by blending hydrophobic SMMs into a hydrophilic polymer solution, which is the cornerstone of this work.

6.3

SURFACE-MODIFYING MACROMOLECULES

It is well documented that the surface chemistry and morphology of the membranes play an important role in the transmembrane transport of permeates (Khayet and Matsuura 2003a,b). To enhance the overall performance of a membrane, it is often necessary to modify the membrane material or its structure. Generally, the objective of modiication is not only to increase the lux and/or selectivity, but also to control the pore size, eliminate defects, and improve the chemical resistance, for example, the solvent resistance, swelling, or fouling resistance. The irst reported membrane-modiication method involved the annealing of porous membranes by heat treatment (Pinnau and Freeman 2000). In the membrane literature, various techniques were carried out for the surface modiication of

Integrating Hydrophobic SMMs into Hydrophilic Polymers for MD

163

polymer membranes: physical, chemical, or bulk modiication (i.e., polymer blends) (Ganbassi et al. 1996). Recently, Pinnau and Freeman (2000) gave a summary of some of the most commonly practiced membrane-modiication methods, including surface coating, chemical treatment with luorination, cross-linking and pyrolysis, annealing with heat treatment, and solvent treatment. A less common approach to modifying the properties of a polymer is to introduce additives that can migrate to the ilm surface and alter the surface chemistry while leaving the bulk properties intact. Ward et al. (1984) were apparently the irst to synthesize a polyurethane block copolymer, to be used as a surface-modifying additive for the development of a new biomedical polyurethaneurea. They showed that this method is eficient because only a small weight percentage of additives was required to modify the surface properties while maintaining the bulk properties unaltered. There are obvious advantages to using surface luorination, although most commercial luoropolymers are dificult to process. In copolymers, luorinated segments are usually enriched at the surface. This fact was the basis for the development of SMMs, which are oligomeric luoropolymers synthesized by polyurethane chemistry and tailored with luorinated end groups. The SMM has an amphipathic structure, consisting, theoretically, of a main polyurethane chain terminated with two low-polarity polymer chains; the luorine segments. It was found that the SMMblended membranes exhibited low surface energies, high chemical resistance, and higher mechanical strength; showed less fouling in the treatment of cutting machine oil/water emulsion by UF; could be used to remove chloroform from water by pervaporation; and demonstrated good potential in biomedical applications. Tang et al. (1996, 1997a,b) designed a series of SMMs to obtain an effective surface modiication of polyester–urea–urethane with improved additive stability. Matsuura et al. (1999) proposed a membrane surface modiication by blending polymer active additives (SMMs) into the polyethersulfone (PES) base material. The goal was to prepare, in a single casting step, membranes with a high hydrophobicity and a chemical resistivity at a low cost and to be competitive with other hydrophobic membranes, such as polydimethylsiloxane (PDMS) or PTFE. Since then, other series of SMM formulations have been synthesized with different combinations and stoichiometries of different reagents and blended with PES. Attempts were made to determine the effects of various types of SMMs and different membrane-casting conditions on the SMM migration to the top membrane surface. The performance of the SMM-modiied membranes was tested for use in UF, in pervaporation, and in biomedical applications. Various SMMs were synthesized and characterized by their molecular weight, luorine content, and glass transition temperature (Tang et al. 1996). They were used to modify polyurethane membranes to improve blood compatibility and reduce biodegradation (Tang et al. 1997a,b). These SMM-blended PES membranes were found to remove chloroform effectively from water by pervaporation (Fang et al. 1994). They were found to be less fouled in the treatment of cutting machine oil/water emulsion by UF (Hamza et al. 1997). The adsorption of humic acid, a typical component of surface water that causes membrane fouling, was found to be reduced by blending an SMM in the PES membranes (Zhang et al. 2003). It was also found

164

Membrane Modification: Technology and Applications

that the mechanical strength of the PES membranes is enhanced by blending an SMM into PES (Suk et al. 2002). The time required for the SMM to migrate to the membrane surface was measured and the kinetics of the surface migration was established (Suk et al. 2006). It was also found that the liquid penetration pressure of water (LEPw) in the porous PES membrane was increased from 0 to about 40 psi by blending an SMM. Therefore, there is a potential to use these membranes for MD (Suk et al. 2006). This technique was found applicable to the PVDF membrane (Ho 1997; Khayet and Mastuura 2003a). Khayet and Matsuura (2001) synthesized three types of SMMs and blended them with the base polymers of PES, PEI, and PVDF. They investigated the effects of the base polymer concentration as well as the SMM content in the polymer casting solution. They tested the developed membranes for UF and pervaporation. Their results showed that enrichment of luorine on the surface of the SMM-modiied membranes took place and that the surface of the SMM-modiied membranes was more hydrophobic than the unmodiied membranes. Recognizing the drawbacks of conventional surface-modiication methods, an alternative approach for membrane surface modiication was proposed by a Toronto/Ottawa University team (Matsuura et al. 1999). According to the new approach, SMMs were synthesized and blended into polymer solutions. While the polymer solution is cast into a ilm and the solvent is evaporated, the SMM migrates to the membrane surface, to reduce the surface energy, rendering the surface of the membrane ultimately obtained more hydrophobic than the bulk membrane phase. This type of SMM has a copolymeric nature. It has an amphipathic structure consisting of a main polyurethane chain terminated with two hydrophobic luorocarbon chains. When the above SMM was blended in the PES membrane, it was found that the CA of the surface increased from 76° of PES to 116°, which is nearly equal to that of Telon. XPS analysis also conirmed that SMMs were concentrated at the surface of the membrane. Proposals were made by a series of publications to fabricate MD membranes using SMMs. Khayet and Matsuura (2001) prepared PVDF lat-sheet membranes for MD. They used pure water as a pore-forming additive in the casting solution, while dimethylacetamide (DMAc) was used as the solvent. Their polymer solutions were cast over a glass plate or over a nonwoven polyester backing material. They characterized the prepared supported and unsupported PVDF membranes in terms of nonwettability, pore size, and porosity. The prepared membranes were tested in VMD experiments for the separation of chloroform from a chloroform/water solution. They have also studied the dependence of the MD lux and the separation factor on the geometrical properties of the supported and unsupported membranes. Khayet and Matsuura (2003a) and Khayet et al. (2005a,b) proposed a new type of composite hydrophobic/hydrophilic membrane for MD. They modiied the top surface of the PEI lat-sheet membranes using luorinated SMMs (Cheng and Wiersma 1982, 1983). The SMM-modiied and the SMM-unmodiied membranes were prepared by the phase inversion method from casting solutions containing the solvent DMAc and the nonsolvent γ-butyrolactone (GBL). They concluded that the SMM-modiied PEI membranes have the potential to be used in MD. The performance of the new membranes was compared with two commercial PTFE membranes. The authors’ results conirmed that the new membranes are promising for MD.

Integrating Hydrophobic SMMs into Hydrophilic Polymers for MD

165

Other research groups also performed surface modiication of membranes for MD. The following is a brief summary of their efforts. Cheng and Wiersma (1982, 1983) patented the composite MD membranes consisting of a hydrophobic top layer made of PTFE or PVDF and a hydrophilic sublayer made of cellulose acetate, polysulfone (PS), cellulose nitrate, or polyallylamine. Wu et al. (1992) treated the surfaces of the hydrophilic porous membranes, such as cellulose acetate, by radiation graft polymerization of styrene to increase their hydrophobicity and to reach the MD membrane characteristics. Kong et al. (1992) employed a cellulose nitrate membrane modiied via plasma polymerization of both vinyltrimethylsilicone and carbontetraluoride and octaluorocyclobutane for the preparation of MD membranes. Fujii et al. (1992) prepared tubular membranes from PVDF polymer dopes by using the dry-jet wet-spinning technique. Ortiz de Zarate et al. (1995) and Tomaszewska (1996) reported on PVDF lat-sheet membranes prepared for MD by the phase inversion method.

6.4

SMM-MODIFIED MEMBRANES FOR DCMD

In summary, the MD membrane requirements are as follows: 1. Water should be repelled on the membrane surface so that the membrane pores are not wetted. This requires a suficiently high hydrophobicity of the membrane material and small enough pore sizes. 2. On the other hand, the membrane pores should be as large as possible to reduce the barrier resistance. These two are contradictory requirements. 3. Membranes should be as thin as possible to reduce the barrier resistance to mass transfer. 4. Membranes should be as thick as possible to reduce the heat conductance, which will lead to a maximum temperature difference between both sides of the membrane. This is contradictory to the third requirement. A relatively simple way to fulill all of the MD requirements, although some of them are mutually contradictory, is to use porous hydrophobic/hydrophilic composite membranes. The top hydrophobic thin layer will prevent the penetration of water into the pores. On the other hand, resistance to the mass transfer is minimized because of the thinness of this hydrophobic layer. Both the hydrophobic and hydrophilic layers will contribute to the overall resistance to the heat transfer. Hence, the heat conductance can be reduced by using a relatively thick hydrophilic sublayer. The approach we have adopted in designing the novel hydrophobic/hydrophilic composite membranes is based on membrane modiication by blending hydrophobic SMMs into a hydrophilic polymer solution. While the polymer solution is cast into a ilm and the solvent is evaporated, the SMM migrates to the membrane surface, to reduce the surface energy, rendering the surface of the membrane ultimately obtained more hydrophobic than the bulk membrane phase. It was found that the SMM-blended membranes exhibited low surface energies, high chemical resistance, higher mechanical strength, showed less fouling and could be used to

166

Membrane Modification: Technology and Applications

remove chloroform from water by pervaporation, and demonstrated good potential in biomedical applications. Considering the obvious advantages of the membrane surface modiication by SMM and the applicability of this principle in manufacturing hydrophobic/hydrophilic membranes for MD, we synthesized four different types of hydrophobic SMMs (nSMM1, nSMM2, nSMM3, and SMM41). These SMMs were then blended into PEI, which is a hydrophilic polymer, to produce novel hydrophobic/hydrophilic composite membranes by the phase inversion method. Finally, these membranes were tested for the desalination of a 0.5 M NaCl solution by DCMD. In this chapter, the experimental details of the SMM synthesis and characterization as well as the membrane preparation, characterization, and testing are thoroughly described in the following examples. Moreover, the effects of the SMM type on the membrane morphology as well as its performance were clearly identiied. Furthermore, the performance of the newly developed membranes was compared to a commercial PTFE membrane in terms of the permeate water lux and the separation factor.

6.5

CASE STUDY: FABRICATION, CHARACTERIZATION, AND APPLICATION OF SMM-MODIFIED MEMBRANES

All chemicals used in this work and their chemical abstract service (CAS) number are listed in Table 6.1. The hydrophilic polymer used is PEI having a weight average molecular weight (Mw) of 15 kDa and a glass transition temperature (Tg) of 216.8°C. The commercial membrane used is PTFE (FGLP 1425) having a porosity of 0.70 and a nominal pore size of 0.25 μm, supplied by Millipore Corporation, Billerica, MA, USA. Four different types of SMMs were synthesized; the irst SMM was named SMM41, in which polypropylene glycol (PPG) was used to form the main polyurethane chain. The other three SMMs, namely, nSMM1, nSMM2, and nSMM3, used PDMS in three different stoichiometric ratios to form the main polyurea chain. The SMMs were synthesized using a two-step solution polymerization method (Carman 1956; Fang et al. 1994; Khayet et al. 2005a,b; Suk et al. 2006; Qtaishat et al. 2009a,b). The solvent DMAc was distilled at about 25°C under a pressure of 133.3 Pa. Methylene bis(p-phenyl isocyanate) (diphenylmethane diisocyanate; MDI) was also distilled at 150°C under 66.7 Pa (0.5 Torr). PPG, α,ω-aminopropyl PDMS, and 2-(perluoroalkyl)ethanol (FAE) were degassed for 24 h under 66.7 Pa. The irst polymerization step was conducted in a solution with a predetermined composition to form polyurethane from the reaction of MDI with PPG or to form polyurea from the reaction of MDI with PDMS as a prepolymer. In the second polymerization step, the prepolymer was end-capped by the addition of FAE, resulting in a solution of SMM. The composition of SMM41 is 3(MDI):2(PPG):2(FAE). On the other hand, the stoichiometric molar ratio of the nSMM components was altered systematically in nSMM synthesis: 1. nSMM1: 2(MDI):1(PDMS):2(BAL) 2. nSMM2: 3(MDI):2(PDMS):2(BAL) 3. nSMM3: 4(MDI):3(PDMS):2(BAL)

167

Integrating Hydrophobic SMMs into Hydrophilic Polymers for MD

TABLE 6.1 Materials Used in This Work Material Description

CAS No.

4,4′-Methylene bis(phenyl isocyanate) (MDI, 98%) Poly(propylene glycol) (PPG, typical Mn 425 Dalton) α,ω-Aminopropyl poly(dimethyl siloxane) (PDMS) of average molecular weight 900 Zonyl luorotelomer intermediate, 2-(Perluoroalkyl)ethanol, (FAE, BA-L of average Mn 443 and 70 wt% luorine) N,N-Dimethylacetamide (DMAc, anhydrous 99.8%) 1-Methyl-2-pyrrolidinone (NMP, anhydrous 99.5%) γ-Butyrolactone (GBL, 99%+)

Source

101-68-8

Sigma-Aldrich, Inc., St. Louis, MO

25322-69-4

Aldrich Chemical Co. Inc., Milwaukee, WI Shin-Etsu Chemical Co. Ltd., Tokyo, Japan

106214-84-0

Tetrahydrofuran (THF, HPLC grade 99.9%) Polyetherimide (PEI, Ultem 1000, Natural Pallet) speciic gravity: 1.27

678-39-7

DuPont product supplied by Aldrich Chemical Co., Inc., Milwaukee, WI

127-19-5

Sigma-Aldrich, Inc., St. Louis, MO

112-14-1

Sigma-Aldrich, Inc., St. Louis, MO

96-48-0

Aldrich Chemical Co., Inc., Milwaukee, WI Aldrich Chemical Co., Inc., Milwaukee, WI General Electric Co., Pittsield, MA

109-99-9 61128-46-9

The chemical structure of the synthesized SMMs is shown in Figure 6.1. The elemental analysis of the luorine content in the four prepared SMMs was carried out using the standard method in ASTM D3761 (Khayet et al. 2005b; Qtaishat et al. 2009a,b). An accurate weight (10–50 mg) of sample was placed into an oxygen lask bomb combustion calorimeter (Oxygen Bomb Calorimeter, Gallenkamp). After pyrohydrolysis, the luorine (ion) was measured by ion chromatography (Ion Chromatograph, Dionex DX1000). The glass transition temperature (Tg) was H

O H

H

H O

C m C 2O H F

C N

C

N C O

F F

CH3

O H

H

H O

H

CH2 CH O n

C N

C

N C O

C 2 C mF H F

H

q

H

F

SMM41 F

H

F C m C 2O H F

O H C N

H C H

CH3 H H O H CH3 H O H H C 3N C N N C N C 3 Si O y Si H CH3 H CH3 q

H C H

H O

H

N C O

C H

F 2

C mF F

nSMM1, 2 and 3

FIGURE 6.1 The chemical structure of the prepared surface-modifying macromolecules. (From Qtaishat, M., Rana, D., Khayet, M., and Matsuura, T., J. Memb. Sci., 327, 264–273, 2009a. With permission.)

168

Membrane Modification: Technology and Applications

determined by a differential scanning calorimeter (DSC) equipped with the Universal Analysis 2000 program (DSC Q1000, TA Instruments, New Castle, DE). About 10 mg of polymer was crimped into an aluminum pan. The SMMs were annealed at 280°C for 10 min, then quenched to –50°C and scanned at a heating rate of 10°C/min. The Tg value was recorded at the midpoint of the corresponding heat capacity transition. The average molecular weight of the synthesized SMMs was measured by gel permeation chromatography (GPC) using a Waters Associates GPC chromatograph equipped with a Waters 410 refractive index detector. Three Waters UltraStyragel packed columns were installed in the series. The tetrahydrofuran (THF) was iltered and used at 40°C and at a low rate of 0.3 ml/min. First, the calibration of the system was performed using polystyrene (Shodex, Tokyo, Japan) standards with different molecular weights between 1.3 × 103 and 3.15 × 106 g/mol. The standards and SMM samples were prepared in a THF aqueous solution (0.2 w/v%) and iltered prior to injection through a 0.45 μm ilter to remove high-molecular-weight components. Millenium 32 software (Waters) was used for data acquisition. The values of the glass transition temperature (Tg) and the weight average molecular weight (Mw) for the SMMs (SMM41, nSMM1, nSMM2, and nSMM3) are given in Table 6.2. The precise Tg value could not be obtained for nSMM1, nSMM2, and nSMM3 as the samples were only heated up to 280°C due to the higher temperature limit of the equipment. According to the SMMs chemical composition presented in Figure 6.1, the value of m, the repeating unit of CF2, was calculated from the molecular weight of FAE, which is 7.58. The values of the n repeating unit of PPG in SMM41 and the values of y, the repeating unit of dimethylsiloxane in nSMM1, nSMM2, and nSMM3, were calculated from the average molecular weights of PPG and PDMS, which are 7.02 and 9.81, respectively. Based on the SMMs average molecular weight, q (the number of the repeating unit of the polyurethane in SMM41 or the polyurea in nSMM1, nSMM2, and nSMM3) was estimated for each SMM; the q values are shown in Table 6.2. The SMM-modiied PEI membranes were prepared in a single casting step by the phase inversion method (Carman 1956; Cheng and Wiersma 1982, 1983; Khayet et al. 2005b; Qtaishat et al. 2009a). GBL was used as a nonsolvent additive. About 12 wt% amount of PEI was dissolved in an NMP/GBL mixture. The PEI concentration in the casting solution was maintained at 12 wt%, while the amount of GBL was maintained at 10 wt%. Four different types of SMMs, SMM41, nSMM1, nSMM2, and nSMM3, were added to the PEI solution in a concentration of 1.5 wt% of the casting solution. The resulting mixtures were stirred in an orbital shaker at room temperature for at least 48 h. Prior to their use, all the resulting polymer solutions TABLE 6.2 Characteristics of SMMs SMM Type

Tg (oC)

F (wt%)

Mw (104 g/mol)

q

SMM41 nSMM1 nSMM2 nSMM3

19.35 >280 >280 >280

11.45 16.21 11.75 10.06

3.61 2.95 2.71 3.30

50.60 24.67 22.58 27.70

Integrating Hydrophobic SMMs into Hydrophilic Polymers for MD

169

were iltered through a 0.5 μm Telon ilter and degassed at room temperature. The polymer solutions were cast on a smooth glass plate to a thickness of 0.30 mm using a casting rod at room temperature. Then, the cast ilms together with the glass plates were immersed for 1 h in distilled water at room temperature. During gelation, it was observed that the membranes peeled off from the glass plate spontaneously. All of the membranes were then dried at ambient conditions for 3 days. Table 6.3 shows the prepared membranes, their materials of construction, and the preparation conditions. The CA of the SMM-blended membranes was measured to study their hydrophobicity/hydrophilicity. The CA measurements were made using the VCA-Optima (AST Products, Inc., Billerica, MA, USA). Samples of 4 cm2 area (2 × 2 cm) at random positions were prepared from each membrane. The samples were then placed on a glass sample plate and ixed with scotch tape. The equipment syringe, illed with distilled water, was installed to stand vertically. About 2 μl of water was deposited on the membrane surface. The CA was measured at ive different spots on both the top and the bottom surfaces of each membrane sample. The resulted CA data of all the membranes are shown in Table 6.4. It was observed that the CA of the top side of the prepared membranes is higher than that of their bottom side in all of the SMM-blended membranes. The higher CA of the membranes prepared with nSMMs (M2, M3, and M4) compared with that of the membrane prepared with SMM41 (M1), prepared under the same conditions, indicates that the hydrophobicity of the nSMMs is greater than that of SMM41. This is logical, since the top-layer hydrophobicity of nSMM-blended membranes is not a result of the luorine content only, but could be a result of the siloxane group as well. The elemental composition at the surface of each SMM-blended membrane was determined by x-ray photoelectron spectroscopy (XPS; Kratos Axis HS x-ray photoelectron spectrometer, Manchester, UK). Each membrane was cut into samples of 1 cm 2 from random positions of the membrane. Monochromatized Al Kα x-radiation was used for excitation and a 180° hemispherical analyzer with a threechannel detector was employed. The x-ray gun was operated at 15 kV and 20 mA. The pressure in the analyzer chamber was 1.33 × 10 –4 to 1.33 × 10 –5 Pa. The size of the analyzed area was about 1 mm2. All the membrane samples were analyzed for their luorine content on both the top and the bottom sides. TABLE 6.3 Membrane Preparation Details: Casting Solution Composition and Preparation Conditions Membrane Code M0 M1 M2 M3 M4

SMM Type — SMM41 nSMM1 nSMM2 nSMM3

SMM Conc. (wt%) 0 1.5 1.5 1.5 1.5

Note: PEI concentration: 12 wt%; GBL concentration: 10 wt%; NMP: 78 wt%; gelation bath temperature: 20°C.

170

Membrane Modification: Technology and Applications

TABLE 6.4 Top and Bottom Contact Angles (CA) and Fluorine Content of the Prepared Membranes CA (θo) Membrane M0 M1 M2 M3 M4

F (wt%)

Top Surface

Bottom Surface

Top Surface

80.04 ± 4.55 89.31 ± 3.91 93.55 ± 1.054 91.93 ± 0.52 100.17 ± 3.62

72.83 ± 2.62 78.69 ± 2.80 62.84 ± 3.05 67.76 ± 3.29 73.80 ± 4.08

0.00 23.71 30.80 28.93 5.82

Bottom Surface 0.00 0.21 4.62 7.68 4.89

The results of the XPS analysis for both the PEI and the nSMM-blended PEI membranes are also presented in Table 6.4. Fluorine was not detected in the membrane that was prepared without SMMs (M0). This is expected since luorine is associated with the SMMs alone. For all the SMM-blended membranes, the luorine contents were found to be higher at the top side than at the bottom side, indicating the migration of SMMs toward the top layer of the membranes. The nSMM1-blended membrane (M2) exhibited more luorine than the SMM41-blended, nSMM2-blended, and nSMM3-blended membranes (M1, M3, and M4). This is related to the order in the luorine contents of nSMM1 compared with the other SMMs, as shown in Table 6.2. A cross section of the SMM-blended PEI membranes was analyzed by scanning electron microscopy (SEM; JSM-6400 JEOL, Japan). The membranes were cut into pieces (3 mm width and 10 mm length) and subsequently immersed in a liquid nitrogen reservoir for 5 sec. While keeping the pieces in the liquid nitrogen, they were broken into two pieces by pulling from both ends. One of the broken pieces was mounted on a metal plate with carbon paste and was gold-coated prior to use. Finally, the cross section of the membranes at the broken parts was examined by SEM. The SEM images of the cross section of the SMM-blended membranes are presented in Figure 6.2. All the membranes are of an asymmetric structure with ingerlike structures at the top surface, whereas the structure at the bottom surface varies depending on the SMM type. For instance, nSMM-blended membranes (see Figure 6.2b–d) exhibited a inger-like structure at the bottom with fully developed macropores. The macropore size of the nSMM2-blended membrane (M3) was larger than those of the nSMM1/PEI (M2) and nSMM3/PEI (M4) membranes. The SMM41blended membrane (M1) exhibited a structure different from the nSMM-blended membranes, since unlike the nSMM-blended membranes, a sponge-like structure was found at the bottom of the cross section. The measurement of the LEPw and the gas permeation test were carried out for the prepared surface-modiied PS membranes. The gas permeation test was performed prior to the measurement of LEPw. Figure 6.3 shows a schematic diagram of the testing setup (Qtaishat et al. 2009a). The effective area of the membrane was 9.6 cm2. First, the equipment was made ready for the gas permeation test by illing the pressure vessel with nitrogen at a pressure of 5 bars. Then, the stainless steel membrane holder inlet valve was opened and the inlet nitrogen pressure was

Integrating Hydrophobic SMMs into Hydrophilic Polymers for MD

(a)

(b)

(c)

(d)

171

FIGURE 6.2 SEM images of the cross section of SMM-blended membranes: (a) M1, (b) M2, (c) M3, and (d) M4. (From Qtaishat, M., Rana, D., Khayet, M., and Matsuura, T., J. Memb. Sci., 327, 264–273, 2009a; Qtaishat, M., Rana, D., Matsuura, T., and Khayet, M., AIChE J. 55(12), 3145–3151, 2009b. With permission.)

Pressure regulator Pressure gauge P P

Flowmeter Pressure vessel Filtration cell

Nitrogen cylinder

FIGURE 6.3 A schematic diagram of the experimental setup for the gas permeation and liquid entry pressure of water tests. (From Qtaishat, M., Khayet, M., and Matsuura, T., J. Memb. Sci., 341, 139–148, 2009c. With permission.)

172

Membrane Modification: Technology and Applications

controlled with the help of the pressure gauge. The lux of the nitrogen lowing from the bottom of the membrane was measured by an air low meter. The permeation lux of the nitrogen through each dry membrane was measured at various transmembrane pressures, in the range of 10–100 kPa. In general, the gas permeance (B) for a porous medium contains both a diffusive term (Knudsen diffusion) and a viscous term, the contribution of which depends on the applied pressure, as reported by Carman (1956): B=

4 2  3  πMRT 

0.5

rε P r 2ε , + m Lp 8µRT Lp

(6.1)

where R is the gas constant, T is the absolute temperature, M is the molecular weight of the gas, μ is the gas viscosity, Pm is the mean pressure within the membrane pore (obtained as the average of the feed and the permeate pressure), r is the membrane pore radius, ɛ is the porosity, and L p is the effective pore length. Throughout all of the gas permeation experiments, it was noticed that the gas permeance was independent of the feed pressure; in other words, the gas permeance was independent of Pm. Therefore, the diffusive mechanism seems to dominate the gas transport through the membrane pores, revealing the fact that the prepared membranes in this study have small pore sizes. Accordingly, the gas permeance will be described after omitting the viscous term in Equation 6.1 as (Equation 6.2) (Khayet et al. 2005a; Qtaishat et al. 2009a,b): B=

4 2  3  πMRT 

0.5

rε . Lp

(6.2)

This test was therefore useful in evaluating the ratio (rɛ/L p). Some of the gas permeation experiments were duplicated using different membrane sheets made from the same casting solution batch in order to evaluate the variance of the obtained values from one batch to another. Moreover, for each membrane, the measurement of the gas low rate was made three times at a given gas pressure and the average value was reported as the membrane permeance. The equipment was then made ready for measuring the LEPw; the pressure vessel was illed with 2 l of water. Then, pressure was applied from the nitrogen cylinder to the water. The pressure was increased by 0.1 bar incrementally until the water penetrated through the membrane and left the iltration cell. As soon as the water started to low, the pressure was recorded and this was considered as the LEPw for that membrane. The experiment was performed three times using three different sheets made from the same casting solution batches. The results were averaged to obtain LEPw. The data for LEPw and (rɛ/L p) are summarized in Table 6.5. The related discussions together with the DCMD data will be reported later on. The prepared SMM-blended PEI membranes were tested by the DCMD setup shown in Figure 6.4 (Qtaishat et al. 2009a). The performance of each membrane was compared with that of the commercial PTFE membrane in terms of the water vapor

173

Integrating Hydrophobic SMMs into Hydrophilic Polymers for MD

TABLE 6.5 LEPw and Effective Porosity/Pore Size Ratio of the Prepared Membranes LEPw (bar)

ɛr/Lp (106)

M1

3.7

70.9

M2

4.5

M3

4.0

M4

4.7

Membrane

Manometer

Tf, out

6.02 15.3 2.74

Membrane module Tp, in

Membrane cells Tp, out

Flowmeter

Thermostat Manometer Tf, in Thermostat Flowmeter

Heat exchanger

Feed

Feed pump

Pump

Permeate

FIGURE 6.4 The DCMD experimental setup. (From Qtaishat, M., Rana, D., Khayet, M., and Matsuura, T., J. Memb. Sci., 327, 264–273, 2009a. With permission.)

174

Membrane Modification: Technology and Applications

lux and the NaCl separation factor. The membrane modules were a system of three circular stainless steel cells each composed of two cylindrical chambers. Each cell had an O-ring to prevent water leakage, one inlet and one outlet at the feed side and one inlet and one outlet at the permeate side. The diameter of each cell was around 10 cm, which resulted in a total effective area (of three cells) of 235.6 cm 2. Each cell had two supportive compartments with a thickness of 2.5–3 cm. The feed chamber was connected to a heating system through its jacket to control the temperature of the liquid feed. The permeate chamber was connected to a cooling system to control the temperature of the permeate side stream. The membranes were placed between the two chambers (feed side and permeate side). The feed and permeate low rates were held constant at 1 l/min for each cell (the total low rate was 3 l/min). The inlet and the outlet temperatures of both the feed and permeate solutions were measured, after steady state was reached, using thermocouples connected to a digital meter with an accuracy of ±0.05°C. The MD module cells and all the tubes were insulated to prevent heat loss. The permeate lux was measured by monitoring the water level in both the feed and permeate cylindrical graduated containers. The loss in the amount of water in the feed container should be equal to the gain in the amount of water in the permeate container when there is no leakage of water in the DCMD setup. Different sets of DCMD experiments were carried out using distilled water and a 0.5 M NaCl aqueous solution as the feed. When distilled water was used, the feed temperature was varied from 35°C to 65°C, while the permeate temperature was maintained at 11°C–15°C. When 0.5 M of the NaCl solution was used as the feed, the feed temperature was 65°C and the permeate temperature was 15°C. Figure 6.5 shows the DCMD luxes of the SMM-blended membranes along with that of the commercial membrane (FGLP 1425). Figure 6.5a shows the DCMD lux vs. the feed inlet temperature when distilled water was used as the feed. Figure 6.5b shows the DCMD lux of the same membranes when using a 0.5 M NaCl aqueous solution as the feed. As can be observed, both the commercial membrane and the SMM-blended membranes exhibit an exponential increase of the DCMD lux with an increase in the feed inlet temperature. Figure 6.5a and b show that the order in the DCMD lux is M1 > M3 > FGLP 1425 > M2 > M4. The SMM41-blended membrane (M1) lux is superior to that of the nSMMs blended for DCMD application. In particular, the DCMD lux of the M1 membrane was found to be 55% higher than that of the commercial membrane. The permeate lux, when using an NaCl aqueous solution as the feed, was 25%– 30% lower than the obtained permeate lux when distilled water was used as the feed, relecting the lower vapor pressure of the salt solution. Another reason for the decrease in the DCMD lux is the concentration polarization due to the presence of the NaCl solute in the feed membrane side (Carman 1956; Khayet et al. 2005b; Qtaishat et al. 2009a,b). Referring to the experiments with a salt solution, the solute separation factor (Equation 6.3) is deined as  C  α =  1 − p  × 100 ,  Cf 

(6.3)

Integrating Hydrophobic SMMs into Hydrophilic Polymers for MD

175

9 M1

8

M2 M3 M4

Jw (10–6 m/s)

7 6

FGLP 1425

5 4 3 2 1 0 30

35

40

45

50

55

60

Tf (°C)

(a)

7

Jw (10–6 m/s)

6 5 4 3 2 1 0 (b)

FGLP 1425

M1

M2

M3

M4

FIGURE 6.5 SMM-blended membrane performance in DCMD: (a) feed temperature effect on the DCMD lux of a distilled water feed solution and (b) the water vapor lux of a 0.5 M NaCl feed solution at a Tf of 65°C and a Tp of 15°C.

where Cp and Cf are the NaCl concentrations in the permeate and in the bulk feed solution, respectively. It was observed that α was above 99% for the SMM-blended and commercial membranes. This indicates that the SMM-blended membranes are promising MD membranes. On the other hand, the M0 membrane, prepared without blending the SMMs, did not show any salt rejection capacity, indicating that SMMs blending is essential for preparing workable MD membranes. As shown in Table 6.5, LEPw increases on the order of M1 (3.7 bar) < M3 (4.0 bar) < M1 (4.5 bar) < M4 (4.7 bar), indicating that the SMM41-blended membrane has the maximum pore size according to the Laplace equation (Qtaishat et al. 2008). As for the ɛr/Lp values, the observed decreasing order is M1 > M3 > M1 > M4. Therefore, it can be concluded that the membrane exhibiting a higher ratio (ɛr/Lp) will have a higher DCMD lux. This is not unexpected since an increase in the ratio (ɛr/Lp) means an

176

Membrane Modification: Technology and Applications

increase in either the porosity or the pore radius or a decrease in the effective pore length. Considering that the value of the ratio (ɛr/Lp) was obtained from the gas permeation experiments, a parallel relationship was found between the gas transport and the vapor transport. Furthermore, according to the SEM images shown in Figure 6.2, the M1 membrane exhibited a sponge-like structure in the bottom sublayer, which could have participated in producing the higher lux of this membrane when compared with the other membranes that exhibited a inger-like structure in the bottom sublayer. The sponge-like structure will reduce the barrier resistance toward mass transfer and enhance the lux eventually. Indeed, the M1 membrane showed the highest lux among the SMM-blended membranes. It is generally accepted that the top skin layer of the asymmetric membrane controls the barrier resistance and the porous sublayer only provides the mechanical strength to the membrane. The above observation, however, calls our attention to the contribution of the porous sublayer to the overall membrane resistance. Another interpretation is that there is a close relationship between the structure of the top skin layer, which cannot be observed by SEM, and that of the porous sublayer.

6.6

CONCLUSION

Novel composite hydrophobic/hydrophilic membranes made speciically for the MD process were prepared and successfully tested for desalination application by DCMD. This was achieved through blending different types of SMMs into PEI, a hydrophilic polymer. A better and instructive understanding of the performance of the hydrophobic/ hydrophilic membranes in MD has been obtained by investigating the relationship between the membrane morphology and its performance in MD. The linkage between the membrane characteristics and the membrane performance was coherent. It was veriied that the characteristics of the top skin layer (the hydrophobic layer) highly inluence the DCMD lux. These characteristics are, in particular, the LEPw, the product of the average pore size, and the effective porosity per unit effective pore length (rɛ/L p). Moreover, it was shown that the cross-sectional structure of the membrane played a role in enhancing the DCMD lux in such a way that the membrane with a sponge-like structure at the bottom layer (the hydrophilic layer) exhibited a higher lux. Among the tested SMM-blended membranes, it was found that the SMM41-blended membrane achieved a better lux than those of the nSMM-blended membranes, since it exhibited a larger value of the product of average pore size, an effective porosity per unit effective pore length (rɛ/L p), and a lower LEPw than those of the nSMM-blended membranes. Generally, most of the SMM-modiied PEI membranes exhibited higher permeate luxes than those obtained using the commercial PTFE membrane, although the nSMMmodiied membranes have a considerably lower pore size and porosity. Moreover, the separation factor was found to be higher than 99% for all the tested membranes. It was proved that the SMMs are necessary to produce workable membranes in MD.

Integrating Hydrophobic SMMs into Hydrophilic Polymers for MD

177

REFERENCES Alklaibi, A.M. and Lior, N. 2000. Membrane-distillation desalination: Status and potential. Desalination 171: 111–131. Burgoyne, A. and Vahdati, M.M. 2000. Direct contact membrane distillation. Sep. Sci. Technol. 35: 1257–1284. Carman, P.C. 1956. Flow of Gases Through Porous Media. Butterworth: London. Cheng, D.Y. and Wiersma, S.J. 1982 and 1983. Composite membranes for a membrane distillation system. US Patents 4,316,772 and 4,419,242. Curcio, E. and Drioli, E. 2005. Membrane distillation and related operations: A review. Sep. Purif. Rev. 34: 35–86. El-Bourawi, M.S., Ding, Z., Ma, R. and Khayet, M. 2006. A framework for better understanding membrane distillation separation process. J. Memb. Sci. 285: 4–29. Fang, Y., Pham, V.A., Matsuura, T., Santerre, J.P. and Narbaitz, R.M. 1994. Effect of surface-modifying macromolecules and solvent evaporation time on the performance of polyethersulfone membranes for the separation of chloroform/water mixtures by pervaporation. J. Appl. Polym. Sci. 54: 1937–1943. Fujii, Y., Kigoshi, S., Iwatani, H. and Aoyama, M. 1992. Selectivity and characteristics of direct contact membrane distillation type experiments. I: Permeability and selectivity through dried hydrophobic ine porous membranes. J. Memb. Sci. 72: 53–72. Ganbassi, F., Morra, M. and Occhiello, E. 1996. Polymer Surfaces from Physics to Technology. Wiley: New York. Hamza, A.V., Pham, V.A., Matsuura, T. and Santerre, J.P. 1997. Development of membranes with low surface energy to reduce the fouling in ultrailtration applications. J. Memb. Sci. 131: 217–227. Ho, J.Y. 1997. The effect of surface modifying macromolecules on the blood compatibility of polyethersulfone membrane intended for biomedical applications. MSc dissertation, University of Toronto. Khayet, M. 2011. Membranes and theoretical modeling of membrane distillation: A review. Adv. Colloid Interface Sci. 164: 56–88. Khayet, M. and Matsuura, T. 2001. Preparation and characterization of polyvinylidene luoride membranes for membrane distillation. Ind. Eng. Chem. Res. 40: 5710–5718. Khayet, M. and Matsuura, T. 2003a. Progress in membrane surface modiication by surface modifying macromolecules using polyethersulfone, polyetherimide and polyvinylidene luoride base polymers: Applications in the separation processes ultrailtration and pervaporation. Fluid Particle Sep. J. 15(1): 9–21. Khayet, M. and Matsuura, T. 2003b. Application of surface modifying macromolecules for the preparation of membranes for membrane distillation. Desalination 158: 51–56. Khayet, M., Matsuura, T. and Mengual, J.I. 2005a. Porous hydrophobic/hydrophilic composite membranes: Estimation of the hydrophobic layer thickness. J. Memb. Sci. 266: 68–79. Khayet, M., Mengual, J.I. and Matsuura, T. 2005b. Porous hydrophobic/hydrophilic composite membranes: Application in desalination using direct contact membrane distillation. J. Memb. Sci. 252: 101–113. Khayet, M., Matsuura, T., Mengual, J.I. and Qtaishat, M. 2006. Design of novel direct contact membrane distillation membranes. Desalination 192: 105–111. Kong, Y., Lin, X., Wu, Y., Cheng, J. and Xu, J. 1992. Plasma polymerization of octaluorocyclobutane and hydrophobic microporous composite membranes for membrane distillation. J. Appl. Polym. Sci. 46: 191–199. Lawson, K.W. and Lloyd, D.R. 1997. Membrane distillation. J. Membrane Sci. 124: 1–25. Matsuura, T., Santerre, P., Narbaitz, R., Fan, Y., Pham, V., Mahmud, H. and Baig, F. 1999. US Patent 5,954,966.

178

Membrane Modification: Technology and Applications

Mosqueda-Jimenez, D.B. 2003. Impact of manufacturing conditions of polyethersulfone membranes on inal characteristics and fouling reduction. Ph.D. dissertation, University of Ottawa. Ortiz de Zarate, J.M., Peña, L. and Mengual, J.I. 1995. Characterization of membrane distillation membranes prepared by phase inversion. Desalination 100: 139–148. Pinnau, I. and Freeman, B.D. 2000. Membrane Formation and Modiication. ACS Symposium Series 744. American Chemical Society: Washington, DC. Qtaishat, M., Matsuura, T., Kruczek, B. and Khayet, M. 2008. Heat and mass transfer analysis in direct contact membrane distillation. Desalination 219: 272–292. Qtaishat, M., Rana, D., Khayet, M. and Matsuura, T. 2009a. Preparation and characterization of novel hydrophobic/hydrophilic polyetherimide composite membranes for desalination by direct contact membrane distillation. J. Memb. Sci. 327: 264–273. Qtaishat, M., Rana, D., Matsuura, T. and Khayet, M. 2009b. Effect of surface modifying macromolecules stoichiometric ratio on composite hydrophobic/hydrophilic membranes characteristics and performance in direct contact membrane distillation. AIChE J. 55(12): 3145–3151. Qtaishat, M., Khayet, M. and Matsuura, T. 2009c. Novel composite hydrophobic/hydrophilic polysulfone membranes for desalination by direct contact membrane distillation. J. Memb. Sci., 341: 139–148. Suk, D., Chowdhury, G., Matsuura, T., Narbaitz, R.M., Santerre, P., Pleizier, G. and Deslandes, Y. 2002. Study on the kinetics of surface migration of surface modifying macromolecules in membrane preparation. Macromolecules 35: 3017–3021. Suk, D.E., Matsuura, T., Park, H.B. and Lee, Y.M. 2006. Synthesis of a new type of surface modifying macromolecules (nSMM) and characterization and testing of nSMM blended membranes for membrane distillation. J. Memb. Sci. 277: 177–185. Tang, Y.W., Santerre, J.P., Labow, R.S. and Taylor, D.G. 1996. Synthesis of surface-modifying macromolecules for use in segmented polyurethanes. J. Appl. Polym. Sci. 62: 1133–1145. Tang, Y.W., Santerre, J.P., Labow, R.S. and Taylor, D.G. 1997a. Application of macromolecular additives to reduce the hydrolytic degradation of polyurethanes by lysosomal enzymes. Biomaterials 18: 37–45. Tang, Y.W., Santerre, J.P., Labow, R.S. and Taylor, D.G. 1997b. Use of surface-modifying macromolecules to enhance the biostability of segmented polyurethanes. J. Biomed. Mater. Res. 35: 371–381. Tomaszewska, M. 1996. Preparation and properties of lat-sheet membranes from poly(vinylidene luoride) for membrane distillation. Desalination 104: 1–11. Ward, R., White, K. and Hu, C. 1984. Use of surface modifying additives in the development of a new biomedical polyurethaneurea. In: Polyurethanes in Biomedical Engineering, H. Planck, G. Egbers and I. Syre (eds), pp. 181–200. Elsevier: Amsterdam. Wu, Y., Kong, Y., Lin, X., Liu, W. and Xu, J. 1992. Surface modiied hydrophilic membranes in membrane distillation. J. Memb. Sci. 72: 189–196. Zhang, L., Chowdhury, G., Chaoyang, F., Matsuura, T. and Narbaitz, R. 2003. Effect of surface-modifying macromolecules and membrane morphology on fouling of polyethersulfone ultrailtration membranes. J. Appl. Polym. Sci. 88: 3132–3138.

7

Plasma Modification of Polymer Membranes Marek Bryjak and Irena Gancarz

CONTENTS 7.1 7.2

Introduction .................................................................................................. 179 Plasma Modiication ..................................................................................... 180 7.2.1 Plasma Used for Cleaning and Etching of Membranes .................... 181 7.2.2 Modiication of Membranes in Plasma of Nonpolymerizing Gases .....183 7.3 Inert Gases Case ........................................................................................... 183 7.4 Reactive Gases Case ..................................................................................... 185 7.4.1 Use of Oxygen Plasma ...................................................................... 185 7.4.2 Use of Air Plasma ............................................................................. 186 7.4.3 Use of CO2 Plasma............................................................................ 186 7.4.4 Use of H2O Plasma ........................................................................... 188 7.5 Use of Nitrogen-Based Plasma Systems (N2, NH3) ...................................... 189 7.5.1 Use of Hydrogen and Methane Plasma ............................................. 191 7.5.2 Use of Chlorocompound and Fluorocompound Plasmas ................. 191 7.6 Plasma-Induced Graft Copolymerization ..................................................... 192 7.6.1 Grafting-To Technique ...................................................................... 192 7.6.2 Grafting-From Technique ................................................................. 193 7.6.3 Grafting in Monomer Vapor ............................................................. 194 7.6.4 Grafting in Monomer Solution ......................................................... 195 7.6.5 Grafting through the Plasma Attachment of Adsorbed Polymer ..... 198 7.7 Plasma Polymerization .................................................................................200 7.8 Conclusion ....................................................................................................204 References ..............................................................................................................204

7.1

INTRODUCTION

Plasma is considered to be the fourth distinct state of matter. It can broadly be deined as a gas containing charged and neutral species, including electrons, ions, radicals, atoms, and molecules. The presence of charge carriers makes the plasma electrically conductive. Depending on the method of plasma generation, it reaches temperatures between 1 and 10 eV, electron densities from 103 to 1012 cm–3, and degrees of ionization from 10 –6 to 0.3. In the ield of surface modiication, only cold (no equilibrium) plasma is used. Its degree of ionization is very low (below 10 –4) and the thermal motion of the ions can be ignored. Hence, such plasmas offer very good conditions 179

180

Membrane Modification: Technology and Applications

for chemical reactions. Cold plasma may be generated at low pressure (e.g., glow discharge, capacitively or inductively coupled plasmas) or at atmospheric pressure (corona or dielectric barrier discharges) (Chan et al. 1996). Each substrate in contact with the plasma undergoes surface reconstruction, but the depth of these changes is only several hundred angstroms. This means that the bulk properties of the material stay unchanged. Depending on the conditions, the plasma can remove certain types of contamination from the surface, can smooth it or make it rougher, and can introduce various functionalities on it. All of these processes are very fast and the time of modiication usually takes from a few seconds up to a few minutes. The method uses chemicals in the gaseous form and produces very small amounts of wastes. Hence, among all the techniques of surface modiication, plasma treatment seems to be the most versatile and environmentally friendly; however, this method also has some disadvantages. The process parameters are highly system-dependent. It is not easy to scale it up from an experimental setup to a large reactor. Additionally, the process is extremely complex and its inal effect is the result of many fragmentary processes that can go on simultaneously in plasma. It is not the authors’ intention to present here the problems related to plasma action, its complexity, and its diversity. We would like to focus on the inal effects and demonstrate some beneits that result from plasma treatment. It is well known that the surface chemical and physical properties play a dominant role in the separation characteristics of a membrane. Most of the currently used membranes are made of polymers because they have excellent bulk physical and chemical properties, they are inexpensive, and are easy to process. However, the surface properties of polymers, their hydrophobicity, and their lack of functional groups stand in the way of many other applications (Chan et al. 1996). So far, various polymers have been used for membrane fabrication. However, due to the limited number of polymeric materials on the market, one cannot expect any signiicant increase in the variety of the membranes offered. What is more, large-scale production of brandnew polymers has not been commercialized during the last decade, nor is it expected to be launched in the near future. These observations have forced material scientists to search for alternative methods to increase the number and variety of membranes being prepared. There are two directions for new membrane manufacturing: (i) to modify a polymer in bulk and then prepare the membrane from it or (ii) to prepare the membrane from a standard polymer and then modify its surface. The irst method needs the optimization of the membrane formation for the particular derivative separately. The second seems to be less complicated and less expensive, and it can offer a wide variety of new membranes based on one starting matrix. The authors’ intention is to present the plasma methods for membrane modiication and tailor them based on the end-user requests.

7.2

PLASMA MODIFICATION

As was pointed out in the introduction, the plasma treatment of the polymer membranes becomes an interesting issue for three reasons: the technique is fast, effective, and meets most of the ecological regulations for clean technology. When plasma acts on any polymer surface, two competing processes can take place:

Plasma Modification of Polymer Membranes

181

ablation and deposition. The irst is related to scission and the degradation of the polymer chains, while the second describes the deposition of the polymer layer. The balance between them depends on the kind of plasma used and the selected process parameters. It should be pointed out that both processes result in the creation of functional groups on the surface and, in consequence, the formation of a new polymer membrane. Generally, there are four particular phenomena that can result from the interactions of plasma with the polymer substrate: 1. Surface cleaning and etching 2. Surface modiication with gas plasma (cross-linking and the creation of new functional groups) 3. Plasma-initiated polymerization/grafting 4. Plasma polymerization During exposure to plasma, competition between all of these processes occurs and the inal result strongly depends on the kind of plasma used and the treatment parameters. The above-mentioned processes were commonly used to improve and change the performance of both the homemade and the commercial membranes.

7.2.1

PLASMA USED FOR CLEANING AND ETCHING OF MEMBRANES

Plasma removes the contaminants (that might originate from polymer production, modiication, or storage) from the polymer surface, etching them to form volatile products. The process runs through the physical sputtering and/or chemical reactions and mostly depends on the plasma gas used. When unreactive plasma gases, such as Ar or He, are applied, ablation occurs by the momentum-exchange process. This is a well-known physical sputtering phenomenon. In the case of reactive gases, such as O2, CO2, N2, and H2O, ablation also occurs by chemical etching. The latter process is stronger, causing a higher mass loss. For example, when polyacrylonitrile (PAN) ultrailtration (UF) membranes were exposed to oxygen plasma, the mass loss during treatment was about six times higher than for the argon or helium plasmas (Tran et al. 2007). It was also shown that oxygen plasma treatment has a much greater effect on the surface roughness than the Ar or even the N2 plasmas. Plasma cleaning is often used as a special pretreatment step for the subsequent implantation and deposition of certain species. It is a common procedure in biomedical applications. Usually, inert gases such as helium, neon, and argon are selected for this purpose. Deep plasma etching replaces many wet processes in the production of electronic devices (dry lithography). Plasma cleaning as the irst step in other processes (Inagaki and Katsuoka 1987; Inagaki et al. 1988; Inagaki 1988; Kita et al. 1989; Ogumi et al. 1990; Danilich et al. 1994; Touik et al. 2002; Vallois et al. 2003; Greene and Tannenbaum 2004b; Basarir et al. 2005; Valdes et al. 2008; Ali et al. 2010) was also applied in membrane science. The authors of several papers (Inagaki and Katsuoka 1987; Inagaki et al. 1988; Inagaki 1988; Kita et al. 1989; Danilich et al. 1994; Touik et al. 2002; Valdes et al. 2008) used argon plasma to eliminate water and other contaminants

182

Membrane Modification: Technology and Applications

adsorbed on the substrate surface before the deposition of some plasma polymers (PPs) on the substrate surface. Oxygen plasma was used to clean the PP (Basarir et al. 2005) or the Naion (Ogumi et al. 1990) membranes before the plasma polymerization of various monomers. It was shown that this process improved the adhesion between the membrane surface and the deposited layers. In the work of Vallois et al. (2003), oxygen or oxygen/Ar plasma was applied before the plasma polymerization of an ethylene–ammonia mixture or before the electrodeposition of poly(ethylene imine) on the polymer membranes. Argon plasma was used to clean the porous polyethylene (PE) membranes before oxygen plasma treatment was used to improve the adsorption of polyelectrolytes on the PE membranes (Greene and Tannenbaum 2004a; Greene et al. 2005). Naion etched in argon plasma before palladium sputtering showed a signiicant reduction of methanol crossover to the positive electrode in a direct methanol fuel cell (Choi et al. 2001). The authors speculated that surface roughening, caused by plasma etching, increased the contact resistance between the electrocatalyst layer and the membrane. Plasma treatment also improved the performance of the nonpolymeric membranes. Hydrogen plasma removed the oxidized layer of the nickel membrane, thereby signiicantly improving the permeation of deuterium through it (Hatano et al. 1998). Nanoporous silica membranes (Vycor) modiied by a cross-linked silicone were treated by oxygen plasma. As a result, an SiO2 layer with many micropores and nanopores was formed on the surface. Such material was appropriate for application in gas/vapor separations, reverse osmosis (RO), and nanoiltration (Beltsios et al. 1999). Another beneit from etching is the possibility of tailoring the shape of the pores and hence changing the performance of the whole membrane. For example, the size of the pores in Nylon-6 membranes increased with the progress of the nitrogen plasma treatment (Villeger et al. 2006). When the poly(ethylene terephthalate) (PET) track membranes were etched in the plasma of air, nitrogen, or oxygen, some asymmetric pores were formed and the membranes gained a higher porosity. Eventually, permeate luxes were raised and the authors noted an improvement in the iltration process (Dmitriev et al. 1998, 2002; Kravets et al. 2002, 2003). Moreover, it was possible to focus the etching process either on selected parts of the pores or along their whole length (Kravets et al. 2003). It is worth mentioning here that etching is accompanied by chemical reactions, hence not only the structure of the membrane but also the nature of the surface is altered (Dmitriev et al. 2002). Such properties of the asymmetric cellulose acetate membranes as wettability, permeability, selectivity, or fouling were greatly changed by simultaneous surface etching and modiication (Olde Riekerink et al. 2002). The etching process is sometimes assisted by redeposition of the ablated material in the form of a highly cross-linked layer. In the case of the porous membranes, this can lead to completely or partially closed surface pores. This process was applied to produce RO membranes from porous PAN and to change the gas permeabilities through hollow iber PE membranes (Kramer et al. 1989). Strong etching was observed when polysulfone (PSU) membranes were treated with CO2 plasma. The bulk of the pores were plugged by redeposited fragments of the polymer matrix (Gancarz et al. 1999a; Wavhal and Fisher 2002b). Hydrophilization of the modiied surface was also observed (Gancarz et al. 1999a; Wavhal and Fisher 2002b; Pal et al.

Plasma Modification of Polymer Membranes

183

2008a). Hydrophilic and smoother membranes are less prone to fouling, easier to clean, and show better permeability (Pal et al. 2008b).

7.2.2 MODIFICATION OF MEMBRANES IN PLASMA OF NONPOLYMERIZING GASES Three main phenomena affecting the membrane properties take place during treatment with the plasma of nonpolymerizing gases: 1. Ablation and etching, resulting in an increase in the pore diameter and the porosity. 2. Chemical reconstruction of the surface layer by introducing a variety of functional groups. 3. Deposition of the polymer fragments formed by the volatile products of surface etching, which can result in a lower porosity. Therefore, the pore size and the pore size distribution of a membrane treated by plasma can become larger or smaller, depending on which of the two competing processes prevails (ablation or deposition). The most important advantage of plasma modiication is the fact that changes are present only at the surface layer due to the small depth of penetration of the active particles. This is a very important issue, as it allows the preservation of the mechanical and physicochemical properties of the membranes. In a typical plasma process, hydrogen is irst abstracted from the polymer chain, creating radicals that react to give the various groups introduced onto the surface. The character of the membrane surface is altered and becomes, depending on the plasma gas, more hydrophilic or hydrophobic. Some of the grafted anchor groups can be useful for further reactions or modiications. The changes induced by the plasmas of the nonpolymerizing gases are, however, not stable with time. Chain migration in the membrane surface can result in a gradual deterioration of the plasma-induced surface properties. This process, called aging or hydrophobicity recovery, restores the original character of the surface to the extent that it adapts the composition to interfacial forces. Possible causes of aging include: (i) Migration of the polar groups on the surface to the internal regions of the polymer, whereas the untreated polymer chains move in the reverse direction; (ii) Structural rearrangement that covers the chemical groups introduced at the surface; and (iii) Chemical reactions occurring on exposure to atmospheric oxygen and moisture (Kull et al. 2005; Van der Bruggen 2009). The extent of the aging process depends on the sample storage environment and, in some cases, a high percentage of the generated functionalities remains, even after longtime storage (Dmitriev et al. 2002).

7.3

INERT GASES CASE

Generally speaking, all gases used during plasma treatment can be divided into two groups: inert gases and reactive gases. Helium, neon, and argon are inert gases and their use for cleaning purposes was discussed in Chapter 1. When these gases are

184

Membrane Modification: Technology and Applications

used, the physical degradation of the polymers and the formation of free radicals on the polymer surface are dominating processes. Ablation occurs by a momentum exchange process involving the impinging neutral and ionic particles (e.g., Ar metastables and Ar+ in an argon plasma) and surface atoms or molecules (Kramer et al. 1989). The abstraction of hydrogen from the polymeric substrate generates free radicals at the surface. These in contact with the air create many and various functional groups on the surface, making it more hydrophilic. The interaction of the polymeric radicals can generate a cross-linked layer on the top of the polymer. In contact with the monomers, these radicals may initiate surface-graft copolymerization. Improving the O2/N2 (Kramer et al. 1989; Houston et al. 2002), the CO2/CH4 (Kramer et al. 1989; Matsuyama et al. 1995), and the CO2/He (Hu et al. 2004) separation properties of membranes was one of the applications for Ar plasma. Usually, a reduction of the gas permeability coeficient as an effect of surface cross-linking was observed. A composite membrane made from a porous PSU UF membrane coated with a silicone polymer (Kramer et al. 1989), a natural rubber membrane (Kramer et al. 1989), or a poly(dimethylsiloxane) (PDMS) ilm (Kramer et al. 1989; Houston et al. 2002) treated with the Ar plasma showed higher O2/N2 permeability ratios. This is thought to be an effect of the increase in oxygen solubility in the hydrophilized polymer surface. The permeation of both the CO2 and the CH4 through PDMS (Kramer et al. 1989; Matsuyama et al. 1995) decreased and their selectivity was remarkably improved after the plasma action. The poly(methyl methacrylate) (PMMA) membrane was treated with plasma to suppress the undesirable plasticization effects in a high-pressure gas separation process. The Ar plasma reduced the permeability to a much lower degree than did other cross-linking modiication methods, at the same time signiicantly improving the He/CO2 separation selectivity (Hu et al. 2004). Only polyphosphazene membranes seem to be practically unaffected by the plasma treatment (Houston et al. 2002). A signiicant increase in the polarity of the PDMS nanoiltration membranes after Ar plasma treatment (also in the admixture with H2 or O2) was observed (Aerts et al. 2006). This lowered the retention of two charge-bearing dyes, while the retention for a neutral component increased. The solvent permeability was also affected. The decreased permeability was attributed to the cross-linking and the hydrophilization of the membrane surface. The UF PAN membranes modiied with the Ar or He plasma were characterized by a slightly increased lux and an almost unchanged albumin retention (Tran et al. 2007). The He plasma signiicantly increased the hydrophilicity of the PAN membranes and that effect was relatively stable with time (Ulbricht and Belfort 1995). Argon (alone or with SF6) plasma was used in the preparation of the protonexchange membranes to improve the adhesion between the microporous PE matrices and the sulfonated, hydrogenated, butadiene–styrene block copolymers (Navarro et al. 2008). The properties of the poly(2-chloro-aniline) ion-exchange membranes were improved by the Ar or He plasma treatment. The membrane pore diameter and the number of pores increased, giving better conditions for anion and molecular transportation (Kir et al. 2006). Modiication of the membranes with argon plasma was also used in the area of biomedicine, where often the surface properties decide on biocompatibility. Argon plasma-treated chitosan membranes (extensively studied

Plasma Modification of Polymer Membranes

185

for tissue engineering of the skin) exhibited excellent attachment, migration, and proliferation of the human-skin–derived ibroblasts (Zhu et al. 2005).

7.4

REACTIVE GASES CASE

Gases such as N2, O2, NH3, CO2, H2, H2O, carbon monoxide, nitrogen dioxide, and nitric oxide are thought to be reactive in plasmas. The mechanism of their action is the same as for nonreactive gases—the surface is bombarded with ionized plasma components to generate radical sites. These subsequently react with gas molecules, creating various functionalities, depending on the plasma conditions. Most polymers used for producing membranes are hydrophobic. Conversion of their surfaces to be hydrophilic usually improves the adhesion strength, biocompatibility, and other pertinent properties. The formation of oxygen functionalities is one of the most useful and effective processes in plasma modiication. Oxygen and oxygencontaining gases (O2, air, CO, CO2, H2O, NO, and NO2) are most commonly employed to modify the polymer surfaces. They can produce a variety of highly oxidized functional groups, including C–O, C=O, O–C=O, C–O–O, and –CO3, on the surfaces. Water plasma may be used to incorporate many hydroxyl and CO2–carboxyl functionalities onto the polymer surface. These gases can strongly etch the polymers and the balance of these two processes depends mainly on the plasma parameters used in an experiment.

7.4.1

USE OF OXYGEN PLASMA

It has been shown that UF PSU membranes treated for 20 sec with oxygen plasma showed increased hydrophilicity. X-ray photoelectron spectroscopy (XPS) analysis proved that this improvement was caused by the presence of the hydroxyl, carbonyl, and carboxyl groups on the surface. For such modiied membranes, the low rate of pure water and gelatin increased and the membranes showed fewer fouling properties (Kim et al. 2002a). By using O2 plasma treatment, the UF property of the PAN (Tran et al. 2007) and PET (Touik et al. 2002) track membranes could be improved with the enhancement of the membrane lux. Meanwhile, their rejection of albumin and dextrans was almost maintained. Oxygen plasma treatment has often been used as a irst step in achieving stable modiied-membrane surfaces with controllable and signiicantly high chemical activity. The reactive charged centers (carboxylic functionalities) created by plasma made possible the strong, irreversible adsorption of oppositely charged polyelectrolytes. Cellulose acetate membrane was modiied in such a way with an ultrathin layer of polyallylamine and the pervaporation performance for the water–ethanol mixture was investigated. The selectivity of the membrane greatly increased and it was not accompanied by any large decrease in the water lux. Both the permeate lux and the selectivity depended on the oxygen plasma parameters. The membrane performance was further improved by the deposition of a second polyelectrolyte layer— poly(acrylic acid) (Kusumocahyo et al. 2002). In this way, porous PE membranes were modiied successively with polyethylenimine and poly(acrylic acid) (Greene and Tannenbaum 2004a,b; Greene et al. 2004). These self-assembled bilayers

186

Membrane Modification: Technology and Applications

exhibited excellent stability over time and in various chemical environments. O2 plasma made possible the covalent bonding of immunoglobulin (Greene et al. 2005) or laminin (Huang et al. 2007) on the membrane surface. Adhesion and proliferation of the vero cells were improved in the case of the polyurethane membranes treated with O2 plasma (Ozdemir et al. 2002).

7.4.2

USE OF AIR PLASMA

Air plasma acts more or less like oxygen plasma. It mainly generates carboxyl groups on the polymer surface (Dmitriev et al. 2002). Air plasma is very effective in hydrophilic modiication accompanied by extensive etching and by the implantation of both oxygen- and nitrogen-containing polar groups. The surface can be made amphoteric, that is, being connected to the degrees of ionization of the amino and carboxyl groups at different pH values (Yu et al. 2008a). For PET track membranes treated in air plasma, a decrease in their thickness and an increase in the effective pore diameter were observed. Additionally, the pores became asymmetric. The permeability increased and depended on the pH of the iltered solution. The membrane surface was no longer smooth, because of the faster etching of the amorphous areas than of the crystalline areas (Dmitriev et al. 2002). The surface of the PET membrane becomes hydrophilic and, in properly chosen conditions, the surface properties are stable (Dmitriev et al. 1995). In polypropylene hollow iber microporous membranes (PPHFMMs), both the O and the N functionalities were found and numerous cracks could be seen on the surface. Generally, a decrease in the low rate was observed as a result of faster cake formation and its compaction. The main positive result of the plasma treatment was a signiicant improvement in the membrane regeneration characteristics (Yu et al. 2008b). These results conirm the conclusions from an earlier paper dealing with the UF PE membranes (Bryjak and Gancarz 1994). Bovine serum albumin (BSA) was deposited on the air plasma-modiied membrane to a larger extent than on the untreated membranes, but its removal from the fouled membrane was easier and more effective (up to 97% of lux recovery). The simultaneous development of two processes in the air plasma-surface etching and the formation of chemical active centers caused the increase in both components (polar and dispersive) of the surface tension of the poly(vinyl chloride) microiltration membranes. Etching has been localized at the surface pore outline and does not signiicantly change the pore dimensions (Vladkova et al. 2003).

7.4.3

USE OF CO2 PLASMA

CO2 plasma treatment leads to the incorporation of oxygen into the membrane surface. For the poly(ether sulfone) (PES) membrane, the oxygen concentration increased by ~47% after 30 sec of treatment (Wavhal and Fisher 2002b). The presence of the hydroxyl, acid, ester aldehyde, and ketone groups yielded a signiicant increase in hydrophilicity (Gancarz et al. 1999a; Wavhal and Fisher 2002b; Pal et al. 2008b). The water drop disappeared into the porous membrane immediately (Wavhal and Fisher 2002b). The surface wettability measurements showed a sharp decrease in

Plasma Modification of Polymer Membranes

187

the water contact angle (from 87° to 36° after modiication) following the irst minute of plasma treatment (Gancarz et al. 1999a). Surface titration (contact angle as a pH function) proved the acidic character of the surface (Gancarz et al. 1999a). CO2 plasma is a relatively strong etching medium and sometimes causes redeposition of the etched material on the surface (Gancarz et al. 1999a; Wavhal and Fisher 2002b). Prolonged plasma excitation can cause signiicant damage to the membrane surface, pore enlargement, and widening of the pore size distribution (Gancarz et al. 1999a). The results are, however, contradictory. The results of atomic force microscopy (AFM) measurements described by Pal et al. (2008a) clearly show smoothing of the membrane surface after the treatment. The average roughness of the PES membrane declined from a value of 111.25 to 19.8 nm after long CO2 plasma treatment. Plasma treatment has signiicantly enhanced the permeabilities (up to 28%) of the membranes, compared with the untreated membrane and they retain their enhanced permeabilities for a suficient time; the permeabilities remain almost unchanged, even after 90 days, for the membranes exposed to the plasma for 12 min (Pal et al. 2008b). Membranes modiied in plasma in appropriate conditions show a signiicant improvement in their performance in protein iltration. The fouling is less intense, the lux recovery is better, and the lux reduction during iltration shows lower values (Gancarz et al. 1999a; Pal et al. 2008b). These effects depend on the pH of the iltered solution and are most pronounced for the system in which both the molecules and the membrane surface bear the same charge (Gancarz et al. 1999a). The hydrophilicity effect gained seems to be relatively persistent (Wavhal and Fisher 2002b; Pal et al. 2008b). Very similar effects were attained in the CO2 plasma treatment of hydrophobic membranes made of polypropylene (Yua et al. 2005). Carbonyl and carboxyl groups generated by CO2 plasma treatment increased the membrane hydrophilicity—the contact angle decreased sharply within 30 sec and prolonged treatment did not cause further changes. It was found that the pore size and the porosity increased with the plasma treatment at the beginning of the process and then decreased with the treatment time, which is connected with etching and further deposition of the etched material. Both the tensile strength and the rate of elongation decreased quickly with the increase in plasma treatment time to 2 min, which can be attributed to the scission of the molecular chains on the membrane surface. The relative pure water luxes for all the CO2 plasma-treated PP samples were higher than that for the untreated membranes. To observe the inluence of the surface modiication on membrane performance, iltration experiments of activated sludge in a submerged membrane bioreactor (SMBR) were carried out. It was found that the more hydrophilic the membrane surface, the more easily was the iltration cake washed off the membrane surface. Generally, the CO2 plasma-modiied membranes showed a better iltration performance than the unmodiied PP membranes (Yua et al. 2005). In the work of Olde Riekerink et al. (2002), one type of UF (cellulose triacetate, CTA) membrane and one type of RO (cellulose acetate blend, CAB) membrane were selected for CO2 plasma modiication. CTA ultrailters are well known for their low fouling properties and CAB membranes are widely used for desalination. CO2 plasma treatment resulted in the gradual etching of the membrane’s dense top layer.

188

Membrane Modification: Technology and Applications

In the case of CAB, the top layer could be neatly tailored from nonporous to nanoporous, up to complete removal of the skin. CO2 plasma treatment strongly inluenced the permeation of both types of membranes. A longer treatment time clearly resulted in a higher lux, although the effects on the CTA and CAB membranes were quite different. In the case of the CTA membranes, a tenfold increase of lux was observed within the irst 2 min of plasma treatment. The CAB membranes needed about 15 min of treatment to obtain a measurable water lux improvement.

7.4.4

USE OF H2O PLASMA

H2O plasma treatment also leads to the incorporation of oxygen-containing groups on the surface; therefore, it is used to increase the hydrophilicity of the membranes. Microiltration and UF PSU (Steen et al. 2001b) and PES (Steen et al. 2002) membranes were rendered completely water-wettable as a result of the H2O plasma and no measurable decrease in hydrophilicity was observed, even after long-term aging. An increase in the pore sizes of the membranes was observed after plasma treatment. These results suggest that the increase in hydrophilicity observed for the H2O plasma-treated membranes is indeed a result of the formation of covalently bound hydrophilic functional groups. XPS analysis performed on the treated samples, aged 1 year, shows that the membranes retained their increased oxygen content and the same functional groups were detected in the C 1s spectrum. H2O plasma similarly modiies these membranes regardless of their material, hence this seems to be a general method for rendering asymmetric polymeric membranes permanently and completely hydrophilic (Steen et al. 2001b). By comparison, O2 plasma does not hydrophilize the membranes permanently. Signiicant differences in the degree and permanency of the hydrophilic modiication were observed for the PE membranes. This modiication was not permanent, with changes in the contact angle measurements observed within 48 h after plasma treatment. The penetration of the plasma through the PE membrane was worse, so the two sides of the membrane were not equally changed. Also, in contrast to the PSU membranes, the structural integrity of the PE membranes was not adversely affected by prolonged exposure to the plasma. These differences were explained as the result of the varied structure of the polymers. The energy requirements for bond cleavage increase in the following way: C–S < C–O < C–C < C–H. Consequently, the easiness of the polymer modiication follows a similar trend, and PSU and PES are easier to modify than PE. The permanence of the plasma treatment can also be attributed to the composition of the polymer. Speciically, the rigidity of the aromatic backbone of PSU and PES may reduce the polymer chain migration from the surface to the bulk, thereby retaining the modiied polymer chains at the surface (Steen et al. 2002). A detailed analysis of the reactive species in the H2O plasma and the mechanisms occurring at the plasma–polymer surface interfaces for four porous membranes (PSU, PES, PE, PTFE) are described in the paper by Steen et al. (2001a). H2O plasma was used to modify poly(methyl pentene) hollow iber membranes, making it possible to immobilize an enzyme (carbonic anhydrase) on them. Such treatment signiicantly improved the respiratory assistance devices for CO2 removal

Plasma Modification of Polymer Membranes

189

in lung failure patients. In artiicial lung devices, HFMs are used as the interface between the blood and gas pathways (Kaar et al. 2007). Water was also used as a plasma medium in an admixture with other gases. The PAN UF membrane was treated with a mixture of He/water plasma (Ulbricht and Belfort 1995). The results obtained were as for the He plasma alone, followed by exposure of the membranes to the air. The surface structure of PAN underwent fast and intensive chemical changes—the conversion of the nitrile group to cyclic (ladder) and inally to aromatic structures and the generation of oxygen functionalities. Polymer etching was observed, but compared with other polymers, the rate was relatively low. No water lux increase due to plasma was observed. The reduced permeability after extended treatment was probably caused by the deposition of etched material. The hydrophilicity gained during the plasma action seemed not to change with storage in the air plasma (Ulbricht and Belfort 1995). Sometimes, hydrogen peroxide was used for the modiication of the membranes (Dmitriev et al. 1995). PET track membranes showed signiicant hydrophilization after such treatment.

7.5

USE OF NITROGEN-BASED PLASMA SYSTEMS (N2, NH3)

After treatment with nitrogen and nitrogen-containing gases plasma, the nitrogen moieties were incorporated in the membrane and the presence of such groups as C–N, C=N, and amide, has been observed. Oxygen-bearing functionalities also appear. Surface etching with these plasmas is minimal except for prolonged exposure to the plasma. PET track membranes treated with nitrogen plasma showed an increased water low rate and a larger pore size. The changes however, were smaller than observed for air or oxygen plasma treatment (Kravets et al. 2003). N2 plasma was found to induce cross-linking of the nonporous poly(vinyl alcohol) (PVA) membranes, leading to the effective improvement in water–isopropanol separation (Upadhyay and Bhat 2004). This effect was stronger than in the case of air or oxygen plasma. Membrane selectivity from 10 to 12 for the unmodiied membrane increased to values of about 1500 after nitrogen plasma treatment. The PSU UF membrane under nitrogen plasma treatment (Gancarz et al. 2000) changes its morphology. During the irst 2 min of plasma action, the pore diameter becomes larger and the pore size distribution is wider. Longer exposure to plasma no longer affects the pores, and the rates of ablation and redeposition of the etched material seem to be balanced. The modiied UF membranes are less prone to protein fouling than their untreated analogs. Importantly, this protection is extended over a wide range of pH values of retentate. The reason for this effect is the presence of various kinds of surface functionalities (acidic and basic) after plasma treatment (Gancarz et al. 2000; Bryjak et al. 1999). The hydrophilization gained during plasma treatment was not stable. However, after a preliminary loss during the irst few days of storage, the polar component of the surface tension stays much higher than for virgin PSU. The same PSU membranes were treated with ammonia—pure or mixed with Ar (Bryjak et al. 2002). Both plasmas introduced hydrophilic moieties on the polymer surface. The NH3+Ar plasma was more eficient (XPS showed more oxygen and nitrogen functionalities) but also more aggressive—etching and redeposition of the etched material inside the pores were observed. On the contrary, ammonia

190

Membrane Modification: Technology and Applications

plasma was soft and caused cleaning of the surface and pore enlargement. The BSA iltration performance was greatly improved after NH3 plasma and was independent of the solution pH, which proved the amphoteric character of the surface. The addition of Ar to the plasma gas leaves the membranes with a rather acidic surface and worse iltration indices. The results of the work dealing with commercial PSU UF membranes (Vidaurre et al. 2001) were different. The authors observed severe etching of the membrane surface, leaving many cracks on it. No substantial decrease in the permeability properties for both the N2 and CO2 gases was noticed. Kull et al. (2005) investigated the process of N2, NH3, NH3+Ar, and NH3+O2 plasmas treatment on microporous PES membranes. Their goal was to obtain the complete and permanent hydrophilic structure of these membranes, including the downstream side of the membrane, involving the modiication of the entire structure of the membrane. Only the NH3+O2 plasma fulilled these demands. On the membrane surface modiied with such a plasma, even after 1 year, the water drop disappeared within ~1 sec for both sides of the membrane, proving the high hydrophilicity of the membrane. From XPS data, it can be seen that the elemental composition of both sides of the membrane was the same. Samples aged for 1 year had considerably more oxygen (22.2%–39.5%) and slightly more nitrogen (7.9%–9.1%) than just after treatment (Kull et al. 2005). The performance of the modiied membranes was signiicantly enhanced. Increased lux was observed for membranes treated with different O2/ NH3 gas ratios. Protein fouling was reduced by even 76% on the plasma-treated membranes. Flux recovery from 63% for control membranes increased to even 90% for the 5:3 O2:NH3 plasma treatments (Kull et al. 2005). A microporous PP membrane was treated with plasma in a mixture of nitrogen and hydrogen (1:2 in volume) (Tang et al. 2004). The presence of the amino functional groups on the surface after treatment was proved by XPS, ATR-FTIR, and UV spectra. The density of these groups is ~0.5 μmol/cm2. Relatively extensive etching was observed, which resulted in an increase in the population of micropores per unit area and a more than sevenfold increase in the surface area. In a PPHFMM after NH3 plasma treatment, two obvious peaks appear, namely, the peak at 531.6 eV corresponding to O 1s and the peak at 402.3 eV corresponding to N 1s (Yu et al. 2008a). The water contact angle decreases from 129° for the PP to 90° after the plasma treatment. The average pore size and the porosity increase, showing the appearance of ablation. The surface-modiied membranes show better performances in the iltration of activated sludge. After continuous operation in the SMBR for about 300 h, the lux recoveries after water cleaning are 6.4% higher, while the lux recoveries after caustic cleaning are 13% higher, and the lux ratios after fouling are 17% higher than those of the nascent PPHFMM. These results, although good, are worse than those obtained for the CO2 plasma treatment (Yu et al. 2008b). Treating a microporous polyvinylidene luoride membrane with ammonia or a mixture of nitrogen and hydrogen, Muller and Oehr (1999) were able to introduce some amine groups on the surface, but their amount was not correlated with the total surface-bound nitrogen. Nanoiltration membranes based on poly(ether ether ketone) modiied with a mixture of ammonia with Ar or hydrogen show much better dye iltration parameters than the unmodiied or commercial membranes (Buonomenna et al. 2009).

Plasma Modification of Polymer Membranes

191

Poly(phenylene oxide) (Kumazawa and Yoshida 2000), PMMA (Yamamoto et al. 2003), and PES (Iwa et al. 2004) membranes were treated with NH3 plasma and the gas separation properties of the modiied membranes were examined. The treatment resulted in an increase in the CO2/N2 separation factor as well as the permeability to CO2. In the case of the PDMS gas, the membrane permeability of CO2 decreased, but the selectivity of CO2 over CH4 was found to be remarkably improved irrespective of the plasma gas used (NH3, Ar, N2, O2). The nitrogen plasma treatment seemed to give better selectivity than the ammonia plasma (Matsuyama et al. 1995). The NH3 and N2 plasma treatment of the dense PE (Nakata and Kumazawa 2006) and PP (Teramae and Kumazawa 2007) membranes increased both the permeation coeficient for CO2 and the ideal separation factor for CO2 relative to N2. The effects of both plasma gases are very similar. It is worth noting that the amino groups introduced by the NH3 plasma treatment made possible the immobilization of various compounds on the membrane surfaces (Lévesque et al. 2002; Lopez et al. 2006; Liu et al. 2006; Buonomenna et al. 2007). Also, the action of the nitrogen plasma on the PP membranes was investigated in detail to apply it to bacteria sterilization (Ricard and Canal 2010).

7.5.1

USE OF HYDROGEN AND METHANE PLASMA

Flat porous poly(vinylidene luoride) hydrophobic membranes were used for the removal and the recovery of CO2 from the emission sources (Lin et al. 2009b). Methane plasma treatment enhanced the hydrophobicity of this polymer—the elemental F/C ratio at the surface increased and was almost twice as big as the starting value. The water contact angle increased from 132° for virgin polyvinylidene luoride (PVDF) to 155° after plasma treatment. Some effects of etching were observed, especially for longer plasma treatment time. The H2 plasma was used to prepare poly(ether ether ketone)–polyurethane membranes (Salerno et al. 2009) before ammonia plasma treatment. Such pretreatment cross-links and stabilizes the surface of the membrane before grafting of the N-groups, so that the hydrophobic recovery of the inal surface is signiicantly reduced.

7.5.2

USE OF CHLOROCOMPOUND AND FLUOROCOMPOUND PLASMAS

The plasmas of these compounds were used to improve the properties of the gas separation membranes. Poly(trimethylsilylpropyne) membranes were treated with CF4 (Lin et al. 1993) or CCl4 (Lin et al. 1995) plasma. The structural changes caused by the introduction of the chlorine or luorine atoms and by the surface cross-linking enhanced the O2/N2 selectivity. The ratio of oxygen permeability to nitrogen permeability increased from 1.40 before modiication up to 6.80 after 7 min of CHCl3 plasma treatment (Lin et al. 1995). A similar effect was achieved by CHCl3 plasma treatment of the polybutadiene/polycarbonate composite membrane (Chen et al. 1997). The membrane became less porous after plasma treatment and the gas permeation lux decreased in the irst few minutes of the process. At the same time, the selectivity of

192

Membrane Modification: Technology and Applications

O2/N2 increased from 3 to about 7. The separation eficiency was related mainly to the physical structure of the membrane surface. The contact angle of the water increased with the storage time and the membrane selectivity also became worse (Chen et al. 1997). Polypropylene hollow ibers modiied with CF4 plasma were used as gas absorption membranes (Lin et al. 2009b). The water contact angle increased during treatment— from 117° to 143°. XPS analysis showed more than 50% of the luorine on the membrane surface (F/C from 0 to 1.3 after 5 min of plasma treatment). The increase in the water contact angle resulted from both the increase in the F/C ratio and the increase in the surface roughness following the plasma treatment. The CO2 absorption lux was enhanced and the membrane durability was greatly improved by the treatment of the membrane with short-term exposure to CF4 plasma (Lin et al. 2009a). Another luorine-containing reagent—SF6—was used to obtain material similar to Naion that could be used as a proton-exchange membrane in fuel cells (Navarro et al. 2008). The CF4 plasma treatment of the CTA and the CTA blend (Olde Riekerink et al. 2002) membranes led to hydrophobic surfaces (a water contact angle of about 100°) due to extensive luorination of the top layer. The highest F/C ratio is observed after 2 min of plasma action. The etching effect is much less pronounced than in the case of the CO2 plasma, and the CF4 plasma treatment also has much less impact on the permeation of the CTA membranes. Prolonged treatment was necessary to obtain a relatively small lux increase. CF4 plasma treatment caused an increase in the membrane’s molecular weight cutoff.

7.6

PLASMA-INDUCED GRAFT COPOLYMERIZATION

Plasma-induced polymerization is applied to graft the desired functionalities onto a membrane surface. Plasma-induced graft polymerization is an attractive way of modifying the surface chemistry and morphology of the polymeric membranes. A desired monomer may be polymerized onto the surface of a membrane to form a grafted brush layer. The advantage of the grafting techniques is that the graft chains are chemically bonded to the membrane matrix and they cannot be detached from the membrane by a permeating solvent. Hence, the generated effect, such as hydrophilicity or the presence of speciic groups, is stable and does not deteriorate with time. This is very important in the case of the membrane used in iltration processes, as a hydrophobic interaction is often the major cause of fouling of the membranes. The other factor is electrostatic repulsion and this is the reason for the use of basic or acidic monomers for grafting. Introducing polymer chains by plasma onto the membrane surface can be conducted in several ways: “grafting to,” “grafting from,” and adsorption of the desired polymer on the polymer surface following its treatment with plasma. All these methods have found great application in membrane science.

7.6.1

GRAFTING-TO TECHNIQUE

In the grafting-to technique, the functional groups of the polymer to be grafted react with the functional groups generated by the plasma at the membrane surface (Figure 7.1).

193

Plasma Modification of Polymer Membranes A AC A A B Membrane

C

Plasma Treatment

B

X

A,B,C X

Functionalities generated by plasma Functional groups able to react with A

FIGURE 7.1 An illustration of the “grafting-to” technique.

This method was applied by Lévesque et al. (2002) to bind methoxy poly(ethylene glycol) (MPEG) as its succinimidyl-activated ester with the amine surface group of the microporous polytetraluoroethylene (PTFE) membranes. The amine groups were created by ammonia plasma. The adsorption of the lipid on the modiied membrane was signiicantly reduced by the presence of an MPEG layer, regardless of the molecular weight of the latter. When a macromonomer is used, the grafting reaction goes through the radicals introduced on the surface by the plasma. Such a technique was used in the work of Chang et al. (2008)—poly(ethylene methacrylate) was grafted onto the expanded PTFE membrane treated with hydrogen plasma. Modiied membranes provide good biofouling resistance, thereby reducing the plasma protein and blood platelet adsorption.

7.6.2

GRAFTING-FROM TECHNIQUE

The grafting-from process, in which the polymer chains are built from the surfaceinitiating centers, goes through two successive processes—the surface activation of the substrate by plasma and the graft polymerization of a monomer. The latter stage can be realized in two ways. The radicals on the polymer surface may immediately make contact with the monomer, directly initiating the growth of the grafted polymer chain (Figure 7.2a). This usually takes place in the plasma reactor, where a monomer is put in as a vapor; however, sometimes a monomer solution is used as well (Xie et al. 2005; Yang et al. 2010). The molecular weights of the grafted chains should be almost equal as they start to grow in this same moment (no kinetics of peroxides dissociation) and grow in the same conditions. Chain termination takes place after exposing the sample to the air. Then, the surface-grafting density could be regulated by the plasma parameters, while the length of the grafted chains depends on the grafting parameters, mainly time. If the plasma-treated membranes are allowed to react with air irst, the mechanism of grafting is different (Figure 7.2b). Radicals react with the components of air, resulting in various functional groups; among them, peroxides and hydroperoxides are formed. These, under the inluence of temperature, redox reactants, or UV radiation, decompose, being the source of the radicals initiating the graft polymerization of the chosen monomer. Sometimes, to create peroxides, ozone is used instead of air (Tu et al. 2006).

194

Membrane Modification: Technology and Applications

Membrane

Monomer Vapor

Gas Plasma

Surface radicals

Surface-grafted membrane

(a)

Membrane

HO O O O

Gas Plasma

COOH

Air Surface radicals

Surface peroxides

T or UV

Monomer Solution Surface-grafted membrane

Surface radicals

(b)

FIGURE 7.2 An illustration of the “grafting-from” technique: (a) created radicals immediately react with the monomer and (b) radicals react with air and the created peroxides initiate polymerization of the monomer.

Lately, surface-initiated atom transfer radical polymerization (ATRP) has been used to obtain a surface-grafted membrane (Liu et al. 2010). A porous PTFE membrane was treated by the hydrogen plasma and the C–F groups of the modiied surface became effective initiators of ATRP. PEG methacrylate or its copolymer with N-isopropylacrylamide was grafted in such a way and the modiied membranes showed temperature-responsive and protein repulsion features (Liu et al. 2010).

7.6.3

GRAFTING IN MONOMER VAPOR

A microporous PP membrane was activated in Ar plasma and grafted with N,Ndimethyl-acrylamide (DMAA) and 2-methoxyethylacrylate (MEA) to reduce the hemolysis of blood (Onishi et al. 1995). The modiied membrane from DMAA became extremely hydrophilic, while the latter was only weakly hydrophilic. A blood–plasma separator equipped with a PP-g-PMEA membrane had good hemocompatibility and an excellent separation capacity. The eficient modiication of the porous PES membranes was achieved by Ar plasma treatment followed by graft copolymerization with acrylamide in the vapor phase (Wavhal and Fisher 2003). No surface damage and only a slight alteration in the pore structure were observed. As a result of such a modiication, the membrane surface was less susceptible to adsorption of BSA. The grafted membranes also gave greater lux recoveries after cleaning, indicating that the protein layer was reversibly attached to the surface. Allylamine and acrylic acid (AAc) were grafted by oxygen plasma activation of the surface of the PET track membranes (Touik et al. 2002). Such treatment increased the water low without affecting the rejection percentage of different molecular weight dextrans.

Plasma Modification of Polymer Membranes

195

AAc is without doubt the most frequently used monomer in plasma-initiated grafting. It is cheap and easily polymerizable and introduces many carboxyl groups onto its surface, which are pH-sensitive and are capable of further chemical reactions. The AAc grafting yield of the PSU UF membrane reached 261 μg/cm2 (Gancarz et al. 1999b). Modiied membranes were characterized with a high surface tension, mainly their polar component (an increase from 0.9 to ~30 mN/m). A signiicant water lux decrease is observed only when the degree of grafting approaches a value of 50 μg/cm2. All parameters of the BSA iltration are better than those of the untreated membrane. Very similar results were obtained for AAc vapor grafted onto commercial polysulphone (Zhan et al. 2004) or PES (Wavhal and Fisher 2002a). This is mainly due to an enhanced repulsive electrostatic force that hinders the adsorption or deposition of protein molecules onto the membrane surface. During the grafting process, two changes on the membrane surface contribute to the permeability—the pore size (usually a decrease is observed) and hydrophilicity. The former decreases the water permeability of the membrane; the latter can increase it. When the effect of hydrophilization is more signiicant than the decrease in the pore size, an increase in the lux can be seen. Prolonged plasma treatment caused etching and resulted in an increase in the pore size. Such effects were observed for the poly(acrylonitrile) UF membrane grafted with AAc (Zhao et al. 2004). The pores in the modiied membranes became smaller and their distribution narrowed. However, the lux of a saccharose solution increased. The best modiied membranes showed a saccharose retention of 76%. The same membrane was the subject of grafting of styrene in the vapor phase (Zhao et al. 2005). In this case, the hydrophobicity of the membrane signiicantly increased, which made it possible to recover the dewaxed oil (M ~ 450 g/mol, 72.8% of rejection) from dewaxed lube oil iltrates.

7.6.4

GRAFTING IN MONOMER SOLUTION

The literature contains many examples of using this method. Plasma-induced grafting was used to control the surface pore size as in the case of PVDF grafted with styrene (Chen et al. 2009). The pores in the modiied membrane get smaller and the distribution of the pores narrows with the increase in grafting time. When the degree of grafting is high, the membrane pores could be totally illed with the polymer. This process is called “plasma-graft illing polymerization.” The morphology of the modiied samples closely depended on the plasma within the membrane pores, which in turn was determined by the plasma power and the gas pressure. The study of the PE and PP membranes plasma-grafted with N-isopropylacrylamide revealed that plasma exists even in the submicropores (~0.4 μm in diameter) (Choi et al. 2003). Special attention is given to these membranes when they are used in the pervaporation process. The composite membrane appears to have the expected properties—it can transport one component and retain another component. A porous high-density PE membrane grafted with poly(methyl acrylate) (pMA) (Yamaguchi et al. 1991, 1997; Wang et al. 1998; Kai et al. 2000a), poly(ethyl acrylate) or poly(butylacrylate) (Yamaguchi et al. 1994), and poly(glycidyl methacrylate) (pGMA) (Wang et al. 1998) was applied with success as the membrane for separating benzene/cyclohexane (Yamaguchi et al. 1991; Wang et al. 1998; Kai et al. 2000a), benzene/chloroform

196

Membrane Modification: Technology and Applications

(Yamaguchi et al. 1997), and chloroform/1,1,2-trichloroethane mixtures (Yamaguchi et al. 1994). For pGMA-grafted membranes, the permeation lux of 0.3 kg/m 2 h and the separation factor of 22 were obtained for a 6:4 mixture of benzene–cyclohexane at 70°C. For the pMA-grafted PE, the lux was higher but the separation factor was much lower (3–6) (Wang et al. 1998). Yamaguchi et al. (1997) developed a model to express the behavior of a illing-type membrane. Methyl acrylate was also used as a grafting monomer for porous glass, giving a membrane that exhibited not only high selectivity in the pervaporation separation of a chloroform-n-hexane mixture, but also high thermal stability (Kai et al. 2000b). The membrane prepared by cross-linking a pore-illing graft polymerization of methyl acrylate onto high density polyethylene (HDPE) was used to compare the pervaporation and RO of a chloroform-n-hexane mixture (Kai et al. 2005). Both the permeation rate and the separation factor in RO are lower than those in PV, but considering the effect of the osmotic pressure present in RO, the membrane showed a reasonable separation performance in RO. Porous poly(vinylidene luoride) ilms were grafted with styrene followed by sulfonation of the grafted membranes (Ihm and Ihm 1995). It was found that the water– ethanol permeation rate was highest for the sodium salt form of the sulfonated graft membrane and was equal to 6.6 kg/m2 h with a separation factor of 21 for a 60 wt% ethanol solution at 50°C. For the same purpose, PSU (Lai et al. 1996) and aromatic polyamide (Teng et al. 2000) membranes were grafted with acrylamide. Both types of membrane modiication improved the permeation properties of the membranes. For the former membrane, the ionized form possessed a separation factor of 10.4 and a permeation rate of 718.1 g/m2 h. For the polyamide membrane with a degree of grafting of 20.5% for a 90 wt% ethanol feed concentration, these values were equal to 200 and 325 g/m2 h, respectively. PTFE membranes were plasma grafted with acrylamide and sodium 4-styrenesulfonate (NaSS), and showed a superior performance to pervaporation dehydration for various aqueous solutions of organic compounds (Tu et al. 2006). In particular, the protonized PTFE-g-NaSS membrane exhibited excellent properties, giving a permeation lux of 422 g/m2 h and a separation factor of 4491 in the pervaporation of a 90 wt% aqueous solution of isopropanol at 65°C. UF membranes were plasma-grafted with hydrophilic monomers to increase their surface hydrophilicity and hence improve the iltration of the protein. Ulbricht and Belfort (1995) used He or He/water plasma to graft AAc, methacrylic acid, and 2-hydroxyl-ethyl methacrylate (HEMA) onto polyacrylamide and PSU UF membranes. A PVDF nanoiber membrane graft copolymerized with methacrylic acid seemed to retain a high lux performance while its pore size was signiicantly smaller (Kaur et al. 2007). Commercial PES UF membranes were treated with He plasma followed by grafting of the N-vinyl-2-pyrrolidone (Chen and Belfort 1999) and PSU membranes modiied with AAc using Ar plasma-initiated graft polymerization (Gancarz et al. 1999b). A loss of hydraulic permeability due to the modiication may be compensated for by a higher iltrate permeability due to the reduced fouling (Ulbricht and Belfort 1996). Modiied membranes are notably less susceptible to protein fouling than untreated membranes (Chen and Belfort 1999). When the degree of grafting is high, the water

Plasma Modification of Polymer Membranes

197

lux becomes very low and the membranes lose their UF properties (Gancarz et al. 1999b). Ulbricht and Belfort (1995) claim that by applying this method, they are able to simultaneously adjust the hydrophilicity and the permeability/selectivity of the UF membranes. PTFE porous membranes grafted with methacrylic acid and 2-(dimethylamino)ethyl methacrylate were applied to the iltration of saccharides and dextrans (Yamada et al. 1996). The irst monomer caused an increase in the membrane permeability with the increase in the grafted chains. The second monomer, by contrast, made the permeability increase for shorter grafted chains. The permeability of the grafted membranes depends not only on the amount grafted, but also on the distribution of the grafted chains, especially when they are located in the pores. These modiied membranes belong to the group of “smart polymers”— stimuli-responsive materials. The porous luoropolymer grafted with 3-carbamoyl-1(p-vinylbenzyl) pyridinium chloride or its copolymer with acrylamide also belongs to the group of smart polymers (Ito et al. 1997a). A change in the water permeability occurred immediately with a change in the ionic strength, which was reversible. The thermoresponsive properties gained the polycarbonate track membranes plasma grafted with poly(N-isopropylamide) (NIPAM) (Xie et al. 2005). Thermoresponsive and molecular-recognizable membranes were obtained from the porous Nylon-6 membrane by a plasma graft pore-illing NIPAM-co-glycidyl methacrylate followed by further reactions with β-cyclodextrin (Yang et al. 2010). The permeation of ribolavin by a poly(vinylidene luoride) membrane grafted with AAc (Lee and Shim 1996) and the permeation of water through the polycarbonate grafted with methacrylic acid (Ito et al. 1997b) depend on the pH of the solution. At a low pH, the grafted chains are protonated and contracted, hence the pores are open and the permeate lux is high. With an increase in the pH value, the chain of polyacids swells and plugs the pore lumens. The permeation of water through the polypropylene membrane grafted with N-isopropylacrylamide depends on both the graft yield and the temperature (Kim et al. 2002b). Plasma grafting of poly{2-(N,N-dimethyl) aminoethyl methacrylate} on a microporous PE substrate resulted in the membrane having amine moieties (Matsuyama et al. 1996). This membrane showed high permselectivity for CO2 over N2 (130 for swollen membrane). Polypropylene porous membranes grafted with AAc with a high degree of grafting showed good electric properties (Choi and Moon 2007; Ciszewski et al. 2006, 2007; Gancarz et al. 2008, 2010). The very low electric resistance and the very good chemical and mechanical stabilities make them suitable for use as an effective separator for high-power alkaline batteries. Biomedicine is an area where plasma-grafted membranes seem to be of great interest. Porous polypropylene ilms grafted with AAc were applied for the immobilization of invertase (Tanioka et al. 1998). Silicone rubber membrane was plasma-grafted with AAc from one side and HEMA from the other, and this heterobifunctional membrane was used as the material for artiicial cornea (Chang et al. 1998). Grafting the PMMA intraocular lenses with a mixture of AAc and acrylamide made the surface hydrophilic and smooth, leading to an appreciable improvement in its properties (Suh et al. 2002). AAc grafted onto the surface of a nonporous poly(ɛcaprolactone) ilm remarkably improved the attachment and proliferation of human

198

Membrane Modification: Technology and Applications

dermal ibroblasts and myoblasts, thereby making this material ideal for membrane tissue engineering (Cheng and Teoh 2004). The sulfonic groups introduced on the surface of a chitosan membrane by vinyl sulfonic acid improved the adhesion and proliferation of the human osteosarcoma cell line (Lopez-Perez et al. 2007). Grafting the PVDF membrane with glycidyl methacrylate followed by a reaction with lysine or 1,6-hexanediamine gave a material with an enhanced neuronal cell culture (Young et al. 2010). A novel antibacterial material was developed by plasma-induced graft polymerization of 4-vinylpyridine followed by quaternization with hexylbromide, on electrospun polyurethane ibrous membranes (Yao et al. 2008) and a poly(vinylidene luoride-co-hexaluoropropylene) membrane (Yao et al. 2009). The killing eficiency of the latter was as high as 99.99995. The grafting of polymers onto the membrane surface was also used to obtain special kinds of membranes, for example, the chelating or afinity types. Thus, the nonporous polypropylene membrane was grafted with N-acryloyl glycine and this ilm showed high selectivity toward the chelation of silver (Poncin-Epaillard et al. 2000). A nonwoven ibrous PSU membrane grafted with poly(methacrylic acid) was used to couple a number of protein ligands (Ma et al. 2006). Filtration analysis showed that such membranes have a much smaller pressure drop and a higher lux compared with the conventional microiltration membranes and are potential candidates for use as afinity membranes.

7.6.5

GRAFTING THROUGH THE PLASMA ATTACHMENT OF ADSORBED POLYMER

The third method of plasma-induced grafting includes the adsorption of the desired polymer onto the polymer surface following its gas plasma treatment. In this case, the chains of the deposited polymer are attached to the membrane surface (Figure 7.3). This method could be described as “grafting to,” but its mechanism is different. The most probable is that the plasma generates radicals both on the membrane surface and on the adsorbed polymer to be grafted. A recombination of these two kinds of radicals makes covalent bonding between the membrane material and the adsorbed polymer possible, hence grafting is achieved. The deposition of the chosen polymer on a membrane surface is achieved most often by dipping the membrane in the polymer solution and subsequently drying of the solvent. Usually, the plasma of Ar, N2, or air is used for the treatment. Extraction with the polymer solvent after the treatment removes all of the free (ungrafted) polymer. PEG or its acrylate derivative was grafted by this method to PSU (Iwata et al. 1994; Song et al. 2000), poly(vinylidene luoride) (Wang et al. 2002), and PTFE membranes (Zhang et al. 2002). XPS and ATR-FTIR spectra showed without doubt

Membrane

Polymer

Plasma and Extraction

FIGURE 7.3 An illustration of the process of grafting a membrane through adsorption and the subsequent plasma action.

Plasma Modification of Polymer Membranes

199

that grafting occurred. Such membranes become signiicantly more hydrophilic. For PSU membranes, the water contact angle decreases from 74° to 37°, when a degree of grafting of about 3% is achieved. Modiied membranes are more resistant to oil staining (Iwata et al. 1994). In the case of the PVDF membranes, the water lux decreased with the increasing surface concentration of the grafted PEG polymer, while the pore size remained almost unchanged. Protein adsorption revealed that these membranes exhibited good antifouling properties (Wang et al. 2002). PEG is soluble in water and ethanol and is nontoxic and nonimmunogenic, making it one of the best polymers used to improve the biocompatibility of the polymer surfaces. PEG-modiied PSU shows higher hemocompatibility (Song et al. 2000) and 3% of the PEG-600-immobilized PTFE membrane appears to be a good candidate for blood ilters. The polypropylene hollow iber microiltration membrane was modiied in a similar way, taking α-allyl glucoside (AG) as an adsorbed material to be ixed onto the membrane surface (Kou et al. 2003; Deng et al. 2004, 2005). The authors claim that polymerization of AG took place during this process, thus PP grafted with poly(AG) was obtained as a result. The static contact angle of pure water on the grafted membrane decreased signiicantly from 120° to 36° with an increase in the degree of grafting of AG from 0 to 3.46 wt%, which indicated that the membrane surface was distinctly changed from hydrophobic to hydrophilic. The results of the BSA adsorption and iltration measurements indicated that the antifouling property of the membrane was signiicantly improved (Kou et al. 2003). The polypropylene microiltration membrane was also the subject of the postadsorption plasma treatment with poly(N-vinyl-2-pyrrolidone) as the grafting polymer (Liu et al. 2005; Yu et al. 2006). Membranes that gained signiicant hydrophilicity during this treatment were notably less susceptible to protein fouling than the untreated membranes. PPHFMMs grafted with poly(N-vinyl-2-pyrrolidone) worked well in an SMBR for wastewater treatment (Yu et al. 2006). A signiicant improvement of the hemocompatibility of the modiied PP was observed. Porous PE membranes on which a layer of acrylonitrile was deposited were treated with argon plasma. The modiied membranes used as a separator for a lithium-ion battery showed improved wettability, electrolyte retention, and interfacial adhesion between the electrodes and the separator, and hence the performance of the battery was also better (Kim et al. 2009). This method is not limited to polymer membranes. Applying it, γ-aminopropyl triethoxysilane was successfully bound to the surface of a ceramic membrane (Ida et al. 2000a,b). The aminopropyl groups that were introduced made possible the immobilization of a model enzyme—glucoamylase—on the membrane surface. The term “grafting” is often used not only for polymer chains, but also for some lowmolecular-weight molecules. In that case, covalent bonds are created through a reaction between the groups present on the membrane surface and those present in the grafted compounds. In this process, ammonia was the most often used gas plasma because it is known to introduce many amine groups on the polymer surface (Lopez et al. 2006; Buonomenna et al. 2007; Fontananova et al. 2006). Other gases, such as Ar, N2, and H2O, were also used. This method was used to graft not only polymers but also other compounds onto the membrane surface. The groups grafted by the

200

Membrane Modification: Technology and Applications

ammonia plasma onto the surface (principally NH2, together with OH, CN, NH, and CO) acted as the same number of anchor sites for the immobilization of the phosphotungstic acid hydrate (W12) (Fontananova et al. 2006) and other tungsten-based catalysts (Lopez et al. 2006; Buonomenna et al. 2007) on the poly(vinylidene luoride) membrane. After the immobilization of W12 (Fontananova et al. 2006), the membrane pore size increased, which together with the higher hydrophilicity increased the water lux through the modiied membranes. Sodium tungstate (W1) (Lopez et al. 2006; Buonomenna et al. 2007) and W10 (decatungstate) or W12-modiied catalytic membranes (Lopez et al. 2006) were used for the oxidation of the secondary amines to nitrones and for the degradation of phenol in water, respectively. The catalytic activity of such membranes was higher than has been observed in a homogeneous catalyst phase. The prepared catalytic membranes are stable and recyclable without loss of activity in successive catalytic runs. A similar method, using N2 or N2 + H2 plasma, was applied to attach diethylenetriamine pentaacetic acid (DTPA) (Poncin-Epaillard et al. 2000) to the polypropylene membrane to obtain a material with chelating properties. The DTPA-grafted membranes have a high selectivity toward the chelation of silver.

7.7

PLASMA POLYMERIZATION

In plasma polymerization, the transformation of low-molecular-weight monomers into high-molecular-weight polymers occurs with the assistance of energetic plasma species, such as electrons, ions, and radicals. In many cases, polymers formed by plasma polymerization have different chemical compositions as well as different chemical and physical properties from those formed by conventional polymerization, even if the same monomers are used in both processes. PPs do not comprise of repeating monomer units, but are, instead, complicated units containing cross-linked, fragmented, and rearranged units from the monomers. Almost any organic gas or vapor can be used as a “monomer” for plasma polymerization, resulting in very thin, pinhole-free ilms. The deposited PPs usually exhibit good adhesion to the substrate and are characterized by their thermal and chemical stabilities, which is a result of their highly cross-linked structure. The inal effect of the plasma polymerization depends on the monomer used, the reactor geometry, and the process parameters (pressure, power, monomer low rate, time). In a mild plasma condition, the partial retention of the monomer’s chemical structure and functionality is possible. Plasma polymerization processes are widely used to produce composite membranes of improved or new and interesting properties that could be applied in all membrane ields. One of the irst applications was the separation of gases. Porous polypropylene was treated with the plasma of organosilicon compounds—tetramethylsilane, hexamethylenedisiloxane, and octamethyl-cyclotetrasiloxane (Sakata et al. 1986). The best oxygen permeability was achieved for octamethyl-cyclotetrasiloxane, while the best O2/N2 ratio was for pp-tetramethylsilane. Inagaki investigated the plasma-polymerized ilms of luorocompounds (Inagaki et al. 1988) or their mixtures with silanes (Inagaki and Katsuoka 1987) as gas separation membranes. The plasma polymerization of luorocompounds gave ilms in which the luorine content increased with an increasing CF4 concentration in the plasma gas mixture.

Plasma Modification of Polymer Membranes

201

The oxygen and nitrogen permeation coeficient for the ilms decreased with the CF4 content, but the separation factor was improved. The best ratio that PO2/PN2 achieved was 7.15 (Inagaki et al. 1988). The plasma polymerization of silane/luorocarbon mixtures gave thin ilms containing both silicon and luorine moieties (Inagaki 1988). The gas separation ability of these plasma ilms was enhanced by the addition of CF4, although the permeability was diminished somewhat. Plasma-polymerized cyclic perluoroamines—perluoro(N-ethylpyrrolidine) and perluoro(N-ethyl morpholine)—also showed good oxygen permeation ratios (106 cm3/cm2 sec cmHg) and oxygen permselectivity to nitrogen (~3.7) (Hayakawa et al. 1996). Porous polypropylene membranes with a plasma-polymerized layer of luoroalkyl acrylates and methacrylates were also suitable for O2/N2 separation (Kita et al. 1989). The permeation rate and the separation factor were strongly inluenced by the kind of monomer and the plasma conditions. The highest separation factor that RO2/RN2 obtained was 3.8. Polybutadiene/polycarbonate membranes with a pp-ethylenediamine layer had an increased gas permeability (in comparison with the unmodiied one) due to surface etching. Their selectivity was closely connected with the chemical composition of the top layer. A high nitrogen content was required for high O2 selectivity (Ruaan et al. 1998). The presence of the amine groups on the membrane surface also enhanced the capacity for CO2/CH4 separation. The plasma-polymerized diisopropylamine on the surface of the composite membrane—porous polyimide (support)/ silicone (skin)— made the separation coeficient as high as 17 for a permeation rate of 4.5 × 10 –4 cm3/cm2 sec cmHg (Matsuyama et al. 1994). In the case of the iltration membranes, the deposition of a plasma-polymerized layer resulted in a smaller water lux (Raik et al. 1997; Kang et al. 2001) or even in micropore blockage (Kang et al. 2001); however, in optimal conditions, one can get suficient lux together with good retention of the iltered substance. The porous poly(vinyl luoride) membrane with a layer of pp-allylamine showed an excellent performance in the iltration of a sugarcane solution (Raik et al. 1997). The surface of the PVDF membrane was changed by the plasma polymerization of the AAc or nonaluorobutyl ethylene. These two membranes, despite their various wettabilities, show similar iltration properties for a dextran solution (Raik et al. 2000). A cellulose acetate UF membrane covered with pp-ethylenediamine showed altered physicochemical characteristics of the surface as well as narrowing of the pore size. Such membranes display a possibility of fractionation between monosaccharides and disaccharides (Gulec et al. 2010). Both the etching and deposition of thin diamond-like ilms were observed for PET track membranes treated with a mixture of N2 and cyclohexane plasma. The effective pore diameter increases and the membrane remains asymmetric. Such a modiication improves the hydrodynamic characteristics of the track membranes— an increase in the iltration rate constant was observed; however, this did not cause a drastic change in the value of the water permeability (Kravets et al. 2002). The plasma polymerization of the allylamine and AAc deposited a nonconsistent layer that decreased and partially blocked the pores of the PET track membranes (Touik et al. 2002). Polymerization of 1H,1H,2H-perluoro-1-octene on the track PET membranes was aimed at protecting one side of the membrane against etching by the alkali (Troimov et al. 2009) and in that way to change the pore shape

202

Membrane Modification: Technology and Applications

from cylindrical to conical. As a result, more eficient iltration membranes were obtained. PSU with a pp-(AAc) UF membrane shows a good iltration performance in a basic solution—lower values of the fouling index, a reduction in the lux in protein iltration, and a higher lux recovery after cleaning than for the unmodiied PSU (Gancarz et al. 1999b). A variation of the pH also had a serious effect on the microporous PP membrane modiied by the plasma polymerization of AAc and allylamine (Kang et al. 2001). An adequate plasma treatment enhanced both the water lux and the eficiency of the solute removal. The fouling during BSA iltration was greatly reduced (Kang et al. 2001). Sulfonated poly(phenylene oxide) membranes covered with a layer of plasma-polymerized allylamine retained their UF structure. The application of the 2,4-d herbicide in the micellar-enhanced UF appeared to be much better than PPO or sulfonated PPO with a water lux of 68.2 dm3/m2 h and a herbicide rejection of up to 92% (Poźniak et al. 2006). Any porous substrate (ilter paper, polyester textile) covered by the PP of hexamethyldisilazane or the double layer of hexamethyldisilazane and n-hexane becomes the membrane selective for the water and hydrophobic liquids (Bankovic et al. 2004). Petroleum ether vapor plasma polymerization was used for the surface modiication of the chitosan membranes to control their permeation rate of water-soluble drugs and metabolites (Wang et al. 2001). This treatment also signiicantly improved the mechanical properties of the membrane. The stability of the supported liquid membranes with a microporous PP membrane as a substrate was signiicantly enhanced after the deposition of a layer of PP of hexamethyldisiloxane and heptylamine (Yang et al. 2000). Plasma polymerization was also applied to obtain an RO membrane, taking as a support a microiltration PP membrane or a UF PSU membrane. All used monomers (AAc, acrylonitrile, allylamine, ethylenediamine, n-propylamine, and methyl methacrylate) in proper plasma conditions improved the RO performance and the membrane resistance of chlorine (Kim and Kim 2001, 2006). For many applications, for example, in batteries, fuel cells, sensors, and electrolyzers, ion-exchange membranes are necessary. The main problem in obtaining such materials in plasma polymerization is the proper selection of the monomer and adjusting such plasma parameters that keep the functional groups from signiicant alteration. Anion-exchange membranes were obtained by the plasma polymerization of γ-aminopropylethoxydimethylsilane (APEMS) (Sakata and Wada 1988) or 4-vinylpyridine (Matsuoka et al. 2008). The latter greatly improved the performance of alkaline fuel cells. Anion-exchange functionalities have often been introduced onto the surface of cation-exchange membranes to improve their performance. Thus, the Naion membrane was covered with a layer of pp-triethylamine or 4-vinylpyridine. As a consequence, the transport of the Fe2+ ions was signiicantly reduced, which increased the eficiency of the redox-low battery (Ogumi et al. 1990). The layer deposited from a mixture of ethylene and ammonia enhanced the selectivity for protons of the Naion (Vallois et al. 2003; Zeng et al. 2000) and sulfonate polyimide membranes (Vallois et al. 2003). The transport number of multivalent cations decreased by 74% and 54% factors, respectively (Vallois et al. 2003). Perluorosulfonic acid membranes with a C7F16/argon plasma-deposited layer show

Plasma Modification of Polymer Membranes

203

the suppressed methanol permeability, which offers advantages for direct methanol fuel cell operations at a high methanol concentration (Lue et al. 2007). For the production of cation-exchange membranes, the plasma polymerization of such compounds as AAc (Basarir et al. 2005; Gancarz et al. 1999b; Ciszewski et al. 2006), triluoromethanesulfonic acid (Danilich et al. 1994), and maleic anhydride (Won et al. 2003) was applied. The last named membranes, after anhydride hydrolysis, were used as proton-exchange membranes for fuel cells. The plasmapolymerized copolymers of triluoromethanesulfonic acid with chlorotriluoroethylene (Danilich et al. 1994) were ionically conductive materials for biomedical sensors, and the AAc plasma polymerized on polypropylene membranes gave good electrochemical properties (Basarir et al. 2005). However, these were worse than from the PP membrane plasma-grafted with the AAc (Ciszewski et al. 2006, 2007; Gancarz et al. 2008, 2010). A thiophene layer deposited onto the cellulose acetate membrane showed better selectivity in the pervaporation of an isopropanol–water mixture (Bhat and Wavhal 2000) and a polyurethane membrane with pp-AAc gave much higher selectivity for the separation of methanol from methyl-t-butyl ether (Weibel et al. 2007). Also, the composite membrane consisting of a porous PSU support and pp-hexamethyldisiloxane seems to be perfect for pervaporation, especially when the concentration of the organic component is low (Zuri et al. 1997). Membranes modiied by plasma polymerization have often been addressed for biomedical applications. The functionalities introduced in this treatment are potential anchor sites for the covalent immobilization of enzymes. The amine group present in plasma-polymerized amines—n-butylamine (Gancarz et al. 2003b; Biederman et al. 2001), allylamine (Gancarz et al. 2003a, 2006), and ethylene diamine (Biederman et al. 2001)—made possible with glutaraldehyde activation, successfully immobilized such enzymes as glucose oxidase (Biederman et al. 2001), glucose isomerase (Gancarz et al. 2003b), or invertase (Gancarz et al. 2006). The results obtained suggest that the concentration of the NH2 groups present at the surface is not the only decisive parameter for successful immobilization, and the correlation between the activity of the enzyme and the amount of surface amine groups created is not straightforward. Plasma polymerization of allyl alcohol on the UF PSU membrane gave a material with some amount of hydroxyl groups, which served as the anchoring sites for the immobilization of xylose isomerase (Gancarz et al. 2003a). PES (De Bartolo et al. 2005) and PES–polyurethane blend (De Bartolo et al. 2007) membranes were modiied by the surface plasma polymerization of AAc, and short peptide arginine– glycine–aspartic acid was immobilized on the PPs. The membranes obtained are suitable substrates for cell culture. They elicit speciic cellular responses and induce hepatocytes to enhance the synthesis of albumin and urea. Nafion is a membrane material of excellent physicochemical properties. Unfortunately, its surface properties promote protein adsorption, which eliminates it from many biomedical applications. To change it, the PP of tetraethylene glycol dimethyl ether (tetraglyme)/HEMA was deposited on the Naion surface (Valdes et al. 2008). Nonfouling surfaces were obtained with a controllable number of hydroxyl groups to which biomimetic molecules could be linked. The problem of

204

Membrane Modification: Technology and Applications

the low adhesion of the expanded PTFE membrane was solved by the deposition of pp-acetylene (in the presence of nitrogen) on the surface (Tu et al. 2005). After modiication, the pore size diameter and the water contact angle were decreased. The deposition of a pp-n-heptylamine layer onto a porous HFM of poly(ɛ-caprolactone) signiicantly enhanced the cell adhesion, making possible the application of this material for tissue engineering scaffolds (Hadijizadeh and Mohebbi-Kalhori 2010). Membranes composed of the PET track membrane and a layer of pp-pyrrole showed asymmetric conductivity. Such membranes can be used to create chemical and biochemical sensors (Kravets et al. 2010).

7.8

CONCLUSION

The above survey of the plasma modiication of polymer membranes shows the power of this method. In a very short time, using very small quantities of reactants, almost each individual researcher is able to tailor suitable membranes. What is more, the plasma-modiied membranes show a unique character. Their bulk properties stay unchanged, while their surfaces differ from one another. By applying this method, it is possible to create a whole range of new membranes having in stock only one type of porous substrate. The simple diagram given below shows some possibilities for obtaining new membranes.

Membranes with different structure of pores

Membranes with antifouling property

Membranes for immobilization of enzymes

Membranes with different surface chemistry

Porous membrane matrix

Nanofiltration membranes

Porous hydrophobic membranes

Membranes for pervaporation

Membranes for electrodialysis and donnan dialysis

REFERENCES Aerts, S., Vanhulsel, A., Buekenhoudt, A., Weyten, H., Kuypers, S., Che, H., Bryjak, M., Gevers, L.E.M., Vankelecom, I.F.J. and Jacobs, P.A. 2006. Plasma-treated PDMSmembranes in solvent resistant nanoiltration: Characterization and study of transport mechanism. J. Memb. Sci. 275: 212–219. Ali, S.T., Leabek, J., Nielsen, L.P., Mathiasen, C., Moller, P. and Kaer, S.K. 2010. Thin ilm thermocouples for in situ membrane electrode assembly temperature measurements in a polybenzimidazole-based high temperature proton exchange unit cell. J. Power Sources 195: 4835–4841.

Plasma Modification of Polymer Membranes

205

Bankovic, P., Demarquette, N.R. and daSilva, M.L.P. 2004. Preparation of selective membranes for water and hydrophobic liquids by plasma enhanced chemical vapor deposition on porous substrates. Mater. Sci. Eng. B 112: 165–170. Basarir, F., Choi, E.Y., Moon, S.H., Song, K.C. and Yoon, T.H. 2005. Electrochemical properties of PP membranes with plasma polymer coatings of acrylic acid. J. Memb. Sci. 260: 66–74. Beltsios, K., Charalambopoulou, G., Romanos, G. and Kanellopoulos, N. 1999. A Vycor® membrane with reduced size surface pores I. Preparation and characterization. J. Porous Mater. 6: 25–31. Bhat, N.V. and Wavhal, D.S. 2000. Characterization of plasma-polymerized thiophene onto cellulose acetate membrane and its application to pervaporation. Sep. Sci. Technol. 35: 227–242. Biederman, H., Boyaci, I.H., Bilkova, P., Slavinska, D., Mutlu, S., Zemek, J., Trchova, M., Klimovic, J. and Mutlu, M. 2001. Characterization of glow-discharge-treated cellulose acetate membrane surfaces for single-layer enzyme electrode studies. J. Appl. Polym. Sci. 81: 1341–1352. Bryjak, M. and Gancarz, I. 1994. Plasma treatment of polyethylene ultrailtration membranes. Angew. Makromol. Chem. 219: 117–124. Bryjak, M., Gancarz, I. and Pozniak, G. 1999. Surface evaluation of plasma-modiied polysulfone (Udel P-1700) ilms. Langmuir 15: 6400–6404. Bryjak, M., Gancarz, I., Pozniak, G. and Tylus, W. 2002. Modiication of polysulfone membranes 4. Ammonia plasma treatment. Eur. Polym. J. 38: 717–726. Buonomenna, M.G., Lopez, L.C., Barbieri, G., Favia, P., d’Agostino, R. and Drioli, E. 2007. Sodium tungstate immobilized on plasma-treated PVDF membranes: New eficient heterogeneous catalyst for oxidation of secondary amines to nitrones. J. Mol. Catal. A Chem. 27: 32–38. Buonomenna, M.G., Lopez, L.C., Davoli, M., Favia, P., d’Agostino, R. and Drioli, E. 2009. Polymeric membranes modiied via plasma for nanoiltration of aqueous solution containing organic compounds. Micropor. Mesopor. Mater. 120: 147–153. Chan, C.M., Ko, T.M. and Hiraoka, H. 1996. Polymer surface modiication by plasmas and photons. Surf. Sci. Rep. 24: 1–54. Chang, P.C.-T., Lee, S.-D. and Hsiue, G.-H. 1998. Heterobifunctional membranes by plasma induced graft polymerization as an artiicial organ for penetration keratoprosthesis. J. Biomed. Mater. Res. 39: 380–389. Chang, Y., Cheng, T.-Y., Shih, Y.-J., Lee, K.-R. and Lai, J.-Y. 2008. Biofouling-resistance expected poly(tetraluoroethylene) membrane with a hydrogel-like layer of surfaceimmobilized poly(ethylene glycol) methacrylate for human plasma protein repulsions. J. Memb. Sci. 323: 77–84. Chen, H. and Belfort, G. 1999. Surface modiication of poly(ether sulfone) ultrailtration membranes by low-temperature plasma-induced graft polymerization. J. Appl. Polym. Sci. 72: 1699–1711. Chen, S.-H., Chuang, W.-H., Wang, A.A., Ruaan, R.-C. and Lai, J.-Y. 1997. Oxygen/nitrogen separation by plasma chlorinated polybutadiene/polycarbonate composite membrane. J. Memb. Sci. 124: 273–281. Chen, J., Li, J. and Chen, C. 2009. Surface modiication of polyvinylidene luoride (PVDF) membranes by low-temperature plasma with grafting styrene. Plasma Sci. Technol. 11: 42–47. Cheng, Z. and Teoh, S.-H. 2004. Surface modiication of ultra thin poly (ɛ-caprolactone) ilms using acrylic acid and collagen. Biomaterials 25: 1991–2001. Choi, E.-Y. and Moon, S.-H. 2007. Characterization of acrylic acid-grafted PP membranes prepared by plasma-induced graft polymerization. J. Appl. Polym. Sci. 105: 2314–2320.

206

Membrane Modification: Technology and Applications

Choi, W.C., Kim, J.D. and Woo, S.I. 2001. Modiication of proton conducting membrane for reducing methanol crossover in a direct-methanol fuel cell. J. Power Sources 96: 411–414. Choi, Y.-J., Moon, S.-H., Yamaguchi, T. and Nakao, S.-I. 2003. New morphological control for thick, porous membranes with a plasma graft-illing polymerization. J. Polym. Sci. A Polym. Chem. 41: 1216–1224. Ciszewski, A., Gancarz, I., Kunicki, J. and Bryjak, M. 2006. Plasma-modiied polypropylene membranes as separators in high-power alkaline batteries. Surf. Coat. Technol. 201: 3676–3684. Ciszewski, A., Kunicki, J. and Gancarz, I. 2007. Usefulness of microporous hydrophobic polypropylene membranes after plasma-induced graft polymerization of acrylic acid for high-power nickel-cadmium batteries. Electrochim. Acta 52: 5207–5212. Danilich, M.J., Gervasio, D.F. and Marchant, R.E. 1994. Radiofrequency plasma polymerization of perluoroionomer membrane materials. J. Appl. Polym. Sci. Symp. 54: 93–105. De Bartolo, L., Morelli, S., Lopez, L.C. et al. 2005. Biotransformation and liver-speciic functions of human hepatocytes in culture on RGD-immobilized plasma-processed membranes. Biomaterials 26: 4432–4441. De Bartolo, L., Morelli, S., Piscioneri, A., Lopez, L.C., Favia, P., d’Agostino, R. and Drioli, E. 2007. Novel membranes and surface modiication able to activate speciic cellular responses. Biomol. Eng. 24: 23–26. Deng, H.-T., Xu, Z.-K., Wu, J., Ye, P., Liu, Z.-M. and Seta, P. 2004. A comparative study on lipase immobilized polypropylene microiltration membranes modiied by sugar-containing polymer and polypeptide. J. Mol. Catal. B Enzym. 28: 95–100. Deng, H.-T., Xu, Z.-K., Dai, Z.-W., Wu, J. and Seta, P. 2005. Immobilization of Candida rugosa lipase on polypropylene microiltration membrane modiied by glycopolymer: Hydrolysis of olive oil in biphasic bioreactor. Enzyme Microb. Technol. 36: 996–1002. Dmitriev, S.N., Kravets, L.I., Simakina, N.V. and Sleptsov, V.V. 1995. Plasma modiication of polyethyleneterephthalate track membranes. Radiat. Meas. 25: 723–726. Dmitriev, S.N., Kravets, L.I. and Sleptsov, V.V. 1998. Modiication of track membrane structure by plasma etching. Nucl. Instrum. Methods Phys. Res. B 142: 43–49. Dmitriev, S.N., Kravets, L.I., Sleptsov, V.V. and Elinson, V.M. 2002. Water permeability of poly(ethylene terephthalate) track membranes modiied in plasma. Desalination 146: 279–286. Fontananova, E., Donato, L., Drioli, E., Lopez, L.C., Favia, P. and d’Agostino, R. 2006. Heterogenization of polyoxometalates on the surface of plasma-modiied polymeric membranes. Chem. Mater. 18: 1561–1568. Gancarz, I., Pozniak, G. and Bryjak, M. 1999a. Modiication of polysulfone membranes 1. CO2 plasma treatment. Eur. Polym. J. 35: 1419–1428. Gancarz, I., Pozniak, G., Bryjak, M. and Frankiewicz, A. 1999b. Modiication of polysulfone membranes 2. Plasma grafting and plasma polymerization of acrylic acid. Acta Polym. 50: 317–326. Gancarz, I., Pozniak, G. and Bryjak, M. 2000. Modiication of polysulfone membranes 3. Effect of nitrogen plasma. Eur. Polym. J. 36: 1563–1569. Gancarz, I., Bryjak, J., Bryjak, M., Pozniak, G. and Tylus, W. 2003a. Plasma modiied polymers as a support for enzyme immobilization. 1. Allyl alcohol plasma. Eur. Polym. J. 39: 1615–1622. Gancarz, I., Bryjak, J., Pozniak, G. and Tylus, W. 2003b. Plasma modiied polymers as a support for enzyme immobilization. II. Amines plasma. Eur. Polym. J. 39: 2217–2224. Gancarz, I., Bryjak, J., Bryjak, M., Pozniak, G. and Tylus, W. 2006. Poly(phenylene oxide) ilms modiied with allylamine plasma as a support for invertase immobilization. Eur. Polym. J. 42: 2430–2440.

Plasma Modification of Polymer Membranes

207

Gancarz, I., Kunicki, J. and Ciszewski, A. 2008. Application of plasma-induced grafting for modiication of alkaline battery separator. Chem. Listy 102: 467–472. Gancarz, I., Bryjak, M., Kunicki, J. and Ciszewski, A. 2010. Microwave plasma-initiated grafting of acrylic acid on Celgard 2500 membrane to prepare alkaline battery separators – Characteristics of process and product. J. Appl. Polym. Sci. 116: 868–875. Greene, G. and Tannenbaum, R. 2004a. Adsorption of polyelectrolyte multilayers on plasmamodiied porous polyethylene. Appl. Surf. Sci. 238: 101–107. Greene, G. and Tannenbaum, R. 2004b. Chemical and thermal stability of surface-modiied porous polyethylene membranes. J. Disper. Sci. Technol. 25: 609–617. Greene, G., Yao, G. and Tannenbaum, R. 2004. Deposition and wetting characteristics of polyelectrolyte multilayers on plasma-modiied porous polyethylene. Langmuir 20: 2739–2745. Greene, G., Radhakrishna, H. and Tannenbaum, R. 2005. Protein binding properties of surfacemodiied porous polyethylene membranes. Biomaterials 26: 5972–5982. Gulec, H.A., Topacli, A., Topacli, C., Albayrak, N. and Mutlu, M. 2010. Modiication of cellulose acetate membrane via low-pressure plasma polymerization for sugar separation applications; Part 1. Membrane development and characterization. J. Memb. Sci. 350: 310–321. Hadijizadeh, A. and Mohebbi-Kalhori, D. 2010. Porous hollow membrane sheet for tissue engineering applications. J. Biomed. Mater. Res. A 93A: 1140–1150. Hatano, Y., Nakamura, H., Furuya, H. and Sugisaki, M. 1998. Inluence of surface impurities on plasma-driven permeation of deuterium through nickel. J. Vac. Sci. Technol. A 16: 2078–2083. Hayakawa, Y., Terasawa, N., Hayashi, E. and Abe, T. 1996. Plasma polymerization of cyclic perluoroamines and composite membranes for gas separation. J. Appl. Polym. Sci. 62: 951–954. Houston, K.S., Weinkauf, D.H. and Steward, F.F. 2002. Gas transport characteristic of plasma treated poly(dimethylsiloxane) and polyphosphazene membrane materials. J. Memb. Sci. 205: 103–112. Hu, C.C., Tu, C.Y., Wang, Y.C., Li, C.L., Lee, K.R. and Lai, J.Y. 2004. Effects of plasma treatment on CO2 plasticization of poly(methyl methacrylate) gas-separation membranes. J. Appl. Polym. Sci. 93: 395–401. Huang, Y.-C., Huang, C.-C., Huang, Y.-Y. and Chen, K.-S. 2007. Surface modiication and characterization of chitosan or PLGA membrane with laminin by chemical and oxygen plasma treatment for neural regeneration. J. Biomed. Mater. Res. A 82A: 842–851. Ida, J., Matsuyama, T. and Yamamoto, H. 2000a. Immobilization of glucoamylase on ceramic membrane surfaces modiied with a new method of treatment utilizing SPCP-CVD. Biochem. Eng. J. 5: 179–184. Ida, J., Matsuyama, T. and Yamamoto, H. 2000b. Surface modiication of a ceramic membrane by the SPCP-CVD method suitable for enzyme immobilization. J. Electrostat. 49: 71–82. Ihm, C.-D. and Ihm, S.-K. 1995. Pervaporation of water–ethanol mixtures through sulfonated polystyrene membranes prepared by plasma graft-polymerization. J. Memb. Sci. 98: 89–96. Inagaki, N. 1988. Gas separation membranes plasma-polymerized from mixtures of silanes and luorocarbons. J. Appl. Polym. Sci. Symp. 42: 327–338. Inagaki, N. and Katsuoka, H. 1987. Gas separation membranes made by plasma polymerization of mixtures of silanes and luoromethane. J. Memb. Sci. 34: 297–305. Inagaki, N., Kobayashi, N. and Matsushima, M. 1988. Gas separation membranes made by plasma polymerization of perlorobenzene/CF4 and pentaluorobenzene/CF4 mixtures. J. Memb. Sci. 38: 85–95. Ito, Y., Nishi, S., Park, Y.S. and Imanishi, Y. 1997a. Oxidoreduction-sensitive control of water permeation through a polymer brushes-grafted porous membrane. Macromolecules 30: 5856–5859.

208

Membrane Modification: Technology and Applications

Ito, Y., Park, Y.S. and Imanishi, Y. 1997b. Imaging of pH-sensitive polymer brush on a porous membrane using atomic force microscopy in aqueous solution. Macromol. Rapid Commun. 18: 221–224. Iwa, T., Kumazawa, H. and Bae, S.Y. 2004. Gas permeabilities of NH3-plasma-treated polyethersulfone membranes. J. Appl. Polym. Sci. 94: 758–762. Iwata, H., Ivanchenko, M.I. and Miyaki, Y. 1994. Preparation of anti-oil stained membrane by grafting polyethylene glycol macromer onto polysulfone membrane. J. Appl. Polym. Sci. 54: 125–128. Kaar, J.L., Oh, H.-I., Russell, A.J. and Federspiel, W.J. 2007. Towards improved artiicial lungs through biocatalysis. Biomaterials 28: 3131–3139. Kai, T., Tsuru, T., Nakao, S. and Kimura, S. 2000a. Preparation of hollow-iber membranes by plasma-graft illing polymerization for organic-liquid separation. J. Memb. Sci. 170: 61–70. Kai, T., Yamaguchi, T. and Nakao, S. 2000b. Preparation of organic/inorganic composite membranes by plasma-graft illing polymerization technique for organic-liquid separation. Ind. Eng. Chem. Res. 39: 3284–3290. Kai, T., Goto, H., Yamaguchi, T. and Nakao, S. 2005. Development of crosslinked plasmagraft illing polymer membranes for the reverse osmosis of organic liquid mixtures. J. Memb. Sci. 265: 101–107. Kang, M.S., Chun, B. and Kim, S.S. 2001. Surface modiication of polypropylene membrane by low-temperature plasma treatment. J. Appl. Polym. Sci. 81: 1555–1566. Kaur, S., Ma, Z., Gopal, R., Singh, G., Ramakrishna, S. and Matsuura, T. 2007. Plasma-induced graft copolymerization of poly(methacrylic acid) on electrospun poly(vinylidene luoride) nanoiber membrane. Langmuir 23: 13085–13092. Kim, H.I. and Kim, S.S. 2001. Fabrication of reverse osmosis membrane via low temperature plasma polymerization. J. Memb. Sci. 190: 21–33. Kim, H.I. and Kim, S.S. 2006. Plasma treatment of polypropylene and polysulfone supports for thin ilm composite reverse osmosis membrane. J. Memb. Sci. 286: 193–201. Kim, K.S., Lee, K.H., Cho, K. and Park, C.E. 2002a. Surface modiication of polysulfone ultrailtration membrane by oxygen plasma treatment. J. Memb. Sci. 199: 135–145. Kim, S.O.Y., Kanamori, T. and Shinbo, T. 2002b. Preparation of thermal-responsive poly(propylene) membranes grafted with n-isopropylacrylamide by plasma-induced polymerization and their water permeation. J. Appl. Polym. Sci. 84: 1168–1177. Kim, J.Y., Lee, Y. and Lim, D.Y. 2009. Plasma-modiied polyethylene membrane as a separator for lithium-ion polymer battery. Electrochim. Acta 54: 3714–3719. Kir, E., Oksuz, L. and Helhel, S. 2006. Preparation of poly(2-chloroaniline) membrane and plasma surface modiication. Appl. Surf. Sci. 252: 3574–3579. Kita, H., Shigekuni, M., Kawafune, I., Tanaka, K. and Okamoto, K. 1989. Preparation of plasma-polymerized membranes from luoroalkyl acrylates and methacrylates and gas permeability through the membranes. Polym. Bull. 21: 371–376. Kou, R.-Q., Xu, Z.-K., Deng, H.-T., Liu, Z.-M., Seta, P. and Xu, Y. 2003. Surface modiication of microporous polypropylene membranes by plasma-induced graft polymerization of α-allyl glucoside. Langmuir 19: 6869–6875. Kramer, P.W., Yeh, Y.-S. and Yasuda, H. 1989. Low temperature plasma for the preparation of separation membranes. J. Memb. Sci. 46: 1–28. Kravets, L.I., Dmitriev, S.N., Sleptsov, V.V. and Elinson, V.M. 2002. Production of asymmetric track membranes with a high permeability and separation selectivity. Desalination 144: 27–34. Kravets, L.I., Dmitriev, S.N., Sleptsov, V.V. and Elinson, V.M. 2003. Production of asymmetric track membranes by gas-discharge method. Surf. Coat. Technol. 174–175: 821–825. Kravets, L., Dmitriev, S., Lizunov, N., Satulu, V., Mitu, B. and Dinescu, G. 2010. Properties of poly(ethylene terephthalate) track membranes with a polymer layer obtained by plasma polymerization of pyrrole vapors. Nucl. Instrum. Methods Phys. Res. B 268: 485–492.

Plasma Modification of Polymer Membranes

209

Kull, K.R., Steen, M.L. and Fisher, E.R. 2005. Surface modiication with nitrogen-containing plasmas to produce hydrophilic, low-fouling membranes. J. Memb. Sci. 246: 203–215. Kumazawa, H. and Yoshida, M. 2000. Mechanism of gas transport of NH3-plasma-treated poly(phenylene oxide) membranes. J. Appl. Polym. Sci. 78: 1845–1852. Kusumocahyo, S.P., Kanamori, T., Iwatsubo, T., Sumaru, K. and Shinbo, T. 2002. Development of polyion complex membranes based on cellulose acetate modiied by oxygen plasma treatment for pervaporation. J. Memb. Sci. 208: 223–231. Lai, J.-Y., Tseng, C.-W. and Lee, K.-R. 1996. Plasma graft polymerization membrane of acrylamide for pervaporation separation of aqueous alcohol mixtures. J. Appl. Polym. Sci. 61: 307–313. Lee, Y.M. and Shim, J.K. 1996. Plasma surface graft of acrylic acid onto a porous poly(vinylidene luoride) membrane and its ribolavin permeation. J. Appl. Polym. Sci. 61: 1245–1250. Lévesque, S., Thibault, J., Castonguay, M., Gaudreault, R.C. and Laroche, G. 2002. Modiication of lipid transport through a microporous PTFE membrane wall grafted with poly(ethylene glycol). Colloids Surf. B Biointerfaces 25: 205–217. Lin, X., Xiao, J., Yu, Y., Chen, J., Zheng, G. and Xu, J. 1993. Gas permeabilities of poly(trimethylsilylpropyne) membranes surface modiied with CF4 plasma. J. Appl. Polym. Sci. 48: 231–236. Lin, X., Qiu, X., Zheng, G. and Xu, J. 1995. Gas permeabilities of poly(trimethylsilylpropyne) membranes surface modiied with CCl4 plasma. J. Appl. Polym. Sci. 58: 2137–2139. Lin, S.-H., Tung, K.-L., Chen, W.-J. and Chang, H.-W. 2009a. Absorption of carbon dioxide by mixed piperazine–alkanolamine absorbent in a plasma-modiied polypropylene hollow iber contactor. J. Memb. Sci. 333: 30–37. Lin, S.-H., Tung, K.-L., Chang, H.-W. and Lee, K.-R. 2009b. Inluence of luorocarbon latmembrane hydrophobicity on carbon dioxide recovery. Chemosphere 75: 1410–1416. Liu, Z.-M., Xu, Z.-K., Wan, L.-S., Wu, J. and Ulbricht, M. 2005. Surface modiication of poly-propylene microiltration membranes by the immobilization of poly(N-vinyl-2pyrrolidone): A facile plasma approach. J. Memb. Sci. 249: 21–31. Liu, Z., Xu, Z.-K. and Ulbricht, M. 2006. Surface modiication of polypropylene microporous membrane by tethering polypeptides. Chin. J. Polym. Sci. 24: 529–538. Liu, Y.L., Han, C.C., Wei, T.C. and Chang, Y. 2010. Surface-initiated atom transfer radical polymerization from porous poly(tetraluoroethylene) membrane using the C–F groups as initiators. J. Polym. Sci. A Polym. Chem. 48: 2076–2083. Lopez, L.C., Buonomenna, M.G., Fontananova, E., Iacoviello, G., Drioli, E., d’Agostino, R. and Favia, P. 2006. A new generation of catalytic poly(vinylidene luoride) membranes: Coupling plasma treatment with chemical immobilization of tungsten-based catalysts. Adv. Funct. Mater. 16: 1417–1424. Lopez-Perez, P.M., Marques, A.P., da Silva, R.M.P., Pashkuleva, I. and Reis, R.L. 2007. Effect of chitosan membrane surface modiication via plasma induced polymerization on the adhesion of osteoblast-like cells. J. Mater. Chem. 17: 4064–4071. Lue, S.J., Hsiaw, S.-Y. and Wei, T.-C. 2007. Surface modiication of perluorosulfonic acid membranes with perluoroheptane (C7F16)/argon plasma. J. Memb. Sci. 305: 226–237. Ma, Z., Kotaki, M. and Ramakrishna, S. 2006. Surface modiied nonwoven polysulphone (PSU) iber mesh by electrospinning: A novel afinity membrane. J. Memb. Sci. 272: 179–187. Matsuoka, K., Chiba, S., Iriyama, Y., Abe, T., Matsuoka, M., Kikuchi, K. and Ogumi, Z. 2008. Preparation of anion-exchange membrane by plasma polymerization and its use in alkaline fuel cells. Thin Solid Films 516: 3309–3313. Matsuyama, H., Hirai, K. and Teramoto, M. 1994. Selective permeation of carbon dioxide through plasma polymerized membrane from diisopropylamine. J. Memb. Sci. 3: 257–265.

210

Membrane Modification: Technology and Applications

Matsuyama, H., Teramoto, M. and Hirai, K. 1995. Effect of plasma treatment on CO2 permeability and selectivity of poly(dimethylsiloxane) membrane. J. Memb. Sci. 99: 139–147. Matsuyama, H., Teramoto, M. and Sakakura, H. 1996. Selective permeation of CO2 through poly{2-(N,N-dimethyl) aminoethyl methacrylate} membrane prepared by plasma-graft polymerization technique. J. Memb. Sci. 114: 193–200. Muller, M. and Oehr, C. 1999. Plasma aminofunctionalization of PDVF microiltration membranes: Comparison of the plasma modiications with a grafting method using ESCA and an amino-selective lorescent probe. Surf. Coat. Technol. 116–119: 802–807. Nakata, M. and Kumazawa, H. 2006. Gas permeability and permselectivity of plasma-treated polyethylene membranes. J. Appl. Polym. Sci. 101: 383–387. Navarro, A., del Rio, C. and Acosta, J.L. 2008. Pore-illing electrolyte membranes based on microporous polyethylene matrices activated with plasma and sulfonated hydrogenated styrene butadiene block copolymer. Synthesis, microstructural and electrical characterization. J. Polym. Sci. B Polym. Phys. 46: 1684–1695. Ogumi, Z., Uchimoto, Y., Tsujikawa, M., Takehara, Z. and Foulkes, F.R. 1990. Modiication of ion exchange membrane surface by plasma process. I. H+ ion perm-selective membrane from Naion for redox-low battery. J. Electrochem. Soc. 137: 1430–1435. Olde Riekerink, M.B., Engbers, G.H.M., Wessling, M. and Feijen, J. 2002. Tailoring the properties of asymmetric cellulose acetate membranes by gas plasma etching. J. Colloid Interface Sci. 245: 338–348. Onishi, M., Shimura, K., Seita, Y. and Yamashita, S. 1995. Design of a new plasma separation membrane by graft copolymerization. Radiat. Phys. Chem. 46: 219–223. Ozdemir, Y., Hasirci, N. and Serbetci, K. 2002. Oxygen plasma modiication of polyurethane membranes. J. Mater. Sci. Mater. Med. 13: 1147–1152. Pal, S., Ghatak, S.K., De, S. and DasGupta, S. 2008a. Evaluation of surface roughness of a plasma treated polymeric membrane by wavelet analysis and quantiication of its enhanced performance. Appl. Surf. Sci. 255: 2504–2511. Pal, S., Ghatak, S.K., De, S. and DasGupta, S. 2008b. Characterization of CO2 plasma treated polymeric membranes and quantiication of lux enhancement. J. Memb. Sci. 323: 1–10. Poncin-Epaillard, F., Médard, N. and Souti, J.C. 2000. Reactivity of surface groups attached on a plasma treated poly(propylene) ilm. Application to a new concept of a chelating membrane. Macromol. Chem. Phys. 201: 212–219. Poźniak, G., Gancarz, I. and Tylus, W. 2006. Modiied poly(phenylene oxide) membranes in ultrailtration and micellar-enhanced ultrailtration of organic compounds. Desalination 198: 215–224. Raik, M., Mas, A., Elhari, A. and Schue, F. 1997. Depot par polymerisation par plasma d’allylamine sur membrane poreuse. Application a la decoloration de solutions sucrees. Eur. Polym. J. 33: 641–648. Raik, M., Mas, A., Elhari, A. and Schue, F. 2000. Modiication de membrane PVDF par plasma d’acide acrylique et de nonaluorobutylethylene. Eur. Polym. J. 36: 1911–1919. Ricard, A. and Canal, C. 2010. Transmission of N atoms through PP membranes for nitrogen lowing post discharge sterilization processes. Plasma Chem. Plasma Process. 30: 373–379. Ruaan, R.-C., Wu, T.-H., Chen, S.-H. and Lai, J.-Y. 1998. Oxygen/nitrogen separation by polybutadiene/polycarbonate composite membranes modiied by ethylenediamine plasma. J. Memb. Sci. 138: 213–220. Sakata, J. and Wada, M. 1988. Preparation of ion exchange membranes by plasma polymerization. I. J. Appl. Polym. Sci. 35: 875–884. Sakata, J., Yamamoto, M. and Hirai, M. 1986. Plasma polymerized membranes and gas permeability. J. Appl. Polym. Sci. 31: 1999–2006. Salerno, S., Piscioneri, A., Laera, S., Morelli, S., Favia, P., Bader, A., Drioli, E. and De Bartolo, L. 2009. Improved functions of human hepatocytes on NH3 plasma-grafted PEEK-WC-PU membranes. Biomaterials 30: 4348–4356.

Plasma Modification of Polymer Membranes

211

Song, Y.-Q., Sheng, J., Wei, M. and Yuan, X.-B. 2000. Surface modiication of polysulfone membranes by low-temperature plasma-graft poly(ethylene glycol) onto polysulfone membranes. J. Appl. Polym. Sci. 78: 979–985. Steen, M.L., Butoi, C.I. and Fisher, E.R. 2001a. Identiication of gas-phase reactive species and chemical mechanisms occurring at plasma–polymer surface interfaces. Langmuir 17: 8156–8166. Steen, M.L., Hymasa, L., Havey, E.D., Capps, N.E., Castner, D.G. and Fisher, E.R. 2001b. Low temperature plasma treatment of asymmetric polysulfone membranes for permanent hydrophilic surface modiication. J. Memb. Sci. 188: 97–114. Steen, M.L., Jordan, A.C. and Fisher, E.R. 2002. Hydrophilic modiication of polymeric membranes by low temperature H2O plasma treatment. J. Memb. Sci. 204: 341–357. Suh, T.S., Joo, C.-K., Kim, Y.C., Lee, M.S., Lee, H.K., Choe, B.Y. and Chun, H.J. 2002. Surface modiication of poly(methyl methacrylate) intraocular lenses with the mixture of acrylic acid and acrylamide via plasma-induced graft copolymerization. J. Appl. Polym. Sci. 85: 2361–2366. Tang, J.X., He, N.Y., Tan, M.J., He, Q.G. and Chen, H. 2004. A novel substrate for in situ synthesis of oligonucleotide: Plasma-treated polypropylene microporous membrane. Colloids Surf. A Physicochem. Eng. Asp. 242: 53–60. Tanioka, A., Yokoyama, Y. and Miyasaka, K. 1998. Preparation and properties of enzymeimmobilized porous polypropylene ilms. J. Colloid Interface Sci. 200: 185–187. Teng, M.-Y., Lee, K.-R., Liaw, D.-J., Lin, Y.-S. and Lai, J.-Y. 2000. Plasma deposition of acrylamide onto novel aromatic polyamide membrane for pervaporation. Eur. Polym. J. 36: 663–672. Teramae, T. and Kumazawa, H. 2007. Gas permeability and permselectivity of plasma-treated polypropylene membranes. J. Appl. Polym. Sci. 104: 3236–3239. Touik, M., Mas, A., Shkinev, V., Nechaev, A., Elhari, A. and Schue, F. 2002. Improvement of performances of PET track membranes by plasma treatment. Eur. Polym. J. 38: 203–209. Tran, T.D., Mori, S. and Suzuki, M. 2007. Plasma modiication of polyacrylonitrile ultrailtration membrane. Thin Solid Films 515: 4148–4152. Troimov, D.A., Shkinev, V.M., Spivakov, B.Ya. and Schué, F. 2009. Improvement of pore geometry and performances of poly(ethylene terephthalate) track membranes by a protective layer method using plasma-induced graft polymerization of 1H,1H,2H-perluoro1-octene monomer. J. Memb. Sci. 326: 265–269. Tu, C.-Y., Wang, Y.-C., Li, C.-L., Lee, K.-R., Huan, J. and Lai, J.-Y. 2005. Expanded poly(tetraluoroethylene) membrane surface modiication using acetylene/nitrogen plasma treatment. Eur. Polym. J. 41: 2343–2353. Tu, C.-Y., Liu, Y.-L., Lee, K.-R. and Lai, J.-Y. 2006. Hydrophilic surface-grafted poly(tetraluoroethylene) membranes using in pervaporation dehydration processes. J. Memb. Sci. 274: 47–55. Ulbricht, M. and Belfort, G. 1995. Surface modiication of ultrailtration membranes by low temperature plasma. I. Treatment of polyacrylonitrile. J. Appl. Polym. Sci. 56: 325–343. Ulbricht, M. and Belfort, G. 1996. Surface modiication of ultrailtration membranes by low temperature plasma. II. Graft polymerization onto polyacrylonitrile and polysulfone. J. Memb. Sci. 111: 193–215. Upadhyay, D.J. and Bhat, N.V. 2004. Pervaporation studies of gaseous plasma treated PVA membrane. J. Memb. Sci. 239: 255–263. Valdes, T.I., Ciridon, W., Ratner, B.D. and Bryers, D. 2008. Surface modiication of a perluorinated ionomer using a glow discharge deposition method to control protein adsorption. Biomaterials 29: 1356–1366. Vallois, C., Sistat, P., Roualdes, S. and Pourcelly, G. 2003. Separation of H+/Cu2+ cations by electrodialysis using modiied proton conducting membranes. J. Memb. Sci. 216: 13–25.

212

Membrane Modification: Technology and Applications

Van der Bruggen, B. 2009. Chemical modiication of polyethersulfone nanoiltration membranes: A review. J. Appl. Polym. Sci. 114: 630–642. Vidaurre, E.F.C., Achete, C.A., Simao, R.A. and Habert, A.C. 2001. Surface modiication of porous polymeric membranes by RF-plasma treatment. Nucl. Instrum. Methods Phys. Res. B 175–177: 732–736. Villeger, S., Sixou, M., Durand, J. and Ricard, A. 2006. Interaction of n atoms through a nylon membrane in nitrogen lowing post discharges. J. Phys. D. Appl. Phys. 39: 3826–3830. Vladkova, T.G., Dineff, P., Stojcheva, R. and Tomerova, B. 2003. Ion-plasma modiication of polyvinylchloride microiltration membranes. J. Appl. Polym. Sci. 90: 2433–2440. Wang, H., Lin, X., Tanaka, K., Kita, H. and Okamoto, K.-I. 1998. Preparation of plasmagrafted polymer membranes and their morphology and pervaporation properties toward benzene/cyclohexane mixtures. J. Polym. Sci. A Polym. Chem. 36: 2247–2259. Wang, H., Fang, Y.-E. and Yan, Y. 2001. Surface modiication of chitosan membranes by alkane vapor plasma. J. Mater. Chem. 11: 1374–1377. Wang, P., Tan, K.L., Kang, E.T. and Neoh, K.G. 2002. Plasma-induced immobilization of poly(ethylene glycol) onto poly(vinylidene luoride) microporous membrane. J. Memb. Sci. 195: 103–114. Wavhal, D.S. and Fisher, E.R. 2002a. Hydrophilic modiication of polyethersulfone membranes by low temperature plasma-induced graft polymerization. J. Memb. Sci. 209: 255–269. Wavhal, D.S. and Fisher, E.R. 2002b. Modiication of porous poly(ether sulfone) membranes by low-temperature CO2-plasma treatment. J. Polym. Sci. B Polym. Phys. 40: 2473–2488. Wavhal, D.S. and Fisher, E.R. 2003. Membrane surface modiication by plasma-induced polymerization of acrylamide for improved surface properties and reduced protein fouling. Langmuir 19: 79–85. Weibel, D.E., Vilani, C., Habert, A.C. and Achete, C.A. 2007. Surface modiication of polyurethane membranes using acrylic acid vapour plasma and its effects on the pervaporation processes. J. Memb. Sci. 293: 124–132. Won, J., Choi, S.W., Kang, Y.S., Ha, H.Y., Oh, I.-H., Kim, H.S., Kim, K.T. and Jo, W.H. 2003. Structural characterization and surface modiication of sulfonated polystyrene– (ethylene–butylene)–styrene triblock proton exchange membranes. J. Memb. Sci. 214: 245–257. Xie, R., Chu, L.-Y., Chen, W.-M., Xiao, W., Wang, H.-D. and Qu, J.-B. 2005. Characterization of microstructure of poly(N-isopropylamide)-grafted polycarbonate track-etched membranes prepared by plasma-graft pore-illing polymerization. J. Memb. Sci. 258: 157–166. Yamada, K., Ebihara, T., Gondo, T., Sakasegawa, K. and Hirata, M. 1996. Membrane properties of porous and expanded poly(tetraluoroethylene) ilms grafted with hydrophilic monomers and their permeation behavior. J. Appl. Polym. Sci. 61: 1899–1912. Yamaguchi, T., Nakao, S. and Kimura, S. 1991. Plasma-graft illing polymerization: Preparation of a new type of pervaporation membrane for organic liquid mixtures. Macromolecules 24: 5522–5527. Yamaguchi, T., Yamahara, S., Nakao, S.-I. and Kimura, S. 1994. Preparation of pervaporation membranes for removal of dissolved organics from water by plasma-graft illing polymerization. J. Memb. Sci. 95: 39–49. Yamaguchi, T., Nakao, S. and Kimura, S. 1997. Swelling behavior of the illing-type membrane. J. Polym. Sci. B Polym. Phys. 35: 469–477. Yamamoto, Y., Maegawa, M. and Kumazawa, H. 2003. Gas permeability of NH3-plasmatreated poly(methyl methacrylate) membranes. J. Appl. Polym. Sci. 87: 1068–1072. Yang, X.J., Fane, A.G., Bi, J. and Griesser, H.J. 2000. Stabilization of supported liquid membranes by plasma polymerization surface coating. J. Memb. Sci. 168: 29–37.

Plasma Modification of Polymer Membranes

213

Yang, M., Xie, R., Wang, J.Y., Ju, X.J., Yang, L.H. and Chu, L.Y. 2010. Gating characteristics of thermo-responsive and molecular-recognizable membranes based on poly(N-isopropyl-acrylamide) and beta-cyclodextrin. J. Memb. Sci. 355: 142–150. Yao, C., Li, X., Neoh, K.G., Shi, Z. and Kang, E.T. 2008. Surface modiication and antibacterial activity of electrospun polyurethane ibrous membranes with quaternary ammonium moieties. J. Memb. Sci. 320: 259–267. Yao, C., Li, X., Neoh, K.G., Shi, Z. and Kang, E.T. 2009. Antibacterial activities of surface modiied electrospun poly(vinylidene luoride-co-hexaluoropropylene) (PVDF-HFP) ibrous membranes. Appl. Surf. Sci. 255: 3854–3858. Young, T.H., Chang, H.H., Lin, D.J. and Cheng, L.P. 2010. Surface modiication of microporous PVDF membranes for neuron culture. J. Memb. Sci. 350: 32–41. Yu, H.-Y., Xua, Z.-K., Xie, Y.-J., Liu, Z.-M. and Wang, S.-Y. 2006. Flux enhancement for polypropylene microporous membrane in a SMBR by the immobilization of poly(Nvinyl-2-pyrrolidone) on the membrane surface. J. Memb. Sci. 279: 148–155. Yu, H.-Y., Liu, L.-Q., Tang, Z.-Q., Yan, M.-G., Gu, J.-S. and Wei, X.-W. 2008a. Surface modiication of polypropylene microporous membrane to improve its antifouling characteristics in an SMBR: Air plasma treatment. J. Memb. Sci. 311: 216–224. Yu, H.-Y., Liu, L.-Q., Tang, Z.-Q., Yan, M.-G., Gu, J.-S. and Wei, X.-W. 2008b. Mitigated membrane fouling in an SMBR by surface modiication. J. Memb. Sci. 310: 409–417. Yua, H.-Y., Xie, Y.-J., Hu, M.-X., Wang, J.-L., Wang, S.-Y. and Xu, Z.-K. 2005. Surface modiication of polypropylene microporous membrane to improve its antifouling property in MBR: CO2 plasma treatment. J. Memb. Sci. 254: 219–227. Zeng, R., Pang, Z. and Zhu, H. 2000. Modiication of a Naion® ion exchange membrane by a plasma polymerization process. J. Electroanal. Chem. 490: 102–106. Zhan, J., Liu, Z., Wang, B. and Ding, F. 2004. Modiication of a membrane surface charge by a low temperature plasma induced grafting reaction and its application to reduce membrane fouling. Sep. Sci. Technol. 39: 2977–2995. Zhang, Q., Wang, C., Babukutty, Y., Ohyama, T., Kogoma, M. and Kodama, M. 2002. Biocompatibility evaluation of ePTFE membrane modiied with PEG in atmospheric pressure glow discharge. J. Biomed. Mater. Res. 60: 502–509. Zhao, Z.-P., Li, J., Zhang, D.-X. and Chen, C.-X. 2004. Nanoiltration membrane prepared from polyacrylonitrile ultrailtration membrane by low-temperature plasma I. Graft of acrylic acid in gas. J. Memb. Sci. 232: 1–8. Zhao, Z.-P., Li, J., Chen, J. and Chen, C.-X. 2005. Nanoiltration membrane prepared from polyacrylonitrile ultrailtration membrane by low-temperature plasma 2. Grafting of styrene in vapor phase. J. Memb. Sci. 251: 239–245. Zhu, X., Chian, K.S., Chan-Park, M.B. and Lee, S.T. 2005. Effect of argon-plasma treatment on proliferation of human-skin-derived ibroblast on chitosan membrane in vitro. J. Biomed. Mater. Res. A 73A: 264–274. Zuri, L., Narkis, M. and Silverstein, M.S. 1997. Film formation and crack development in plasma polymerized hexamethyldisiloxane. Polym. Eng. Sci. 37: 1188–1194.

8

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes Carmen García-Payo and Mohamed Khayet

CONTENTS 8.1

Introduction .................................................................................................. 215 8.1.1 Electrospinning Process ................................................................... 217 8.1.2 Polymers Used in Electrospinning ................................................... 218 8.1.3 Electrospinning Parameters .............................................................. 219 8.2 Surface Modiication Techniques ................................................................. 220 8.2.1 Blending ............................................................................................ 221 8.2.2 Surface Layers ..................................................................................224 8.2.2.1 Coating ...............................................................................224 8.2.2.2 Layer-by-Layer Assembly .................................................. 227 8.2.3 Plasma Treatment ............................................................................. 231 8.2.4 Chemical Methods ............................................................................ 232 8.2.5 Surface Graft Copolymerization ...................................................... 235 8.2.5.1 Radiation-Induced Graft Copolymerization ...................... 235 8.2.5.2 Plasma-Induced Graft Copolymerization .......................... 237 8.2.5.3 Ce(IV)-Induced Graft Copolymerization .......................... 238 8.2.6 Molecule Immobilization on Nanoiber Surface .............................. 239 8.2.6.1 Physical Adsorption ........................................................... 239 8.2.6.2 Nanoparticles Assembly on the Surface of Nanoibers .....240 8.2.6.3 Covalent Bonding or Chemical Immobilization ................ 243 8.3 Conclusion ....................................................................................................246 References .............................................................................................................. 247

8.1

INTRODUCTION

Many researchers are interested in studying the unique properties of nanoscale materials, especially because of the emergence of nanoscience and nanotechnology. Fiber materials with diameters within the nanometer range when compared with microscale materials have several important characteristics, such as a very large 215

216

Membrane Modification: Technology and Applications

surface area to volume ratio, which for a nanoiber can be as large as 103 times that of a microiber, tunable porosity, pore size ranging from 10 nm to several micrometers, lexibility in surface functionalities, good mechanical performance, and malleability to construct a wide variety of iber sizes and shapes. These outstanding properties make nanoibers the optimal candidates for many important applications. Polymer nanoibers have been fabricated using physical, chemical, thermal, and electrostatic techniques, such as drawing (Xing et al. 2008), template synthesis (Tao and Desai 2007; Grimm et al. 2008), liquid–liquid phase separation (Ma and Zhang 1999), self-assembly (Hartgerink et al. 2001; Zhang 2003), vapor-phase polymerization (Rollings et al. 2007; Rollings and Veinot 2008), and electrospinning (Fong and Reneker 2001; Li and Xia 2004). It should be noted that self-assembly generally permits the production of nanoibers with diameters ranging from 5 to 8 nm. However, the complexity, the low productivity, and the nonstructured soft matrices formed by this technique limit its application (Ma et al. 2005b; Beachley and Wen 2010). Unlike self-assembly, phase separation is a simple technique that does not require specialized equipment. However, this method is effective with only a selected number of polymers and is not feasible for scale-up (Beachley and Wen 2010). Electrospinning is a unique nanoiber approach that has attracted the attention of various researchers because of its simplicity, versatility, economics, and scale-up (Fong and Reneker 2001; Li and Xia 2004; Huang et al. 2003). Moreover, a wide range of polymers and copolymers can be used to produce ibers with different diameters from a few micrometers down to tens of nanometers with controllable structures by adjusting the process parameters (Deitzel et al. 2001b; Bhardwaj and Kundu 2010; Baji et al. 2010). In the early 1900s, Cooley (1902) and Morton (1902) patented an apparatus and a process in which electric charges were used to electrospray luids. From 1934 to 1944, Formhals iled a series of patents describing an experimental setup for the production of polymer ibers using an electrostatic force (Formhals 1934, 1939, 1940, 1943, 1944). A polymer (cellulose acetate, CA) solution was introduced in an electric ield and the polymer ibers were formed from the solution between two electrodes bearing electrical charges of opposite polarity. About 50 patents on electrospinning have been iled in the last 70 years (Li and Xia 2004). Since the 1980s and especially in the last 10 years, electrospinning has regained attention probably due to the emerging interest in nanotechnology, as ultraine ibers and three-dimensional nanostructures of various polymers can be easily fabricated with this technique (Huang et al. 2003; Frenot and Chronakis 2003; Ramakrishna et al. 2006; Teo and Ramakrishna 2006; Bhardwaj and Kundu 2010). Electrospinning has also gained widespread interest as a potential polymer processing technique for applications in tissue engineering and drug delivery, because polymer nanoibers can provide three-dimensional architecture, modulate cell behavior, and have the potential to deliver biomolecules (Li et al. 2002; Ma et al. 2005a; Agarwal et al. 2008; Sill and von Recum 2008; Zhang et al. 2009). A number of electrospun polymeric nanoibers have been fabricated for applications in the diverse ields of afinity membrane (Iskandar 2009; Zhang et al. 2010a,b), biosensor, optical and chemical sensors (Wang et al. 2002; Gouma 2003; Ding et al. 2004; Li et al. 2006a), bioreactors (Huang et al. 2008), wound dressing (Liu et al. 2010), air iltration (Gibson et al. 2001), clean energy (Ramakrishna et al.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 217

2010), electronic and semiconductive materials (Choi et al. 2003; Kim and Yang 2003; Miao et al. 2010; Laforgue 2011), stimuli-responsive or “smart” materials (Fu et al. 2009), reinforced nanocomposites (Bergshoef and Vancso 1999), and membrane distillation (Feng et al. 2008). More details on the electrospun nanoibers and their applications may be found in a recent review by Fang et al. (2008).

8.1.1

ELECTROSPINNING PROCESS

The basic concept of electrospinning is to employ electrostatic repulsion forces in a highly charged polymer jet to produce randomly oriented or aligned nanoibers deposited on a collector surface. A typical electrospinning setup is shown in Figure 8.1. Basically, the setup consists of three major components: a high-voltage power supply; a capillary tube with a needle of small diameter, sometimes called a spinneret; and a grounded collector, which is usually a metal target, a plate, or a rotating mandrel. In addition, a metering syringe pump can be used to control the low rate of the polymer solution. The needle of the syringe typically serves as an electrode to electrically charge the polymer solution. The repulsion between the charges at the free surface of the needle tip works against the surface tension and the polymer liquid elasticity to deform the hemispherical surface of the polymeric droplet into a conical shape, called a Taylor cone (Taylor 1964). When the electric voltage applied between the injector and the collector exceeds the surface tension force of the Taylor cone, the charged jets are eventually sprinkled to the ground. Every time the polymer iber loops in the air gap, its diameter is reduced. For example, Figure 8.2 shows several Taylor cone images for different low rates (Q) and applied electric ields (E) using 2% polyethylene oxide (PEO) in water as the polymer solution. When the ibrous jet travels through the air gap between the needle and the collector, the solvent evaporates, generating a solidiied nanoibrous structure on the collector. Generally, a grounded target is used to collect the resultant ibers deposited in the form of a nonwoven mesh. Although the jet is stable near the tip of the needle, it soon starts bending, and whipping instabilities occur with further stretching of the solution jet under the electrostatic ield. These instabilities increase owing to the charge– charge repulsion between the excess charges present in the jet, which leads to the Syringe High voltage supply Needle tip or spinneret

Nanofibers Collector

FIGURE 8.1 Schematic diagram of an electrospinning setup.

Membrane Modification: Technology and Applications

218

(a)

(b)

(c)

(d)

FIGURE 8.2 Cone jets of 2 wt% solutions of polyethylene oxide (MW = 920 kg/mol) in water. D = 45 cm in all cases. (a) Q = 0.02 ml/min, E = 0.282 kV/cm; (b) Q = 0.10 ml/min, E = 0.344 kV/cm; (c) Q = 0.50 ml/min, E = 0.533 kV/cm; and (d) Q = 1.00 ml/min, E = 0.716 kV/cm. (Reprinted from Rutledge, G.C. and Fridrikh, S.V., Adv. Drug Deliv. Rev. 59, 1384, 2007. With permission.)

thinning and elongation of the jet (Yarin et al. 2001; Shin et al. 2001a,b; Rutledge and Fridrikh 2007). For iber assembly, two main methods can be applied: one is to control the light of the electrospinning jet by manipulation of the electric ield and the other is to use a dynamic collection device (Deitzel et al. 2001a; Kidoaki et al. 2005). Using different static collection devices, it is possible to assemble the ibers into a speciic form (Teo and Ramakrishna 2006).

8.1.2

POLYMERS USED IN ELECTROSPINNING

Almost any soluble polymer with a suficiently high molecular weight can be subjected to electrospinning. Nanoibers made of synthetic and natural polymers, polymer blends as well as melts, nanoparticle-impregnated or drug-impregnated polymers, and ceramic precursors have been successfully electrospun. Typically, a wide range of polymers such as those used in conventional spinning have been used in electrospinning, including polyurethanes (Schreuder-Gibson and Gibson 2002), polyamides (Tsai et al. 2002), polyester (Reneker and Chun 1996), polystyrene (Megelski et al. 2002; Lu et al. 2008), polyvinylidene luoride (Yang et al. 2009), poly(ether imide) (Koombhongse et al. 2001), styrene–butadiene–styrene triblock copolymer (Fong and Reneker 1999), and poly(vinylidene luoride-co-hexaluoropropylene) (Yao et al. 2009). Biopolymers such as proteins, DNA, collagen, polypeptides, or others such as electric-conducting and photonic polymers (Matthews et al. 2002; Li et al. 2005, 2006b; Zhong et al. 2006) and silk ibroin (SF) (Ohgo et al. 2003; Alessandrino et al. 2008; Meinel et al. 2009; Lee and Kim 2010) have also been used in electrospinning. Poly(a-hydroxy acids), especially lactic acids, glycolic acids, and their copolymers with 3-caprolactone, are the most commonly used among all the biodegradable polymers for the fabrication of nanoibers for medical use (Piskin et al. 2007; Nie et al. 2010). Different iber morphologies have also been shown, such as beaded, ribbon, porous, and core–shell ibers (Ramakrishna et al. 2005; Lee et al. 2010;

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 219

Wang et al. 2010). Although the most commonly electrospun material is polymer, ceramic precursors have also been electrospun without the addition of polymers (Son et al. 2006; Wang and Santiago-Aviles 2004).

8.1.3

ELECTROSPINNING PARAMETERS

The parameters involved in the electrospinning processes that affect the nanoiber geometry and structure can be divided into two groups: (i) System parameters such as polymer molecular weight, molecular weight distribution, polymer architecture (branched, linear), concentration of the polymer solution and its properties, including viscosity, electrical conductivity, and surface tension; and (ii) Process parameters such as applied electric voltage, polymer low rate, distance between the needle tip and the collector, ambient parameters such as temperature, humidity, and air velocity in the chamber, and motion of the collector (Frenot and Chronakis 2003). It is worth noting that 200 research papers have been published in the last decade on the fundamentals of electrospinning (Hohman et al. 2001a,b; Yarin et al. 2001; Rutledge and Fridrikh 2007; Jaworek and Sobczyk 2008) and on the effects of the electrospinning conditions on the nanoiber diameter and morphology (Tan et al. 2005; Deitzel et al. 2001a; Zhang et al. 2005a; Fridrikh et al. 2003; Haghi and Akbari 2007; Casper et al. 2004; Krishnappa et al. 2003). The parameters that affect the iber diameter have been discussed in several studies (Theron et al. 2004; Ramakrishna et al. 2005; Tan et al. 2005; Baji et al. 2010; Bhardwaj and Kundu 2010). Regarding the system parameters, one of the most signiicant factors inluencing the iber diameter is the polymer solution viscosity, which increases with an increasing polymer concentration. The higher the solution viscosity or the polymer concentration, the greater is the iber diameter. Several researchers found that the iber diameter increased with an increasing polymer concentration according to a power law relationship (Deitzel et al. 2001a; Ki et al. 2005). However, it should be pointed out that the viscosity range of a given polymer solution necessary for iber formation depends on the type of polymer. The molecular weight of the polymer is another important factor affecting the morphology of a nanoiber. Generally, highmolecular-weight polymer solutions have been used in electrospinning (Haghi and Akbari 2007). The polymer molecular weight relects the number of entanglements of the polymer chains in a solution and is directly related to the solution viscosity (Koski et al. 2004; Mit-uppatham et al. 2004). The surface tension also plays a critical role in the electrospinning process. Basically, the surface tension determines the upper and lower boundaries of the polymer solution suitable to form nanoibers (Fong and Reneker 1999; Supaphol et al. 2005; Haghi and Akbari 2007). In electrospinning processes, a crucial process parameter is the electrical voltage applied to the polymer solution. However, the level of signiicance varies with the polymer solution and the distance between the needle and the collector (Yördem et al. 2008). Reneker and Chun (1996) showed that the applied voltage did not have much of an effect on the diameter of the electrospun PEO nanoibers. Other researchers suggested that a higher applied voltage ejects more luid in a jet and this favors the formation of a larger iber diameter (Zhang et al. 2005a). On the contrary, other authors reported that the electrostatic repulsive forces in the formed luid jet increased with

220

Membrane Modification: Technology and Applications

an increasing applied voltage and the iber diameter decreased (Buchko et al. 1999; Deitzel et al. 2001a; Megelski et al. 2002). Few systematic studies on the relationship between the low rate of the polymer solution and the nanoiber diameter have been carried out (Megelski et al. 2002; Zong et al. 2002; Son et al. 2004b). Megelski et al. (2002) studied the effects of the low rate of a polystyrene/tetrahydrofuran (THF) solution on the structure of the electrospun ibers and found that both the iber diameter and the pore size increased with the increasing low rate. The distance between the needle and the collector has been examined as another process parameter to control the iber diameter and its morphology. By reducing this distance, the electrospun ibers tended to stick to the collecting device and to one another, due to the incomplete solvent evaporation, with the formation of a mesh with interconnected ibers (Buchko et al. 1999). Therefore, it was observed that a minimum distance is required to give the ibers enough time to dry before reaching the collector. However, it should be pointed out that the effect of this distance on the iber morphology is not as signiicant compared with the effects of the other previously cited parameters (Zhang et al. 2005a; Li et al. 2005; Chen et al. 2006b; Zhao et al. 2005). Various collectors and methods have been adopted to obtain aligned ibers: a dynamic collector such as a rotating drum (Xu et al. 2004); a pair of parallel conducting electrodes to create an electric ield so that the electrospun ibers are preferentially aligned across the gap between the electrodes (Li and Xia 2004); and a wire screen (Wang et al. 2005). Tubular electrospun ibers can be fabricated on a rotating tube and the deposited iber layer subsequently extracted from the tube.

8.2

SURFACE MODIFICATION TECHNIQUES

The surface modiication of polymers is a classic research topic, and it is still receiving extensive attention because new applications of polymeric materials have emerged, especially in the ields of biotechnology and bioengineering and most recently in nanotechnology. In order to successfully apply electrospun nanoibers in such wide ields, which include adhesion, membrane iltration, coatings, friction and wear, composites, microelectronic devices, thin-ilm technology, biomaterials, and biomedical uses, their surfaces should be chemically and/or physically modiied after electrospinning in order to provide the surface with the necessary properties for a successful application. These surface properties include chemical composition, hydrophilicity, roughness, crystallinity, conductivity, lubricity, and antibacterial activity. In membrane technology, surface modiication is a powerful tool that is used to enhance the performance of the membranes. The main purpose of surface modiication is to change the surfaces by altering the atoms/molecules in the existing surface, changing the surface topography, or coating over the existing surface with a material having a new composition. There are several surface modiication techniques that can be broadly categorized as physical or chemical approaches (Kato et al. 2003; Ma et al. 2007b; Gopal et al. 2007; Yoo et al. 2009). The deinition of these broad categories depends on how the process actually affects the surface. Briely, physical methods, in most cases, do not change the chemical composition of the surface. These methods may change the surface roughness, grain sizes, grain boundaries,

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 221

surface pore size, and surface porosity. In some cases, the physical surface modiication methods can lead to changes in the chemical composition of the functional groups at the surface of the nanoiber due to the removal or addition of a chemical group or the activation of a chemical reaction on the surface of a material. For example, selective or ion-beam sputtering or selective cross-linking in the presence of plasma can induce chemical changes, such as chain scissions, cross-linking, surface oxidation, and depletion of the low-molecular-weight fragments, leading to the formation of a signiicant amount of hydroxyl groups, and even carbonyl groups. The chemical methods introduce a change in the chemical composition of the surface of a material. The surface may or may not possess chemical properties that are different from that of the bulk material. Physical processes take advantage of surface segregation, radiation, and oxidation with gases, while chemical modiications use wet treatment, coating, and blending or co-electrospinning of the surface-active agents and the polymers. Other chemical methods include treatment with UV radiation and reactive plasmas. In the following sections, some interesting nanoiber surface modiication techniques are reviewed.

8.2.1

BLENDING

Blending is the simplest and easiest method employed to functionalize a polymer. This is a physical approach with the addition of blending ligand molecules into the polymer solution and then electrospinning the polymer solution. No chemical bonding or attachment is involved between the polymer material and the modiied species (Figure 8.3). It is a simple mixing of two or more materials that has been proven to be an effective method for polymer nanoiber modiication. Nevertheless, blend molecules are susceptible to detachment and the technique is neither reproducible nor controllable. An example of blending was when phenylcarbomylated or azido phenylcarbomylated β-CD was successfully blended with polymethyl methacrylate (PMMA) and electrospun into nanoibrous membranes for organic waste treatment and water puriication (Kaur et al. 2006). The presence of the β-CD derivatives on the surface of the nanoibers was conirmed by attenuated total relectance–Fourier transform infrared spectrometry (ATR-FTIR) and x-ray photoelectron spectroscopy (XPS). A solution containing phenolphthalein (PHP) was used to determine the ability of the functionalized membranes to capture small organic molecules. The results showed

Biologically or therapeutically functional agents

Blend solution

Electrospinning Surface orientation

FIGURE 8.3 A schematic representation of the blending technique for the surface modiication of electrospun nanoibers. (Adapted from Yoo, H.S., et al., Adv. Drug Deliv. Rev., 61, 1033, 2009. With permission.)

222

Membrane Modification: Technology and Applications

that the functionalized nanoibrous membranes were able to capture the PHP molecules effectively. Polymer blends and modiied polymers can also be synthesized to include the desired exposed functional groups on their structure. A poly(ɛ-caprolactone)/ poly(ethylene glycol) (PCL–PEG) block copolymer was synthesized with the functional amine groups on the surface via PEG linkers and electrospun into nanoibers that were further functionalized (Choi et al. 2008). Bellan et al. (2006) electrospun blended PEO and DNA and demonstrated the ability to produce polymeric nanoibers containing isolated stretched DNA molecules. The embedded DNA molecules were imaged with luorescence microscopy by incorporating a dilute concentration of luorescently labeled DNA molecules into an electrospinning process. The direct observation of the degree to which the DNA molecules had been stretched could give information on the luid dynamic behavior of the jet and the mechanical properties of the nanoibers. Electrospun SF-based ibers were prepared from aqueous regenerated silkworm silk (Bombyx mori)/PEO solutions to be used as scaffolds for tissue engineering (Jin et al. 2004). PEO supplied good mechanical properties to the electrospun ibers. An MeOH posttreatment induced an amorphous to silk β-sheet conformational transition. The electrospun silk membrane was washed with water to remove PEO in order to improve the cell adhesion and proliferation. These silk ibrous membranes were nonimmunogenic, biocompatible, and capable of supporting bone marrow stromal cell (BMSC) attachment. In another work, electrospun wool keratin/silk ibroin (WK/SF) blend nanoibers exhibited higher Cu2+ adsorption capacity than SF nanoibrous membrane (Ki et al. 2007). Blends allow for the fabrication of collagen containing nanoibers with a greater range of mechanical properties and iber diameters. Ma et al. (2007a) reported that collagen-blended poly(lactic-co-glycolic acid) (PLGA) nanoibers have great potential to facilitate wound healing in skin tissue engineering. Mechanical tests showed that the ultimate strain value and tensile modulus of the blended nanoibers were comparable to those of human skin. Contact angle assessment showed their decreased hydrophobicity compared with pure PLGA nanoibers and suggested an improved capability for cell attachment. In another study, collagen was directly blended into the poly(l-lactic acid)-co-poly(ɛ-caprolactone) PLLA-co-PCL (70:30) nanoibers (He et al. 2005a). The blended nanoibers, with different weight ratios of polymer to collagen, were fabricated by electrospinning a mixture of collagen and the copolymer solution. The morphology of the electrospun nanoibers was investigated, proving the existence of collagen molecules on the surface of the nanoibers. Five characteristic endothelial cell (EC) markers were studied by RT-PCR, and the results showed that the collagenblended polymer nanoibers preserved the EC’s normal phenotype and enhanced cell viability, spreading, and attachment, indicating a potential application as a vascular graft in tissue engineering. To mimic the ratio of collagen and elastin in the native blood vessel, electrospun nanoibers of collagen/elastin/synthetic polymer (e.g., PLGA, PLLA, and PCL) blends were prepared (Lee et al. 2007). The electrospun nanoiber membrane showed no cytotoxicity, was dimensionally stable, and its mechanical properties were similar to the native blood vessels. Liu et al. (2010) prepared PLGA/collagen electrospun nanoiber membrane

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 223

for wound dressing. In the cell activity assessment, the electrospun (50/50 PLGA/ collagen ratio) nanoibers exhibited signiicant levels of cell proliferation during culture. This may be a consequence of the high surface area available for cell attachment due to the three-dimensional features of the iber construct and the restoration of the biological and structural properties of the natural extracellular matrix (ECM) proteins. Several studies directly comparing the electrospun collagen and the gelatin nanoibers indicate that the electrospun collagen nanoibers may retain some favorable bioactivity when compared with the gelatin nanoibers, despite solvent denaturation of the collagen. The gelatin nanoiber blends with a variety of iber diameters can also be electrospun by combining gelatin and other polymers in one solution. The cells attached and proliferated better on the nanoibers when they were blended with gelatin (Ghasemi-Mobarakeh et al. 2008; Kim et al. 2008). Increases in cell attachment and proliferation have been shown to be a function of the ratio of gelatin in the nanoiber blends (Jeong et al. 2008). PCL ibers blended with gelatin also enhanced nerve differentiation as compared with plain PCL nanoibrous scaffolds. Chitosan nanoibers have been electrospun as blends with very low concentrations of PEO to enhance cell attachment and promote the liver function of hepatocytes cultured in bioreactors (Chu et al. 2009; Bhattarai et al. 2005). Electrospun nanoibers of a chitosan–hydroxybenzotriazole (CS–HOBt)/polyvinyl alcohol (PVA) blend were prepared without organic solvent or organic acids (Charernsriwilaiwat et al. 2010). The morphologies of the ibers and their diameters were strongly inluenced by the weight ratio in this blend. Cytotoxicity tests showed that CS–HOBt/PVA nanoiber membranes were nontoxic to ibroblast cells. Chuangchote et al. (2008) fabricated ultraine poly[2-methoxy-5-(20-ethylhexyloxy)1,4-phenylenevinylene]/poly(vinyl pyrrolidone) (PVP) composite ibers for conductive nanoibers by electrospinning blended solutions in a mixed solvent of chlorobenzene and methanol. The average diameter of the electrospun ibers was found to decrease with the reduction of the PVP concentration and/or the addition of the volatile organic salt, pyridium formate. Vasita et al. (2010) reported the surface hydrophilization of the electrospun membranes of PLGA (lactic/glycolic acid, 85:15) by blending with small quantities (0.5%–2% w/v) of a hydrophilic polymer, Pluronic F-108 (PF-108). The blended iber meshes showed a decrease in the surface contact angle when compared with the pure PLGA iber meshes, demonstrating an improvement in the surface hydrophilicity due to blending. Moreover, by varying the type and the concentration of the PF-108 used for blending with PLGA, it was possible to modulate the surface properties of the electrospun PLGA micro/nanoibrous meshes without signiicantly inluencing the bulk properties of the PLGA micro/nanoibers. PVP was successfully introduced to PLLA microiber ilms, which were fabricated by electrospinning in order to improve their hydrophilicity (Xu et al. 2009). The PVPblended membranes were claimed to be beneicial by improving the compatibility of the vascular smooth muscle cells (VSMCs) with the PLLA substrates. The platelet adhesion test showed that the membranes had excellent blood compatibilities with the addition of PVP in PLLA.

224

Membrane Modification: Technology and Applications

Blending was also used for the bulk modiication of the polymer nanoibers for drug delivery applications. Luu et al. (2003) used electrospinning to fabricate synthetic polymer/DNA composite scaffolds for therapeutic application in gene delivery for tissue engineering. A PLGA random copolymer and a poly(d,l-lactide)– poly(ethylene glycol) (PLA–PEG) block copolymer were used. The release of plasmid DNA from the scaffolds was sustained over a 20-day study period, with a maximum release occurring after 2 h. In another example, blending a hydrophilic but water-insoluble polymer poly(ethylene glycol)-g-chitosan (PEG-g-CHN) with PLGA could assist in the release of a poor water-soluble drug, Ibuprofen (Jiang et al. 2004).

8.2.2 SURFACE LAYERS The formation of ultrathin surface layers (10 nm thick or less), either through bottomup approaches such as self-assembly or the application of surface coatings through synthetic chemistry on different types of surfaces, is one of the major research areas for membrane surface modiication. The main reason for such interest is the number of applications requiring relatively defect-free surfaces with systematically engineered properties. Coatings are predominantly used to enhance the performance of nanoibrous scaffolds. 8.2.2.1 Coating Collagen (type I), ibronectin, and laminin have been coated on an electrospun SF nanoiber surface to promote cell adhesion (Min et al. 2004). The cell activities of normal human keratinocytes and ibroblasts on coating the SF nanoiber were investigated. Poly(p-xylylene) (PPX)-coated poly(vinyl alcohol)/bovine serum albumin (PVA/BSA) nanoibers were prepared by chemical vapor deposition (CVD) (Zeng et al. 2005). The release of BSA from uncoated PVA nanoibers was fast under physiological conditions. By contrast, PPX-coated nanoibers exhibited a signiicantly retarded release of BSA, depending on the coating thickness of PPX. This continuous release of the intact protein from the immersed ibers meets a fundamental prerequisite for the application of proteins such as enzymes or other sensitive agents released from electrospun nanoibers under physiological conditions. Zhang et al. (2005c) studied the effect of nanoiber surface coatings on cell proliferation behavior. They prepared PCL nanoibers with a rough collagen-coated surface and individually collagen-coated PCL nanoibers (i.e., collagen-r-PCL in the form of a core–shell structure) by a coaxial electrospinning technique. It was found that the coatings of collagen on the PCL nanoibrous matrix deinitely favored cell proliferation and the eficiency of human dermal ibroblast (HDF) proliferation was coating-dependent. Collagen-r-PCL was signiicantly more favorable for cell proliferation than the rough collagen-coated PCL. The results indicated that the individual nanoiber coated with collagen tends to resemble the natural ECM rather than the rough collagen coating or the pristine PCL nanoibers. Araujo et al. (2008) prepared PCL electrospun nanoiber meshes coated with a biomimetic calcium phosphate (BCP) layer that mimics the extracellular microenvironment found in the human bone structure. The deposition of a calcium phosphate layer, similar to the inorganic phase of bone, on the PCL nanoiber meshes was

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 225

achieved by means of a surface modiication followed by treatment with solutions containing calcium and phosphate ions. This coating enhanced the proliferation of osteoblasts for long culture periods. Surface modiication by plasma treatment has also been shown to improve the coating effectiveness by improving the wetting of the hydrophobic polymer surfaces with the collagen solution (He et al. 2005a,b). The collagen-coated poly(l-lactic acid)-co-poly(ɛ-caprolactone) P(LLA-CL 70:30) nanoiber was fabricated by plasma treatment of the electrospun P(LLA-CL) nanoiber followed by collagen coating, as shown in Figure 8.4. Air plasma treatment of the P(LLA-CL) nanoiber membrane was carried out in an inductively coupled radio-frequency glow discharge plasma cleaner. It was found that the nanoiber P(LLA-CL) membrane was a potential material for a tissue-engineered vascular graft. The collagen-coated P(LLA-CL) nanoiber membrane also showed mechanical properties suitable for vascular graft. Plasma treatment can also increase the surface hydrophilicity of materials (Chen et al. 2003; Wan et al. 2003). The collagen-coated P(LLA-CL) nanoiber membrane enhanced the spreading cell morphology and increased the viability and the attachment of the human coronary artery endothelial cells (HCAECs) (He et al. 2005b). The same researchers fabricated collagen-coated random and collagen-aligned poly(l-lactic acid)-co-poly(ɛ-caprolactone) (P(LLA-CL)) nanoiber membranes by electrospinning (He et al. 2006). Mechanical testing showed that the tensile modulus and strength were greater for the aligned P(LLA-CL) nanoiber membrane than Collagen solution cannot penetrate into the mesh

P(LLA-CL) nanofiber mesh (hydrophobic) Collagen solution can be absorbed into the mesh immediately Air plasma–treated P(LLA-CL) nanofiber mesh (hydrophilic)

Dried Collagen layer

FIGURE 8.4 A schematic representation of the plasma treatment and collagen coating of the electrospun P(LLA-CL) nanoiber membrane. (Reprinted from He, W., et al., Biomaterials, 26, 7606, 2005. With permission.)

226

Membrane Modification: Technology and Applications

for the random nanoiber membrane. Both the random and the aligned P(LLA-CL) nanoiber membranes preserved their phenotypes with the expression of platelet EC adhesion molecule-1, ibronectin, and collagen type IV at the protein level. The cell functions of HCAECs were also maintained, as demonstrated by complementary DNA microarray analysis of 112 genes relevant to the EC functions. It was observed that the HCAECs grew along the direction of the nanoiber alignment and showed elongated morphology that simulated ECs in vivo under blood low. The coating of (luoroalkyl)silane (FAS) on the electrospun inorganic silica nanoiber membranes was found to be responsible for the change in the surface wettability of the silica nanoiber membrane from amphiphilic to amphiphobic (Guo et al. 2010). The inorganic silica nanoiber membranes were obtained via electrospinning of the blend solutions of PVA and silica gel, followed by calcination to remove the organic component. Additionally, the luorinated inorganic nanoiber membrane exhibited a high heat resistance and high hydrophobicity (water contact angle of 154°) and oleophobicity (oil contact angle of 144°). In recent years, depositing a metal oxide onto electrospun precursors has been found to exhibit some interesting properties, such as electrical conductivity and optical transparency in the visible range (Deng et al. 2007). There are some techniques employed to deposit zinc oxide onto electrospun precursors, for example, sol–gel coating (Kumar et al. 2007), sputter coating (Deng et al. 2007), and a pulsed laser deposition method (Villanueva et al. 2006). The sol–gel technique has several advantages over other processes because of its simplicity, easy control of the ilm composition, safety, and the low cost of the necessary apparatus and raw materials. Shao et al. (2008) prepared polyacrylonitrile (PAN) nanoibers by electrospinning with a functional layer of zinc oxide (ZnO) deposited by the sol–gel technique. The effects of coating, preoxidation, and carbonization on the surface morphology and structures of the nanoibers were characterized. The ZnO-coated PAN nanoibers were carbonized to form functional carbon nanoibers. FTIR scans corroborated that ZnO was deposited onto the surface of the carbon nanoibers, and the SEM observations revealed that ZnO was clearly distributed on the surface of the carbon nanoibers as nanoclusters. Figure 8.5 shows an example of the SEM images of ZnO-coated PAN nanoibers after preoxidation and carbonization. A signiicant increase in the size of the ZnO nanograins on the surface of the nanoibers was observed. To use nanoiber membranes as conductimetric sensors, they should be functionalized with metal oxide semiconductors. Drew et al. (2003) reported the fabrication of novel metal oxide–coated polymeric nanoibers using the electrospinning technique. Metal oxide coatings on the nanoibrous structures were created by a technique known as liquid-phase deposition (LPD). A metal precursor is irst hydrolyzed in an aqueous solution, which subsequently forms a metal oxide by the condensation of water. Both tin dioxide (SnO2) and titanium dioxide (TiO2) were successfully coated on PAN electrospun nanoibers. This technique effectively coated the individual electrospun ibers, leaving the inherent high surface area of the electrospun membrane intact. Such metal oxide–coated nanoibrous membranes are expected to provide unusual and highly reactive surfaces for improved sensing, catalysis, and photoelectric conversion applications.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 227

(a)

(b)

(c)

(d)

FIGURE 8.5 SEM images of PAN nanoibers (a) at a magniication of 10,000; (b) at a magniication of 90,000 and for ZnO-coated PAN nanoibers after the treatments of coating, preoxidation, and carbonization; (c) at a magniication of 10,000; and (d) at a magniication of 90,000. (Adapted from Shao, D., et al., Appl. Surf. Sci. 254, 6543, 2008.)

Gouma produced molybdenum oxide (MoO3)-containing PEO nanoibers by electrospinning a mixture of MoO3 sol–gel and a PEO solution (Gouma 2003). A gas-sensing test of the electrospun nanoiber mat was carried out using ammonia and nitroxide as the harmful model gas. The electrical resistance of the sensing ilm as a function of the gas concentration was measured, and the results showed that the nanoscale metal oxide ibers offered a high sensitivity and a fast response to the gas sensing of harmful chemical species. PCL electrospun nanoibers were coated onto a PCL membrane to obtain a nanotopographical surface (Chen et al. 2007a). The modiied PCL nanoiber membrane showed an almost zero water contact angle due to the capillary action on the highly rough surface. A favorable cellular morphology and strong cell attachment were observed on the nanoibrous substrate after NaOH treatment. 8.2.2.2 Layer-by-Layer Assembly A versatile surface modiication method that allows surface coating with a thickness from a few nanometers to several micrometers through precise control has been realized by a layer-by-layer (LbL) polyelectrolyte multilayer assembly (Delcorte et al. 1997; Thierry et al. 2003; Zhang et al. 2005b; Tang et al. 2006). The method

228

Membrane Modification: Technology and Applications

comprises of the alternative LbL deposition of polyanions and polycations principally driven by an electrostatic force on charged substrates, resulting in a self-assembled multilayer coating. The combination of LbL electrostatic assembly and electrospinning takes advantage of the high speciic surface area, lightweight, and high porosity of the electrospun iber mats. It also provides the versatility to incorporate different functional polyelectrolytes to achieve multifunctional coatings for simultaneous chemical and biological protection. This technique has attracted considerable attention due to the ease of synthesis, universality of any complex structure of substrate, possibility of using any composition for the coating layer, and the fact that it forms highly tailored polymer thin-ilm structures. Wang et al. (2004) reported a new fabrication approach to highly sensitive optical sensors by combining the techniques of electrospinning and electrostatic LbL assembly. A luorescent probe assembled by the LbL method onto electrospun CA membranes resulted in a signiicant increase in the sensitivity of the optical sensors. For the application in protective fabrics, Lee et al. (2009a) demonstrated that electrospun ibers used as a substrate for titanium dioxide (TiO2) nanoparticle coatings resulted in high speciic substrate surface areas of about 104 times that of the lat ilm, which enhanced the photocatalytic decomposition of toxic industrial chemicals. Krogman et al. (2009) employed a newly developed spray-assisted LbL assembly technique to functionalize electrospun nylon ibers with titanium dioxide nanoparticles for protective fabrics. Individual nanoibers within the matrix can be conformally functionalized for ultrahigh-surface-area catalysis or bridged to form a networked sublayer with complimentary properties. The LbL-functionalized iber mats improved the photocatalytic capability without sacriicing the water vapor permeability of the electrospun nanoiber. In another example, Li et al. (2008b) fabricated ibrous scaffolds of poly(ɛ-caprolactone) in the form of nonwoven mats. The surfaces of the ibers were then coated with gelatin through LbL self-assembly, followed by functionalization with a uniform coating of bone-like calcium phosphate by mineralization. They found that the presence of gelatin facilitated a homogeneous calcium phosphate coating. It was also demonstrated that the incorporation of gelatin promoted the nucleation and growth of calcium phosphate. The porous scaffolds could mimic the structure, composition, and biological function of bone ECM. Moreover, by selecting the appropriate polyelectrolytes of opposite charges and varying the functionalities, multifunctional electrospun iber-based fabrics for both chemical and biological protection can be fabricated. Muller et al. (2006) fabricated polystyrene (PS) nanoibers by electrospinning with the surface of the electrospun ibers negatively charged by sulfonation of the phenyl groups. LbL assembly with poly(allylamine hydrochloride) (PAH) and poly(styrene sulfonate) (PSS) was also conducted on the surface of the electrospun ibers, followed by the dissolution of the PS core ibers, producing hollow polyelectrolyte complex nanotubes. Di-block oligonucleotide (polyA15G15 and polyT15C15) multilayers were also built up based on a DNA hybridization mechanism and the LbL technique. The multilayer structure of the DNA grew in a linear manner with a smoother surface compared with the iber surface prepared by PAH/PSS multilayers. It was suggested that this surface modiication of the electrospun ibers with the synthetic polymers (PAH and PSS) and the biopolymer (DNA) could provide the opportunity for creating a variety of drug-releasing surfaces for biomedical applications.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 229

A nucleophilic and chemically reactive polyanion, polyhydroxamic acid (PHA), and an antimicrobial polycation, poly-(N-vinylguanidine) (PVG), were synthesized and assembled onto prefabricated PAN iber mats (Yang et al. 1992). The performance of the functionalized mats in organophosphorous decomposition was tested with diisopropyl luorophosphate (DFP). This is commonly used as a stimulant of G-type nerve agents. The antibacterial properties of these functionalized mats were examined with Escherichia coli and Staphylococcus epidermidis, Gram-negative and Gram-positive microorganisms, respectively. The electrospun PAN nanoiber membranes were plasma-treated prior to the LbL deposition of an aqueous solution of PHA and PVG (Chen et al. 2010). Multilayer polyelectrolyte coatings on the PAN ibers were produced by alternately dipping the nanoiber membrane in 10 mM PVG and PHA aqueous solutions. The dipping time in each polyelectrolyte solution was 60 min, followed by rinsing in deionized water for 3 min. The PVG/PHA-functionalized nanoiber membrane demonstrated potent bactericidal capability over representative strains of E. coli and S. epidermidis, when used as biological protective fabrics. Yarin and Zussman (2004) developed a technique for the formation of upward needleless electrospinning of multiple nanoibers. The approach is based on a combination of normal magnetic and electric ields acting on a two-layer system. The lower layer, a ferromagnetic luid, and the upper layer, a PEO solution, were subjected to a normal magnetic ield provided by a permanent magnet or coil. Under the action of the magnetic ield, numerous steady spikes were generated at the free surface of the magnetic luid. As a result, the interlayer interface and the free surface of the uppermost polymer layer were perturbed. The magnitude of the perturbation forces is such that the peak height of the magnetic luid spikes can exceed the uppermost polymer layer. When a normal electric ield was applied in addition to the magnetic ield, multiple jets of the polymer solution are ejected towards the grounded lat counter electrode. Multiple electriied jets underwent strong stretching by the electric ield with bending instability. To stabilize the process, a grounded piece of metal saw with teeth oriented downward was used as a counter electrode. The production rate of the jets increased by a factor of 12 when this method was used instead of separate nozzles. Ogawa et al. (2007) tried to mimic the topography of silver ragwort leaves with a FAS-modiied, LbL-structured, ilm-coated, electrospun nanoiber membrane. The rough iber surface caused by the electrostatic LbL coating of the TiO2 nanoparticles and the poly(acrylic acid) (PAA) was used to imitate the rough surface of the nanosized grooves along the silver ragwort leaf iber axis. A schematic diagram illustrating the preparation of superhydrophobic surfaces via LbL coating and FAS surface modiication is shown in Figure 8.6. The LbL-structured ilms were assembled on a CA nanoiber surface by alternating the adsorption of the positively charged TiO2 nanoparticles and the negatively charged PAA. The results showed that the FAS modiication was the key process for changing the membrane surface property from superhydrophilic to hydrophobic and even to superhydrophobic. Other researchers also tried to imitate the micro/nanostructure of the silver ragwort leaf by forming a rough surface on the smooth CA nanoibers using an electrostatic LbL self-assembly technique (Ding et al. 2005, 2006). The wettability of the electrospun ibers (CA)

230

Membrane Modification: Technology and Applications Cross section of cellulose acetate nanofiber (−) (1) TiO2(+) adsorption

(2) PAA(−) adsorption

Repeat procedure (1) and (2)

LbL films–coated nanofiber

FAS modification

Super hydrophobic nanofiber

FIGURE 8.6 A schematic diagram illustrating the preparation of superhydrophobic surfaces via layer-by-layer (LbL) coating of TiO2 nanoparticles and poly(acrylic acid) (PAA) and luoroalkylsilane (FAS) surface modiication. (Reprinted from Ogawa, T., et al., Nanotechnology 18, 165607, 2007. With permission.)

was changed from a hydrophilic to a hydrophobic surface (water contact angle of 156°) with a simple sol–gel coating of decyltrimethoxysilane (DTMA) and tetraethyl orthosilicate. After coating, the ibrous and the porous structures were maintained. The superhydrophobicity was attributed to the combined effects of the high surface roughness of the electrospun nanoiber membrane and the hydrophobic DTMS sol– gel coating. In chemiluminescent-based assay systems, Mark et al. (2008) found that noncovalent immobilization of alkaline phosphatase (ALP) on the electrospun nylon 6 nanoiber membrane, using a multistacked LbL technique with the cationic polymer Sapphire II, resulted in the highest enzyme loading when compared with other covalent immobilization methods based on glutaraldehyde cross-linking.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 231

8.2.3

PLASMA TREATMENT

The plasma treatment of polymer substrates has been commonly employed to tailor the surface adhesion and wetting properties by changing the surface chemical composition (Ladizesky and Ward 1995; Grace and Gerenser 2003). The appropriate selection of a plasma source enabling the introduction of diverse functional groups on a membrane surface to improve its biocompatibility or to allow subsequent covalent immobilization of bioactive molecules is important since plasma treatments with oxygen, ammonia, or air can generate carboxyl groups or amine groups on the surface (Park et al. 2007a; Yoo et al. 2009). For example, the number of polar groups, which contributed to the enhanced surface hydrophilicity, could be more effectively increased by the ammonia gas–plasma treatment on a PLGA nanoiber (Park et al. 2007a). A schematic representation of a plasma treatment for the surface modiication of electrospun nanoibers is shown in Figure 8.7. The plasma treatment is also used as a pretreatment for the deposition of a macromolecular structure, graft copolymerization, functionalization, etching, roughening, or cross-linking. A variety of ECM protein components, such as gelatin, collagen, laminin, and ibronectin, could be immobilized onto a plasma-treated surface to enhance cellular adhesion and cell proliferation (He et al. 2005b, 2006; Ma et al. 2005d; Baek et al. 2008). Electrospun nanoibers composed of poly(glycolic acid) (PGA), PLLA, or PLGA were modiied with the carboxylic acid groups through a plasma glow discharge with oxygen and ammonia (Park et al. 2007a,b). The surface hydrophilicity of the electrospun PLGA nanoibers was increased signiicantly by plasma treatment. Such hydrophilized nanoibers were found to enhance the ibroblast adhesion and proliferation without altering the physical and mechanical bulk properties. Air or argon plasma treatment has been widely used as a facile surface modiication technique for many biomaterials since its surface hydrophilicity can be easily increased with this treatment. Yoon et al. (2009) investigated the preparation and

Biocompatible nanofibers Plasma or wet chemical treatment Electrospun nanofibers

Functionalized surface

Immobilization of protein, enzyme, growth factor, drug Biologically or therapeutically functionalized nanofibers

FIGURE 8.7 A schematic representation of plasma treatment (valid also for the chemical method) for the surface modiication of electrospun nanoibers. (Adapted from Yoo, H.S., et al., Adv. Drug Deliv. Rev., 61, 1033, 2009.)

232

Membrane Modification: Technology and Applications

characterization of lotus leaf–like micro/nanoibrous mats with cellulose triacetate (CTA). To improve their hydrophobicity, the CTA electrospun ibers were treated with CF4 plasma for 0–300 sec. Fluorine atoms could be introduced onto the CTA iber surface, resulting in a decrease in the surface energy and an increase in the hydrophobicity of the CTA ibrous mats. Various electrospun nanoibers made of PCL, PCL/hydroxyapatite, polystyrene, and SF were surface-modiied by air or argon plasma, resulting in an improved cell adhesion and proliferation (Venugopal et al. 2008; Prabhakaran et al. 2008; Yang et al. 2008). PCL nanoibrous scaffolds were surface-modiied by a simple plasma treatment process to enhance the Schwann cell adhesion, proliferation, and interactions with the nanoibers necessary for nerve tissue formation (Prabhakaran et al. 2008). The results showed that plasma-treated PCL nanoibrous scaffolds are a costeffective material compared with PCL/collagen scaffolds and can potentially serve as an ideal tissue-engineered scaffold, especially for peripheral nerve regeneration. A poly-ɛ-caprolactone/hydroxyapatite (PCL/HA) nanoibrous surface was modiied by oxygen plasma treatment, observing a 0° contact angle for the adhesion and mineralization of the osteoblast cells (Venugopal et al. 2008). The osteoblast proliferation rate was signiicantly increased at the modiied surface of the nanoibers because HA acted as a chelating agent for the mineralization of the osteoblasts for bone tissue engineering. It should be mentioned that plasma treatment can also be used to improve the hydrophilicity of the nanoibrous membranes to improve the covalent bonding modiications (Duan et al. 2007). This will be explained later in Section 8.2.6.3.

8.2.4 CHEMICAL METHODS The plasma treatment of a nanoibrous mesh cannot effectively modify the surface of the hidden nanoibers located deep within the mesh because of the limited penetration depth of plasma, thus wet chemical methods can be more effective surface modiication techniques because of their lexibility for surface modiication of thick nanoibrous membranes (Croll et al. 2004). Chemical modiication involves the introduction of one or more chemical species to a given surface in order to produce a surface with enhanced chemical and physical properties. Wet chemical oxidation treatments are commonly employed to introduce oxygen-containing functional groups, such as carbonyl, hydroxyl, and carboxyl groups, at the surface of the nanoibrous membranes (Agarwal et al. 2010). The oxygen-containing functional groups increase the polarity and the ability of hydrogen bond, which favors wettability and adhesion. Chemical reactions can also be carried out on nanoiber membranes that are vulnerable to electrophilic or nucleophilic attack by means of structures such as benzene rings, hydroxyl groups, double bond, and halogen (Agarwal et al. 2010). Functional groups may already be present in the polymer structures of the chemicals or can be added to the surface by various chemical treatments, depending on the molecular coniguration of the speciic polymer required for modiication. For example, the polymer PMMA, which consists of methyl ester groups, can be chemically modiied by a simple reduction reaction to alcohols, with lithium aluminum hydride in ether-based solutions, followed by the widely used organosilane

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 233

chemistry to introduce diverse functional groups, such as perluoroalkyl (–CnF2n+2), amino (–NH2), and sulfhydryl (–SH) for surface passivation or further biomolecule immobilization (Cheng et al. 2004). Alternatively, amino functionalities may also be placed on the PMMA surface through aminolysis of the ester groups by treatment with a solution of N-lithio diaminopropane in cyclohexane (Henry et al. 2000). The functional groups most commonly used for modifying the surface of the polymer nanoibers are the carboxyl and amine groups. For example, the carboxyl groups can be introduced on the surface of the PLLA or PCL nanoibers by treatment with sodium hydroxide (NaOH) (Chen et al. 2006a, 2007; Yu et al. 2009). The duration of the hydrolysis and the concentration of the hydrolyzing agents are important to optimally produce surface functional groups with a minimum change in the bulk property of the nanoibrous material. The NaOH-treated PLLA nanoiber membrane was used for hydroxyapatite (HAp) mineralization (Chen et al. 2006a). The mineralization process with a modiied PLLA nanoiber membrane was signiicantly enhanced because the calcium ions can bind to the carboxyl groups on the nanoiber surface. Amine groups were used on the surface of the polyester nanoibers by treatment with a 1,6-hexanediamine/propanol solution or ethylenediamine (ED) in order to ensure a positive surface charge and to create an amine functionalized surface (Croll et al. 2004; Zhu et al. 2002, 2007). An electrospun poly(l-lactide-co-caprolactone) copolymer (PLLC) nanoibrous scaffold was surface-modiied via aminolysis, followed by immobilization of a cell adhesive protein, ibronectin, by glutaraldehyde (Zhu et al. 2007). Surface modiication with chitosan on a poly(acrylonitrile-co-acrylic acid) (PANCAA) nanoiber membrane was accomplished by the coupling reaction of the carboxylic groups of PANCAA and the primary amino groups of chitosan. Platelet adhesion experiments were further carried out to evaluate the hemocompatibility (Che et al. 2008). When the desired functional groups cannot be easily grafted directly onto the polymer nanoiber surface, then a linker molecule with the desired functional groups can be attached to the iber surface. Acrylic acid (AA) was grafted onto a PGA, PLLA, and PLGA nanoiber membrane using oxygen plasma treatment to add the hydrophilic functional groups (carboxyl groups) (Park et al. 2007b). An additional di-amino-poly(ethylene glycol) (di-NH2-PEG) was used as a linker molecule to add the functional amine groups to the PLLA nanoibers that were previously soaked in NaOH to increase the density of the carboxylic acid groups on the PLLA nanoiber membrane (Patel et al. 2007). Di-NH2-PEG molecules were then covalently attached to the carboxylic acid groups on the PLLA nanoibers using zero-length crosslinkers 1-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC) and N-hydroxysulfosuccinimide (sulfo-NHS). Subsequently, heparin molecules were covalently attached to the free amines on the di-NH2-PEG molecules via EDC and sulfo-NHS. Finally, an ECM protein (laminin) and a basic ibroblast growth factor (bFGF) were immobilized on a heparin-functionalized surface of PLLA nanoiber membrane, as shown in Figure 8.8. In another work, polyethersulfone (PES) nanoibers were surface-modiied with the carboxyl groups by UV-initiated PAA grafting as a linker molecule (Chua et al. 2006). Amino or hydroxyl groups were subsequently introduced to the nanoiber surface by reacting the ED or ethanolamine with the surface carboxylic acid groups using carbodiimide chemistry. An aminated PES

Membrane Modification: Technology and Applications

234

bFGF

Laminin

–

OH O

NHSO–3

OH

NHSO3 O

CH2SO–3

OH

Heparin

OH

OH O

O

O

CH2SO–3

C=O

OH O C=O

NH

NH

PEG

PEG

NH C=O

NH Nanofiber

C=O

FIGURE 8.8 The immobilization of the basic ibroblast growth factor (bFGF) and laminin on poly(l-lactic acid) (PLLA) nanoibers using di-amino-poly(ethylene glycol) (di-NH2-PEG) and heparin as linkers. (Reprinted from Patel, S., et al., Nano Lett., 7, 2122, 2007. With permission.)

nanoiber showed the highest expansion eficiency of human umbilical cord blood cells (hematopoietic stem/progenitor cells, HSPCs). In addition, aminated nanoibers also enhanced the adhesion of HSPCs. Polymer blends and modiied polymers can also be electrospun to produce surface-functionalized nanoibers. A PCL–PEG block copolymer was synthesized with the functional amine groups on the surface via the PEG linkers and electrospun into nanoibers that were further functionalized (Choi et al. 2008). A copolymer of methyl methacrylate (MMA) and AA was synthesized and electrospun into nanoibers with different carboxyl group contents by varying the ratio of MMA to AA (Li et al. 2008a). In another case, a blend mixture of a biodegradable PCL and a poly(d,l-lactic-co-glycolic acid)–poly(ethylene glycol)–NH 2 (PLGA–b–PEG– NH2) block copolymer was electrospun to produce surface-functionalized nanoibers with amino groups (Kim and Park 2006). The resulting nanoibrous membrane with the primary amine groups on the surface was applied for the immobilization of biologically active molecules using lysozyme as a model enzyme. The lysozyme was immobilized via covalent conjugation using a homobifunctional coupling agent. The PAN nanoiber membrane was modiied by hydroxylamine, resulting in the functionalization of the surfaces of the nanoibers to form polyacrylamidoxime (PAAO) (Chen et al. 2009). The nucleophilic amidoxime groups are capable of hydrolyzing esters and reacting with the organophosphate pesticides or chemical warfare agents as DFP. The hydrolytic degradation of DFP occurs only in the presence of free water, which serves as a medium to promote the nucleophilic action of the amidoxime groups in the nanoibers by facilitating proton transfer and stabilizing the transition state. The surface modiication of the electrospun PAN nanoiber

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 235

membrane has the advantages of a higher surface density of oxime functional groups and an enhanced reactivity compared with the PAN/PAAO blending technique (Chen et al. 2007b).

8.2.5

SURFACE GRAFT COPOLYMERIZATION

Surface graft copolymerization is an easy and controllable introduction of graft chains to the surfaces of the nanoibrous membranes without changing the bulk properties. This technique can be employed to modify the polymer surface to obtain distinct properties through the choice of different monomers (Mori et al. 1994; Ma et al. 2005c; Yao et al. 2008). Surface graft copolymerization is used to confer surface hydrophilicity and also to introduce multifunctional groups on nanoibrous surfaces for the covalent immobilization of the bioactive molecules. The surface graft copolymerization can be initiated via direct chemical modiication, ozone, γ-rays, electron beams, plasma discharge, or UV radiation treatment to generate free radicals for the polymerization (Ikada 1992). Radiation-induced graft copolymerization and plasma-induced graft copolymerization are the two main approaches to the graft copolymerization of nanoibers. Figure 8.9 illustrates, as an example, the graft copolymerization (radiation-induced or plasma-induced) on a nanoiber surface that was later used to immobilize proteins, drugs, enzymes, or growth factor for further applications. 8.2.5.1 Radiation-Induced Graft Copolymerization Radiation-induced graft copolymerization is the irradiation of the polymer surfaces with a high-energy source, permitting monomer (or monomers) grafting onto the surface. Radiation can generally be classiied into two categories: highenergy radiation and low-energy radiation. Absorption of high-energy radiation by the polymers induces excitation and ionization. The excited and ionized species are the initial chemical reactants for graft copolymerization. Among the lowenergy radiation, UV radiation is the most commonly used. UV radiation interacts with the polymer by a mechanism of excitation of its atoms and molecules. It can

Induced by plasma or radiation

Monomer Electrospun nanofibers

Surf ce-graft polymerization

Immobilization of protein, enzyme, growth factor, drug

Biologically or therapeutically functionalized nanofibers

FIGURE 8.9 A schematic representation of graft copolymerization (radiation-induced or plasma-induced) on an electrospun nanoiber surface. (Adapted from Yoo, H.S., et al., Adv. Drug Deliv. Rev., 61, 1033, 2009.)

236

Membrane Modification: Technology and Applications

be used successfully for nanoiber surface modiication as an initiator for graft copolymerization. UV-induced graft copolymerization can cause cross-linking and grafting on the surface of the nanoiber membrane. On the other hand, it may cause chain scission and may even destroy the polymer nanoibers easily when the polymer is biodegradable. A moderate number of cross-links can often improve the physical properties of the polymers, whereas the scission processes usually produce damage effects, resulting in soft and weak nanoiber membranes. Therefore, the reaction condition for the surface modiication of the polymer nanoibers must be optimized to preserve the nanoibrous morphology. The degradation of the nanoibers also depends on the types of polymers used. Aromatic groups, such as polyimides and polyetherimides, are more resistant to irradiation compared with other groups, such as C–Cl in PVC or C–O–C linkage in polyoxymethylene. Polytetraluoroethylene (PTFE) and other luorinated polymers, which have excellent resistance to chemical attack and high temperature, are unfortunately the least radiation-resistant polymers (Gopal et al. 2007). Chua et al. (2005) used radiation-induced graft copolymerization to develop a biofunctional poly(ɛ-caprolactone-co-ethyl ethylene phosphate) (PCLEEP) nanoiber for hepatocyte culture. PAA was grafted onto the nanoiber scaffold via UV-induced copolymerization using AA monomers. Then, galactose ligands (AHG) conjugation onto a PAA-grafted PCLEEP nanoiber membrane was achieved using sulfo-NHS and EDC. The surface-conjugated galactose ligands promoted the formation of hepatocyte spheroids around the nanoibers, resulting in an integrated spheroid-nanoiber construct in the nanoiber membrane. In another study (Chua et al. 2006), PAA was grafted onto a PES nanoiber mesh surface by UV radiation. The PAA-grafted PES nanoiber mesh and the ilms were further conjugated with ethylene diamine (EtDA) using a two-step carbodiimide cross-linking method. The aminated nanoiber mesh could further enhance the adhesion of HSPCs for allogeneic and autologous hematopoietic stem cell transplantation. A combination of argon plasma pretreatment, UV-induced surface graft copolymerization with poly(4-vinyl-N-alkylpyridinium bromide) (4VP), and quaternization reaction of the grafted pyridine groups with hexylbromide was used for the surface modiication of the poly(vinylideneluoride-hexaluoropropylene) (PVDF-HFP) nanoibrous membranes for antibacterial activities (Yao et al. 2009). PVDF-HFP nanoiber membranes were subjected to argon plasma pretreatment and were subsequently immersed in the 4VP solution and exposed to UV irradiation. The nanoiber membranes with graft-copolymerized 4VP were then placed in a heptane–hexylbromide solution. The surface-modiied PVDF-HFP nanoiber exhibited very high antibacterial eficiency against both Gram-positive and Gram-negative bacteria. Broad-energy (nonmonochromatic) ion beam implantation is often used in polymer modiication because, compared with the monochromatic beam, it provides a wider range of ion doping depths and, therefore, distributes energy over a greater depth range and minimizes thermal damage. Recently, ion implantation with a broad energy range ion beam (N+ and He+ beams) was used as a postprocessing treatment technique to modify the elastic modulus as well as the chemical structure of the electrospun PVA nanoibers without damaging the surface morphological features

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 237

(Wong et al. 2009). Wong et al. (2009) converted a high-energy monoenergetic ion beam, generated from a conventional ion implanter, to a low polyenergetic beam using a tantalum (Ta) foil as an energy diffuser. The energy of the beam was reduced and broadened through the collision/straggling process with the Ta atoms. Generally, the ion implantation process recoils the hydrogen atoms at a rate higher than that of the carbon and oxygen atoms from both the hydroxyl groups and the carbon backbone of the PVA molecules. An increase of 30% in the elastic modulus was observed after ion beam treatment, while the iber diameter decreased to 11%. As the ion– polymer interaction induces an irreversible chain scission and cross-linking effect in the polymer matrix, new amine and amide functional groups are created using the N+ beam, which could improve the cell compatibility of the PVA nanoibers. Jeon et al. (2008) fabricated thermoresponsive nanoibrous surfaces using electron beam irradiation. Poly(3-hydroxybutyrate-co-3-hydroxyvalerate) (PHBV) nanoibers were modiied by grafting poly(N-isopropylacrylamide) (PIPAAm) using electron beam irradiation. The results showed that PIPAAm surfaces, which are hydrophobic at a higher temperature, became markedly more hydrophilic in response to a temperature reduction due to the spontaneous hydration of the surface-grafted PIPAAm. 8.2.5.2 Plasma-Induced Graft Copolymerization Plasma-induced graft copolymerization is an eficient and versatile way to introduce a selective polymeric layer (usually a monomer) on the surface of a hydrophobic membrane. Plasma-induced graft copolymerization treatment is limited to the surface, and hence, the bulk properties of the membranes are still maintained. The thickness of the modiied layer can be controlled up to the angstrom level. In addition, it is a powerful technique for transforming a membrane with a symmetrical structure to an asymmetrical structure, which increases selectivity without signiicantly increasing the hydrodynamic resistance. Kaur et al. (2007) studied the effect of the plasma exposure time on the grafting of the electrospun poly(vinylidene) luoride (PVDF) nanoiber membrane and a commercial hydrophobic PVDF (HVHP, Millipore) membrane. The surface of the PVDF nanoiber was exposed to argon plasma and subsequently graft-polymerized with a hydrophilic monomer methacrylic acid (MAA) to develop an asymmetric membrane. The degree of grafting increased steeply with an increase in the plasma exposure time of the grafted-PVDF nanoiber membrane, whereas the increase in the grafting density on the surface of the HVHP membrane was not strong, reaching a plateau after 60 sec. Water iltration studies suggested that the grafted-PVDF nanoiber membrane can be successfully fabricated by surface modiication to decrease the pore size and increase the permeation lux. The hydrophilicity and the protein fouling resistance of the PVDF-HFP nanoibrous membranes were improved by plasma-induced graft copolymerization of the hydrophilic monomers (Sun et al. 2010). The surface modiication involves an atmospheric pressure glow discharge plasma (APGDP) pretreatment, followed by graft copolymerization of poly(ethylene glycol) methyl ether methacrylate (PEGMA). The grafting of gelatin onto nonwoven poly(caprolactone) nanoibers (PCL NF) and aligned PCL nanoibers (APCL NF) was done to improve the spreading and proliferation of the ECs on the nanoibers (Ma et al. 2005d). To graft gelatin onto the nanoiber surface, the PCL nanoibers were irst treated with air plasma to introduce

238

Membrane Modification: Technology and Applications

–COOH groups on the nanoibrous surface, followed by covalent grafting of the gelatin molecules, using water-soluble carbodiimide as a coupling agent. It is interesting to note that gelatin-grafted APCL NF readily oriented ECs along the ibers, whereas unmodiied APCL NF did not. The electrospun PGA, PLLA, and PLGA nanoibers were chemically modiied using oxygen plasma treatment and in situ grafting of the hydrophilic AA (Park et al. 2007b). The surface properties of the AA-grafted nanoiber membrane revealed higher ratios of oxygen to carbon, lower contact angles, and the presence of carboxylic (–COOH) groups. Fibroblast proliferation was found to be much better on the AA-grafted surface as compared with the unmodiied PGA, PLLA, and PLGA nanoiber membranes. Polyurethane (PU) nanoiber surfaces were also modiied by grafting AA with plasma treatment, and the modiied nanoiber membranes were employed in a poly(dimethylsiloxane)-based microluidic chip (Lee et al. 2009b). After modiication, a terminal carboxyl group formed on the nanoiber surface and the wettability increased signiicantly. 8.2.5.3 Ce(IV)-Induced Graft Copolymerization The cerium (IV) ion is a powerful oxidation agent for alcohols containing the 1,2-glycol groups. The mechanism of ceric ion reaction involves a redox system of ceric ion (Ce4+) as an oxidant and organic reductant that decomposes to generate free radicals. In the presence of glycidyl methacrylate (GMA) and MAA monomers, these free radicals will generate graft polymerization on the surface of the ibers (Mishra et al. 2003). Ma et al. (2005c) prepared poly(ethylene terephthalate) (PET) nanoiber membranes by electrospinning and modiied their surfaces to mimic the ibrous proteins in the native ECM and to design a biocompatible surface for the ECs. The electrospun PET nanoiber membranes were irst treated in formaldehyde to yield hydroxyl groups on the surface, followed by the grafting copolymerization of MAA initiated by Ce(IV). Finally, the poly(methacrylic acid) (PMAA)-grafted PET nanoiber membranes were grafted with gelatin using water-soluble carbodiimide as a coupling agent. The gelatin-grafting method could obviously improve the spreading and proliferation of the ECs on the PET nanoiber membranes and, moreover, could preserve the EC phenotype. In another study, the same researchers used Ce(IV)-induced graft copolymerization at the surface of polysulfone (PSU) nanoibers (Ma et al. 2006a). After electrospinning, the PSU iber mesh was heat-treated under 188°C to signiicantly improve the mechanical strength of the nanoiber membrane. For surface modiication, the carboxyl groups were introduced onto the PSU iber surfaces through graft copolymerization of MAA initiated by Ce(IV) after air plasma treatment of the PSU nanoiber membrane. Toluidine Blue O (TBO), a dye that is able to form a stable complex with the carboxyl groups, was used as a model target molecule to be immobilized by the PMAA-grafted PSU nanoiber membrane. This will be explained in the following section. Chen and Hsieh (2005) have grafted PAA onto ultraine cellulose nanoiber surfaces by either ceric ion (Ce(IV))– initiated copolymerization or methacrylation of the cellulose with methacrylate chloride (MACl) and subsequent free-radical polymerization of AA. The thickness of the PAA layer grafted by the ceric ion–initiated polymerization

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 239

increased with an increasing reaction time, monomer AA, and initiator concentrations. The thickness of the PAA layer grafted onto the methacrylated cellulose nanoibers also increased with the increasing molar ratios of MACl to cellulosic hydroxyl groups and monomer AA. The adsorbed lipase (immobilized protein) on the ceric ion–initiated grafted surface possessed a greatly improved organic solvent stability over the crude lipase.

8.2.6 MOLECULE IMMOBILIZATION ON NANOFIBER SURFACE Surface-active agents can be embedded in a nanoiber membrane by chemical functionalization, by postspinning modiication, by physical adsorption, or by nanoparticle polymer composites. 8.2.6.1 Physical Adsorption Physical surface adsorption is the simplest approach for loading target molecules onto nanoibrous meshes. Generally, electrostatic interaction, hydrogen bonding, hydrophobic interaction, or van der Waals interaction can be used as a driving force for surface adsorption. The eficiency of the physical adsorption on the hydrophobic nanoiber membranes can be increased by air plasma treatment to render them more hydrophilic and allow greater iltration of the aqueous solutions containing watersoluble biomolecules. Wang et al. (2006) immobilized the lipase enzyme from Candida rugosa by physical adsorption on the surface of polysulfone (PSF) composite nanoibrous membranes. PVP and PEG were used as additives, aiming to tailor the surface properties of the PS nanoibers, increasing their hydrophilicity. The results showed that the activity of the immobilized lipase increased with the content of PVP or PEG, whereas the adsorbed amount of lipase was unchanged. These results were attributed to the enrichment of PVP or PEG at the nanoiber surface. In another study, lipase from Pseudomonas cepacia was immobilized by physical adsorption onto electrospun PAN ibers and used for the conversion of (S)-glycidol with vinyl n-butyrate to glycidyl n-butyrate in isooctane (Sakai et al. 2010). Biologically active, functionalized electrospun membranes permit immobilization and long-term delivery of growth factors. Heparin, a highly sulfated glycosaminoglycan, has a strong binding afinity with various growth factors, such as basic bFGF, vascular endothelial growth factor (VEGF), heparin-binding epidermal growth factor (HBEGF), and transforming growth factor-β (TGF-β). This approach offers preservation of the biological activity by preventing early degradation of the growth factors. The heparin-functionalized electrospun nanoibers were prepared by Casper et al. (2005) using bFGF. In this study, PEG functionalized with low-molecular-weight heparin (PEG–LMWH) was fabricated by electrospinning using a carrier polymer, either PEO or PLGA. The incorporation of heparin into the electrospun PEO and PLGA nanoibers did not affect the surface morphology or the iber diameters. Improvements in the binding of bFGF to the electrospun ibers were also observed for ibers functionalized with PEG–LMWH over those functionalized with LMWH alone. Ahmad et al. (2010) reported a novel fabrication approach of a highly sensitive amperometric glucose biosensor based on a single ZnO nanoiber (ZONF).

240

Membrane Modification: Technology and Applications

ZONFs with diameters in the range of 350–195 nm were obtained by the calcination of electrospun PVP/zinc acetate composite nanoibers. A single ZONF on a gold electrode was functionalized with glucose oxidase (GOx) by physical adsorption. Electrochemical measurements revealed that the biosensor exhibited a good antiinterference ability and favorable stability during long-term storage. 8.2.6.2 Nanoparticles Assembly on the Surface of Nanofibers Many functionalized composite nanoibers were produced directly by electrospinning polymer solutions containing nanoparticles of different types, including metal alkoxide precursors. The incorporation of metal nanoparticles in polymer matrices has allowed the development of materials with unique properties due to the nanoscale size and shape of the dispersed nanoparticles. For example, the electrospun nanoibrous membranes, acrylonitrile and AA copolymers (PAN-AA) containing catalytic palladium (Pd) nanoparticles, were prepared by electrospinning from homogeneous solutions of PAN-AA and PdCl2, followed by reduction with hydrazine (Demir et al. 2004). Dodecanethiol-capped Au nanoparticles were mixed with PEO prior to electrospinning, and one-dimensional arrays of Au nanoparticles within the electrospun nanoibers were obtained using the intrinsic nature of semicrystalline PEO as a template to arrange the Au nanoparticles (Kim et al. 2005b). In another study, various metal oxide nanoparticles were deposited on PET with CA or cellulose nanoiber surfaces using LPD (Sundarrajan and Ramakrishna 2010). It was found that the nucleation density of the coated nanoparticles was higher on PET/CA and PET/cellulose-blended surfaces than the PET surface, due to the hydroxyl functional groups of the nanoiber surface. Jun et al. (2008) prepared polyaniline nanoibers with undoped ZnO nanoparticles or Ga-doped ZnO (ZnO:Ga) nanoparticles and studied their structural and electrical properties. Polyaniline nanoibers embedded with ZnO:Ga nanoparticles showed higher electrical conductivities compared with polyaniline nanoibers embedded with undoped ZnO nanoparticles. In another study, iron nanoparticles were immobilized on a PAA/PVA nanoiber membrane previously cross-linked by heat treatment (Xiao et al. 2009). The PAA/PVA nanoiber membranes were immersed in an aqueous solution of ferric trichloride to allow ferric cations to complex with the available free carboxyl groups on PAA through ion exchange. Subsequently, the sodium borohydride solution was gradually dropped onto the nanoiber membrane until no hydrogen gas was produced to reduce the ferric iron to zerovalent iron (ZVI). The ZVI nanoparticles with a mean diameter of 1.6 nm were uniformly distributed on the nanoiber membrane. The produced ZVI nanoparticles containing polymer nanoiber membrane exhibited a superior capability to decolorize the acid fuchsine solution, a model dye in the wastewater of the printing and dyeing industry. Nanomaterials based on silver nanoparticles have received signiicant attention from various authors in different research ields, such as biomaterials, medical devices, and electronics. Apart from their unique wound healing ability, silver nanoparticles exhibit antibacterial properties that make them potential candidates for use in infection-resistant biomaterials (Jang 2006). To prepare silver polymer nanoibers, silver nanoparticles are deposited onto the electrospun polymer nanoibers by sputter coating or polymeric solutions containing silver ions are electrospun.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 241

Kong and Jang (2006) fabricated silver nanoparticle–embedded poly(vinyl alcohol)– poly(methyl methacrylate) (PVA–PMMA) nanoibers by one-step radical-mediated dispersion polymerization using 2,29-azobis(isobutyronitrile) (AIBN), which played a role in initiating the radical polymerization, as well as reducing the silver ions. In this methodology, PVA played a role in producing the polymer nanoibers and in protecting the silver nanoparticles from aggregation since PVA acted as a stabilizing agent to prohibit the silver clusters from sintering. Figure 8.10 illustrates, as an example, the overall synthetic procedure for the fabrication of silver nanoparticle–embedded polymer nanoibers as well as TEM and SEM images. The organic radicals produced by the decomposition of AIBN were adsorbed onto the surfaces of

PVA

Ag+

Shear

Ag+/PVA

Polymerization

Ag/PVA-PMMA nanofibers

Ag/PVA/MMA

AIBN

MMA

Ag/PVA/AIBN

20 nm

1 µm

FIGURE 8.10 A schematic representation of the overall synthetic procedure for the fabrication of polymer nanoibers with embedded silver nanoparticles, including TEM and SEM images. (Adapted from Kong, H. and Jang, J., Chem. Comm., 28, 3010, 2006.)

242

Membrane Modification: Technology and Applications

the silver ion–PVA complexes, which were relatively hydrophobic compared with the hydrophilic aqueous medium. Consequently, the silver nanoparticles were embedded in the polymer nanoibers and assembled in one direction. Jin et al. (2005) prepared Ag–PVP nanoibers using two methods. In the irst method, the PVP nanoibers containing the Ag nanoparticles were directly prepared from PVP solutions, whereas in the second method, the PVA aqueous solutions were electrospun with the PVP containing the Ag nanoparticles. The Ag nanoparticles were distributed in the PVA nanoibers and the PVP containing the Ag nanoparticles could be used to introduce the Ag nanoparticles to other polymer nanoibers that are miscible with PVP. Son et al. (2004a) reported for the irst time that polymer nanoibers containing Ag nanoparticles on th eir surface could be produced by UV irradiation of the polymer electrospun nanoibers with small amounts of silver nitrate (AgNO3). A CA solution with small amounts of AgNO3 was electrospun, followed by UV radiation treatment. Ag nanoparticles in ultraine CA ibers (with an average size of 22 nm) were stabilized by interactions with the carbonyl oxygen atoms in CA. Ultraine CA nanoibers with Ag nanoparticles were found to exhibit a very strong antimicrobial activity. In another study, antimicrobial PVA nanoibers containing Ag nanoparticles were fabricated by electrospinning PVA/AgNO3 aqueous solutions, followed by either a short heat treatment at 155°C for 3 min or by UV irradiation (Hong 2007). After posttreatment, the PVA/AgNO3 nanoibers became insoluble in water; whereas the Ag+ ions therein were reduced to produce a large number of Ag nanoparticles situated preferentially on the surface of the electrospun nanoibers. It was noted that the average diameter of the Ag nanoparticles was 5.9 nm after the heat treatment and increased slightly to 6.3 nm after UV irradiation. Other surface modiications of nanoibrous membranes can be achieved by either surface mineralization or by incorporating bioactive ceramic components into the nanoiber membrane. Surface modiication using inorganic components is a promising approach, enhancing the dispersion in the polymers and the interaction and adhesion between the ceramic particles and the polymers. Composite nanoibers of PCL with calcium carbonate were seeded with osteoblasts (Fujihara et al. 2005). HAp was used as an inorganic material for the surface mineralization of the polymeric nanoiber membrane because it is the major inorganic component of bone and teeth (Ito et al. 2005). The incorporation of HAp on the surface layers of the electrospun biocompatible and biodegradable polyester poly(3-hydroxy butyrate valerate-co3-hydroxyvalerate) (PHBV) was carried out by soaking in a simulated body luid nearly equal to that of human blood plasma. The combination with HAp enhanced the hydrophilicity and thus accelerated the biodegradation rate of the nanoibrous ilm. The effects of the surface modiication of β-CaSiO3 on the electrospun poly(butylene succinate)/β-CaSiO3 composite ibers were investigated (Zhang and Chang 2009). Dodecyl alcohol can be esteriied on the surface of β-CaSiO3 to modify the bioceramics and prepare the composite ibrous materials with improved properties. Therefore, PBSU/β-CaSiO3 composite nanoiber membranes were fabricated by incorporating the surface-modiied and surface-unmodiied β-CaSiO3 into the polymer matrix.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 243

As inspired by many natural plants, such as lotus leaves and butterly wings, superhydrophobicity is generally deined for surfaces having water contact angles larger than 150° and the sliding angle smaller than 5°. The wettability of the liquid droplets on the solid surfaces is governed mainly by two parameters: chemical composition and geometrical microstructure. Recently, considerable efforts have been made to fabricate such superhydrophobic surfaces by combining these two parameters. The superhydrophobic nanoiber membranes are potentially applicable to iltration processes with high eficiency, enabling emerging applications such as self-cleaning surfaces, antifouling coatings, coatings for microluidic channels and biosensors, and smart materials and devices. Lim et al. (2007) developed a simple method to conine colloidal particles inside electrospun nanoibers and assembled the particles during iber thinning. Electrospun nanoibers with undulated surfaces at micrometer and nanometer scales were prepared with water-soluble polymers (poly(acryl amide) [PAM] or PEO) and different colloids, including monodisperse silica or polystyrene microspheres for larger particles and monodisperse silica nanoparticles for smaller particles. A schematic diagram of the procedure and SEM images of the undulated surface nanoibers are shown in Figure 8.11. Selective removal of the organic materials by calcination and subsequent treatment with a luorinated silane coupling agent by vapor-phase reactions enabled the creation of superhydrophobic surfaces with an extremely low sliding angle. 8.2.6.3 Covalent Bonding or Chemical Immobilization Covalent immobilization of active molecules on a polymer surface ensures the longterm chemical stability of introduced chains in contrast to what occurs in physically coated polymer chains and enhances cell adhesion and cell proliferation (Mori et al. 1994; Ma et al. 2005c; Yao et al. 2008). Surface modiication has long been recognized as a potential tool for enhancing biocompatibility. This is because the hydrophilic and protein-containing surfaces are known to promote cellular growth,

Fiber thinning by electrospinning

Fiber thinning by electrospinning

Calcination at 500°C to Superhydrophobicity remove polymer matrix (Contact angle >150°)

Fluorination

Calcination at 500°C to remove polymer matrix and large colloid

Fluorination

FIGURE 8.11 A schematic representation of the procedure for multiple-scale ibrous structures from electrospinning including SEM images of the undulated surface nanoibers. (Reprinted from Lim, J.M., et al., Langmuir 23, 7981, 2007. With permission.)

244

Membrane Modification: Technology and Applications

and many researches have focused on immobilizing biomolecules, such as collagen, gelatin, laminin, chitosan, Arg-Gly-Asp (RGD) containing peptide, lysozyme, and oligopeptide, onto the surfaces of the polymeric nanoibers to improve their cytocompatibility (Wang et al. 2009). Covalent bonding may require the activation of functional groups on the polymer surface or on the biomolecules or on both. Carboxyl groups can be activated to react with primary amines by treating with EDC as the coupling agent (Ma et al. 2005d; Duan et al. 2007). The eficiency of EDC-initiated bonding can be improved by adding NHS, which converts an unstable amine-reactive intermediate formed by the EDC reaction into an amine-reactive NHS ester (Casper et al. 2007). Kim and Park (2006) modiied an electrospun poly(lactic acid-co-glycolic acid)-block-poly(ethylene glycol) nanoiber membrane by covalently attaching a lysozyme using a homobifunctional coupling agent (ethyleneglycol-bis(succinimidylsuccinate)). These surfacemodiied nanoiber membranes could immobilize a far greater amount of lysozyme on their surface and exhibited greater catalytic activities to those of the solvent casting ilm, mainly as a result of their enlarged surface area. Sun et al. (2007) demonstrated that polar oligopeptide sequences that were covalently bound to a host polymer migrated to the nanoiber surface during electrospinning. Similarly, self-complementary hydrogen-bonding groups were observed to migrate to the iber surface of the PMMA copolymers (McKee et al. 2004). Jia et al. (2002) prepared polystyrene synthesized nanoibers with reactive surfaces comprised of a hydroxyl-containing initiator. Then, α-chymotrypsin was covalently attached to these surfaces. In another study, seed enzyme molecules were covalently attached to the nanoibers electrospun from a mixture of polystyrene and poly(styrene-co-maleic anhydride) (Kim et al. 2005a). Deitzel et al. (2002) investigated the surface segregation of luorine in the electrospun ibers of random copolymers of MMA and tetrahydroperluorooctyl acrylate (PMMA-r-TAN). The atomic luorine contents at the beaded iber surface were two to three times higher than in the bulk and were comparable to the surface luorine content in the solution cast ilms of the same copolymers. A partially perluorinated hyperbranched polyethyleneimine additive (PFA) was observed to selectively segregate to the surface of the electrospun PMMA nanoibers and this depended primarily on the additive concentration (Hunley et al. 2008). The presence of the additive on the iber surface allowed surface functionalization with the Ag nanoparticles, promoting further surface migration of the hyperbranched surface-active additives during an electrospinning process. Nanoibrous sugar sticks with linear poly[acrylonitrile-co-(d-gluconamidoethyl meth-acrylate)] (PANCGAMA) and cyclic poly[acrylonitrile-co-(α-allyl glucoside)] (PANCAG) glucose pendants were fabricated by electrospinning solutions of acrylonitrile-based glycopolymers (Yang et al. 2006). Cyclic glucose pendants on the PANCAG nanoibrous “sugar sticks” demonstrated great afinity for separating the concanavalin A (Con A) protein from Con A/BSA protein mixtures. However, the PANCGAMA nanoibers showed almost no selectivity for these two proteins due to the poor speciicity between the linear glucose pendants and Con A. An interesting phenomenon is that the PSU nanoibers become curled after PMAA grafting by Ce(IV) and the protein ligands such as Cibacron blue F3GA (CB) can be covalently immobilized on the PMAA-grafted PSU nanoiber membrane

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 245

surface using the carboxyl (–COOH) groups as coupling sites (Ma et al. 2006b). By using carbodiimide as a coupling agent, the carboxyl groups were reacted with diamino-dipropylamine (DADPA) to form amino groups. Finally, CB was covalently coupled to the PSU–PMAA–DADPA nanoiber membrane through a reaction between its triazine chloride and the primary amino groups (–NH2) on the PSU– PMAA–DADPA nanoiber membrane surface under alkaline conditions. The CB immobilized nanoiber membrane showed the ability to speciically capture BSA with a capturing capacity of 22 mg/g, as well as good mechanical strength and good reusability. Filtration analysis showed that these modiied electrospun membranes exhibited lower pressure drop and higher permeate lux compared with conventional microiltration membranes and, consequently, can be used for afinity separation. A poly(acrylonitrile-co-2-hydroxyethyl methacrylate) (PANCHEMA) nanoiber membrane was used as the support for the covalent immobilization of lipase from C. rugosa. The hydroxyl groups were activated on the nanoiber membrane surface with epichlorohydrin, cyanuric chloride, or p-benzoquinone (Huang et al. 2008). The activity retention for the immobilized lipase varied between 32.5% and 40.6%, depending on the method used for activating the hydroxyl groups. To maximize the enzyme activity, the support must provide a biocompatibility and/or inert environment for enzyme protein. A biofriendly interface on the support surface for enzyme immobilization may reduce some nonbiospeciic enzyme–support interaction and protein denaturalization, creating a speciic microenvironment for the enzyme, thereby beneiting the enzyme activity (Huang et al. 2006). In this sense, Mattanavee et al. (2009) proposed the modiication of the PCL nanoiber membrane with 1,6-hexamethylenediamine to introduce amino groups on its surface in order to render PCL more suitable for tissue engineering. Various biomolecules, including collagen, chitosan, and the Gly-Arg-Gly-Asp-Ser (GRGDS) peptide, were immobilized on the PCL nanoiber surface, with N,N′-disuccinimidylcarbonate as a coupling agent. Two biomacromolecules, collagen and protein hydrolysate from egg skin, were tethered on the PANCAA nanoiber membrane in the presence of 1-ethyl3-(dimethyl-aminopropyl) carbodiamine (EDC)/N-hydroxysuccinimide (NHS) to create a biofriendly microenvironment for enzyme immobilization (Huang et al. 2009). The lipase from C. rugosa was then immobilized on the protein-modiied nanoiber membrane by covalent binding using glutaraldehyde as the coupling agent and on the nascent PANCAA nanoiber membrane using EDC/NHS as the coupling agent. The enhancement of both the activity retention and the stabilities, such as thermal stability, reusability, and storage stability, of the immobilized lipases was found on egg skin hydrolysate-modiied and collagen-modiied nanoibrous membranes compared with that on the nascent PANCAA. In another study, an adhesive protein, ibronectin, was grafted onto the PLLC nanoiber surface via a two-step reaction (Zhu et al. 2007). Polyester aminolysis followed ibronectin coupling via glutaraldehyde. Porcine esophageal epithelial cells were seeded on the ibronectingrafted nanoiber to test the cell growth promotion against the control unmodiied PLLC nanoiber. The results showed that ibronectin grafted onto the PLLC nanoiber greatly promotes epithelium regeneration. Enzyme loading on a hydrophobic support is often limited because of the low accessibility of the enzymes to the support surface. Hydrophobic polystyrene/

Membrane Modification: Technology and Applications

246

polystyrene-containing maleic anhydride (PS/PSMA) nanoibers were treated with an aqueous alcohol solution. The tightly aggregated nanoibers could be dispersed in water to form a loosely entangled structure, which was stable enough for enzyme immobilization (Nair et al. 2007). Combinational techniques using both covalent bonding and physical absorption can also be used to biofunctionalize the polymer nanoiber surfaces. The physical absorption of HAp coatings with simulated body luid can be made more eficient if the carboxyl groups are introduced by NaOH prior to incubation (Chen et al. 2006a; Yu et al. 2009). Growth factors, such as basic ibroblast growth factor (bFGF or FGF2), VEGF, hepatocyte growth factor (HGF), and ECM, can be bound to polymer nanoibers using a combinational technique where molecules such as heparin are covalently bonded to act as reservoirs that stabilize, store, and protect growth factors introduced by physical absorption (Casper et al. 2007; Patel et al. 2007). This technique allows for the strong and stable covalent binding of heparin to the nanoiber surface without exposing the growth factors to harsh conditions that could reduce their biofunctionality.

8.3

CONCLUSION

Electrospun polymer nanoibers are receiving extensive research interest for applications in diverse ields such as separation technology and biotechnology because of their important and interesting characteristics, which are a very large surface area to volume ratio, excellent mechanical properties, and a highly open porous structure. The surface chemical and physical properties of the nanoiber membrane play a crucial role in every speciic application. However, most of the polymer nanoibers do not possess the required speciic functional groups and therefore must be functionalized for successful applications. Various surface modiication techniques have been applied to render nanoiber membranes suitable as afinity iltration membranes, tissue engineering scaffolds, drug delivery carriers, biosensors/chemosensors, and protective air iltration cloth. The methods used to modify the surface of the polymer nanoibers usually depend strongly on the nature of the iber-forming polymer. Because of their size, nanoibers require less harsh reaction conditions and controlled reactions to prevent their morphology from being destroyed. In order to develop electrospun nanoibers as useful nanobiomaterials, the surfaces of the electrospun nanoibers must be chemically functionalized to achieve sustained delivery through the physical adsorption of diverse molecules. The surface modiication techniques of nanoibers include plasma treatment, wet chemical methods, surface graft polymerization, LbL assembly, and co-electrospinning of the surface-active agents and polymers. A variety of bioactive molecules, including anticancer drugs, enzymes, cytokines, and polysaccharides as well as nanoparticles, have been entrapped within the interior or physically immobilized on the surfaces of the nanoibers. The surfaces of the electrospun nanoibers were also chemically modiied by immobilizing the cell-speciic bioactive ligands to enhance cell adhesion and cell proliferation by mimicking the morphology and biological function of ECM.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 247

REFERENCES Agarwal, S., Wendorff, J.H. and Greiner, A. 2008. Use of electrospinning technique for biomedical applications. Polymer 49: 5603–5621. Agarwal, S., Wendorff, J.H. and Greiner, A. 2010. Chemistry on electrospun polymeric nanoibers: Merely routine chemistry or a real challenge? Macromol. Rapid Commun. 31: 1317–1331. Ahmad, M., Pan, C., Luo, Z. and Zhu, J. 2010. A single ZnO nanoiber-based highly sensitive amperometric glucose biosensor. J. Phys. Chem. C 114: 9308–9313. Alessandrino, A., Marelli, B., Arosio, C., Fare, S., Tanzi, M.C. and Freddi, G. 2008. Electrospun silk ibroin mats for tissue engineering. Eng. Life Sci. 8: 219–225. Araujo, J.V., Martins, A., Leonor, I.B., Pinho, E.D., Reis, R.L. and Neves, N.M. 2008. Surface controlled biomimetic coating of polycaprolactone nanoiber meshes to be used as bone extracellular matrix analogues. J. Biomater. Sci. Polym. Ed. 19: 1261–1278. Baek, H.S., Park, Y.H., Ki, C.S., Park, J.C. and Rah, D.K. 2008. Enhanced chondrogenic responses of articular chondrocytes onto porous silk ibroin scaffolds treated with microwave-induced argon plasma. Surf. Coat. Technol. 202: 5794–5797. Baji, A., Mai, Y.W., Wong, S.C., Abtahi, M. and Chen, P. 2010. Electrospinning of polymer nanoibers: Effects on oriented morphology, structures and tensile properties. Composites Sci. Technol. 70: 703–718. Beachley, V. and Wen, X. 2010. Polymer nanoibrous structures: Fabrication, biofunctionalization, and cell interactions. Prog. Polym. Sci. 35: 868–892. Bellan, L.M., Cross, J.D., Strychalski, E.A., Moren-Marbel, J. and Craighead, H.G. 2006. Individually resolved DNA molecules stretched and embedded in electrospun polymer nanoibers. Nano Lett. 6: 2526–2530. Bergshoef, M.M. and Vancso, G.J. 1999. Transparent nanocomposites with ultrathin, electrospun nylon-4,6 iber reinforcement. Adv. Mater. 11: 1362–1365. Bhardwaj, N. and Kundu, S.C. 2010. Electrospinning: A fascinating iber fabrication technique. Biotechnol. Adv. 28: 325–347. Bhattarai, N., Edmondson, D., Veiseh, O., Matsen, F.A. and Zhang, M. 2005. Electrospun chitosan-based nanoibers and their cellular compatibility. Biomaterials 26: 6176–6184. Buchko, C.J., Chen, L.C., Shen, Y. and Martin, D.C. 1999. Processing and microstructural characterization of porous biocompatible protein polymer thin ilms. Polymer 40: 7397–7407. Casper, C.L., Stephens, J.S., Tassi, N.G., Chase, D.B. and Rabolt, J.F. 2004. Controlling surface morphology of electrospun polystyrene ibers: Effect of humidity and molecular weight in the electrospinning process. Macromolecules 37: 573–578. Casper, C.L., Yamaguchi, N., Kiick, K.L. and Rabolt, J.F. 2005. Functionalizing electrospun ibers with biologically relevant macromolecules. Biomacromolecules 6: 1998–2007. Casper, C.L., Yang, W., Farach-Carson, M.C. and Rabolt, J.F. 2007. Coating electrospun collagen and gelatin ibers with perlecan domain I for increased growth factor binding. Biomacromolecules 8: 1116–1123. Charernsriwilaiwat, N., Opanasopit, P., Rojanarata, T., Ngawhirunpat, T. and Supaphol, P. 2010. Preparation and characterization of chitosan-hydroxybenzotriazole/polyvinyl alcohol blend nanoibers by the electrospinning technique. Carbohydr. Polym. 81: 675–680. Che, A.F., Huang, X.J., Wang, Z.G. and Xu, Z.K. 2008. Preparation and surface modiication of poly(acrylonitrile-co-acrylic acid) electrospun nanoibrous membranes. Aust. J. Chem. 61: 446–454. Chen, F., Lee, C.N. and Teoh, S.H. 2007a. Nanoibrous modiication on ultra-thin poly(epsiloncaprolactone) membrane via electrospinning. Mater. Sci. Eng. C 27: 325–332.

248

Membrane Modification: Technology and Applications

Chen, H. and Hsieh, Y.L. 2005. Enzyme immobilization on ultraine cellulose ibers via poly(acrylic acid) electrolyte grafts. Biotechnol. Bioeng. 90: 405–413. Chen, J., Chu, B. and Hsiao, B.S. 2006a. Mineralization of hydroxyapatite in electrospun nanoibrous poly(l-lactic acid) scaffolds. J. Biomed. Mater. Res. A 79: 307–317. Chen, L., Bromberg, L., Hatton, T.A. and Rutledge, G.C. 2007b. Catalytic hydrolysis of p-nitrophenyl acetate by electrospun polyacrylamidoxime nanoibers. Polymer 48: 4675–4682. Chen, L., Bromberg, L., Schreuder-Gibson, H., Walker, J., Hatton, T.A. and Rutledge, G.C. 2009. Chemical protection fabrics via surface oximation of electrospun polyacrylonitrile iber mats. J. Mater. Chem. 19: 2432–2438. Chen, L., Bromberg, L., Lee, J.A., et al. 2010. Multifunctional electrospun fabrics via layerby-layer electrostatic assembly for chemical and biological protection. Chem. Mater. 22: 1429–1436. Chen, M., Zamora, P.O., Som, P., Pena, L.A. and Osaki, S. 2003. Cell attachment and biocompatibility of olytetraluoroethylene (PTFE) treated with glow-discharge plasma of mixed ammonia and oxygen. J. Biomater. Sci. Polym. Ed. 14: 917–935. Chen, M., Patra, P.K., Warner, S.B. and Bhowmick, S. 2006b. Optimization of electrospinning process parameters for tissue engineering scaffolds. Biophys. Rev. Lett. 1: 153–178. Cheng, J.Y., Wei, C.W., Hsu, K.H. and Young, T.H. 2004. Direct-write laser micromachining and universal surface modiication of PMMA for device development. Sensor Actuator B Chem. 99: 186–196. Choi, S.W., Jo, S.M., Lee, W.S. and Kim, Y.R. 2003. An electrospun poly(vinylidene luoride) nanoibrous membrane and its battery applications. Adv. Mater. 15: 2027–2032. Choi, J.S., Leong, K.W. and Yoo, H.S. 2008. In vivo wound healing of diabetic ulcers using electrospun nanoibers immobilized with human epidermal growth factor (EGF). Biomaterials 29: 587–596. Chu, X.H., Shi, X.L., Feng, Z.Q., Gu, Z.Z. and Ding, Y.T. 2009. Chitosan nanoiber scaffold enhances hepatocyte adhesion and function. Biotechnol. Lett. 31: 347–352. Chua, K.N., Lim, W.S., Zhang, P. et al. 2005. Stable immobilization of rat hepatocyte spheroids on galactosylated nanoiber scaffold. Biomaterials 26: 2537–2547. Chua, K.N., Chai, C., Lee, P.C. et al. 2006. Surface-aminated electrospun nanoibers enhance adhesion and expansion of human umbilical cord blood hematopoietic stem/progenitor cells. Biomaterials 27: 6043–6051. Chuangchote, S., Sagawa, T. and Yoshikawa, S. 2008. Fabrication and optical properties of electrospun conductive polymer nanoibers from blended polymer solution. Jpn. J. Appl. Phys. 47: 787–793. Cooley, J.F. 1902. Apparatus for electrically dispersing luids. US patent 692631. Croll, T.I., O’Connor, A.J., Stevens, G.W. and Cooper-White, J.J. 2004. Controllable surface modiication of poly(lactic-co-glycolic acid) (PLGA) by hydrolysis or aminolysis I: Physical, chemical, and theoretical aspects. Biomacromolecules 5: 463–473. Deitzel, J.M., Kleinmeyer, J., Harris, D. and Tan, N.C.B. 2001a. The effect of processing variables on the morphology of electrospun nanoibres and textiles. Polymer 42: 261–272. Deitzel, J.M., Kleinmeyer, J., Hirvonen, J.K. and Beck, T.N.C. 2001b. Controlled deposition of electrospun poly(ethylene oxide) ibers. Polymer 42: 8163–8170. Deitzel, J.M., Kosik, W., McKnight, S.H., Beck Tan, N.C., DeSimone, J.M. and Crette, S. 2002. Electrospinning of polymer nanoibers with speciic surface chemistry. Polymer 43: 1025–1029. Delcorte, A., Bertrand, P., Wischerhoff, E. and Laschewsky, A. 1997. Adsorption of polyelectrolyte multilayers on polymer surfaces. Langmuir 13: 5125–5136. Demir, M.M., Gulgun, M.A., Menceloglu, Y.Z. et al. 2004. Palladium nanoparticles by electrospinning from poly(acrylonitrile-co-acrylic acid)− PDCL2 solutions. Relations between preparation conditions, particle size, and catalytic activity. Macromolecules 37: 1787–1792.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 249 Deng, B.Y., Yan, X., Wei, Q.F. and Gao, W.D. 2007. AFM characterization of nonwoven material functionalized by ZnO sputter coating. Mater. Charact. 58: 854–858. Ding, B., Gong, J., Kim, J. and Shiratori, S. 2005. Polyoxometalate nanotubes from layer-by-layer coating and thermal removal of electrospun nanoibres. Nanotechnology 16: 785–790. Ding, B., Kim, J., Fujimoto, K. and Shiratori, S. 2004. Electrospun nanoibrous polyelectrolyte membranes for advanced chemical sensors. Chem. Sensors 20: 264–265. Ding, B., Li, C.R., Fujita, S. and Shiratori, S. 2006. Layer-by-layer self-assembled tubular ilms containing polyoxometalate on electrospun nanoibers. Colloids Surf. A Phys. Eng. Aspect. 284: 257–262. Drew, C., Liu, X., Ziegler, D., Wang, X., Bruno, F.F., Whitten, J., Samuelson, L.A. and Kumar, J. 2003. Metal oxide-coated polymer nanoibers. Nano Lett. 3: 143–147. Duan, Y., Wang, Z., Yan, W., Wang, S., Zhang, S. and Jia, J. 2007. Preparation of collagencoated electrospun nanoibers by remote plasma treatment and their biological properties. J. Biomater. Sci. Polym. Ed. 18: 1153–1164. Fang, J., Niu, H., Lin, T. and Wang, X. 2008. Applications of electrospun nanoibers. Chin. Sci. Bull. 53: 2265–2286. Feng, C., Khulbe, K.C., Matsuura, T., Gopal, R., Kaur, S., Ramakrishna, S. and Khayet, M. 2008. Production of drinking water from saline water by air-gap membrane distillation using polyvinylidene luoride nanoiber membrane. J. Memb. Sci. 311: 1–6. Fong, H. and Reneker, D.H. 1999. Elastomeric nanoibers of styren–butadiene–styrene triblock copolymer. J. Polym. Sci. B Polym. Phys. 37: 3488–3493. Fong, H. and Reneker, D.H. 2001. Electrospinning and formation of nanoibers. In: Structure Formation in Polymeric Fibers, ed. D.R. Salem, pp. 225–246. Hanser: Munich. Formhals, A. 1934. Process and apparatus for preparing artiicial threads. US patent 1975504. Formhals, A. 1939. Method and apparatus for spinning. US patent 2160962. Formhals, A. 1940. Artiicial thread and method of producing same. US patent 2187306. Formhals, A. 1943. Apparatus for producing artiicial ibers from iber-forming liquids by an “electrical spinning” method. US patent 2323025. Formhals, A. 1944. Spinner for synthetic ibers. US patent 2349950. Frenot, A. and Chronakis, I.S. 2003. Polymer nanoibers assembled by electrospinning. Curr. Opin. Colloid Interface Sci. 8: 64–75. Fridrikh, S.V., Yu, J.H., Brenner, M.P. and Rutledge, G.C. 2003. Controlling the iber diameter during electrospinning. Phys. Rev. Lett. 90: 144502–144504. Fu, G.D., Xu, L.Q., Smart, F.Y., Li, G.L. and Kang, E.T. 2009. Nanoibers with a photoresponsive surface for controlled release. Appl. Mater. Interfaces 1: 2424–2427. Fujihara, K., Kotaki, M. and Ramakrishna, S. 2005. Guided bone regeneration membrane made of polycaprolactone/calcium carbonate composite nano-ibers. Biomaterials 26: 4139–4147. Ghasemi-Mobarakeh, L., Prabhakaran, M.P., Morshed, M., Nasr-Esfahani, M.H. and Ramakrishna, S. 2008. Electrospun poly(epsiloncaprolactone)/gelatin nanoibrous scaffolds for nerve tissue engineering. Biomaterials 29: 4532–4539. Gibson, P., Gibson, H.S. and Rivin, D. 2001. Transport properties of porous membranes based on electrospun nanoibers. Colloid Surf. A Physicochem. Eng. Aspect 187/188: 469–481. Gopal, R., Zuwei, M., Kaur, S. and Ramakrishna, S. 2007. Surface modiication and application of functionalized polymer nanoibers. Topics Appl. Phys. 109: 72–91. Gouma, P.I. 2003. Nanostructured polymorphic oxides for advanced chemosensors. Rev. Adv. Mater. Sci. 5: 147–154. Grace, J.M. and Gerenser, L.J. 2003. Plasma treatment of polymers. J. Dispers. Sci. Technol. 24: 305–341. Grimm, S., Giesa, R., Sklarek, K. et al. 2008. Nondestructive replication of self-ordered nanoporous alumina membranes via cross-linked polyacrylate nanoiber arrays. Nano Lett. 8: 1954–1959.

250

Membrane Modification: Technology and Applications

Guo, M., Ding, B., Li, X., Wang, X., Yu, J. and Wang, M. 2010. Amphiphobic nanoibrous silica mats with lexible and high-heat-resistant properties. J. Phys. Chem. C 114: 916–921. Haghi, A.K. and Akbari, M. 2007. Trends in electrospinning of natural nanoibers. Phys. Status Solidi. 204: 1830–1834. Hartgerink, J.D., Beniash, E. and Stupp, S.I. 2001. Self-assembly and mineralization of peptide-amphiphile nanoibers. Science 294: 1684–1688. He, W., Yong, T., Teo, W.E., Ma, Z.W. and Ramakrishna, S. 2005a. Fabrication and endothelialization of collagen-blended biodegradable polymer nanoibers: Potential vascular graft for blood vessel tissue engineering. Tissue Eng. 11: 1575–1588. He, W., Ma, Z.W., Yong, T., Teo, W.E. and Ramakrishna, S. 2005b. Fabrication of collagencoated biodegradable polymer nanoiber mesh and its potential for endothelial cells growth. Biomaterials 26: 7606–7615. He, W., Yong, T., Ma, Z.W., Inai, R., Teo, W.E. and Ramakrishna, S. 2006. Biodegradable polymer nanoiber mesh to maintain functions of endothelial cells. Tissue Eng. 12: 2457–2466. Henry, A.C., Tutt, T.J., Galloway, M. et al. 2000. Surface modiication of poly(methyl methacrylate) used in the fabrication of microanalytical devices. Anal. Chem. 72: 5331–5337. Hohman, M.M., Shin, M., Rutledge, G. and Brenner, M. 2001a. Electrospinning and electrically forced jets. I. Stability theory. Phys. Fluids 13: 2201–2220. Hohman, M.M., Shin, M., Rutledge, G. and Brenner, M. 2001b. Electrospinning and electrically forced jets. II. Applications. Phys. Fluids 13: 2221–2236. Hong, K.H. 2007. Preparation and properties of electrospun poly (vinyl alcohol)/silver iber web as wound dressings. Polym. Eng. Sci. 47: 43–49. Huang, X.J., Xu, Z.K., Wan, L.S., Innocent, C. and Seta, P. 2006. Electrospun nanoibers modiied with phospholipid moieties for enzyme immobilization. Macromol. Rapid Commun. 27: 1341–1345. Huang, X.J., Yu, A.G., Jiang, J., Pan, C., Qian, J.W. and Xu, Z.K. 2009. Surface modiication of nanoibrous poly(acrylonitrile-co-acrylic acid) membrane with biomacromolecules for lipase immobilization. J. Mol. Catal. B Enzym. 57: 250–256. Huang, X.J., Yu, A.G. and Xu, Z.K. 2008. Covalent immobilization of lipase from Candida rugosa onto poly(acrylonitrile-co-2-hydroxyethyl methacrylate) electrospun ibrous membranes for potential bioreactor application. Bioresour. Technol. 99: 5459–5465. Huang, Z.M., Zhang, Y.Z., Kotakic, M. and Ramakrishna, S. 2003. A review on polymer nanoibers by electrospinning and their applications in nanocomposites. Composites Sci. Technol. 63: 2223–2253. Hunley, M.T., Harber, A., Orlicki, J.A., Rawlett, A.M. and Long, T.E. 2008. Effect of hyperbranched surface-migrating additives on the electrospinning behavior of poly(methyl methacrylate). Langmuir 24: 654–657. Ikada, Y. 1992. Comparison of surface modiication of polymers by different methods. Radiat. Phys. Chem. 39: 509–511. Iskandar, F. 2009. Nanoparticle processing for optical applications – A review. Adv. Powder Technol. 20: 283–292. Ito, Y., Hasuda, H., Kamitakahara, M. et al. 2005. A composite of hydroxyapatite with electrospun biodegradable nanoibers as a tissue engineering material. J. Biosci. Bioeng. 100: 43–49. Jang, J. 2006. Conducting polymer nanomaterials and their applications. Adv. Polym. Sci. 199: 189–259. Jaworek, A. and Sobczyk, A.T. 2008. Electrospraying route to nanotechnology: An overview. J. Electrostatics 66: 197–219. Jeon, H.A., Oh, H.H., Kim, Y.J. et al. 2008. Fabrication and characterization of thermoresponsive nanoibrous surfaces using electron beam irradiation. Polymer-Korea 32: 359–365.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 251 Jeong, S.I., Lee, A.Y., Lee, Y.M. and Shin, H. 2008. Electrospun gelatin/poly(l-lactide-coepsilon-caprolactone) nanoibers for mechanically functional tissue-engineering scaffolds. J. Biomater. Sci. Polym. Ed. 19: 339–357. Jia, H.F., Zhu, G.Y., Vugrinovich, B., Kataphinan, W., Reneker, D.H. and Wang, P. 2002. Enzyme-carrying polymeric nanoibers prepared via electrospinning for use as unique biocatalysts. Biotechnol. Prog. 18: 1027–1032. Jiang, H., Fang, D., Hsiao, B., Chu, B. and Chen, W. 2004. Preparation and characterization of ibuprofen-loaded poly(lactide-co-glycolide)/poly(ethylene glycol)-g-chitosan electrospun membranes. J. Biomater. Sci. Polym. Ed. 15: 279–296. Jin, H.J., Chen, J., Karageorgiou, V., Altman, G.H. and Kaplan, D.L. 2004. Human bone marrow stromal cell responses on electrospun silk ibroin mats. Biomaterials 25(6): 1039–1047. Jin, W.J., Lee, H.K., Jeong, E.H., Park, W.H. and Youk, J.H. 2005. Preparation of polymer nanoibers containing silver nanoparticles by using poly (N-vinylpyrrolidone). Macromol. Rapid Commun. 26: 1903–1907. Jun, J.H., Cho, K., Yun, J., Suh, K.S., Kim, T.Y. and Kim, S. 2008. Enhancement of electrical characteristics of electrospun polyaniline nanoibers by embedding the nanoibers with Ga-doped ZnO nanoparticles. Org. Electron. 9: 445–451. Kato, K., Uchida, E., Kang, E.T., Uyama, Y. and Ikada, Y. 2003. Polymer surface with graft chains. Prog. Polym. Sci. 28: 209–259. Kaur, S., Kotaki, M., Ma, Z., Gopal, R., Ramakrishna, S. and Ng, S.C. 2006. Oligosaccharide functionalized nanoibrous membrane. Int. J. Nanosci. 5(1): 1–11. Kaur, S., Ma, Z., Gopal, R., Singh, G., Ramakrishna, S. and Matsuura, T. 2007. Plasma-induced graft copolymerization of poly(methacrylic acid) on electrospun poly(vinylidene luoride) nanoiber membrane. Langmuir 23: 13085–13092. Ki, C.S., Baek, D.H., Gang, K.D., Lee, K.H., Um, I.C. and Park, Y.H. 2005. Characterization of gelatin nanoiber prepared from gelatin-formic acid solution. Polymer 46: 5094–5102. Ki, C.S., Gang, E.H., Um, I.C. and Park, Y.H. 2007. Nanoibrous membrane of wool keratose/ silk ibroin blend for heavy metal ion adsorption. J. Memb. Sci. 302: 20–26. Kidoaki, S., Kwon, I.K. and Matsuda, T. 2005. Mesoscopic spatial designs of nano and microiber meshes for tissue-engineering matrix and scaffold based on newly devised multilayering and mixing electrospinning techniques. Biomaterials 26: 37–46. Kim, B.C., Nair, S., Kim, J. et al. 2005a. Preparation of biocatalytic nanoibres with high activity and stability via enzyme aggregate coating on polymer nanoibres. Nanotechnology 16: S382–388. Kim, C. and Yang, K.S. 2003. Electrochemical properties of carbon nanoiber web as an electrode for supercapacitor prepared by electrospinning. Appl. Phys. Lett. 83: 1216–1218. Kim, G.M., Wutzler, A., Radusch, H.J. et al. 2005b. One-dimensional arrangement of gold nanoparticles by electrospinning. Chem. Mater. 17: 4949–4957. Kim, H.W., Yu, H.S. and Lee, H.H. 2008. Nanoibrous matrices of poly(lactic acid) and gelatin polymeric blends for the improvement of cellular responses. J. Biomed. Mater. Res. A 87: 25–32. Kim, T.G. and Park, T.G. 2006. Surface functionalized electrospun biodegradable nanoibers for immobilization of bioactive molecules. Biotechnol. Prog. 22: 1108–1113. Kong, H. and Jang, J. 2006. One-step fabrication of silver nanoparticle embedded polymer nanoibers by radical-mediated dispersion polymerization. Chem. Comm. 28: 3010–3012. Koombhongse, S., Liu, W. and Reneker, D.H. 2001. Flat ribbons and other shapes by electrospinning. J. Polym. Sci. B Polym. Phys. 39: 2598–2606. Koski, A., Yim, K. and Shivkumar, S. 2004. Effect of molecular weight on ibrous PVA produced by electrospinning. Mater. Lett. 58: 493–497.

252

Membrane Modification: Technology and Applications

Krishnappa, R.V.N., Desai, K. and Sung, C. 2003. Morphological study of electrospun polycarbonates as a function of the solvent and processing voltage. J. Mater. Sci. 38: 2357–2365. Krogman, K.C., Lowery, J.L., Zacharia, N.S., Rutledge, G.C. and Hammond, P.T. 2009. Spraying asymmetry into functional membranes layer-by-layer. Nat. Mater. 8: 512–518. Kumar, N., Kaur, R. and Mehra, R.M. 2007. Photoluminescence studies in sol–gel derived ZnO ilms. J. Lumin. 126: 784–788. Ladizesky, N.H. and Ward, M. 1995. A review of plasma treatment and the clinical application of polyethylene ibers to reinforcement of acrylic resins. J. Mater. Sci. Mater. Med. 6: 497–504. Laforgue, A. 2011. All-textile lexible supercapacitors using electrospun poly(3,4-ethylenedioxythiophene) nanoibers. J. Power Sour. 196: 559–564. Lee, G.H., Song, J.C. and Yoon, K.B. 2010. Controlled wall thickness and porosity of polymeric hollow nanoibers by coaxial electrospinning. Macromolecular 18: 571–576. Lee, H. and Kim, G. 2010. Biocomposites electrospun with poly(epsilon-caprolactone) and silk ibroin powder for biomedical applications. J. Biomat. Sci. Polym. Ed. 21: 1687–1699. Lee, J.A., Krogman, K.C., Ma, M., Hill, R.M., Hammond, P.T. and Rutledge, G.C. 2009a. Highly reactive multilayer-assembled TiO2 coating on electrospun polymer nanoibers. Adv. Mater. 21: 1252–1256. Lee, K.H., Kwon, G.H., Shin, S.J. et al. 2009b. Hydrophilic electrospun polyurethane nanoiber matrices for hMSC culture in a microluidic cell chip. J. Biomed. Mater. Res. 90A: 619–628. Lee, S.J., Yoo, J.J., Lim, G.J., Atala, A. and Stitze, J. 2007. In vitro evaluation of electrospun nanoiber scaffolds for vascular graft application. J. Biomed. Mater. Res. A 83: 999–1008. Li, D. and Xia, Y. 2004. Electrospinning of nanoibers: Reinventing the wheel. Adv. Mater. 16: 1151–1170. Li, D., Frey, M.W. and Baeumner, A.J. 2006a. Electrospun polylactic acid nanoiber membranes as substrates for biosensor assemblies. J. Memb. Sci. 279: 354–363. Li, J., He, A., Zheng, J. and Han, C.C. 2006b. Gelatin and gelatin–hyaluronic acid nanoibrous membranes produced by electrospinning of their aqueous solutions. Biomacromolecules 7: 2243–2247. Li, M., Mondrinos, M.J., Gandhi, M.R., Ko, F.K., Weiss, A.S. and Lelkes, P.I. 2005. Electrospun protein ibers as matrices for tissue engineering. Biomaterials 26: 5999–6008. Li, W., Guo, Y., Wang, H. et al. 2008a. Electrospun nanoibers immobilized with collagen for neural stem cells culture. J. Mater. Sci. Mater. Med. 19: 847–854. Li, W.J., Laurencin, C.T., Caterson, E.J., Tuan, R.S. and Ko, F.K. 2002. Electrospun nanoibrous structure: A novel scaffold for tissue engineering. J. Biomed. Mater. Res. 60: 613–621. Li, X., Xie, J., Yuan, X. and Xia, Y. 2008b. Coating electrospun poly(ɛ-caprolactone) ibers with gelatin and calcium phosphate and their use as biomimetic scaffolds for bone tissue engineering. Langmuir 24: 14145–14150. Lim, J.M., Yi, G.R., Moon, J.H., Heo, C.J. and Yang, S.M. 2007. Superhydrophobic ilms of electrospun ibers with multiple-scale surface morphology. Langmuir 23: 7981–7989. Liu, S.J., Kau, Y.C., Chou, C., Chen, J.K., Wu, R.C. and Yeh, W.L. 2010. Electrospun PLGA/ collagen nanoibrous membrane as early-stage wound dressing. J. Memb. Sci. 355: 53–59. Lu, X., Zhou, J., Zhao, Y., Qiu, Y. and Li, J. 2008. Room temperature ionic liquid based polystyrene nanoibers with superhydrophobicity and conductivity produced by electrospinning. Chem. Mater. 20: 3420–3424. Luu, Y.K., Kim, K., Hsiao, B.S., Chu, B. and Hadjiargyrou, M. 2003. Development of a nanostructured DNA delivery scaffold via electrospinning of PLGA and PLA–PEG block copolymers. J. Control. Release 89: 341–353.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 253 Ma, K., Yong, T., Chan, K.C. and Ramakrishna, S. 2007a. Collagen-blended biodegradable polymer nanoibers: Potential substrates for wound healing in skin tissue engineering. In: Proceedings of the Fifth IASTED International Conference on Biomedical Engineering, pp. 262–266. Acta Press: Anaheim, CA. Ma, P.X. and Zhang, R. 1999. Synthetic nano-scale ibrous extracellular matrix. J. Biomed. Mater. Res. 46: 60–72. Ma, Z., Kotaki, M. and Ramakrishna, S. 2005a. Electrospun cellulose nanoiber as afinity membrane. J. Memb. Sci. 265: 115–123. Ma, Z., Kotaki, M., Inai, R. and Ramakrishna, S. 2005b. Potential of nanoiber matrix as tissue engineering scaffolds. Tissue Eng. 11: 101–109. Ma, Z., Kotaki, M., Yong, T., He, W. and Ramakrishna, S. 2005c. Surface engineering of electrospun polyethylene terephthalate (PET) nanoibers towards development of a new material for blood vessel engineering. Biomaterials 26: 2527–2536. Ma, Z.W., He, W., Yong, T. and Ramakrishna, S. 2005d. Grafting of gelatin on electrospun poly(caprolactone) nanoibers to improve endothelial cell spreading and proliferation and to control cell orientation. Tissue Eng. 11: 1149–1158. Ma, Z., Kotaki, M. and Ramakrishna, S. 2006a. Surface modiied nonwoven polysulphone (PSU) iber mesh by electrospinning: A novel afinity membrane. J. Memb. Sci. 272: 179–187. Ma, Z.W., Masaya, K. and Ramakrishna, S. 2006b. Immobilization of Cibacron blue F3GA on electrospun polysulphone ultra-ine iber surfaces towards developing an afinity membrane for albumin adsorption. J. Memb. Sci. 282: 237–244. Ma, Z., Mao, Z. and Gao, C. 2007b. Surface modiication and property analysis of biomedical polymers used for tissue engineering. Colloids Surf. B Biointerfaces 60: 137–157. Mark, S.S., Stolper, S.I., Baratti, C., Park, J.Y., Taku, M.A., Santiago-Avilés, J.J. and Kricka, L.J. 2008. Bioconjugation of alkaline phosphatase to mechanically processed, aqueous suspendible electrospun polymer nanoibers for use in chemiluminescent detection assays. Macromol. Biosci. 8: 484–498. Mattanavee, W., Suwantong, O., Puthong, S., Bunaprasert, T., Hoven, V.P. and Supaphol, P. 2009. Immobilization of biomolecules on the surface of electrospun polycaprolactone ibrous scaffolds for tissue engineering. Adv. Mater. Interface 1: 1076–1085. Matthews, J.A., Wnek, G.E., Simpson, D.G. and Bowlin, G.L. 2002. Electrospinning of collagen nanoibers. Biomacromolecules 3: 232–338. McKee, M.G., Elkins, C.L. and Long, T.E. 2004. Inluence of self-complementary hydrogen bonding on solution rheology/electrospinning relationships. Polymer 45: 8705–8715. Megelski, S., Stephens, J.S., Chase, D.B. and Rabolt, J.F. 2002. Micro- and nanostructured surface morphology on electrospun polymer ibers. Macromolecules 35: 8456–8466. Meinel, A.J., Kubowb, K.E., Klotzsch, E. et al. 2009. Optimization strategies for electrospun silk ibroin tissue engineering scaffolds. Biomaterials 30: 3058–3067. Miao, J.J., Miyauchi, M., Simmons, T.J., Dordick, J.S. and Linhardt, R.J. 2010. Electrospinning of nanomaterials and applications in electronic components and devices. J. Nanosci. Nanotechnol. 10: 5507–5519. Min, B.M., Lee, G., Kim, S.H., Nam, Y.S., Lee, T.S. and Park, W.H. 2004. Electrospinning of silk ibroin nanoibers and its effect on the adhesion and spreading of normal human keratinocytes and ibroblasts in vitro. Biomaterials 25: 1289–1297. Mishra, A., Srinivasan, R. and Gupta, R. 2003. P. psyllium-g-polyacrylonitrile: Synthesis and characterization. Colloid Polym. Sci. 281: 187–189. Mit-uppatham, C., Nithitanakul, M. and Supaphol, P. 2004. Ultraine electrospun polyamide-6 ibers: Effect of solution conditions on morphology and average iber diameter. Macromol. Chem. Phys. 205: 2327–2338. Mori, M., Uyama, Y. and Ikada, Y. 1994. Surface modiication of polyethylene iber by graft polymerization. J. Polym. Sci. Polym. Chem. 32: 1683–1690.

254

Membrane Modification: Technology and Applications

Morton, W.J. 1902. Method of dispersing luids. US patent 705691. Muller, K., Quinn, J.F., Johnston, A.P.R., Becker, M., Greiner, A. and Caruso, F. 2006. Polyelectrolyte functionalization of electrospun ibers. Chem. Mater. 18: 2397–2403. Nair, S., Kim, J., Crawford, B. and Kim, S.H. 2007. Improving biocatalytic activity of enzymeloaded nanoibers by dispersing entangled nanoiber structure. Biomacromolecules 8: 1266–1270. Nie, H., He, A., Jia, B., Wang, F., Jiang, Q. and Han, C.C. 2010. A novel carrier of radionuclide based on surface modiied poly-(lactide-co-glycolide) nanoibrous membrane. Polymer 51: 3344–3348. Ogawa, T., Ding, B., Sone, Y. and Shiratori, S. 2007. Super-hydrophobic surfaces of layer-bylayer structured ilm-coated electrospun nanoibrous membranes. Nanotechnology 18: 165607. Ohgo, K., Zhao, C., Kobayashi, M. and Asakura, T. 2003. Preparation of non-woven nanoibers for Bombyx mori silk, Samia cythia ricini silk and recombinant hybrid silk with electrospinning method. Polymer 44: 841–846. Park, H., Lee, K.Y., Lee, S.J., Park, K.E. and Park, W.H. 2007a. Plasma-treated poly(lacticcoglycolic acid) nanoibers for tissue engineering. Macromol. Res. 15: 238–243. Park, K., Ju, Y.M., Son, J.S., Ahn, K.D. and Han, D.K. 2007b. Surface modiication of biodegradable electrospun nanoiber scaffolds and their interaction with ibroblasts. J. Biomater. Sci. Polym. Ed. 18: 369–382. Patel, S., Kurpinski, K., Quigley, R. et al. 2007. Bioactive nanoibers: Synergistic effects of nanotopography and chemical signaling on cell guidance. Nano Lett. 7: 2122–2128. Piskin, E., Bolgen, N., Egri, S. and Isoglu, I. 2007. Electrospun matrices made of poly(alphahydroxy acids) for medical use. Nanomedicine 2: 441–457. Prabhakaran, M.P., Venugopal, J., Chan, C.K. and Ramakrishna, S. 2008. Surface modiied electrospun nanoibrous scaffolds for nerve tissue engineering. Nanotechnology 19: 455102. Ramakrishna, S., Fujihara, K., Teo, W.E., Lim, T.C. and Ma, Z. 2005. An Introduction to Electrospinning and Nanoibers. World Scientiic Publishing: Singapore. Ramakrishna, S., Fujihara, K., Teo, W.E., Yong, T., Ma, Z. and Ramaseshan, R. 2006. Electrospun nanoibers: Solving global issues. Mat. Today 9: 40–50. Ramakrishna, S., Jose, R., Archana, P.S. et al. 2010. Science and engineering of electrospun nanoibers for advances in clean energy, water iltration, and regenerative medicine. J. Mater. Sci. 45: 6283–6312. Reneker, D.H. and Chun, I. 1996. Nanometer diameter ibers of polymer produced by electrospinning. Nanotechnology 7: 216–223. Rollings, D.A., Tsoi, S., Sit, J.C. and Veinot, J.G. 2007. Formation and aqueous surface wettability of polysiloxane nanoibers prepared via surface initiated vapor-phase polymerization of organotrichlorosilanes. Langmuir 23: 5275–5278. Rollings, D.A.E. and Veinot, J.G.C. 2008. Polysiloxane nanoibers via surface initiated polymerization of vapor phase reagents: A mechanism of formation and variable wettability of iber-bearing substrates. Langmuir 24: 13653–13662. Rutledge, G.C. and Fridrikh, S.V. 2007. Formation of ibers by electrospinning. Adv. Drug Delivery Rev. 59: 1384–1391. Sakai, S., Liu, Y., Yamaguchi, T., Watanabe, R., Kawabe, M. and Kawakami, K. 2010. Immobilization of Pseudomonas cepacia lipase onto electrospun polyacrylonitrile ibers through physical adsorption and application to transesteriication in nonaqueous solvent. Biotechnol. Lett. 32: 1059–1062. Schreuder-Gibson, H.L. and Gibson, P. 2002. Protective textile based on electrospun nanoibers. J. Adv. Mater. 34: 44–55. Shao, D., Wei, Q., Zhang, L., Cai, Y. and Jiang, S. 2008. Surface functionalization of carbon nanoibers by sol–gel coating of zinc oxide. Appl. Surf. Sci. 254: 6543–6546.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 255 Shin, Y.M., Hohman, M.M., Brenner, M.P. and Rutledge, G.C. 2001a. Experimental characterization of electrospinning: The electrically forced jet and instabilities. Polymer 42: 9955–9967. Shin, Y.M., Hohman, M.M., Brenner, M.P. and Rutledge, G.C. 2001b. Electrospinning: A whipping luid jet generates submicron polymer ibers. Appl. Phys. Lett. 78: 1149–1151. Sill, T.J. and von Recum, H.A. 2008. Electrospinning: Applications in drug delivery and tissue engineering. Biomaterials 29: 1989–2006. Son, W.K., Youk, J.H., Lee, T.S. and Park, W.H. 2004a. Preparation of antimicrobial ultraine cellulose acetate ibers with silver nanoparticles. Macromol. Rapid Commun. 25: 1632–1637. Son, W.K., Youk, J.H., Lee, T.S. and Park, W.H. 2004b. The effects of solution properties and polyelectrolyte on electrospinning of ultraine poly (ethylene oxide) ibers. Polymer 45: 2959–2966. Son, W.K., Cho, D. and Park, W.H. 2006. Direct electrospinning of ultraine titania ibres in the absence of polymer additives and formation of pure anatase titania ibres at low temperature. Nanotechnology 17: 439–443. Sun, F.Q., Li, X.S., Xu, J.K. and Cao, P.T. 2010. Improving hydrophilicity and protein antifouling of electrospun poly(vinylideneluoride-hexaluoropropylene) nanoiber membranes. Chin. J. Polym. Sci. 25: 705–713. Sun, X.Y., Shankar, R., Borner, H.G., Ghosh, T.K. and Spontak, R.J. 2007. Field-driven biofunctionalization of polymer iber surfaces during electrospinning. Adv. Mater. 19: 87–91. Sundarrajan, S. and Ramakrishna, S. 2010. Fabrication of functionalized nanoiber membranes containing nanoparticles. J. Nanosci. Nanotechnol. 10: 1139–1147. Supaphol, P., Mit-uppatham, C. and Nithitanakul, M. 2005. Effects of solvent system and emitting electrode polarity on morphology and average iber diameter. Macromol. Mater. Eng. 290: 933–942. Tan, S.H., Inai, R., Kotaki, M. and Ramakrishna, S. 2005. Systematic parameter study for ultra-ine iber fabrication via electrospinning process. Polymer 46: 6128–6134. Tang, Z.Y., Wang, Y., Podsiadlo, P. and Kotov, N.A. 2006. Biomedical applications of layerby-layer assembly: From biomimetics to tissue engineering. Adv. Mater. 18: 3203–3224. Tao, S.L. and Desai, T.A. 2007. Aligned arrays of biodegradable poly(epsiloncaprolactone) nanowires and nanoibers by template synthesis. Nano Lett. 7: 1463–1468. Taylor, G.I. 1964. Disintegration of water drops in an electric ield. Proc. R. Soc. A 280: 383–397. Teo, W.E. and Ramakrishna, S. 2006. A review on electrospinning design and nanoibre assemblies. Nanotechnology 17: R89–R106. Theron, S.A., Zussman, E. and Yarin, A.L. 2004. Experimental investigation of the governing parameters in the electrospinning of polymer solutions. Polymer 45: 2017–2030. Thierry, B., Winnik, F.M., Merhi, Y., Silver, J. and Tabrizian, M. 2003. Bioactive coatings of endovascular stents based on polyelectrolyte multilayers. Biomacromolecules 4: 1564–1571. Tsai, P.P., Schreuder-Gibson, H.L. and Gibson, P. 2002. Different electrostatic methods for making electret ilters. J. Electrostatic 54: 333–341. Vasita, R., Mani, G., Agrawal, C.M. and Katti, D.S. 2010. Surface hydrophilization of electrospun PLGA micro-/nano-ibers by blending with Pluronic® F-108. Polymer 51: 3706–3714. Venugopal, J., Low, S., Choon, A.T., Kumar, A.B. and Ramakrishna, S. 2008. Electrospun modiied nanoibrous scaffolds for the mineralization of osteoblast cells. J. Biomed. Mater. Res. A 85A: 408–417. Villanueva, Y.Y., Liu, D.R. and Cheng, P.T. 2006. Pulsed laser deposition of zinc oxide. Thin Solid Films 501: 366–369.

256

Membrane Modification: Technology and Applications

Wan, Y.Q., Yang, J., Yang, J.L., Bei, J.Z. and Wang, S.G. 2003. Cell adhesion on gaseous plasma modiied poly(l-lactide) surface under shear stress ield. Biomaterials 24: 3757–3764. Wang, X., Drew, C., Lee, S.H., Senecal, K.J., Kumar, J. and Samuelson, L.A. 2002. Electrospun nanoibrous membranes for highly sensitive optical sensors. Nano Lett. 11: 1273–1275. Wang, X., Kim, Y.G., Drew, C., Ku, B.C., Kumar, J. and Samuelson, L.A. 2004. Electrostatic assembly of conjugated polymer thin layers on electrospun nanoibrous membranes for biosensors. Nano Lett. 4: 331–334. Wang, X., Um, I.C., Fang, D., Okamoto, A., Hsiao, B.S. and Chu, B. 2005. Formation of water-resistant hyaluronic acid nanoibers by blowing-assisted electro-spinning and non-toxic post treatments. Polymer 46: 4853–4867. Wang, Y. and Santiago-Aviles, J.J. 2004. Synthesis of lead zirconate titanate nanoibres and the Fourier-transform infrared characterization of their metallo-organic decomposition process. Nanotechnology 15: 32–36. Wang, Y., Ramos, I. and Santiago-Aviles, J. 2010. Electrical characterization of a single electrospun porous nanostructured tin oxide ribbon. J. Nanosci. Nanotechnol. 10: 1884–1890. Wang, Z.G., Wang, J.Q. and Xu, Z.K. 2006. Immobilization of lipase from Candida rugosa on electrospun polysulfone nanoibrous membranes by adsorption. J. Mol. Catal. B Enzym. 42: 45–51. Wang, Z.G., Wan, L.S., Liu, Z.M., Huang, X.J. and Xu, Z.K. 2009. Enzyme immobilization on electrospun polymer nanoibers: An overview. J. Mol. Catal. B Enzym. 56: 189–195. Wong, K.K.H., Hutter, J.L., Zinke-Allmang, M. and Wan, W.K. 2009. Physical properties of ion beam treated electrospun poly(vinyl alcohol) nanoibers. Eur. Polym. J. 45: 1349–1358. Xiao, S., Shen, M., Guo, R., Wang, S. and Shi, X. 2009. Immobilization of zerovalent iron nanoparticles into electrospun polymer nanoibers: Synthesis, characterization, and potential environmental applications. J. Phys. Chem. C 113: 18062–18068. Xing, X., Wang, Y. and Li, B. 2008. Nanoibers drawing and nanodevices assembly in poly(trimethylene terephthalate). Opt. Express 16: 10815–10822. Xu, C.Y., Inai, R., Kotaki, M. and Ramakrishna, S. 2004. Aligned biodegradable nanoibrous structure: A potential scaffold for blood vessel engineering. Biomaterials 25: 877–886. Xu, F., Cui, F.Z., Jiao, Y.P., Meng, Q.Y., Wang, X.P. and Cui, X.Y. 2009. Improvement of cytocompatibility of electrospinning PLLA microibers by blending PVP. J. Mater. Sci. Mater. Med. 20: 1331–1338. Yang, C.R., Jia, Z.D., Guan, Z.C. et al. 2009. Polyvinylidene luoride membrane by novel electrospinning system for separator of Li-ion batteries. J. Power Sources 189: 716–720. Yang, F., Wolke, J.G.C. and Jansen, J.A. 2008. Biomimetic calcium phosphate coating on electrospun poly (epsilon-caprolactone) scaffolds for bone tissue engineering. Chem. Eng. J. 137: 154–161. Yang, Q., Wu, J., Li, J.J., Hu, M.X. and Xu, Z.K. 2006. Nanoibrous sugar sticks electrospun from glycopolymers for protein separation via molecular recognition. Macromol. Rapid Commun. 27: 1942–1948. Yang, Y.C., Baker, J.A. and Ward, J.R. 1992. Decontamination of chemical warfare agents. Chem. Rev. 92: 1729–1743. Yao, C., Li, X.S., Neoh, K.G., Shi, Z.L. and Kang, E.T. 2008. Surface modiication and antibacterial activity of electrospun polyurethane ibrous membranes with quaternary ammonium moieties. J. Memb. Sci. 320: 259–267. Yao, C., Li, X.S., Neoh, K.G., Shi, Z.L. and Kang, E.T. 2009. Antibacterial activities of surface modiied electrospun poly(vinylidene luoride-co-hexaluoropropylene) (PVDF-HFP) ibrous membranes. Appl. Surface Sci. 255: 3854–3858. Yarin, A.L. and Zussman, E. 2004. Upward needleless electrospinning of multiple nanoibers. Polymer 45: 2977–2980.

Surface Modification of Electrospun Nanofiber and Nanofibrous Membranes 257 Yarin, A.L., Koombhongse, S. and Reneker, D.H. 2001. Bending instability in electrospinning of nanoibers. J. Appl. Phys. 89: 3018–3026. Yoo, H.S., Kim, T.G. and Park, T.G. 2009. Surface-functionalized electrospun nanoibers for tissue engineering and drug delivery. Adv. Drug Deliv. Rev. 61: 1033–1042. Yoon, Y.I., Moon, H.S., Lyoo, W.S., Lee, T.S. and Park, W.H. 2009. Superhydrophobicity of cellulose triacetate ibrous mats produced by electrospinning and plasma treatment. Carbohydr. Polym. 75: 246–250. Yördem, O.S., Papila, M. and Menceloğlu, Y.Z. 2008. Effects of electrospinning parameters on polyacrylonitrile nanoiber diameter: An investigation by response surface methodology. Mater. Des. 29: 34–44. Yu, H.S., Jang, J.H., Kim, T.I., Lee, H.H. and Kim, H.W. 2009. Apatite-mineralized polycaprolactone nanoibrous web as a bone tissue regeneration substrate. J. Biomed. Mater. Res. A 88: 747–754. Zeng, J., Aigner, A., Czubayko, F., Kissel, T., Wendorff, J.H. and Greiner, A. 2005. Poly(vinylalcohol) nanoibers by electrospinning as a protein delivery system and the retardation of enzyme release by additional polymer coatings. Biomacromolecules 6: 1484–1488. Zhang, C., Yuan, X., Wu, L., Han, Y. and Sheng, J. 2005a. Study on morphology of electrospun poly (vinyl alcohol) mats. Eur. Polym. J. 41: 423–432. Zhang, D. and Chang, J. 2009. Effects of surface modiication of β-CaSiO3 on electrospun poly(butylenes succinate)/β-CaSiO3 composite ibers. Mater. Chem. Phys. 118: 379–384. Zhang, H., Wu, C., Zhang, Y. et al. 2010a. Elaboration, characterization and study of a novel afinity membrane made from electrospun hybrid chitosan/nylon-6 nanoibers for papain puriication. J. Mater. Sci. 45: 2296–2304. Zhang, H., Nie, H., Yu, D., Wu, C., Zhang, Y., White, C.J.B. and Zhu, L. 2010b. Surface modiication of electrospun polyacrylonitrile nanoiber towards developing an afinity membrane for bromelain adsorption. Desalination 256: 141–147. Zhang, J., Senger, B., Vautier, D., Picart, C., Schaaf, P., Voegel, J.C. and Lavalle, P. 2005b. Natural polyelectrolyte ilms based on layer-by layer deposition of collagen and hyaluronic acid. Biomaterials 26: 3353–3361. Zhang, S. 2003. Fabrication of novel biomaterials through molecular self-assembly. Nat. Biotechnol. 21: 1171–1178. Zhang, X., Reagan, M.R. and Kaplan, D.L. 2009. Electrospun silk biomaterial scaffolds for regenerative medicine. Adv. Drug Deliv. Rev. 61: 988–1006. Zhang, Y.Z., Venugopal, J., Huang, Z.M., Lim, C.T., and Ramakrishna, S. 2005c. Characterization of the surface biocompatibility of the electrospun PCL-collagen nanoibers using ibroblasts. Biomacromolecules 6: 2583–2589. Zhao, Z.Z., Li, J.Q., Yuan, X.Y., Li, X., Zhang, Y.Y., and Sheng, J. 2005. Preparation and properties of electrospun poly(vinylidene luoride) membranes. J. Appl. Polym. Sci. 97: 466–474. Zhong, S., Teo, W.E., Zhu, X., Beuerman, R.W., Ramakrishna, S. and Yung, L.Y.L. 2006. An aligned nanoibrous collagen scaffold by electrospinning and its effects on in vitro ibroblast culture. J. Biomed. Mater. Res. A 79: 456–463. Zhu, Y.B., Gao, C.Y., Liu, X.Y. and Shen, J.C. 2002. Surface modiication of polycaprolactone membrane via aminolysis and biomacromolecule immobilization for promoting cytocompatibility of human endothelial cells. Biomacromolecules 3: 1312–1319. Zhu, Y., Leong, M.F., Ong, W.F., Chan-Park, M.B. and Chian, K.S. 2007. Esophageal epithelium regeneration on ibronectin grafted poly(l-lactideco-caprolactone) (PLLC) nanoiber scaffold. Biomaterials 28: 861–868. Zong, X., Kim, K., Fang, D., Ran, S., Hsiao, B.S. and Chu, B. 2002. Structure and process relationship of electrospun bioadsorbable nanoiber membrane. Polymer 439: 4403–4412.

9

Development of Membranes for Pervaporation by Membrane Surface Modification and Incorporation of Inorganic Particles Kailash Chandra Khulbe, Chaoyang Y. Feng, and Takeshi Matsuura

CONTENTS 9.1 9.2

9.3 9.4

Introduction ..................................................................................................260 Theory........................................................................................................... 262 9.2.1 Membrane Productivity .................................................................... 262 9.2.2 Membrane Selectivity ....................................................................... 263 9.2.3 Membrane Stability .......................................................................... 263 9.2.4 Solution–Diffusion Model ................................................................ 263 9.2.5 Pore Flow Model...............................................................................266 Use of Pervaporation Technology .................................................................266 Membranes for PV ........................................................................................ 268 9.4.1 Hydrophilic Membranes ................................................................... 269 9.4.2 Hydrophobic Membranes.................................................................. 276 9.4.3 Organophilic Membrane ................................................................... 283 9.4.4 Membranes Incorporation of Inorganic–Organic Particles.............. 287 9.4.4.1 Inorganic–Organic Particles Composite Membranes ........ 287 9.4.5 Multilayer Composite Membranes ...................................................304 9.4.6 PV Membrane Reactors ....................................................................309 9.4.7 Pervaporation by Ion-Exchange Membrane ..................................... 310

259

260

Membrane Modification: Technology and Applications

9.4.8 Sulfur Removal from Gasoline ......................................................... 313 9.4.9 Supported Liquid Membrane for Pervaporation ............................... 314 9.5 Conclusion .................................................................................................... 317 References .............................................................................................................. 318

9.1

INTRODUCTION

Pervaporation (PV) is the most promising technology in molecular-scale liquid–liquid separations existing in bioreinery, petrochemical, and pharmaceutical industries, as the process is highly selective, economic, safe, and ecofriendly. It is a rapidly developing technology. PV is a very mild process and hence very effective for the separation of less stable compounds. It is a method for separating mixtures of liquids by partial vaporization through a very dense nonporous membrane. The name of this membrane-based process is derived from the two basic steps of the process: irst, the permeation through the membrane by the permeate and then its evaporation into the vapor phase. The concept of PV has been recognized for a long time (Mitchell 1831). However, only in the past 50 years has the development of this technology as a process taken place. Binning et al. (1961a,b) were the irst to suggest, based on their experiments with organic liquid mixtures, that PV had commercial potential. During the following two decades, most of the work on PV was focused on ethanol (EtOH)–water separation. The reason was that PV could achieve improved separation—something that was not possible by ordinary distillation-breaking of the EtOH–water azeotrope. In the 1980s, SETEC and GFT (now part of Le CarboneLorraine) in Germany commercialized PV membrane systems based on composite polyvinyl alcohol (PVA) membranes for EtOH–water separation (Bruschke et al. 1985). At present, by using this technology, there are numerous commercial facilities around the world with capacities as large as 150,000 l/day. Fleming (1992) suggested that PV is also good for solvent dehydration. Recently, many reviews on PV have been written (Smitha et al. 2004; Shao and Huang 2007; Kujawski 2000; Jiang et al. 2009; Peng et al. 2003). Kujawski (2000) wrote a review on the “Application of Pervaporation and Vapor Permeation in the Environmental Protection.” In his article, he suggested following possible application areas for PV and vapor permeation: 1. Dewatering of organic luids such as alcohols, ketones, and ethers. 2. Separation of mixtures from narrow boiling temperatures to constant (azeotrope) boiling temperatures. 3. Removal of organic pollutants from water and air streams. 4. Separation of fermentation products. 5. Separation of organic–organic liquid mixtures. Wee et al. (2008) wrote a review on “Membrane Separation Process-Pervaporation Through Zeolite Membrane.” The focus of this review was on zeolite membrane covering: (i) Synthesis of zeolite membranes; (ii) Membrane characterization; (iii) PV studies; and (iv) Applications in alcohol dehydration, organic–organic separations,

261

Development of Membranes for Pervaporation Membrane Feed

Retentate

Permeate

FIGURE 9.1 Overview of the pervaporation process.

and acid separations. The transport mechanism during PV was discussed and the issues related with PV were addressed. Innovation and future development of zeolite membrane in PV were also presented. In scientiic terms, PV is a method for separating mixtures of liquids by partial vaporization through a nonporous or porous membrane. In other words, it is a technique whereby the components of a mixture of liquids are separated by selective permeation through a semipermeable membrane, the component that passes through the membrane being removed by evaporation. Thus, PV involves the separation of two or more components across the membrane by differing rates of diffusion through a thin polymer and an evaporative phase change comparable to a simple lash step. Figures 9.1 and 9.2 show an overview of the PV process. In PV, the liquid mixture to be separated (feed) is placed in contact with one side of a membrane and the permeated product (permeate) is removed as a low-pressure vapor from the other side. The permeate vapor can be condensed and collected or released as desired. The chemical potential gradient across the membrane is the driving force for the mass transport. The driving force can be created by applying either a vacuum pump or an inert purge (normally air or steam) on the permeate side to maintain the permeate vapor pressure lower than the partial pressure of the feed liquid. Vacuum PV, which is customarily referred to as the standard PV, is the most widely utilized mode of operation, while inert purge PV is normally of interest if the permeate can be discharged without condensation. In addition to these two Charge mixtures (liquid phase)

Membrane

More permeable molecules Less permeable molecules

Nonpermeant (retentate) Membrane support

Permeate (vapor phase)

FIGURE 9.2 Schematic of liquid permeation.

262

Membrane Modification: Technology and Applications

modes of operation, there are several other process variants, including thermal PV, perstraction or osmotic distillation, saturated vapor permeation, and pressure-driven PV (Feng and Huang 1997). PV shares some characteristics with another membranebased separation known as membrane distillation (MD). Urtiaga et al. (2001) studied the parallelism and the differences in PV and vacuum membrane distillation (VMD) in the removal of volatile organic compounds (VOCs) from aqueous streams. In their study, two gas–liquid separation processes, PV and VMD, were compared in the application to the separation of chloroform–water mixtures. Both technologies include the transfer of separated compounds initially in liquid phase through a membrane to a low-pressure gas phase. The use of a solid membrane enhances the separation eficiencies. However, PV and VMD are based on different mechanisms and employ membranes of different characteristics. Selective membranes need to be used in PV processes, while the VMD process requires the use of microporous nonselective membranes. In PV, the properties of the membrane material play a part. If the membrane is hydrophobic, then organic compounds will preferentially permeate relative to water and the permeate will be enriched in the organic compounds. On the other hand, if the membrane is hydrophilic, then water will be enriched in permeate and the organic compound in the feed liquid will be dehydrated. PV separation is governed by the chemical nature of the macromolecules that comprise the membrane, the physical structure of the membrane, the physiochemical properties of the mixtures to be separated, and the permeate–permeate and permeate– membrane interactions. Characteristics of the PV process include: 1. 2. 3. 4.

Low energy consumption No entrainer required without any contamination Permeate must be volatile at operating conditions Functions independent of vapor–liquid equilibrium

PV can be operated at low feed pressures and at ambient temperature or even below this, and no additional chemicals are needed for separation.

9.2

THEORY

PV is a rate-controlled separation process. For the development of a PV membrane, there are three main issues that must be addressed: 1. Membrane productivity 2. Membrane selectivity 3. Membrane stability

9.2.1

MEMBRANE PRODUCTIVITY

It is a measure of the quantity of a component that permeates through a speciic area of the membrane surface in a given unit of time. Membrane productivity is characterized

Development of Membranes for Pervaporation

263

by the permeation lux (J), which relates the product rate to the membrane area required to achieve the separation.

9.2.2

MEMBRANE SELECTIVITY

For the separation of a mixture composed of components A and B, the separation factor (Equation 9.1) is deined as  Y  1− X  α= ,  1 − Y   X 

(9.1)

where X and Y are the molar fractions of the more permeable component A in the feed and in the permeate, respectively. When the separation factor is unity, no separation occurs. When it approaches ininity, the membrane becomes perfectly “semipermeable.” It is the membrane selectivity that forms the basis for separating a mixture. Generally, membrane permeability and selectivity have to be determined experimentally. In general, membrane selectivity is expressed in terms of the enrichment factor (β), which is simply deined as the ratio of the concentrations of the preferentially permeating species in the permeate and in the feed. Unlike α, the numerical value of β depends on the concentration units used. Huang and Yeom (1990) introduced a composite parameter to evaluate the overall performance of a membrane, namely, pervaporation separation index (PSI), which is expressed as the product of the permeation lux and the separation factor. Consider that when α = 1, no separation occurs, but the corresponding PSI may still be very large, depending on the lux, as is the case of a highly porous membrane. Therefore, PSI is redeined as J multiplied by α − 1; a PSI of zero means either zero lux or zero separation (Huang and Feng 1993).

9.2.3

MEMBRANE STABILITY

Membrane stability is the ability of a membrane to maintain both the permeability and the selectivity under speciic system conditions for an extended period of time. Membrane stability is affected by the chemical, mechanical, and thermal properties of the membrane. When considering polymeric membranes for the separation of anhydrous organic mixtures, the membrane stability is the main issue. Shao and Huang (2007) and Feng and Huang (1997) wrote reviews on polymeric membrane PV, which are focused on the fundamental understanding of the membrane. There are principally two approaches to describing mass transport in PV: (i) The solution–diffusion model and (ii) The pore low model.

9.2.4

SOLUTION–DIFFUSION MODEL

Solution–diffusion model is the generally accepted mechanism of mass transport through nonporous membranes (Figure 9.3). According to this mechanism, PV consists of three consecutive steps:

264

Membrane Modification: Technology and Applications Membrane Membrane

Dissolution

Permeate vapor

δ

n sio ffu

Di

δa

Feed liquid

Feed (liquid)

δb Permeate (vapor)

Evaporation δ

(a)

(b)

FIGURE 9.3 Schematic representation of the pervaporation transport mechanism: (a) solution-diffusion model and (b) pore low model.

1. Sorption of the permeate from the feed liquid to the membrane 2. Diffusion of the permeate in the membrane 3. Desorption of the permeate to the vapor phase on the downstream side of the membrane In general, solubility and diffusivity are concentration-dependent. A number of mathematical equations for mass transport have been formulated on the basis of Fick’s diffusion equation using different empirical expressions of concentration dependency of solubility and/or diffusivity. However, these equations cannot be taken for granted unless they are used within the experimentally established range for which the relationships expressed for diffusion and thermodynamic equilibria are applicable. When a PV membrane is in contact with a liquid mixture, the thermodynamic equilibrium reaches instantly at the membrane-feed surface. Cm = K, C

(9.2)

where Cm and C are the concentrations of a species in the membrane surface and in the feed, respectively, and K is thus the partition coeficient of a species between the membrane and the feed phase. Membrane transport is a rate-controlling process, which is generally governed by Fick’s irst law: N = −D

dCm , dδ

(9.3)

where N, D, and δ are the permeation lux of the species through the membrane, diffusion coeficient of the species in the membrane, and the position variable, respectively. When Equations 9.2 and 9.3 are combined and Equation 9.3 is integrated from the feed to the permeate side of the membrane, we get Equation 9.4:

265

Development of Membranes for Pervaporation

N = DK

∆C DK = ∆C , δ δ

(9.4)

where both the diffusion and the partition coeficients are treated as constants. If the transmembrane concentration difference (ΔC) is taken as the driving force for the mass transport, the permeability (Equation 9.5) of the species in the membrane can be deined as (9.5)

P = DK

where P is the permeability. Thus, permeability is an index measuring the intrinsic mass transport capability of a membrane for a species. The ideal separation factor for species i and j can be deined as α ij =

Pi D K = i i = α ij Pj D j K j

( ) (α )

i j K

D

(9.6)

Research efforts in PV are thus devoted to seeking the right membrane materials to maximize the differences in D, K, and P parameters. Experimentally, N (Equation 9.7) and the separation factor can be obtained, respectively, by N=

(α ) i j

Q , A∆t

permselectivity

=

(9.7)

(Y Y ) , (X X ) i i

j

(9.8)

j

where Q is the quantity (in gram or mole) of the permeate collected in a time interval Δt, A is the effective area of the membrane, and X and Y represent the fractions of the components in the feed and in the permeate. Since the downstream pressure in PV operation is negligibly small, the permeation lux of each species through the membrane is essentially proportional to its intrinsic permeability as well as its activity in the feed. As such, the separation factor deined in Equation 9.8 is equivalent to the ideal separation factor deined in Equation 9.6. PV performance of a polymeric membrane is deeply related to the swelling of the membrane. When a membrane is swollen or plasticized by transporting species, the interactions between the polymer chains tend to be diminished, and the membrane matrix will therefore experience an increase in the free volume. The fractional free volume (FFV) (Equation 9.9) of polymers is deined as follows: FFV =

Specific free volume . Polymer specific volume

(9.9)

266

Membrane Modification: Technology and Applications

It is generally true that in a given membrane, increased free volumes correspond to increased diffusion coeficients of the penetrants. When a membrane is plasticized by more than one species, the diffusion coeficient of a species is facilitated by all the plasticizants. The free volume theory developed by Vrentas and Duda (1977) and Vrentas et al. (1986) offers a perfect background for modeling the mass transport in the plasticized membranes. Binary transport in PV with moderate degrees of swelling is discussed in detail by Shao and Huang (2007). Takashi discussed the relation of a water–alcohol separation membrane on the basis of a dissolution–diffusion model (Takashi 1999). Their model provides several guiding principles for designing PV membranes. Tyagi et al. developed the transport model for PV. The transport model assumes the presence of an imaginary phase (liquid or vapor), which is in thermodynamic equilibrium with the membrane phase. Theoretical proiles were compared with experimental concentration proiles. Their work was the irst theoretical work that could reconstruct penetrant proiles inside the membrane showing concentration polarization (Tyagi et al. 1995).

9.2.5

PORE FLOW MODEL

Okada and Matsuura (1991, 1992) proposed a transport model applicable to PV based on a pore low mechanism. It is assumed that there are a bundle of straight cylindrical pores at the membrane surface. The mass transport by the pore low mechanism also consists of three steps (Figure 9.3): 1. Liquid transport from the pore inlet to the liquid–vapor phase boundary 2. Evaporation at the phase boundary 3. Vapor transport from the boundary to the pore outlet The distinguishing feature of the pore low model is that it assumes a liquid–vapor phase boundary inside the membrane, and PV is considered to be a combination of liquid transport and vapor transport in series.

9.3

USE OF PERVAPORATION TECHNOLOGY

The following are the main applications of PV in industry: 1. 2. 3. 4. 5.

The treatment of wastewater contaminated with organics Pollution control application Recovery of valuable organic compounds from process side streams Separation of 99.5% pure EtOH–water systems Harvesting of organic substances from fermented broth

PV technology can be used for product separation or puriication (Table 9.1). Membrane PV process was also used for the enhanced separation of fuel oils (Kujawski et al. 2002). For this process, a commercial hydrophilic membrane was used. PV technology is being developed for the treatment of brackish groundwater

Alcohol

Aromatic

Ethanol and methanol Propanol

Benzene

Ketone

Amines

Aliphatic

Ether

Acetone

Triethylamine

Toluene

Butanone

Pyridine

Chlorinated hydrocarbons (various) Dichloro methane

Isopropanol

Phenol

Methyl isobutyl ketone (MIBK)

Aniline

Perchloroethylene

Methyl tert-butyl ether (MTBE) Ethyl tert-butyl ether (ETBE) Diisopropyl ether (DIPE)

Butanol Pentanol Cyclohexanol Benzyl alcohol

Aromas

Ester Methyl acetate Ethyl acetate Butyl acetate

Organic Acid

Inorganic

Acetic acid

Ammonia Hydrazine

Development of Membranes for Pervaporation

TABLE 9.1 Products that Can Be Treated or Purified by PV Technology

Tetrahydrofuran Dioxane

267

268

Membrane Modification: Technology and Applications

(Quińones-Bolańos et al. 2005) and for the removal of VOCs from groundwater (Shah et al.2004) and heat pump systems (Riffat et al. 2004). In recent years, there has been rapid development in the ield of organic–organic separation. This is focused on the isolation of aromatic compounds from aliphatic hydrocarbons or gasoline. In the petroleum industry, the most important and most dificult process is the separation of benzene from cyclohexane (CYH). Desulfurization of gasoline is one of the potential applications of organophilic PV. The reduction of sulfur-containing components in FCC naphtha using PV is feasible (White and Lesemann 2002; Song 2003). PV is also used to recover aroma compounds lost during evaporation from fruit juices such as apple juice. The vapor from the evaporation process is further processed using PV. The recovered, concentrated apple juice can be combined with the product solution to help the apple juice retain its aromatic and taste qualities. PV systems can be divided into two categories: (i) Batch PV and (ii) Continuous PV. The batch PV system is simple with great lexibility; however, a buffer tank is required for batch operation. On the other hand, continuous PV consumes very little energy, operates best with low impurities in the feed, and is best for larger capacities. Vapor phase permeation is preferred for direct feeds from distillation columns or for streams with dissolved solids.

9.4

MEMBRANES FOR PV

The composition and the morphology of the membranes are key to effective use of membrane technology. The choice of membranes strongly depends on the types of applications (Koops and Smolders 1991). It is important to know which of the components should be separated from the mixture and whether the component is water or an organic liquid. In general, the component with the smallest weight fraction in the mixture should preferentially be transported across the membrane. The PV membrane can be considered as a dense homogeneous medium in which diffusion of species takes place in the free volume that is present between the macromolecular chains of the polymeric membrane material. PV membranes fall into two categories: the homogeneous membrane and the composite membrane. Higher permeation lux can be obtained by composite membranes in comparison with the homogeneous ones, due to the much smaller thickness of the homogeneous membrane supported on a porous substrate. Surface modiication of membranes is an attractive approach to change the surface properties of the membrane in a deined selective way while preserving its macroporous structure. The surface-modiied membranes are considered as one type of composite membranes. Modiication of membranes can be done with grafting, coating, illing, blending two polymers, and using high energy particles. Different composite membranes consisting of polyethersulfone (PES) as a sublayer and polydimethyl siloxane (PDMS) coating as a top layer have been prepared by Gudernatsch et al. (1991). By varying the polymer concentration in the casting solution of the sublayer, different properties were obtained. These membranes were used for PV.

Development of Membranes for Pervaporation

269

The membranes used in PV processes are classiied according to the nature of the separation being performed. There are three kinds of PV membranes: (i) Hydrophilic membranes, (ii) Hydrophobic membranes, and (iii) Organophilic membranes. 1. Hydrophilic membranes To remove water from organic solutions, hydrophilic membranes are used. These types of membranes are typically made of polymers with glass transition temperatures above room temperatures. PVA is an example of a hydrophilic membrane material. 2. Hydrophobic membranes Hydrophobic membranes can be used to extract organic solvents or VOCs from water. The membranes are made of hydrophobic cross-linked polymers. These compounds are repelled by water and are usually neutral (no charge). Membranes with hydrophilic character can be changed to hydrophobic by grafting or coating the hydrophobic molecules on the surface and by other methods as well. These membranes are generally used for the removal of VOCs from aqueous stream. 3. Organophilic membranes Organophilic membranes can be used to extract organic solvents from other organic solvents. However, in general, organophilic membranes are hydrophobic. These membranes are typically made up of elastomeric materials (polymers with glass transition temperatures below room temperature). The lexible nature of these polymers makes them ideal for allowing organic solvents to pass through. Examples include nitrile, butadiene rubber, and styrene butadiene rubber. Buckley-Smith (2006) discussed the use of solubility parameters to select membrane materials for PV of organic mixtures. His research showed that Hansen solubility parameters (HSP) are a good screening method for PV membranes, especially where the molecules being separated are of comparable size. Polymers that have HSP close to the desired component and not to other components tend to have the best selectivity and lux characteristics. However, diffusion is an important factor and is not completely accounted for by HSP.

9.4.1

HYDROPHILIC MEMBRANES

Hydrophilic compounds have an afinity to water and are usually charged or have polar side groups that attract water. A hydrophilic membrane will not be fouled with oil, grease, or other hydrophobic substances. The membrane is highly attractive to water so the water molecules will push away other molecules in order to gain access to the membrane. Once formed, hydrogen bonds are quite stable and reluctant to break apart. This keeps contaminants away from the membrane so it remains clean and functioning for longer. Hydrophilic coatings or materials become wet very easily and maintain wetness for longer than a hydrophobic equivalent would. Hydrophilic membranes are generally used for dehydration of organic liquids by using PV technique. They are mainly used for dehydration of aqueous azeotrope

270

Membrane Modification: Technology and Applications

EtOH. The common hydrophilic membranes are poly(amide sulfonamide) and PVA, sodium alginate, polyimide (PI), cellulose, and chitosan. The hydrophobic character of the membranes can be changed to hydrophilic by grafting or coating hydrophilic molecules on the surface. PV using PVA membranes were studied extensively (Yeom and Lee 1996; Li et al. 2007; Rhim et al. 2002). Nonporous PVA membranes were modiied in gaseous plasma to improve the separation eficiency for water–isopropanol (IPA) mixtures via PV. Membranes were exposed to the plasma of air, nitrogen, and oxygen for 2–15 min and their performance was evaluated at 75°C. Modiication of membranes in a nitrogen plasma environment leads to increase in selectivity by about 1477 at 25°C. The cross-linking on the membrane surface induced by the plasma was found to be responsible for the observed improvement (Upadhyay and Bhat 2004). Hyder et al. (2006) studied the surface and bulk properties of hydrophilic PVA membranes subjected to PV. The PVA membranes were cross-linked in two ways by heating at 125°C or by chemical reaction with glutaraldehyde at room temperature. These membranes were used for the dehydration of EtOH–water mixtures over a range of EtOH concentrations (10%–70%) in feed solution and at varied temperatures (from 25°C to 50°C). The PV results showed that the thermally cross-linked membrane was more hydrophilic than the chemically cross-linked membranes, and this helped transport water at a higher lux through the membrane. However, the selectivity of the thermally cross-linked membrane was lower and water lux through the membrane became higher when compared with the chemically crosslinked membranes. The dehydration results were correlated with the results of the physiochemical measurements of the membranes. Aminabhavi et al. (2006) modiied the surface of the polyethylene glycol (PEG)/ PVA membrane by grafting poly(acrylonitrile) (PAN). These membranes are useful in removing trace amounts of water from organic compounds via PV. PV of water–EtOH mixtures through cross-linked and surface-modiied PVA membranes was reported by Kang et al. (1990). In order to enhance the permselectivity of water through the cross-linked membrane, the surface of the cross-linked PVA membrane was hydrophilically modiied by reaction with monochloroacetic acid. The cross-linked and surface-modiied membranes exhibited a permselectivity for water enhanced by a factor of nearly 2 compared with that for the only cross-linked membrane, while maintaining the total lux. Park et al. (2000) studied PV of pyridine–water mixture through a poly(acrylonitrile-co-monoacryloxyethyl phosphate) (PANPH) membrane for the dehydration of aqueous pyridine solution. All the PAN-based phosphoric acid– containing membranes were very selective toward water. The PV performance depended on the content of the phosphoric acid moiety in the membrane, operating temperature, and feed concentration. The PV performances of the water–pyridine mixture through the PANPH membrane showed good PV performance. PV performance at 75°C was such that the water concentration in the permeate was >99.8% and the total lux was about 120 g/m2 h. PV of IPA–water mixtures through the wet-spinning and then heat-treated PAN hollow iber (hf) membranes was studied by Tsai et al. (2005). Compared with PAN precursor hf membranes, the PV performances of heat-treated PAN hf membranes

Development of Membranes for Pervaporation

271

improved. The porous structure of PAN precursor hf membranes becomes denser after heat treatment. The PV results of permeation rate and water content in the permeate for the PV of a 90 wt% aqueous IPA solution through a PAN hf membrane with heat treatment at 120°C for 12 h were 186 g/m2 h and 99.2 wt%, respectively. Superiority of PI membranes subjected to PV over other polymeric membranes has been highlighted by Jiang et al. (2009). Membranes prepared from soluble PIs derived from 4,4′-diamino-3,3′-dimethyldiphenylmethane was used as a PV membrane for EtOH–water mixture (Wang et al.2006). The average value of the separation factor and total permeation lux were 46–108 and 660–1380 g/m2 h, respectively. Yanagishita et al. (1995) used asymmetric PI (aromatic) membranes (lat sheet) prepared by a phase inversion process, for the separation of alcohol solution by PV. The membrane exhibited separation factors of α(H2O/EtOH) = 900 and α(H2O/2-PrOH) = 11,000 with a lux of 0.45 kg/m 2 h for 95 vol% alcohol aqueous solution at 60°C. The membrane showed stable performance for 3 months. Torlon 4000TF polyamidoimide (PAI) and Ultem 1010 polyetherimide (PEI) membranes were fabricated and studied for IPA dehydration by PV (Wang et al. 2008). The properties of these two materials and the membranes fabricated were tested and compared through different characterizations including differential scanning calorimetry (DSC), thermogravimetric analysis (TGA), goniometer, and x-ray diffraction, and compared with PEI dense membranes, PAI dense membranes show a much higher separation factor (up to 6000 at 60°C) and comparable lux. The PV results of the asymmetric membranes show that the polymer concentration for membrane casting is very important. The separation using a membrane with porous structure facing against the feed solution showed a much higher separation factor with only a slight decrease in lux. The lower swelling of the dense selective layer was believed to be responsible for the higher separation performance of the reversed membrane. Chitosan is widely used in membrane applications based on its high hydrophilicity, good ilm-forming character, and excellent chemical-resistant properties. However, chitosan membranes are highly swollen in water and alcohol solutions, and such swollen membranes usually lose their permselectivity and show poor longterm stability of operation. Much endeavor was made to improve the stability and mechanical properties of chitosan-based membranes such as blending chitosan with other polymers, casting chitosan on other polymer substrate, bringing cross-linked structure to membrane, and adding inorganic reinforcements into chitosan membranes (Liu et al. 2005). Masaru et al. (1985) were the irst to use chitosan membranes for EtOH dehydration via PV. It was noticed that chitosan is a very promising material for solvent dehydration, and it demonstrated better separating performance than the cross-linked PVA (Shao and Huang 2007). Blended PVA and chitosan membranes (varied from 0% to 100%) were used for dehydration of IPA (Svang-Ariyaskul et al. 2006). After heating the blended PVA– chitosan membranes at 150°C for an hour, the membranes were cross-linked by glutaraldehyde and sulfuric acid in acetone (AC) aqueous solution. The membranes were tested at 30°C–60°C for dehydration performance of 50%–95% IPA aqueous solution. The membrane containing 75% of chitosan performed the best. For a feed solution containing 90% IPA at 60°C, the permeate lux was 644 g/m2 h with water content of nearly 100% in the permeate. With 55% IPA in the feed at 60°C,

272

Membrane Modification: Technology and Applications

the permeate lux was 3812 g/m2 h. Novel diisocyanate cross-linked chitosan membranes for PV separation of water–IPA mixtures were also developed (Choudhari et al. 2007). The highest separation selectivity observed was 5918 with a lux of 2.20 × 10−2 kg/m2 h at 30°C for 5 mass% of water. The total lux and the lux of water were found to be overlapping particularly for higher cross-linked membranes, suggesting that the membranes developed with higher amount of blocked diisocyanate could be used effectively to break the azeotropic point of the water–IPA mixture, so as to remove a small amount of water from IPA. Nawawi and Hassan (2003) prepared homogeneous (unmodiied) and crosslinked chitosan (extracted from shrimp shells) membranes. Chitosan membranes were modiied via a chemical cross-linking technique. PV experiments were done using an IPA–water system. From the PV experiments, it was observed that chitosan membranes exhibited preferential permeation to water. The modiied chitosan membrane showed a lower permeation lux, but a higher separation factor than the unmodiied membranes. The modiied chitosan membrane had a better PV performance than the unmodiied membranes in terms of a PSI. Chitosan-based polyelectrolyte complexes (PECs) were developed as PV membranes by incorporating phosphotungstic acid (PTA) (Rachipudi et al. 2009). Membranes were prepared by immersing the chitosan membrane in 0.015, 0.025, and 0.045 M solutions of PTA at ambient temperature and the resulting membranes were washed with distilled water. These membranes were tested for their ability to separate water–IPA mixtures by PV in the temperature range of 30°C–50°C. The experimental results demonstrated that both lux and selectivity increased simultaneously with the increasing PTA content in the membrane. The permeation lux of pure chitosan membrane increased dramatically from 4.13 to 11.7 × 10−2 kg/m2 h and correspondingly its separation factor increased from 4,490 to 11,241 and then decreased to 7,490 at 30°C for 10 wt% of water in the feed, with the increase in PTA content. Chitosan was also cross-linked with 3-aminopropyl-triethoxysilane (APTEOS) to prepare chitosan–silica hybrid membranes (CSHMs) (Chen et al. 2007a). Both the permeation lux and the water permselectivity increased remarkably with increasing APTEOS content. CSHM-10 containing 10 wt% APTEOS has the highest separation factor of 597 with a lux of 0.887 kg/m2 h in PV of 85% feed EtOH at 323 K. Xiao et al. (2007) prepared trimesoyl chloride (TMC) cross-linked PVA/chitosan membranes and synthetic PI membranes for PV dehydration of IPA and gas separation. PVA membranes were interfacially cross-linked with different amounts of TMC/ hexane. Cross-linking time was varied from 20 to 80 min. Water permeation and PV dehydration of IPA were conducted. The results showed that PVA–3TMC (60 min cross-linking time) had the best overall PV properties among the four PVA–TMC membranes studied (20, 40, 60, and 80 min cross-linking time). Huang et al. (1999, 2000, 2001) used cross-linked chitosan/polysulfone (PSf) composite membranes for the PV dehydration of alcohol mixtures. It was noticed that the poor wettability of PSf substrate could be improved by immersing the substrate into dilute polymer solution before the top layer casting. In addition to the improvement of structural stability, the permeation lux also increased due to the reduced hydrophobicity of the PSf layer produced by coating with the ultrathin PVA layer. It was suggested that the

Development of Membranes for Pervaporation

273

porous substrate in the composite membrane acts not only as a mechanical support but also as a part of a selective layer affecting the PV performance. Schauer et al. (2003) used poly(2,6-dimethyl-1,4-phenylene oxide) (PPO) membranes to separate water–EtOH mixtures by PV. Asymmetric membranes were prepared from solutions containing chloroform as a solvent and 1-butanol as a nonsolvent via the phase inversion technique. Dense membranes were prepared from chloroform solution by evaporation. Nonporous membranes (membranes precipitated from solutions with a small amount of the nonsolvent or prepared by evaporation) were preferentially permeable to water. Microporous membranes (prepared from solutions with a large amount of the nonsolvent) were preferentially permeable to EtOH, provided the membrane was not wetted by the feed solution. Yoshikawa et al. (2002) used green polymer, agarose, as a membrane material and studied PV of aqueous organic mixtures, such as water–methanol, water–EtOH, water–1-propanol, water–2-propanol, and water–acetic acid. The agarose membranes preferentially permeated water from aqueous organic mixtures by PV. From their data, they concluded that agarose can be used as one of the promising membranes for the dehydration of water-miscible organics. Polyurethane (PU) PV membranes were reported by Schauer et al. (1999). PU membranes were prepared by the reaction of toluene-2,4-diisocyanate with hydroxylterminated oligomers. Oligomers were either liquid polybutadiene (PB) (MW 3000) or propylene oxide–based PEs (MW 420 and 4800). The prepared membranes were used in PV of binary mixtures of water–EtOH, water–dioxane, and EtOH– toluene, respectively. Membranes obtained from the polymer quaternized poly[3(N′,N′-dimethyl)aminopropylamide-co-acrylonitriles] showed selective separation of water from aqueous EtOH solution by PV (Yoshikawa et al. 1991). The separation factor toward water reached over 15,000. Membrane performance showed a good correlation to membrane polarity. DSC melting endotherms of the water-swollen membranes were studied to clarify the state of water in the membrane. The results suggested that there were two states of water in the membrane: bound and free. The higher the fraction of bound water in the membrane, clearly, the more preferentially was water permeated. Lokaj and Bilá (2003) studied the PV of EtOH–water mixtures through styrenesubstituted N-phenylmaleimide copolymer membranes. The object was to estimate the effect of incorporated maleimide units on the PV properties on the polymeric membranes. For this, a number of copolymers of styrene with substituted N-phenylmaleimide were synthesized and their solutions in chloroform were used in the casting of homogeneous membranes. These membranes were characterized by the separation factor related to transported water and by the lux of the permeate. In contrast to the membranes made from copolymers of styrene with N-phenylmaleimide, the separation factor of the membranes containing substituted N-phenylmaleimide increased with increasing amount of EtOH in the feed solution. Sulfonate-containing aromatic polyamides PV membranes were also studied for dehydration of organic solvents (Kirsh et al. 1995). Films were prepared from aromatic polyamides (condensation of 4-4′-diaminodiphenylamino-2-sulfonate (DAPDS) and 4-4′-diaminodiphenyloxide, taken in different ratios, with isophthaloyl chloride).

274

Membrane Modification: Technology and Applications

Sulfonation improved the PV performance of the PSf membrane. Introduction of a sulfuric group in the polymer unit increased the hydrophilicity of the PSf membrane (Chen et al.2001a). The separation performance of water–EtOH mixtures was shown to strongly depend on the degree of substitution of the PSf membrane. Klamklang et al. (2002) improved the hydrophilic property of natural rubber latex by combining it with hydrophilic polymers (such as polyacrylamide (PAM), phthalated chitosan) using a mechanical blending technique. The membranes were subjected to PV dehydration of aqueous azeotrope EtOH. The polymer network of prevulcanized latex and PAM (at ratio 60:40 and using 1% of ethylene glycol dimethacrylate (EGDM)) was the most appropriate membrane in view of strength and water sorption selectivity, as a dehydration PV membrane for azeotropic EtOH solution. The characteristics of the above membrane were as follows: • Tensile strength = 16 ± 0.81 N/mm2 • Total sorption = 0.036 ± 0.003 g/g of dry membrane • Water sorption selectivity = 231 ± 8.9 Mahmud et al. (2001) studied the PV performance of PES membranes prepared by incorporating surface-modifying macromolecules (SMMs). The experimental data showed that the PES–SMMs membrane appeared to be water-selective as a signiicant depletion of chloroform was achieved in the permeate. Li et al. (2008a) introduced solvent-resistant multifunctional PV membranes based on segmented polymer networks (SPNs). Hydrophilic bis(acrylate)-terminated poly(ethylene oxide) was used as a macromolecular cross-linker of different hydrophobic polyacrylates for the synthesis of amphiphilic SPNs. Multifunctional composite membranes with thin SPN top layers were prepared by in situ polymerization. The support consisted of hydrolyzed PAN. These membranes were tested for dehydration of EtOH and isopropyl alcohol. The selectivity of the membranes greatly depended on the composition or the ratio of the hydrophilic and hydrophobic phases of the SPN. Wang et al. (2002) demonstrated a composite membrane (subjected to PV) of an asymmetric poly(4-methyl-1-pentene) (TPX) membrane dip-coated with poly(acrylic acid) (PAA). To improve the interface peeling of the PAA/TPX composite membranes, the surfaces of TPX membranes were modiied by residual air plasma in a tubulartype reactor. The plasma treatments were effective in rendering the asymmetric TPX membrane hydrophilic. Optimal results were obtained with PAA/TPX composite membrane prepared from the PAA/ethylene glycol (EG)/aluminum nitrate = 1/2/0.05 coating solution at 5 W/30 s plasma treatment condition. Concentration of the water in the permeate was nearly 100%, and a permeate lux of 960 g/m2 h was obtained with a 3 wt% feed acetic acid concentration. Neidlinger et al. (1987) used a polyethylenimine/PSf composite membrane for PV in the separation of EtOH–water mixtures. Polyethylenimine solution was coated onto a PSf support. Cross-links were then generated by limited exposure to toluene2,4-diisocyanate solution, after which the prepared membrane was heat-cured. The resulting membrane structures showed high selectivity in permeating EtOH or water over a wide range of feed concentration.

Development of Membranes for Pervaporation

275

Lai et al. (1995) introduced chemically modiied TPX membrane for PV separation of acetic acid–water mixtures. The introduction of high hydrophilic 4-vinylpyridine (4-VP) monomer into the TPX matrix was done by free-radical polymerization to form the TPX/P4-VP membrane. The separation factor and the permeation rate of TPX/P4-VP membranes were higher than those of the unmodiied TPX membranes for PV of aqueous acetic acid solutions. The surface of poly(tetraluoroethylene) (e-PTTE) membrane was modiied by using acetylene/nitrogen plasma. The variation in the surface morphology of e-PTTE membranes was conirmed by FTIR-ATR, SEM, and contact angle measurements. It was noticed that the surface hydrophilicity increased with increasing nitrogen content in the feed gas mixture, RF power, and plasma treatment time. The surface pore size decreased with increasing RF power and plasma treatment time. The water contact angles of modiied e-PTTE membrane decreased from 125.8° to 34.1° through the acetylene/nitrogen plasma treatment. The vapor permeation results were obtained using a plasma-modiied membrane, giving a permeation rate of 666 g/m2 h and water concentration in the permeate of 72 wt% from the feed solution containing 90 wt% of EtOH (Tu et al. 2005). Ulbricht and Schwarz (1997) applied for the irst time the “pore-illing” heterogeneous graft copolymerization concept to asymmetric membranes. Asymmetric PAN membranes (average pore sizes of 7 or 12 nm) were used as matrix of polymeric PV separation phases, which were prepared in situ by heterogeneous photoinitiated graft copolymerization. By this means, defect-free and stable layers were synthesized from moderately hydrophilic methacrylates such as poly(ethylene glycol methacrylates). Xu et al. (2003) used polypropylene hf membranes grafted with PAA for EtOH– water separation by PV. The separation factor increased with the degree of grafting of PAA in the range of 20–70 wt%. Incorporating different counterions as well as a multifunctional comonomer into the grafted chain had obvious inluence on the PV properties. The separation factor of the counterion-containing membrane decreased in the following order: Al3+ > K+ > Ca2+ > Na+ > Li+, and the permeation lux increased following the sequence Al3+ > Ca2+ > K+ > Na+ > Li+. Remarkable increase in the permeation lux without reduction in the separation factor could be reached by incorporating Al3+ counterion in the membrane. Hhairnar and Pangarkar (2004) separated glycerol–water mixtures via PV by using hydrophilic poly(acrylonitrile-co-methacrylic acid) (PANMAC), poly(acrylonitrileco-hydroxyethylmethacrylate), PVA GFT-1001, and PVA cross-linked with a maleic anhydride (PVAManh) series. All membranes were found to be highly water-selective. PVAManh membrane yielded the highest permeation lux for water over the entire range of water concentrations studied. Separation of hydrazine–water was carried out via PV through various blended membranes (ethyl cellulose (EC) and acrylonitrile–butadiene–styrene (ABS)) as well as pure EC and ABS polymeric membranes (Mandal et al. 2008). Results of PV showed that normalized lux decreases and selectivity increases with increasing ABS in the blended membrane. Through contact angle measurements, it was observed that increasing content of ABS increases the hydrophobic characteristic.

276

Membrane Modification: Technology and Applications

Lee et al. (1990) studied the PV of EtOH–water mixtures by cationic–anionic interpenetrating polymer network (IPN) membranes. The cationic PU prepared by quaternizing the tertiary amine from N-methyldiethanolamine and the anionic acrylic copolymer from acrylic acid were used as the cationic and anionic components. The separation factor increased as the content of the anionic component in the membrane increased. The ion content in the membrane was also found to be an important factor for the optimum permseparation of water. Better performance was achieved by PEG-based PU as the cationic component of the IPN membrane. The highest selectivity was observed for the IPN membrane with 10 wt% of PEGbased PU. El-Gendi et al. (2010) studied the PV properties of block ether aromatic copolyimide series membranes where the ether soft block acted both as a selective and a permeable block. The PV results showed that rubbery copolyimides can lead to promising asymmetric membranes for liquid–liquid separations (water–EtOH). Kühn and Maser (1990) patented hydrophilic collagen ilms as PV membranes for the selective separation of small polar molecules from liquid mixtures. They claimed that the collagen ilm can be used for the removal of water from chemicals, separation of azeotropic mixtures, and concentration of beverages, especially juice and wine. Maser et al. (1991) tested various ilms based on collagen for use as hydrophilic PV membranes. Aqueous solutions of EtOH, IPA, ethylene glycol, and AC were used as model mixtures to study the separation properties of these membranes. Collagen ilms proved to be promising PV membranes with good selectivity and, most of all, surprisingly high luxes at low operation temperatures.

9.4.2

HYDROPHOBIC MEMBRANES

Hydrophobic compounds are repelled by water and are usually neutral (no charge). Hydrophilic character membranes can be changed to hydrophobic by grafting or coating the hydrophobic molecules on the surface. These membranes are generally used for the removal of VOCs from aqueous streams. As PDMS is hydrophobic, mostly hydrophobic PV membranes are made of PDMS or are PDMS-based. Peng and Liu (2003) used PDMS membranes to remove 1,1,1-trichloroethane (TCA) from contaminated ground water using a bench-scale membrane PV unit. The effect of temperature on PV performance was profound; the TCA lux and the water lux increased as the temperature in the unit was raised. EtOH-permselective PDMS/PVDF composite membranes were prepared by curing PDMS with various cross-linking reagents, such as tetraethoxylsilane (TEOS), γ-aminopropyltriethoxylsilane (APTEOS), phenyltrimethoxylsilane (PTMOS), and octyltrimethoxylsilane (OTMOS) (Zhan et al. 2009b). The cross-linking density and surface properties of the PDMS active layer were adjusted by varying the crosslinking reagents. These membranes were used for the separation of EtOH from its aqueous solution. The experimental results revealed that the total lux of the four kinds of membranes increased with the increase in temperature. However, the separation factor followed a reversed order. It was found that the composite membrane cured by PTMOS showed much better separation performance compared with other

Development of Membranes for Pervaporation

277

membranes. It was reported that measured diffusivities of EtOH and water in PDMS had a magnitude of 10−10 m2/s and 10−11 m2/s at 313.15 K, respectively. Kim et al. (2001) used cross-linked PDMS membranes for the removal of chlorinated organic compounds. Mishima and Nakagawa (2000) used a grafting technique to introduce 1H-, 1H-, 9H-hexadecaluorononyl methacrylate (HDFNMA) into a PDMS membrane and then evaluated this membrane for the separation of aromatic and halogenated hydrocarbons from water by PV. They reported that a 7.0% degree of grafting gave a higher permselectivity for toluene. Uragami et al. (2001) used poly(methylmethacrylate)–poly(dimethylsiloxane) (PMMA-g-PDMS), poly(ethylmethacrylate)–PDMS (PEMA-g-PDMS), and poly(nbutylmethacrylate–PDMS (PBMA-g-PDMS))-grafted copolymers to remove VOCs such as benzene and chloroform in aqueous benzene and chloroform solution. When aqueous solutions of dilute VOCs were permeated through the PMMA-g-PDMS and PEMA-g-PDMS membranes, these membranes showed benzene and chloroform permselectivity. Permeation and separation characteristics of the PMMA-gPDMS and PEMA-g-PDMS membranes changed drastically at a DMS content of about 40 and 70 mol%, respectively. Permeation rate and VOC permselectivity of the PBMA-g-PDMS membrane, however, increased gradually with the increasing DMS content, unlike those of the PMMA-g-PDMS and PEMA-g-PDMS membranes. The high benzene and chloroform permselectivity of microphase-separated membranes was due to high diffusivity of benzene and chloroform in the continuous PDMS phase. The preferential afinity to EtOH depends on the balance between the hydrophilicity and the hydrophobicity of the membrane’s material (Huang 1991). Qiu and Peinemann (2006) developed novel organic nanocomposite membrane for PV. The basic polymers were PDMS and poly(1-trimethylsilyl-1-propyne) (PTMSP). By implanting the hydrophobic organic molecules in PTMSP and PDMS, permselectivity to EtOH was enhanced. For example, PDMS with 20 wt% α-cyclodextrins provides a separation factor of 12 for EtOH (5 wt%)–water (95 wt%). Similarly, PTMSP with only 8 wt% α-cyclodextrins improved the enrichment of the low concentration of EtOH from 5 to 48 wt% and maintained the lux at 9 kg μm/m2 h. They claimed that the increased performance in EtOH–water separation with this organic nanocomposite membrane may lead to the practical industrial application by means of the PV process to produce bioethanol. Novel composite cross-linked PDMS/PSf PV membrane was reported by Liang et al. (2005) to remove 1,2-dichloroethane (1,2-DCE) from aqueous solution. Composite PV membranes were prepared by the cross-linking of H-terminated oligosilylstyrene (oligo-SiH3) and vinyl-terminated PDMS (vinyl-PDMS) on the surface of PSf UF membrane. The membrane showed the characteristic of preferential selection to 1,2-DCE and the separation factor was 600–2300. In another article, Liang et al. (2006) studied the mass transfer of dilute 1,2-dimethoxyethane (1,2-DME) aqueous solution during PV on vinyl-PDMS/PSf membrane. The effect of a boundary layer on the mass transfer of dilute 1,2-DME aqueous solution was observed during the PV of the composite membranes constructed from cross-linked PDMS as the top layer and PSf UF membrane as the substrate. In the lavor industry, concentrated lavor distillates and extracts are important products for simulating natural lavors (Manjuan 2005). Manjuan (2005) used

278

Membrane Modification: Technology and Applications

hydrophobic membranes (four PDMS and one POMS) to extract solute (lavor organics) from solvent (water). Experimental results obtained by Manjuan can help lavor or membrane companies to select optimum operating conditions and membranes for application of the PV technique. Diban et al. (2008) used a PV hf module with a PDMS commercial membrane for the separation of key components of bilberry aroma. Grape juice aroma (methyl anthranilate) was separated from its juice by using PV and PDMS/PC membranes (Rajagopalan and Cheryan 1995). Hoshi et al. (1999) developed a copolymer membrane of poly n-butyl acrylate-coacrylic acid (BA-co-AA) by a cross-linking method in order to overcome the drawback of weak mechanical strength when long-chain ester residues were introduced to obtain higher afinity for VOCs such as TCE. In another study, Hoshi et al. (2000) used PU membranes. PUs were synthesized by polyaddition reaction of poly(alkylene glycols) (polyethylyeneglycol, polytetramethyleneglycol, polyhexamethyleneglycol, polyoctamethyleneglycol, and polydecamethyleneglycol) and 1,6-hexamethylenediisocyanate (HMDI) using a dibutyltin dilaurate catalyst. The phenol permselectivity (the phenol concentration in the permeate liquid and phenol partial lux) increased with increasing methylene group length in the poly(alkylene glycol), and the water lux slightly decreased. Dutta and Sikdar (1999) studied a block copolymer membrane made from styrene and butadiene (S-B-S) for the removal of chlorinated hydrocarbons from water. The polystyrene block phase provides mechanical strength due to its high glass transition temperature (95°C), while the PB block provides good selectivity for the organic compounds relative to water. A thin ilm of this copolymer coated on a porous polytetraluoroethylene (PTFE) support yielded an organic water separation factor as high as 5000 for TCA, trichloroethylene (TCE), and perchloroethylene (PCE) at near-ambient temperatures, but this decreased substantially at higher temperature. Knapp et al. (2009) introduced polynorbornene PV membrane ilm for the separation of organics (butanol) from water. Wang et al. (2005) used the carbazolefunctionalized norbornene membrane for the separation of aqueous alcohol solution by PV. A higher separation factor and lower permeation rate were achieved for a higher-molecular-weight alcohol. Jian and Pintauro (1997) used asymmetric PVDF hfs for organic–water PV separations. They reported that organic–water separation properties depend on the spinning solution viscosity and bore luid composition. Variations in the membrane performance (benzene separation factor and lux) with the changes in organic-feed concentration, downstream pressure, and feed temperature were qualitatively similar to those observed for lat sheet asymmetric PVDF membranes and elastomeric PV membranes. O’Brien and Craig (1996) used commercially available polydimethyl–siloxane membrane for EtOH production in a continuous fermentation/membrane PV system. The PV module contained 0.1 m2 of a commercially available polydimethylsiloxane membrane and consistently produced permeate of 20%–23% (w/w) EtOH while maintaining a level of 4%–6% EtOH in a stirred-tank fermentor. PV lux and EtOH selectivities were 0.31–0.79 l/m2 h and 1.8–6.5 respectively. Lee et al. (1989) studied the effect of irradiation treatment on polyethylene (PE) membranes used for the separation of chlorobenzene from a chlorobenzene

279

Development of Membranes for Pervaporation

(10 ppm)–water system. The PE is partially crystalline in nature at temperatures below 105°C. The crystalline phase does not absorb the penetrants. Irradiation induced cross-linking between the PE molecules and reduced the degree of crystallinity. The permeability of chlorobenzene in PE exposed to a moderate irradiation dose of about 35 megaroentgen was found to be 40% larger and the total lux showed a 50% increase relative to the nonirradiated membrane. Aneja (2006) reported that biocomponent/composite hfs have major potential for the separation of organic and aqueous–organic mixtures via PV. They investigated the transport of isopentane through a composite of silicone rubber and PSf asymmetric hf membrane. A one-parameter mathematical model of the PV process was developed. The parameter was related to the sorption and transport properties of the penetrant through the membrane layers. They claimed that the PV parameter is a useful tool for membrane characterization and process design. Acrylamide (AAm)–plasma graft–aromatic polyamide (AAm-p-aramide) membrane was prepared by plasma graft polymerization. The membrane was subjected to PV (water–EtOH mixtures) (Teng et al. 2000). The effects of the degree of grafting, feed composition, feed temperature, and surface properties on the PV performances were studied. The water contact angle of the AAm-p-aramide membrane decreases with the increase in the degree of grafting, as shown in Figure 9.4. The separation factor and permeation rate of AAm-p-aramide membranes were higher than those of the unmodiied aramide membrane. Optimum PV was obtained by an AAm-p-aramide membrane with a degree of grafting of 20.5% for a 90 wt% EtOH feed concentration, giving a separation factor of 200 and permeation rate of 325 g/m2 h. CF4 plasma and CO2 plasma were used to modify a polyamide (PA) membrane, resulting in different surfaces; one is rather hydrophobic (with CF4), whereas the 100

Contact angle (deg.)

90

80

70

60

50 0

5 10 15 20 Degree of grafting (wt %)

25

FIGURE 9.4 Effect of degree of grafting on the water contact angle for the aramide membrane. (Reprinted from European Polymer Journal, 36, Teng, M.Y., Lee, K.R., Liaw, D.J., Lin, Y.S., and Lai, J.Y., Plasma deposition of acrylamide onto novel aromatic polyamide membrane for pervaporation, 663–672, Copyright (2000), with permission from Elsevier.)

280

Membrane Modification: Technology and Applications

other is more hydrophilic (with CO2). The effect of this modiication on the permeametric properties was investigated by liquid water and liquid toluene permeation measurements. The results showed two opposite effects of the two different treatments. CF4 plasma treatment led to a reduction of water and toluene permeabilities. With CO2 plasma treatment, in terms of permeation, two different behaviors were observed, an increase and a decrease in permeancy for water and toluene, respectively. The results were in full agreement with those obtained for the surface characterization (contact angle measurement, XPS and AFM) and conirmed the change in the polymeric surface afinity for the permeant, leading to a variation in materials permeancy (Dreux et al. 2003). Krea et al. (2004) developed new copolyimide membranes with high siloxane content designed to remove polar organic compounds from water by PV. Two series of segmented polydimethylsiloxane-imide (PSI) copolymers were synthesized from α,ω-(bisaminopropyl) dimethylsiloxane oligomers (ODMS) and aromatic dianhydrides (PMDA, 6FDA). Synthesized PSI was obtained with a high content of siloxane block up to 94 wt%. To improve the average molecular weight of siloxane copolymers prepared from the longer-chained ODMS, the particular extender 1,3-bis(3-aminopropyl) tetraethyldisiloxane (MDMS) was used, thus giving the materials with enhanced mechanical properties. PV results showed that the PSI led to a selective recovery of the organic compounds from their aqueous mixtures. PSI having a large number of siloxane blocks gave the best PV results, and optimum PV performances were obtained with membranes incorporating aromatic luorinated residues ((feed = 5 wt% PhOH): α = 22 and J10μm = 2.4 kg/m 2 h; (feed = 10 wt% EtOH): α = 10.6 and J10μm = 0.56 kg/m 2 h). PV results showed clearly that the triluoromethyl groups strongly enhanced the EtOH or phenol luxes. Khayet and Matsuura (2004) studied the separation of chloroform–water mixture via PVDF lat membrane by using both PV and VMD techniques. Both PV and VMD membranes were prepared using the phase inversion method and the same polymer material. VMD membranes with different pore sizes were prepared using pure water as a pore-forming additive in the PVDF/dimethylacetamide (DMAC) casting solution, whereas PV membranes were obtained with higher polymer concentration, without nonsolvent additives (water) and with solvent evaporation before gelation. A comparative study was made between both. Hirabayashi (2002) used regenerated cellulose and chitosan membrane for the removal of ammonia from aqueous solution of urine component. The removal of ammonia through the chitosan membrane was from 57% to 59%. Adsorption of ammonia from the downstream vapor by silica gels was carried out. Ji et al. (2005) described a process for enriching a crude ammonia stream having moisture therein by PV. Membranes were made from various polymers, including polyvinylidene luoride (PVDF), Telon (PTFE), and polypropylene. Removal of ammonia by PV technique, especially sweeping/stripping PV, is common. Cole and Genetelli (1971) reported that PV through hfs has great potential for the treatment of water and waste water containing ammonia. Tan et al. (2006) used PVDF hf membranes for ammonia removal from water by using stripping PV technique. PVDF hf membranes with asymmetric structures and good hydrophobicity have been

Development of Membranes for Pervaporation

281

prepared by the phase inversion method and have been applied to the removal of ammonia from water. A mathematical model was presented to stimulate the ammonia removal in PVDF hf modules. It was observed that the posttreatment with EtOH was useful to improve both the hydrophobicity and the effective surface porosity of the resulting PVDF hf membranes and thus favors the ammonia removal. Aqueous solution containing sulfuric acid was used as stripping solution to accelerate the removal of ammonia. González-Rodriguez et al. (2002) developed an automatic low injection PV method for monitoring the ammonia in wine. Vorotyntsev et al. (2006) separated ammonia by using absorbing PV technique. Different types of membranes including rubbery and glassy polymers were used. Pure water, glycol, PE glycol, and water solution of PE glycol with different molar masses (up to 2000) were used as absorbents. It was reported that it is possible to achieve a high degree of ammonia puriication from most of the impurities by the absorption PV method. Mixa and Staudt (2008) used unmodiied as well as cross-linked ethylene– methacrylic acid (E-MAA) copolymers with different amounts of methacrylic acid for the separation of phenol–water mixtures with concentrations between 3 and 8 wt% in the feed. These researchers used cross-linking agents, aluminum acetyl acetonate, which leads to ionically cross-linked membranes, and 2,3,5,6-tetramethyl-1,4-phenylene diamine and glycerine diglycidylether, leading to covalently cross-linked membranes. In general, it was observed that with increasing phenol content in the feed, the total lux was increasing, whereas the enrichment factor was decreasing. With nonmodiied membranes with higher methacrylic acid monomer content in the polymer, lower luxes and higher enrichment factors were noted. For the ionically cross-linked E-MAA copolymer membrane, a strong increase in the enrichment factor and an exponential increase in the lux with increasing temperature were observed. Yamaguchi et al. (1991, 1998) proposed a new type of membrane for liquid separation, called a pore-illing membrane. This membrane is composed of two materials: a porous substrate and a illing polymer that ills the pore of the substrate. The porous substrate is completely inert to organic liquids, and the illing polymer is soluble in one component in the feed. The illing polymer exhibits permselectivity due to a solubility difference, and the porous substrate matrix prevents the swelling of the illing polymer due to its mechanical strength. The pore-illing membrane can be prepared by the plasma graft polymerization technique. These membranes show high selectivity and permeability for organic mixture separation. Using ethylacrylate (EA) or BA as the grafting monomer, illing polymerized membranes were prepared with the aid of an emulsion system by the plasma graft polymerization technique (Yamaguchi et al. 1994). The illing membrane was composed of two different polymers: porous high density PE substrate and grafted polymer illing the substrate pores. Both monomers were grafted eficiently with the surfactant, and it was veriied that the grafted polymer illed the substrate pores. PV separations of dilute aqueous solution of chloroform or 1,1,2-TCE through these membranes were carried out. The chloroform selectivity of the BA-grafted membrane was higher than that of the EA or methylacrylate membranes. For TCE, BA-grafted membrane showed high permselectivity and lux. A permeation rate

282

Membrane Modification: Technology and Applications

of 0.232 kg/m 2 h and a separation factor of 1096 were achieved for a feed stream consisting of 0.135 wt% TCE. Yamaguchi et al. (2001) used hf-type pore-illing membranes made by plasma graft polymerization for the removal of chlorinated organics from water via PV. Laurylacrylate (LA)- or n-BA-grafted layers were formed inside the porous hf substrate, and the pores were illed with the grafted chains formed from plasma-initiated graft polymerization. The grafted layer thickness of both lat sheet and hf substrates were around 20 μm. The hf-type LA-grafted membranes showed extremely high separation properties; a 0.09 wt% TCE aqueous solution was condensed to 99 wt% TCE in the permeate. The membrane can remove TCE from a water stream, and at the same time, the membrane can purify the TCE for reuse. The membrane also showed high separation performance for an aqueous dichloromethane (DM) solution. In another study, Yamaguchi et al. (1996) used more hydrophobic polymers such as poly(ethylhexylacrylate) (EHA), poly(LA), and poly(stearylacrylate) (SA) for the grafted polymer. Porous high-density polyethylene (HDPE) was used as a substrate. TCE was used as a model organic of halogenated hydrocarbon contaminants. The hf-type LA-grafted membranes showed extremely high separation properties; a 0.09 wt% TCE aqueous solution was condensed to 99 wt% TCE in permeate. The membrane also showed high separation performance for an aqueous DM solution. Recovery of ethyl acetate (EA) from aqueous solution was done by using ethyl vinyl acetate (EVA) copolymer with 30 wt% of vinyl acetate content membranes (EVA38) via PV (Bai et al. 2007). Bai et al. (2007) investigated the effect of casting solvent and posttreatment on PV performances of EVA38 membranes for recovery of EA from water. The membrane showed high performance of both separation factor and permeation lux. It was proposed by them that the casting solvent changes the chain shape of EVA38 in solvents and alters its physical structure and PV performance. This effect of casting solvent gradually vanished with enhancing posttreatment, that is, increasing the temperature and time of annealing. Satyanarayana et al. (2006) used EC, ABS, and hydrophobic modiied EC membranes (dense) for hydrazine–water system PV. Hydrophobic modiied ethylcellulose membrane (ECNCO) was prepared by reacting ethylcellulose with isocyanate. These membranes were used in the PV of a hydrazine–water system. Both are highly polar liquids. The results are given in Table 9.2. The PV lux through the EC membrane is somewhat larger compared with ECNCO and ABS. These data were correlated with TABLE 9.2 Pervaporation of Hydrazine–Water (Concentration of Hydrazine: 64%; T = 50°C; Downstream Pressure: p + 0.1 mmHg; Batch Mode) Membranea EC ECNCO ABS a

Flux (g/m2 h)

Selectivity

9.65 9.35 7.81

1.72 1.87 5.07

Thickness of the membrane was 50 μm.

Development of Membranes for Pervaporation

283

the free volume of the membranes, which were measured by a positron annihilation technique. Reddy et al. (2008) cross-linked sodium alginate (SA)- and chitosan (CS)-blended membranes, pretreated with calcium chloride, with maleic anhydride for the separation of a 1,4-dioxane water mixture at 30°C by PV. The membrane performance exhibited a reduction in selectivity and an improvement in lux due to increased swelling with increasing feed water compositions. It was claimed by Reddy et al. that the membrane has a good potential for breaking the aqueous azeotrope 1,4-dioxane.

9.4.3

ORGANOPHILIC MEMBRANE

Separation of aromatic hydrocarbons from aliphatic hydrocarbons is a very important target in the membrane separation process of organic–organic mixtures. In particular, it is dificult to separate benzene from CYH in a mixture, due to their close boiling points and molecular sizes (Lu et al. 2006; Yildirin et al. 2001). Separation of organic–organic mixtures is also one of the most important processes in the chemical and petrochemical industries. Currently, thermal processes such as distillation or liquid–liquid extraction are mainly applied. But there are problems regarding close boiling points or formation of azeotropes. Furthermore, high amounts of extractive agents are necessary in the case of liquid–liquid extraction and they must be further processed. The application of organophilic PV for the separation of organic–organic mixtures has attracted signiicant attention from industry, which is now considering PV as a potential technology. Smitha et al. (2004) have given an overview to the research activities in the ield of the separation of organic– organic mixtures by PV. Membranes made by poly(vinyl chloride) (PVC) were used for PV separation of benzene–CYH mixture. Polymer was dissolved in THF to give 4 and 8 wt% solution at room temperature. PVC exhibits a strong afinity for benzene only. The casted ilms were dried under a nitrogen atmosphere for 3 days and then completely dried under reduced pressure. Thus, the membrane made by PVC was more permeable to benzene than to CYH. Increasing the concentration of benzene resulted in increasing lux as well as decreasing selectivity. Flux depended on the amount of polymer in the casting solution. Thickness of the polymeric membranes affected the permeation rate (Yildirin et al. 2001). Tabe-Mohammadi et al. (2001) studied the effects of solvent in the casting solution on the performance of CA membranes in methanol–methyl tert-butyl ether (MTBE) separation. The polymeric solvents were AC, dimethylformamide (DMF), and N-methylpyrrolidone (NMP). It was observed that the lux and selectivity of the membrane were affected by the type of solvent used to prepare the casting solution. The sample with DMF consistently gave the highest selectivity and lowest lux, followed by the samples with NMP and AC. The selectivity increased and the permeability decreased with increasing polymer concentration in the casting solution. The degree of CA acetylation inluenced the PV properties for methyl alcohol (MeOH)– MTBE mixture (Cao et al. 2000). It was reported that the acetylation degrees had great inluence on the PV properties because the quantity of hydroxyl groups in the polymer matrix leads to different levels of crystallinity.

284

Membrane Modification: Technology and Applications

Qian et al. (2005) used CA lat sheet membranes, prepared by dissolving CA in different solvent mixtures, for separating MeOH–MTBE mixture with 5 wt% MeOH via PV. The CA PV membranes were prepared from AC, AC/tetrahydrofuran (THF), AC/chloroform, and AC/CYH. The lux of the CA membranes from the solvent mixture was always higher than the lux of the CA membranes from the pure solvent AC for separating MeOH–MTBE mixtures. Solvent mixtures changed the morphology of the CA membrane surfaces. New copolymer membranes of acrylonitrile for the separation of benzene–CYH mixtures by PV were developed by Ray et al. (1997). The monomers in the copolymers were selected on the basis of their solubility parameter values relative to those of benzene and CYH. These were styrene, methyl methacrylate, and vinyl acetate. Copolymers of acrylonitrile with methyl methacrylate and vinyl acetate showed good selectivity and moderate lux: 60–70 and 0.075 kg/m2 h, respectively, with a membrane of 10 μm thickness for a feed mixture containing 5% benzene. Copolymer of acrylonitrile with styrene showed comparatively higher lux but lower selectivity. Ray et al. claimed that the selectivities obtained with these membranes were better than those reported in the literature. Matuschewski and Schedler (2008) showed that molecular surface engineering (MSE)-modiied membranes are capable of separating organic mixtures by organophilic PV. MSE technology allows the development of a large spectrum of composite membranes and the solution of many separation problems. MSE is a surface functionalization of porous membranes in an innovative procedure to design surfaces with desired properties. There are different ways for the functionalization of porous membranes, which are schematically presented in Figure 9.5. In one case, the entire outer surface is functionalized, resulting in membranes for nanoiltration, ultrailtration, and microiltration. In the other case, the pores are illed with a functional polymer. This leads to dense membranes for PV or gas separation. Commercially available UF membranes are used as support for the surface modiication to produce membranes for the organophilic PV. A thin, defect-free, cross-linked polymeric functional matrix is covalently anchored in the separation layer of the support. Based on the application of the pore-illing concept, the swelling of the functional polymer is negligible. Kuznetsov et al. (2001) used composite membranes, coated with poly(2-dimethylamino ethyl)methacrylate (PDMA) and ladder polyorganosiloxane (LP) onto microporous aromatic PAI support and PAN, as PV membranes for separating mixtures of methanol and MTBE. It was noticed that the PAI and PAN supports of the multilayer membranes have opposite effects on the total selectivity. High membrane permeability of the multilayer membranes was attained by virtue of the thinness of the interlayer or coating diffusion layer of PDMA with ultrahigh molecular weight. Initiated by ceric ammonium nitrate, PVA-graft-PAA membrane was prepared by graft polymerization of acrylic acid onto PVA (Yoshikawa et al. 1999a). These membranes permeated benzene preferentially from benzene–CYH mixtures. Permselectivity toward benzene increased with the increase in the PAA content in the membrane, and the separation factor reached around 10. Matsui and Paul (2003) separated aromatic–aliphatic hydrocarbons by a series of ionically cross-linked poly(n-alkyl acrylate) membranes using a PV technique.

285

Development of Membranes for Pervaporation

Functionalization process based on MSE technology

Membrane

Pore

Filtration membranes with pores for nanofiltration, ultrafiltration, and microfiltration

Composite membrane without pores for pervaporation gas permeation

FIGURE 9.5 Functionalization with MSE technology. (Reprinted from Matuschewski, H. and Schedler, U., Desalination, 224, 124–131, 2008. With permission.)

Copolymers of methyl-, ethyl- or n-BA with acrylic acid were synthesized and used to fabricate ionically cross-linked, via aluminum acetylacetonate, membranes for the separation of toluene–i-octane mixtures by PV at high temperatures. At 100°C for a feed containing 50 wt% toluene, the normalized lux of the permeate increased from 20 to over 1000 kg μm/m2 h in going from the methyl to n-BA, while the toluene over i-octane selectivity decreased from 13 to about 2.5. Uragami et al. (1998) synthesized benzoyl chitosan as a membrane material for the separation of benzene–CYH mixtures via PV. When the benzoyl chitosan membrane was applied for the separation of benzene–CYH mixtures in PV, both the permeation rate and benzene concentration in the permeate increased with increasing benzene concentration in the feed and thus this membrane showed benzene permselectivity. It was reported that the benzene permselectivity was dependent on both the sorption and diffusion selectivity but was signiicantly governed by the latter. A tentative model for benzene permselectivity was discussed. Huang et al. (2001) used chitosan/anionic surfactant complex membranes for the PV separation of methanol–MTBE. When anionic surfactants were added into the cationic chitosan solution, the viscosity was drastically decreased due to the collapsed chain conformation. High lux and good separation eficiency were achieved

286

Membrane Modification: Technology and Applications

for the chitosan–surfactant complex membrane. Huang et al. (2001) suggested that the chitosan–surfactant complex membrane has potential application as a PV membrane for separating alcohol–organic mixtures. Yoshikawa’s research group (Yoshikawa and Kitao 1997; Yoshikawa et al. 1997, 1999b, 2000; Okushita et al. 1995) developed novel nylon-6 membranes for PV separation of aromatic–aliphatic mixtures. It was reported that the polyoxyethylenegrafted nylon-6 membrane (membrane G790) was effective for a selective separation of cyclohexanone and cyclohexanol from a CYH–cyclohexanone–cyclohexanol mixture by PV. Membrane G790 showed high permselectivity toward cyclohexanol over the CYH when the feed mole fraction of cyclohexanol was 0.9. From the FTIR analysis, it was revealed that there was a hydrogen-bonding interaction between the hydroxyl group in cyclohexanol and that in the polyoxyethylene grafting chain. Aromatic hydrocarbons can be separated from mixtures with nonaromatics by PV through a membrane cross-linked polyconjugated diene rubber containing from 15 to 50 wt% C = N groups. The thin ilm of nitrile rubber can be prepared by using a casting solution containing 1–25 wt%, most preferably 5–18 wt%, of the rubber in a dissolving solvent. The casting solution is spread on a suitable substrate for support, permitting the solvent to evaporate and then cross-linking the polymer by exposing it to 140°C–200°C from 1 to 12 h (Black 1989). Schucker’s (2001) invention relates to a method for the separation of a multicomponent feed stream into a permeate stream rich in one or more components of the feed stream and a retentate stream lean in those same components. The study on PV is mainly concentrated on binary mixtures; in fact, most products to be separated are multiple mixtures in the petrochemical industry, such as methanol–MTBE–C5 and CYH–cyclohexanol–cyclohexanone. There is a large difference of PV performance between ternary and binary mixtures because of the more complicated effects accompanying ternary mixtures. The PV performances for methanol–MTBE–C5 ternary mixtures and corresponding binary mixtures through a CA membrane were reported by Zhang et al. (2002). It was found that there were very high PV properties for methanol–C5; the lux of methanol was greater than 430 g/m2 h. On the other hand, the lux of methanol–MTBE is lower than 90 g/m2 h. From the results of ternary mixtures, the lux of methanol decreased from about 450 to 100 g/m2 h with an increase in MTBE concentration in the feed from 5 to 50 wt%, and there was a strong accompanying effect between methanol and MTBE. PV of binary mixtures (CH2Cl2–CHCl3, CHCl3–CCl4, CH2Cl2–CCl4) through two amorphous copolymers of 2,2-bis-triluoromethyl-4,5-diluoro-1,3-dioxole and tetraluoroethylene was studied at different temperatures, feed composition, and downstream pressures (Polyakov et al. 2004). It was noticed that the copolymer with a larger content of the dioxole comonomer and greater free volume (AF 2400) was more permeable, whereas the selectivities of the two copolymers were similar for some regimes. The strongest changes in PSI are induced by the content of CH2Cl2 in the mixtures and by temperatures in AF 2400 ilms. PV separation indices found for amorphous Telfons AF showed that these membrane materials, especially AF 2400, exhibit good performance for the separation of organic mixtures and offer an attractive combination of permeation lux and selectivity of separation.

Development of Membranes for Pervaporation

287

Kai et al. (2005) prepared cross-linked pore-illing membranes using plasma graft polymerization. An HDPE was used as a porous substrate (pore diameter = 0.02 μm) and the monomers were methyl acrylate (MA), vinyl acrylate (VA), and N-N′-methylene bis(acrylamide) (MBAAm). The PV of a 72/28 wt% chloroform–n-hexane mixture was used to examine the PV performance of the prepared membranes. The permeation rate was decreased and the separation factor increased as the concentration of the cross-linker in the monomer solution increased. This could be due to the restriction in the mobility of the grafted chain by cross-linking. PV properties of homogeneous membranes based on composites containing polyaniline (PANI) and copolymer of aniline with anthranilic acid (coPANI) in a matrix of aromatic PI were investigated by Pulyalina et al. (2010) for binary organic mixtures with azeotropic points. It was established that PI membranes exhibited the highest selectivity to methanol during the separation of methanol–toluene mixtures. Addition of coPANI in the PI matrix leads to an increase in the hydrophilic properties and consequently the equilibrium sorption of water during separation of water– IPA mixture. PI–coPANI exhibited higher transport properties, lux, and selectivity, as compared with those of PI membrane. Polotskaya et al. (2010) developed polymeric membranes modiied by fullerene C60 for PV. Two polymers, PPO and poly(phenylene isophthalamide), were modiied by fullereneC60. The PV of the reacting mixture of EtOH–acetic acid–water– ethylacetate by fullerene–PPO membranes and that of the methanol–CYH mixture azeotropic point by fullerene–PA membranes were demonstrated on these new membranes. Assis et al. (2005) extracted and concentrated the volatile compounds of the coffee beverage via PV using a lat membrane of ethylene propylene diene terpolymer (EPDM). Shepherd et al. (2002) used PDMS hfs for orange juice aroma recovery. Zhang et al. (2006) used CA membranes for PV separation of a methanol–C5 mixture.

9.4.4

MEMBRANES INCORPORATION OF INORGANIC–ORGANIC PARTICLES

9.4.4.1 Inorganic–Organic Particles Composite Membranes Inorganic–organic membranes are also known as hybrid membranes. The simplest approach to the formation of hybrid inorganic membranes is to disperse the inorganic material in the polymer solution. 9.4.4.1.1 Hydrophilic Membranes Zhang et al. (2008) prepared organic–inorganic hybrid membranes with high separation performance by incorporation of polysilisesquioxane (PSS) into a PVA matrix in order to solve the trade-off relationship between the selectivity and the permeability. The incorporation of the PSS resulted in a change in the physical and chemical structures of the hybrid membranes. The crystalline region in the hybrid membranes decreased with increasing PSS content. The hydrophilicity of the hybrid membranes increased when the PSS content was below 3 wt% and then decreased. The membranes were used for the dehydration of THF (aqueous THF solution, 90%). The permselectivity and the lux of the hybrid membranes increased when the PSS content

288

Membrane Modification: Technology and Applications

was below 2 wt%, whereas the permselectivity decreased when the PSS content was above 2 wt%. The hybrid membrane containing 2 wt% PSS had the highest separation factor of 1810. Using a sol–gel technique, organic–inorganic hybrid membranes were developed by Varghese et al. (2010) that consisted of chitosan and mixed silica precursors such as tetraethoxysilane (TEOS) and γ-glycidoxypropyltrimethoxysilane. The γ-(glycidoxypropyl)trimethoxy-silane (GPTMS) acted as a coupling agent to enhance the compatibility between the organic (chitosan) and the inorganic (TEOS) phases. These membranes were used to separate water–IPA mixtures by PV in the temperature range of 303–323 K. Both the lux and the selectivity increased simultaneously when increasing the amount of GPTMS. The membrane containing 0.25 mass fraction of GPTMS exhibited the highest separation selectivity of 18,981 with a thickness-normalized lux of 7.45 × 10−7 m kg/m 2 h at 303 K for 0.5 mass fraction of water in the feed. With an increase in temperature, the permeation lux increased while decreasing the selectivity and this was attributed to decreased interaction between the permeants. The activation energy values obtained for water permeation were signiicantly lower than those of IPA, suggesting that the developed membranes had an excellent separation performance for water–IPA systems. Kreiter et al. (2011) demonstrated that large numbers of organic–inorganic hybrid silica (HybSi) sols can be prepared and screened, allowing tailoring of sol structures toward expected membrane performance in PV. Kreiter’s group also demonstrated the broad applicability of HybSi nanosieve membranes in the separation of binary (bio) alcohol–water mixtures (Kreiter et al. 2009). It was also discovered that membranes based on a mixture of 1,2-bis(triethoxysilylethane (BTESE) and methyltriethoxysilane (MTES) have unprecedented lifetimes of up to 2 years in the dehydration of n-butanol at 150°C (Castricum et al. 2008a,b). Membranes prepared from only BTESE showed a further improved separation performance with separation factor (α) in the range of 1,000–10,000 and water luxes of 10–20 kg/m2 h for this separation (Castricum et al. 2008c). Liu et al. (2007) also developed chitosan/poly(tetraluoroethylene) (PTFE) by casting GPTMS-containing chitosan solution on poly(styrene sulfuric acid)-grafted expended PTFE ilm surfaces. The adhesion between the chitosan skin layer and the PTFE substrate was good enough to give high performance of the chitosan/PTFE composite membranes for dehydration of IPA by PV. The chitosan/PTFE membrane exhibited a permeation lux of 1730 g/m2 h and a separation factor of 775 at 70°C on PV dehydration of a 70% IPA aqueous solution. The membrane maintained performance after a long-term test of 45 days. Kariduraganavar et al. (2004) developed novel PV membranes for the separation of water–IPA mixtures using sodium alginate and NaY zeolite composite membrane. Incorporation of NaY zeolite in sodium alginate (sodium salt of alginic acid, NaC6H7O6) had shown signiicant improvement in the membrane performance while separating water–IPA mixtures. An increase in the zeolite content in the membrane resulted in a simultaneous increase in both the permeation lux and the selectivity. The highest separation selectivity was found to be 614.33 with a lux of 14.49 × 10−2 kg/m 2 h for a higher loading of zeolite in the membrane at 30°C for

289

Development of Membranes for Pervaporation

5 mass% of water in feed. Experimental data revealed that the total lux and the lux of water were close to each other throughout the investigated range, indicating that the NaY zeolite–loaded membranes were highly water-selective, which was also predicted by the diffusion study. The research of Wolter et al. (2004) was related to the development of semipermeable membranes comprising organically modiied silicic acid polycondensates and a process for preparing them for their use in gas exchange and in separation techniques, especially in gas separation, dialysis, and PV. The membranes of their invention can be lat or tubular. The membranes are cured by addition polymerization or polyaddition of the C=C double bonds. A high-performance hybrid PV membrane with superior hydrothermal stability was developed by Castricum et al. (2008a). They developed a new organic–inorganic hybrid membrane with exceptional performance in dewatering applications. The membrane consists of organically bridged silica moieties and has an open microporous structure. The precursor used in the sol–gel synthesis of the selective layer was originally linked 1,2-bis(triethoxysilyl)ethane (BTESE). In the dehydration of n-butanol with 5% water, the membrane showed a high separation factor of over 4000 and ultrafast water transport at a rate of more than 20 kg/m2 h at 150°C. Kittur et al. (2003) studied the separation of water–IPA mixtures using ZSM-5 zeolite–incorporated PVA membranes via PV. Membranes were prepared by mixing zeolite in a PVA solution in water. The amount of zeolite in the polymer was varied as 2, 4, and 6 mass% and the resulting membranes were designated as M-1, M-2, and M-3, respectively. Table 9.3 shows the data for PV lux and separation selectivity for different mass% of water in the feed mixture at 30°C for different membranes. Selectivity increased signiicantly from membrane M-1 to M-3 due to a reduction in free volume as well as packing density with increasing zeolite content. Highest separation selectivity was found to be 216 for 6 mass% zeolite loading in the membrane at 30°C, without sacriicing the permeation lux signiicantly. The PSI data also indicated that the higher the degree of zeolite loading, the better is the membrane performance.

TABLE 9.3 PV Flux and Separation Selectivity Data for Different Mass% of Water in the Feed Mixture at 30°C for Different Membranes Mass% Water 10 20 30 40 50

αsep

Jp102 (kg/m2 h) M-1 0.45 1.86 3.24 2.00 1.50

M-2 0.34 1.50 2.85 1.55 1.23

M-3 0.32 0.89 2.39 1.26 1.07

M-1 91.09 15.06 5.71 2.07 0.32

Jp is the total lux and αsep is the separation factor (water over isopropanol).

M-2 141.14 26.31 18.00 8.07 4.97

M-3 216.22 46.00 30.25 11.00 5.65

290

Membrane Modification: Technology and Applications

Cross-linked PVA membranes were also used for the PV separation of the liquid mixtures of water–acetic acid and water–acetic acid–n-butanol–BuAc (Liu et al. 2001). The permeation luxes of water and acetic acid as a function of the composition were studied. The esteriication of acetic acid with n-butanol catalyzed by Zr(SO4) ⋅ 4H2O was carried out at a temperature range of 60°C–90°C. It was noticed that mostly water, less acetic acid, much less n-butanol, and actually no BuAc permeated through the membrane during PV separation of the quaternary mixture of water–acetic acid–n-butanol–BuAc. From the results obtained from this study, Liu et al. (2001) developed a kinetic model equation for the esteriication; then, it was taken as a model reaction to study the coupling of PV with esteriication. Chen et al. (2001b) reported the preparation and characterization of HY zeoliteilled chitosan membrane for the dehydration of organic–water mixtures using PV. A lux of 353 g/m2 h and a selectivity of 102 were obtained for an EtOH–water mixture with 10 wt% of water at room temperature, when the HY zeolite content in the membrane was 20 wt%. Liu et al. (2005) used cross-linked organic–inorganic hybrid chitosan membranes for PV dehydration of IPA–water mixtures. Cross-linked organic–inorganic hybrid chitosan membranes were obtained from blending chitosan and GPTMS in acetic acid aqueous solution. The hydrophilicity of the modiied membrane was not signiicantly decreased, which resulted in good permselectivity and high permeation lux in PV dehydration of a 70 wt% IPA–water solution. A lux of 1730 g/m2 h and a separation factor of 694 were observed with the chitosan membrane containing 5 wt% GPTMS. Both of the cross-linked and the organic–inorganic hybrid structures contributed to stabilize the membranes to maintain their performance in a long-term operation for 140 days. They claimed that the performance of the studied membrane was superior over other reported chitosan-based membranes, which will ensure its commercial application. Swelling of poly(vinyl alcohol-co-acrylic acid) (P(VA-co-AA)) membranes in aqueous alcohol solution operated under PV conditions leads to low water–EtOH selectivity. To reduce swelling, organic–inorganic hybrid membranes composed of P(VA-co-AA) and tetraethoxysilane (TEOS) were introduced by Uragami et al. (2005). When an aqueous EtOH solution was permeated through P(VA-co-AA)/ TEOS hybrid membranes by PV, the permeation rate increased and the water–EtOH selectivity decreased with increasing TEOS content. The increase in the permeation rate and the decrease in the water–EtOH selectivity were caused by an increase in the degree of swelling of the membrane and a decrease in the membrane density with increasing TEOS content. These effects resulted from insuficient formation of hydrogen bonds between the silanol groups by hydrolysis of TEOS and the hydroxyl and carboxyl groups of P(VA-co-AA). When the P(VA-co-AA) hybrid membranes were annealed, the water–EtOH separation factor increased with increasing annealing time and TEOS content. Bifunctional organic–inorganic charged nanocomposite membrane for dehydration of EtOH by PV was used by Tripathi et al. (2010). Chitosan was modiied to N-p-carboxy benzyl chitosan (NCBC) by introducing an aromatic ring grafted with an acidic–COOH group to stabilize the structure, and cross-linked nanostructure NCBC–silica composite membranes were prepared for PV dehydration of a

Development of Membranes for Pervaporation

291

water–EtOH mixture. These membranes were tailored to have three regions, namely a hydrophobic region, a highly charged region, and a selective region in which a weak acidic group (–COOH) was grafted at the organic segment, while a strong acidic group (–SO3H) was grafted at the inorganic segment to achieve high stability and less swelling in a water–EtOH mixture. Among the prepared membranes, the nanocomposite membrane with 3 h cross-linking time and 90% (w/w) of NCBC–silica content (PCS-3-3) exhibited 1.66 × 10−4 cm3 (STP) cm/cm2 s cmHg water permeability (Pw) with 1.35 × 10−7 cm3 (STP) cm/cm2 s cmHg EtOH permeability (PEtOH) and 1231 PV selectivity factor at 30°C for separating water from 90% (w/w) EtOH mixture. Song and Hong (1997) prepared a tubular membrane for the dehydration of EtOH and IPA in PV. The membrane was formed either on the inner or on the outer surface of a porous ceramic support with CA by a dip-coating and rotation-drying technique. In PV using a membrane coated on the outer surface of the ceramic support, the selectivities of water were in the range of 4–11 in the EtOH–water system and 8–240 in the IPA–water system. Li et al. (2008b) improved the PV separation performance of an organic– inorganic hybrid membrane by adding zeolite 13X into the PI(BAPP-BODA), which was synthesized using one-step polycondensation polymerization of 2,2-bis(4-(4-aminophenoxy)phenyl)propane (BAPP) with bicycle(2.2.2)oct-7ene-2,3,5,6-tetracarboxylicdianhydride (BODA). Compared with the BAPP-BODA membrane, the BAPP-BODA/13X hybrid membrane had good PV performance for dehydrating aqueous alcohol solutions. In the long-term stability examination, the PV of a 70 wt% IPA mixture at 25°C could be maintained for at least 180 days for the BAPP-BODA/13X hybrid membrane. Ahmad et al. (2005) developed novel NH4Y zeolite–illed chitosan membranes for the dehydration of water–IPA mixture via PV. The presence of NH4Y zeolite in the chitosan membranes caused nonhomogeneous dispersion of the NH4Y zeolite crystals and membrane swelling due to its hydrophilic properties. The presence of NH4Y zeolite increased the total permeation lux and the separation factor simultaneously. The PSI showed that 0.2 wt% of NH4Y zeolite– illed membrane gave the optimum performance in the PV process. Gao et al. (1996) prepared composite hydrophilic membranes, consisting of KA, NaA, CaA, and NaX zeolites and PVA polymers. The PV and separation characteristics of different alcohol–water systems through these membranes were studied at temperatures ranging from 20°C to 50°C. Performance of hydrophilic PV membrane (PVA) was enhanced by adding zeolite iller to the membrane. It was also observed that zeolite facilitates the permeation of smaller molecules, but resists that of larger molecules. Thus, it improved the lux and separation factor of the membrane. AC–water separation by PV on KA and CaA zeolite-illed PVA membranes gave better separation results than that of the EtOH–water system. It was also reported that zeolite-illed membranes perform better in PV-aided reactions because of the enhancement of the water transport of the membranes. Hilmioglu reported that a zeolite 13X–illed CA membrane may have better PV performance than a zeolite 4A–illed CA membrane. Since the zeolite 13X–illed CA membrane had a lower swelling degree, it will have lower PV lux and higher selectivity (Hilmioglu 2009). Hilmioglu (2009) suggested that the application of

292

Membrane Modification: Technology and Applications

CA membranes in industrial-scale PV units may be feasible for the separation of EtOH–water mixtures. EtOH–water azeotropy especially will be broken by PV. Using distillation and PV hybrid systems, bioethanol can be produced economically. 9.4.4.1.2 Hydrophobic Membranes In modern membrane technology, the hydrophilic character of ceramic membranes can be changed to hydrophobic by grafting the hydrophobic molecules on the surface. Kujawski et al. (2007a) studied the PV properties of luoroalkylsilane (FAS)grafted ceramic membranes. FAS are a group of compounds that can be eficiently used to create the hydrophobic character of different surfaces. The grafting process may be performed by a reaction between –OH surface groups of the ceramics and ethoxy groups (–O–Et) present in organosilane compounds. In order to prepare the hydrophobic ceramic membranes, 1H,1H,2H,2H-perluorodecyltriethoxysilane molecules C8F17C2H4Si(OEt)3 (C8 solution—Lancaster) were grafted on the surface of ceramic membranes by a condensation reaction. In addition, Kujawski et al. (2007a) also grafted on two types of commercial NF tubular membranes with an average pore diameter of 5 nm. The irst membrane possessed a support layer made of alumina (denoted after grafting as FAS-Al/x). The selective layer of the other one (FAS-Ti/x) was prepared from titania. Both FAS-Al/x and FAS-Ti/x were prepared by immersing the membranes in C8 solution at a temperature of 25°C for a given period of time. FAS-grafted membranes were tested in PV of water–electrolyte and water–organic solvent solutions. It was found that the transport and selective properties of grafted membranes depend on the type of ceramics and the grafting conditions. In the case of PV of water–electrolyte solutions, only water was found in the permeate. The PV lux was inversely proportional to the salt concentration. Both FAS-Al/x and FAS-Ti/x membranes showed high selectivity toward the organic component of water–organic mixtures. The selectivity factors of FAS-Ti/x (grafted in 0.1M C8), FAS-TiO2 (grafted in 0.01 M C8), and FAS-Al/x (grafted in 0.01 M C8) membranes in contact with water–MTBE (2 wt% MTBE) solution were 95, 45, and 23, respectively. Yoshida and Cohen (2003) developed asymmetric tubular alumina-supported poly(vinyl acetate) (PVAc) and poly(vinyl pyrrolidone) (PVP) membranes for the separation of binary mixtures of methanol and MTBE by using the PV technique. The active separation layer was created by free-radical graft polymerization of PVAc and PVP onto a vinylsilane-modiied alumina substrate with an average native pore diameter of 50 Å. The separation layer consisted of a surface phase of terminally anchored polymer chains with an estimated radius of gyration about a factor of 4.5–6.8 times larger than the membrane pore radius. The ceramic-supported polymer (CSP) membranes were found to be methanol-selective for the PV of methanol– MTBE mixtures, achieving separation factors up to 100 (PVAc) or 26 (PVP). Total permeate lux attained with the PVAc- and PVP-based membranes ranged from 0.055 to 1.26 and 0.55 to 6.19 kg/m2 h. Their results suggested that the CSP membrane can be tailored for speciic separations of organic mixtures with the proper choice of polymer chemistry. However, increase in selectivity also required careful optimization of grafted polymer chain spacing and chain length relative to the support pore size.

Development of Membranes for Pervaporation

293

Peng et al. (2007) prepared novel nanocomposite PV membranes (PVA–CNT(CS)) composed of PVA and chitosan-wrapped carbon nanotubes. The PV performance of PVA–CNT(CS) nanocomposite membranes was investigated using permeation lux and separation factor as evaluating parameters. For benzene–CYH (50/50, w/w) mixtures at 323 K, the permeation lux and the separation factor of pure PVA membrane were only 20.3 g/m2 h and 9.6, respectively, while the corresponding values of PVA–CNT(CS) (CNT content: 1%) nanocomposite membrane were 60.93 g/m2 h and 53.4. Cohen (2002) developed a ceramic-supported polymeric membrane. The porous ceramic membrane support of average pore size not larger than 500 Å was activated by attaching a vinyl-terminated lower alkoxy silane to the surface of the ceramic membrane pores. These ceramic-supported membranes are useful for PV separation of liquid mixtures that are suficiently different in their vapor pressure. Jou et al. (1999) developed a novel asymmetric ceramic-supported polymer (CSP) PV membrane by using free-radical graft polymerization of PVAc onto a porous tubular silica substrate. The membrane was used to remove TCE and chloroform from dilute aqueous solution via PV. These membranes can be tailored for a variety of separation applications through a proper selection of the graft polymer phase and optimizing polymerization conditions. The active separation layer in these membranes consists of a thin macromolecular layer of terminally anchored polymer chains. The enrichment factor of separation of TCE and chloroform from water varied from 69 to 106. However, the enrichment factor increased with increase in polymer graft. The mass transfer resistance from the membrane was found to be negligible compared with the resistance caused by concentration polarization on the feed side of the membrane when separating TCE and chloroform from aqueous solution. Song and Lee (2005) modiied the surface of alumina substrate with a silane coupling agent (perluoroalkyl silane). The membrane showed superhydrophobicity and the water drop contact angle on the surface-modiied alumina membrane was about 162°. Using PV technique, recovery of esters, volatile organic lavor compounds, EA, propyl acetate, and BuAc was attempted with these modiied membranes. The luxes of EA, propyl acetate, and BuAc at 0.6 wt% (6000 ppm) feed concentration and 40°C were 254, 296, and 318 g/m2 h, which were much higher than those obtained with polymer membranes. Compared with nonporous PDMS, the surface-modiied membranes showed a much higher lux and enough selectivity of esters. Table 9.4 shows a comparison of polymer membranes with the surface-modiied membranes. Yoshida and Cohen (2004) studied the removal of MTBE from MTBE–water mixture by PV using ceramic-supported polymeric membranes (CSP). The resulting membranes consisted of PVAc chains terminally and covalently anchored to the membrane surface. These membranes were prepared by the free-radical graft polymerization of vinyl acetate onto a vinyltrimethoxysilane-activated porous silica substrate with a native average pore size of 500 Å. The separation factor varied from 68 to 577 with a corresponding total lux range of 0.31–0.72 kg/m2 h, depending on the Reynolds number and the extent of membrane modiication. The PVAc– silica membrane had negligible resistance to MTBE permeation, while the feed-side concentration boundary layer provided the major resistance to MTBE transport.

294

Membrane Modification: Technology and Applications

TABLE 9.4 Comparison of Polymer Membranes with the Surface-Modified Membrane Flavor EA

Membrane PDMS GFT* PDMS DC* This study

Butyl acetate

PDMS GFT* This study

Condition 90–4800 ppm, 25°C 100 ppm, 30°C 0.15–0.6 wt% (6000 ppm), 30°C–50°C 10–15 ppm, 5°C 0.15–0.6 wt% (6000 ppm), 30°C–50°C

Flavor Flux (g/m2 h)

Enrichment Factor β

1.1–5.8

85.0–145.0

1.0 31.9–380.3

368.0 48.5–62.8

— 103.7–317.6

112.0 97.7–101.5

From Klamklang, S., Soontarapa, K., and Damronglerd, S., Sci. Asia, 28, 135–143, 2002.

*

Yosida and Cohen (2004) suggested that the separation performance of PVAc–silica membrane can be increased by reducing the water lux via denser surface chain coverage and by increasing the grafted polymer chain size relative to the native membrane pore size. A PV process characterized by excellent separation eficiency using a membrane that comprised an elastomeric polymer matrix containing zeolite was patented by Hennepe et al. (1990). A preferred group of such elastomeric polymers are silicone rubbers, especially polysiloxane rubbers and in particular polydimethylsiloxane rubber; the nitrilebutadiene rubbers (NBR); polyisobutylene, and the polyisoprene, and styrenebutadiene copolymer rubbers. Zeolites were incorporated into the membranes to make them as hydrophobic as possible. These membranes were particularly suitable for the separation of hydrocarbons, alcohols, esters, ethers, and amines from aqueous solutions containing these impurities by PV. PDMS is the most well-known membrane material for the extraction of VOC from aqueous waste stream by PV. Although it is quite permeable and selective to many VOCs in water, its selectivity can be improved further with appropriate zeolite illers. Such improvement may be needed for polar solutes such as aroma and fermentation products, whose high value makes the PV process attractive. Films of zeolite-illed silicone/PVDF composite membranes were prepared by incorporating zeolite particles into PDMS membranes (Zhan et al. 2009a). These membranes were used for the PV of EtOH–water mixtures. It was noticed that with the increase of zeolite loading from 10% to 30%, the lux increased signiicantly from 265.0 g/m2 h to 820.7 g/m2 h with 5% EtOH feed concentration at 50°C, and the separation factor increased from 11.3 to 13.7. The zeolite-illed PDMS/PVDF composite membrane with 30% zeolite loading was EtOH-permselective over a wide range of EtOH feed concentration (5–90 wt%), and especially showed excellent PV performance in the low concentration range.

295

Development of Membranes for Pervaporation

TABLE 9.5 Flux and Selectivity of Filled and Unfilled PDMS Membranes in the PV of 1.23% Ethyl Acetate–Water Mixtures at 50°C Membranes

Selectivity, α

Flux (Total) (g/m2 h 100 μ)

Flux (Water) (g/m2 h 100 μ)

Flux (Ester) (g/m2 h 100 μ)

PDMS 20% SiO2 20% SY-2 20% ZSM-5

182.4 188.6 239.0 249.5

320 292 341 314

97.5 87.0 85.6 76.4

22.5 205.0 255.4 237.6

Zeolite-illed hydrophobic membrane in a bioreactor for the separation of alcohol during fermentation was introduced by Hennepe et al. (1987). It was reported by them that adding a sorptive iller with a high selectivity toward alcohols to the membrane appears to improve both selectivity and lux. Yang et al. (2001) studied PV properties of different zeolite-illed PDMS membranes. They reported that the incorporation of hydrophobic zeolites into PDMS enhances the permeation selectivity toward the VOCs, but decreases the permeation rate in the corresponding membranes. The decrease in the permeation rate results from the cross-linking effect of the zeolite particles and also from the increase in the diffusion path; the zeolite particles act as solvent reservoirs in the sorption, but as obstacles for the permeation diffusion. Table 9.5 shows the lux and selectivity of illed and unilled PDMS membranes in the PV of 1.23% EA–water mixture at 50°C. Vankelecom et al. (1997a) studied three types of hydrophobic porous illers (carbon blacks, in situ methylated silicas, and silylated silicas) incorporated in PDMS membranes in order to ind out under which conditions these membranes were advantageous for PV of aqueous solutions. The properties of these illers were changed systematically in order to maximize the luxes and selectivities of the PV of aqueous EtOH, tert–butyl alcohol (TBA), or aroma solution. The effect of incorporation of carbon black in PDMS for the PV of a 6 wt% alcohol solution in water is based on the level of the PDMS. When using PDMS membranes illed with hydrophobic silicas, the best results were obtained with silylated silicas. Vankelecom et al. (1997b) investigated the PV of aroma compounds using zeolite-illed PDMS composite membranes. Zeolite-illed PDMS membranes were supported on the PAN asymmetric membrane coated on a nonwoven polyester. Table 9.6 shows the inluence of the iller on the luxes and overall enrichment factors. Zeolite Y clearly shows the most extreme results with an exceptionally high total lux, due entirely to the transport of water, leading to an extremely low enrichment factor. Among all other membranes, unilled PDMS is one of the best. For the more hydrophobic illers—USY and silicalite—the enrichment factors are high due to the exclusion of water from the pores. Study of the inluence of temperature revealed the importance of diffusion limitations on the transport of organics through the membranes. This effect was stronger when more zeolite was incorporated.

296

Membrane Modification: Technology and Applications

TABLE 9.6 Influence of the Filler on the Fluxes and Overall Enrichment Factors Filler

Organic Flux (g/m2 h)

Water Flux (g/m2 h)

Total Flux (g/m2 h)

Overall Enrichment Factor

PDMS USYa Zeo Y Sil

7.34 5.80 5.77 5.52

66.35 46.24 722.00 49.5

73.69 52.04 727.77 54.17

125 139 10 126

a

Ultrastable zeolite Y.

Adnadjević et al. (1997) studied the effect of three different types of hydrophobic zeolites (ultrastable zeolite type Y, pentasyl-type zeolite (ZSM-5), and ALPO-5 type zeolite) on the PV properties of zeolite-illed PDMS membranes. The physiochemical properties of the zeolite used, primarily the degree of hydrophobicity, as well as the sorption capacity for EtOH, the speciic pore volume, speciic area, and mean crystallite size of the zeolite, signiicantly inluence the membrane’s PV properties. An increase in the zeolite content results in an increase in both membrane permeability and membrane selectivity, while an increase in the PV temperature results in an increase in the permeability and a decrease in the selectivity, as opposed to the effect of membrane thickness. For the irst time, porous magnesium oxide (MgO) particles were used to generate mixed matrix membranes (MMM) for the dehydration of IPA via PV by Jiang et al. (2007). The precursor was a commercially available PI, which is Matrimid. It was observed that Matrimid/MgO MMMs generally have higher selectivity, but lower permeability relative to the precursor Matrimid dense membrane. The highest selectivity was obtained with MMM containing 15 wt% MgO. In the dehydration of IPA aqueous solution with 10 wt% water, the selectivity of MMMs was around 2000, which was more than twice that of the precursor polymeric membrane (900). This makes MMMs extremely suitable for breaking the azeotropes of water–IPA. Composite separation membranes of PDMS containing well-dispersed silicate particles of 50 nm were successfully prepared by Lu et al. (2001) and applied to the preferential PV of acetic acid over water. The nanocomposite membrane showed improvement on both the separation factor and the permeation lux for the PV process, as compared with plain PDMS membranes and composite membranes containing silicate particles of 5 μm. The improvement was attributed to higher readily accessible speciic surface area and higher sorption selectivity toward acetic acid of the nano-size silicate particles. Moermans et al. (2000) prepared PDMS membranes illed with colloidal silicate-1 and applied these membranes in the PV of EtOH–water mixtures. The nano-sized zeolites showed much improved PV results compared with the micron-sized silicate membranes. Both the lux and the selectivity were drastically increased at the highest loading. The best results were obtained with the PDMS membrane containing 40 wt% of silycated nano-sized silicate-1.

Development of Membranes for Pervaporation

297

Okumus et al. (1994) developed a mixed-matrix polymer–zeolite membrane for PV. In the preparation of these membranes, cellulose acetate (CA) as base polymer, AC, or DMF as solvent, and 13X or 4A zeolites as illers were used. It was observed that the addition of zeolite to the membrane matrix improved the lux value twofold with respect to its homogeneous membranes with some loss in their selectivity. For example, for a feed concentration of 74% EtOH at 50°C and 1 mmHg, the lux value for the unilled membrane was 0.6 l/m2 h, and for a 30% zeolite-illed membrane, the lux was increased to 1.33. For these cases, the selectivities were 7.76 and 5.0 for the unilled and illed membranes, respectively. Novel membranes for PV were prepared from carbon graphite (CG) and nylon 6 (N6) by Yoshikawa’s research group (Yoshikawa et al. 2000). Membranes were prepared from a mixture of 2,2,2-triluoroethanol (TFE) and CG. These composite membranes showed permselectivity toward benzene from benzene–CYH mixture. Benzene was permeated in preference to CYH from benzene–CYH mixtures and the separation factor of benzene reached 235 at the weight fraction of benzene in a feed of 0.1. Their work suggested that CG-nylon 6 composite membrane has the potential for petroleum-reining process. Sikdar et al. (2000) developed adsorbent-illed PV membranes for removing VOCs from waste water. These membranes were prepared by dispersing at least one hydrophobic adsorbent uniformly into a polymer matrix. Polymeric membrane was made of rubbery polymer selected from the group consisting of PDMSs, PTMSP, PUs, polycarbonates (PCs), PE-block-polyamides, silicon PCs, styrene butadiene rubber, nitrile butadiene rubber, and ethane–propene terpolymer. The hydrophobic adsorbent was selected from the group consisting of hydrophobic zeolites, hydrophobic molecular sieves, activated carbon, hydrophobic polymer resin adsorbents, and mixtures thereof. Highly hydrophobic siloxane–urethane membranes were introduced by Czerwiński et al. (2004) for the removal of volatile organic solvents by PV. Aqueous dispersion of siloxane–urethane copolymers (polysiloxane urethanes) containing 0–49% of siloxane moieties were synthesized by the modiied prepolymer–ionomer method. Dense membranes were prepared from these dispersions by cross-linking with a multifunctional aziridine derivative. Membranes were tested in vacuum PV of the following water–organic mixtures: water–MTBE (1.8 wt% MTBE) and water–BuAc (0.25 wt% BuAc) at a temperature of 313 K. It was observed that PV properties of polysiloxane–urethane membranes depended on the content of siloxane moieties. The best separation and transport properties were observed for the membrane containing 49% of siloxane groups. For this, the membrane separation factor was equal to 750 and 1370 for water–MTBE and water–BuAc mixtures, respectively. The ratio of organic and water partial molar luxes (Jorg/Jwater) was 2.8 and 0.5, respectively. 9.4.4.1.3 Inorganic Membranes Membranes that have high selectivity and high lux are not commonly available. Polymeric membranes have been limited to dehydration of solvents due to insuficiency of their thermal, mechanical, and chemical stabilities. The development of zeolite membranes has made it possible to overcome this limitation.

298

Membrane Modification: Technology and Applications

Inorganic membranes are widely used in UF and MF of liquid mixtures. Applications of inorganic membranes to PV, however, are still very limited, mainly because of insuficient permeation rates and/or insuficient selectivities of the membranes, although PV has attracted much interest because of its ability to separate azeotropic, close boiling, or aqueous organic mixtures with energy consumption (Fleming 1992). Inorganic membranes are versatile. They can be operated at elevated temperatures, with metal membranes stable at temperatures ranging from 500°C to 800°C and with many ceramic membranes operable at over 1000°C. A ceramic is an inorganic, nonmetallic solid prepared by the action of heat and subsequent cooling. Ceramic materials may have a crystalline or partly crystalline structure, or may be amorphous, for example, glass. Ceramic membranes are produced from inorganic materials such as aluminum oxides, silicon carbide, and zirconium oxide. Ceramic membranes are very resistant to the action of aggressive media such as acids and strong solvents. Zeolites are also used for making membranes. Zeolite is a naturally occurring mineral group consisting of over 50 different minerals. Made up of a special crystalline structure that is porous but that remains rigid in the presence of water, zeolites can be adapted for a variety of uses. Zeolite membranes are used for PV both industrially and in laboratory studies. A review has been written by Bowen et al. (2004) on the Fundamentals and Applications of Pervaporation Through Zeolite Membranes (more than 240 references). These membranes are polycrystalline zeolite layers deposited on porous inorganic supports, and they offer several advantages over polymeric membranes, which are as follows: 1. Zeolite membranes do not swell, whereas polymeric membranes do. 2. Zeolite has uniform, molecular-sized pores that cause signiicant differences in the transport rates for some molecules and allow molecular sieving in some cases. 3. Most zeolite structures are chemically more stable than polymeric membranes, allowing separations of strong solvents or low pH mixtures. 4. Zeolites are stable at high temperatures (as high as 1270 K in some zeolites) (Bekkum et al. 1994). Due to the previously mentioned advantages, zeolite membranes are attractive alternatives for separating mixtures whose components have adsorption or size differences but are dificult to perform using polymeric membranes and other conventional methods. Zeolite membranes usually have a higher separation factor than zeolite-illed membranes because the zeolite layer is continuous. Zeolite membranes can adsorb impurities during PV and also from the atmosphere during storage, and these impurities can signiicantly affect the permeation and separation performance. Zeolite membranes are both hydrophobic and hydrophilic in nature. Hydrophilic zeolite membranes, such as NaA, have effectively dehydrated alcohols with high separation factors, and hydrophobic zeolite membranes, such as silicate-1, have organic compounds removed from water (Bowen et al. 2004). Type A zeolite membranes are suited for organic dehydration because they are highly hydrophilic (Kondo et al. 1997; Okamoto et al. 1996, 2001; Morigami

Development of Membranes for Pervaporation

299

et al. 2001). Okamoto and coworkers (Okamoto et al. 1996, 2001; Morigami et al. 2001) and Van den Berg et al. (2003) developed zeolite membranes for PV. It is reported that zeolite A membranes synthesized on a UV-irradiated TiO2-coated metal support showed high PV performance for water–EtOH separation. At 45°C, the selectivities (54 × 103) and luxes (0.86 kg/m2 h) were obtained (Van den Berg et al. 2003). Their exceptional performance was ascribed to the pretreatment of the supporting titania with UV photons, which improves the hydrophilicity of the support, implying that the amount of surface Ti–OH groups increased. Okamoto et al. (1996) developed zeolite A type ilms for PV by depositing seed crystals on porous supports such as alumina, silica, zirconia, silicon carbide, stainless steel, PE, polypropylene, PI, PSf, and PTFE. Shah et al. (2000) used hydrophilic zeolite NaA membranes for the separation of alcohol–water, methanol–water, EtOH–water, IPA–water, and DMF–water mixtures. The total lux for EtOH–water mixture was found to vary from 2 to 0.05 kg/m2 h at 60°C as the feed solvent concentration was increased from 0 to 100 wt%. The total luxes for methanol–water and IPA–water mixtures were observed to vary from 2 to 0.15 and 2 to 0.21 kg/m2 h, respectively, as the alcohol concentration was changed from 0 to 100 wt%. Both water-to-EtOH and water-to-IPA separation factors were observed to lie between 1000 and 5000 over a wide range of solvent concentrations. The water-to-methanol separation factor was found to lie in the range of 500–1000. It was observed that ionic Na+ sites in the NaA zeolite matrix play a very important role in the water transport through the membrane. These sites act both as water sorption and transport sites. Peters et al. (2005a–d) developed a continuous composite catalytic PV membrane. Composite catalytic membranes were prepared by applying a zeolite coating on top of a ceramic hf silica membrane. The performance of the composite catalytic membrane was examined in the esteriication reaction between acetic acid and butanol (esteriication coupling). In the PV-assisted esteriication reaction, the catalytic membrane was able to couple catalytic activity and water removal. Zeolite membranes are widely used for the separation of alcohol. Kita et al. (1995) synthesized zeolite NaA membrane on porous ceramic support for separation of water–organic liquid mixtures by PV. Zeolite membrane was grown hydrothermally on the surface of a porous cylindrical substrate of α-alumina. Figure 9.6 shows the PV lux and the selectivity of zeolite membrane for water–EtOH mixtures as a function of the feed concentration at 75°C. Silica (SiO2) exhibits unique properties related to the ability of its SiO4 tetrahedral elemental structure to be connected together to give rise to a large number of different amorphous or crystallized solids that can be microporous, mesoporous, or macroporous. In comparison with other single oxides such as alumina (Al2O3), titania (TiO2), or zirconia (ZrO2), silica is easier to prepare as ultramicroporous or supermicroporous thin layers and thus used for molecular sieving applications. Zeolites are mainly silica or silica-based crystallized solids with structural ultramicroporosity. It is established that silicate membranes are hydrophilic. Sano et al. (1991) prepared pure silicate membranes on porous support of sintered stainless steel or alumina discs. Individual crystals were intergrown in three dimensions into the

300

Membrane Modification: Technology and Applications 5

105

104

3

αW/E

Q (kgm2 h)

4

2 103 1 0

0

50

60

70 80 XE (wt%)

90

100

FIGURE 9.6 (∙) Pervaporation lux and (⚪) selectivity of zeolite membrane for water (W)–ethanol (E) mixtures as a function of the feed concentration at 73°C. (Reprinted from Kita, H., Horii, K, Ohtoshi, Y., Tanaka, K., and Okamoto, K.I., J. Mater. Sci. Lett., 14(3), 206–208, 1995. With permission.)

polycrystalline phase. It was reported that pure silicate membranes exhibited high EtOH permselectivity with a separation factor (EtOH–H2O) of more than 60 for a 5 vol% aqueous EtOH solution at 30°C. Duke et al. (2008) used molecular sieving membranes to enrich lactic acid by selectivity, depleting water through the membrane via PV. Silica, molecular sieving, membranes exhibited a water selectivity factor up to 220 and a rejection coeficient of 0.995, with lactic acid in the permeate as low as 0.08 wt% after regeneration with an overall stable lux of 0.2 kg/m2 h. Peters et al. (2005b) developed microporous silica membranes on the outer surface of hf ceramic substrate. It was reported that PV performance of these membranes in the dehydration of n-butanol was better compared with that of commercially available tubular membranes. In the dehydration of n-butanol (80°C, 5 wt% water), initial high lux and selectivity were observed (2.9 kg/m2 h and 1200, respectively). A similar behavior was observed for the dehydration of DMF. Asaeda et al. (2002) developed porous silica–zirconia composite membranes for PV of aqueous organic solutions. The sol–gel methods have been applied for the fabrication of porous SiO2–ZrO2 composite membranes of various amounts of zirconia content to test their stability against water and to examine the PV performance for IPA–water and THF–water mixtures at normal boiling points. Zirconia was coated on the porous substrate by two methods: hot-coating and conventional dip-coating. The hot-coating procedures were effective for the fabrication of porous SiO2–ZrO2 composite membranes of high stability against water, while conventional dip-coating procedures gave a SiO2–ZrO2 (ZrO2: 50%) membrane of poor stability against water because of structural heterogeneity. The membranes ired at lower temperature (400°C) gave a higher lux of 450–500 mol/m2 h (9 kg/m2 h) and a separation factor greater than 1500 at 10 wt% of IPA (10 wt% of water) in PV of aqueous IPA solution at normal boiling point.

301

Development of Membranes for Pervaporation

Water flux (kg/m2 h)

14 12 10 8 6

Pervatech PVP 70°C

4

SiO2–ZrO2-50% (1) 75°C SiO2–ZrO2-50% (2) 75°C

2 0

SiO2–ZrO2-50% (3) 75°C

0

(a)

10 20 30 Water content in feed (wt%)

40

Water flux (kg/m2 h)

7 6 5 4 3

Pervatech PVP 50°C

2

Sulzer SMS 50°C

1

SiO2–ZrO2-50% (1) SiO2–ZrO2-50% (2)

0 0 (b)

10 20 30 Water content in feed (wt%)

40

FIGURE 9.7 Water lux as a function of water content in the feed through SiO2–ZrO2 50% membrane and commercial SiO2 membranes for the dehydration of (a) water–isopropanol and (b) water–acetone mixtures. (Reprinted from Urtiaga, A., Casado, C., Asaeda, M., and Ortiz, I., Desalination, 193, 97–102, 2006. With permission.)

Urtiaga et al. (2006) prepared several SiO2–ZrO2 50% tubular PV membranes by the sol–gel and hot-coating method. Several colloidal sols were prepared with tetraethoxy silane and zirconium tetra-n-butoxide as precursors, controlling the concentration (2.0, 1.5, 1.0, 0.8, 0.5 wt% alkoxides) by adding water and acid. Sols (SiO2–ZrO) were coated on a commercial tubular support made of alumina. Their PV performance was investigated regarding the separation of water–IPA and water– AC mixtures in terms of water lux and selectivity. Figure 9.7a shows the water lux as a function of water content in the feed for the SiO2–ZrO2 50% membranes during PV experiments of water–IPA at 75°C, and Figure 9.7b shows the water lux as a function of water content for the PV experiments of water–AC at 55°C. The results obtained using the two commercial silica membranes are included for comparison. These commercial membranes were Pervatech BV from Pervatech PVP, The Netherlands, and Sulzer SMS from Sulzer Chemtech GmbH, Germany. From Figures 9.7a and 9.7b, it can be observed that SiO2–ZrO2 50% membranes have a higher water lux than the commercial membranes, both for IPA and AC dehydration. Ishida et al. (2005) studied the PV of acetic acid–water mixture by porous silica membranes. Because of the small membrane thickness of 0.5–1 μm, the membrane

302

Membrane Modification: Technology and Applications

showed a larger water lux of about 400 mol/m2 h (7.2 kg/m2 h) with a separation factor of about 500 at 73 mol% (90 wt%) of acetic acid. Highly hydrophobic zeolite membranes, such as silicalite-1 (Sano et al. 1994), Ge-ZSM-5 (Li et al. 2001a), and β-type (Tuan et al. 2003), have been used for separating organic compounds from water. Matsuda et al. (2001) developed silicate PV membranes with a high separation factor for an EtOH–water system. The silicate membranes were prepared on porous supports of sintered stainless steel (2 μm) by hydrothermal synthesis. Hexane isomers and benzene–p-xylene mixtures have been separated by PV through an MFI-type zeolite membrane by Matsufuji et al. (2000). The PV tests for n-hexane, 2-methylpentane (2-MP), and 2,3-dimethylbutane (2,3-DMB) were performed using an MFI-type zeolite membrane at 303 K. n-Hexane preferentially permeated through the MFI membrane in the PV tests for binary mixtures of n-hexane–2-MP and n-hexane–2,3-DMB. The separation factors (α(n-hexane– 2-MP) and α(n-hexane–2,3-DMB)) were always greater than their ideal selectivities. The ideal selectivities of n-hexane–2-MP and n-hexane–2,3-DMB were 37 and 50, respectively. It was observed that separation factor α(n-hexane/2,3-DMB) was as high as 270 when the feed concentration of n-hexane was 10 mol%. Matsufuji et al. (2000) claimed that the MFI membranes have promising potential to separate n-hexane and branched hexane–isomer mixtures. Chen et al. (2005) synthesized high-reproducibility silicate-1 membranes on silica tubes, and all membranes showed high separation performance toward EtOH– water mixtures by PV. However, the membranes synthesized on alumina tubes showed much lower separation performance than the membranes on silica tubes; this was caused by the decrease in hydrophobicity resulting from the dissociation of α-alumina tubes during hydrothermal synthesis. It was concluded that silica supports are more suitable for preparing high-performance and high-reproducibility silicalite-1 membranes. A few studies have been made on PV with acidic feeds using zeolite membranes. Hydrophilic zeolites, in general, are not suitable in low pH environments because acid leaches Al from the framework. Zeolite membranes used for low pH PV, therefore, need to have relatively high Si/Al ratios so that the framework is not destroyed when Al is removed. Stainlesssteel supports are usually used for these applications because Al2O3 supports are susceptible to degradation by acids. A Ge-ZSM-5 membrane removed acetic acid from a 5 wt % acetic acid–water mixture at 363 K with α = 14 and a 16.8 mol/m2 h lux (Tan et al. 2006). Kalipcilar et al. (2002) reported that SSZ-13 zeolite membranes (CHA structure, Si/Al = 14) dehydrated HNO3, which is a stronger acid than acetic acid, and exhibited an azeotrope at approximately 69 wt% HNO3. The H2O–HNO3 separation factors were approximately 3.3 using a 69 wt% HNO3 feed at 298 K, and the luxes were about 2.8 mol/m2 h. About 5 wt% of initial Al was leached out from SSZ-13 powder during 3 days of exposure to 69 wt% HNO3, but a membrane continued to separate with a lux of 8.5 mol/m2 h and α = 2.6, even after 13 days of PV with approximately 69 wt% HNO3 feed. Li and coworkers (2001a,b,c, 2002) used silicate-1, ZSM-5, ZSM-11, mordenite, and X-, Y-, and β-type zeolite membranes to remove 1-3-propanediol from glycerol

Development of Membranes for Pervaporation

303

and glucose in aqueous mixtures. X-type zeolite has a faujasite (FAU) structure and its typical composition is Na2O ⋅ Al2O3 ⋅ 2.5SiO2 ⋅ 6H2O. It is used commercially as an adsorbent and as a catalyst. Li et al. (2001a–c) prepared X-type zeolite membranes by a template-free method on porous tubular support. The best membrane was prepared as one zeolite layer on a γ-Al2O3 support, and it had a tri-isopropyl benzene PV lux of 2.3 g/m2 h at 300 K. The membrane separated 1,3-propanediol from glycerol in aqueous mixtures by PV (Li et al. 2001c, 2002). An Na-ZSM-5 zeolite membrane was effective for separating 1,3-propanediol from glycerol and glucose by PV (Li et al. 2001a). In another study to see the effects of zeolite membrane structure on the separation of 1,3-propanediol from glycerol and glucose by PV, Li et al. (Li et al. 2001b) synthesized seven types of zeolite membranes on the inside surfaces of alumina and stainlesssteel supports. It was observed that the zeolite structure had a signiicant effect on the PV lux; the larger pore membranes had higher luxes. At 308 K, the X-type membrane had the largest 1,3-propanediol lux of 62 g/m2 h, with a 1,3-propanediol/glycerol selectivity of 59. Mordenite membranes were prepared by seeded hydrothermal synthesis onto commercial ceramic tubular supports by Casado et al. (2003) and used for the PV of alcohol–water mixtures. It was reported by them that selective adsorption of water on zeolite pores and small intercrystalline defects controlled the separation mechanism in the mordenite. Wang et al. (2009) modiied the zeolite surface to prepare high-performance hydrophilic zeolite LTA (Linda Type A) PV membranes on ceramic hfs by dipping–wiping seed deposition. Zeolite LTA membranes on hf supports were made by direct but very careful rubbing of powdered seed crystals (average size 0.8 μm). Wang et al. (2009) reported that by one single hydrothermal synthesis, a zeolite LTA membrane with a high lux of 9.0 kg/m 2 h and high water–EtOH separation factor of 10,000 could be formed on a ceramic hf. Pera-Titus et al. (2008) studied the modeling PV of EtOH–water mixtures within “real” zeolite NaA membranes and described their dehydration behavior. It was suggested that there is a certain role of grain boundaries between adjacent zeolite NaA crystals to mass transfer. Grain boundaries might behave as fast diffusion paths due to anisotropy of the zeolite layer. Kita et al. (1997) noticed that an NaY zeolite membrane grown hydrothermally on the surface of a porous alumina substrate showed high alcohol selectivity in PV. The membrane could be used for liquid separation process of alcohol–benzene, CYH, or MTBE. Zeolite membranes have been successfully applied to remove water from aqueous solutions at the large scale. Morigami et al. (2001) synthesized NaA zeolite membranes composed of continuous intergrowth of NaA crystals on the surface of a porous tubular support (12 mm outer diameter, 80 cm length, and 1 μm average pore size). The membrane was highly selective for permeating water preferentially with the high permeation lux in PV. The irst large-scale PV plant with this membrane produces 530 l/h of solvents (EtOH, IPA, MtOH) at less than 0.2 wt% of water from 90 wt% solvent at 120°C. Microporous silica and zeolite membranes are being developed for the separation of water from organic process stream by PV (Gallego-Lizon et al. 2002; Jonquières et al. 2002). By incorporating methyl groups in the silica membrane structure, water

304

Membrane Modification: Technology and Applications

stability in PV can be improved. vanVeen et al. (2004) found that these membranes can also be used in the separation of methanol from organics such as THF, toluene, MTBE, and several alcohols by PV. Experimental results showed that the luxes obtained from these membranes were much higher than for polymeric membranes with good selectivities and that the membranes can be used up to at least 125°C. Kita et al. (2003) reported on a tubular-type PV and vapor permeation module with zeolite membranes for fuel EtOH production. They used two types of zeolite membranes: (i) NaA-type zeolite membrane, which was grown on the surface of a porous cylindrical mullite support; and (ii) T-type zeolite membrane, which was also grown hydrothermally on the mullite support. Both membranes were studied for the lux and the separation factor of PV and vapor permeation for water–alcohol mixtures at 50°C and 75°C. The membranes were selective for permeating water preferentially with the high permeation lux. The separation factor of the T-type zeolite membrane was slightly smaller than the NaA zeolite membrane. They also claimed that this can provide more energy-eficient concentration of the EtOH to fuel grade EtOH. Yang et al. (1999) prepared two types of hydrophobic zeolites Y by treating the NaY-type zeolite with SiCl4, with or without subsequent hydrothermal treatment. It was reported that the hydrophobic zeolite Y as iller had a signiicant effect on the silicone membrane properties, even at a iller content level as low as 5 wt%. In addition, the ester sorption, permeation selectivity, and lux of the illed membranes increased with the iller Si/Al ratio in the EA extraction from water by PV. To improve the afinity between silicalite and PDMS, silicate-1 particles were modiied by a novel silane coupling agent vinyltriethoxysilane (VTES) and incorporated into the PDMS matrix for the preparation of silicalite/PDMS hybrid membranes. The effect of silicate loading on the PV performances of the hybrid membranes with dilute EtOH solutions was investigated. As compared with the unmodiied hybrid membranes, the silane-modiied silicate/PDMS hybrid membranes effectively improved the PV selectivity at different silicate loadings. With increasing silicate loading, membrane selectivity increased for both unmodiied and VTES-modiied silicate/PDMS hybrid membranes and a selectivity of 33 for EtOH separation from dilute aqueous solution was obtained when VTES-modiied silicate loading was 67%. It was also found that with increasing silicate loading, the total lux of both hybrid membranes increased. With increasing the feed EtOH concentration at a given temperature, the total lux, EtOH lux, and EtOH concentration in the permeate increased almost proportionally, while the separation factor decreased slightly and the water lux decreased indistinctively at lower feed temperature and sharply at higher temperature. An increase in temperature increased the permeation lux (Wan et al. 2009).

9.4.5

MULTILAYER COMPOSITE MEMBRANES

Recently, researchers’ interest is increasing toward multilayer composite membranes for PV. Surfaces of PV membranes were modiied by formation of an ultrathin polyion complex layer (Kusumochyo et al. 2002, 2004) to separate the organic isomers via PV.

Development of Membranes for Pervaporation

305

Top layer

Support layer

Middle layer

FIGURE 9.8 Schematic drawing cross section of a three-layer structure composite membrane.

An ultrathin polyion complex (PIC) layer containing β-cyclodextrin (β-CD) was formed on the surface of a charged base membrane. A positively charged copolymer containing β-CD was synthesized by the radical copolymerization of β-CD monomer and allylamine and was used to modify the surface of a Naion membrane containing negatively ixed ions. The membrane showed good selectivity toward butanol isomers. Wang et al. (2000) studied three-layer composite membranes for the separation of alcohol–water mixtures. The top layer was a thin ilm of chitosan cross-linked with glutaraldehyde, and the support layer was made of microporous PAN. Between the dense and microporous layers, there was an intermolecular cross-linking layer (Figure 9.8). The performance data showed that this was an excellent PV membrane for alcohol dehydration. However, the structure and the performance of these novel composite membranes varied with conditions of membrane preparation, such as hydrolysis degree of PAN membrane, content of cross-linking agent, and heat-curing temperature. The results indicated that the separation factor and the permeation rate of this novel composite membrane increase with the increase in operating temperature. At the same time, the PV properties can be adjusted by changing the structure of the top layer and the middle layer. The membrane had a better PV performance of a separation factor more than 1410 and 0.33 kg/m2 h for the 90 wt% EtOH aqueous solution and a separation factor of 5000 and lux of 0.43 kg/m2 h for the 90 wt% i-PrOH aqueous solution. PV separation of EtOH–water mixtures was done by using composite membranes prepared by coating a thin ilm of a polystyrene sulfonate (PSSf) across the surface of a microporous alumina support membrane (PSSf/Al2O3) (Chen et al. 1995). These membranes were able to break up the azeotrope of the EtOH–water mixture by PV. Water was the preferred permeate for all membranes and all feed compositions. Membranes in the Na+ counterion form (PSSf-Na) showed higher separation factors than membranes having Mg2+ as the counterion (PSSf-Mg). The separation factor for the membrane of 10.5 mol% sulfonate was one order of magnitude greater than the membranes of 27.5 mol% sulfonate. The membranes of 10.5 mol% of sulfonate showed overall permeate luxes one order magnitude lower than the membranes of 27.5 mol% sulfonate. The highest separation factor (400) was observed for the 10.5 mol% PSSf-Na/Al2O3 composite membrane at the azeotropic feed composition.

306

Membrane Modification: Technology and Applications

EtOH weight fraction in permeate was always less than 6 wt% for this composite membrane and all feed compositions. For the irst time, silica-illed poly(1-trimethylsilyl-1-propyne) (PTMSP) layers on top of UF membranes for the pervaporative separation of EtOH–water mixtures was reported by Claes et al. (2010). Reduction of the thickness of the separating PTMSP top layer and addition of hydrophobic silica particles resulted in a clear lux increase as compared with dense PTMSP membranes. The performances of the supported PTMSP–silica nanohybrid membranes were signiicantly better than the best commercially available organophilic PV membranes. The developed composite PTMSP–silica nanohybrid membranes exhibited EtOH–water separation factors around 12 and luxes up to 3.5 kg/m2 h, establishing a sevenfold to ninefold lux increase as compared with dense PTMSP membranes. A thin, high lux, and highly selective cross-linked PVA water-selective layer on top of hf ceramic supports was prepared by Peters et al. (2006). The support was an α-Al2O3 hf substrate and an intermediate γ-Al2O3 layer, which provided a suficiently smooth surface for the deposition of ultrathin PVA layer. The thickness of the PVA layer formed on top of the γ-Al2O3 intermediate layer was on the order of 0.3–0.8 μm. In the dehydration of 1-butanol (80°C, 5 wt% water), the membrane exhibited a high water lux (0.8–2.6 kg/m 2 h), combined with a high separation factor (500–10,000). The values for the lux and separation factor exceeded typical values obtained for cross-linked PVA membranes on polymeric supports. In the dehydration of 2-propanol and 1-butanol, a simultaneous increase in both the water lux and the separation factor was observed with increasing temperature or water concentration. Microporous silica membranes prepared by sol–gel processing and comprising a three-layer system consisting of a support prepared from α-alumina powder, γ-alumina intermediate layer, and a molecular sieving silica top layer have been described by Benes et al. (2000). The surface polarity of sol–gel materials can be controlled by copolycondensation of MeSi(OR)3, with Si(OR)4. A multilayer PDMS/PVDF/nonwoven iber/PVDF/PDMS composite membrane was prepared by Zhan et al. (2010) and used for the recovery of EtOH from its aqueous solution via PV. The porous PVDF substrate was obtained by casting PVDF solution on both sides of nonwoven iber with an immersion precipitation phase inversion method. Pure PDMS and cross-linking PTMOS were dissolved in n-hexane with vigorous stirring and the organometallic catalyst was added into the solution. The resulting homogeneous PDMS solution was coated onto both sides of the porous PVDF substrate one side after the other and dried at room temperature for 48 h. A schematic diagram of the coniguration of the multilayer PDMS/PVDF composite membrane is shown in Figure 9.9. The composite membrane signiicantly surpassed the limit of pure PDMS selectivity for EtOH, with the total lux decreasing a little compared with the single-layer PDMS/PVDF membrane. The maximum separation factor of the multilayer PDMS/ PVDF composite membrane was obtained at 60°C, and the total lux increased exponentially along with the increase in temperature. The composite membrane gave the best PV performance with a separation factor of 15 and a permeation rate of 450 g/m2 h with a 5 wt% EtOH concentration at 60°C.

307

Development of Membranes for Pervaporation PDMS PVDF

Nonwoven fiber

FIGURE 9.9 Schematic diagram of a multilayer PDMS/PVDF composite membrane.

Kuznetsov et al. (2001) used multilayer composite membranes for separating mixtures of methanol and MTBE. The membranes were prepared using microporous PAI as support with poly(2-methylaminoethyl)methacrylate (PDMA) and an LPas interlayer and coating diffusion layers, respectively. Commercial PAN was also used as support. It was noticed that PAI and PAN supports of the multilayer membranes have opposite effects on the total selectivity. High membrane permeability of the multilayer membrane was attained by the thinness of the interlayer or the coating diffusion layer of PDMA having ultrahigh molecular weight. Multilayer composite PV membranes were prepared via an interfacial polymerization reaction between ethylenediamine (EDA) and trimesoyl chloride (TMC) on the surface of modiied polyacrlonitrile (mPAN) (Huang et al. 2008). Positron annihilation spectroscopy study on these membranes revealed that the polymerized layer near the support surface gradually became dense with increasing EDA concentration, causing dificulty for the EDA monomer to penetrate through the polymerized layer in order to react with the acyl chloride. This resulted in a looser structure in the polymerized layer farther from the mPAN support surface. The resulting multilayer structure was composed of the top layer, which was EDA–TMC polyamide; the second layer, which was a transition layer of (EDA–TMC) + mPAN; and the third layer, which was the porous mPAN. A high-density EDA–TMC active layer was obtained at high concentration of EDA solution, resulting in a decrease in the permeation rate and an increase in the concentration of water in the permeate with an increasing EDA concentration from 0.5 to 10 wt%. Table 9.7 shows the effect of the concentration of aqueous EDA solution on PV performance of 90 wt% aqueous IPA mixture at 25°C. Budd et al. (2004) prepared multilayer membranes by altering the adsorption of cationic polyelectrolyte, chitosan, and an anionic polyelectrolyte, poly(4-styrene sulfonate) onto zeolite. The ilms were formed on porous supports in sheet and tube forms. The multilayer membranes were shown to exhibit high selectivity for water over alcohols associated with zeolite A but to be considerably more stable under acidic conditions than the pure zeolite A membrane. The membranes were successfully applied to the selective removal of water by PV in the esteriication of lactic acid with EtOH, catalyzed by p-toluenesulfonic acid. For esteriication at 70°C, yields of

308

Membrane Modification: Technology and Applications

TABLE 9.7 Effect of Concentration of Aqueous EDA Solution on PV Performance of 90 wt% Aqueous Isopropanol Mixture at 25°Ca EDA Concentration (wt%) 0.5 2 5 10 a

Permeation Rate (g/m2 h)

Water Concentration in Permeate (wt%)

384 ± 45 252 ± 32 213 ± 9 195 ± 37

52.5 ± 4.5 76.8 ± 2.3 92.1 ± 1.5 91.3 ± 1.8

Polyamide thin-ilm composite membranes (contact time of aqueous EDA solution: 30 min; concentration/immersion time of organic TMC solution: 1 wt%/3 min).

about 60% without the membrane were increased to >80% with PV, and for esteriication at 100°C, yields of about 70% without the membrane were increased to 90% with PV. Meier-Haack et al. (2001) developed composite membranes based on polyelectrolyte multilayer assemblies for the PV separation of water–alcohol. The supporting membrane was prepared from both unmodiied polyamide-6 and a comb-like polymer with carboxyl-terminated polyamide-6 side chains. Up to 20 layers were deposited onto the membrane surface by dipping the membranes in aqueous solutions containing oppositely charged polyelectrolytes. The polyanions used were PAA, poly(styrene sulfonic acid), and alginic acid. The polycations used were poly (diallyldimethylammoniumchloride), chitosan, and poly(ethylenimine). The performance of these membranes depends strongly on the layer number and on the type of polyelectrolytes. In general, membranes modiied with two weak polyelectrolytes of high charge density gave the best separation properties while those modiied with strong polyelectrolytes of low charge density led to poorer separation properties. However, the highest separation factor ≥10,000 for a water–2-propanol mixture (12/88 w/w) at a permeate lux of 300g/m2 h was obtained with six double layers of poly(ethylenimine) and alginic acid. These composite membranes were stable over an operating period of at least 400 h. Due to the broad variety of polyelectrolytes, the consecutive alternating adsorption of oppositely charged polyelectrolytes is an easily accessible and powerful method to tailor the membranes to the desired properties. van Veen et al. (2001) developed a new tubular microporous membrane based on hydrophilic silica for the dewatering of organic solvents. The support system for this particular membrane consisted of four layers and was basically made in the following way. The α-alumina macroporous support tubes that were used as structural carriers for the actual membrane were made by ceramic paste extrusion followed by a sintering procedure. Before the inal membrane layers could be applied, two intermediate layers were applied to the support. The intermediate layer was coated by using an α-alumina colloidal suspensions. After drying, a “gamma” layer was applied onto the second intermediate layer by slip-coating of a boehmite sol. After drying and during heat treatment, this boehmite was transformed to γ-alumina.

Development of Membranes for Pervaporation

309

The silica membrane, which was the inal separation layer, was made by means of sol–gel processing systems. A silicon alcoxide was hydrolyzed, from which a polymeric silica sol was obtained. This sol was coated onto the support followed by drying and calcination. All layers were applied on the outside of the tube. This membrane could be used above 100°C, even up to 300°C. The acid stability of the membrane was much better than zeolite A PV membranes. Peters et al. (2005c) developed composite catalytic membranes for PV-assisted esteriication processes. Catalytic zeolite H-USY layers were deposited on silica membranes by dip-coating using tetra ethyl oxysilane (TEOS) and Ludox AS-40 (colloidal silica, 40% suspension in water). In the PV-assisted esteriication reaction, the catalytic membrane was able to couple catalytic activity and water removal. The collected permeate consisted mainly of water, and the loss of acid, alcohol, and ester through the membrane was negligible as binder material.

9.4.6

PV MEMBRANE REACTORS

PV reactors are the systems in which the separation and the reaction are carried out simultaneously in order to increase the conversion of reactants by removing one or more products formed during equilibrium reactions. PV is a promising option to enhance the conversion of reversible condensation reactions in which water is formed as a by-product. Peters et al. (2005) prepared composite catalytic membranes by a dip-coating technique. Composite catalytic membranes have been prepared by applying a zeolite coating on top of ceramic hf silica membranes. In the PV-assisted esteriication reaction, the catalytic membrane was able to couple catalytic activity and water removal. A reactor evaluation proved that the outlet conversion of the catalytic PV-assisted esteriication reaction exceeded the conversion of a conventional inert PV membrane reactor, with the same loading of catalyst dispersed in the bulk liquid. Further, the performance of the zeolite-coated PV membranes can be increased by optimization of the catalytic layer thickness or by an increase in catalytic activity. Kiatkittipong et al. (2002) investigated a PV membrane reactor for the synthesis of ethyl tert-butyl ether (ETBE) from a liquid phase reaction between EtOH and TBA. Supported β-zeolite and PVA membrane were used as catalyst and as membrane in the reactor, respectively. The permeation studies of water–EtOH binary system revealed that the membrane worked effectively for water removal for the mixtures containing water lower than 62 mol%. The permeation studies of quaternary mixtures (water–EtOH–TBA–ETBE) were performed at three temperature levels of 323, 333, and 343 K. It was found that the membrane was preferentially permeable to water. Mori and Inaba (1990) applied a PV technique to attain both high productivity and eficient recovery of EtOH from a fermentation broth. The membrane bioreactor consisted of a jar fermenter and a PV system for the direct production of EtOH from uncooked starch with a thermophilic anaerobic bacterium, Clostridium thermohydrosulfuricum. From the four types of EtOH-selective membranes tested, microporous PTFE membrane, the pores of which were impregnated with silicone rubber, was chosen for its large lux, high EtOH selectivity, and high stability. During

310

Membrane Modification: Technology and Applications

feed-batch fermentation with PV in the membrane bioreactor, EtOH was continuously extracted and concentrated in two traps with concentrations at 5.6%–6.2% (w/w) in trap 1 (20°C) and 27%–32% (w/w) in trap 2 (liquid N2), while the EtOH concentration in the broth was maintained at 0.85%–0.9% (w/w). Zhu and Chen (1998) prepared cross-linked PVA composite catalytic membranes on porous ceramic plate for PV separation and PV–esteriication coupling. The composite catalytic membrane was evaluated through the PV and a model system of n-butyl alcohol esteriication coupling with the PV. The conversion of n-butyl alcohol reached 95% when a cross-linked PVA catalytic membrane was used. The order of permeation luxes was water > acid > alcohol > acetate and the total lux was greater than 0.5 kg/m2 h during the reaction time. The order of the separation selectivities of membranes was water–acetate > water–alcohol > water–acid. The parameters such as temperature, initial molar ratio of acid to alcohol, and catalyst concentration could be changed in order to attain the optimum of the PV–esteriication coupling operation. A polyimide/clay hybrid nanocomposite membrane was utilized by Wang et al. (2004) in the PV of aqueous EtOH mixtures. The polyimide/clay hybrid polymers were prepared by direct polycondensation of 4,4′-methylenedianiline (MDA) and 4,4′-hexaluoroisopropylidenedibenzoic acid (6FDAc) in the presence of organomodiied montmorillonite (organoclay) in N-methyl-2-pyrrolidinone (NMP). The organoclay (SDS-clay) was prepared using an ion-exchange reaction between the montmorillonite silicate layers and the sodium dodecyl sulfate (SDS, intercalating agent). The heat stability of the polyimide/SDS-clay nanocomposite membrane was higher than that of pure polyimide. Compared with pure polyimide membranes, the separation factor for the polyimide/SDS-clay membranes exhibited higher values in the 10%–90% EtOH feed concentration range. Hasanoğlu et al. (2009) used cross-linked hydrophobic PDMS membranes, which were permselective to EA, for the esteriication reaction of acetic acid and EtOH to produce EA and water, in a batch PV reactor. Temperature has a strong inluence on the membrane reactor performance because it acts in kinetics of both esteriication and PV.

9.4.7

PERVAPORATION BY ION-EXCHANGE MEMBRANE

Ion-exchange membrane is a membrane comprising ixed ions in its polymer matrix. It preferentially adsorbs and dissolves ions of opposite charge but repels those of the same charge. Membranes with ixed ions or dipoles will preferentially adsorb and dissolve molecules with dipoles, for example, water, but will repel nonpolar molecules. Ion exchange is a reversible chemical reaction between an insoluble solid and a solution during which ions may be interchanged. Ion-exchange membranes are a class of membranes that bear ionic groups and therefore have the ability to selectively permit the transport of ions through themselves. In biological systems, cell membranes and many other biological membranes contain ionic groups, and the conduction of ions is essential to their function. Synthetic ion-exchange membranes are used in fuel cells, electrochemical processes for chlorine manufacture and desalination,

Development of Membranes for Pervaporation

311

membrane electrodes, and separation processes. Ion-exchange membranes typically consist of a thin-ilm phase, usually polymeric, to which ionizable groups have been attached. Numerous polymers have been used, including polystyrene, PVA, PE, PSf, and luorinated polymers. Ionic groups attached to the polymer include sulfonate (SO3−), carboxylate (COO−), tetralkylammonium (N(CH3)4+), phosphonate (PO3H−), and many others. Ionic membranes can be used not only in separation processes involving ionic species such as Donnan dialysis, electrolysis, or electrodialysis, but also in the dehydration of organic solvents by PV, that is, as hydrophilic membranes, or in the facilitated transport of neutral solute in aqueous media via a complexation mechanism. Nguyen et al. (2003) prepared ionic membranes by blending PVA with an ionic polymer. Semi-interpenetrating polymer networks (sIPNs) in which the crosslinked PVA chains trap the ionic polymer were obtained by using a heat treatment at 180°C (for purely thermal cross-linking) or dibromoethane vapor at 140°C (crosslinking reagent). Poly(styrene sulfonic acid) (PSSH), poly(sodium styrene sulfonate) (PSSNa), PAA, or poly(dimethyl dimethylene piperidinium chloride) (PDMeDMPCl) was used as the ionic polymer. These membranes could be used in esteriication between n-propanol and propionic acid by PV. These tailored membranes also have other applications. The asymmetric aluminum ion-exchange PSf membranes were successfully used for the dehydration of EtOH–water mixtures (Chen et al. 2007a). The experimental results showed that the separation performance of those membranes was increased upon increasing the degree of aluminum ion exchange in PSf membranes. Lue et al. (2004) studied the PV of benzene–CYH mixtures using ion-exchange membrane (Neosepta–CMX exchange membrane, Tokuyama Corp, Japan) containing copper ions (II). This membrane contained 45%–65% sulfonated styrene/ divinylbenzene random copolymer and 45%–55% PVC. Results obtained with this membrane were compared with those of an Na(I)-containing membrane. During the PV experiments, the Na(I) and Cu(II) membranes demonstrated preferential transport properties for benzene throughout the entire feed composition range. The separation factors using Cu(II) membranes were consistently higher than those with the Na(I) form. PV of water–EtOH through synthetic polymer membranes having cationic charge sites was reported by Yoshikawa et al. (1986). Kujawski and Pozniak (2005) studied the PV properties of polyethylene– sulfonated polystyrene (PESS) ion-exchange membranes in contact with water– aliphatic alcohol mixtures. PESS ion-exchange membranes were prepared by chemical modiication of the IPN system PE–poly(styrene-co-divinylbenzene). PESS membranes were also loaded with different alkali metal ions (H+, Li+, or K+) as counterions. The experimental data showed that the properties of the PESS membranes depended strongly on the kind of counterions, degree of cross-linking, and difference in the polarities between water and the organic component of the binary mixtures. Results obtained for PESS membranes were compared with data obtained for a Naion 117 ion-exchange membrane. Kujawski et al. (2007b) also reported that diffusivity of EtOH was greatly affected by the presence of water in the membrane. The increase in alcohol polarity increases their diffusion properties.

312

Membrane Modification: Technology and Applications

Kujawski and Krajewski (2004) studied the selective and transport properties of an ion-exchange hf membrane (Pervasiv Ltd., Israel) in the sweeping gas PV used for dehydration of different organic solvents. The hf membranes were prepared by grafting anion-exchange groups on the PE. These membranes possessed hydrophilic properties. It was found that the separation factor increased exponentially with the decrease in the dielectric constant of the organic solvent. The molar partial luxes of the organic solvents were much higher for polar solvents (EtOH) when compared with solvents of lower polarity such as isobutanol and methyl acetate. Boucher-Sharma et al. (1999) studied the removal of n-butanol from aqueous solutions by ion-exchange membranes containing organic counterions. The PV butanol– water mixture was investigated by using thin-ilm composite membranes composed of a poly(vinylidene luoride) substrate coated with a sulfonated poly(2,6-dimethyl-1, 4-phenylene oxide) polymer. The polymer was ion-exchanged with quaternary ammonium cations having aliphatic substituents of various chain lengths. The membrane was alcohol-selective. The separation factor increased and the permeate lux decreased as the chain lengths of the aliphatic substituents were increased. They suggested that the mass transport properties of such membranes can be controlled or altered to yield desired permselectivity by the introduction of relevant counterions. Modiied Naion membranes, prepared by charging Naion 117 membranes with different long-chained counterions were used for PV of acetic acid–water mixture (Kusumochyo and Sudoh 1999). It was noticed that the selectivity of the Naion membranes was enhanced by charging with long-chained counterions. Among the modiied Naion membranes, Naion(C8H17)4 was the most suitable membrane with a lux of 0.18 kg/m2 h and a selectivity of 243 for a feed concentration of 90% acetic acid. Oren et al. (2004) developed a new method for preparing heterogeneous ionexchange membranes by the application of an electric ield during curing of the polymeric matrix, which resulted in the agglomeration of the ion-exchange particles in long linear chains extending across the membranes. As a result, the percolation threshold for ion conductance was reduced to a value more than three to four times smaller than those for membranes with randomly distributed particles. Membranes prepared by this technique will be expected to be more shape-stable, with good conductivities at lower water content and with improved ionic versus nonionic permeabilities. These features are important for both ED and fuel cell applications. Shevachman et al. (2001) bound jojoba wax to PE membranes and hfs for use in ion exchange and PV processes. To change the hydrophobicity of the material, chlorosulfonation with SO2 and Cl2 was used to introduce SO2Cl groups into the polymeric chain (PE–SO2Cl). These groups were hydrolyzed to form strong acids such as –SO3H, which was utilized in ion exchange processes. By binding jojoba wax, a hydrophobic material, to PE–SO2Cl, followed by hydrolysis, produced a modiied membrane with unique properties for ion exchange or dialysis processes. The modiied material in hf form was suitable for application in PV processes for the separation between the polar and nonpolar molecules. The PV experiment was conducted with hfs made of jojoba-bound PE–SO2Cl (41% (w/w) of jojoba) with dioxane–water separation. Dioxane and water are miscible and have very close boiling points (dioxane 101.3°C, water 100°C). Preliminary results of PV of a dioxane–water mixture

Development of Membranes for Pervaporation

313

showed the separation of the dioxane from the water with a separation factor of 6. This technique can be applied to remove residual organic solvents in the puriication of industrial wastewater. Choi et al. (2006) used cation-exchange membrane for dehydration by PV for HIx solution (HI–H2O–I2 mixture). The permeation lux was dependent on the I2 in the feed having an I2/HI molar ratio.

9.4.8

SULFUR REMOVAL FROM GASOLINE

The presence of sulfur in gasoline is one of the causes for pollution of the environment. PV techniques are also used in the removal of sulfur from gasoline. PV with PDMS membrane was proved to be an alternative route for the removal of sulfur impurities out of gasoline (Minhas et al. 2003; Saxton et al. 2004). Zhao et al. (2009) prepared PDMS/PEI composite membranes by casting a PDMS solution onto porous PEI substrate. The membranes were used for sulfur removal from gasoline by PV. Experimental results showed that higher feed temperature yielded higher total lux and lower sulfur enrichment factor. The total lux varied little with the increase in sulfur content in the feed, but the sulfur enrichment factor irst increased with the amount of thiophene added into gasoline, and then the variation became small. The results indicated that the PDMS/PEI composite membranes have promising future for desulfurization by PV. In another study, Zhao et al. (2008) reported data for the PV separation of n-heptane–sulfur species mixtures with PDMS/PEI composite membrane. Gasoline is a complex mixture containing various sulfur species. A series of sulfur species (thiophene, 2-methylthiophene, 2,5-dimethylthiophene, n-butyl mercaptan, and n-butyl sulide) were used by Zhao et al. (2008). Figure 9.10 shows the inluence of the feed sulfur content on the sulfur enrichment factor of different binary n-heptane–sulfur mixtures at the feed temperature of 353.15 K. It can be seen that variation in feed sulfur content had nearly negligible inluence on the sulfur enrichment factor for every sulfur species. As shown in Figure 9.11, partial luxes of sulfur species are proportional to the feed sulfur content. It was found that the different sulfur species had signiicant impact on the desulfurization eficiency. Cross-linked PDMS/PEI composite membranes were used in PV separation of thiophene–n-heptane, 2-methylthiophene–n-heptane, 2,5-dimethylthiophene–nheptane, n-butyl mercaptan–n-heptane, and n-butyl sulide–n-heptane mixtures (Chen et al. 2008). PV results indicated that as the feed temperature increased from 50°C to 90°C, the partial lux increased and the enrichment factor decreased. The partial lux and the enrichment factor of sulfur species had the same order: n-butyl sulide < n-butyl mercaptan < 2,5-dimethyl thiophene < 2-methyl thiophene < thiophene. As the organic S content increased from 50 ng/μl to 250 ng/μl, the partial lux of organic sulfur increased and the enrichment factor decreased slightly but the total lux was scarcely inluenced by the concentration of organic sulfur as very low content of organic sulfur led to minor contribution to the total lux. Qi et al. (2006a) developed PDMS/PAN membranes for sulfur removal from gasoline by PV. PDMS, ethyl orthosilicate, dibutyltin dilaurate, and n-heptane were used for the preparation of membranes, and asymmetric microporous PAN membranes

314

Membrane Modification: Technology and Applications n-Heptane/thiophene n-Heptane/2-methylthiophene n-Heptane/2,5-dimethylthiophene n-Heptane/n-butyl mercaptan n-Heptane/n-butyl sulfide

16

Sulfur enrichment factor

14 12 10 8 6 4 2 0

0

300

600 900 1200 Sulfur content in feed (ng/µl)

1500

FIGURE 9.10 Effect of feed sulfur content on sulfur enrichment factor for binary n-heptane–sulfur mixtures. (Reprinted from Separation and Puriication Technology, 63, Zhao, C., Li. J., Qi, R., Chen, J., and Luan, Z., Pervaporation separation of n-heptane/sulfur species mixtures with polydimethylsiloxane membranes, 220–225, Copyright (2008), with permission from Elsevier.)

were employed as supports. Solution of PDMS, cross-linking agent ethyl orthosilicate, and initiator dibutyltin dilaurate in heptane were cast onto the PAN membrane to form the skin layer. Experimental results of alkane–thiophene mixtures revealed that the total luxes decreased with an increase in carbon number in alkane, while the enrichment factor of thiophene increased simultaneously. With the membrane having a normalized PDMS layer of 15 μm, the total lux for n-octane–thiophene mixture was found to be about 1.04 kg/m2 h, with the corresponding enrichment factor of thiophene 4.4 at 31°C. PV results of n-hexene/n-hexane/thiophene demonstrated that the rise in the concentration of n-hexane in the feed would yield a larger total lux, but a smaller enrichment factor for thiophene at the same time. A quaternary mixture of n-heptane, n-octane, n-nonane, and thiophene was used to simulate the desulfurization process of gasoline. With the membrane having a PDMS layer of 11 μm, the total lux was about 1.65 kg/m2 h with the enrichment factor of thiophene 3.9 at 30°C (Qi et al. 2006b).

9.4.9

SUPPORTED LIQUID MEMBRANE FOR PERVAPORATION

PV by supported liquid membranes (SLMs) is a process for separating VOCs from their dilute aqueous solution. PV through supported liquid membranes has been suggested especially for the separation and concentration of fermentation products such as EtOH, butanol, diacetyl, and acetic acid (Ishii et al. 1995; Komada and

315

Development of Membranes for Pervaporation

40 Thiophene 2-Methylthiophene 2,5-Dimethylthiophene n-Butyl mercaptan n-Butyl sulfide

Partial flux (10–3 kg/m2 h)

35 30 25 20 15 10 5 0

0

300

600 900 1200 Sulfur content in feed (ng/µl)

1500

FIGURE 9.11 Dependence of partial luxes on feed sulfur content for binary n-heptane– sulfur mixtures. (Reprinted from Separation and Puriication Technology, 63, Zhao, C., Li. J, Qi, R., Chen, J., and Luan, Z., Pervaporation separation of n-heptane/sulfur species mixtures with polydimethylsiloxane membranes, 220–225, Copyright (2008), with permission from Elsevier.)

Honda 1985). Oleyl alcohol and isotridecanol were used as the liquid membranes. Compared with the large number of PV studies using solid membranes, investigations on liquid membranes for PV are rare. Liquid membranes have stability problems due to various losses. Sirkar et al. (1997) introduced hf-contained liquid membrane PV for the removal of VOCs from aqueous solutions. Sirkar (2009) patented liquid membrane systems for use in PV techniques that can achieve high selectivity, ensure stability, and prevent contamination of the fermentation broth. Tri-n-octylamine (TOA), trilaurylamine, or tri-decylamine as liquid membrane (Figure 9.12) was immobilized in the pores of a hydrophobic hf substrate. Further, a nanoporous hydrophobic Feed Coating layer Porous support Pore partially filled with liquid membrane Hollow fiber

Thin liquid membrane Fiber bore has a high vacuum Coating layer Feed

FIGURE 9.12

Representation of a porous hollow iber membrane with a nanoporous coating.

316

Membrane Modification: Technology and Applications

coating was applied on the broth side. The liquid membrane in the coated hfs demonstrated high selectivity and reasonable mass luxes of the solvents in PV. The mass luxes were substantially increased with the same selectivity of solvents when an ultrathin liquid membrane was used. On adding butanol into the feed solution, the selectivity of membrane was increased. Thongsukmak and Sirkar (2007) developed a new liquid membrane–based PV technique to achieve high selectivity, ensure stability, and prevent contamination of the fermentation broth. TOA as a liquid membrane was immobilized in the pores of a hydrophobic hf substrate having a nanoporous coating on the broth side and studied for the PV-based removal of solvents (AC, EtOH, and butanol) from their dilute aqueous solution. The liquid membrane (LM) of TOA in the coated hfs demonstrated high selectivity and reasonable mass luxes of the solvents in PV. The selectivities of butanol, AC, and EtOH achieved were 275, 220, and 80, respectively, with 11.0, 5.0, and 1.2 g/m2 h for mass luxes of butanol, AC, and EtOH, respectively, at a temperature of 54°C for a feed solution containing 1.5 wt% butanol, 0.8wt% AC, and 0.5 wt% EtOH. The mass luxes were increased by as much as ive times with a similar selectivity of solvents when an ultrathin liquid membrane was used. The TOA-based LM present throughout the pores of the coated substrate demonstrated excellent stability over many hours of experiment and essentially prevented the loss of liquid membrane to the feed solution and the latter’s contamination by the liquid membrane. Qin et al. (2003) used SLMs for the separation of acetic acid and butyric acid from their aqueous solution. Polypropylene hfs and silicon-coated, microporous hydrophobic polypropylene membranes were used as support. In another study, Qin et al. (2002) demonstrated PV by using a liquid membrane consisting of nonvolatile hydrocarbons immobilized in the micropores of hydrophobic hfs on the outer surface of the ibers. TCE was separated and concentrated from its aqueous solution at 25°C and essentially atmospheric pressure. The feed TCE concentration was varied between 50 and 950 ppm; the permeate pressure range was 0.6–70 mmHg. A 78-iber, 30–33 cm long module could achieve as much as 98% removal of TCE. It was reported by them that the hexadecane SLM was permselective for TCE: the experimental selectivity was 30,000 and the intrinsic selectivity could be as high as 2 × 105, much higher than the values obtained by any solid membranes. Christen et al. (2004) developed an SLM system for the extraction of EtOH during semicontinuous fermentation of Saccharomyces bayanus. The membrane was a porous Telon sheet as support, soaked with isotridecanol. The removal of EtOH from the cultures led to decreased inhibition and, thus, to gain in conversion of 452 g/l glucose versus 293 g/l glucose without extraction. At the same time, the EtOH volumetric productivity was enhanced 2.5 times, due to an improvement of yeast viability, while the substrate conversion yield was maintained above 95% of its theoretical value. In addition to these improvements in the fermentation performances, the process resulted in EtOH puriication, since the separation was selective toward microbial cells and carbon substrate and likely selective to mineral ions present in the fermentation broth. For PV, a concentration of EtOH four times greater was obtained in the collected permeates.

Development of Membranes for Pervaporation

9.5

317

CONCLUSION

PV technique is energy-eficient and full of environmental beneits. PV is an attractive technology because of the potential to selectively separate alcohols and water. It is important to note that phase change occurs in standard PV processes. PV is effective for dilute solutions containing trace or minor amounts of the component to be removed. Based on this, hydrophilic membranes are used for dehydration of alcohols containing small amounts of water and hydrophobic membranes are used for the removal/recovery of trace amounts of organics from aqueous solutions. PV is a very mild process and hence very effective for the separation of those mixtures that cannot survive the harsh conditions of distillation. PV continues to evolve as a feasible separation technology for many different applications. As a proven method of separation at low temperatures and pressures, further application development for food processing is likely. Using PV to clean wastewater streams by removing a variety of organic compounds also holds much promise. More speciically, PV can be used in the following: • Solvent dehydration: dehydrating the EtOH–water and IPA–water azeotropes • Continuous water removal from condensation reactions such as esteriications to enhance conversion and rate of the reaction • Removing organic solvents from industrial wastewaters • Combination of distillation and PV/vapor permeation • Concentration of hydrophobic lavor compounds in aqueous solutions (using hydrophobic membranes) Organophilic PV membranes can be used for the separation of organic–organic mixtures, for example: • • • • •

Reduction of the aromatic content in reinery streams Breaking of azeotropes Puriication of extraction media Puriication of product stream after extraction Puriication of organic solvents.

Hydrophobic membranes are often polydimethylsiloxane-based where the actual separation mechanism is based on the solution-diffusion model. The commercially most successful PV membrane system to date is based on PVA. More recently, membranes based on PI have also become available. To overcome the intrinsic disadvantages of polymeric membrane systems, ceramic membranes have been developed over the last decade. These ceramic membranes consist of nanoporous layers on top of a macroporous support. The pores must be large enough to let the water molecules pass through and retain any other solvents that have a larger molecular size such as EtOH. As a result, a molecular sieve with a pore size of about 4 Å is obtained. The most widely available member of this class of membranes is that based on zeolite. PV membranes with incorporation of inorganic–organic particles and multilayer membranes have a promising future.

318

Membrane Modification: Technology and Applications

REFERENCES Adnadjević, B., Jonaović, J. and Gajinov, S. 1997. Effect of different physicochemical properties of hydrophobic zeolites on the pervaporatin properties of PDMS-membrane. J. Memb. Sci. 136: 173–179. Ahmad, A.L., Mohd. Nawawi, M.G. and So, L.K. 2005. Development of novel NH4 zeolite illed chitosan membranes for the dehydration of water–isopropanol mixture using pervaporation. Sep. Sci. Technol. 40(15): 3071–3091. Aminabhavi, T.M., Kulkarni, P.V. and Kurkuri, M. 2006. Pervaporation membranes and methods of use. US 7045062. Aneja, A.P. 2006. Bicomponent hollow ibers for pervaporation. J. Ind. Textil. 36(1): 17–30. Asaeda, M., Sakou, Y., Yang, J. and Shimasaki, K. 2002. Stability of porous silica–zirconia composite membranes for pervaporation of aqueous organic solutions. J. Memb. Sci. 209: 163–175. Assis, A.R., Saraiva, S.H., Matta, V. et al. 2005. Recovery of coffee aromatic extracts by pervaporation. 2nd Mercosur Congress on Chemical Engineering, Village Rio das Pedras, Reo de Janerio, Brazil, August 14–18. Bai, Y.X., Qian, J.W., Zhao, Q., Zhu, Z.H. and Zhang, P. 2007. Effect of recovery conditions of poly(ethylene-co-vinyl acetate) membrane on the membrane physical structure and pervaporation properties in the recovery of ethyl acetate from aqueous solution. J. Memb. Sci. 299: 45–53. Bekkum, H., Geus, E.R. and Kouwenhoven, H.W. 1994. Supported zeolite systems and applications. Stud. Surf. Sci. Catal. 85: 509–542. Benes, N., Nijmeijer, A. and Verweij, H. 2000. Microporous silica membranes. In: Recent Advanced in Gas Separation by Microporous Ceramic Membranes, ed. N.K. Kanellopoulos, pp. 335–372. Elsevier: Amsterdam. Binning, R.C., Johnston, J. and Walker, F. 1961a. Aromatic separation process. US 2970106. Binning, R.C., Lee, R.J., Jennings, J.F. and Martin, E.C. 1961b. Separation of liquid mixtures by permeation. Ind. Eng. Chem. 53(1): 45–50. Black, L.E. 1989. Aromatics-non aromatics separation by permeation through thermally crosslinked nitrile rubber membranes. US 4885096. Boucher-Sharma, A.P., Chowdhury, G. and Matsuura, T. 1999. Removal of n-butanol from aqueous solutions by ion-exchange membranes containing organic counterions. J. Appl. Polym. Sci. 74: 47–58. Bowen, T.C., Noble, R.D. and Falconer, J.L. 2004. Fundamentals and application through zeolite membranes. J. Memb. Sci. 245: 1–33. Bruschke, H.E.A., Tusel, G.F. and Rautenbach, R. 1985. Pervaporation membranes: Application in the chemical process industry. ACS Symp. Ser. 281: 467–478. Buckley-Smith, M.K. 2006. The use of solubility parameters to select membrane materials for pervaporation of organic mixtures. Ph.D. Thesis, University of Waikato, Hamilton, New Zealand. Budd, P.M., Ricardo, N.M.P.S., Jafar, J.J., Stephenson, B. and Hughes, R. 2004. Zeolite/polyelectrolyte multilayer pervaporation membranes for enhanced reaction yield. Ind. Eng. Chem. Res. 43: 1863–1867. Cao, S., Shi, Y. and Chen, G. 2000. Inluence of acetylation degree of cellulose acetate on pervaporation properties for MeOH/MTBE mixture. J. Memb. Sci. 165: 89–97. Casado, L., Mallada, R., Téllez, C., Coronas, J., Menéndz, M. and Santamaría, J. 2003. Preparation, characterization and pervaporation performance of mordenite membranes. J. Memb. Sci. 216: 135–147. Castricum, H.L., Kreiter, R., van Veen, H.M., Blank, D.H.A., Vente, J.F. and ten Elshof, J.E. 2008a. High-performance hybrid pervaporation membranes with superior hydrothermal and acid stability. J. Memb. Sci. 324: 111–118.

Development of Membranes for Pervaporation

319

Castricum, H.L., Sah, A., Kreiter, R., Blank, D.H.A., Vent, J.F. and ten Elshof, J.E. 2008b. Hybrid ceramic nanosieves: Stabilizing nanopores with organic links. Chem. Commun. 1103–1105. Castricum, H.L., Sah, A., Kreiter, R., Blank, D.H.A., Vent, J.F. and ten Elshof, J.E. 2008c. Hydrothermally stable molecular separation membranes from organically linked silica. J. Mater. Chem. 18: 2150–2158. Chen, W.J., Aranda, P. and Martin, C.R. 1995. Pervaporation separation of ethanol/water mixtures by polystyrenesulfonate/alumina composite membranes. J. Memb. Sci. 107: 199–207. Chen, S.H., Yu, K.C., Lin, S.S., Chang, D.J. and Liou, R.M. 2001a. Pervaporation separation of water/ethanol mixture by sulfonated polysulfone membrane. J. Memb. Sci. 183: 29–36. Chen, X., Yang, H., Gu, Z.Y. and Shao, Z.Z. 2001b. Preparation and characterization of HY zeolite illed chitosan membranes for pervaporation separation. J. Appl. Polym. Sci. 79: 1144–1149. Chen, H.L., Li, Y.S., Zhu, G.Q., Liu, J. and Yang, W.S. 2005. Synthesis and pervaporation performance of high-reproducibility silicate-1 membranes. Chin. Sci. Bull. 53(22): 3505–3510. Chen, J.H., Liu, Q.L., Zhang, X.H. and Zhang, Q.G. 2007a. Pervaporation and characterization of chitosan membranes cross-linked by 3-aminopropyltriethoxysilane. J. Memb. Sci. 292: 125–132. Chen, S.H., Liou, R.M., Shih, C.Y., Lai, C.L. and Lai, J.Y. 2007b. Pervaporation separation of water/ethanol mixture using asymmetric ion-exchange membranes containing aluminium ions. J. Appl. Polym. Sci. 106(2): 1412–1420. Chen, J., Li, J., Qi, R., Ye, H. and Chen, C. 2008. Pervaporation performance of crosslinked polydimethylsiloxane membranes for deep desulfurization of FCC gasoline 1. Effect of different sulfur species. J. Memb. Sci. 322: 113–121. Choi, H.S., Hong, S.D., Hwang, G.J., Park, C.S., Bae, K.K. and Onuki, K. 2006. Dehydration through pervaporation from HIx solution (HI–H2O–I2 mixture) using a cation exchange membrane for thermochemical water-splitting iodine–sulfur process. Kor. J. Chem. Eng. 23(2): 288–291. Choudhari, S.K., Kittur, A.A., Kulkarni, S.S. and Kariduraganavar, M.Y. 2007. Development of novel blocked diisocyanate crosslinked chitosan membranes for pervaporation separation of water–isopropanol mixtures. J. Memb. Sci. 302: 197–206. Christen, P., Minier, M. and Renon, H. 2004. Ethanol extraction by supported liquid membrane during fermentation. Biotechnol. Bioeng. 36(2): 116–123. Claes, S., Vandezande, P., Mullens, S. et al. 2010. High lux composite PTMSP-silica nanohybrid membranes for the pervaporation of ethanol/water mixtures. J. Memb. Sci. 351: 160–167. Cohen, Y. 2002. Ceramic-supported polymer (CSP) pervaporation membrane. US 6440309. Cole, C.A. and Genetelli, E.J. 1971. Pervaporation of volatile pollutants from water using selective hollow ibers. J. Water Pollut. Control Fed. 42(8) (Research Supplement to: 42, 8, Part II): R290–R298. Czerwiński, W., Ostrowska-Gumkowska, B., Kozakiewicz, J., Kujawski, W. and Warszawski, A. 2004. Siloxane-urethane membranes for removal of volatile organic solvents by pervaporation. Desalination 163: 207–214. Diban, N., Urtiaga, A. and Ortiz, I. 2008. Recovery of key components of bilberry aroma using a commercial pervaporation membrane. Desalination 224: 34–39. Dreux, F., Marais, S., Poncin-Epaillard, F., Metayer, M., Labbe, M. and Saiter, J.M. 2003. Water and toluene barrier properties of a polyamide 12 modiied by a surface treatment using cold plasma. Mater. Res. Innovat. 7: 183–190.

320

Membrane Modification: Technology and Applications

Duke, M.C., Lim, A., da Luz, S.C. and Nielsen, L. 2008. Lactic acid enrichment with inorganic nanoiltration and molecular sieving membranes by pervaporation. Food Bioprod. Process. 86: 290–295. Dutta, B.K. and Sikdar, S.K. 1999. Separation of volatile organic compounds from aqueous solutions by pervaporation using S-B-S block copolymer membrane. Environ. Sci. Technol. 33: 1709–1716. El-Gendi, A., Roizard, D., Grignard, J. and Favre, E. 2010. Investigations of rubbery copolyimides for the preparation of asymmetric pervaporation membranes. Desalin. Water Treat. 14: 67–77. Feng, X. and Huang, R.Y.M. 1997. Liquid separation by membrane pervaporation: A review. Ind. Eng. Chem. Res. 36: 1048–1066. Fleming, H.L. 1992. Consider membrane pervaporation. Chem. Eng. Prog. 88: 46–52. Gallego-Lizon, T., Edwards, E., Lobiundo, G. and dos Santos, L.F. 2002. Dehydration of water/t-butanol mixtures by pervaporation: Comparative study of commercially available polymeric, microporous silica and zeolite membranes. J. Memb. Sci. 197: 309–319. Gao, Z., Yue, Y. and Li, W. 1996. Application of zeolite-illed pervaporation membrane. Zeolites 16: 70–74. González-Rodriguez, J., Pérez-Juan, P., Luque de Castro, M.D. 2002. Method for monitoring urea and ammonia in wine and must by low injection-pervaporation. Anal. Chim. Acta 471: 105–111. Gudernatsch, W., Menzel, Th. and Strathmann, H. 1991. Inluence of composite membrane structure on pervaporation. J. Memb. Sci. 61: 19–30. Hasanoğlu, A., Salt, Y., Keleşer, S. and Dinçer, S. 2009. The esterication of acetic acid with ethanol in a pervaporation membrane reactor. Desalination 245: 662–669. Hennepe, H.J.C., Bergeman, D., Mulder, M.H.V. and Smolders, C.A. 1987. Zeolite illed silicone rubber membranes, Part 1. Membrane preparation and pervaporation results. J. Memb. Sci. 35: 39–55. Hennepe, T., Hendrikus, J.D., Mulder, V., Marcellinus, H., Smolders, C.A. and Schroder, G.A.T. 1990. Pervaporation process and membrane. US 4925562. Hhairnar, D.B. and Pangarkar, V.G. 2004. Dehydration of glycerine/water mixtures by pervaporation using homo and copolymer membranes. J. Am. Oil Chem. Soc. 81(5): 505–510. Hilmioglu, N.D. 2009. Bioethanol recovery using the pervaporation separation technique. Manag. Environ. Qual. 20(2): 165–174. Hirabayashi, Y. 2002. Pervaporation membrane system for the the removal of ammonia from water. Mater. Trans. JIM 42: 1074–1077. Hoshi, M., Saitoh, T., Yoshioka, C., Higuchi, A. and Nakagawa, T. 1999. Pervaporation separation of 1,1,2-trichloroethane-water mixture through crosslinked acrylate copolymer composite membranes. J. Appl. Polym. Sci. 74: 983–984. Hoshi, M., Ieshige, M., Saitoh, T. and Nakagawa, T. 2000. Separation of aqueous phenol through polyurethane membranes by pervaporation. III. Effect of the methylene group length in poly(alkylene glycols). J. Appl. Polym. Sci. 76: 654–664. Huang, R.Y.M. 1991. Pervaporation Membrane Separation Processes. Elsevier: Amsterdam. Huang, RY.M. and Feng, X. 1993. Dehydration of isopropanol by pervaporation using aromatic polyetherimide membrane. Sep. Sci. Technol. 28: 2035–2048. Huang, R.Y.M. and Yeom, C.K. 1990. Pervaporation separation of aqueous mixtures using crosslinked poly(vinyl alcohol) (PVA). II. Permeation of ethanol–water mixtures. J. Memb. Sci. 51: 273–292. Huang, R.Y.M., Pal, R. and Moon, G.Y. 1999. Crosslinked chitosan composite membrane for the pervaporation dehydration of alcohol mixtures and enhancement of structural stability of chitosan/polysulfone composite membranes. J. Memb. Sci. 160: 17–30.

Development of Membranes for Pervaporation

321

Huang, R.Y.M., Moon, G.Y. and Pal, R. 2000. N-Acetylated chitosan membranes for the pervaporation separation of alcohols/toluene mixtures. J. Memb. Sci. 176: 223–231. Huang, R.Y.M., Moon, G.Y. and Pal, R. 2001. Chitosan/anionic surfactant complex membranes for the pervaporation separation of methanol/MTBE and characterization of the polymer/surfactant system. J. Memb. Sci. 184: 1–15. Huang, S.H., Hung, W.S., Liaw, D.J. et al. 2008. Investigation of multilayer pervaporation membrane by positron annihilation spectroscopy. Macromolecules 41: 6438–6443. Hyder, M.N., Huang, R.Y.M. and Chen, P. 2006. Correlation of physicochemical characteristics with pervaporation performance of poly(vinyl alcohol) membranes. J. Memb. Sci. 283: 281–390. Ishida, M., Tasaka, Y. and Aseada, M. 2005. A study on vapor permeation and pervaporation of acetic acid/water mixtures by porous silica membranes. J. Chem. Eng. Jpn. 38(11): 903–907. Ishii, N., Matsumura, M., Tanaka, H. and Araki, K. 1995. Diacetyl fermentation coupled with pervaporation using oleyl alcohol supported liquid membrane. Bioprocess Biosyst. Eng. 13: 119–123. Ji, W., Shirley, A.I. and Sadkowski, P.J. 2005. Process and apparatus for enriching ammonia. US 2005/0034479 A1. Jian, K. and Pintauro, P.N. 1997. Asymmetric PVDF hollow-iber membranes for organic/ water pervaporation separations. J. Memb. Sci. 135: 41–53. Jiang, L.Y., Chung, T.S. and Rajagopalan, R. 2007. Matrimid®/MgO mixed matrix membranes for pervaporation. AIChE J. 53(7): 1745–1757. Jiang, L.Y., Wang, Y., Chung, T.S., Qiao, X.Y. and Lai, J.Y. 2009. Polyimides membranes for pervaporation and biofuels separation. Prog. Polym. Sci. 34: 1135–1160. Jonquières, A., Clément, R., Lochon, P., Néel, J., Dresch, M. and Chrétien, B. 2002. Industrial state-of-the-art of pervaporation and vapour permeation in the western countries. J. Memb. Sci. 206: 87–117. Jou, J.D., Yoshida, W. and Cohen, Y. 1999. A novel ceramic-supported polymer membrane for pervaporation of dilute volatile organic compounds. J. Memb. Sci. 162: 269–284. Kai, T., Goto, H., Shimizu, Y., Yamaguchi, T., Nakao, S.I. and Kimura, S. 2005. Development of crosslinked plasma-graft illing membranes for the reverse osmosis of organic liquid mixtures. J. Memb. Sci. 265: 101–107. Kalipcilar, H., Bowen, T.C., Noble, R.D. and Falconer, J.L. 2002. Synthesis and separation performance of SSZ-13 zeolite membranes on tubular supports. Chem. Mater. 14: 3458–3464. Kang, Y.S., Lee, S.W., Kim, U.Y. and Shim, J.S. 1990. Pervaporation of water–ethanol mixtures through crosslinked and surface-modiied poly(vinyl alcohol) membrane, J. Memb. Sci. 51: 215–226. Kariduraganavar, M.Y., Kittur, A.A., Kulkarni, S.S. and Ramesh, K. 2004. Development of novel pervaporation membranes for the separation of water–isopropanol mixtures using sodium alginate and NaY zeolite. J. Memb. Sci. 238: 165–178. Khayet, M. and Matsuura, T. 2004. Pervaporation and vacuum membrane distillation processes: Modeling and experiments. AIChE J. 50(8): 1697–1712. Kiatkittipong, W., Assabumrungrat, S., Praserthdam, P. and Goto, S. 2002. A Pervaporation membrane reactor for liquid phase synthesis of ethyl tert-butyl ether from tret-butyl alcohol and ethanol. J. Chem. Eng. Jpn. 35(6): 547–556. Kim, Y.W., Hong, Y.K. and Hong, W.H. 2001. Removal of chlorinated organic compounds using crosslinked PDMS membrane. Clean Technol. 7(3): 195–202. Kirsh, Y.E., Fedotov, Y.A., Semenova, S.I. et al. 1995. Sulfonate containing aromatic polyamides as materials of pervaporation membranes for dehydration of organic solvents: Hydration, sorption, diffusion and functioning. J. Memb. Sci. 103(1–2): 95–103.

322

Membrane Modification: Technology and Applications

Kita, H., Horii, K., Ohtoshi, Y., Tanaka, K. and Okamoto, K.I. 1995. Synthesis of a zeolite NaA membrane for pervaporation of water/organic liquid mixtures. J. Mater. Sci. Lett. 14(3): 206–208. Kita, H., Inoue, T., Asamura, H., Tanaka, K. and Okamoto K. 1997. NaY zeolite membrane for the pervaporation separation of methyl–methyl tert-butyl ether mixtures. Chem. Commun. 45–46. Kita, H., Okamoto, K., Yamamura, T. and Abe, J. 2003. Zeolite membranes for fuel ethanol production. Fuel Chem. Div. Preprints 48(1): 438. Kittur, A.A., Kariduranavar, M.Y., Toti, U.S., Ramesh, K. and Aminabhavi, T.M. 2003. Pervaporation separation of water–isopropanol mixtures using ZSM-5 zeolite incorporated poly(vinyl alcohol) membranes. J. Appl. Polym. Sci. 90: 2441–2448. Klamklang, S., Soontarapa, K. and Damronglerd, S. 2002. Preparation and characterization of hydrophilic pervaporation membranes from natural rubber latex based polymer. Sci. Asia 28: 135–143. Knapp, B., Elce, E., Bedwell, B., Langsdorf, L.J. and Wilks, R. 2009. Polynorbornene membrane ilms, preparation and use thereof. USPTO Application # 20090188863. Komada, H. and Honda, Z. 1985. Organic acid permselective membrane. JP 60-40319. Kondo, M., Komori, M., Kita, H. and Okamoto, K.I. 1997. Tubular-type pervaporation module with zeolite NaA membrane. J. Memb. Sci. 133: 133–141. Koops, G.H. and Smolders, C.A. 1991. Estimation and evaluation of polymeric materials for pervaporation membranes. In: Pervaporation Membrane Separation Processes, ed. R.Y.M. Huang, p. 253. Elsevier: New York. Krea, M., Roizard, D., Moulai-Mostefa, N. and Sacco, D. 2004. New copolyimide membranes with high siloxane content designed to remove polar organics from water by pervaporatio. J. Memb. Sci. 241: 55–64. Kreiter, R., Rietkerk, M.D.A., Casticum, H.L., van Veen, H.M., ten Elshof, J.E. and Vente, J.F. 2009. Stable hybrid silica nanosieve membranes for the dehydration of lower alcohols. ChemSusChem 2: 158–160. Kreiter, R., Rietkerk, M.D.A., Castricum, H.L., van Veen, H.M., ten Elshof, J.E. and Vent, J.F. 2011. Evaluation of hybrid silica sols for stable microporous membranes using highthroughput screening. J. Sol Gel Sci. Technol. 57: 245–252. Kühn, J. and Maser, F. 1990. (EN) collagen ilm for use as a pervaporation membrane. WO/1990/006166. Kujawski, W. 2000. Application of pervaporation and vapor permeation in environmental protection. Pol. J. Environ. Stud. 9: 13–26. Kujawski, W. and Krajewski, S.R. 2004. Sweeping gas pervaporation with hollow-iber ionexchange membranes. Desalination 162: 129–135. Kujawski, W. and Pozniak, G. 2005. Transport properties of ion-exchange membranes during pervaporation of water–alcohol mixtures. Sep. Sci. Technol. 40(11): 2277–2295. Kujawski, W., Capala, W., Palczewska-Tulinska, M., Ratajczak, W., Linkiewicz, D. and Michalak, B. 2002. Application of membrane pervaporation process to the enhanced separation of fusel oils. Chem. Pap. 56(1): 21–25. Kujawski, W., Krajewska, S., Kujawski, M., Gazagnes, L., Larbot, A. and Persin, M. 2007a. Pervaporation properties of luoroalkylsilane (FAS) grafted ceramic membranes. Desalination 205: 75–86. Kujawski, W., Staniszewski, M. and Nguyen, T.Q. 2007b. Transport parameters of alcohol vapors through ion-exchange membrane. Sep. Purif. Technol. 57: 476–482. Kusumochayo, S.P. and Sudoh, M. 1999. Dehydration of acetic acid by pervaporation with charged membranes. J. Memb. Sci. 161: 77–83. Kusumochyo, S.P., Kanamori, T., Iwatsubo, T., Sumaru, K. and Shinbo, T. 2002. Surface modiication of pervaporation membranes by formation of ultra-thin polyion complex layer. 1st Conference of Aseanian Membrane Society, Accession No 4A00244206, Tokyo, Japan, pp. 58–59.

Development of Membranes for Pervaporation

323

Kusumochyo, S.P., Sumaru, K., Kanamori, T., Iwatsubo, T. and Shinbo, T. 2004. Synthesis and characterization of an ultrathin polyion complex membrane containing β-cyclodextrin for separation of organic isomers. J. Memb. Sci. 230: 171–174. Kuznetsov, Y.P., Kononova, S.V., Kruchinina, E.V., Romashkova, K.A., Svetlichnyi, V.M. and Molotkov, V.A. 2001. Pervaporation membranes for separating mixtures of methanol and methyl tert-butyl ether. Russ. J. Appl. Chem. 74(8): 1338–1342. Lai, J.Y., Yin, Y.L. and Lee, K.R. 1995. Chemically modiied poly(4-methyl-1-pentene) membrane for pervaporation separation of acetic acid–water mixtures. Polym. J. 27(8): 813–818. Lee, G.T., Krovvidi, K.R. and Greenberg, D.B. 1989. Pervaporation of trace chlorinated organics from water through irradiated polyethylene membrane. J. Memb. Sci. 47: 183–202. Lee, Y.K., Tak, T.M., Lee, D.S. and Kim, S.C. 1990. Cationic/anionic interpenetrating polymer network membranes for the pervaporation of ethanol–water mixture. J. Memb. Sci. 52: 157–172. Li, S.G., Tuan, V.A., Noble, R.D. and Falconer, J.L. 2001a. A Ge-substituted ZSM-5 zeolite membrane for the separation of acetic acid from water. Ind. Eng. Chem. Res. 40: 6165–6171. Li, S., Tuan, V.A., Falconer, J.L. and Noble, R.D. 2001b. Separation of 1,3-propanediol from aqueous solutions using pervaporation through an X-type zeolite membrane. Ind. Eng. Chem. Res. 40: 1952–1959. Li, S., Tuan, V.A., Falconer, J.L. and Noble, R.D. 2001c. Effects of zeolite membrane structure on the separation of 1,3-propanediol from glycerol and glucose by pervaporation. Chem. Mater. 13: 1865–1873. Li, S., Tuan, V.A., Falconer, J.L. and Noble, R.D. 2002. X-type zeolite membranes: Preparation, characterization, and pervaporation performance. Microporous Mesoporous Mater. 53: 59–70. Li, G., Zhang, W., Yang, J. and Wang, X. 2007. Time-dependence of pervaporation performance for the separation of ethanol/water mixtures through poly(vinyl alcohol) membrane. J. Colloid Interface Sci. 306(2): 337–344. Li, X., Basko, M., Prez, F.D. and Vankelecom, F.J. 2008a. Multifunctional membranes for solvent resistant nanoiltration and pervaporation application based on segmented polymer networks. J. Phys. Chem. B 112: 16539–16545. Li, C.L., Huang, S.H., Hung, W.S. et al. 2008b. Study on the inluence of the free volume of hybrid membrane on pervaporation performance by positron annihilation spectroscopy. J. Memb. Sci. 313: 68–74. Liang, L., Dickson, J.M., Zhu, Z., Jiang, J. and Brook, M.A. 2005. Removal of 1,2-dichloroethane from aqueous solutions with novel composite polydimethylsiloxane pervaporation membranes. J. Appl. Polym. Sci. 98: 1477–1491. Liang, L., Dickson, J.M., Jiang, J. and Brook, M.A. 2006. Mass transfer of dilute 1,2-dimethoxyethane aqueous solutions during pervaporation process. J. Appl. Polym. Sci. 100: 2075–2084. Liu, Q., Zhang, Z. and Chen, H. 2001. Study on the coupling of esteriication with pervaporation. J. Memb. Sci. 182: 173–181. Liu, Y.L., Su, Y.H., Lee, K.R. and Lai, J.Y. 2005. Crosslinked organic–inorganic hybrid chitosan membranes for pervaporation dehydration of isopropanol–water mixtures with a long term stability. J. Memb. Sci. 251: 233–238. Liu, Y.L., Yu, C.H., Lee, K.R. and Lai, J.Y. 2007. Chitosan/poly(tetraluoroethylene) composite membranes using in pervaporation dehydration processes. J. Memb. Sci. 287: 230–236. Lokaj, J. and Bilá, J. 2003. Pervaporation of ethanol–water mixtures through styrenesubstituted N-phenylmaleimide copolymer membranes. J. Appl. Polym. Sci. 46(11): 2039–2042.

324

Membrane Modification: Technology and Applications

Lu, S.Y., Huang, H.Y. and Wu, K.H. 2001. Silicalite/poly(dimethylsiloxane) nanocomposite pervaporation membrabnes for acetic acid/water separation. J. Mater. Res. 16(11): 3053–3059. Lu, L., Peng, F., Jiang, Z. and Wang, J. 2006. Poly(vinyl alcohol)/chitosan blend membranes for pervaporation of benzene/cyclohexane mixtures. J. Appl. Polym. Sci. 101(1): 167–173. Lue, S.J., Wang, F.J. and Hsiaw, S.Y. 2004. Pervaporation of benzene/cyclohexane mixtures using ion-exchange membrane containing copper ions. J. Memb. Sci. 240: 149–158. Mahmud, H., Minnery, J., Fang, Y. et al. 2001. Evaluation of membranes containing surface modifying macromolecules: Determination of the chloroform separation from aqueous mixtures via pervaporation. J. Appl. Polym. Sci. 79: 183–189. Mandal, M.K., Dutta, S.D. and Bhattacharya, P.K. 2008. Characterization of blended polymeric membranes for pervaporation of hydrazine hydrate. Chem. Eng. J. 138: 10–19. Manjuan, S. 2005. Concentration of lavor distillates and extracts by pervaporation. Ph.D. Thesis, Chemical Engineering Department, University of Cincinnati, Cincinnati, OH. Masaru, M., Reikichi, I., Seiich, M., Shuzo, Y. and Akira, M. 1985. Chitosan membranes for separation of water–ethanol by pervaporation. Kobunshi Ronbunshu 42(2): 139–142. Maser, F., Ströher-Glowienka, C., Kimmerle, K. and Gudernatsch, W. 1991. Collagen ilm as a new pervaporation membrane. J. Memb. Sci. 61: 269–278. Matsuda, H., Yanagishita, H., Kitamoto, D. et al. 2001. Preparation of silicate pervaporation membrane with ethanol permselectivity by qa 2-step hydrothermal synthesis. Sep. Sci. Technol. 36(15): 3305–3310. Matsufuji, T., Watanabe, K., Nishiyama, N., Egashira, Y., Matsukata, M. and Ueyama, K. 2000. Permeation of hexane isomers through an MFI membrane. Ind. Eng. Chem. Res. 39: 2434–2438. Matsui, S. and Paul, D.R. 2003. Pervaporation separation of aromatic/aliphatic hydrocarbons by a series of ionically crosslinked poly(n-alkyl acrylate) membrane. J. Memb. Sci. 213: 67–83. Matuschewski, H. and Schedler, U. 2008. MSE-modiied membranes in organophilic pervaporation for aromatics/aliphatics separation. Desalination 224: 124–131. Meier-Haack, J., Lenk, W., Lehmann, D. and Lunkwitz, K. 2001. Pervaporation separation of water/alcohol mixtures using composite membranes based on polyelectrolyte multilayer assemblies. J. Memb. Sci. 184: 233–243. Minhas, B.S., Chuba, M.R. and Saxton, R.J. 2003. Membrane process for separating sulfur compounds from FCC light naphtha. US 6,649,061. Mishima, S. and Nakagawa, T. 2000. Sorption and diffusion of volatile organic compounds in luoroalkyl methacrylate-grafted PDMS membrane. J. Appl. Polym. Sci. 75: 773–783. Mitchell, J.K. 1831. On the penetrativeness of luids. J. Roy. Inst. Great Britain 2: 101–307. Mixa, A. and Staudt, C. 2008. Membrane-based separation of phenol/water mixtures using ionically and covalently crosslinked ethylene–methacylic acid copolymer. Int. J. Chem. Eng. ID 319392: 12. Moermans, B., Beuckelaer, W.D., Vankelecom, I.F.J., Ravishankar, R., Martens, J.A. and Jacobs, P.A. 2000. Incorporation of nanosized zeolites in membranes. J. Roy. Soc. Chem. (Chem. Commun.) 2647–2648. Mori, Y. and Inaba, T. 1990. Ethanol production from starch in a pervaporation membrane bioreactor using clostridium thermohydrosulfuricum. Biotechnol. Bioeng. 36(8): 849–853. Morigami, Y., Kondo, M., Abe, J., Kita, H. and Okamoto, K. 2001. The irst large-scale pervaporation plant using tubular-type module with zeolite NaA membrane. Sep. Purif. Technol. 25: 251–260. Nawawi, M.G.B.M. and Hassan, H.B. 2003. Pervaporation separation of isopropanol–water mixtures using crosslinked chitosan membranes. J. Teknol. 39(A) (Keluaran Khas Dis): 55–64.

Development of Membranes for Pervaporation

325

Neidlinger, H.H., Schissel, P.O. and Orth, R.A. 1987. Pervaporation separation of ethanol– water mixtures using polyethylenimine composite membranes. US 06/746592. Nguyen, Q.T., M’Bareck, C.O., David, M.O., Métayer, M. and Alexandre, S. 2003. Ionexchange membranes made of semi-interpenetrating polymer networks, used for pervaporation assisted esteriication and ion transport. Mater. Res. Innovat. 7: 212–219. O’Brien, D.J. and Craig, J.C. Jr.1996. Ethanol production in a continuous fermentation membrane pervaporation system. Appl. Microbiol. Biotechnol. 44: 699–704. Okada, T. and Matsuura, T. 1991. A new transport model for pervaporation. J. Memb. Sci. 59: 133–150. Okada, T. and Matsuura, T. 1992. Predictability of transport equations for pervaporation on the basis of pore-low mechanism. J. Memb. Sci. 70: 163–175. Okamoto, K.I., Kita, H., Kondo, M., Miyake, N. and Matsuo, Y. 1996. Membrane for liquid separation. US 5,554,286. Okamoto, K., Kita, H., Horii, K., Tanaka, K. and Kondo, M. 2001. Zeolite NaA membrane: Preparation single gas permeation, and pervaporation and vapor permeation of water/ organic liquid mixtures. Ind. Eng. Chem. Res. 40: 163–175. Okumus, E., Gurkan, T. and Yilmaz, L. 1994. Development of a mixed-matrix membrane for pervaporation. Sep. Sci. Technol. 29(18): 2451–2473. Okushita, H., Yoshikawa, M. and Shimidzu, T. 1995. Pervaporation of cyclohexane/cyclohexanone/cyclohexanol mixture through polyoxyethylene grafting nylon 6 membrane. J. Memb. Sci. 105: 51–53. Oren, Y., Ferger, V. and Linder, C. 2004. Highly conductive ordered heterogeneous ionexchange membranes. J. Memb. Sci. 239: 17–26. Park, C.H., Nam, S.Y., Lee, Y.M. and Kujawski, W. 2000. Pervaporation of pyridine–water mixture through poly(acrylonitrile-co-monoacryloxyethyl phosphate) membrane. J. Memb. Sci. 164(1–2): 121–128. Peng, M. and Liu, S. 2003. VOC removal from contaminated groundwater through membrane pervaporation, (II): 1,1,1-Trichloroethane-SDS surfactants. J. Environ. Sci. 15(6): 815–820. Peng, M., Vane, L.M. and Liu, S.L.J. 2003. Recent advances in VOCs removal from water by pervaporation. J. Hazard. Mater. 98(1–3): 69–90. Peng, F., Pan, F., Sun, H., Lu, L. and Jiang, Z. 2007. Novel nanocomposite pervaporation membranes composed of poly(vinyl alcohol) and chitosan-wrapped carbon nanotube. J. Memb. Sci. 300: 13–19. Pera-Titus, M., Fité, C., Sabastián, V., Lorente, F., Llorens, J. and Cunill, F. 2008. Modeling pervaporation of ethanol/water mixtures within ‘real’ zeolite NaA membranes. Ind. Eng. Chem. Res. 47(9): 3213–3224. Peters, T.A., Benes, N.E. and Keurentjes, J.T.F. 2005a. Zeolite-coated ceramic pervaporation membranes; pervaporation-esteriication coupling and reactor evaluation. Ind. Eng. Chem. Res. 44: 9490–9496. Peters, T.A., Fontalvo, J., Vorstman, M.A.G. et al. 2005b. Hollow iber microporous silica membranes for gas separation and pervaporation: Synthesis, performance and stability. J. Memb. Sci. 248: 73–80. Peters, T.A., van der Tuin, J., Houssin, C. et al. 2005c. Preparation of zeolite-coated pervaporation membranes for the integration of reaction and separation. Catal. Today 104: 288–295. Peters, T.A., Poeth, C.H.S., Benes, N.E., Buijs, H.C.W.M., Vercauteren, F.F. and Keurentjes, J.T.F. 2006. Ceramic-supported thin PVA pervaporation membranes combining high lux and high selectivity; contradicting the lux-selectivity paradigm. J. Memb. Sci. 276(1–2): 42–50. Polotskaya, G.A., Penkova, A.V., Pientka, Z. and Toikka, A.M. 2010. Polymer membranes modiied by fullerene C60 for pervaporation of organic mixtures. Desalin. Water Treat. 14: 158–164.

326

Membrane Modification: Technology and Applications

Polyakov, A.M., Starannikova, L.E. and Yampolskii, Y.P. 2004. Amorphous Telfons AF as organophilic pervaporation materials: Separation of mixtures of chloromethanes. J. Memb. Sci. 238: 21–32. Pulyalina, A.Y., Polotskaya, G.A., Suschenko, I.G., Meleshko, T.K., Kalyuzhnaya, L.M. and Toikka, A.M. 2010. Pervaporation membrane based on composites of polyimide with poyaniline or its copolymers. Desalin. Water Treat. 14: 158–164. Qi, R., Wang, Y., Li, J. and Zhu, S. 2006a. Sulfur removal from gasoline by pervaporation: The effect of hydrocarbon species. Sep. Purif. Technol. 51: 258–264. Qi, R., Wang, Y., Li, J., Zhao, C. and Zhu, S. 2006b. Pervaporation separation of alkane/thiophene mixtures with PDMS membrane. J. Memb. Sci. 280: 545–552. Qian, J.W., An, Q.F., Wang, L.N., Zhang, L. and Shen, L. 2005. Inluence of the dilute-solution properties of cellulose acetate solvent mixtures on the morphology and pervaporation performance of their membrane. J. Appl. Polym. Sci. 97: 1891–1898. Qin, Y., Sheth, J.P. and Sirkar, K.K. 2002. Supported liquid membrane-based pervaporation for VOC removal from water. Ind. Eng. Chem. Res. 41(14): 3413–3428. Qin, Y., Sheth, J.P. and Sirkar, K.K. 2003. Pervaporation membranes that are highly selective for acetic acid. Ind. Eng. Chem. Res. 42: 582–595. Qiu, J. and Peinemann, K.V. 2006. Novel organic nanocomposite membrane for pervaporation. Desalination 200: 435–436. Quińones-Bolańos, E., Zhou, H. and Parkin, G. 2005. Membrane pervaporation for waste water reuse in microirrigation. J. Environ. Eng. 131(12): 1633–1643. Rachipudi, P.S., Kittur, A.A., Choudhari, S.K., Varghese, J.G. and Kariduraganavar, M.Y. 2009. Development of polyelectrolyte complexes of chitosan and phosphotungstic acid as pervaporation membranes for dehydration of isopropanol. Eur. Polym. J. 45: 3116–3126. Rajagopalan, N. and Cheryan, M. 1995. Pervaporation of grape juice aroma. J. Memb. Sci. 104: 243–250. Ray, S.K., Sawant, S.B., Joshi, J.B. and Pangarkar, V.G. 1997. Dehydration of acetic acid by pervaporation. Ind. Eng. Chem. Res. 36(12): 5265–5276. Reddy, A.S., Kalyani, S., Kumar, N. S., Boddu, V.M. and Krishnaiah, A. 2008. Dehydration of 1,4-dioxane by pervaporation using crosslinked calcium alignate-chitosan blend membranes. Polym. Bull. 61(6): 779–790. Rhim, I.W., Lee, S.W. and Kim, Y.K. 2002. Pervaporation separation of water–ethanol mixtures using metal-ion-exchanged poly(vinyl alcohol (PVA)/sulfosuccnic acid (SSA)) membranes. J. Appl. Polym. Sci. 85: 1867–1873. Riffat, S.B., Wu, S. and Bol, B. 2004. Pervaporation membrane process for vapor absorption system. Int. J. Refrig. 27: 604–611. Sano, T., Yanagishita, H., Kiyozumi, Y., Mizukami, F. and Haraya, K. 1991. Separation of ethanol/water mixture by silicate membrane on pervaporation. J. Memb. Sci. 95: 221–238. Sano, T., Hasegawa, M., Kawakami, Y. et al. 1994. Potentials of silicate membranes for the separation of alcohol/water mixtures. Stud. Surf. Sci. Catal. 84: 1175–1182. Satyanarayana, S.A.V., Subrahmanyam, V.S., Verma, H.C., Sharma, A. and Bhattacharya, P.K. 2006. Application of positron annihilation: Study of pervaporation dense membranes. Polymer 47: 1300–1307. Saxton, R.J., Robert, J. and Minhas, B.S. 2004. Ionic membranes for organic sulfur separation from liquid hydrocarbon solutions. US 6,702,945. Schauer, J., Bartz, D. and Maroušek, V. 1999. Polyurethane pervaporation membranes. Angew. Makromol. Chem. 268(1): 41–45. Schauer, J., Schwarz, H.H. and Eisold, C. 2003. Pervaporation and membrane distillation through membranes made of poly(2,6-dimethyl-1,4-phenylene oxide). Angew. Makromol. Chem. 206(1): 193–198. Schucker, R.C. 2001. Membrane pervaporation and vapor permeation system. US 6273937.

Development of Membranes for Pervaporation

327

Shah, D., Kissick, K., Ghorpade, A., Hannah, R. and Bhattacharya, D. 2000. Pervaporation of alcohol–water and dimethylformamide–water mixtures using hydrophilic zeolite NaA membranes: Mechanism and experimental results. J. Memb. Sci. 179: 185–205. Shah, M.R., Noble, R.D. and Clough, D.E. 2004. Pervaporation-air stripping hybrid process for the removal of VOCs from ground water. J. Memb. Sci. 241: 257–263. Shao, P. and Huang, R.Y.M. 2007. Polymeric membrane pervaporation. J. Memb. Sci. 287: 162–179. Shepherd, A., Habert, A.C. and Borges, C.P. 2002. Hollow iber modules for orange juice aroma recovery using pervaporation. Desalination 148: 111–114. Shevachman, M., Belfer, S., Binman, S. and Shani, A. 2001. Application of jojoba wax bound to polyethylene membranes and hollow ibers in ion-exchange and pervaporation processes. J. Appl. Polym. Sci. 80: 763–768. Sikdar, S.K., Ji, W. and Wang, S.T. 2000. Adsorbent-illed membranes for pervaporation. US 6117328. Sirkar, K.K. 2009. Pervaporation membranes highly selective for volatile solvents present in fermentation broths. US 2009/0114594. Sirkar, K.K., Yang, D., Majumdar, S., Kovenklioglu, S. and Sengupta, A. 1997. Hollow iber contained liquid membrane pervaporation for removal of volatile organic compounds from aqueous solutions. US 5637224. Smitha, B., Suhanya, D., Sridhar, S. and Ramakrishna, M.J. 2004. Separation of organic– organic mixtures by pervaporation – A review. J. Memb. Sci. 241: 1–21. Song, C. 2003. An overview of new approaches to deep desulfurization for ultra-clean gasoline, diesel fuel, and jet fuel. Catal. Today 86: 211–263. Song, K.M. and Hong, W.H. 1997. Dehydration of ethanol and isopropanol using tubular type cellulose acetate membrane with ceramic support in pervaporation process. J. Memb. Sci. 123: 27–33. Song, K.H. and Lee, K.R. 2005. Pervaporation of lavors with hydrophobic membrane. Kor. J. Chem. Eng. 22(5): 735–739. Svang-Ariyaskul, A., Huang, R.Y.M., Douglas, P.L. et al. 2006. Blended chitosan and polyvinyl alcohol membranes for the pervaporation dehydration of isopropanol. J. Memb. Sci. 280: 815–823. Tabe-Mohammadi, A., Garcia Villaluenga, J.P., Kim, H.J., Chan, T. and Rauw, V. 2001. Effects of polymer solvents on the performance of cellulose acetate membranes in methanol/ methyl tertiary butyl ether separation. J. Appl. Polym. Sci. 82: 2882–2895. Takashi, I. 1999. Membrane swelling and pervaporation performance. J. Natl. Inst. Mater. Chem. Res. 7(3): 129–141. Tan, X., Tan, S.P., Teo, W.K. and Li, K. 2006. Polyvinylidene luoride (PVDF) hollow iber membranes for ammonia removal from water. J. Memb. Sci. 271: 59–68. Teng, M.Y., Lee, K.R., Liaw, D.J., Lin, Y.S. and Lai, J.Y. 2000. Plasma deposition of acrylamide onto novel aromatic polyamide membrane for pervaporation. Eur. Polym. J. 36: 663–672. Thongsukmak, A. and Sirkar, K.K. 2007. Pervaporation membranes highly selective for solvents present in fermentation broths. J. Memb. Sci. 302: 45–58. Tripathi, B.P., Kumar, M., Saxena, A. and Shahi, V.K. 2010. Bifunctionalized organic– inorganic charged nanocomposite membrane for pervaporation dehydration of ethanol. J. Colloid Interface Sci. 346: 54–60. Tsai, H.A., Ciou, Y.S., Hu, C.C., Lee, K.R., Yu, D.G. and Lai, J.Y. 2005. Heat-treatment effect on the morphology and pervaporation performances of asymmetric PAN hollow iber membranes. J. Memb. Sci. 255: 33–47. Tu, C.Y., Wang, Y.C., Li, C.L., Lee, K.R., Huang, J. and Lai, J.Y. 2005. Expanded poly(tetraluoroethylene) membrane surface modiication using acetylene/nitrogen plasma treatment. Eur. Polym. J. 41(10): 2343–2353.

328

Membrane Modification: Technology and Applications

Tuan, V.A., Weber, L.L., Falconer, J.L. and Noble, R.D. 2003. Synthesis of B-substituted β-zeolite membranes. Ind. Eng. Chem. Res. 42: 3019–3021. Tyagi, R.K., Fouda, A.E. and Matsuura, T. 1995. A pervaporation model: Membrane design. Chem. Eng. Sci. 50(19): 3105–3114. Ulbricht, M. and Schwarz, H.H. 1997. Novel high performance photo-graft composite membranes for separation of organic liquids by pervaporation. J. Memb. Sci. 136: 25–33. Upadhyay, D.J. and Bhat, N.V. 2004. Pervaporation studies of gaseous plasma treated PVA membrane. J. Memb. Sci. 239: 255–263. Uragami, T., Tsukamoto, K., Inui, K. and Miyata, T. 1998. Pervaporation characteristics of a benzoylchitosan membrane for benzene–cyclohexane mixtures. Macromol. Chem. Phys. 199: 49–54. Uragami, T., Yamada, H. and Miyata, T. 2001. Removal of dilute volatile organic compounds in water through graft copolymer membranes consisting of poly(alkylmethacrylate) and poly(dimethylsiloxane) by pervaporation and their membrane morphology. J. Memb. Sci. 187: 225–269. Uragami, T., Matsugi, H. and Miyata, T. 2005. Pervaporation characteristics of organic– inorganic hybrid membranes composed of poly(vinyl alcohol-co-acrylic acid) and tetraethoxysilane for water/ethanol separation. Macromolecules 38: 8440–8446. Urtiaga, A.M., Gorri, E.D., Ruiz, G. and Ortiz, I. 2001. Parallelism and differences of pervaporation and vacuum membrane distillation in the removal of VOCs from aqueous streams. Sep. Purif. Technol. 22–23: 327–337. Urtiaga, A., Casado, C., Asaeda, M. and Ortiz, I. 2006. Comparison of SiO3–ZrO2–50% and commercial SiO2 membrane on the pervaporative dehydration of organic solvents. Desalination 193: 97–102. Van den Berg, A.W.C., Gora, L., Jansen, J.C., Makkee, M. and Maschmeyer, T. 2003. Zeolite A membranes synthesized on UV-irradiated TiO2 coated metal support: The high pervaporation performance. J. Memb. Sci. 224: 29–37. van Veen, H.M., veen Delft, Y.C., Engelen, C.W.R. and Pex, P.P.A.C. 2001. Dewatering of organics by pervaporation with silica membranes. Sep. Purif. Technol. 22–23: 361–366. van Veen, H.M., Delft, Y.C., Bongers, B., Engelen, C.W.R. and Pex, P.P.A.C. 2004. Methanol separation from organics by pervaporation with modiied silica membranes. Proceedings of the 8th International Conference on Inorganic Membranes, pp. 266–270 and Proceedings, 2004, and 10th Anchen Membrane Colloquiium, Cincinnati, USA, pp. 557–563. Vankelecom, I.F.J., Kinderen, J.D., Dewitte, B.M. and Uytterhoeven, J.B. 1997a. Incorporation of hydrophobic porous illers in PDMS membranes for use in pervaporation. J. Phys. Chem. B 101: 5182–5185. Vankelecom, F.J., Beukelae, S.D. and Uytterhoeven, J.B. 1997b. Sorption and pervaporation of aroma compounds using zeolite-illed PDMS membranes. J. Phys. Chem. 101: 5186–5190. Varghese, J.G., Karuppannan, R.S. and Kariduraganavar, M.Y. 2010. Development of hybrid membranes using chitosan and silica precursors for pervaporation separation of water + isopropanol mixtures. J. Chem. Eng. Data 55: 2084–2092. Vorotyntsev, I.V., Drozdov, P.N., Shablikin, D.N. and Gamajunova, T.V. 2006. Ammonia separation by absorbing pervaporation. Desalination 200: 379–380. Vrentas, J.S. and Duda, J.L 1977. Diffusion in polymer-solvent systems. 1. Reexamination of the free-volume theory. J. Polym. Sci. (Poymer Physics Edition) 15: 403–416. Vrentas, J.S., Duda, J.L. and Huang, W.J. 1986. Regions of Fickian diffusion in polymersolvent system. Macromolecules 19: 1718–1724. Wan, Y., Yi, S. and Su, Y. 2009. Preparation of vinyltriethoxysilane (VTES) modiied silicate/ PDMS hybrid pervaporation membranes: Membrane characterization and its application in ethanol separation from dilute aqueous solutions. AIChE Annual Meeting, Nashville, TN, November 8–13.

Development of Membranes for Pervaporation

329

Wang, X.P., Feng, Y.F. and Shen, Z.Q. 2000. Pervaporation properties of a three layer structure composite membrane. J. Appl. Polym. Sci. 75: 740–745. Wang, Y.C., Li, C.L., Chang, P.F., Fan, S.C., Lee, K.R. and Lai, J.Y. 2002. Separation of water–acetic acid mixture by pervaporation through plasma-treated asymmetric poly(4methyl-1-pentene) membrane and dip-coated with polyacrylic acid. J. Memb. Sci. 208: 3–12. Wang, Y.C., Fan, S.C., Lee, K.R. et al. 2004. Polyamide/SDS-clay hybrid nanocomposite membrane application to water–ethanol mixture pervaporation separation. J. Memb. Sci. 239: 219–226. Wang, Y.C., Li, C.L., Lee, K.R. and Liaw, D.J. 2005. Pervaporation separation of aqueous alcohol solution through a carbazole-functionalized norbornenederivative membrane using living ring-opening metathesis polymerization. J. Memb. Sci. 246: 59–65. Wang, L., Li, J., Zhao, Z. and Chen, C. 2006. Synthesis and characterization of soluble polyimides derived from 4,4′-diamino-3-3′-dimethyldiphenylmethane and their pervaporation performances. J. Macromol. Sci. A 43: 305–314. Wang, Y., Jiang, L., Matsuura, T., Chung, T.S. and Goh, S.H. 2008. Investigation of the fundamental differences between polyamide-imide (PAI) and polyetherimide (PEI) membranes for isopropanol via pervaporation. J. Memb. Sci. 318: 217–226. Wang, Z., Ge, Q., Shao, J. and Yan, Y. 2009. High performance zeolite LTA pervaporation membranes on ceramic hollow ibers by dipcoating-wiping seed deposition. J. Am. Chem. Soc. 131(20): 6910–6911. Wee, S.L., Tye, C.T. and Bhatia, S. 2008. Membrane separation process-pervaporation through zeolite membrane. Sep. Purif. Technol. 63(3): 500–516. White, L.S. and Lesemann, M. 2002. Sulfur reduction of naphtha with membrane technology. Am. Chem. Soc. Div. Petrol. Preprints 47: 45–47. Wolter, H., Ballweg, T. and Storch, S. 2004. Semipermable membranes. US 6818133. Xiao, S., Feng, X. and Huang, R.Y.M. 2007. Trimesoyl chloride crosslinked chitosan membranes for CO2/N2 separation and pervaporation dehydration of isopropanol. J. Memb. Sci. 306: 54–63. Xu, Z.K., Dai, Q.W., Liu, Z.M., Kou, R.Q. and Xu, Y.Y. 2003. Microporous polypropylene hollow iber membranes Part II. Pervaporation serparation of water/ethanol mixtures by the poly(acrylic acid) grafted membranes. J. Memb. Sci. 214: 71–81. Yamaguchi, T., Nakao, S. and Kimura, S. 1991. Plasma-graft illing polymerization: Preparation of a new type of pervaporation membrane for organic liquid mixtures. Macromolecules 24: 5522–5527. Yamaguchi, T., Yamahara, S., Nakao, S. and Kimura, S. 1994. Preparation of pervaporation membranes for removal of dissolved organics from water by plasma-graft illing polymerization. J. Memb. Sci. 95: 39–49. Yamaguchi, T., Tominaga, A., Nakao, S. and Kimura, S. 1996. Chlorinated organics removal from water by plasma-graft illing polymerized membranes. AIChE J. 42(3): 892–895. Yamaguchi, T., Miyazaki, Y., Nakao, S., Tsuru, T. and Kimura, S. 1998. Membrane design for pervaporation or vapor permeation separation using a illing-type membrane concept. Ind. Eng. Chem. Res. 37: 177–184. Yamaguchi, T., Suzuki, T., Kai, T. and Nakao, S. 2001. Hollow-iber type pore-illing membranes made by plasma-graft polymerization for the removal of chlorinated organics from water. J. Memb. Sci. 194: 217–228. Yanagishita, H., Kitamoto, D. and Nakane, T. 1995. Separation of alcohol aqueous solution by pervaporation using asymmetric polyimide membrane. High Perform. Polym. 7(3): 275–281.

330

Membrane Modification: Technology and Applications

Yang, H., Ping, Z., Long, Y. and Nguyen, Q.T. 1999. Improved hydrophobic zeolites to ill silicone membranes for ethyl acetate extraction from water by pervaporation. Can. J. Chem. 77: 1671–1677. Yang, H., Nguyen, Q.T., Ping, Z., Long, Y. and Hirata, Y. 2001. Desorption and pervaporation properties of zeolite-illed poly(dimethylsiloxane) membranes. Mater. Res. Innovat. 5: 101–106. Yeom, C.K. and Lee, K.H. 1996. Pervaporation separation of water–acetic acid mixtures through poly(vinyl alcohol) membranes crsosslinked with glutaraldehyde. J. Memb. Sci. 109: 257–265. Yildirin, A.E., Hilmioglu, N.D. and Tulbentci, S. 2001. Pervaporation separation of benzene/cyclohexane mixtures by (polyvinyl chloride) membranes. Chem. Eng. Technol. 24(3): 275–279. Yoshida, W. and Cohen, Y. 2003. Ceramic-supported polymer membranes for pervaporation of binary organic/organic mixtures. J. Memb. Sci. 213: 145–157. Yoshida, W. and Cohen, Y. 2004. Removal of tert-butyl ether from water by pervaporation using ceramic-supported polymer membrane. J. Memb. Sci. 229: 27–32. Yoshikawa, M. and Kitao, T. 1997. Speciality of polymeric membranes-VI. Pervaporation separation of benzene/cyclohexane mixtures through nylon 6-graft-poly(ethyl methacrylate) membranes. Eur. Polym. J. 33: 25–31. Yoshikawa, M., Yukoshi, T., Sanui, K. and Ogata, N. 1986. Pervaporation of water–ethanol through synthetic polymer membranes having cationic charge sites. Polym. J. (The Society of Polymer Science, Japan) 18(5): 447–450. Yoshikawa, M., Ochiai, S., Tanigaki, M. and Eguchi, W. 1991. Application and development of synthetic membranes. VI. Pervaporation of aqueous ethanol solution through quaternized poly[3-(N′,N′-dimethyl) aminopropylacrylamide-co-acrylonitile] membrane. J. Appl. Polym. Sci. 43(11): 2021–2023. Yoshikawa, M., Takeuchi, S. and Kitao, T. 1997. Speciality of polymeric membranes, 7. Pervaporation separation of benzene/cyclohexane mixtures with nylon 6-graftpoly(oxyethylene) membranes. Angew. Makromol. Chem. 245: 193–202. Yoshikawa, M., Motoi, T. and Tsubouchi, K. 1999a. Speciality polymeric membranes. 11. Pervaporation of benzene/cyclohexane mixtures through poly(vinylalcohol)-graftpoly(acrylic acid) membranes. J. Macromol. Sci. Pure Appl. Chem. A36(4): 621–631. Yoshikawa, M., Tsubouchi, K. and Kitao, T. 1999b. Specialty polymeric membranes. VIII. Separation of benzene/cyclohexane mixtures with nylon 6-graft-poly(butyl methacrylate) membranes. Sep. Sci. Technol. 34: 403–422. Yoshikawa, M., Shimada, H., Tsubouchi, K. and Kondo, Y. 2000. Specialty polymeric membranes 12. Pervaporation of benzene–cyclohexane mixtures through carbon–graphite– nylon 6 composite membranes. J. Memb. Sci. 177: 49–53. Yoshikawa, M., Masaki, K. and Ishikawa, M. 2002. Pervaporation separation of aqueous organic mixtures through agarose membranes. J. Memb. Sci. 205: 293–300. Zhan, X., Li, J.D., Chen, J. and Huang, J.Q. 2009a. Pervaporation of ethanol/water mixtures with high lux by zeolite-illed PDMS/PVDF composite membranes. Chin. J. Polym. Sci. 27(6): 771–780. Zhan, X., Li, J.D., Huang, J.Q. and Chen, C.X. 2009b. Pervaporation properties of PDMS membranes cured with different cross-linking reagents for ethanol concentration from aqueous solutions. Chin. J. Polym. Sci. 27(4): 533–542. Zhan, X., Li, J., Huang, J. and Chen, C. 2010. Enhanced pervaporation performance of multilayer PDMS/PVDF composite membrane for ethanol recovery from aqueous solution. Appl. Biochem. Biotechnol. 160: 632–642. Zhang, L., Chen, H.L., Zhou, Z.J., Lu, Y. and Gao, C.J. 2002. Pervaporation of methanol/ MTBE/C5 ternary mixtures through the CA membrane. Desalination 149: 73–80. Zhang, L., Huang, J.X., Zhou, Z.J., Chen, H.L. and Pan, Z.R. 2006. Effect of pervaporation module structure on separation performance. Desalination 193: 166–170.

Development of Membranes for Pervaporation

331

Zhang, Q.G., Liu, Q.L., Shi, F.F. and Xiong, Y. 2008. Structure and permeation of organic– inorganic hybrid membranes composed of poly(vinyl alcohol) and polysilisesquioxane. J. Mater. Chem. 18: 4646–4653. Zhao, C., Li, J., Qi, R., Chen, J. and Luan, Z. 2008. Pervaporation separation of n-heptane/ sulfur species mixtures with polydimethylsiloxane membranes. Sep. Purif. Technol. 63: 220–225. Zhao, C., Li, J., Chen, J., Qi, R. and Luan, Z. 2009. Separation of sulfur/gasoline mixture with polydimethylsiloxane/polyetherimide composite membranes by pervaporation. Chin. J. Chem. Eng. 17(4): 707–710. Zhu, Y. and Chen, H. 1998. Pervaporation separation and pervaporation-esteriication coupling using crosslinked PVA composite catalytic membranes on porous ceramic plate. J. Memb. Sci. 138: 123–134.

10

Tailor-Made Polymeric Membranes for Advanced Crystallization of Biomolecules Efrem Curcio, Enrica Fontananova, and Gianluca Di Profio

CONTENTS 10.1 10.2 10.3 10.4 10.5 10.6

Membrane Crystallization: Principles and Applications .............................. 333 Membranes ................................................................................................... 336 Effect of Contact Angle ................................................................................ 338 Effect of Membrane Porosity ........................................................................344 Effect of Membrane Roughness ................................................................... 347 Polymorph Selection ..................................................................................... 352 10.6.1 Effect of Physicochemical Parameters of the Membrane ................. 352 10.6.2 Effect of the Transmembrane Flux ................................................... 356 10.7 Conclusion .................................................................................................... 359 References ..............................................................................................................360

10.1

MEMBRANE CRYSTALLIZATION: PRINCIPLES AND APPLICATIONS

Crystallization is an excellent technique for the puriication of chemical species by solidiication from liquid mixtures. Many materials are marketed and sold in crystalline form, and a large amount of a product may be obtained from impure solutions even in a single step. Crystallization is widely applied in the electronic industry for manufacturing semiconductor silicon, GaAs, InP, GaP, CdTe, and its alloys, and scintillation optical crystals (Scheel and Fukuda 2003). In the pharmaceutical industry, crystallization is extensively applied to chiral discrimination and polymorphism (Collins et al. 1997). In biochemistry, crystallization is used for detailed description of protein structures at the atomic level, which is mostly achieved by x-ray diffraction analysis of single biomolecular crystals (McPherson 2003). The possibility of using nanostructured hydrophobic membranes in order to support crystallization operations has been investigated in recent years. In this innovative methodology, polymeric membranes are used in order to (i) generate supersaturation 333

334

Membrane Modification: Technology and Applications

Flux

Flux

Flux

Flux

Stripping solution

Low concentration

Retentate solution

Flux

Flux

Flux

High concentration

by transferring the vaporized solvent from the mother solution to an hypertonic salt solution, under a partial pressure gradient; and (ii) promote heterogeneous nucleation, which permits a decrease in the energy barrier required for the aggregation of critical nuclei and, in ultimate analysis, to drastically reduce the induction time and to accelerate the crystallization process even at low supersaturation (Curcio et al. 2003). Membrane crystallization systems make use of hydrophobic membranes; the hydro-repellent character avoids the passage of solvent in liquid state, while sustaining a liquid–vapor interface at pore mouths. The resulting gradient of partial pressure between the two sides of the membrane, activated by a concentration difference to avoid thermal degradation of biomolecules (Figure 10.1), is the driving force to the evaporation of the solvent that generates a supersaturated solution (Curcio et al. 2001). A well-documented advantage of membrane crystallization techniques over conventional methods is represented by the accelerated rate of the crystallization process, as demonstrated by induction time analysis and nucleation/growth rate measurements (Di Proio et al. 2003). In fact, the membrane surface acts as a promoter of heterogeneous nucleation. As will be detailed in the next paragraphs, this increases the probability of forming stable crystalline nuclei even in supersaturation conditions that would not be adequate for homogeneous nucleation. This aspect is particularly

FIGURE 10.1 Schematic representation of a membrane crystallization system: a high concentrated stripping solution (usually an aqueous solution of polyelectrolyte salt) is used to induce the solvent to migrate—in vapor phase—from the retentate compartment (a solution of biomolecules). The progressive removal of solvent drives the retentate solution toward supersaturation.

Tailor-Made Polymeric Membranes

335

relevant in the case of macromolecules like proteins. Crystallization represents today a topic of strategic relevance in proteomics, since it serves as the basis for x-ray diffraction analysis, the most applied technique to yield atomic-level structural images of proteins. Unfortunately, due to the poor capacity of these biomolecules to aggregate in an ordered lattice, crystallization is often referred as the bottleneck for their three-dimensional structure resolution. In this context, membrane crystallization experiments carried out on various hydrosoluble proteins (i.e., lysozyme, trypsin, catalase, β-glucosydase) showed the possibility of obtaining a crystalline product at lower induction times with respect to those measured when using conventional vapor diffusion techniques (Di Proio et al. 2005). In addition, x-ray diffraction analysis by synchrotron radiation conirmed the high order of the crystal structure grown on microporous hydrophobic membranes (Figure 10.2). Crystallization experiments on hybrid metals–protein systems have also resulted in a new orthorhombic form P21212 of the complex Lys-Co2+ never reported before in literature (Simone et al. 2006). Both theoretical and experimental investigations have conirmed the potential to modulate the crystallization kinetics by controlling the physicochemical characteristics of the polymeric membrane. The transmembrane lux and, consequently, the supersaturation can be precisely controlled by modulating the morphological characteristics of the membrane, such as porosity, pore size, and thickness, or by acting on the process parameters such as type and concentration of the stripping solution, temperature, and low rates. This allows operation in a kinetically or thermodynamically controlled nucleation regime of the nucleation stage, resulting in the production of crystals with speciic structural properties. In particular, the crystallization of stable or metastable polymorphs can be preferentially addressed by controlling the level and rate of supersaturation (Di Proio et al. 2007a,b, 2009), as detailed in Section 10.6.

100µm

FIGURE 10.2 From left to right, hen egg white lysozyme (HEWL) crystal nucleated on the surface of a microporous polypropylene membrane; x-ray diffraction spots of a single HEWL crystal (space group: P 43212); a portion of the electron density map around the tryptophan residue 28 (resolution limit: 1.9Å, mosaicity: 0.165°). (Adapted from Journal of Crystal Growth, 257, Di Proio, G., Curcio, E., Cassetta, A., Lamba, D., and Drioli, E., Membrane crystallization of lysozyme: Kinetic aspects, 359–369, Copyright (2003), with permission from Elsevier.)

336

Membrane Modification: Technology and Applications

10.2

MEMBRANES

When producing polymeric membranes for crystallization purposes, the selection of the material is mainly driven by the necessity to achieve a good hydrophobicity (low surface energy), high chemical stability, controlled porosity, and thickness. The typology and main characteristics of the polymers frequently used as starting material for microporous hydrophobic membranes are given in Table 10.1. The possibility of controlling a crystallization process by a suitable tuning of the physicochemical properties of the polymeric substrate has enhanced interest toward the preparation of speciically modiied membranes. In particular, theoretical and experimental investigations that are the subject of this chapter refer to: (i) Membranes prepared from copolymers to modulate the hydrophobicity and (ii) Microporous hydrophobic membranes modiied by using additives in the casting solution to modulate the morphology in terms of pore size and porosity. Asymmetric and composite membranes commercially known as HYFLON AD are obtained from copolymers of tetraluoroethylene (TFE) and 2,2,4-triluoro5-triluoromethoxy-1,3-dioxole (TTD); these membranes show a high hydrophobic character with contact angles to water greater than 120° (Arcella et al. 1999). Hydrophobic membranes from copolymers of polytetraluoroethylene (PTFE) and polyvinylideneluoride (PVDF) were prepared by a phase inversion process (Feng et al. 2004); these membranes exhibit excellent mechanical properties and good hydrophobicity (contact angle to water of about 87°). Asymmetric hydrophobic membranes exhibiting interesting physicochemical properties can be prepared from poly(vinylideneluoride-co-hexaluoropropylene) copolymer (PKF) and PVDF homopolymer (PK) by a phase inversion process induced by a nonsolvent. Membranes from PVDF copolymer generally exhibit lower gas permeance

TABLE 10.1 Typical Polymeric Materials for Membrane Crystallization Membranes Polymer Polypropylene (PP)

Polyvinylideneluoride (PVDF)

Polytetraluoroethylene (PTFE)

Chemical Structure H

CH3

C

C

H

H

F

H

C

C

F

H

F

F

C

C

F

F

Surface Energy (×10−3 N/m) ∼30

∼30

∼9

337

Tailor-Made Polymeric Membranes

compared to those prepared from PVDF homopolymer. The dissimilar behavior is explained by the different phase inversion mechanisms. For the copolymer, it mainly depends on the balance of liquid–liquid demixing, whereas for the homopolymer, which has higher crystallinity, it is mostly affected by the solid–liquid demixing. The average contact angle to water shows that membranes from Kynarlex 2800 are more hydrophobic (90° ± 3°) than those prepared from Kynar 460 (82° ± 5°). Hydrophilic additives, such as LiCl and polyvinylpyrrolidone (PVP), are used to modify the morphology of PVDF membranes by acting on the thermodynamic and kinetic mechanisms occurring during the phase inversion process. These additives are soluble both in N,N-dimethylacetamide (DMA) and H2O, and are leached out of the cast solution exchanging with the coagulation bath. The scanning electron micrographs (SEMs) of the membrane cross sections are reported in Figure 10.3.

PKF

PKF-PVP 2.5%

PK-LiCl 7.5%

PKF-LiCl 2.5%

PK

PK-PVP 5.0%

FIGURE 10.3 Cross section of the membranes made of poly(vinylideneluoride-co-hexaluoropropylene) or PKF and PVDF homopolymer or PK, with different concentrations of LiCl and PVP in the casting solution. (Adapted from Desalination, 192, Fontananova, E., Jansen, J.C., Cristiano, A., Curcio, E., and Drioli, E., Effect of additives in the casting solution on the formation of PVDF membranes, 190–197, Copyright (2006), with permission from Elsevier.)

338

Membrane Modification: Technology and Applications

TABLE 10.2 Composition (wt%) of the Casting Solutions, Identification Codes, and Morphological Properties of PVDF Membranes

DMA

LiCl

ID CODE

Mean Pore Size (μm)

Porosity (−)

Thickness (μm)

— — 2.5

PKF PKF-L2 PKF-P2

0.034 0.072 0.044

0.0060 0.043 0.11

41.4 69.4 103

— 5.0

PK-L7 PK-P5

0.056 0.20

0.27 0.45

101 150

PVP

Kynarflex 2800 (20 wt%) 80.0 77.5 77.5

— 2.5 —

Kynar 460 (20 wt%) 72.5 75

7.5 —

Kynarlex 2800: poly(vinylideneluoride-co-hexaluoropropylene); Kynar 460: polyvinylideneluoride homopolymer; DMA: N,N-dimethylacetamide; PVP: polyvinylpyrrolidone (MW = 10,000); LiCl: lithium chloride.

Figure 10.3 shows that both membranes prepared from PKF and PK without any additive contain macrovoids. These macrovoids become more accentuated and extend over the whole membrane cross section when using PVP in the casting solution. On the other hand, a relatively high LiCl concentration inhibits the macrovoid formation. This different behavior is explained by the competitive thermodynamic and kinetic effects of the additives on the phase inversion process. Both PVP and LiCl act as nonsolvents, reducing the miscibility of the casting solution (thermodynamic effect) and inducing the enhancement of liquid–liquid phase separation. Moreover, the additives increase the solution viscosity and delay the mutual diffusion between the solvent in the coagulation bath and the nonsolvent in the cast ilm, thus inducing a phase separation delay (kinetic effect). Systems with a rapid phase inversion rate tend to form macrovoids with a inger-like structure, whereas systems with slow phase inversion rates result in a sponge-like structure (Kimmerle and Strathmann 1990). The main structural properties of the membranes shown in Figure 10.3, prepared according to procedures detailed in Fontananova et al. (2006) and identiied by a code for convenience, are reported in Table 10.2.

10.3

EFFECT OF CONTACT ANGLE

A crystallizing solution can be imagined as a certain number of solute molecules moving among the molecules of solvent and colliding with one another, so that some of them converge to form clusters (or growth units). In general, these clusters have a greater probability of being dissolved than of continuing to grow, but, under speciic conditions, they achieve a critical size having the same probability (50%) to grow or dissolve.

339

Tailor-Made Polymeric Membranes

Gibbs free energy of the crystalline phase ΔG is the sum of the contributions from the bulk ΔGvol and from the surface ΔGsurf, and the total energy balance gives ∆G = ∆Gvol + ∆Gsurf = −

∆µ V + γ L AL − ( γ s − γ i ) ASL Ω

(10.1)

where Δμ is the chemical potential gradient between the crystalline phase and the mother solution, Ω the molar volume, and γL, γi, and γs the surface tensions of the nucleus–liquid, nucleus–substrate, and liquid–substrate interfaces, respectively. Referring to Figure 10.4, showing a sphere cap nucleating on a solid substrate (e.g., a polymeric dense membrane), V is the volume of the sphere cap, AL is the contact area of the nucleus–solution surface, and ASL is the contact area of the nucleus–membrane surface. Simple geometrical considerations suggest that V=

πR 3 (1 − cos θ)2 (2 + cos θ) 3

(10.2a)

AL = 2πR 2 (1 − cosθ )

(10.2b)

ASL = πR 2 sin 2 θ

(10.2c)

If Equations 10.2a through 10.2c are substituted into Equation 10.1, then ∆G = −

πR 3 ∆µ (1 − cos θ)2 (2 + cos θ) + 2πR 2 γ L (1 − cos θ) − πR 2 ( γ s − γ i ) sin 2 θ 3 Ω (10.3)

In Equation 10.3, the gradient of chemical potential Δμ (Equation 10.4) is approximated as h = R(1–cos )



R Rsin

FIGURE 10.4 Geometry of a sphere cap nucleating on a solid surface. The cluster of radius R forms a contact angle θ with the porous surface.

340

Membrane Modification: Technology and Applications

∆µ ≈ kT ln

c c*

(10.4)

where k the Boltzmann’s constant, T the absolute temperature, c the concentration of the mother liquor (supersaturated), and c* the equilibrium concentration (saturation). When nucleation occurs on ideal and nonporous surfaces, the Young equation expressing the mechanical equilibrium between interfaces is applied:

( γ s − γ i ) = γ L cos θ

(10.5)

An exempliicative plot of ΔG as a function of the nucleus radius is shown in Figure 10.5 at three different values of the contact angle. In general, a maximum ΔG* corresponds to the critical cluster nucleus R*, and further growth of the cluster leads to a decrease in free energy. Clusters of critical size are called critical nuclei, and all those clusters with a size larger than the size of the critical nucleus will be likely to grow spontaneously. Therefore, crystallization can be considered an activated process, and an energy barrier ΔG* must be crossed in order to induce the formation of stable nuclei (Volmer and Weber 1926). If the following maximization condition d ( ∆G ) =0 dR

(10.6)

is applied to Equation 10.3, then calculations show that, for a critical cluster nucleating in the homogeneous phase (i.e., θ = 180°),

θ = 180° (homogeneous)

4

∆G (kJ/mol)

θ = 110° θ = 90°

2

0

–2 0

0.2

R* 0.4 0.6 Nucleus radius (nm)

0.8

1

FIGURE 10.5 Diagram of the Gibbs free energy versus nucleus radius at three different contact angles (θ = 90°, 110°, 180°). The curves attain their maximum value in correspondence of the critical radius R*. Input data: c/c* = 1.5; Ω = 1.9 × 10−28 m3; γL = 0.004 J/m2.

341

Tailor-Made Polymeric Membranes

* = Rhomogeneous

2γ L Ω ∆µ

(10.7a)

By substituting Equation 10.7a in Equation 10.3: * ∆Ghomogeneous =

16 3  Ω  πγ L  3  ∆µ 

2

(10.7b)

* where ∆Ghomogeneous is the value of the Gibbs energy threshold for homogeneous nucleation. Figure 10.5 demonstrates that the presence of a foreign interface in the crystallizing system—speciically a polymeric membrane with θ < 180°—decreases the work required to generate a critical nucleus and increases locally the probability of nucleation with respect to other locations in the bulk of the system. This phenomenon is known as heterogeneous nucleation. In agreement with the classical nucleation theory (CNT), calculations show that the Gibbs free energy barrier of a critical nucleus developing on a solid surface is * ∆Gheterogeneous 1  1 3 =  − cos θ + cos3 θ *  2 4 4 ∆Ghomogeneous

(10.8)

In order to exemplify the results of Equation 10.8, it can be observed that the work of nucleation occurring on a nonporous membrane having a contact angle θ = 90° is half with respect to that required to form a critical cluster in a homogeneous phase. As expected, the values of Gibbs free energy barrier for homogeneous and heterogeneous nucleation become equal when θ = 180°. CNT elucidates the strict link between ΔG* and nucleation rate J: J = Γe

−

∆G * kT

(10.9)

where Γ is a pre-exponential kinetic factor. Therefore, because of the dependence of the energy barrier on the contact angle, heterogeneous nucleation rate can be controlled by modulating the interaction between the crystallizing solution and the membrane. When investigating nucleation kinetics, an important parameter to take into account is the induction time, deined as the time elapsed from the attainment of a given supersaturation up to the formation of critical nuclei (Garside et al. 2002). As a general rule, a higher nucleation rate results in a shorter induction time. Measurements of induction times conirm the ability of the membrane to promote heterogeneous nucleation. This is particularly evident in protein crystallization. In homogeneous systems, macromolecules display a rather asymmetric and weak bonding coniguration at their surfaces, and tend to aggregate in n-mers that diversify the

342

Membrane Modification: Technology and Applications

shape and size of lattice units, thus making ordered attachment for growing units less probable. Crystallization experiments carried out on hen egg white lysozyme (HEWL) demonstrated that, under comparable or even lower supersaturation ratio, induction times for protein crystals grown on polypropylene membranes are lower than those reported in literature for conventional vapor diffusion techniques (Di Proio et al. 2003). Turbidity measurements obtained by spectrophotometer methods represent a useful tool for investigating the early stage of nucleation. Speciically, for a scattering path length d, the relationship between turbidity τ (Equation 10.10) and absorbance A of a suspension of particles is given by τ=

2.303 A d

(10.10)

Typical turbidity spectra for HEWL solutions are shown in Figure 10.6. During the time interval preceding the formation of stable nuclei (induction period) no signiicant changes in turbidity are detected. As soon as the nucleation process starts, the slope of turbidity proile increases rapidly; the subsequent intermittent peaks are

1

Turbidity (cm–1)

0.8

Polypropylene membrane 2 (transmembrane flux: 0.020 µl/mm h)

0.6

Polypropylene membrane 2 (transmembrane flux: 0.015 µl/mm h)

Induction time

0.4

0.2 Homogeneous system

0

0

20,000

40,000

60,000

Time (sec)

FIGURE 10.6 Turbidity proiles of lysozyme during membrane crystallization by using polypropylene membranes (20 mg/ml HEWL in AcNa/AcH buffer 0.1 M pH 4.6; stripping solution: MgCl2 16–22 wt%). (Adapted from Journal of Crystal Growth, 257, Di Proio, G., Curcio, E., Cassetta, A., Lamba, D., and Drioli, E., Membrane crystallization of lysozyme: Kinetic aspects, 359–369, Copyright (2003), with permission from Elsevier.)

343

Tailor-Made Polymeric Membranes

due to sedimentation of largest crystals. The induction time is estimated by the intersect values, on the time axis, of the tangent to the irst rapid increase in the turbidity proile curve. Membrane surfaces act as a promoter of crystallization by lowering the activation barrier to the nucleation stage, thus allowing molecules to aggregate in conditions of supersaturation that would not be adequate for the spontaneous nucleation. The relatively small elapsed time for the appearance of biomolecular crystals demonstrates the existence of the molecule–membrane interactions that favorably affect the mechanisms of nucleation. A short list of biomolecules tested by the authors is reported in Figure 10.7. Today, the irst step to the development of new drug molecules is represented by crystallographic studies, usually carried out by x-ray diffraction analysis, on single protein crystals. However, proteins are quite reluctant to crystallize and, consistently with the high complexity of their atomic structure, they exhibit lower crystallization kinetics with respect to inorganic substances. An extensive multivariable screening is often necessary to ind the optimal crystallization condition. As a consequence, a money-consuming (because of the high amount of protein requested) and timeconsuming (because of the large number of process parameters to be investigated) procedure is needed to obtain well-diffracting crystals. In this respect, the production of high-quality protein crystals by using low initial amounts of substance, in conjunction with rapid crystallization kinetics, represents an extremely attractive feature of a membrane crystallizer. Induction time (day) 1

2

3

4

5

Glycine (75 Da)

Lysozyme (14 kDa)

Trypsin (24 kDa)

β-Glucosidase (135 kDa)

Catalase (240 kDa)

FIGURE 10.7 Typical ranges of induction times for biomolecules crystallized by hydrophobic membranes.

344

10.4

Membrane Modification: Technology and Applications

EFFECT OF MEMBRANE POROSITY

The inluence of membrane porosity on the nucleation kinetics was recently investigated (Curcio et al. 2006). Referring to Equation 10.1 and considering the geometry of the system under investigation (a sphere cap laying on a porous surface) as depicted in Figure 10.8, we obtain V=

πR 3 (1 − cos θ)2 (2 + cos θ) 3

(10.11a)

AL = 2πR 2 (1 − cos θ ) + ε (1 + cos θ )

(10.11b)

ASL = πR 2 (1 − ε ) sin 2 θ

(10.11c)

where R is the radius of the cluster, θ is the contact angle, and ɛ the porosity of the membrane; according to Equation 10.11a the volume of the small sphere caps penetrating into the pores was considered negligible with respect to the overall volume of the droplet nucleating on the surface. If Equations 10.11a through 10.11c are substituted into Equation 10.1: ∆G = −

πR 3 ∆µ (1 − cos θ)2 (2 + cos θ) + 2πR 2 γ L (1 − cos θ) + ε (1 + cos θ) 3 Ω

− πR 2 ( γ s − γ i ) (1 − ε ) sin 2 θ

(10.12)

In case of nucleation on porous membranes, Equation 10.5 is not applicable because it is based on the Young equilibrium condition strictly valid only for ideal and nonporous surfaces. For a porous membrane, a modiied form of the Young

R(1–cos )




R

Rsin  



FIGURE 10.8 Geometry of a sphere cap nucleating on a porous membrane. The cluster of radius R forms a contact angle θ with the porous surface. (Adapted with permission from Curcio et al., J. Phys. Chem. B, 114, 13650–13655. Copyright 2006 American Chemical Society.)

345

Tailor-Made Polymeric Membranes

equation correlating the surface porosity to the measured and equilibrium contact angles is used: 

4ε (1 + cos θ )  1 − ε ) (1 − cos θ ) 

( γ s − γ i ) = γ L  cos θ + ( 

(10.13)

The practical validity of Equation 10.13 was successfully tested on microporous PTFE membranes (Troger et al. 1997). After substituting Equation 10.13 in the last term of Equation 10.3, and imposing the maximization condition to the variation of the free Gibbs energy as by Equation 10.6, the value of radius R* for a critical cluster nucleated on a surface of a porous substrate is R* =

 2Ωγ  (1 + cos θ)2  = R* (1 + cos θ)2  1 − ε 1 − ε  homogeneous ∆µ   (1 − cos θ)2  (1 − cos θ)2  

(10.14)

For homogeneous nucleation (ɛ = 1 and θ = 180°), as well as for a solid surface (ɛ = 0), the value of R* becomes equal to the value of critical radius expressed by Equation 10.7a in agreement with provisions of the CNT. In order to calculate the energy barrier to heterogeneous nucleation occurring on a porous membrane, Equation 10.14 is substituted in Equation 10.12; after appropriate simpliications: *  ∆Gheterogeneous 1 (1 + cos θ)2  2 = + cos cos − − θ θ ε 2 1 1  ( ) ( ) * 4 ∆Ghomogeneous  (1 − cos θ)2 

3

(10.15)

If ɛ = 0, Equation 10.15 reduces to the mathematical form describing the heterogeneous nucleation on nonporous surfaces as expressed by Equation 10.8. The maximum of the Gibbs free energy function associated to critical cluster formation, which corresponds to the energy barrier that nuclei must overcome to become stable, is plotted in Figure 10.9 as a function of membrane porosity and measured contact angle. Points are located on the graph according to the contact angle measured for lysozyme solution (40 mg/ml, 2% w/v NaCl) and porosity of PVDF membranes described in Section 10.2 (data reported in Table 10.2). In general, it was observed that the energy barrier decreases at higher porosity and, therefore, nucleation rate speeds up when using highly porous substrates. For instance, the predicted ΔGheterogeneous/ΔGhomogeneous ratio for PKF–P2 membrane (ɛ = 0.11) is 0.30, that is, 35% lower than the value calculated for an ideal dense polymeric matrix having the same contact angle (87.4°). For PKF membrane exhibiting a very low porosity (ɛ = 0.0059) and high contact angle (100° ± 2.5°), the ΔGheterogeneous/ΔGhomogeneous ratio differs only by 1.5% with respect to the value of 0.64 predicted by the CNT. Empirical evidence of a favorable effect of porosity on the crystallization rate is present in the literature. It was observed that the stress energy in forming crystals

346

Membrane Modification: Technology and Applications

∆G*heterogeneous/∆G*homogeneous

1 ε=0 ε = 0.1 ε = 0.4 Slow nucleation

0.8 PKF

ε = 0.2

0.6 PKF-L2 0.4 PKF-P2

Fast nucleation

0.2 PK-L7 PKF-P5 0 80

100

120 140 Contact angle (°)

160

180

FIGURE 10.9 Ratio of Gibbs energy barrier for heterogeneous and homogeneous nucleation as a function of the membrane porosity and the contact angle of lysozyme solution (protein: 40 mg/ml; NaCl as precipitant: 2% w/v). Lines are from Equation 10.15; abbreviations as in Table 10.2. (Adapted with permission from Curcio et al., J. Phys. Chem. B, 114, 13650–13655. Copyright 2006 American Chemical Society.)

can be drastically reduced due to pores existing on the substrate surface (Luryi and Suhir 1986). The enhancement of macromolecular nucleation on porous silicon surface (PSS) was theoretically investigated by considering the substrate as a fractal object (Stolyarova et al. 2005). It was observed that, although the lattice constant of typical protein crystals (on the order of 10 nm or more) is similar to the pore size of the PSS used, pores support the formation of nuclei for further crystallization on the surface. In the nucleation process associated with capillary condensation of a vapor in a hydrophobic cylindrical pore, the liquid−vapor transition was described within the framework of a simple lattice model, and the phase properties were characterized both at the mean-ield level and with Monte Carlo simulations. It was found that the reduction of energy barrier predicted by simulation depends on system porosity and pore size (Saugey et al. 2005). The nucleation of thaumatin, trypsin, lobster α-crustacyanin, lysozyme, c-phycocyanin, myosin-binding protein-C, and α-actinin actin binding is enhanced in the presence of a porous medium; nonporous surfaces are less successful at promoting nucleation (Chayen et al. 2006). Monte Carlo simulations conirmed that nucleation out of a illed pore is always faster than on a perfectly smooth surface, and the log of the rate varies almost linearly with pore size (Page and Sear 2006). In principle, the possibility to correlate the membrane morphology to the kinetics of nucleation might offer the opportunity of a more rational design of a crystallization process. Connection between theoretical predictions and experimental results was veriied by measuring the nucleation rate of lysozyme on membranes characterized by different contact angles and, therefore, different interfacial free energies (Curcio et al. 2006).

347

Tailor-Made Polymeric Membranes 0.003

Nucleation rate (#/cm3 sec)

PVDF Kynarflex 2800

0.002

0.001 PVDF Kynar 460

0

0

0.2

0.4 Porosity

0.6

FIGURE 10.10 The increasing trend of HEWL heterogeneous nucleation rate with porosity for modiied PVDF Kynar and PVDF Kynarlex membranes (protein dissolved in 0.05 M sodium acetate buffer pH 4.5 and mixed 1:1 to NaCl solution as precipitant. Final concentrations: protein: 40 mg/ml; NaCl: 2.5%–3% w/v). (Adapted with permission from Curcio et al., J. Phys. Chem. B, 114, 13650–13655. Copyright 2006 American Chemical Society.)

Kinetic data for various modiied PVDF Kynar/Kynarlex membranes characterized by different morphologies, measured for HEWL solutions with different precipitant concentrations, are cumulatively plotted in Figure 10.10. A qualitative increase of the nucleation rate with porosity is observed, in agreement with the tendency of the nucleation barrier ΔG* to decrease at higher porosity.

10.5

EFFECT OF MEMBRANE ROUGHNESS

It was recently conirmed that modulating the roughness of a polymeric membrane enhances the control of the nucleation stage, thus increasing the potential to obtain biomolecular crystals with appropriate size and high structural order. In order to theoretically support the experimental indings that provide a clear connection between the crystallization kinetics and the geometrical proile of the membrane surface, researchers implemented a Metropolis Monte Carlo algorithm to study the Ising energy function of a inite square lattice (Curcio et al. 2010). This approach had been already used to investigate collective effects occurring in the nucleation of a new phase undergoing a irst-order phase transition. A twodimensional Ising model was successfully applied to study heterogeneous nucleation on a square lattice with nearest-neighbor interactions and free boundary conditions (Cirillo and Lebowitz 1998). The same approach was used to calculate the heterogeneous nucleation rate on microscopic impurities (Sear 2006). Investigations about the nucleation on the pores showed that the formation of critical nuclei often

348

Membrane Modification: Technology and Applications

proceeds via two steps: nucleation of pore illing and successive nucleation out of pore (Page and Sear 2006). Here we refer to a two-dimensional Ising model on a inite square lattice Λ = {1 … N}2 with spin σ in the coniguration σ ∈ Ω, with Ω = {−1, +1}. The energy function of the system for spin–spin and spin–ield interactions is E  −J

∑

′ ij

σi σ j − h

∑ σ −J ∑ i

i

s

″ ij

σi σ j

(10.16)

where σij = ±1 is the state of spin, J is the coupling constant between free spins, h accounts for an eventual external magnetic ield, and Js is the strength of the coupling between a free spin and a ixed spin of the surface. On the right hand side of Equation 10.16, the irst term refers to interactions between free spins, and the single dash on the sum indicates that it is over all nearest-neighbor pairs ij of free spins. The second term accounts for the interactions between free spins and the external magnetic ield. The third term is for interactions between the free spins and the ixed spins of the wall, and the double dash on the sum indicates that it is over all nearest-neighbor pairs ij of free and ixed spins. The surface is formed of a region where spins are ixed to σi = +1. In general, it is assumed that Js = 0, meaning that the polymeric surface does not preferentially attract either phase. Simulation makes use of the standard Metropolis Monte Carlo method for spin lipping. Flip is always accepted if it lowers the energy or, otherwise, accepted with probability of exp (−ΔE/kT), where ΔE is the energy difference due to spin lip, and k is the Boltzmann constant (Chandler 1987). A typical time sequence of simulation snapshots for a two-dimensional Ising lattice is presented in Figure 10.11. The rough proile of the surface offers preferential sites to spin lipping from down to up phase, and, at the beginning, isolated up spins appear not susceptible for aggregation in the bulk of the system. After the complete lipping to spin-up phase of sites adjacent to the membrane surface, the system remains almost stable in this coniguration for several cycles. This is consistent with

FIGURE 10.11 A time sequence of snapshots for a 40 × 40 lattice box emulating a surface with roughness coeficient of 1.6 (J/kT = 0.7; h/kT = 0.05). The ixed spins indicating the membrane surface are black, up spins are gray, and down spins are white. (Adapted with permission from Curcio et al., J. Phys. Chem. B, 114, 13650–13655. Copyright 2010 American Chemical Society.)

349

Tailor-Made Polymeric Membranes

the existence of an activation barrier to nucleation. The corresponding value of (cycles*sites) is usually interpreted as the induction time τIsing required by the system to leave the metastable state. In the nucleation regime, only one cluster grows and engulfs the whole lattice, and the induction time is inversely proportional to the −1 ). nucleation rate NIsing ( N Ising  τ Ising The two-dimensional Ising model permits the evaluation of the nucleation rate, beyond an unpredictable proportionality factor, expressed in the unit of numbers of critical nuclei formed per Monte Carlo step and lattice site. Figure 10.12 compares both theoretical and experimental results for crystallization of lysozyme. The values of NIsing span over two orders of magnitude from 9.1 × 10−7 #/cycle*site at low roughness (r = 1.2) to 5.2 × 10−5 #/cycle*site at r = 1.6. As a reference, Sear obtained a nucleation rate on an impurity of 6 spins on the order of 10−5 #/cycle*site, compared to a homogeneous nucleation rate of ∼10−13 #/cycle*site (Sear 2006). The increasing trend of NIsing with the Wenzel roughness factor (deined as the ratio of a corrugated surface to the projected area) is experimentally conirmed by crystallization tests. The measured nucleation rate using a PVDF membrane characterized by a roughness coeficient of 1.25 was 5 × 10−6 #/μl/sec; this is ive times higher than the value measured on HYFLON membranes (r = 1.1). The faster nucleation process, occurring at a rate of 1 × 10−5 #/μl/sec, was measured on polydimethylsiloxane (PDMS) membranes (r = 1.5). Induction times, reduced for higher nucleation rates, range between 1 and 3 h.

0.0001

0.0001 Experimental data

PDMS

1E–005

1E–005 PVDF

HYFLON

Experimental nucleation rate (#/µl/sec)

Two-dimensional ising nucleation rate (#/cycle*site)

Two-dimensional Ising model

1E–006

1E–006 1.2

1.4

1.6

1.2

1.4

1.6

Wenzel roughness factor (–)

FIGURE 10.12 Theoretical and experimental nucleation rate of lysozyme crystals as a function of the Wenzel roughness coeficient (lysozyme dissolved in 0.05 M NaAc buffer at pH 4.5; a precipitant solution prepared by dissolving NaCl in NaAc buffer. Both solutions iltered and mixed 1:1 to give a inal NaCl concentration of 2 wt% a inal protein concentration of 40 mg/ml). (Adapted with permission from Curcio et al., J. Phys. Chem. B, 114, 13650–13655. Copyright 2010 American Chemical Society.)

350

Membrane Modification: Technology and Applications

An alternative approach is provided by CNT, and the starting point is again Equation 10.1. From simple geometric and trigonometric procedures based on Figure 10.13, we obtain the following (Equations 10.17a through 10.17c): V

π 3 π 2 R (1 − cos θ ) ( 2 + cos θ ) + πR 2h sin 2 θ − ms 2h 3 3

(10.17a)

AL  2πR 2 (1 − cosθ )

(10.17b)

ASL  πR 2 sin 2 θ − mπs 2 + ms s 2 + h2

(10.17c)

where m is the number of asperities on the surface, h is the proile height, l is the distance between two consecutive peaks, s is the half-base length of an asperity and all other symbols as previously deined. As stated, the Young equation is only valid for solid, ideal, and smooth surfaces; thus the wetting behavior of a rough surface is described by the Wenzel equation (Equation 10.18):

( γ s − γ i ) = 1 γ l cos θ

(10.18)

r

where r is the Wenzel roughness factor (r > 1). In the speciic case depicted in Figure 10.13, assuming that asperities on the membrane surface have a conical shape, the roughness factor r (Equation 10.19) can be evaluated as r=

( πl 2 − mπs 2 ) + mπs πl

2

h2 + s 2

(10.19)



I

h R s

 o

FIGURE 10.13 Geometry of a sphere cap nucleus on a rough surface. (Adapted with permission from Curcio et al., J. Phys. Chem. B, 114, 13650–13655. Copyright 2010 American Chemical Society.)

351

Tailor-Made Polymeric Membranes

Figure 10.14 illustrates the ratio between the energy barrier for heterogeneous nucleation occurring on rough surfaces and the energy barrier for homogeneous nucleation, as obtained from the maximization condition expressed by Equation * * / ∆Ghomogeneous increases with the contact angle and, for the 10.6. The ∆Gheterogeneous limiting value of θ = 180°, the curves converge to 1 (nucleation in homogeneous phase). For a perfectly smooth surface, characterized by a roughness coeficient r = 1, the ratio is expressed by Equation 10.15; in particular, for θ = 90° * * ∆Gheterogeneous / ∆Ghomogeneous = 0.5. According to CNT predictions, the activation barrier to the formation of critical nuclei is decreased with increasing surface roughness, thus resulting in an exponentially faster nucleation process. The experimental research activity reported in literature conirms these theoretical indings. The nucleation rate of calcium sulfate dihydrate was increased by the roughness of heated stainless steel surfaces where CaSO4·2H2O is deposited (Gunn 1980). Analogous observations were reported about the inluence of surface topography of copper sheet on the heterogeneous nucleation of isotactic polypropylene (iPP) at the iPP–Cu interface (Lin et al. 2001). It was also proved that an increase of the wettability and of the roughness of mica surfaces promotes nonspeciic and local interactions that contribute to the enhanced nucleation rate of proteins (Falini et al. 2002). HEWL crystallization tests carried out on poly(l-glutamic acid), poly(2-hydroxyethyl methacrylate), and (3-aminopropyl) triethoxysilane surfaces demonstrated that the speciic topography signiicantly affects the nucleation rate (Liu et al. 2007). Figure 10.15 compares the results of the Ising model and CNT predictions at a contact angle of 90° and for different values of roughness coeficient. The agreement

∆G*heterogeneous/∆G*homogeneous

1

0.8 r = 1.0 (flat surface)

0.6

1.2 1.3

0.4

1.6 0.2

0

0

30

60 90 120 Contact angle (°)

150

180

FIGURE 10.14 Ratio of the activation energy of heterogeneous nucleation process occurring on a membrane with different surface roughness and homogeneous nucleation, as a function of the contact angle. (Adapted with permission from Curcio et al., J. Phys. Chem. B, 114, 13650–13655. Copyright 2010 American Chemical Society.)

352

Membrane Modification: Technology and Applications

∆G*heterogeneous/∆G*homogeneous

0.5

Two-dimensional Ising model CNT 0.4

0.3 J = 0.7 kT; h = 0.05 kT Contact angle: 90° Two-dimensional Ising lattice size L = 40

0.2 1

1.2 1.4 Roughness coefficient

1.6

* / FIGURE 10.15 Comparison between Ising model and CNT results for ∆Gheterogeneous * ∆Ghomogeneous as a function of roughness coeficient (contact angle of 90°). (Adapted with permission from Curcio et al., J. Phys. Chem. B, 114, 13650–13655. Copyright 2010 American Chemical Society.)

is satisfactory, although the Ising model seems to underestimate the reduction of the energy barrier to heterogeneous nucleation; for moderately rough surfaces (r < 1.25), the difference between the Ising model and CNT results is lower than 10%, whereas deviations exceed 20% for r > 1.55. In some cases, a substantial agreement of the Ising model with CNT is obtained if surface tension γ is increased ∼4/3 times the macroscopic value (Acharyya and Stauffer 1998).

10.6 10.6.1

POLYMORPH SELECTION EFFECT OF PHYSICOCHEMICAL PARAMETERS OF THE MEMBRANE

Polymorphism plays a crucial role in the preparation of active pharmaceutical ingredients (APIs), and the possibility to predict and control the crystallization of a speciic polymorph is today of great interest for many applications in the pharmaceutical industry. Although identical in their chemical composition, different polymorphs often exhibit important differences in solubility, dissolution rate, stability, melting point, density, and many other properties that signiicantly affect the eficacy, bioavailability, and safety of APIs (Llinàs and Goodman 2008). It is today well accepted that a control of the heterogeneous primary nucleation can favor the formation of different polymorphs. In this context, the ability of membranebased crystallizers to promote the selective formation of stable or metastable polymorphs on polymeric hydrophobic membranes was conirmed by experimental tests: on glycine with selective crystallization of forms α and γ (Di Proio et al. 2007a), on paracetamol with the possibility to discriminate between the monocline stable form of

353

Tailor-Made Polymeric Membranes

polymorph I and the orthorhombic metastable form of polymorph II (Di Proio et al. 2007b), on l-glutamic acid with selective crystallization of forms α and β (Di Proio et al. 2009), on l-histidine with selective crystallization of forms A and B (Di Proio et al. 2010). The high number of phenomena involved in the crystallization process of polymorphs means that the interpretation of these results requires a more systematic analysis of the effects of the different physicochemical parameters on the heterogeneous nucleation stage occurring at the solution–membrane interface. According to CNT, the pre-exponential factor Γ in Equation 10.9—related to the frequency and growth probability for a polymorph cluster at a certain supersaturation (Auer and Frenkel 2004)—is a function of the number density of sites for nucleation ρl, the attachment rate of particles to the critical cluster fc+, and the Zeldovich factor z (Equation 10.20a): Γ = ρl fc+ z

(10.20a)

where (Ter Horst and Jansens 2005):

z=

−1 2πkT

 ∂ 2 ∆G   2   ∂N 

*

(10.20b)

In Equation 10.20b, k is the Botzmann constant, T the temperature, N the number of molecules in the cluster, and the * indicates that the second derivative of the Gibbs free energy is calculated for a critical cluster. The Zeldovich factor represents a thermodynamic correction parameter and takes into account the fact that a cluster having reached the critical size does not necessarily nucleate, but could luctuate in size back into the sub-critical region (Schmelzer 2003). The Zeldovich factor for nucleation rate on a polymeric membrane is a function of porosity and contact angle. According to Equation 10.1: ∂ 2 ∆G ∂ 2 ∆Gvol ∂ 2 ∆Gsurf = + ∂N 2 ∂N 2 ∂N 2

(10.21)

Moreover, considering Equation 10.2a, the number of molecules in the cluster (Equation 10.22) can be estimated as N=

V πR 3 = (1 − cos θ )2 (2 + cos θ ) s Ω 3Ω

(10.22)

In order to simplify mathematical steps, the volumetric (ΔGvol) and surface contributions (ΔGsurf ) to ΔG are discussed separately. For the volumetric contribution (Equations 10.23 and 10.24), using the chain rule, ∂Gvol ∂Gvol  ∂N  = ∂N ∂r  ∂r 

−1

= − µ

(10.23)

354

Membrane Modification: Technology and Applications

and, therefore, ∂ 2 ∆Gvol =0 ∂N 2

(10.24)

If an analogous procedure is applied to the surface term of the Gibbs free energy for a nucleus formation, then ∂∆Gsurf ∂∆Gsurf  ∂N  = ∂N ∂r  ∂r  =

−1

4Ωγ L (1 − cos θ ) + ε (1 + cos θ ) − 2Ω ( γ s − γ i ) (1 − ε ) sin 2 θ R (1 − cos θ ) ( 2 + cos θ ) 2

(10.25)

For the case of nucleation on porous membranes, if Equation 10.13 is applied to Equation 10.25, after appropriate arrangements, the surface term of the Gibbs free energy (Equation 10.26) for a nucleus formation is 2 1 + cos θ )  ∂∆Gsurf 2Ωγ L  ( =  1 − ε R  ∂N (1 − cos θ)2  

(10.26)

Going through the second derivative of the surface term: ∂ 2 ∆G ∂  ∂∆Gsurf   ∂N  = ∂N 2 ∂r  ∂N   ∂r  =−

2Ω 2 γ L πR 4

−1

  1 (1 + cos θ)2   − ε 1    2 (1 − cos θ)2    (1 − cos θ ) ( 2 + cos θ ) 

(10.27)

2 According to Equation 10.21, the term ∂ ∆G must be evaluated for a critical ∂N 2 cluster; the value of the critical radius R * for heterogeneous nucleation on a porous membrane is reported in Equation 10.14. By substituting Equation 10.14 in Equation 10.27, and considering Equation 10.21, we get Equation 10.28:

zheterogeneous =

(

Ω

* 2π Rhomogeneous

)

2

γ kT

 (1 + cos θ)2  1 − ε  (1 − cos θ)2 (2 + cos θ)  (1 − cos θ)2 

−3

4

(10.28) Table 10.3 summarizes the form of the Zeldovich factor for different cases; homogeneous nucleation (ɛ = 0, θ = 180°), heterogeneous nucleation on a nonporous support (ɛ = 0), and heterogeneous nucleation on a microporous membrane.

355

Tailor-Made Polymeric Membranes

TABLE 10.3 Expressions of the Zeldovich Factor (z) for Nucleation on Different Membrane Surfaces Case

Expression for Zeldovich Factor

Homogeneous nucleation

zhomogeneous =

(

Ω

* 2π Rhomogeneous

Heterogeneous nucleation on a nonporous membrane

zheterogeneous = zhomogeneous

Heterogeneous nucleation on a porous membrane

zheterogeneous = zhomogeneous

Comment —

γ kT

)

2

1 ζ

ζ=

* ∆Gheterogeneous as by * ∆Ghomogeneous

Equation 10.8 1 ψ

ψ=

* ∆Gheterogeneous as by * ∆Ghomogeneous

Equation 10.15

Figure 10.16 compares the values of the Zeldovich factor for heterogeneous nucleation on a porous hydrophobic membrane and homogeneous nucleation. The ratio zheterogeneous/zhomogeneous increases at higher porosity, and it approaches unity for a contact angle θ coming close to 180° (homogeneous nucleation). The dependence of z with ɛ is remarkable at moderate hydrophobicity. In addition, highly porous membranes

Zheterogeneous/Zhomogeneous

ε: porosity

2.2

ε = 0.7

1.8

ε = 0.5

1.4

ε = 0.3 ε = 0.1 ε=0

1 100

120

140

160

180

Contact angle (°)

FIGURE 10.16 Zeldovich factor for heterogeneous and homogeneous nucleation versus contact angle at different membrane porosity. (Adapted from Journal of Crystal Growth, 310, Curcio, E., Di Proio, G., and Drioli, E., Probabilistic aspects of polymorph selection by heterogeneous nucleation on microporous hydrophobic membrane surfaces, 5364–5369, Copyright (2008), with permission from Elsevier.)

356

Membrane Modification: Technology and Applications

(ɛ > 0.5) inluence signiicantly the pre-exponential term of the nucleation rate equation. For a contact angle of 100° between mother solution and solid substrate, which is a typical value of θ for membrane materials commonly used in membrane distillation– crystallization applications, the value of zheterogeneous/zhomogeneous for a membrane with porosity of 0.7 is 1.9 times higher than zheterogeneous/zhomogeneous calculated for a nonporous surface (ɛ = 0). This conirms the important role played by the morphological and chemical properties of the membrane in the heterogeneous nucleation stage. Nucleation occurs by random aggregation and detachment of molecular growth units. From a probabilistic point of view, the cluster reaches a critical size n* when both attach and detach frequencies are equal. For n > n*, the clusters tend to grow with a probability P(n) (Equation 10.29) that, according to the kinetic theory of nucleation, is expressed as (Curcio et al. 2008) P ( n) =

1 1 + erf (β ( n − n *)) 2

(10.29)

where the numerical factor β (of value between 0 and 1) (Equation 10.30) is related to the Zeldovich factor z according to β = π1/ 2 z

(10.30)

As an example, the growth probability of two paracetamol polymorphs is reported in Figure 10.17. Here it is observed that both homogeneous and heterogeneous nucleation processes favor the stable monoclinic form I over the metastable orthorhombic form II. However, the differences between form I and form II in terms of growth probability in case of crystallization on a microporous hydrophobic membrane are substantially lower with respect to homogeneous nucleation. Closer P(n) curves result in an increased probability to have both paracetamol polymorphs already at moderate supersaturation. Theoretical predictions agree with the experimental evidence of a simultaneous presence of both I and II forms at a supersaturation ratio of 1.52, and of a progressive shifting toward the metastable form II at higher Δμ.

10.6.2

EFFECT OF THE TRANSMEMBRANE FLUX

CNT predicts a linear dependence between the logarithm of the nucleation rate and the second power of the supersaturation, the latter one representing the driving force of a crystallization process. Therefore, the rate of solvent removal in the vapor phase through a microporous membrane is expected to severely affect the crystallization kinetics. In a porous medium, assuming surface diffusion is negligible, the mass transfer is limited by viscous resistance, resulting from the momentum transferred to the membrane, Knudsen diffusion resistance due to molecule–membrane collisions and ordinary diffusion due to collisions between molecules. Predominance, coexistence, or transitions between these different mechanisms are estimated by the dimensionless Knudsen number (Kn) that compares the mean free path ι of diffusing molecules

357

Tailor-Made Polymeric Membranes 1 Form I

Growth probability, P(n)

0.9

Form II

Form I

Form II

0.8 ε = 0.7

0.7

Homogeneous

0.6 ∆µ/kT = 0.419 0.5

0

20 40 Cluster size, n

60

FIGURE 10.17 Growth probability P(n) for both polymorphic forms I and II of paracetamol, expressed as a function of the cluster size and compared for both homogeneous and heterogeneous nucleation (in the latter case a polypropylene membrane with porosity of 0.7 is considered). Inset: paracetamol crystals grown by using a microporous polypropylene membrane (S = 1.52); circle highlights needle-shaped crystals of the polymorphic form II. (Adapted from Journal of Crystal Growth, 310, Curcio, E., Di Proio, G., and Drioli, E., Probabilistic aspects of polymorph selection by heterogeneous nucleation on microporous hydrophobic membrane surfaces, 5364–5369, Copyright (2008), with permission from Elsevier.)

to the mean pore size of the membrane. Kinetic theory of ideal gases provides the following expression (Equation 10.31) for ι: ι=

kT P 2 πσ 2

(10.31)

where k is the Boltzmann constant (1.380 × 10 −23/J K), and σ is the collision diameter of the molecule (σ = 2.7 Å for water). A situation closer to the real functioning of a membrane crystallizer has a binary mixture of water vapor in air; in this case, the free mean path ι a/w (Equation 10.32) can be evaluated at the average membrane temperature T : ιa/w =

kT

1

π (( σ w + σ a ) 2 ) P 1 + ( M w M a ) 2

(10.32)

where σa (=3.7 Å) and σw are the collision diameters, and Ma and Mw are the molecular weights for air and water, respectively. Usual values of mean free path for water

358

Membrane Modification: Technology and Applications

molecules are around 0.1 μm, and are comparable to the pore size of microporous hydrophobic membranes, both commercially available and laboratory-made. In this context, the Dusty Gas Model (DGM) is conventionally used to describe gaseous molar luxes through porous membranes; the most general form (again neglecting surface diffusion) is expressed as (Kast and Hohenthanner 2000) J iD + Dike

n

∑ j =i≠i

p j J iD − pi J Dj 1 =− ∇pi 0 Dije RgT

(10.33a)

J iv = −

εφ2 pi ∇P 8 RgT τµ

(10.33b)

Dike =

2εφ 8 RgT 3τ πMi

(10.33c)

ε PDij0 τ

(10.33d)

Dij0e =

where JD is the diffusive lux, Jv the viscous lux, Dk the Knudsen diffusion coeficient, D 0 the ordinary diffusion coeficient, p the partial pressure, Rg the gas constant (8.314 J/mol/K), T the temperature, P the total pressure, μ the gas viscosity, ϕ the membrane radius, ɛ the membrane porosity, and τ the membrane tortuosity. Underscript e indicates the “effective” diffusion coeficient, calculated by taking into account the structural parameters of the membrane as shown in Equations 10.33c and 10.33d. In a membrane crystallizer, the membrane surface acts as an interface for the evaporation of the volatile solvent. Evaporation rate can be controlled by modulating morphological properties of the membrane such as porosity, pore size, thickness, or the process parameters including concentration or temperature gradient, retentate and permeate low rates. As a consequence, the rate of the supersaturation achievement can be controlled. The possibility to modulate supersaturation, representing the driving force of the crystallization process, is a key issue in a membrane crystallizer. This provides a unique opportunity to operate directly on the nucleation kinetics and to redirect the system toward speciic metastable zones (MZWs). It is well known that the width of a MZW inluences the rate of mass and energy transfer during the creation of a crystal, and induces speciic variations on the inal properties of the crystalline product in terms of size, size distribution, and shape (Ulrich and Strege 2002). Moreover, the rate of the supersaturation achievement inluences the extent of thermodynamic and kinetic contributions to the nucleation process. Experimental investigations prove that, in a membrane crystallizer operated at a low transmembrane lux (low rate of supersaturation achievement), if the least stable structure or a mix of different polymorphs nucleate at the same time, nuclei of the

359

Tailor-Made Polymeric Membranes

Form γ Static membrane crystallizer

Form α Transmembrane flux (µm/h)

250

500

750

1000

1250

1500

FIGURE 10.18 Control of polymorphism in glycine crystallization by acting on the transmembrane lux: high solvent evaporation rate → kinetic product (form α); low solvent evaporation rate → thermodynamic product (form γ).

more stable structure have time to grow at the expense of the less stable forms. On the other hand, in a membrane crystallizer operated at a high transmembrane lux, the increase in the MZW induces nucleation at higher values of supersaturation, and the rapid attainment of a high level of supersaturation forces the system toward the production of the kinetically favored form(s). This offers the unique opportunity to operate a polymorph selection during a membrane-based crystallization process by switching between kinetic and thermodynamic mechanisms at the nucleation stage, thus allowing the production of either a metastable or stable polymorph. An example of this possibility is shown in Figure 10.18, where the selective crystallization of either α or γ forms of the amino acid glycine was achieved by modulating the rate of solvent evaporation through a microporous hydrophobic polypropylene membrane (Di Proio et al. 2007a).

10.7

CONCLUSION

The possibility of controlling both heterogeneous nucleation kinetics and supersaturation generation rate by modulating the physicochemical parameters of the membrane (i.e., porosity, roughness, hydrophobic/hydrophilic character, etc.) represents a unique feature of membrane-based crystallization techniques. Major beneits are evident in macromolecular crystallization, where hydrosoluble proteins nucleate on the membrane surface at a higher rate with respect to conventional crystallization techniques, while maintaining an excellent structural order. Moreover, the proven ability to select a speciic polymorphic form by varying the energy barrier to the formation of critical nuclei is of huge interest in the pharmaceutical industry. It should be also noted that the mandatory requirement that the membrane cannot be wetted by contacting solutions is quite easily satisied for aqueous solutions by using hydrophobic materials such as polypropylene, PTFE, or PVDF. However, the problem exists when trying to crystallize hydrophobic molecules that are soluble in organic solvents. In this case, the development of new membranes that are resistant to/not wetted by organic media will open the way to many exciting future applications, as, for example, single crystal semiconductors crystallization from organic solvents, production of inorganic (photo)catalysts, etc.

360

Membrane Modification: Technology and Applications

REFERENCES Acharyya, M. and Stauffer, D. 1998. Nucleation and hysteresis in Ising model: Classical theory versus computer simulation. Eur. Phys. J. B 5: 571–575. Arcella, V., Colaianna, P., Maccone, P., Sanguinetti, A., Gordano, A., Clarizia, G. and Drioli, E. 1999. A study on a perluoropolymer puriication and its application to membrane formation. J. Memb. Sci. 163: 203–209. Auer, S. and Frenkel, D. 2004. Quantitative prediction of crystal-nucleation rates for spherical colloids: A computational approach. Annu. Rev. Phys. Chem. 55: 333–361. Chandler, D. 1987. Introduction to Modern Statistical Mechanics. Oxford University Press: New York. Chayen, N.E., Saridakis, E. and Sear, R.P. 2006. Experiment and theory for heterogeneous nucleation of protein crystals in a porous medium. Proc. Natl. Acad. Sci. USA 103: 597–601. Cirillo, E.N.M. and Lebowitz, J.L. 1998. Metastability in the two-dimensional Ising model with free boundary conditions. J. Stat. Phys. 90: 211–226. Collins, A.N., Sheldrake, G.N. and Crosby, J. 1997. Chirality in Industry II. John Wiley & Sons: England. Curcio, E., Criscuoli, A. and Drioli, E. 2001. Membrane crystallizers. Ind. Eng. Chem. Res. 40: 2679–2684. Curcio, E., Di Proio, G. and Drioli, E. 2003. A new membrane-based crystallization technique: Tests on lysozyme. J. Cryst. Growth 247: 166–176. Curcio, E., Fontananova, E., Di Proio, G. and Drioli, E. 2006. Inluence of the structural properties of poly(vinylidene luoride) membranes on the heterogeneous nucleation rate of protein crystals. J. Phys. Chem. B 110: 12438–12445. Curcio, E., Di Proio, G. and Drioli, E. 2008. Probabilistic aspects of polymorph selection by heterogeneous nucleation on microporous hydrophobic membrane surfaces. J. Cryst. Growth 310: 5364–5369. Curcio, E., Curcio, V., Di Proio, G., Fontananova, E. and Drioli, E. 2010. Energetics of protein nucleation on rough polymeric surfaces. J. Phys. Chem. B 114: 13650–13655. Di Proio, G., Curcio, E., Cassetta, A., Lamba, D. and Drioli, E. 2003. Membrane crystallization of lysozyme: Kinetic aspects. J. Cryst. Growth 257: 359–369. Di Proio, G., Curcio, E. and Drioli, E. 2005. Trypsin crystallization by membrane-based techniques. J. Struct. Biol. 150: 41–49. Di Proio, G., Tucci, S., Curcio, E. and Drioli, E. 2007a. Selective glycine polymorph crystallization by using microporous membranes. Cryst. Growth Des. 7: 526–530. Di Proio, G., Tucci, S., Curcio, E. and Drioli, E. 2007b. Controlling polymorphism with membrane-based crystallizers: Application to form I and II of paracetamol. Chem. Mater. 19: 2386–2388. Di Proio, G., Curcio, E., Ferraro, S., Stabile, C. and Drioli, E. 2009. Effect of supersaturation control and heterogeneous nucleation on porous membrane surfaces in the crystallization of l-glutamic acid polymorphs. Cryst. Growth Des. 9: 2179–2186. Di Proio, G., Caridi, A., Caliandro, R., Guagliardi, A., Curcio, E. and Drioli, E. 2010. Fine dosage of antisolvent in the crystallization of l-histidine: Effect on polymorphism. Cryst. Growth Des. 10: 449–455. Falini, G., Fermani, S., Conforti, G. and Ripamonti, A. 2002. Protein crystallisation on chemically modiied mica surfaces. Acta Crystallogr. D 58: 1649–1652. Feng, C., Shi, B., Li, G. and Wu, Y. 2004. Preliminary research on microporous membrane from F2.4 for membrane distillation. Sep. Purif. Technol. 39: 221–228. Fontananova, E., Jansen, J.C., Cristiano, A., Curcio, E., and Drioli, E. 2006. Effect of additives in the casting solution on the formation of PVDF membranes. Desalination 192: 190–197.

Tailor-Made Polymeric Membranes

361

Garside, J., Mersmann, A. and Nyvlt, J. 2002. Measurement of Crystal Growth and Nucleation Rates. IChemE: Rugby, UK. Gunn, D.J. 1980. Effect of surface roughness on the nucleation and growth of calcium sulphate on metal surfaces. J. Cryst. Growth 50: 533–537. Kast, W. and Hohenthanner, C.R. 2000. Mass transfer within the gas-phase of porous media. Int. J. Heat Mass Tran. 43: 807–823. Kimmerle, K. and Strathmann, H. 1990. Analysis of the structure-determining process of phase inversion membranes. Desalination 79: 283–302. Lin, C.W., Ding, S.Y. and Hwang, Y.W. 2001. Interfacial crystallization of isotactic polypropylene molded against the copper surface with various surface roughnesses prepared by an electrochemical process. J. Mater. Sci. 36: 4943–4948. Liu, Y.X., Wang, X.J., Lu, J. and Ching, C.B. 2007. Inluence of the roughness, topography, and physicochemical properties of chemically modiied surfaces on the heterogeneous nucleation of protein crystals. J. Phys. Chem. B 111: 13971–13978. Llinàs, A. and Goodman, J.M. 2008. Polymorph control: Past, present and future. Drug Discov. Today 13: 198–210. Luryi, S. and Suhir, E. 1986. New approach to the high quality epitaxial growth of latticemismatched materials. Appl. Phys. Lett. 49: 140–142. McPherson, A. 2003. Introduction to Macromolecular Crystallography. John Wiley & Sons: Hoboken, NJ. Page, A.J. and Sear, R.P. 2006. Heterogeneous nucleation in and out of pores. Phys. Rev. Lett. 97: 065701/1-4. Saugey, A., Bocquet, L. and Barrat, J.L. 2005. Nucleation in hydrophobic cylindrical pores: A lattice model. J. Phys. Chem. B 109: 6520–6526. Scheel, H.J. and Fukuda, T. 2003. Crystal Growth Technology. John Wiley & Sons: England. Schmelzer, J.W.P. 2003. Kinetic and thermodynamic theories of nucleation. Mater. Phys. Mech. 6: 21–33. Sear, R.P. 2006. Heterogeneous and homogeneous nucleation compared: Rapid nucleation on microscopic impurities. J. Phys. Chem. B 110: 4985–4989. Simone, S., Curcio, E., Di Proio, G., Ferraroni, M. and Drioli, E. 2006. Polymeric hydrophobic membranes as a tool to control polymorphism and protein–ligand interactions. J. Memb. Sci. 283: 123–132. Stolyarova, S., Baskin, E., Chayen, N.E. and Nemirovsky, Y. 2005. Possible model of protein nucleation and crystallization on porous silicon. Phys. Status Solidi A 8: 1462–1466. Ter Horst, J.H. and Jansens, P.J. 2005. Nucleus size and Zeldovich factor in two-dimensional nucleation at the Kossel crystal (0 0 1) surface. Surf. Sci. 574: 77–88. Troger, J., Lunkwitz, K. and Burger, W. 1997. Determination of the surface tension of microporous membranes using contact angle measurements. J. Colloid Interface Sci. 194: 281–286. Ulrich, J. and Strege, C. 2002. Some aspects of the importance of metastable zone width and nucleation in industrial crystallizers. J. Cryst. Growth 237–239: 2130–2135. Volmer, M. and Weber, A. 1926. Keimbildung in ubersettigten gebilden. Z. Phys. Chem. 119: 277–301.

11

Chemical Cross-Linking Modifications of Polymeric Membranes for Gas Separation Applications Ahmad Fauzi Ismail and Farhana Aziz

CONTENTS 11.1 Introduction .................................................................................................. 363 11.2 Modiication Techniques to Improve Membrane Performance .................... 365 11.2.1 Thermal Treatment ........................................................................... 365 11.2.2 Polymer Blending ............................................................................. 366 11.2.3 Cross-Linking ................................................................................... 368 11.2.3.1 Photo Cross-Linking .......................................................... 368 11.2.3.2 Chemical Cross-Linking.................................................... 370 11.3 Experimental ................................................................................................ 373 11.3.1 Physical and Chemical Characterization of Cross-Linked Matrimid Membranes ....................................................................... 374 11.3.2 Gas Transport Properties of Cross-Linked Matrimid Membranes......377 11.4 Conclusion .................................................................................................... 380 References .............................................................................................................. 381

11.1

INTRODUCTION

In the last two decades, various engineered glassy polymers have been developed with both experimental and commercial values. These materials combine high selectivities with acceptable permeability coeficients and, therefore, are suitable for the preparation of gas separation membranes (Barsema et al. 2003). The glassy state offers a more structured sieving matrix than a rubbery state, as well as higher load–bearing properties, allowing high pressure drop across the membrane. Thus, the separation of common gas pairs is accomplished by diffusion control or sizeselective sieving mechanisms (Robeson 1999). In the literature, many articles reporting different types of glassy polymers that have been used for the fabrication of membrane-based gas separation equipment can 363

364

Membrane Modification: Technology and Applications

be found. Such polymers include polyimide (PI) (Barsema et al. 2003; Kanehashi et al. 2007; Kim et al. 2006; Krol et al. 2001; Miyata et al. 2008; Recio et al. 2007; Shao et al. 2004a,b; Wind et al. 2003a,b), polycarbonate (Hacarlioglu et al. 2003; Luccio et al. 2000; Sen et al. 2006), polysulfone (PSF) (Ismail et al. 1997; Ismail and Lorna 2003; Kapantaidakis et al. 1996), and polyethersulfone (PES) (Li et al. 2004; Wang et al. 1995, 1996). There is, however, a common problem in gas separation if a polymeric membrane is used; membranes with high gas permeability often have low separation factors, and, on the contrary, membranes with high separation factor have low permeability factors. Robeson (1999) summarized the relationship between the gas permeability (P) and the ideal separation factor, or the gas selectivity (α), for a larger number of experiments. From his observation, he found that the selectivities of many of the glassy polymers are on the upper bound limit or are very close to the upper bound limit. However, the conventional polymeric membranes have limited ability in gas separation, where there is a trade-off line between permeability and selectivity. In order to overcome this situation, concerted effort has been made to overcome these limitations by conducting extensive research on membrane fabrication methods, module design, membrane surface modiication, and development of membrane materials with enhanced properties and separation performance in process environments. Ismail et al. (1997) fabricated asymmetric PSF hollow iber membranes for gas separation. They investigated the possibility of enhancing the selectivity of gas separation hollow iber membranes by increasing the shear rates experienced in the spin line. The results suggest that increased shear during spinning increases the molecular orientation, which in turn enhances selectivity, but lowers the pressure-normalized lux. Hacarlioglu et al. (2003) investigated the effect of preparation parameters on the performance of dense homogeneous polycarbonate gas separation membranes. They demonstrated that the formation of the inal structures of the membranes varied according to their thermal history, casting type, and casting solvent. Thus, their transport properties are also affected by these parameters. Sen et al. modiied the chemical structure of polycarbonate membranes by incorporating an antiplasticizer into the polymer matrix to modify the membrane performance (Sen et al. 2006). The permselective properties of modiied membranes were improved even at lower concentrations of the antiplasticizer. However, the permeability was slightly decreased. Antiplasticization was described as increasing the stiffness of the polymer with the addition of antiplasticizer, reducing the segmental motion in the polymer chain. The effect of the antiplasticizer was related to the reduction of the free volume in the polymer, which restricted the diffusion of gas molecules through the modiied membranes. Therefore, the modiied membranes demonstrated increasing permselective properties while the permeability properties reduced. Surface modiication is a valuable tool for the design of appropriate membrane, as the interfacial characteristics required can rarely be achieved by bulk modiication of the membrane-forming polymer without complications during membrane fabrication (He et al. 2009). Surface modiication methods have been employed by the membrane manufacturers to produce hydrophilic, low-binding membranes (Peinemann and Nunes 2008) with improved membrane performances and properties (Pandey and Chauhan 2001; Robeson 1999). These methods include UV

Chemical Cross-Linking Modifications of Polymeric Membranes

365

treatment, thermal treatment, plasma polymerization, luorination, luorooxidation, chemical cross-linking modiication, and others. Thus, this chapter aims to give an overview on the recent developments in implementing surface modiication techniques, especially cross-linking, to improve the membrane performances. Other modiication techniques such as polymer blending and thermal treatment are also discussed. The results and discussion on the experimental work that has been done in AMTEC on chemical cross-linking modiication of PI membrane are presented.

11.2

MODIFICATION TECHNIQUES TO IMPROVE MEMBRANE PERFORMANCE

Membrane gas separation is now considered to be a proven technology, which has gained an important place in chemical technology and is being widely used in a broad range of applications. One of the most recent developments in the ield of membrane technology has been the successful commercialization of polymer membrane processes for gas separation/puriication. However, there are still challenges to make membrane technology competitive with other conventional separation technologies. Achieving higher selectivity for the relevant application with at least equivalent productivity is one of the main challenges. The other challenges include maintaining membrane performance, thermal stability, and chemical resistance in the presence of aggressive feeds. There is a lot of research reported on the modiication of polymeric membranes in order to improve their performances (Hacarlioglu et al. 2003; Ismail and Yaacob 2006; Kapantaidakis et al. 1996, 2003; Kawakami et al. 1996; Kim et al. 2006; Wind et al. 2003a; Sen et al. 2006). The most widely studied modiication techniques are thermal treatment, polymer blending, and cross-linking, which are discussed next with reference to gas separation applications.

11.2.1

THERMAL TREATMENT

Heat treatment causes the polymer matrix of the membrane to become denser and results in a reduction of the chain mobility and simultaneously minimizes the membrane plasticization. A higher density was observed for a thermally treated membrane when compared with the density of an untreated one (Ismail and Lorna 2003). This indicated a better packing of polymer chains and a decrease in free volume. Kanehashi et al. (2007) reported the effect of thermal treatment on the packing of polymer segments in a membrane. Thermal treatment induced the densiication of glassy polymer membranes. As a result, the critical plasticization pressure shifted from nearly 10–30 atm with the thermal treatment of membranes at 250°C. Compared with the as-cast membranes, the heat-treated membranes were denser and, therefore, the fractional free volume (FFV) values were less than those of the as-cast membranes. These results suggested the formation of charge transfer complexes (CTCs), because the membrane color changed from transparent to a light yellow-brown after thermal treatment. In general, the gas permeability of aromatic PIs decreased with thermal treatment due to polymer-chain

366

Membrane Modification: Technology and Applications

packing effects induced by CTC. However, the d-spacing value was the same for all membranes, indicating that the mean distance between the polymer segments was not inluenced by heat treatment, even though the total amount of free volume space decreased. Krol et al. (2001) demonstrated that in order to suppress plasticization, it was not necessary to have a high cross-linking density of the polymer; a simple annealing effect forming CTCs appeared to be suficient. In this work, Matrimid hollow iber membranes were given different heat treatments, both above and below Tg, to study the possibility of suppressing propylene plasticization. From gas permeation tests, they found that for the untreated ibers, propylene plasticization occurred for pressures greater than 1 bar, while for ibers with mild heat treatment (thermal curing at 100°C or 150°C), the propylene permeance slightly increased up to 3 bar feed pressure, followed by a strong increase at higher pressures. These heat treatments are, therefore, not suficient to suppress plasticization completely at higher feed pressures. However, when a heat treatment at 200°C or higher (but still below Tg) was applied, the propylene permeability decreased slightly or remained constant over the feed pressure range investigated, while the reduction in permeability, compared with untreated ibers, was less than 50%. In addition, only when heat treatment at 350°C was applied did the dissolution time clearly increase. This might be related to the cross-linking phenomena. However, the ibers still readily dissolved in NMP after a 5 min heat treatment at 250°C or 300°C. It can, therefore, be concluded that thermal curing at temperatures below Tg was successful in suppressing propylene plasticization. Kawakami et al. (1996) thermally cured aromatic PI membranes at 150°C, 200°C, and 250°C to study the gas transport properties of the modiied membranes. The gas permeability of CO2, O2, N2, and CH4 for the modiied membranes was measured at 35°C and at pressure up to 10 bar. The gas selectivities for the PI membranes cured at 250°C were enhanced due to an increase in the diffusivity selectivities. Formation of CTCs observed in the PI increased the packing density of the polymer chain and decreased the free volume, thereby leading to an increase in diffusivity selectivities. Thus, it is considered that the suppression of chain mobility by the CTCs provides high size and shape discrimination between the gas molecules. Through extensive research, it has been shown that thermal treatment can be applied successfully to improve the membrane performance and suppress plasticization. However, there are still drawbacks such as the relatively long heat treatment time and the dificulty of uniform application to hollow ibers, which limit the potential for modiication of hollow iber membranes (Cao et al. 2003).

11.2.2

POLYMER BLENDING

Membranes must meet the following performance criteria in order to have a broad and successful applicability in various gas separation environments (Xiao et al. 2007): • Resistant to chemical attack • Thermally stable up to several hundred degrees celsius • Mechanically strong in gas separation environments

Chemical Cross-Linking Modifications of Polymeric Membranes

367

• Easily processed into the asymmetric structures of high-performance membranes • Able to selectively separate gases while providing high throughput of one component However, no single material possesses all of these qualities. Those that are chemically resistant are dificult to process, and vice versa. These limitations have been overcome by the formation of polymer blends for membrane application. Polymer blends of many types have become the dominant material class of polymers in commercial practice over the last 30 years with much research in this area. Rezac et al. (1997) have investigated the transport properties of cross-linkable PI blends. They have reported on the gas transport, chemical resistance, and bulk properties of a series of PI blends. The blends were prepared from a standard PI, which is known to have good transport properties but limited chemical resistance, and a diacetylene-functionalized imide oligomer, which provides a mechanism for solid-state cross-linking. They found that the density of the bulk polymer samples increased with increasing 1,1-6FDA-DIA content in the blend. In addition, gas permeability decreased as the content of 1,1-6FDA-DIA in the blend increased. Nitrogen permeability, for example, decreased by 50% as the 1,1-6FDA-DIA content increased from 0 to 10 wt%. Compared with other polymer systems documented by Robeson (1999), the PI blends evaluated here exhibited a great combination of gas permeabilities and permselectivities (Rezac et al. 1997). Research by Kapantaidakis et al. (1996) investigated the gas permeability (He, H2, CO2, N2, and O2) through PSF/PI-miscible blend membranes. In this work, gas separation membranes were prepared from miscible PSF/PI blends of various compositions (0/100, 20/80, 50/50, 80/20, 100/0). They found that for gases that do not interact with the polymer matrix, such as He, H2, O2, and N2, the permeability coeficients do not vary with the change in the component polymer concentration. In the case of CO2, its permeability through the blend membranes is in all cases lower than the one for the pure polymers. This behavior could be attributed to a strong interaction between CO2 and in the blend membranes, or to possible polymer matrix densiication and consolidation caused by CO2 permeation at high pressures. From this work, they found that the critical plasticization pressure for pure PI is 15 atm, while for pure PSF, it is more than 50 atm. Another important result is that, by using moderately high PSF concentrations (20%–50%), the critical CO2 pressure of plasticization of PSF/PI blend membranes is signiicantly increased, compared with that of pure PI. Moreover, the PI portion of PSF/PI blend membranes offers additional thermal stability and chemical resistance compared with those of pure PSF gas separation membranes. Thus, the blending of PI with PSF yields membranes that can be applied to high-pressure operation or to gas mixtures with high CO2 content. Recently, polymer blending has also been applied in the preparation of a polymeric precursor for carbon membranes. Hosseini and Chung (2009) prepared polymer precursors for carbon membranes by blending polybenzimidazole (PBI) with different types of PIs (Matrimid, Torlon and P84). PBI is a high-performance polymer with the interesting capabilities of withstanding elevated temperatures and high resistance toward aggressive feeds. However, the limited solubility of PBI in common

368

Membrane Modification: Technology and Applications

solvents and its brittleness limit the application of PBI as a polymeric membrane. They observed that all blend precursors exhibited a single composition–dependent glass transition temperature located in the range between the values obtained for the blend constituents. This relected that all the prepared polymer blends were miscible. In addition, the Fourier transform infrared spectroscopy (FTIR) analysis also conirmed the miscibility of blends, as there is an interaction at the molecular level. The gas permeation results showed that the permeability of the blend precursor falls in the range between that of corresponding values obtained for individual polymers. However, for each pair, different effects could be observed, which relect the nature and intensity of the interactions involved. The PBI/Matrimid blend had a signiicant drop in permeability to, on average, half of the values obtained for Matrimid. This was less than the reduction in permeability obtained for other blends. The selectivity was improved for certain gas pairs for each blend, and no deinite trend could be observed. On the other hand, the carbon membranes derived from the polymeric blend precursors exhibited much larger permeability compared with their corresponding precursors. Carbon membrane that was derived from PBI/Matrimid possessed the largest selectivity for the majority of gas pairs including H2/N2, N2/CH4, CO2/CH4, and H2/CO2. The observed microstructural properties of the precursor and carbon membrane demonstrated that PBI/Matrimid constructs have the highest regularity and packing order, thus suggesting that precursor morphology has an effect on the carbon membrane properties. Most research on polymer blending has reported the successful application of the polymer-blending technique in the production of high-performance membrane with improved membrane properties. However, there are some major issues involved with polymer blending that need to be introduced. Blending of the immiscible polymer pairs is one of the main problems. This contributes to three inherent problems: (1) Poor dispersion of one polymer phase in the other; (2) Weak interfacial adhesion between the two phases; and (3) Instability of immiscible polymer blends (Baker et al. 2001).When all of these problems occur, the objective of polymer blending will not be achieved.

11.2.3

CROSS-LINKING

Cross-linking has received much attention since it appears to provide a promising approach toward improving the selectivity and chemical and plasticization resistance. Cross-linking can be achieved by thermal, ion beam, and UV-irradiation treatment or by reactions with added chemicals (Kim et al. 2006; Wind et al. 2003a). The cross-linking modiication tends to increase chain packing and inhibits the intrasegmental and intersegmental mobilities, resulting in higher gas selectivity (Xiao et al. 2007). Cross-links will tie all the polymer molecules together, thereby improving the thermal stability of the polymer. The schematic diagram in Figure 11.1 shows conceptually the polymer chains after cross-linking. 11.2.3.1 Photo Cross-Linking Photo cross-linking is of great interest for the modiication of already prepared polymer membranes to increase their selectivity or permeability or to introduce other

Chemical Cross-Linking Modifications of Polymeric Membranes Polymer

369

Cross-links

Cross-linking

FIGURE 11.1

Schematic diagram of un–cross-linked and cross-linked polymer chains.

desired properties. Cross-links can be formed by chemical reactions that are initiated by heat, ion beam, or UV-irradiation treatment. The energy source employed to promote the cross-linking has been shown to inluence the membrane properties. Based on the literature, two trends can be observed: thermal processing results in nearly uniform distribution of cross-links within the polymer matrix and irradiative processing results in higher reactivity at the irradiated surface (Dudley et al. 2001). For gas separation membrane, such a treatment was proved to improve the membrane performances through optimization of free volume. Polymers with larger FFV have greater permeabilities and diffusivities and are mainly used for reducing the plasticization or swelling tendency in the presence of aggressive feeds. Xiao et al. (2007) produced new membrane materials by the combination of thermal cross-linking and carbonization to fabricate new membrane materials, in order to achieve high gas separation performance and better physical/chemical stability. They observed that thermal cross-linking of PI increased the polymer chain stiffness and decreased the polymer chain spacing, with little change occurring in the solubility selectivity. Although thermal cross-linking modiication cannot improve the gas separation performance, it can signiicantly enhance the plasticization resistance of PIs against CO2 because the resultant polymers have lower chain segmental mobility. For the carbonized membranes, Xiao et al. (2007) found that carbonized membranes produced from cross-linked PI precursors led to relatively higher permselectivities compared with the un–cross-linked precursors, since there was a greater degree of chain packing and increased rigidity. Dudley et al. (2001) studied the inluence of the cross-linking technique on the physical and transport properties of ethynyl-terminated monomer (ETM)/ polyetherimide asymmetric membranes. The polyetherimide was cross-linked by blending with an ETM. Various energy sources have been employed to promote cross-linking in polymeric materials including thermal annealing and UV, γ-, and electron-beam irradiation. They were able to separate the results into two categories, depending upon the activation procedure employed: surface treatments and thermal treatments. Surface treatments resulted in improvement of separation selectivity with only modest reduction in fast gas lux. Irradiation using an electronbeam appears to have led to polymer chain scission, resulting in a reduced thermal stability and glass transition temperature. Thermal treatments were carried out at a temperature well below the maximum reaction temperature of the ethynyl

370

Membrane Modification: Technology and Applications

unit (180°C) and at a temperature near the peak reaction temperature (270°C). The reaction temperature was higher than the glass transition temperature of the fully reacted blend (270°C vs. 215°C) and collapse of the membrane substructure was observed. This was accompanied by a reduction in membrane permeance. Conversely, thermal treatment at 180°C resulted in a marked increase in the permeance of the membranes with a small reduction in selectivity. Based on these results, they summarized that the treatments provide methods for enhancing the resistance to chemical dissolution while maintaining or enhancing thermal stability and enhancing gas selectivity while maintaining lux. However, no single treatment simultaneously achieved all three. Wright and Paul (1997) thermally cross-linked the synthesized polyarylates from 1,2-dihydrocyclobutabenzene-3,6-dicarbonyl dichloride (XTA-Cl), 5-t-butyl isophthalic acid dichloride (tBIA), and luorene bisphenol (FBP). The cross-linking was performed by heating the ilms in a gas-chromatograph oven at 350°C for 3 h under N2 purge. The mechanism of cross-linking initiated with benzocyclobutene undergoes a ring-opening reaction to form o-quinodimethane at high temperatures. Based on their observation, there is a temperature/time trade-off for cross-linking; lower temperatures require longer times for cross-linking. The differential scanning calorimeter (DSC) results indicated that heating increased Tg by 10°C for the copolymer (FBP/XTA-Cl/ tBIA) partly due to cross-linking, while the Tg of homopolymer (FBP/tBIA) increased only by 4°C after heating. The density of the homopolymer did not change on heating, while the density of the copolymer increased slightly on heating. The reason for this phenomenon was the very low intersegmental mobility, indicated by the high Tg values. The gas transport properties of the heated membrane showed some results that were the reverse of those of a typical cross-linked membrane: increase in permeability with decrease in permselectivity. Based on these properties, Wright and Paul (1997) concluded that cross-linking is not the only consequence of heat treatment. It is possible that heat treatment in the absence of chemical changes could lead to physical changes in the chain packing density. The polymer was quenched from above Tg and could have a slightly larger free volume, hence permeability, than the as-cast polymer. It is more likely that these gas transport changes are the results of degradation reactions related with the heating process due to the insigniicant free volume changes. 11.2.3.2 Chemical Cross-Linking Chemical cross-linking is the process where cross-linking is achieved by adding chemical or cross-linker agents to the polymer or membrane. There are different types of cross-linker agents used to cross-link different types of polymers that have been reported in the literature. Table 11.1 shows the list of polymers and cross-linker agents that have been used before in membrane research. Wind et al. (2003a) studied the effects of the cross-linking agent structure and thermal treatment on pure gas permeation and sorption of CO2 and CH4 at 35°C. The PI 6FDA-DAM: DABA 2:1 was cross-linked with ethylene glycol, 1,4-butylene glycol, 1,4-cyclohexanedimethanol, and 1,4-benzenedimethanol to illustrate these effects. This cross-linking approach is quite general, since the cross-link density may be tuned by the amount of DABA incorporated into the polymer backbone and by the selection of the diol cross-linking agent. The cross-linking is carried out in

Chemical Cross-Linking Modifications of Polymeric Membranes

371

TABLE 11.1 List of Polymer and Cross-Linker Agents Reported in the Literature References

Polymer

Cross-Linker Agents

Wind et al. (2003a)

Polyimide

Cao et al. (2003) Liu et al. (2001) Shao et al. (2004a)

Polyimide Polyimide Polyimide

Yi et al. (2006) Shao et al. (2005a) Aziz and Ismail (2010) Peter and Peinemann (2009)

PVAm/PEG Polyimide Polyimide Polyimide PTMSP Polyimide Polyimide

Ethylene glycol, 1,4-butylene glycol, 1,4-cyclohexanedimethanol, 1,4-benzenedimethanol p-Xylenediamine, m-xylenediamine p-Xylenediamine Polypropylenimine tetraamine (DAB-AM-4;G1), polypropylenimine octaamine (DAB-AM-8; G2), propylenimine octaamine (DAB-AM-16; G3) dendrimers Glutaraldehyde Ethylenediamine Para-phenylenediamine 1,2-Xylylenediamine, Bis(2azidophenyl) sulfone Buthylenediamine or Jeffamine 1,3-Propanediamine

Nistor et al. (2008) Jiang et al. (2008)

the solid state, thus minimizing the processing complications. From the research, Wind et al. (2003a) found that cross-linking with butylene glycol and cyclohexanedimethanol is effective in controlling the CO2 permeability up to 40 atm feed pressure by controlling the polymer chain mobility, not necessarily by suppressing sorption. Thus, the physical properties of the cross-linking agent play a major role in determining the inal degree of cross-linking. In addition, they observed that the annealing temperature has an effect on the degree of cross-linking and the permeability. The advantages of this approach are increased membrane productivity, implementation within commercial membrane formation processes, and decrease in the possibility of membrane damage caused by solvent treatments needed in previous approaches (Kim et al. 2006; Wind et al. 2003a). Yi et al. (2006) developed ixed-carrier composite membranes with poly(vinylamine) (PVAm)/poly(ethylene glycol) (PEG) as the separation layer and a PES ultrailtration membrane as the support. In this work, the membranes prepared were cross-linked with glutaraldehyde, and the effect of cross-linking on the performance of the membranes was investigated. They observed that the cross-linkage reduces the mobility of the polymer and produces a denser membrane, which leads to a decrease in diffusion of CH4 and CO2, but the CO2/CH4 selectivity increased to about 20%, compared with that in un–cross-linked membranes. There are several factors that need to be considered when selecting polymer materials for the preparation of membrane-based gas separation. For industrial applications, robust, processable, glassy polymers with high glass transition temperatures, Tg, are typically desired. PIs are one of the glassy polymers that have been identiied as materials with high selectivities and permeabilities for gas separation. In addition

372

Membrane Modification: Technology and Applications

to having high gas separation performance, PIs also possess high glass transition temperatures (Tg > 200°C), which means that they have high thermal stability. A PI that is available in the market is Matrimid 5218 (3,3′,4,4′-benzophenone tetracarboxylic dianhydride and 5(6)-amino-1-(4′-aminophenyl-1,3-trimethylindane), BTDADAPI). Its permeation properties, combined with its processability (i.e., solubility in common solvents), make it a potential candidate for gas separation applications. By far, the largest amount of polymer cross-linking research has been performed on PI membranes. Chung and his group have been active in this area in recent years, conducting cross-linking studies on a variety of different PIs, including: 6FDA-durene, 6FDA-ODA/NDA, Matrimid, and 6FDA-2, 6-DAT (Cao et al. 2003; Liu et al. 2001; Shao et al. 2005b; Xiao et al. 2007). Liu et al. (2001) have developed an extremely simple cross-linking modiication technology at ambient temperature for PIs. The modiication was carried out by immersing the membranes in a cross-linker agent solution for a certain period of time, then the membranes were taken out, followed by washing with fresh methanol and drying at ambient temperature. Based on observation, the chemical mechanism of the cross-linking modiication can be described as follows: the amino groups in p-xylenediamine react with imide groups to form amide groups, thus forming the cross-linking. The swelling of the PI ilms in the p-xylenediamine methanol solution is a prerequisite for the formation of the cross-linking. When p-xylenediamine was added to 2% (w/v) 6FDA-durene dichloromethane, gel particles appeared immediately, so the reaction between the amine groups in p-xylenediamine and imide groups takes place rapidly. The swelling of PI ilms is the rate-determining step. Although methanol is a nonsolvent for the PIs, FTIR studies indicate that all the imide groups can be transferred to amide groups as long as the PI ilms are completely swollen by the p-xylenediamine/methanol solution and the immersion time is long enough. The cross-linking modiication results in signiicant decreases in gas permeabilities of cross-linked PIs to He, O2, N2, and CO2. The permselectivities of He/N2 and O2/N2 increased from 10 to 86 and from 4.1 to 5.9, respectively, but CO2/N2 selectivity decreased with an increase in the degree of cross-linking, thus suggesting that the proposed approach is useful for the application of He/N2 and O2/N2 separations. The results of Liu et al. (2001), in which hollow ibers were predominantly used instead of dense ilms, indicated decreases in permeability with increasing degrees of cross-linking. Selectivity for the investigated gas pairs tends to increase with cross-linking, and the same phenomenon was observed by other researchers (Jiang et al. 2008; Liu et al. 2003; Powell et al. 2007; Tin et al. 2003; Shao et al. 2004a). Liu et al. (2003) also investigated how the amidation alone affects the gas separation properties of PIs by using a diamine reagent that has a tertiary amino group as one of its two amino groups. As a result, the effects of cross-linking on PIs may be removed. The amidation was carried out by immersing the ilms in a 10% (w/v) hexane solution of N,N-dimethylaminoethyleneamine (DMEA) for a certain period of time, followed by taking the ilms out of the solution, washing with fresh hexane, and drying at ambient temperature. ATR-FTIR was used to monitor the surface chemical structure changes of the PI dense ilms during the modiication. The results showed that when the immersion time is long enough (72 h), the characteristic peaks of imide

Chemical Cross-Linking Modifications of Polymeric Membranes

373

groups almost disappear, indicating that all the imide groups had been converted into amide groups. The amidation of both 6FDA-durene/mPDA (50:50) and 6FDAdurene dense ilms lowers the gas permeabilities and improves the permselectivities for the gas pairs of He/N2 and O2/N2. The effect of amidation on permselectivities of CO2/N2 and CO2/CH4 depends on the PI chemistry. These observations are in agreement with those published for research using p-xylenediamine as the cross-linking agent with a long immersion time. Hence, it is reasonable to conclude that the modiication of imide groups to amide groups in the cross-linking process has remarkable effects on the gas separation properties of PIs. Shao et al. (2004a) have modiied PI membranes through chemical cross-linking by using different generations of diaminobutane (DAB) dendrimers, namely polypropylenimine tetraamine (DAB-AM-4; G1), polypropylenimine octaamine (DAB-AM-8; G2), and propylenimine octaamine (DAB-AM-16; G3) dendrimers. The pure gas tests show that the maximum selectivity increased by about 400%, 300%, and 265% for the gas pairs of He/N2, H2/N2, and H2/CO2, respectively, after 60 min of cross-linking with G1 dendrimers. For the gas pair of CO2/CH4, the maximum increment was 74% after 20 min of G1 cross-linking. When compared with different generations of DAB dendrimers, the gas permeability decreased in the order of G1 > G2 > G3, which is consistent with the increasing order of the degree of cross-linking. Shao et al. (2005a) also used ethylenediamine (EDA) to cross-link PIs based on its unique linear structure, small molecular volume, and the potential of its functional groups to react with PIs. Thermal annealing processes at 100°C and 200°C were also performed on the membranes in order to study their effects on the separation performance of EDA–cross-linked PIs. The permeation results demonstrated that the EDA cross-linking can effectively decrease the permeability and increase the selectivity of PIs for He/N2 and H2/N2 separations. However, there were no such increases for O2/N2 selectivity, and CO2/CH4 selectivity was found to decrease with cross-linking. The gas transport properties of the EDA–cross-linked and then thermally treated PIs also show high selectivity for He/N2 and H2/N2 separations. In cross-linked PI membranes examined by a few researchers (Cao et al. 2003; Liu et al. 2001; Shao et al. 2005a,b; Aziz and Ismail 2010; Peter and Peinemann 2009; Nistor et al. 2008; Jiang et al. 2008), who used cross-linking modiication technology at ambient temperature, gas permeabilities decreased while gas selectivities increased with cross-linking and were strongly dependent on the types of cross-linking agents, cross-linking agent concentrations and cross-linking time. Different materials may need different cross-linking agents. Therefore, there is an urgent need to identify suitable cross-linker agents for PI membranes and to fundamentally understand the science of the effects of molecular structure, chain length, and lexibility of cross-linking agents on the cross-linking process and membrane performance.

11.3

EXPERIMENTAL

This section will provide results on the effects of para-phenylenediamine (pPD) as a cross-linking agent on PI membranes. The effects of varying cross-linking

374

Membrane Modification: Technology and Applications

TABLE 11.2 Physical and Chemical Properties of p-Phenylenediamine Physical and Chemical Properties Molecular structure

NH2

H2 N Molecular weight (g/mol) Melting point (°C) Boiling point (°C) Flash point (°C) Stability

108.14 140 271 154 Stable under ordinary conditions

agent concentrations will also be considered. The effects of immersion time on the degree of cross-linking and membrane performance have been discussed earlier by Aziz and Ismail (2010). pPD was chosen based on its properties, which are high temperature stability, high strength, and chemical and electrical resistance (Moeller et al. 2008). Since using pPD as a cross-linker agent for PI membrane is a new approach, this research should in principle give new valuable knowledge to the scientiic community. Through the literature review, diamine–cross-linked Matrimid membranes showed a remarkable improvement in O2/N2 separation; hence, the pPD cross-linked Matrimid membranes developed in this study should have the potential for O2/N2 separation (Shao et al. 2005a; Aziz and Ismail 2010; Liu et al. 2003). The asymmetric lat sheet Matrimid membranes were prepared through the phase inversion technique using a pneumatically controlled lat sheet membrane casting machine. The cross-linking agent, pPD, was purchased from Sigma-Aldrich and used as received. Table 11.2 shows the physical and chemical properties of pPD. Cross-linking was performed by immersing the membrane in 5%, 10%, and 15% pPD/methanol solution for 120 min. The pPD-modiied ilms were then washed with fresh methanol immediately after removal from the pPD solution in order to wash away all residual pPD, followed by drying naturally at room temperature for at least 1 day to ensure the complete removal of the methanol solvent.

11.3.1

PHYSICAL AND CHEMICAL CHARACTERIZATION OF CROSS-LINKED MATRIMID MEMBRANES

The physical properties of the membrane after cross-linking are presented in Table 11.3. The weight of the membrane increased after cross-linking and was consistent with the cross-linker concentration (Table 11.3). In addition, the changes in membrane thickness after cross-linking were insigniicant. This revealed that the

375

Chemical Cross-Linking Modifications of Polymeric Membranes

TABLE 11.3 Physical Properties of Cross-Linked Polyimide Membrane Sample

Weight (%)

0% pPD 5% pPD 10% pPD 15% pPD

Color

0.00 4.34 29.80 43.04

Yello Brown Dark brown Dark brown

cross-linker agents (pPD) had diffused into the membrane and performed the crosslinking reaction. In addition, the color of the membrane also changed with the crosslinker concentration. After cross-linking, the color of the membrane changed from yellow to brown (for 5% pPD) and became darker when immersed in higher crosslinker solution (10% and 15% pPD). The reason for this phenomenon was the concentration of pPD solution. Based on this physical evidence, it was proven that the cross-linker agents had successfully diffused into the membrane after a given period of time (120 min). The change in the chemical structure of PI membranes after cross-linking modiication was monitored using ATR-FTIR. The cross-linking technique used in this study changed the imide groups to amide groups. Thus, by using ATR-FTIR analysis, the cross-linking network could be detected from the appearance of amide groups. Figure 11.2 presents the ATR-FTIR spectra of pPD-modiied Matrimid membranes at different pPD concentrations. It can be seen that 0% pPD samples

Imide group

15% pPD

Amide group

10% pPD

5% pPD

Amine group

0% pPD

3600

3400

3200

3000

2800

2600

2400

2200

2000

1800

1600

1400

1200

Wave numbers (cm–1)

FIGURE 11.2 Comparison of FTIR spectra of the cross-linked BTDA-DAPI polyimide membranes at different cross-linker concentrations.

376

Membrane Modification: Technology and Applications

(unmodiied membrane) had spectra with imide peaks at around 1378.9 (attributed to C–N stretching vibration of imide groups), 1718.3 (attributed to C=O symmetric stretching vibration of imide groups), and 1776.1/cm (attributed to C=O asymmetric stretching vibration of imide groups). After cross-linking, the following peaks in the spectra were observed: the C=O stretch band at 1664.3/cm of the amide groups, the C–N stretching vibration of the C–N–H group at 1540.9/cm, the NH stretching vibration of the amide groups at 3342.1/cm, and the NH2 bend of the amine groups at 1601.1/cm. By referring to Figure 11.2, a common trend can be observed in the decreasing intensities of the imide peaks and the increasing intensities of the amide peaks with the cross-linker concentrations. This conirms that the cross-linking of PI with pPD is a concentration-dependent process. This has been reported earlier by Bulgarelli et al. (1999), who have also concluded that major parameters affecting cross-linking are the cross-linking time and the cross-linker percentage, both if the cross-linker is introduced in the preparative emulsion or if cross-linking is achieved by postsynthesis treatment. The mechanism of chemical cross-linking by using pPD is illustrated elsewhere (Aziz and Ismail 2010). Since the ATR-FTIR for the chemical composition analysis is limited to the surface of the membranes, a gel test was carried out on the cross-linked membranes. In this study, the gel contents of the cross-linked membranes were measured by soaking the membranes in THF for 2 days at 25°C. The gel contents increased with the cross-linker concentrations (Table 11.4). This indicates that the degree of cross-linking increased with the cross-linker concentrations, which supports the FTIR results. The changes in the microstructural properties of cross-linked Matrimid membranes were observed using X-ray diffractometer (XRD). Based on the XRD spectra, the peaks for all samples are broad, indicating the presence of amorphous structures in the samples. The most prominent XRD peak in the amorphous glassy polymer spectra is often used to estimate the d-space (Powell et al. 2007; Ghosal and Freeman 1994). The Bragg equation was used to calculate the value of d-space. The increasing order of d-space for Matrimid and cross-linked Matrimid membranes follows the sequence: pure Matrimid (d-space: 5.46Å) < 10% pPD (5.59Å) < 5% pPD (5.75Å) < 15% pPD (5.90Å). The larger d-space for cross-linked Matrimid compared with pure Matrimid membranes could be due to a few possible reasons. The methanolswelling effect that formed the interstitial spaces between the nodules is one possible reason (Shao et al. 2005b). The second reason is that the cross-linking tightens the TABLE 11.4 The Gel Contents of Cross-Linked Matrimid Membrane at Various Cross-Linker Concentrations Samples 0% pPD 5% pPD 10% pPD 15% pPD

Gel Contents (%) No insoluble fraction 5.33 17.04 28.04

Chemical Cross-Linking Modifications of Polymeric Membranes

377

structure of the polymer matrix and the nodules cannot return to their original positions. The same phenomenon was reported by Aziz and Ismail (2010) and Tin et al. (2003). Lastly, the insertion of pPD cross-linking agents into the membranes, which may increase the space between the chains, could increase the d-space of the crosslinked Matrimid membranes. The same results were obtained as discussed by Aziz and Ismail (2010) when they varied the immersion time. This relects that both the parameters that affect the cross-linking reactions, immersion time and cross-linker concentrations, had changed the ordered dimensions and the interchain spacing of cross-linked PI membranes.

11.3.2

GAS TRANSPORT PROPERTIES OF CROSS-LINKED MATRIMID MEMBRANES

Table 11.5 presents the gas transport properties of unmodiied and modiied Matrimid membranes at different cross-linker concentrations. According to Table 11.5, for 5% pPD membrane, the gas permeance for all tested gases increases compared with the unmodiied membrane except for CO2. This behavior is mainly due to the swelling effects induced by methanol at the early stage of the cross-linking process. The swelling effect induced by the methanol forms interstitial spacing between the nodules in the polymeric chains, consequently allowing the gas molecules to permeate through it. The deviation of CO2 is attributed to the cross-linking effects, which converts imide groups to amide groups and enhances the interaction between CO2 and polymer chains, consequently hindering the transport of the gas across the membrane (Shao et al. 2008). The same phenomenon could be observed for 15% pPD membrane, though permeance largely increases compared to 5% pPD membrane. In contrast to 10% pPD membrane, the pressure-normalized lux slightly decreases compared with the unmodiied membrane. However, this result is not in agreement with that of Shao et al. (2008), who observed that a higher cross-linker concentration results in a greater decline in gas pressure–normalized lux. The main reason for this contradictory observation is that different cross-linker agents and different PI membranes were used. This implies that the diamino cross-linking using different types of cross-linking agents on different PI membranes results in different gas transport properties. The d-space of a polymeric membrane is generally related to its gas pressure–normalized lux, and a larger d-space commonly results in a higher pressure-normalized lux in amorphous polymers (Bulgarelli et al. 1999; Ghosal and Freeman 1994; Liu et al. 2003; Moeller et al. 2008; Powell et al. 2007; Shao et al. 2004, 2005a, 2008; Tin et al. 2003). Based on the XRD results, the increasing d-space follows the order: pure Matrimid (d-space: 5.46Å) < 10% pPD (5.59Å) < 5% pPD (5.75Å) < 15% pPD (5.90Å), while the increasing gas pressure–normalized lux for tested gases (i.e., O2) follows the order: pure Matrimid (0.71 GPU) < 10% pPD (0.91 GPU) < 5% pPD (1.15 GPU) < 15% pPD (2.62 GPU). Based on both these trends, the relation between the d-space and the gas permeance can be clearly explained: a larger d-space results in a higher pressure-normalized lux. In order to further clarify the relationship between the gas permeation properties and the membrane morphology in the crosslinked membrane, Figure 11.3 presents schematically, at the nodular level, the effect

378

Samples Cross-Linker Concentration% (w/v) 0 5 10 15 a

Standard deviation.

Permeance (GPU)

Selectivity

O2

N2

CH4

CO2

O2/N2

CO2/CH4

0.71 ± (0.027a) 1.15 ± (0.002) 0.91 ± (0.068) 2.62 ± (0.003)

0.11 ± (0.001) 1.07 ± (0.002) 0.09 ± (0.001) 2.68 ± (0.001)

0.26 ± (0.011) 1.23 ± (0.001) 0.20 ± (0.013) 4.77 ± (0.013)

4.50 ± (0.065) 0.97 ± (0.003) 0.65 ± (0.010) 4.29 ± (0.009)

6.19 1.07 10.01 0.98

17.07 0.78 3.29 0.90

Membrane Modification: Technology and Applications

TABLE 11.5 Gas Transport Properties of Unmodified and Modified BTDA-DAPI Polyimide Membranes

Chemical Cross-Linking Modifications of Polymeric Membranes

379

Nodules

(a) Without cross-linking pPD

Tie chains

(b) After cross-linking with 5% pPD

(c) After cross-linking with 10% pPD

(d) After cross-linking with 15% pPD

FIGURE 11.3 Phenomenological models of Matrimid membranes (a) without cross-linking; after cross-linking with (b) 5% pPD, (c) 10% pPD, and (d) 15% pPD.

of swelling and cross-linking on the membrane morphology. Note that each nodule includes a few hundreds of macromolecules and its size is a few hundreds of angstrom (Å). As can be seen in Figure 11.3a, there is a “tie-chain” by which the nodules are interconnected (Ismail and Lorna 2003). Macromolecules are intertwined at the boundary of two nodules and such intertwinement can be considered as the “tie-chain.” For 5% pPD samples, the swelling effect by methanol dominates the reactions. This is due to the low concentrations of pPD in cross-linking solutions, which lowers the cross-linking reactions that occur. In the diamine cross-linking method used in this study, the modiication process was initiated by the swelling of methanol, followed by the attack of the imide functional groups of Matrimid by the amino groups in pPD, resulting in the formation of a cross-linked structure (Tin et al. 2003). Upon swelling, the “tie-chain” is stretched, and the degree of stretching depends on the swelling power of methanol and the elasticity of the “tie-chain” (Figure 11.3b). Therefore, at this point, the pressure-normalized lux is found to attain the maximum value as the swelling of methanol increases the free volume, which consequently causes the membrane to lose its selectivity. However, for 10% pPD samples, the cross-linking

380

Membrane Modification: Technology and Applications

reaction is stronger than the swelling effect by methanol. The cross-linking reaction brings the nodules back nearly to their original positions (Figure 11.3c). Moreover, the cross-linking network starts to develop and results in a more compact nodular assembly and reduced free volume due to the space-illing by the cross-linker pPD. As a result, the gas pressure–normalized lux starts to decrease and improves the gas selectivities. Similar interpretation has been made by Xiao et al. (2007) and Tin et al. (2003). In contrast to the 15% pPD sample, the excess pPD units that diffused into the membrane created steric hindrances, which prevented excess densiication of polymeric chain packing by hindering the return of nodules back to their original positions (Figure 11.3d; Xiao et al. 2007). This contributed to an increased d-space for the 15% pPD sample, which consequently enhanced the gas pressure–normalized lux. The comparison of selectivities shows that all pPD–cross-linked samples have lower CO2/CH4 selectivity than those of the unmodiied Matrimid samples. These observations are consistent with the CO2 permeance results, which demonstrate that all pPD–cross-linked samples have lower CO2 permeance compared with the unmodiied membrane. This may be attributed to the cross-linking effects that hinder the CO2 gas transport across the membrane, which consequently affects the separation performance. However, a remarkable O2/N2 selectivity enhancement was observed for the 10% pPD samples, which showed an increase of the O2/N2 selectivity from 6.19 (pure Matrimid samples) to 10.01 (10% pPD samples). Compared with other studies reported on diamino cross-linking of PI membranes using different types of cross-linking agents such as EDA (Shao et al. 2005a, 2008), 1,3-propane diamine (PDA) (Shao et al. 2008), 1,4-butane diamine (BuDA) (Shao et al. 2008), p-xylenediamine (Tin et al. 2003), and 1,3-cyclohexanebis(methylamine) (CHBA) (Shao et al. 2005b), the pPD–cross-linking PI membrane used in this study had the highest O2/N2 selectivity. The impressive selectivity enhancement clearly demonstrates that pPD cross-linking can effectively modify the Matrimid membranes for O2/N2 separation. In addition, we also conclude that 10% (w/v) pPD/methanol solution is the optimum cross-linker solution concentration for Matrimid membrane modiication.

11.4

CONCLUSION

The chemical cross-linking modiication using pPD on Matrimid membranes has been successful. This success is based on the separation performance results; the cross-linking modiication enhanced the O2/N2 selectivity without much loss in pressure-normalized lux, especially for 10% (w/v) pPD/methanol solution. The characterization results demonstrated that the degree of cross-linking increased with cross-linker concentrations. In addition, the microstructural properties of the membranes are also altered due to methanol-swelling effects and structure-tightening effects induced by cross-linking. Clearly, from this chapter, different cross-linking mechanisms produce different effects on a given membrane system. Selecting the most suitable cross-linking or modiication technique can be dificult, but a detailed literature review, together with

Chemical Cross-Linking Modifications of Polymeric Membranes

381

a good knowledge of the application of the cross-linked membrane, will narrow the cross-linking method and polymer selection signiicantly. However, the problem of the depletion of gas permeability after cross-linking needs to be solved as soon as possible by further research, to make this modiication technique effective for producing a gas separation membrane with the upper bound limit performance.

REFERENCES Aziz, F. and Ismail, A.F. 2010. Preparation and characterization of cross-linked Matrimid® membranes using para-phenylenediamine for O2/N2 separation. Sep. Purif. Technol. 73: 421–428. Baker, W., Scott, C. and Hu, G-H. 2001. Reactive Polymer Blending. Carl Hanser Verlag: Munich, Germany. Barsema, J.N., Kapantaidakis, G.C., van der Vegt, N.F.A., Koops, G.H. and Wesling, M. 2003. Preparation and characterization of highly selective dense and hollow iber asymmetric membranes based on BTDA-TDI/MDI co-polyimide. J. Memb. Sci. 216: 195–205. Bulgarelli, E., Forni, F. and Bernabei, M.T. 1999. Casein:gelatin beads: I. Cross-linker solution composition effect on cross linking degree. Int. J. Pharm. 190: 175–182. Cao, C., Chung, T.S., Liu, Y., Wang, R. and Pramoda, K.P. 2003. Chemical cross-linking modiication of 6FDA-2,6-DAT hollow iber membranes for natural gas separation. J. Memb. Sci. 216: 257–268. Dudley, C.N., Schoberl, B., Sturgill, G.K., Beckham, H.W. and Rezac, M.E. 2001. Inluence of crosslinking technique on the physical and transport properties of ethynyl-terminated monomer/polyetherimide asymmetric membranes. J. Memb. Sci. 191: 1–11. Ghosal, K. and Freeman, B.D. 1994. Gas separation using polymer membranes: An overview. Polym. Adv. Technol. 5: 673–697. Hacarlioglu, P., Toppare, L. and Yilmaz, L. 2003. Effect of preparation parameters on performance of dense homogenous polycarbonate gas separation membranes. J. Appl. Polym. Sci. 90: 776–785. He, D., Susanto, H. and Ulbricht, M. 2009. Photo-irradiation for preparation, modiication and stimulation of polymeric membranes. Prog. Polym. Sci. 34: 62–98. Hosseini, S.S. and Chung, T-S. 2009. Carbon membranes from blends of PBI and polyimides for N2/CH4 and CO2/CH4 separation and hydrogen puriication. J. Memb. Sci. 328: 174–185. Ismail, A.F. and Lorna, W. 2003. Suppression of plasticization in polysulfone membranes for gas separations by heat-treatment technique. Sep. Purif. Technol. 30: 37–46. Ismail, A.F. and Yaacob, N. 2006. Performance of treated and untreated asymmetric polysulfone hollow iber membrane in series and cascade module conigurations for CO2/CH4 gas separation system. J. Memb. Sci. 275: 151–165. Ismail, A.F., Shilton, S.J., Dunkin, I.R. and Gallivan, S.L. 1997. Direct measurement of rheologically induced molecular orientation in gas separation hollow ibre membranes and effects on selectivity. J. Memb. Sci. 126: 133–137. Jiang, L.Y., Chung, T.-S. and Rajagopalan, R. 2008. Dehydration of alcohols by pervaporation through polyimide Matrimid asymmetric hollow ibers with various modiications. Chem. Eng. Sci. 63: 204–216. Kanehashi, S., Nakagawa, T., Nagai, K., Duthie, X., Kentish, S. and Stevens, G. 2007. Effects of carbon dioxide-induced plasticization on the gas transport properties of glassy polyimides membranes. J. Memb. Sci. 298: 147–155. Kapantaidakis, G.C., Kaldis, S.P., Dabou, X.S. and Sakellaropoulus, G.P. 1996. Gas permeation through PSF-PI miscible blend membranes. J. Memb. Sci. 110: 239–247.

382

Membrane Modification: Technology and Applications

Kapantaidakis, G.C., Koops, G.H. and Wessling, M. 2003. CO2 plasticization of polyethersulfone/polyimide gas-separation membranes. AIChE J. 49: 1702–1711. Kawakami, H., Mikawa, M. and Nagaoka, S. 1996. Gas transport properties in thermally cured aromatic polyimide membranes. J. Memb. Sci. 118: 223–230. Kim, J.H., Koros, W.J. and Paul, D.R. 2006. Effects of CO2 exposure and physical aging on the gas permeability of thin 6FDA- based polyimide membranes. Part 2. with crosslinking. J. Memb. Sci. 282: 32–43. Krol, J.J., Boerrigter, M. and Koops, G.H. 2001. Polyimide hollow iber gas separation membranes: Preparation and the suppression of plasticization in propane/propylene environments. J. Memb. Sci. 184: 275–286. Li, Y., Cao, C., Chung, T-S. and Pramoda, K.P. 2004. Fabrication of dual-layer polyethersulfone (PES) hollow iber membranes with an ultrathin dense-selective layer for gas separation. J. Memb. Sci. 245: 53–60. Liu, Y., Wang, R. and Chung, T-S. 2001. Chemical cross-linking modiication of polyimide membranes for gas separation. J. Memb. Sci. 189: 231–239. Liu, Y., Chng, M.L., Chung, T-S. and Wang, R. 2003. Effects of amidation on gas permeation properties of polyimide membranes. J. Memb. Sci. 214: 83–92. Luccio, M.D., Nobrega, R. and Borges, C.P. 2000. Microporous anisotropic phase inversion membranes from bisphenol-A polycarbonate: study of a ternary system. Polymer 41: 4309–4315. Miyata, S., Sato, S., Nagai, K., Nakagawa, T. and Kudo, K. 2008. Relationship between gas transport properties and fractional free volume determined from dielectric constant in polyimide ilms containing the hexaluoroisopropylidene group. J. Appl. Polym. Sci. 107: 3933–3944. Moeller, R., Lichter, J. and Blomeke, B. 2008. Impact of para-phenylenediamine on cyclooxygenases expression and prostaglandin formation in human immortalized keratinocytes (HaCaT). Toxicology. 249: 167–175. Nistor, C., Shishatskiy, S., Popa, M. and Nunes, S.P. 2008. Composite membranes with cross-linked Matrimid selective layer for gas separation. Environ. Eng. Manage. J. 7: 653–659. Pandey, P. and Chauhan, R.S. 2001. Membranes for gas separation. Prog. Polym. Sci. 26: 853–893. Peinemann, K-V. and Nunes, S.P. 2008. Membranes for life sciences. Wiley-VCH: Weinheim, German. Peter, J. and Peinemann, K-V. 2009. Multilayer composite membranes for gas separation based on crosslinked PTMSP gutter layer and partially crosslinked Matrimid® 5218 layer. J. Memb. Sci. 340: 62–72. Powell, C.E., Duthie, X.J., Kentish, S.E., Qiao, G.G. and Stevens, G.W. 2007. Reversible diamine cross-linking of polyimide membranes. J. Memb. Sci. 291: 199–209. Recio, R., Palacio, L., Pra’danos, P., Herna’ndez, A., Lozano, A.E., Marcos, A., de la Campa, J.G. and de Abajo, J. 2007. Gas separation of 6FDA–6FpDA membranes. Effect of the solvent on polymer surfaces and permselectivity. J. Memb. Sci. 293: 22–28. Rezac, M.E., Sorensen, E.T. and Beckham, H.W. 1997. Transport properties of crosslinkable polyimide blends. J. Memb. Sci. 136: 249–259. Robeson, L.M. 1999. Polymer membranes for gas separation. Curr. Opin. Solid State Mater. Sci. 4: 549–552. Sen, D., Kalipcilar, H. and Yilmaz, L. 2006. Gas separation performance of polycarbonate membranes modiied with multifunctional low molecular-weight additives. Sep. Purif. Technol. 41: 1813–1828. Shao, L., Chung, T.-S., Goh, S.H. and Pramoda, K.P. 2004a. Transport properties of crosslinked polyimide membranes induced by different generations of diaminobutane (DAB) dendrimers. J. Memb. Sci. 238: 153–163.

Chemical Cross-Linking Modifications of Polymeric Membranes

383

Shao, L., Chung, T.-S., Wensley, G., Goh, S.H. and Pramoda, K.P. 2004b. Casting solvent effects on morphologies, gas transport properties of a novel 6FDA/PMDA-TMMDA copolyimide membrane and its derived carbon membranes. J. Memb. Sci. 244: 77–87. Shao, L., Chung, T.-S., Goh, S.H. and Pramoda, K.P. 2005a. Polyimide modiication by a linear aliphatic diamine to enhance transport performance and plasticization resistance. J. Memb. Sci. 256: 46–56. Shao, L., Chung, T.-S., Goh, S.H. and Pramoda, K.P. 2005b. The effects of 1,3-cyclohexanebis(methylamine) modiication on gas transport and plasticization resistance of polyimide membranes. J. Memb. Sci. 267: 78–89. Shao, L., Liu, L., Cheng, S.X., Huang, Y.D. and Ma, J. 2008. Comparison of diamine cross linking in different polyimide solutions and membranes by precipitation observation and gas transport. J. Memb. Sci. 312: 174–185. Tin, P.S., Chung, T.-S., Liu, Y., Wang, R., Liu, S.L. and Pramoda, K.P. 2003. Effects of crosslinking modiication on gas separation performance of Matrimid membranes. J. Memb. Sci. 225: 77–90. Wang, D., Li, K. and Teo, W.K. 1995. Effects of temperature and pressure on gas permselection properties in asymmetric membranes. J. Memb. Sci. 105: 89–101. Wang, D., Li, K. and Teo, W.K. 1996. Polyethersulfone hollow iber gas separation membranes prepared from NMP/alcohol solvent systems. J. Memb. Sci. 115: 85–108. Wind, J.D., Bickel, C.S., Paul, D.R. and Koros, W.J. 2003a. Solid-state covalent cross-linking of polyimide membranes for carbon dioxide plasticization reduction. Macromolecules 36: 1882–1888. Wind, J.D., Sirard, S.M., Paul, D.R., Green, P.F., Johnston, K.P. and Koros, W.J. 2003b. Carbon dioxide-induced plasticization of polyimide membranes: Pseudo-equilibrium relationships of diffusion, sorption and swelling. Macromolecules 36: 6433–6441. Wright, C.T. and Paul, D.R. 1997. Feasibility of thermal crosslinking of polyarylate gasseparation membranes using benzocyclobutene-based monomers. J. Memb. Sci. 129: 47–53. Xiao, Y., Chung, T.-S., Guan, H.M. and Guiver, M.D. 2007. Synthesis, cross-linking and carbonization of co-polyimides containing internal acetylene units for gas separation. J. Memb. Sci. 302: 254–264. Yi, C., Wang, Z., Li, M., Wang, J. and Wang, S. 2006. Facilitated transport of CO2 through polyvinylamine/polyethlene glycol blend membranes. Desalination 193: 90–96.

12

Development of Fuel Cell Polymer Electrolyte Membranes by Radiation-Induced Grafting with ElectronBeam Irradiation Mohamed Mahmoud Nasef

CONTENTS 12.1 Introduction .................................................................................................. 385 12.2 Polymer Electrolyte Membranes .................................................................. 386 12.2.1 Polymer Electrolyte Membrane Fuel Cell ........................................ 387 12.2.2 Current Status of Polymer Electroltye Membranes .......................... 388 12.3 Sources for High-Energy Radiation ............................................................. 390 12.3.1 Electron-Beam Accelerators ............................................................. 391 12.4 Radiation-Induced Grafting.......................................................................... 392 12.4.1 Methods of Radiation-Induced Grafting .......................................... 393 12.5 Preparation of Polymer Electrolyte Membranes Using Electron Beam ....... 394 12.5.1 Composite Sulfonic Membranes for Direct Methanol Fuel Cell ...... 395 12.5.2 Sulfonic Acid Membrane for Polymer Electrolyte Membrane Fuel Cell by a Single Grafting Route ............................................... 399 12.6 Conclusion ....................................................................................................404 Acknowledgment ...................................................................................................405 References ..............................................................................................................405

12.1

INTRODUCTION

Membrane systems are one of the most effective and fast-growing technologies that can address industrial and research separation and puriication needs in the twentyirst century. The attraction of membrane systems stems from their energy-eficient operation, and modularity, combined with their low operational cost, as compared with conventional techniques. Since membrane is the heart of the system, there is 385

386

Membrane Modification: Technology and Applications

an increasing necessity to develop versatile and economical membrane materials (Kariduraganavar et al. 2006). Development of membrane materials and membrane processes continues to receive increasing attention at laboratory and industrial scales. This is due to the versatility of the membrane materials and processes in various ields of industrial interest, including separation and puriication in chemical and biochemical industries, environmental remediation, and energy production. Among all these, energy production systems and devices based on membranes have the potential to reduce cost, enhance eficiency, eliminate environmental pollution, and give an overall performance beyond current conventional benchmarks (Ciuffa et al. 2004). The performance of the membranes is a key factor in determining their application for speciic processes. The suitability of a membrane for a certain application is determined by its speciic response to the separation medium, which is dictated by a combination of the physical and chemical nature of the developed membrane during its preparation. Furthermore, both physical and chemical factors are essential to the control of the critical relationship between the membrane structure and performance in speciic operations (Gupta et al. 2004). Today, various types of membrane materials are available, which can be easily divided into two categories based on their origin: organic and inorganic. The organic membranes can be further subdivided into neutral and ionically charged conductive membranes. The former class is normally used in processes such as reverse osmosis, ultrailtration, pervaporation, and gas separation, whereas the latter is mainly used in electrochemical processes. Among all membranes, polymer electrolyte membranes (PEMs) are an advanced category of polymeric materials that play an important role in the development of practical electrochemical power devices (Kalhammer 2000). Currently, PEMs are still dominated by Naion, one of the most important commercial membranes, despite various efforts to develop alternative analogous materials (Banerjee and Curti 2004). Naion membrane was developed by Dupont and has a perluorinated structure with pendant sulfonic acid groups in side chains. Its performance and stability have been excellent in systems operating under highly corrosive environments such as in chloralkali electrodialyzers, water electrolyzers, and fuel cells (Heitner-Wirguin 1996).

12.2

POLYMER ELECTROLYTE MEMBRANES

PEMs are semipermeable membranes made from charged (ion-containing) polymers commonly available in copolymer or composite form and designed to have transport properties comparable with that of common liquid ionic solutions. These membranes have generated considerable interest in various ields of industries, including electrochemical processes such as water electrolyzers and dialyzers and devices such as fuel cells, batteries, sensors, and actuators. Figure 12.1 shows a schematic representation of various applications of PEMs. The PEMs have common functions when applied in such electrochemical devices: (1) Mechanical function to prevent mixing of reactants and (2) Chemical function to selectively transport speciic ions from the anode to the cathode. These polymer electrolytes have several advantages over their liquid counterparts. For instance, they have no internal shorting and no leakage of

Development of Fuel Cell Polymer Electrolyte Membranes

387

Polymer electrolyte membrane applications Electrochemical/ separation processes

Energy conversion

Sensing and actuating

Water electrolyzers

Fuel cells

Sensors

Electrodialyzers

Batteries

Actuator

FIGURE 12.1 Various applications of polymer electrolyte membranes.

electrolyte and noncombustible reaction products at the electrode surface as found in the liquid electrolytes (Scrosati 1993). Membranes conducting lithium and protons are the most famous among this class of materials, which are currently undergoing intensive research efforts to promote the commercialization of fuel cell and polymer electrolyte lithium batteries (Stephan 2006). The performance of PEMs mainly depends on the nature and content of the incorporated ixed ionic groups, together with the microstructure developed during the preparation procedure. Homogeneous distribution of ixed ionic groups across the membranes renders them highly conductive with low internal resistance, whereas concentration of ionic groups on or near surface layers causes an increase in the internal resistance, which eventually leads to a reduction in the electrochemical system eficiency. Thus, the diversity in applications of PEMs implies a strong demand to design these membranes to attain particular properties suitable for speciic applications. The most common requirements for most applications of PEMs despite their diversity include: (1) Operation at high temperature, (2) Chemical stability, (3) Low resistance, and (4) High selectivity (Nasef and Hegazy 2004).

12.2.1

POLYMER ELECTROLYTE MEMBRANE FUEL CELL

PEM or proton exchange membrane fuel cells (PEMFCs) are emerging power sources that have high power density and eficiency, low emission, no moving parts, and silent low-temperature operation, making them suitable for stationary, mobile, and portable power-generation applications. PEMFC converts the chemical energy liberated from fuel such as pure H2 or reformed hydrocarbons through an electrochemical reaction in the presence of an oxidant, oxygen or air, to electrical energy in the form of DC current. It consists mainly of PEM sandwiched between two gas diffusion electrodes loaded with a small amount of platinum catalyst at the surfaces facing the membranes. A schematic representation of the PEMFC basic unit is shown in Figure 12.2. PEM is an essential component in PEMFC, which has crucial functions that include (1) Prevention of the bulk mixing of H2 and O2, (2) Transportation of protons

388

Membrane Modification: Technology and Applications Electrical current Excess fuel

e−

e−

Water and heat out

e− e−

H+ H+

H2

H+

H2O

O2

H+ H2

O2

Anode 2H2

Electrolyte

4H+ + 4e− Overall: 2H2 + O2

Cathode O2 + 4H+ + 4e−

2H2O

2H2O + Energy

FIGURE 12.2 A schematic representation of a basic PEMFC unit.

(H+) Produced by the dissociation of H2 at the anode to the cathode, (3) Prevention of the associated electron low through the membranes forcing them to low in the external circuit to the cathode to produce DC current, and (4) Support for the catalyst loaded on the electrodes. When H2 is replaced by methanol as a fuel in liquid form in direct methanol fuel cell (DMFC), the dissociation of methanol solution at the anode produces protons that are transported through the hydrated PEM to the cathode, where a reduction of O2 produces water in the presence of the protons. To qualify PEM for commercial application in PEMFC and DMFC, it should have a combination of properties including (Maiyalagan and Pasupathi 2010; Neburchilov et al. 2007; Nagarale et al. 2010): • • • • • • •

Hydromechanical and thermal stability at 70°C–130°C High proton conductivity (≥10−2 S/cm) with low area resistance (0.1–0.5 Ω cm2) Electronic inductivity Low gas permeability (for H2) Low methanol crossover (in DMFC) High oxidative stability Long-term stability (≥5000 h mobile applications and ≥10,000 h stationary application) • Low cost

12.2.2 CURRENT STATUS OF POLYMER ELECTROLTYE MEMBRANES Several PEMs are commercially available, including Naion (DuPont), Dow (Dow Chemicals), Aciplex (Asahi Chemicals Co.), Aciplex-S (Asahi Kasei) based on a weak

Development of Fuel Cell Polymer Electrolyte Membranes

389

functional acid –COOH, Flemion (Asahi Glass Co.), Gore-Tex (Gore and Associate), BAM 3G (Ballard), CRA and CRS (Solvay), and Dais membranes (Dais Co.) (Nasef 2008). Of all these membranes, Naion is the most established membrane that has been widely tested, and the majority of the available fuel cell systems are based on it. However, Naion is regarded as expensive (U$700–1000/m2), has high methanol permeability (in DMFC), and loses water during operations at high temperature. Alternatively, considerably cheaper sulfonated poly(ether ether ketone) membranes (SPEEK) are also available (U$375/m2) (Agro et al. 2005). Nonetheless, other alternative cost-effective and highly conductive membranes are still being sought. Thus, different PEM materials are under intensive research in many laboratories, with novel membranes at various stages of development ranging from laboratory scale to near commercialization. Various approaches have been considered to develop new alternative membranes, all of which aim at cost reduction and circumvention of the drawbacks associated with current commercial membranes. The irst approach includes formation of Naion composites or modiication of Naion membranes involving the use of surface coatings and loading. The second approach involves direct sulfonation of nonluorinated polymer backbones such as polystyrene (PS), polyphosphazene, polyphenylene oxide, polysulfone, polyether sulfone, polyether ether ketone, polybenzimidazole, and polyimides. The challenge in this approach is to achieve suficient sulfonation for high proton conductivity in the membranes without the polymer becoming soluble. The third approach involves sulfonation of pendant aromatic rings attached to a variety of grafted copolymer ilms obtained by chemical, plasma, thermal, or radiochemical graft copolymerization of styrenic monomers. The latest progress on the development of new membranes for fuel cells using such approaches and their different classes has recently been reviewed in several articles (Maiyalagan and Pasupathi 2010; Neburchilov et al. 2007; Nagarale et al. 2010; Nasef 2008; Gubler and Günther 2010). Of all these, radiation-induced grafting (RIG) of styrene onto various luoropolymer ilms has been found to be the most effective method for the preparation of alternative PEMs for fuel cells (Nasef and Hegazy 2004; Nasef 2008; Gürsel et al. 2008). The use of RIG for the preparation of membranes provides several advantages such as (1) Simplicity of the procedure; (2) Proper control over the composition of the obtained membranes, leading to the achievement of tailored properties; (3) Use of various high-energy radiation sources such as γ-rays and electron beam (EB) for grafting initiation; and (4) Absence of shaping problems as the reaction starts from a polymer substrate in sheet form, which makes it attractive for addressing the challenging task of developing PEMs for fuel cells. The objective of this chapter is to review the latest progress in the preparation of radiation-grafted PEMs for fuel cells using EB. In particular, the results associated with the preparation and the properties of two types of membranes having polystyrene sulfonic acid (PSSA) grafted to poly(vinylidene luoride) (PVDF) ilms are reviewed. The fundamentals of RIG operation with EB accelerators are also briely reviewed to furnish a clear understanding of the mechanism of PEM formation and how membrane composition and, consequently, their properties can be controlled.

390

12.3

Membrane Modification: Technology and Applications

SOURCES FOR HIGH-ENERGY RADIATION

Different types of high-energy radiation sources are commercially available and can be utilized for polymer processing including grafting, surface curing, and crosslinking. They fall into two categories: (1) Electromagnetic radiation or photons such as γ-rays and x-rays and (2) Particulate radiation (charged particles) such as electrons and swift heavy ions. Radioactive isotopes such as cobalt-60 (60Co) and cesium-137 (137Cs)—ission products of nuclear plants—are the main sources of γ-rays, with 60Co being more advantageous and preferred practically due to its higher energy emission (1.25 MeV, compared with 0.66 MeV for 137Cs), simplicity, ease of preparation, and low cost (Nasef and Hegazy 2004). Two different types of γ-ray sources are commercially available: cave for industrial scale and cavity for laboratory scale operations. More details on the setup of such facilities and their operation mechanism can be found elsewhere (Ivanov 1992). On the other hand, particulate radiations such as electrons are normally obtained from EB accelerators, which are commercially available with different designs and acceleration energies. Considering application, γ-rays are usually most economical at lower doses (~80 kGy and below) and for large, high-density parts due to their deep penetration. EB is commonly used for irradiating small parts, particularly low-density parts, and linear products processed reel-to-reel such as wire, cable, and tubing (Cleland and Parks 2003). A comparison between the two high-energy radiation sources of isotopes and electron accelerators is presented in Table 12.1. A brief account of EB principle of operation, types, and application is given in the next section.

TABLE 12.1 Comparison of the Sources of High-Energy Radiation: Isotopes and Electron Accelerators Features

Radioisotopes

Electron Accelerators

Penetration depth

High

Operation Mode of operation

Simple Single

Limited to acceleration energy Relative complex Dual (x-ray and electrons)

On/off function Throughput Radiation direction Power, energy, and geometry Dose rate Processing time Nuclear waste Cost Industrial application

No Low Scattered No control Low Long Yes Cheap Established

Yes High Well-directed Well-controlled High Short No Relatively high Growing

Source: From Nasef, M.M. and Hegazy, E.A., Prog. Polym. Sci., 29, 499–561, 2004. With permission.

Development of Fuel Cell Polymer Electrolyte Membranes

12.3.1

391

ELECTRON-BEAM ACCELERATORS

EB accelerator is a machine that uses electrical energy to generate free electrons, accelerates them to high speeds, as schematized in Figure 12.3, and then directs them at materials passing through the accelerator on a conveyor or in another type of low-through system. Like a TV cathode tube, EB produces a cloud of free electrons by heating a negative cathode inside a vacuum chamber. Once generated, the negatively charged electrons are attracted by a positive electrical potential (≅10 kV) applied to an attracting plate (anode). The electrons are accelerated by traveling through the electric ield, thereby gaining energy. These accelerated electrons are collimated through a window in the anode plate and directed toward the materials to be treated. The accelerator, which generates the electrons, operates in both pulse and continuous beam modes (Bly 1988). Industrial electron accelerators are classiied into three categories depending on the acceleration energy and application: (1) Low-energy accelerators (0.1–0.5 MeV), (2) Medium-energy accelerators (0.5–5.0 MeV), and (3) High-energy accelerators (5.0–10.0 MeV). Low-energy accelerators are used when low penetration is needed as in surface curing, whereas medium- and high-energy accelerators are used in applications requiring more penetration including grafting, cross-linking, degradation, sterilization, and food processing (Mondelaers 1998). The amount of energy absorbed by a material is traditionally expressed as dose in “rad,” which is the measure of the amount of energy deposition in 1 g of the matter and equals 102 erg/g or 10−2 J/kg. The SI unit is Gray (Gy), which is equal to 104 erg/g or 1 J/kg. So, the dose rate, which is the absorbed dose per unit time, can be expressed as Gy/s, and kilogray (kGy/s) is used to express the large amount of absorbed doses. However, some sources continue to use the old unit Megarad (Mrad), which is equal to 10 kGy, as a dose expression. The absorbed dose affects the amount of radical produced in the polymer and can initiate grafting in the presence of a monomer or cross-linking together with other competing reactions such as chain scission that leads to lower-molecular-weight

−

Electrical source

Hot filament

+ Positive plate (anode) Electron gun

Electrons emission Cathode

FIGURE 12.3 accelerator.

Electrical field (vacuum chamber)

Positive plate + (anode)

Schematic diagram of the principle of operation of an electron-beam

392

Membrane Modification: Technology and Applications

products (Nasef and Hegazy 2004). Generally, an increase in the absorbed dose causes an increase in the formation of radicals in the polymers. Particularly, electrons produced in polymers by EB are generally most effective for performing RIG and when the preirradiation method is sought. It is also most effective when industrial usage and pilot-scale production are desired because of its high irradiation dose rate, short processing time, ease of generation of free radicals in many polymers, and moderate initiation reaction conditions. Depending on the acceleration energy of EB, current throughput, and reaction parameters, it is possible to extend the chemical modiication to speciic depths ranging from the surface to the bulk of the polymer ilms (Nasef 2003).

12.4

RADIATION-INDUCED GRAFTING

Grafting or graft copolymerization is a process in which a side chain (graft) is covalently attached to a polymer backbone forming a graft copolymer. Grafting takes place as a result of formation of active sites initiated on the polymer backbone, which can be free radicals or ionic chemical groups responsible for initiating polymerization reaction. The formation of active sites can be achieved using several initiating methods such as plasma treatment, ultraviolet, decomposition of chemical initiator, and high-energy radiation (Prakash et al. 2004; Qiu et al. 2003; Finsterwalder and Hambitzer 2001). Of all these methods, grafting with high-energy radiation, also known as radiation-induced grafting (RIG), provides the most versatile method for modiication of preexisting polymers available in various morphologies such as resins, ibers, ilms, and fabrics without changing their inherent properties. It offers a unique way to combine two highly incompatible polymers and imparts new properties to the obtained graft copolymers (Shin et al. 2005). Applying this method enables changing polymer wettability, adhesion, printability, metalization, antifog properties, antistatic properties, and biocompatibility (Nasef and Hegazy 2004). Grafting with RIG can be initiated in a wide range of temperatures including subambient temperature, in monomers of various states such as bulk solution, emulsion, vapor, and even at solid state. Thus, RIG has been widely investigated for the preparation of PEMs and ion exchange membranes as bulk modiication of polymer ilms can be achieved (Nasef and Hegazy 2004) unlike plasma-induced and UV-induced grafting, which produces surface modiication in polymers (Finsterwalder and Hambitzer 2001). In the RIG method, the reaction usually follows free-radical polymerization mechanism: initiation, propagation, and termination. The active sites are formed on the polymer backbone upon controlled exposure to high-energy radiation, and the irradiated polymer is subsequently brought into contact with the monomer units, which then form macroradicals that propagate to form side-chain grafts of particular molecular weight when terminated. Grafting is often done by a front mechanism (Gupta and Scherer 1994; Gubler et al. 2005), where grafting starts at surface layers by monomer diffusion and moves inward following progressive diffusion. This continues until grafting fronts meet in the middle of the polymer ilm, leading to homogeneous polymer diffusion as depicted in Figure 12.4, which shows cross-sectional

393

Development of Fuel Cell Polymer Electrolyte Membranes

(a)

(b)

(c)

(d)

FIGURE 12.4 Scanning optical microscope images of cross sections of (a) original PVDF and dyed grafted membranes with various G%, (b) 12%, (c) 29%, and (d) 53%, indicating a front mechanism. (From Nasef, M.M., Saidi, H., and Dahlan, K.Z.M., Radiat. Phys. Chem., 80, 66–75, 2011. With permission.)

views obtained by scanning optical microscopy of PVDF ilms grafted with PS sulfonate at different degrees of grafting (Nasef et al. 2011).

12.4.1

METHODS OF RADIATION-INDUCED GRAFTING

Two standard methods of RIG that have been developed as depicted in the schematic diagram shown in Figure 12.5 are simultaneous (direct) irradiation and preirradiation. In the simultaneous method, a polymer is irradiated while immersed in the monomer solution lushed with inert gas or evacuated through freeze–thaw cycles. An inhibitor is often added to the solution to suppress the homopolymerization in the liquid phase. Alternatively, in the preirradiation method, the polymer is irradiated irst in an inert gas or vacuum and subsequently brought into contact with the Radiation-induced grafting methods Simultaneous/direct irradiation

+

Preirradiation

Irradiation in O1 or air/peroxidation

Irradiation in vacuum or inert atmosphere

Monomer OO + Monomer

+ + Monomer

Graft copolymer

FIGURE 12.5 A schematic representation of the various radiation-induced grafting methods.

394

Membrane Modification: Technology and Applications

monomer solution under controlled conditions in a grafting apparatus. When the irradiation step is carried out in air, the radicals generated in the polymer backbone react with oxygen, forming peroxide and hydroperoxide radicals and the preirradiation method is then called “peroxidation.” The grafting reaction is then initiated by thermal decomposition of the trapped radicals in the polymer backbone in the presence of monomer units under controlled conditions. Each of the RIG methods has its advantages and disadvantages. For instance, the simultaneous irradiation method is very simple in nature and highly eficient from the viewpoint of polymer radiation chemistry; it produces higher degrees of grafting due to eficient utilization of radicals, provided that the parallel homopolymerization is controlled (Nasef and Hegazy 2004). On the other hand, the preirradiation and peroxidation methods are very effective when highly reactive monomers such as acrylic acid are being grafted and for pilot-scale production in the form of a continuous production line. It is also more convenient as the polymer substrate can be irradiated and preserved at subambient temperatures for a long time after exposure to the radiation source. Nevertheless, obtaining desired grafting levels in both methods requires an optimization of the reaction parameters such as monomer concentration, irradiation dose, dose rate, temperature, and type of solvent. Such optimization leads to a variation in the depth of penetration of the monomer into the bulk of the polymer substrate to achieve uniformity and homogeneity and eventually allow a control over the membrane composition. The degree of grafting can be calculated by: G (%) = 100 × [(W2 − W1)/W1], where W2 and W1 are the weights of the ilm after and before grafting, respectively.

12.5

PREPARATION OF POLYMER ELECTROLYTE MEMBRANES USING ELECTRON BEAM

Many types of membranes have been prepared by grafting of various monomers onto polymer ilms using EB. Ion exchange membranes are the major class that have been prepared using RIG, starting from common polymer ilms. Preferably, the polymer ilms have to be hydrophobic materials of high radiation resistance, thermal stability, and chemical and mechanical strength, such as luorine-containing polymers except poly(tetraluoroethylene), which is radiation-sensitive at low temperatures despite having outstanding properties. The monomer to be grafted determines the type, nature, and functionality of the obtained membrane. Two types of grafting monomers are available for grafting: (1) Functional monomers such as acrylic acid and methacrylic acid, both of which directly confer a weakly acidic (cationic) character to the polymer backbone unlike sodium styrene sulfonate (SSS), which confers a strongly acidic character; and (2) Nonfunctional monomers such as styrene and 4-vinylpyridine, which produce grafted ilms that can be activated in a postgrafting chemical reaction converting them into acidic or basic membranes by sulfonation or quaternization. The use of radiation grafting as an alternative technique for the preparation of various membranes, together with the fundamental concepts of RIG and the effects of grafting conditions, has been extensively reviewed by Nasef and Hegazy (2004).

Development of Fuel Cell Polymer Electrolyte Membranes

Microporous substrate

395

Composite membrane

FIGURE 12.6 A schematic representation of the concept of composite membranes.

12.5.1

COMPOSITE SULFONIC MEMBRANES FOR DIRECT METHANOL FUEL CELL

Composite pore-illed membranes, which have a porous polymer structure illed with a polyelectrolyte as schematized in Figure 12.6, are receiving growing attention in the ield of separation and puriication and recently in fuel cell application (Yamaguchi et al. 2003). In these membranes, the porous polymer substrate acts as an inert host that constrains the swelling of an anchored polyelectrolyte and provides high mechanical strength for the obtained membranes. Recently, Nasef et al. (2006a,b) reported on the preparation of composite membranes using simultaneous irradiation with EB for the irst time and proposed them for DMFC application. The membrane preparation started by loading a porous PVDF ilm (porosity 70%) with styrene monomer, followed by direct irradiation with EB and subsequent sulfonation. The EB was operated in the low-energy region (as presented in Table 12.2) to reduce the effects of the undesirable side reactions— homopolymerization and chain scission of the polymer backbone—that may hinder the copolymerization of styrene or cause the degradation of the PVDF ilm. The porous structure of the obtained membranes was illed with varying amounts of PS in the range of 8–46% by alteration of the absorbed doses in the range of 5–50 kGy. The obtained grafted ilms were treated with 10% (v/v) chlorosulfonic acid in dichloromethane for a period of 12 h at room temperature under an N2 atmosphere to achieve degrees of sulfonation close to 100% in all membranes, and G% represents the content of PSSA. The evidence of PS grafting and subsequent sulfonation is given in the FTIR spectral analysis shown in Figure 12.7. The properties of the obtained membranes having various G% are presented in Table 12.3. The essential electrochemical properties of all membranes, which include water uptake, methanol

TABLE 12.2 Electron-Beam Operating Parameters Applied for the Preparation of Composite Membrane Accelerating voltage Beam current Dose per pass Conveyor speed Atmosphere Temperature

200 keV 2 mA 10 kGy 2.25 m/min N2 28°C

396

Membrane Modification: Technology and Applications

( CF2

CH2 (

n

95 90 85 (a) 80

(CF2

T (%)

75

CH)n ( CH2 CH )m

70 (b) 65 60

(CF2

55

CH)n ( CH2 CH )m

50 SO3H

45 (c)

3800 3600 3400 3200 3000 2800 2600 2400 2200 2000 1800 1600 1400 1200 1000 800 Wave numbers (cm–1)

FIGURE 12.7 Typical FTIR spectra of (a) pristine PVDF ilm, (b) PS composite PVDF ilm (G = 30%), and (c) its corresponding sulfonated membrane. (From Nasef, M.M., Zubir, N.A., Ismail, A.F., Dahlan, K.Z.M., Saidi, H., and Khayet, M., J. Power Sour., 156, 200–210, 2006. With permission.)

uptake, ion exchange capacity (IEC), proton conductivity (σ), and methanol permeability (P), were found to increase with the increase in G% as a result of the increase in PSSA content. Particularly, membranes with G = 40% and 46% demonstrated higher conductivity and remarkably lower methanol permeability than Naion 117 membrane under similar experimental conditions, thus suggesting that they are potential alternative candidates for DMFC. To establish structure–property relations, the variation in the physical, thermal, and structural properties in correlation TABLE 12.3 Properties of PSSA-Containing Composite Membranes Have Various G% G (%) 0 8 19 30 40 46 Naion 117

IEC (mmol/g)

Uptake H2O

Uptake CH3OH

σ (S/cm) × 10−3

P (cm2/s)

— 0.7 1.3 1.8 2.2 2.3 0.91

— 17.2 37.5 51.2 61.0 72.0 39.0

— 6.0 12.8 22.2 26.8 30.5 63.0

— 9 26 43 54 61 53

— 2.6 E−4 1.5 E−5 4.9 E−6 1.6 E−6 1.0 E−6 3.5 E−6

Source: From Nasef, M.M., Zubir, N.A., Ismail, A.F., and Khayet, M., Desalination, 200, 642–644, 2006. With permission.

397

Development of Fuel Cell Polymer Electrolyte Membranes

TABLE 12.4 Variation of the Physical, Thermal, and Structural Properties of Composite Membranes with Different G% G (%) 0 8 19 30 40 46

Thickness (μm)

ρo (g/cm3)

ɛ (%)

Tm (°C)

Tc (°C)

Xc (%)

118 125 132 140 146 152

0.72 0.77 1.10 1.27 1.55 1.63

70 43 32 21 12 5

163.6 162.1 160.1 157.9 156.1 154.2

141.2 139.1 137.7 136.5 135.3 134.4

37.1 18.4 13.2 10.5 8.9 7.2

Source: From Nasef, M.M., Zubir, N.A., Ismail, A.F., and Khayet, M., Desalination, 200, 642– 644, 2006. With permission.

with G% was also investigated and the obtained data are presented in Table 12.4 (Nasef et al. 2006c). As can be observed, there is a decrease in the porosity of the PS pore-illed PVDF ilms, which is accompanied by a simultaneous increase in their apparent densities and thicknesses. This trend is due to the increase in the pore-illing ratio in the PVDF matrix. A sharp drop in the porosity from 70% with no grafting to a value of 5% at G = 46% was observed, suggesting that the effectiveness of the applied procedure in plugging the PVDF porous structure to meet the DMFC requirements. This result is conirmed by the SEM morphological analysis shown in Figure 12.8, in which the emergence of a new phase structure composed of spherical granules of PS aggregates in the form of distinctive gel-like clusters upon sulfonation can be observed. Both melting temperature (Tm) and crystallization temperature (Tc) decrease with the increase in the PSSA content in the membranes. The reduction in Tm is an indication of a crystalline disorder, which is conirmed by the reduction in the degree of crystallinity (Xc) of the membranes with the increase in the PSSA content. The

(a)

(b)

(c)

FIGURE 12.8 SEM images of cross sections of (a) pristine PVDF ilm, (b) grafted polystyrene pore-illed PVDF ilm (G = 30%), and (c) its corresponding sulfonated composite membrane. (From Nasef, M.M., Zubir, N.A., Ismail, A.F., Dahlan, K.Z.M., Saidi, H., and Khayet, M., J. Power Sour., 156, 200–210, 2006. With permission.)

398

Membrane Modification: Technology and Applications

reduction in Tc is an indication of mixing of some of the PSSA grafts into the crystalline fraction of the PVDF melt. However, the PSSA is not compatible with the PVDF polymer and mostly forms a separated phase within the amorphous region of the membrane. The liquid uptake was found to be dependent on Xc despite the argument that grafting of the PS in porous PVDF membranes takes place in the vicinity of the pores of the PVDF matrix starting from the pore walls and continues until the pore is plugged or the monomer is consumed. This indicates that grafting takes place in the amorphous fraction of the PVDF ilm. Therefore, the reduction of crystallinity is suggested to be due to the penetration of the aggressive sulfonic acid to the crystalline structure during the sulfonation step, causing a combined effect of disruption and dilution that grows with the increase in sulfonic acid content. The free water or methanol trapped into the unplugged pores, together with the solvation of the sulfonic acid groups, determines the liquid absorbance capacity of these membranes. However, the increased interaction with water, compared with methanol, shown by these membranes is limited to the solvation of PSSA domains. The proton transfer represented by the ionic conductivity (σ) was found to be a function of collective structural changes: the increase in both hydrophilicity and SO3− content (IEC increase) together with the reduction in both porosity and Xc. Since grafting took place in the pores of the PVDF ilm and in its structure, there is a strong possibility that the hydrated PSSA forms a network combining the pores within the polymer structure for the proton transport at high PSSA contents (e.g., 40 and 46%). The methanol permeation (P) was found to decrease with the reduction in the porosity and crystallinity accompanying the increase in the content of PSSA. This was despite the increase in the methanol uptake and the IEC. The reduction in the porosity marginalizes the effect of the free methanol trapping, which limits the uptake to the PSSA domains, which have preferential selectivity for water over methanol. On the other hand, the reduction in Xc seems to play no signiicant role in methanol permeation as PVDF has a low glass transition temperature, which should enhance the PSSA chain mobility, allowing more methanol uptake. The performance of the PSSA composite membranes in the DMFC under dynamic conditions is depicted in Figure 12.9 and shows typical polarization curves for membranes with G of 40 and 46% at methanol concentration of 5 M and operating temperature of 70°C. The membranes showed clear regions of activation and ohmic and concentration polarization. The performance in terms of the open cell voltage and power density was found to be 0.820 and 0.830 V and 110 and 120 mWA/cm2 for 40 and 46% grafted membranes, respectively. The difference is due to the higher conductivity, enhanced by IEC, and increased water uptake in the 40% grafted membrane, compared with the 60% grafted one. However, such performance is very good for these membranes when compared with other membranes based on the PVDF reported in the literature (Nasef 2003; Qiu et al. 2003; Nasef et al. 2006b). In conclusion, the use of simultaneous EB irradiation grafting method is found to simplify the pore-illing process and reduce the reaction time as well as the monomer consumption. This method provides membranes with a very good combination of properties and reasonable performance in the DMFC.

399

Development of Fuel Cell Polymer Electrolyte Membranes 1000 5 M methanol, Temp 70°C

160

Anode: 5 M method and 0.2 ml/min. Cathode: dry air and 100 ml/min.

800

140

Potential (mV/cm2)

700

120

600 100 500 80 400 300

60

200

40

100

G = 40%

Power density (mWA/cm2)

900

20

G = 46%

0 0

50

100

150

200

250

300

350

Current density (mA/cm2)

FIGURE 12.9 Current–voltage characteristics and power density curves of a DMFC with PVDF-g-PSSA having various G%.

12.5.2

SULFONIC ACID MEMBRANE FOR POLYMER ELECTROLYTE MEMBRANE FUEL CELL BY A SINGLE GRAFTING ROUTE

Radiation-grafted sulfonic acid membranes for PEMFC are commonly prepared by RIG of styrene, styrene/cross-linker mixtures, or substituted styrene monomers onto luorinated polymer ilms, followed by sulfonation reaction to confer the ionic character (Nasef and Hegazy 2004; Nasef 2008). Sulfonation is commonly performed using a strong sulfonating agent such as chlorosulfonic acid diluted with sulfonationresisting solvent (e.g., 1,2-dichloromethane, 1,1,2,2-tetrachloroethane, or carbon tetrachloride) under controlled parameters. However, the challenge in this reaction is to obtain a 100% degree of sulfonation, that is, every benzene ring should contain one pendant sulfonic acid group, without compromising the physical strength and chemical stability of the obtained membranes. Therefore, developing a method to directly introduce sulfonic acid groups to a luorinated ilm is a signiicant contribution that is highly sought. RIG of SSS is an alternative method to produce sulfonic acid membranes in a single-step grafting reaction (Nasef et al. 2009). This method has the advantage of eliminating the hazardous sulfonation reaction and shortening the reaction, so reducing the cost together with setting a basis for continuous production of these membranes. However, RIG of SSS requires overcoming the poor polymerization kinetics of such a monomer caused by the incompatibility between its highly ionized sulfonic

400

Membrane Modification: Technology and Applications (a)

(f )

T (%)

(e) (d)

(c)

(b)

2000

1500

1000

500

(cm–1)

FIGURE 12.10 Typical FTIR of membranes obtained by grafting of SSS onto PVDF at (a) 0% grafting, (b) G = 5%, (c) G = 16%, (d) G = 44%, (e) G = 54%, and (f) G = 65%. (From Nasef, M.M., Saidi, H., and Dahlan, K.Z.M., Radiat. Phys. Chem., 80, 66–75, 2011. With permission.)

acid groups, with their hydration spheres, and the hydrophobic polymer backbone. Such poor kinetics was improved by introducing acrylic acid as a comonomer during or prior to the SSS grafting (Reddy et al. 2005; Zu et al. 2005, 2006, 2007, 2009; Lee et al. 2008a,b). However, the resultant bifunctional (–COOH and –SO3H) membranes are not appropriated for PEMFC (Nasef et al. 2010b). Nasef et al. (2009) reported for the irst time RIG of SSS onto EB-irradiated PVDF ilms without adding acrylic acid monomer to prepare proton exchange membrane. The introduction of PSSA to the PVDF ilms was evident from FTIR spectral analysis shown in Figure 12.10. The grafted membranes displayed features conirming the presence of aromatic ring structures. This was established by the presence of skeletal C–C in plate-stretching vibrations at 1603 and 1495 cm−1, together with an in-plane CH notable band of bending vibration of the substituted benzene ring at 695 cm−1. The band at 1034 cm−1 is due to the in-plane CH bending vibration of the disubstituted benzene ring. The bands at 1128 and 1062 cm−1 are due to the presence of –SO3H groups. The band at 1004 cm−1 represents the in-plane ring vibrations of the aromatic ring caused by the para-substituted sulfonic acid. The intensity of the characteristic bands varies with the degree of grafting. The synergetic role of the addition of various acids with different concentrations and volumes to grafting mixtures, in boosting the grafting level to high values

401

Development of Fuel Cell Polymer Electrolyte Membranes

TABLE 12.5 Summary of the Physicochemical Properties of Graft Copolymer Membranes

G (%) 44 53 65 a

Thickness (μm)

WU (g/g)

IEC (mmol/g)

Proton Conductivity (mS/cm)

56 58 61

57 71 92

2.30 2.63 2.90

63 87 114

Permeability (Barrer)a H2

O2

16.7 21.9 25.8

6.4 9.2 12.4

1 Barrer = 10−10 cm3(STP)/cm/cm2/s/cm/H.

suitable for fuel cell application, has been investigated (Nasef et al. 2010a). The addition of an acid to the monomer solution during grafting was found to be essential for boosting the reaction from negligible levels to values suitable for PEMFC. The grafting reaction was found to proceed by a front mechanism as seen in the scanning optical microscope pictures shown in Figure 12.4. The role of acids in increasing the G% followed the sequence H2SO4 > HCl > CH3COOH > HNO3. The acid concentration and volume were found to heavily affect the G%. The highest G (65%) was achieved with an aqueous solution of sulfuric acid having a concentration of 0.2 mol/l at 10 vol% of the total volume of SSS diluted with DMF. The synergetic acid effect was attributed to increase in the monomer supply to the graft’s growing chains under the inluence of partitioning effect, the inhibition of termination by recombination in the graft propagating chains, and the suppression of the homopolymer by Na-salt formed in the grafting solution. G% was also found to be a function of monomer concentration, absorbed dose, temperature, and ilm thickness under constant values of acid concentration and volume (Nasef et al. 2011). The properties of the obtained membranes including thickness, IEC, water uptake, proton conductivity, and gas permeabilities of H2 and O2 were found to be a function of the ionic moiety content. The permeability of H2 is about twice that of O2 for all membranes, with the value of both permeabilities very suitable for the PEMFC. In conclusion, the membranes were found to attain very good combinations of physicochemical properties suitable for fuel cell application as shown in Table 12.5. The membranes prepared by the single-step grafting method (SSGM) were also found to have superior properties compared with those obtained by the conventional two-step grafting method (CTSGM) as far as fuel cell requirements are concerned (Zu et al. 2005). The evidence for this is shown in the data presented in Table 12.6, which gives a comparison of the properties of two candidate membranes having G = 53% obtained by CTSGM and SSGM methods. Investigation of the thermal properties of the membranes was also evaluated and the obtained data are presented in Table 12.7. The Tm, Xc, and Tc were found to decrease with the increase in PSSA content in the membranes. The decrease in both Tm and Xc clearly indicates a dilution effect by the incorporation of amorphous sulfonated PS. This inding is consistent with the variation of the other properties

402

Membrane Modification: Technology and Applications

TABLE 12.6 Comparison of the Properties of Two Candidate Membranes Having the Same G (53%) Obtained by CTSGM and SSGM Methods

G (%)

WU (g/g)

IEC (mmol/g)

Proton Conductivity (mS/cm)

Tm (°C)

SSGM (53) CTSGM (53)

71 113

2.63 2.30

87 96

160.4 159.3

Xc

Young’s Modulus (MPa)

Elongation (%)

26.8 19.2

511 342

269 163

TABLE 12.7 Variation of the Thermal and Structural Properties of Composite Membranes with Various G% G (%) 0 16 44 53 65

Tm (°C)

Tc (°C)

Xc

167.9 166.4 162.7 160.4 158.8

139.9 138.8 137.7 136.6 135.7

39.2 34.5 28.6 26.8 24.5

such as water uptake, hydration number, and ionic conductivity, which were found to be functions of G%. The reduction in Tc is an indication of mixing of some of the PSSA grafts into the crystalline fraction of the PVDF melt. Nevertheless, the membranes that are most suitable to the PEMFC, with a G of 44–65%, possess reasonable levels of crystallinity, ensuring suficient mechanical stability for the target application. A PEM fuel cell test was carried out at 60°C with three membrane electrode assemblies (MEAs) made of commercial electrodes (E-tek., 0.4 mg/cm2) and PVDFg-PSSA membranes having G of 44, 53, and 65%, respectively. Figure 12.11 shows current–voltage characteristics and power density of MEAs with PVDF-g-PSSA membranes. The current–voltage characteristics of the three MEAs have a similar trend, with a maximum power density of 67–78 mW/cm2 observed with the variation of G% in the range of 44–65%. This indicates that increasing G% improves the polarization characteristics of the membrane, and such behavior may be attributed to the higher proton conductivity and water uptake at high G% values, both of which reduce the membrane resistance to proton transfer. The high open-circuit values in all membranes (982 ± 3 mV) indicate that the tested PVDF-g-PSSA membranes are gas-tight and H2 or O2 permeation is independent of the thickness. A comparison between two membranes having the same degree of grafting (G = 53%) and obtained by SSGM and CTSGM showed that the membrane obtained by SSGM has a better performance in terms of power density, as indicated in Figure 12.12. Moreover, the membrane obtained by SSGM was found to have a better stability than that obtained by the CTSGM, marked by a 30% increase under

403

Development of Fuel Cell Polymer Electrolyte Membranes 1400

G = 65% G = 53% G = 44%

1200

Cell voltage (mV)

1000 60 800 600

40

400 20

Cell temp.: 60°C Electrode area: 5 cm2 Humidifier temp.: TH2/TO2 = 75/65°C Pt loading: 0.4 mg/cm2 Pressure: PH2/PO2 = 0.20/0.2 MPa

200 0 0

50

100

Power density (mW cm–2)

80

0 150

200

250

300

Current density (mA cm–2)

FIGURE 12.11 Current–voltage characteristics and power density curves of a PEM fuel cell with PVDF-g-PSSA membranes having various G%.

SSGM (53%) CTSGM (53%)

1200

80

Cell voltage (mV)

60 800

600

40

400

Power density (mW cm–2)

1000

20 Cell temp.: 60°C Electrode area: 5 cm2 Humidifier temp.:TH2/TO2 = 75/65°C Pt loading: 0.4 mg/cm2 Pressure: PH2/PO2 = 0.20/0.2 MPa

200

0 0

50

100

150

200

250

0 300

Current density (mA cm–2)

FIGURE 12.12 Current–voltage characteristics and power density curves of a PEM fuel cell with PVDF-g-PSSA having the same G (53%) and prepared with two different radiationgrafting methods.

404

Membrane Modification: Technology and Applications 1.0

0.8 Cell voltage (V)

150 mA New method

0.6

0.4 Conventional 0.2

0.0 0

20

40

60

80 100 120 140 160 180 200 Time (h)

FIGURE 12.13 Stability of PVDF-g-PSSA having the same G% and prepared with two different radiation-grafting methods.

dynamic conditions, as shown in Figure 12.13. This conirms the superiority of the former membrane over the latter and suggests that the use of SSGM, which led to the elimination of sulfonation reaction, was crucial for the improvement that took place in the membranes.

12.6

CONCLUSION

Grafting with high-energy radiation is a convenient method to prepare PEMs for fuel cell applications. Particularly, EB accelerator is highly effective in initiating copolymerization reactions required for the preparation of such membranes and has the advantage of high radiation dose, short processing time, and suitability for pilotscale production when commercial applications are sought. The grafting penetration can be varied from the surface to the bulk of the membranes depending on the acceleration energy that can be easily controlled by selection from EB machines currently available with a wide range of acceleration energies. The use of RIG with EB through simultaneous irradiation and preirradiation methods allowed the preparation of two types of PEMs that had similar chemical composition but differed in their morphology: composite PVDF-g-PSSA membrane and graft copolymer counterparts for DMFC and PEMFC. The use of simultaneous EB irradiation was found to simplify the process and reduce the reaction time, together with a reduction in the monomer consumption, whereas the use of the preirradiation method in a singlestep mode provided a shorter route to prepare PEMs with improved properties and reduced cost. This is in addition to setting a basis for designing a continuous line to produce membranes with dedicated EB. Both types of the PEMs were found to have very good combinations of physicochemical properties suitable for fuel cells. The performance of the membranes in DMFC and PEMFC under dynamic conditions was found to be promising, but it requires further improvement to enhance their chemical stability.

Development of Fuel Cell Polymer Electrolyte Membranes

405

ACKNOWLEDGMENT The author wishes to acknowledge the inancial support of the Malaysian Ministry of Science, Technology and Innovation (MOSTI) under the Science Fund program (vote # 79296).

REFERENCES Agro, S., DeCarmine, T., DeFelice, S. and Thoma, L. 2005. Annual progress report for the DOE hydrogen program, US Department of Energy (DOE). http://www.hydrogen.energy.gov (accessed in December 30, 2010). Banerjee, S. and Curti, D.E. 2004. Naion® perluorinated membranes in fuel cells. J. Fluorine Chem. 125: 1211–1216. Bly, J.H. 1988. Electron Beam Processing. International Information Associates: Yardley, PA. Ciuffa, F., Croce, F., D’Epifanio, A., Panero, S. and Scrosati, B. 2004. Lithium and proton conducting gel-type membranes. J. Power Sour. 127: 53–57. Cleland, M.R. and Parks, L.A. 2003. Medium and high-energy electron beam radiation processing equipment for commercial applications. Nucl. Instr. Meth. Phys. Res. B 208: 74–89. Finsterwalder, F. and Hambitzer, G. 2001. Proton conductive thin ilms prepared by plasma polymerization. J. Memb. Sci. 185: 105–124. Gubler, L. and Günther, G. 2010. Scherer, trends for fuel cell membrane development. Desalination 250: 1034–1037. Gubler, L., Gürsel, S.A. and Scherer, G.G. 2005. Radiation grafted membranes for polymer electrolyte fuel cells. Fuel Cells 5: 317–335. Gupta, B. and Scherer, G. 1994. Proton exchange membranes by radiation-induced graft copolymerization of monomers into Telon-FEP ilms. Chimia 48: 127–137. Gupta, B., Anjum, N., Jain, R., Revagade, N. and Singh, H. 2004. Development of membranes by radiation-induced graft polymerization of monomers onto polyethylene ilms. J. Macromol. Sci. C Polym. Rev. C44: 275–309. Gürsel, S.A., Gubler, L., Gupta, B. and Scherer, G.G. 2008. Radiation Grafted Membranes. Adv. Polym. Sci. 215: 157–217. Heitner-Wirguin, C. 1996. Recent advances in perluorinated ionomer membranes: Structure, properties and applications. J. Memb. Sci. 120: 1–33. Ivanov, V.S. 1992. Radiation chemistry of polymer. VSP: Utrecht, the Netherlands. Kalhammer, F.R. 2000. Polymer electrolytes and the electric vehicle. Solid State Ionics 135: 315 –323. Kariduraganavar, M.Y., Nagarale, R.K., Kittur, A.A. and Kulkarni, S.S. 2006. Ion-exchange membranes: Preparative methods for electrodialysis and fuel cell applications. Desalination 197: 225–246. Lee, S.W., Bondar, Y. and Han, D.H. 2008a. Synthesis of a cation-exchange fabric with sulfonate groups by radiation-induced graft copolymerization from binary monomer mixtures. React. Funct. Polym. 68: 474–482. Lee, S.W., Bondar, Y. and Han, D.H. 2008b. Synthesis of polypropylene fabric with sulfonate groups. Radiat. Phys. Chem. 77: 503–510. Maiyalagan, T. and Pasupathi, S. 2010. Components for PEM fuel cells: An overview. Sci. Forum 657: 143–189. Mondelaers, W. 1998. Low-energy electron accelerators in industry and applied research. Nucl. Instr. Meth. Phys. Res. B 139: 43–50. Nagarale, R.K., Shina, W. and Singh, P.K. 2010. Progress in ionic organic-inorganic composite membranes for fuel cell applications. Polym. Chem. 1: 388–408.

406

Membrane Modification: Technology and Applications

Nasef, M.M. 2003. Application of electron beam for membrane preparation. Proceedings of the FNCA 2003 Workshop on the Application of Electron Accelerator–Radiation System for Thin Film. 18–22 August 2003, Kuala Lumpur, Malaysia, pp. 37–54. Nasef, M.M. 2008. Fuel cell membranes by radiation-induced graft copolymerization: Current status, challenges, and future directions. In: Polymer Membranes for Fuel Cells, eds. S.M.J. Zaidi and T. Matsuura, pp. 87–114. Springer Science: New York. Nasef, M.M. and Hegazy, E.A. 2004. Preparation and applications of ion exchange membranes by radiation-induced graft copolymerization of polar monomers onto non-polar ilms. Prog. Polym. Sci. 29: 499–561. Nasef, M.M., Zubir, N.A., Ismail, A.F., et al. 2006a. PSSA pore-illed PVDF membranes by simultaneous electron beam irradiation: Preparation and transport characteristics of protons and methanol. J. Memb. Sci. 268: 96–108. Nasef, M.M., Zubir, N.A., Ismail, A.F., Dahlan, K.Z.M., Saidi, H. and Khayet, M. 2006b. Preparation of radiochemically pore-illed polymer electrolyte membranes for direct methanol fuel cell. J. Power Sour. 156: 200–210. Nasef, M.M., Zubir, N.A., Ismail, A.F. and Khayet, M. 2006c. Sulfonated radiation grafted polystyrene pore-illed poly(vinylidene luoride) membranes for direct methanol fuel cell: Structure–property correlations. Desalination 200: 642–644. Nasef, M.M., Saidi, H. and Dahlan, K.Z.M. 2009. Single-step radiation induced grafting for preparation of proton exchange membranes for fuel cell. J. Memb. Sci. 339: 115–119. Nasef, M.M., Saidi, H. and Dahlan, K.Z.M. 2010a. Acid synergized grafting of sodium styrene sulfonate onto electron beam irradiated poly(vinylidene luoride) ilms for preparation of fuel cell membrane. J. Appl. Polym. Sci. 118: 2801–2809. Nasef, M.M., Saidi, H. and Dahlan, K.Z.M. 2010b. Radiation grafted poly(vinylidene luoride)-graft-polystyrene sulfonic acid membranes for fuel cells: Structure-property relationships. Chin. J. Polym. Sci. 28: 761–770. Nasef, M.M., Saidi, H. and Dahlan, K.Z.M. 2011. Kinetic investigation of Graft copolymerization of sodium styrene sulfonate onto poly(vinylidene luoride) ilms. Radiat. Phys. Chem. 80: 66–75. Neburchilov, V., Martin, J., Wang, H. and Zhang, J. 2007. A review of polymer electrolyte membranes for direct methanol fuel cells. J. Power Sour. 169: 221–238. Prakash, G.K.S., Smart, M.C., Wang, Q-J., et al. 2004. High eficiency direct methanol fuel cell based on poly(styrenesulfonic) acid (PSSA)– poly(vinylidene luoride) (PVDF) composite membranes. J. Fluorine Chem. 125: 1217–1230. Qiu, X., Zhang, W., Li, S., Liang, H. and Zhu, W. 2003. The microstructure and character of the PVDF-g-PSSA membrane prepared by solution grafting. J. Electrochem. Soc. 150: A917–A921. Reddy, P.R.S., Agathian, G. and Kumar, A. 2005. Preparation of strong acid cation-exchange membrane using radiation-induced graft polymerization. Radiat. Phys. Chem. 73: 169–174. Scrosati, B. 1993. Applications of Electroactive Polymers. Chapman Hall: London. Shin, J.-P., Chang, B.-J., Kim, J.-H., Lee, S.-B. and Suh, D.H. 2005. Sulfonated polystyrene/ PTFE composite membranes. J. Memb. Sci. 251: 247–254. Stephan, M.A. 2006. Review on gel polymer electrolytes for lithium batteries. Eur. Polym. J. 42: 21–42. Vie, P., Paronen, M., Strømgård, M., Rauhala, E. and Sundholm, F. 2002. Fuel cell performance of proton irradiated and subsequently sulfonated poly(vinyl luoride) membranes. J. Memb. Sci. 204: 295–301. Yamaguchi, T., Miyata, F. and Nakao, S. 2003. Pore-illing type polymer electrolyte membranes for a direct methanol fuel cell. J. Memb. Sci. 214: 283–292. Zu, J., Wu, M., Fu, H. and Yao, S. 2005. Cation-exchange membranes by radiation-induced graft copolymerization of monomers onto HDPE. Radiat. Phys. Chem. 72: 759–764.

Development of Fuel Cell Polymer Electrolyte Membranes

407

Zu, J., Wu, M., Zhang, J., Yu, C., Liu, X., Tong, L. 2006. Radiation-induced grafting of acrylic acid and sodium styrene sulfonate onto high-density polyethylene membranes. I. Effect of grafting conditions. J. Appl. Polym. Sci. 99: 3401–3405. Zu, J., Hu, Z., Wong, W., Zhang, J., Pino, E., Gu, J. and Tong, L. 2007. The effect of additives on radiation induced grafting of AA and SSS onto HDPE. J. Radioanalyt. Nucl. Chem. 273: 479–484. Zu, J., Zhang, J., Sun, G., Zhou, R. and Liu, Z. 2009. Preparation of cation-exchange membrane containing bi-functional groups by radiation induced grafting of acrylic acid and sodium styrene sulfonate onto HDPE: Inluence of the synthesis conditions. J. Radioanalyt. Nucl. Chem. 279: 185–192.

13

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application Ahmad Fauzi Ismail, Muhammad Noorul Anam Mohd Norddin, Juhana Jaafar, and Takeshi Matsuura

CONTENTS 13.1 13.2 13.3 13.4

Introduction ..................................................................................................409 Type of Fuel Cell .......................................................................................... 410 Direct Methanol Fuel Cell ............................................................................ 411 State-of-the-Art Commercial Polymer Electrolyte Membrane (Naion) ...... 412 13.4.1 Modiication of Naion by Incorporating Inorganic Material .......... 414 13.4.2 Modiication of Naion by Surface Modiication.............................. 415 13.5 Development of Alternative PEM Materials ................................................ 418 13.5.1 Poly(Ether Ether Ketone) .................................................................. 420 13.5.2 Sulfonation of Poly(Ether Ether Ketone) .......................................... 420 13.5.3 Development of Modiied SPEEK for DMFC Application .............. 424 13.5.3.1 Modiication of SPEEK by Incorporating Inorganic Material .............................................................................. 424 13.5.3.2 Modiication of SPEEK by Surface Modiication ............. 438 13.6 Conclusion .................................................................................................... 442 References .............................................................................................................. 442

13.1

INTRODUCTION

One of the most important challenges that our world will face in the twenty-irst century will be continuing to meet the ever-increasing energy needs of its citizens. Global concerns over energy sustainability and environmental impact of fossil fuels have motivated efforts to improve our energy sources. Therefore, a renewable longterm energy source and a more environment-friendly source are highly preferable. One of the promising candidates as a power source solution for the future world energy problem is fuel cells. Fuel cells are ideal candidates for distributed power generation. They can provide a low emission and highly eficient source for cogenerating heat and electricity. 409

410

Membrane Modification: Technology and Applications

Fuel cells used for portable applications are one of the most recent and promising areas that have attracted global interest and have become a promising alternative for niche applications. In addition, fuel cells can offer a higher power density and longer life span compared with batteries for portable applications (Apanel and Johnson 2004). Reliability is another important factor where the beneits of fuel cells outweigh those of batteries.

13.2

TYPE OF FUEL CELL

A fuel cell is an electrochemical device that converts chemical energy directly into electrical energy from an electrochemical reaction. It produces electricity, water, and heat from the reaction of fuel and oxygen without any burning, thus greatly reducing the pollutants and ineficiencies brought about by combustion. Unlike dry-cell batteries, fuel cells do not require recharging and operate as long as the fuel is available (Li et al. 2003a). The fuel is typically hydrogen or methanol. The basic advantages of fuel cells are the potential for a high operating eficiency (up to 50%–70%) and near-zero greenhouse gas emissions. Second, fuel cell systems provide quiet and vibration-free operation. Third, a fuel cell system is a highly scalable design. Finally, fuel cells have multiple choices of potential fuel feedstock from renewable ethanol to biomass hydrogen production and a nearly instantaneous recharge capacity compared with batteries (Li et al. 2003b). Fuel cells consist of two electrodes, anode and cathode, which are separated by an electrolyte. The anode provides an interface between the fuel and the electrolyte, while the cathode provides an interface between the oxygen and the electrolyte. The electrodes contain catalysts for the oxidation of fuel and reduction of oxygen. At the anode, the oxidation reaction produces free proton (H+) and electron e−. The proton travels to the cathode via the electrolyte, while the electron is harnessed as direct current via an external circuit. In addition to completing the electrical circuit by transporting ions between the electrodes, the electrolyte also acts as the separator between the fuel and the oxidant. There are several types of fuel cells, which are classiied primarily by the kind of electrolyte they employ. The materials used for electrolytes have their best conductance only within certain temperature ranges (Hirschenhofer 1994). A few of the most promising types include phosphoric acid fuel cell (PAFC), molten carbonate fuel cell (MCFC), solid oxide fuel cell (SOFC), alkaline fuel cell (AFC), proton exchange membrane fuel cell (PEMFC), and direct methanol fuel cell (DMFC). Among the promising types for small appliances and portable application are PEMFC and DMFC. PEMFC is one of the most promising clean energy technologies under development. It has become increasingly important as an alternative energy source for stationary, automobile, and portable power. The major advantages include the current prototype eficiency of up to 64%, high energy densities (relative to batteries), and the ability to operate on clean fuels while emitting no pollutants. This fuel cell type operates at relatively low temperatures (30°C–150°C) but generates more power for a given volume or weight of a hydrogen–air fuel than any other type of fuel cell (Einsla 2005). In addition, PEMFCs have drawn a lot of attention

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

411

because of their high eficiency, quiet operation, use of fuel from totally renewable resources, and environment-friendliness. Despite these beneits, diffusion of PEMFC technology into the marketplace is being limited by many challenges. These challenges include choice of the appropriate fuel (mostly hydrogen) source and infrastructure, industry regulation, safety and public acceptance, and hydrogen handling problems. Therefore, research into fuel cells has grown exponentially over the last 15 years. In the case of polymer fuel cells, currently, DMFC, which does not use hydrogen as fuel, is gaining more attention. Therefore, this chapter covers only the DMFC type.

13.3

DIRECT METHANOL FUEL CELL

DMFC is a type of PEMFC that uses direct aqueous methanol solution as fuel, which eliminates all the handling problems of hydrogen fuel. DMFC is regarded as a promising candidate for portable power applications due to the ease of transporting, storing, and distributing the methanol fuel. Even though direct use of methanol in DFMC is a topic of considerable interest, there are still a lot of obstacles that DMFC has to overcome in order to achieve commercialization. The most inluential factor is its high cost and low power density. Its high cost is due mainly to the catalyst and the membrane electrode assembly (MEA) parts, which, however, will not be discussed in great detail in this chapter. The relatively low power density or lower cell performance compared with PEMFC is caused mainly by the poor kinetics of the anode electro-oxidation of methanol, by low membrane proton conductivity, and by the crossover of methanol through the polymer electrolyte membrane (PEM). The slow oxidation kinetics of methanol to carbon dioxide is due to the formation of carbon monoxide as an intermediate, which strongly adsorbs on the catalyst surface. The key to resolving proton conductivity and methanol crossover lies mainly in the proton electrolyte membrane (PEM), which must be able to ilter methanol but pass the proton through to the cathode. In a DMFC, as shown in Figure 13.1, the liquid MeOH fuel is oxidized in the presence of water at the Pt–Ru anode electrocatalyst. This results in the generation of CO2, hydrogen ions, and electrons. The electrons travel through the external circuit, as the electric output of the fuel cell, while the hydrogen ions are conducted through the electrolyte (PEM) and react with the electrons from the external circuit to form water at the cathode, to complete the circuit. Thus, the DMFC reactions (Equations 13.1 through 13.3) are as follows (Othman 2006): At anode, CH 3OH + H 2O → CO2 + 6H + + 6e − , Eo = 0.02V

(13.1)

3 O2 + 6H + + 6e − → 3H 2O ( cathode ) , Eo = 1.23V 2

(13.2)

At cathode,

412

Membrane Modification: Technology and Applications

6e−

Load

6e−

Porous carbon anode

Porous carbon cathode 6H+

Methanol (CH3OH)

Air (O2)

Ion-exchange membrane

Pt–Ru catalyst 6e−

6e− Product: H2O, CO2 + waste heat

6H+

CH3OH

Anode reaction: CH3OH + H2O

CO2 + 6H+ + 6e−

Cathode reaction: 3/2 O2 + 6H+ + 6e−

6H+

3/2O2 3H2O

3H2O

FIGURE 13.1 Basic reaction and structure of direct methanol fuel cell (DMFC).

Overall, CH 3OH + H 2O +

3 O2 → CO2 + 3H 2O ( overall ) , Eo = 1.21V 2

(13.3)

At present, the most widely used commercial PEM is Naion produced by DuPont since 1992. Naion is a plain perluorosulfonic membrane that is thermally stable and is excellent for PEMFC because of its high proton conductivity. However, Naion is not suitable for DMFC applications, partly due to its cost. This type of membrane has high permeability toward methanol even at low temperatures, which drastically reduces the DMFC performance (Neburchilov et al. 2007). This is worsened by high water permeability in perluorinated membranes that can cause cathode looding and thus lower cathode performance, which also contributes to lower DMFC performance.

13.4

STATE-OF-THE-ART COMMERCIAL POLYMER ELECTROLYTE MEMBRANE (NAFION)

In PEMFC, the current state-of-the-art fuel cell technology primarily involves the use of perluorosulfonic acid (PFSA) membrane as electrolyte. PFSA membranes are composed of carbon luorine backbone chains with perluoro side chains containing

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

413

sulfonic groups (Doyle and Rajendran 2003). Commercially available ionomers of this sort are the Naion membrane (Du Pont), the Dow membrane (Dow Chemical Company, United States), and Flemion (Asahi Glass Co., Japan) (Costamagna and Srinivasan 2001). The most well-known and established polymer is Naion. Naion is the membrane of choice for H2–air fuel cells and can be considered an industry standard. Naion is a perluorinated hydrocarbon that can be described as an ionomer of nonpolar Telon (poly(tetraluoroethylene)) backbone with polar sulfonic acid groups at the end of the side chains. To synthesize Naion, tetraluoroethylene is reacted with SO3 to form a cyclic sultone. Figure 13.2 shows the chemical structure of Naion. It can be seen that Naion consists of a polytetraluoroethylene (PTFE) backbone (nonpolar), which gives it its high chemical resistance. The side chains consist of perluorinated vinyl polyethers attached to the PTFE backbone through an ether oxygen. The side chains are terminated in sulfonic acid groups, –SO3H (polar), which give Naion its proton exchange capability. The presence of the nonpolar and polar moieties has led to phase segregation, which is desirable for transport properties. The higher phase segregation in Naion has created broader water channels with fewer dead ends and thus more interconnected channels. With the presence of water, this microstructure transports water and proton better, which confers on Naion a high proton conductivity when fully hydrated. However, this microstructure is a drawback for the methanol crossover, which is also very high, to a greater amount than that of sulfonation of poly(ether ether ketone) (SPEEK). Naion has been reported to have relatively high proton conductivity (0.1 S/cm at 25°C, fully hydrated), high thermal stability (280°C), and excellent oxidative stability by the accelerated test method (less than 1 × 10−3/min degradation rate constant at 68°C in Fenton’s reagent (Bae and Kim 2003). Although it is proven that Naion has high proton conductivity and outstanding chemical stability, Naion membranes do not perform well in DMFCs, because of the high methanol permeability of Naion. It has been reported that over 40% of the methanol feed could be wasted during DMFC operation with Naion due to methanol crossover through the membrane (Pivovar et al. 1999). Due to the high methanol crossover, various efforts have been made to either modify Naion or develop new alternative polymers for DMFC applications. CF2 CF2

CF2 x

CF

y O

O

CF2

CF2

O CF

z

CF2

S OH O

CF3

FIGURE 13.2 Chemical structure of Naion. (Reprinted from Curtin, D.E., Lousenberge, R.D., Henry, T.J., Tengeman, P.C., and Tisack, M.E., J. Power Sour., 131, 41–48, 2004. With permission.)

414

Membrane Modification: Technology and Applications

The next section will describe some modiications studied for Naion, followed by discussions on the development of new alternative polymers.

13.4.1

MODIFICATION OF NAFION BY INCORPORATING INORGANIC MATERIAL

Several efforts have been made to improve the performance of Naion, using various modiication methods. One of the methods is to incorporate inorganic materials in the Naion membrane. The incorporation of inorganic materials is believed to improve the proton conductivity and reduce the methanol permeability, because the barrier properties of the membrane are expected to increase as the concentration of rigid backscattering and the tortuous pathways that molecules encounter during the permeation increase due to the presence of inorganic particles (Silva et al. 2005). Modiication of Naion through the addition of silica is a common approach utilized for the improvement of membrane performance in DMFC applications. This is because silica is a hygroscopic oxide and its presence inside the composite membrane enhances water retention at high temperatures, thus extending the operation range of Naion (low humidity and high temperature operation). Studies by Arico et al. showed that by the addition of silica to Naion, the water retention in the membrane was improved, thus enabling the operation of fuel cell up to 130°C (Arico et al. 2003a). Ren et al. also reported that the water uptake by the membrane containing oxide is higher than that by the pristine Naion (Ren et al. 2005). Naion–silica membranes have been prepared employing several methods by casting mixtures such as silica powder, diphenylsilicate (DPS), sol–gel reaction with tetraethylorthosilicate (TEOS), followed by solution casting of the Naion solution, phosphotungstic acid (PTA)-doped composite silica/Naion/PTA, and silica oxide (Wang et al. 2007). Composite Naion membranes have been made by the incorporation of inorganic nanoparticles as methanol barrier components, which include silicon oxide, titanium oxide, or mixed silicon–titanium oxides (Arico et al. 2003; Dimitrova et al. 2002; Jung et al. 2003). One drawback of this approach is a decrease in conductivity due to the addition of the nonconductive oxide (Aparicio et al. 2005). A signiicant improvement in the conductivity of Naion at elevated temperatures was achieved by Bahar et al., by incorporating perluorinated ionomers in a Naion matrix and by doping it with heteropoly acids such as PTA and phosphomolybdenic acid (Bahar et al. 1996). Tazi and Savodogo also incorporated silicatungstic acid (SiWA) and thiophene (TH) in Naion 117 membrane and demonstrated an improvement in the ionic conductivity and achieved a high power density (Tazi and Savodogo 2001). Water uptake in the Naion modiied with SiWA was 60% and with both SA and TH, it was 40% compared with the normal value of 27% for Naion 117. Malhotra and Datta impregnated the Naion membranes with a PTA solution in acetic acid or in molten salt solvent and the resultant membrane could operate up to 110°C–120°C (Malhotra and Datta 1997). In addition, zirconium phosphate (ZrP) was also reported as a promising iller in Naion membrane due to the additional H+ ions of the phosphate moiety and bound crystal water (Bauer and Porada 2004). Bauer and Porada reported that in the

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

415

TABLE 13.1 The Effects of Nafion Modification on Proton Conductivity Combination Naion/α -ZrP Naion/silica Naion/HPA Naion/mordenite Naion/imidazole

Effects on Proton Conductivity Conductivity similar to Naion, improved MEA, and fuel crossover Conductivity similar to Naion, improved fuel crossover Good improvement in conductivity over Naion counteracted by leaching Very small conductivity improvement at high temperatures only Very good conductivity results; however, imidazole poisoned Pt catalyst

References Yang et al. (2001) and Nunes et al. (2002) Mauritz (1998) Zaidi et al. (2000) Kwak et al. (2003) Kreuer (2001) and Glipa et al. 1997

ZrP-modiied Naion membrane, the water content was slightly enhanced as compared with untreated membrane and, consequently, the methanol permeability was reduced. In addition to these studies on Naion, it is worth discussing other composite Naion membranes that have been studied; these are summarized in Table 13.1. The above literature review shows that many inorganic compounds have been incorporated into Naion in order to increase its hydrophilicity and water uptake for the purpose of higher proton conductivity. Even though these objectives can be achieved, normally the methanol permeability will also increase with the water uptake. Therefore, several approaches have been attempted to suppress methanol permeability in Naion for DMFC application and one of the methods is by the surface modiication technique.

13.4.2

MODIFICATION OF NAFION BY SURFACE MODIFICATION

Research on Naion modiication has recently been focused on minimizing the methanol crossover, which is particularly serious in the DMFC system. Some of the modiications made to the Naion membrane are by using surface modiication techniques, which include plasma deposition, palladium sputtering, and polymer coating and grafting. Yoon et al. used Palladium (Pd) to modify Naion membranes by coating them with a different thickness of Pd ilm, using as puttering method (Yoon et al. 2002). The scanning electron microscopy (SEM) micrographs showed that the 10 and 30 nm Pd ilms were dense and appeared to be well attached to the membrane. However, some cracks were found on the surface of the 100 nm Pd ilm. It was believed that the cracks were caused by the difference between the Naion membrane and the Pd ilm in the expansion that took place when the composite membrane was immersed in water. The Naion membrane swells more than the Pd ilm, which eventually develops cracks in the Pd ilm. Therefore, the sputtering procedures for a membrane thicker than 100 nm apparently need to be improved. From this research, a trade-off between proton conductivity and methanol crossover was noted similar to the results

416

Membrane Modification: Technology and Applications

from the fabrication of Naion/silicon oxide composite membrane. The presence of Pd ilm not only diminishes methanol crossover, but also reduces the membrane proton conductivity. Ma et al. reported that the addition of silver (Ag) to Pd improved the proton conductivity of Pd ilm as the Pd–Ag ilm had higher proton conductivity than the pure Pd ilm (Ma et al. 2003). Using the same sputtering method, a thin Pt/Pd–Ag/Pt alloy ilm was deposited on the surface of the Naion membrane. The platinum (Pt) on the membrane surface acts as a bridge to transfer the protons from the Pd–Ag alloy ilm to the Naion membrane due to the different mechanisms of proton transport between the Pd–Ag alloy ilm and the Naion membrane. In the Naion membrane, protons are associated with water and diffuse due to a concentration difference, while only protons can diffuse through the Pd–Ag alloy ilm. Therefore, at the interface of the Naion membrane and the Pd–Ag alloy, a change in the proton transfer mechanism may occur from water-associated proton to pure proton, and the presence of platinum may catalyze this transfer between the two forms of proton. Although it is also not a crack-free ilm, it was proven that this modiied membrane was effective in reducing methanol crossover and gave a higher cell performance than that of an unmodiied Naion membrane, when it was tested in a single-cell DMFC. Choi et al. combined plasma modiication of the membrane surface and a Pd-sputtering technique in an attempt to reduce the methanol crossover problem (Choi et al. 2001). Three types of modiied membranes—plasma-etched membrane, Pd-sputtered membrane, and a combination of plasma-etched and Pd-sputtering membrane—were prepared and their permeabilities, along with the unmodiied Naion membrane, were measured in a diffusion cell. The cell contained two compartments: one compartment was illed with a solution of methanol and 1-butanol in deionized water, and the other compartment was illed with 1-butanol in deionized water only. All membrane modiications were effective in reducing the methanol permeability. The methanol permeability for the Pd-sputtered membrane was lower than that of the plasma-etched membrane, while the membrane modiied by both plasma etching and Pd sputtering exhibited the lowest methanol permeability. As for proton conductivity, the best performance was the Pd-sputtered membrane followed by the plasma-etched membrane, followed by the unmodiied Naion membrane, and the lowest was the combination of plasma-etched and Pd-sputtered membrane. However, the indings of this work disagrees with the work by Yoon et al., who observed that proton conductivity for the modiied membrane decreased as compared with the unmodiied membrane, despite using the same modiication technique (Yoon et al. 2002). Smit et al. have synthesized a thin layer of polypyrrole inside the nanopores of an existing Naion membrane (Smit et al. 2003). However, the polymer was found not to deposit in a very uniform way, and two distinct groups of polypyrrole grains appeared. The result, however, indicates that the unmodiied membrane has the highest methanol permeability while the membrane with polypyrrole has the lowest permeability. Another surface modiication by the layer method is a three-layer membrane that has been examined for DMFCs, where a blended Naion–poly(vinylidene luoride)

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

417

(PVDF) methanol barrier layer was placed between two outer Naion layers (Kim et al. 2004a,b). The trilayer structure was prepared by spraying 5 wt% Naion solution directly onto both sides of a Naion–PVDF ilm. Both proton conductivity and methanol permeability were controlled by the composition and thickness of the Naion–PVDF layer. The DMFC performance of a trilayer system, composed of a 10 μm thick ilm of Naion–PVDF (55 wt% of PVDF) with two 20 μm thick outer layers of Naion, was better than that of Naion 112 (54 μm thick), while the methanol crossover was reduced by 64%, as compared with that of Naion 112, but there was no reduction in crossover, as compared with a thicker Naion 117 membrane (Kim et al. 2004b). Shao et al. have coated polyvinyl alcohol (PVA) ilm on the surface of the Naion membrane to produce a Naion/PVA composite membrane by immersing the Naion membrane in a PVA casting solution (Shao et al. 2002). Subsequent cross-linking and sulfonation treatment were also applied on the composite membrane to enhance the mechanical strength and proton conductivity of the casting ilm. The best result was with the modiied membrane of 50:50 ratio, which exhibited the lowest methanol permeability. It is worth mentioning that several other luorinated polymer membranes have also been developed, including aromatic luorinated polymers such as triluorostyrene (TFS), acid aliphatic luorinated polymers such as PTFE, poly(ethylenealt-tetraluoroethylene) (ETFE), and PVDF. ETFE was also improved using surface modiication, in particular the grafting method. Shen et al. grafted the ETFE membrane with polystyrene sulfonic acid (PSSA) as a proton-conducting group (Shen et al. 2005). By adjusting the degree of grafting and membrane thickness, the modiied ETFE membrane could provide an attractive alternative to the expensive Naion membrane for DMFC application. Another study of luorinated polymer was conducted by Nasef et al. (2005). They illed the pores of the porous PVDF ilms with PSSA using radiation-induced grafting. They reported that the proton conductivity in the membrane increases with the increase in the grafting percentage and exceeds that of the Naion 117 membrane at a grafting percentage of 46%. The methanol permeabilities of 40% and 46% poreilled membranes are lower than that of Naion 117 by 53% and 71%, respectively. This result suggests that the PVDF membrane is also a potential material as electrolyte membrane of DMFC, in addition to TFS, PTFE, and ETFE. Although the surface modiication by adsorption of the modiier onto the membrane surface, chemical or physicochemical posttreatment of the surface via hydrolysis or gas plasma treatment and grafting or cross-linking a modiier on the surface generally result in some improvements suppressing methanol permeability, all of these surface modiications involve multiple steps before the membrane surface can be inally modiied. In addition, the plasma technique and surface luorination cannot be easily controlled, and scale-up of an experimental setup to a large production reactor is not a simple process (Chan 1994). Plasma treatment also requires a vacuum system, which increases the operation cost. Cross-linking is only useful for producing solvent-resistant surfaces and its use in modifying the surface energies is not common. Chemical coating and grafting require speciic conditions, which somewhat limit their use (Pham 1995).

418

13.5

Membrane Modification: Technology and Applications

DEVELOPMENT OF ALTERNATIVE PEM MATERIALS

In order to overcome the problem faced by luorinated membrane (Naion), many researchers proposed the use of less expensive materials. Even though the material lifetime and mechanical properties are lower, they should be acceptable as lower cost is a very important factor in the commercial sector. Therefore, the idea of using hydrocarbon polymers, although previously abandoned due to low thermal and chemical stabilities, has attracted the interest of researchers around the world. Various researches have been aimed to develop alternative membranes that are of lower cost as well as higher proton conductivity and lower methanol permeability. Many promising polymers are based on aromatic polymers such as PEEK and other poly(aryl ether ketone), PES, PBI, and PAEK. The aromatic polymers possess excellent chemical resistance, high thermo-oxidative stability, and good mechanical properties and are of low cost. The following paragraphs discuss the few types of aromatic polymers that have been used as PEM. Recently, sulfonated polyimide membranes were examined as DMFC membranes. Woo et al. synthesized sulfonated polyimide by incorporation of sulfonated diamine as a comonomer (Woo et al. 2003). Polyimide was chosen because it is thermally stable and mechanically strong. However, it was found that there was an upper limit in the proton conductivity (0.041 S/cm), which is about half that of Naion 117. For such membranes, the sulfonation level was high (63 mol%), but the water uptake was low (16 wt%). Additionally, there was concern as to the hydrolytic stability of sulfonated polyimides. PSSA can be synthesized using a sulfonation process. Caretta et al. reported that the membrane cast from sulfonated polystyrene exhibited proton conductivity equal to that of the Naion membrane and the methanol permeability of PSSA is comparatively lower, which is about two times lower than that of the Naion membrane (Caretta et al. 2001). Yang and Manthiram studied SPEEK membranes for use in a DMFC (Yang and Manthiram 2003). It was reported that a membrane with 50% degree of sulfonation (DS) exhibited comparable DMFC performance to Naion 115 at 65°C, while the methanol crossover was two times lower than the Naion membrane. Sulfonated poly(ether ketone) (SPEK), sulfonated poly(ether ketone ketone) (SPEKK), and sulfonated poly(ether ether ketone ketone) (SPEEKK) come from the same family as SPEEK and therefore their characteristics are quite similar. The sulfonation of poly(ether ether ketone) is similar to the sulfonation technique for PSSA using sulfonation agents such as high concentration of sulfuric acid. The DS of SPEEK can be controlled by varying the reaction time and temperature (Xing et al. 2004). The work of Lee et al. showed that the proton conductivity of the SPEEK membrane increased with DS (Lee et al. 2007). The increase in DS will increase the hydrophilicity and water uptake and, consequently, gives an opportunity for the formation of water-mediated pathways for protons. Methanol permeability measurement indicates that its rate increases with the increase of DS but at a much lower rate than the Naion membrane. This is due to the difference in their microstructures (Li et al. 2003a). Libby et al. showed that PVA can easily be formed into membrane and its permeability can be altered with heat (Libby et al. 2003). PVA is also a good methanol

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

419

barrier and the methanol permeability of its membrane decreases with increasing methanol concentration (Pivovar et al. 1999). It does the reverse for Naion. Thus, at higher methanol concentrations, PVA becomes a better methanol barrier, while Naion becomes an even poorer barrier. PVA also has comparatively high conductivity at approximately 10−2 cm/s. However, information in the literature on DMFC performance of a single PVA membrane is limited. Kim et al. observed that sulfonated poly(arylene ether sulfone) membranes had a higher selectivity (i.e., proton conductivity/methanol permeability) than that of Naion because of their lower methanol permeability (up to 10-fold lower than Naion) and relatively high proton conductivity (0.07–0.1 S/cm), for polymers with 40 mol% sulfonation (Kim et al. 2004a). These polymers became brittle in the dry state and were dificult to process and handle. Two types of polyphosphazene (PPh) group such as sulfonated and phosphonated PPh have been developed recently for DMFC application (Gao and Sammes 1999; Allock et al. 2002; Zhou et al. 2003). Zhou et al. reported that the methanol crossover of the sulfonated membrane was about eight times lower than that of Naion 117 at room temperature even though the values were comparable at 120°C (Zhou et al. 2003). The methanol permeability of the phosphonated PPh derivative was about 40 times lower than that of Naion 117 at room temperature and about 9 times lower at 120°C. This is a signiicant improvement over the behavior of Naion 117. Although the conductivities of both PPh membranes were lower than that of the Naion 117 membrane, they are still practical for DMFC applications since the selectivity of both sulfonated PPh (at temperature 60%). Cloisite is a very promising additive in nanocomposite membranes, since it is already being organically modiied from the natural MMT, which is expected to enhance the compatibility with organic polymers (Cervantes-Uc et al. 2007).

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

425

A compatibilizer, 2,4,6-TAP, was introduced into the SPEEK/Cloisite 15A nanocomposite in order to enhance the compatibility between the materials. It was expected that TAP would form hydrogen bonding with both the polymer (SPEEK) and the inorganic iller (Cloisite 15A) to improve the compatibility between them. Figure 13.6 illustrates the expected bonding formed by TAP between the hexadecyl tallow in Cloisite 15A and SPEEK. We also anticipate that the porous layered Cloisite 15A exfoliated in SPEEK matrix can act as a selectively permeable barrier that blocks methanol permeation but does not reduce or may potentially even increase the proton conductivity of the conventional SPEEK. These behaviors occur because of the combination of the sieving effect of the Cloisite pores, which allow permeation of protons while preventing that of methanol, as well as a high aspect ratio that creates highly tortuous pathways for methanol molecules (Hudiono et al. 2009). SPEEK/Cloisite 15A nanocomposite membranes were fabricated by incorporating TAP as compatibilizer and characterized for DMFC application. In order to fabricate homogeneous SPEEK/Cloisite 15A/TAP nanocomposite membranes, the solution intercalation method was employed by preparing the SPEEK and Cloisite 15A/TAP solutions separately before mixing them, where 10 wt% of SPEEK solution was irst prepared by dissolving SPEEK in DMSO. Amounts of 0.1 g of Cloisite 15A and 0.1 g of TAP were dissolved in DMSO at 60°C by vigorously stirring for 2 h and then added to the SPEEK solution. The mixture was vigorously stirred for 24 h at 60°C to produce a homogeneous solution. There were three solutions prepared: SPEEK/Cloisite 15A/TAP (SP/Cl/TAP), SPEEK/Cloisite 15A (SP/Cl), and SPEEK/ TAP (SP/TAP).

O

H

O

H3C(CH2)16 C O CH

O HC O C (CH2)16CH3

O

H3C(CH2)16 C O CH

H

H3C(CH2)16 C O CH H3C(CH2)16 C O CH

H H H N

O

HC O C (CH2)16CH3

O

H Two hydrogenated tallow in Cloisite 15A

N TAP

− Hδ N δ+ H

N

δ−

= Possible interaction

N H + Hδ O

O

δ−

δ− O

C

SO3H SPEEK

FIGURE 13.6 Interaction model between Cloisite 15A clay, 2,4,6-triaminopyrimidine, and SPEEK. (Reprinted from Jaafar, J., Ismail, A.F., and Matsuura, T., J. Memb. Sci., 345, 119– 127, 2009. With permission.)

426

Membrane Modification: Technology and Applications

TABLE 13.3 Physical Properties of SPEEK-Nanocomposite Membranes Water Uptake (wt%) SPEEK SP/Cl SP/Cl/TAP SP/TAP Naion

10 34 20 29 21

Methanol Uptake (wt%) 14 43 30 31 41

Source: Data from Jaafar, J., Ismail, A.F., and Matsuura, T., J. Memb. Sci., 345, 119–127, 2009. With permission.

The uptake of water and that of aqueous methanol solution of SPEEK and its nanocomposite membranes were studied at room temperature. According to Table 13.3, it is observed that among the nanocomposite membranes, which exclude the parent SPEEK membrane, SP/Cl/TAP shows the lowest water and methanol uptake values. This might be due to the partially exfoliated polymer–clay nanocomposite formation. It was suggested that when SPEEK is intercalated or exfoliated in the Cloisite 15A layers, the polymer chain movement is hindered, resulting in tighter chain packing and thus a decrease in the free void in the membrane (Gosalawit et al. 2008). One of the most important characteristics that relect the performance of a PEM in DMFC is proton conductivity. It has been reported that the proton conductivity depends on the DS, pretreatment of the membrane, hydration state, and ambient relative humidity and temperature. For ionomeric membranes, the proton conductivity depends on the amount of acid groups attached to the polymer ring and their dissociation capability in water, which is accompanied by the generation of protons. The high ionic conductivity demonstrated by the membrane at high sulfonation level suggests that the water-swollen ionic domains in the membrane pores were interconnected to form a network structure. Water molecules also dissociate acid functionality and facilitate proton transport. Therefore, it can be deduced that water uptake is an important parameter in proton conductivity tests (Bauer et al. 2000). In general, it is known that the proton conductivity does not directly correlate with either water uptake or IEC for any of the polymers applicable in fuel cell technology (Ismail et al. 2009). Therefore, it can be suggested that the main cause of the high proton conductivity for SP/Cl/TAP was due to the conductivity of the Cloisite 15A itself. It was reported that the proton formed in the natural clay possesses a good proton conductivity of 10−4 S/cm at room temperature (Lin et al. 2007). In addition, the low proton conductivity obtained from the SP/Cl membrane might be due to the agglomeration of Cloisite 15A at certain regions in the polymer matrix, which might retard the mobility of the protons in the membranes (Chuang et al. 2007). Methanol permeability is another crucial fundamental test of a proton exchange membrane for DMFC application. From the permeability results, it was found that

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

427

all SPEEK composite membranes showed methanol permeability lower than that of the Naion 112 membrane. The corresponding methanol permeability for Naion 112 was 15.72 × 10−7 cm2/s. A comparable value was also reported by Ismail et al. (2009). The higher methanol permeability value for the Naion 112 membrane obtained was possibly due to the difference in the microstructures between the composite SPEEK membranes and the Naion 112 membrane. The high number of hydrophobic/hydrophilic domains of the Naion 112 membrane not only allowed water and protons to migrate in the membrane but also smaller polar molecules such as methanol to pass through (Kreuer 2001; Gao et al. 2003). The result also showed that the SP/Cl/TAP membrane possessed the lowest methanol permeability. It was suggested that the addition of Cloisite nanoiller with good compatibility has improved the barrier properties of the parent SPEEK membrane toward methanol. This is because the well-dispersed Cloisite 15A nanoillers increase the tortuous path for methanol across the membranes due to its high length-to-width ratio (Kim et al. 2007). It was reported that Cloisite 15A possesses the aspect ratio of 70:150, which is the highest among nanoclays (Hanley et al. 2001). However, the methanol permeability value for SP/Cl membranes was higher as compared with the native SPEEK and other SPEEK composite membranes tested. This might be due to the poor compatibility between the SPEEK and the Cloisite 15A nanoiller. SP/TAP membranes showed lower methanol permeability than pure SPEEK and SP/Cl membranes, which might be due to the formation of strong hydrogen bonds between SPEEK and TAP. It can be deduced that when the hydrogen bonds are formed between the polymer chains and the low-molecular-weight additives such as TAP, the free volume of polymers will decrease, which will result in a decrease in their permeabilities (Yong et al. 2001). In order to be an excellent PEM for DMFC, high proton conductivity with low methanol permeability PEM is desirable. The ratio of proton conductivity to methanol permeability relects the overall membrane characteristics and can be used as an overall performance evaluation factor (Nasef et al. 2006). Figure 13.7 illustrates the overall membrane characteristic of SPEEK for different SPEEK composite membranes and the Naion 112 membrane. Although Naion 112 possesses the highest proton conductivity, its high methanol permeability restricts its ability to perform as a good PEM for DMFC application. Interestingly, loading a very small amount of Cloisite 15A and TAP into the SPEEK membrane with a low DS has made the membrane a very good candidate for DMFC PEM, owing to the acceptable proton conductivity value and very low methanol permeability. This behavior resulted in a high overall membrane characteristic, which was 75% higher as compared with the Naion 112 membrane. It was believed that by loading a higher and desirable amount of Cloisite 15A and TAP into the native SPEEK membrane, higher proton conductivity was conferred. This was in agreement with the proposed proton-hopping mechanism for proton transport in SPEEK membranes. The transport of proton was promoted by the long chain surfactant tallow of the Cloisite 15A itself and by the good interaction, via hydrogen bonding, that was improved by TAP (Matsuura 1994). It can be concluded that TAP was successfully incorporated into the SPEEK/ Cloisite 15A matrix to improve the compatibility between SPEEK and Cloisite 15A and thus to enhance the homogeneity of SPEEK/Cloisite15A/TAP nanocomposite

428

Membrane Modification: Technology and Applications

Ratio of proton conductivity to methanol permeability

35,000 30,000 25,000 20,000 15,000 10,000 5,000 0 SPEEK

SP/Cl

SP/Cl/TAP SP/TAP

Nafion

FIGURE 13.7 Ratio of the proton conductivity to the methanol permeability for SPEEK nanocomposite membranes and Naion 112 membrane. (Reprinted from Jaafar, J., Ismail, A.F., and Matsuura, T., J. Memb. Sci., 345, 119–127, 2009. With permission.)

membranes. The developed intercalated SPEEK/Cloisite 15A/TAP nanocomposite membranes have improved membrane barrier properties due to the unique feature of Cloisite 15A that contributed to the formation of a longer pathway for methanol transport across the membrane. The stability of the produced membranes in water and in aqueous methanol solution was enhanced by the incorporation of TAP into the SPEEK/Cloisite 15A nanocomposite membrane. The methanol permeability was signiicantly reduced by the incorporation of Cloisite 15A and TAP in the SPEEK membrane. However, the proton conductivity of the all SPEEK nanocomposites loaded with 1 wt% of the inorganic nanoclay was still lower than that of the Naion 112 membrane. Nevertheless, due to the methanol permeability that was much lower than that of the Naion 112 membrane, the nanocomposite membranes showed a signiicant increase in the overall membrane performance as compared with the pristine SPEEK and Naion 112 membranes. 13.5.3.1.2 Modification of SPEEK with Boron Orthophosphate Among several types of additives, boron orthophosphate (BPO4) has been considered as a potential inorganic material for composite membrane in fuel cell applications due to its high proton conduction. BPO4 is categorized in the class of orthophosphates and consists of phosphorus and boron tetrahedrally coordinated by oxygen. It was reported that proton conduction takes place in a more effective way in the polymer composite membrane (SPEEK/PBI/BPO4) that possessed lower water uptake than that of the SPEEK/PBI blend membrane. This was attributed to the presence of conductive BPO4 in the composite membrane. Therefore, it was reported that proton conductivity is more dependent on BPO4 than the water uptake (Zaidi 2005). The most convenient and inexpensive method for preparing BPO4 is by reacting boric acid (H3BO3) with phosphoric acid (H3PO4) (Keary and Moffatt 1992).

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

429

An appropriate amount of H3PO4 and H3BO3 was stirred continuously at 120°C in air until a thick mass was formed. It was then kept at the same temperature for 4 h without stirring and subsequently subjected to calcination for 12 h in air at 410°C. The obtained solids were washed in distilled water (5 g of solid per 50 ml of distilled water) and separated by centrifugation. The dried samples were ground and sieved to obtain BPO4 powder with a 28 μm mesh grade. SPEEK with 63% DS was irst dissolved in N,N-dimethylacetamide (DMAc) to make a 10 wt% solution and the appropriate weight (0–60 wt%) of BPO4 was then added to the solution. The resulting mixture was stirred for 16–24 h prior to normal casting and drying procedures. The performance of the SPEEK/BPO4 composite membrane should be closely related to its internal structure, especially morphology. Figure 13.8 shows the SEM micrographs of the composite membrane at various BPO4 loading. As shown in Figure 13.8a and b, it was evident that the pure SPEEK membrane and composite membrane at low BPO4 loading (≤20 wt%) were free from pore defect. However, a higher loading of BPO4 into the polymer matrix has a signiicant impact on the polymer structure. Figure 13.8c and d indicate the presence of micropores around the BPO4 particles in the composite membranes when 40 and 60 wt% BPO4 were embedded in the microstructure, respectively.

(a)

(b)

20 µm

20 µm

(d)

(c)

20 µm

20 µm

FIGURE 13.8 SEM micrographs of the cross-sectional cryogenic fracture of SPEEK/BPO4 composite membranes (magniication 500×); (a) 0 wt% BPO4, (b) 20 wt% BPO4, (c) 40 wt% BPO4, and (d) 60 wt% BPO4. (Reprinted from Othman, M.H.D., Ismail, A.F., and Mustafa, A., J. Memb. Sci., 299, 156–165, 2007. With permission.)

430

Membrane Modification: Technology and Applications

It is expected that the micropores were formed during the blending of BPO4 powder with SPEEK solution; SPEEK solution was prepared by dissolving polymer in the solvent. Since the BPO4 had no interaction with SPEEK, and the size of 10–28 μm BPO4 powder was not ine enough, the incorporation of BPO4 solid into polymer solution would affect the compatibility of the resultant solution, especially at high content of BPO4. This incompatibility led to the weak attachment of BPO4 in the SPEEK membrane, and thus formation of the micropores around the BPO4 particles. Further evidence of the incompatibility can be seen in Figure 13.8c and d, as the BPO4 was easily pulled out from the polymer matrix, leaving the pores behind it. Since water molecules dissociate acid functionality and facilitate proton transport, the water uptake becomes an important parameter in studying the electrolyte membranes for fuel cell application (Li et al. 2003b). Therefore, it is crucial to fully understand the relationship between the water uptake and the composite membrane variables, such as BPO4 loading and B/P fraction of BPO4. Figure 13.9 shows the water uptake of composite membrane against the BPO4 loading. Interestingly, the water uptake of membranes increased with increasing BPO4 loading up to 20 wt%, but started to decrease after further addition of BPO4. However, the formation of micropores around the BPO4 particles in the composite membrane at BPO4 loading of 40 wt% or more, as proved by SEM micrographs in Figure 13.8, has reduced the retention of water in the membrane. The micropores in the membrane structure have enabled the water molecules to bleed out from the membrane and therefore the water uptake for 40 and 60 wt% of BPO4 in composite membranes was much lower than 20 wt% loading of BPO4 and even lower than pure SPEEK membrane. Methanol permeability of composite membranes at room temperature as a function of BPO4 loading is shown in Figure 13.10 for different B/P ratios of BPO4. It can be observed that the increment in BPO4 content led to an increase in methanol

Water uptake (%)

45 40

BPO4 with 0.8 B/P BPO4 with 1.0 B/P BPO4 with 1.2 B/P

35

Nafion 117

30 25 20 15

0

10

20 30 40 BPO4 loading (wt%)

50

60

FIGURE 13.9 Water uptake of SPEEK/BPO4 composite membranes as a function of BPO4 loading. (Reprinted from Othman, M.H.D., Ismail, A.F., and Mustafa, A., J. Memb. Sci., 299, 156–165, 2007. With permission.)

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

431

Methanol permeability (cm2/s)

1.0E-05

1.0E-06 BPO4 with 0.8 B/P BPO4 with 1.0 B/P BPO4 with 1.2 B/P Nafion 117 1.0E-07

0

10

30 40 20 BPO4 loading (wt%)

50

60

FIGURE 13.10 Methanol permeability of SPEEK/BPO4 composite membranes as a function of BPO4 loading at room temperature. (Reprinted from Othman, M.H.D., Ismail, A.F., and Mustafa, A., J. Memb. Sci., 299, 156–165, 2007. With permission.)

permeability of the composite membrane. At low content of BPO4 (0–20 wt%), the methanol permeability increased signiicantly, before the rate of increase was reduced when 40 wt% of BPO4 was reached. As the content of BPO4 in the membrane became more than 40 wt%, a sharp rise in methanol permeability occurred. However, this trend was found to be contrary to the water uptake correlation with BPO4 content as shown in Figure 13.9, where the water uptake of composite membrane drastically decreased at 40 and 60 wt% of BPO4 content. This contrast can be explained by looking into the factors that inluence methanol permeability and water uptake. The factor that enhances the water absorption in the composite membrane is the presence of nanopores of BPO4, while the porosity of the membrane is noted as the factor to decrease the membrane water uptake. On the other hand, the increase in methanol permeability is caused by both nanopores of BPO4 and membrane porosity. The mutual interaction between the hydrophilic domains and the hydrophobic domains affects the methanol transport; the hydrophilic domain is responsible for high permeability while the hydrophobic domain hinders water/methanol transport. The modiication of composite membrane with nanoporous BPO4 is expected, improving the overall water content of the composite membrane, resulting in enlarging the hydrophilic domains in membranes. The enlarged hydrophilic domains may have overcome the expelling forces of the hydrophobic domains formethanol, which then resulted in increased methanol permeability. Apart from the nanopores in BPO4, the presence of micropores in the composite membrane also contributed to an increase in methanol permeability. Owing to the fact that methanol transport is accomplished through the micropores and the connecting channels, the formation of micropores in the composite membrane led to higher methanol permeability due to less tortuous pathways for methanol molecules.

432

Membrane Modification: Technology and Applications

The effect of BPO4 loading on the proton conductivity of composite membrane at room temperature is shown in Figure 13.11. It can be seen from the igure that the composite membrane always exhibited higher conductivity than the pure SPEEK membrane, although conductivity decreased at the 40 and 60 wt% loading of BPO4. As expected, this trend is in agreement with the water uptake results shown in Figure 13.9, since the proton conductivity, as reported by Kreuer, relies more heavily on the amount of water (Kreuer 2001). The composite membrane with the highest water uptake possessed the highest proton conductivity, and the decrease in water uptake also reduced the conductivity value (Kreuer 2001). This correlation can be explained by looking into the transportation of protons in the hydrophilic domains, which was facilitated by the presence of water molecules. Apart from the facilitation of proton transport by water molecules, the increase in proton conductivity is also associated with the acidic site of BPO4 in the presence of water. Since there are plenty of acidic sites on the surface, the proton transport in the composite membrane has probably originated from the dissociation of adsorbed water molecules by the hydrolysis mechanism. In this mechanism, the impregnation of BPO4 will increase the number of ionic sites in the composite membrane in the presence of H3O+ ions, giving rise to mobile protons. From these results, it can also be seen that proton conductivity was more dependent on the acidic sites of BPO4 rather than the water content. Referring to Figure 13.9, the water uptake at 40 and 60 wt% loading of BPO4 was not as high as pure SPEEK membranes, but in Figure 13.11, these membranes showed higher conductivity values. Therefore, the presence of BPO4 not only increased the proton conductivity of composite membranes but also reduced the dependency of conductivity on water molecules (Mikhailenko et al. 1998). As a conclusion, SPEEK/BPO4 composite membrane has been successfully prepared and characterized, speciically for DMFC application. In terms of the

Proton conductivity (S/cm)

0.1 BPO4 with 0.8 B/P BPO4 with 1.0 B/P BPO4 with 1.2 B/P Nafion 117 0.01

0.001 0

10

20 30 40 BPO4 loading (wt%)

50

60

FIGURE 13.11 Proton conductivity of SPEEK/BPO4 composite membranes as a function of BPO4 loading at room temperature. (Reprinted from Othman, M.H.D., Ismail, A.F., and Mustafa, A., J. Memb. Sci., 299, 156–165, 2007. With permission.)

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

433

membrane performance, the composite membrane showed enhancement in proton conductivity with the incorporation of BPO4 particles into SPEEK up to 20 wt% and decreased suddenly after further BPO4 addition. The highest proton conductivity, which reached 3.35 × 10−3 S/cm, is only slightly lower than that for the Naion 117 membrane. The increase in proton conductivity of this composite membrane was probably associated with the acidic site of BPO4 and the presence of water in the composite membranes. The methanol permeability of this composite membrane proportionally increased with BPO4 loading but was still much lower than that of the Naion 117 membrane. 13.5.3.1.3

Modification of SPEEK by Incorporating Tungstosilicic Acid Supported on Silica–Aluminum Oxide This section describes the use of tungstosilicic acid supported on silica–aluminum oxide as an inorganic additive. Silica can function as a support material and tends to improve the conductivity characteristics of the membrane, probably by increasing the water absorption and creating the path for proton motion. Silica oxides are acidic in nature, whereas aluminum oxides are basic. It has been observed that the surface acid–base properties of the inorganic illers play a fundamental role in determining the conductivity characteristics and fuel cell performance (Arico et al. 2003b). In order to prepare the SPEEK/tungstosilicic acid/silica–aluminum oxide composite membrane, the inorganic proton-conducting solids were irst synthesized. The inorganic proton-conducting solids were prepared by dissolving 30–70 wt% of tungstosilicic acid (SiWA) in 100 ml of deionized water and neutralized by adding a few drops of 1M sodium hydroxide (NaOH) solution. The neutralization of the supported tungstosilicic acid was a necessary step to avoid the precipitation of SPEEK. Then 25–75 wt% of silica oxide (SiO2) and aluminum oxide (Al2O3) were added to make a suspension. The suspension was stirred in an ultrasonic bath for 6–7 h and dried at 100°C to evaporate the water. The solid obtained was ground to make a very ine powder and then dried further. The procedures above were adopted to establish strong HPA and silica–aluminum oxide (SiO2–Al2O3) interaction. Figure 13.12 illustrates the effects of the DS and the incorporation of SiWA and SiO2–Al2O3 composite on water uptake measured at room temperature. In all cases, the water uptake of the composite membrane behaves in a similar fashion, where the membranes that have higher DS, higher SiWA content, and higher SiO2 content in SiO2–Al2O3 composites show higher water uptake. In addition, the water uptake of all prepared membranes is found to be higher than the Naion 112 membrane. This is because the sulfonation process is known to enhance the hydrophilicity of the polymer through the introduction of a sulfonic group. Thus, the resultant SPEEK polymers with higher DS and greater hydrophilicity are able to absorb a larger amount of water (Zaidi et al. 2000; Li et al. 2003c). The water uptake of the membranes was also increased upon increase in the hydrophilic solid SiWA content in the SPEEK matrices. This may be due to the presence of Keggin structures of heteropolyacids, which contain high amounts of water and still can take more water due to the speciic Keggin properties. The density of Keggin structures is very high, most probably due to the clustering or agglomeration. Clustered ionomers absorb more water, and therefore a large increase in water

434

Membrane Modification: Technology and Applications 100.00 90.00

Nafion 112 DS56/25SO/75AO

SPEEK 56 DS56/50SO/50AO

SPEEK 66 DS56/75SO/25AO

DS66/25SO/75AO

DS66/50SO/50AO

DS66/75SO/25AO

80.00

Water uptake (%)

70.00 60.00 50.00 40.00 30.00 20.00 10.00 0.00 0

10

20 30 40 50 60 Tungstosilicic acid compositions (wt%)

70

80

FIGURE 13.12 Water uptake for different contents of SiO2 in SiO2–Al2O3 composite at various SiWA compositions. (Reprinted from Ismail, A.F., Othman, N.H., and Mustafa, A., J. Memb. Sci., 329, 18–29, 2009. With permission.)

uptake was observed, which due to the presence of ion-rich regions proton transfer was faster. It also appears that SiO2 is more hygroscopic than Al2O3, and systematically yields higher water uptake with an increase in the SiO2 content until ~77% is achieved by the (DS66/70SiWA/75SO/25AO) membrane. At this stage, the membrane swells, but the phenomenon is not obvious due to the presence of mesoporous silica and aluminum oxide, which are known to have better water management and retention (Zaidi and Ahmad 2006). This was found true by Watanabe et al. as they reported that water uptake by the oxide-containing membrane is higher than that by the pristine Naion (Watanabe et al. 1994). Studies by Othman et al. also found that the inorganic materials that are porous solid help the membrane to absorb and retain water inside the polymer matrix (Othman et al. 2007). Figure 13.13 illustrates the dependence of the proton conductivity on the incorporation of SiWA and SiO2–Al2O3 composite at room temperature. The result for Naion 112 is also presented for comparison with the membranes prepared. Only pure SPEEK56 membrane without any addition of inorganic materials possessed lower proton conductivity than Naion 112 membrane while other membranes showed higher values of proton conductivity as shown in Figure 13.14. In general, it is known that the proton conductivity does not directly correlate with either water uptake or IEC for any of the polymers applicable in fuel cell technology. Sulfonated hydrocarbon polymers usually have greater water uptake than the perluorosulfonic polymers in order to achieve the same proton conductivity, which is probably because the acid content in the hydrocarbon polymers is much lower

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

435

–10,000

Nafion 112.z DS56.z DS56-30S-25SO-75AO.z DS56-30S-50SO-50AO.z DS56-30S-75SO-25AO.z DS56-50S-25SO-75AO.z DS56-50S-50SO-50AO.z DS56-50S-75SO-25AO.z DS56-70S-25SO-75AO.z DS56-70S-50SO-50AO.z DS56-70S-75SO-25AO.z

Z″

–7,500

–5,000

–2,500

0

0

2,500

7,500

5,000

10,000

12,500

Z′

(a) –10,000

Z″

–7,500

DS66.z DS66-30S-25SO-75AO.z DS66-30S-50SO-50AO.z DS66-30S-75SO-25AO.z DS66-50S-25SO-75AO.z DS66-50S-50SO-50AO.z DS66-50S-75SO-25AO.z DS66-70S-25SO-75AO.z DS66-70S-50SO-50AO.z DS66-70S-75SO-25AO.z

–5,000

–2,500

0 (b)

0

2,500

5,000

7,500

10,000

12,500

Z′

FIGURE 13.13 (a, b) Nyquist plot of the Naion 112 and DS56 membrane at various SiWA and SiO2 contents in SiO2–Al2O3 composite. (Reprinted from Ismail, A.F., Othman, N.H., and Mustafa, A., J. Memb. Sci., 329, 18–29, 2009. With permission.)

than in the PFSA polymers. This shortage of acid content is compromised by adding the inorganic materials to enhance the proton conductivity of the membranes. This effect can be clearly seen where the proton transport is increased linearly with the amount of SiWA and SiO2. It is worth noting that for the membrane with the highest DS, highest content of SiWA, and highest content of SiO2 in SiO2–Al2O3 composite

436

Membrane Modification: Technology and Applications Nafion 112 DS56/25SO/75AO DS66/25SO/75AO

0.0700

SPEEK 56 DS56/50SO/50AO DS66/50SO/50AO

SPEEK 66 DS56/75SO/25AO DS66/75SO/25AO

Proton conductivity (S/cm)

0.0600

0.0500

0.0400

0.0300

0.0200

0.0100

0.0000 0

10

20 30 40 50 60 Tungstosilicic acid compositions (wt%)

70

80

FIGURE 13.14 Proton conductivity for different contents of SiO2 in SiO2–Al2O3 composite at various SiWA compositions. (Reprinted from Ismail, A.F., Othman, N.H., and Mustafa, A., J. Memb. Sci., 329, 18–29, 2009. With permission.)

(DS66/70SiWA/75SO/25AO), the value of proton conductivity becomes very high, achieving up to 6.10 × 10−2 S/cm. The main cause for the increase in proton conductivity with increased amount of SiWA is due to the conductivity of SiWA itself. It has also been noticed that when the SiO2 content in SiO2–Al2O3 composite is increased, the proton conductivity is also increased because SiO2 possesses more hydrophilic character compared with Al2O3. The hygroscopic properties of SiO2 increase the proton conductivity by increasing the water absorption and creating the pathway for proton motion by entrapping the ‘liquid’ due to the mesoporous structure. Results of methanol permeability at various DS, SiWA, and SiO2 are plotted in Figure 13.15. It can be seen that no membrane possesses higher methanol permeability than the Naion 112 membrane. The corresponding methanol permeability for Naion 112 was about 15.07 × 10−7 cm2/s. The methanol permeabilities of composite membranes based on SPEEK are obviously lower than that of Naion 112, which is possibly due to the microstructure differences between the SPEEK and the Naion 112 membrane. Kreuer reported that the Naion membrane has high hydrophobicity of the perluorinated backbone and also high hydrophilicity of the sulfonic groups (Kreuer 2001). In the presence of water, this character is more pronounced and consequently increases the hydrophobic/hydrophilic domains of the Naion membrane.

437

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application 1.80E–06

Nafion 112 DS56/25SO/75AO DS66/25SO/75AO

Methanol permeability (cm2/s)

1.60E–06

SPEEK 56 DS56/50SO/50AO DS66/50SO/50AO

SPEEK 66 DS56/75SO/25AO DS66/75SO/25AO

1.40E–06 1.20E–06 1.00E–06 8.00E–07 6.00E–07 4.00E–07 2.00E–07 0.00E+00 0

10

20 30 40 50 60 Tungstosilicic acid composition, SiWA (%)

70

80

FIGURE 13.15 Methanol permeability for different contents of SiO2 in SiO2–Al2O3 composite at various SiWA compositions. (Reprinted from Ismail, A.F., Othman, N.H., and Mustafa, A., J. Memb. Sci., 329, 18–29, 2009. With permission.)

The smaller hydrophobic/hydrophilic separation and the lesser lexibility of the polymer backbone of SPEEK produce narrow proton channels and a highly branched structure, which bafles the transfer of methanol. It is very interesting to observe that the methanol permeability of composite membrane decreases with increasing DS. This observation can be attributed to the strong interaction between the SPEEK and the inorganic materials at higher DS. The inorganic materials function as barriers by closing the membrane pores and hence reduce the methanol permeability. This was demonstrated by Wang et al., as they reported that the holes in the copolymer matrix formed by HPA particle extraction decreased from 5–10 μm to 80 nm as the DS of the copolymers was increased from 0.2 to 0.6 (Wang et al. 2006). When the DS was increased, fewer hydrogen-bonding sites were attached to the sulfonic acid groups. As a result, the interactions between HPA and sulfonic acid groups gradually became stronger because of the increased number of unattached sulfonic acid groups. The stronger interactions reduce the size of the membrane pores, thus reducing the methanol permeability. However, when the SiO2 composition in SiO2–Al2O3 composite is increased, the methanol permeability also increases. This is most likely due to the hygroscopic character of SiO2 that absorbs water. On the other hand, when the content of Al2O3 is increased, the methanol permeability decreases because the structure of Al2O3, which is much coarser than SiO2, serves as a barrier, thus reducing the pathways for methanol to permeate. Although the methanol permeability of the prepared membranes is considerably lower than that of the Naion 112 membrane, further improvement can be made by performing surface modiication on the inorganic illers.

438

Membrane Modification: Technology and Applications

Based on the experimental results and analysis, the following conclusions are derived. The membranes that possess higher DS, higher SiWA content, and higher SiO2 content in SiO2–Al2O3 composite show higher water uptake and proton conductivity. The water uptake of all prepared membranes was found to be higher than that of the Naion 112 membrane. The methanol permeability of the composite membrane decreases when DS is higher. However, when the SiO2 composition in SiO2–Al2O3 composite is increased, the methanol permeability also increases but the value is still lower than Naion 112. It can be seen that each of the prepared membranes had a higher overall evaluation factor than the Naion 112 membrane. Membrane DS 66/70SiWA/25SO/75AO possessed the highest overall factor, which was 58.95 × 103, while the Naion 112 membrane possessed the lowest overall factor, which was 10.65 × 103. These features enable SPEEK/SiWA/SiO2–Al2O3 membranes to appear as highly favorable and potential candidates for DMFC application. 13.5.3.2 Modification of SPEEK by Surface Modification There are many reported studies on the modiication of SPEEK using surface modiication techniques. Zhong et al. have prepared a series of SPEEK membranes via photochemical cross-linking (Zhong et al. 2007). SPEEK was irst dissolved with benzophenone and triethylamine photoinitiator, then the membrane was cross-linked by exposure to UV light. They reported that the mechanical and thermal stabilities were better for the modiied membrane. Furthermore, the water uptake was reduced, which also reduced the methanol permeability with only slight sacriice in proton conductivities. Then, Chen et al. reported on using a double cross-link technique, where they irst cross-linked the PEEK ilms with doses of more than 33 MGy and then sulfonated in chlorosulfonic solution (Chen et al. 2007). They claimed that for the modiied SPEEK, after further heat treatment, the water uptake and methanol permeability were reduced with a slight reduction in proton conductivity. Ren et al. introduced a layer technique where they immersed the SPEEK membrane in a Naion casting solution, as a result of which a layer membrane of Naion– SPEEK–Naion was prepared (Ren et al. 2005). The proton conductivity of the modiied membrane was greater than the original SPEEK, but, unfortunately, the methanol permeability also increased. This shows that Naion, which has high methanol permeability and is on the top and bottom layers of the modiied SPEEK, has some inluence in increasing the methanol permeability. Even though there have been various methods of SPEEK modiication, there is no reported study on the surface-modifying macromolecule (SMM) approach. The next section looks at surface modiication in a simpler way; that is, by the addition of SMMs so that the membrane surfaces can be modiied via a single casting step. 13.5.3.2.1

Modification of Sulfonation of Poly(Ether Ether Ketone) by Blending with Charged Surface-Modifying Macromolecule There have been many initiatives to modify the membrane surfaces at laboratory scale for various membrane applications. Most of the work attempted to change the hydrophilicity/hydrophobicity of the membrane surface. The addition of a compatible modiier into the casting solution is an excellent option that can modify the membrane surface using a single casting step. The SMMs are polymers tailor-made

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

439

to be compatible and end-grouped with either the hydrophobic or hydrophilic tails (Suk et al. 2002). During casting, SMMs preferably migrate to the membrane surfaces in order to minimize the total free energy of the system (Hamza et al. 1997). This migration then modiies the membrane surface to be either more hydrophilic or hydrophobic depending on the hydrophilicity or hydrophobicity of the SMMs added (Rana et al. 2005). In this chapter, a different kind of modiication with the SMM method is discussed. A new formulation of a charged SMM, cSMM, which was anticipated to increase the hydrophilicity of the base polymer (SPEEK) as well as its sulfonic group contents, was developed. It is well documented that the enhancement of both properties is crucial for a better PEM performance of SPEEK as its proton conductivity is dependent on the water uptake and sulfonic group content in its polymer chain. It is expected that the proton conductivity will be further increased by the interaction of the sulfonic group from cSMM with that of SPEEK. The SMM synthesis consists of the following two steps: (1) Reacting diisocyanate with polyol to form polyurethane prepolymer, and (2) End-capping of the prepolymer with chemicals with appropriate functional groups to endow hydrophobic or hydrophilic or charged properties to the SMMs. Until now, the main diisocyanates used for SMM preparation were methylene bis-p-phenyl isocyanate (methylene diphenyl diisocyanate, MDI) and hexamethylene diisocyanate (HDI) (Qtaishat et al. 2009). The polyol employed was diethylene glycol (DEG) (Kim et al. 2004a). As for the end group, hydrophobic, hydrophilic, and cSMM were made by end-capping with luoroalcohol, poly(ethylene glycol) (PEG), and hydroxyl benzene sulfonate (HBS), respectively. The enrichment of the cSMM at the SPEEK surface would occur as a result of a driving force for a spontaneous surface migration in order to minimize the interfacial energy (Rana et al. 2005). In a polymer blend, thermodynamic incompatibility between the polymers usually causes the separation of the polymers. If an interface is formed between the polymer and the external environment, the polymer with the lowest surface energy will concentrate at the interface and consequently reduce the system’s interfacial tension. It was observed that during the preparation procedure of the SPEEK/cSMM membrane, cSMM migrated to the ilm surface. The migration resulted in the surface becoming more hydrophilic than the bulk membrane phase. Because only a small amount of SMM was added to the casting dope, usually less than 5 wt%, the bulk properties of the base polymer remained relatively unchanged. The synthesis and the chemical structure of cSMM are depicted in Figure 13.16. Mixing SPEEK and cSMM (MDI–DEG–HBS) in NMP formed homogeneous and transparent solution. In the solution, cSMM and SPEEK, due to the presence of sulfonic group in both, are thermodynamically miscible. However, after casting on a glass plate, the modiied SPEEK membrane became more yellowish than the original SPEEK. The yellowish color was probably caused by the migration of the cSMM to the membrane surfaces. The migration of cSMM to the membrane surface was conirmed by the water contact angle (CA)measurement. The hydrophilicity/hydrophobicity of the membrane surface can be evaluated by measuring the CA. Table 13.4 shows the results of the CA measurement. The CA of SPEEK is signiicantly higher than the SPEEK/cSMM blend at the top surface, the

440

HO

(CH2)2

H

O

N

C

NaO3S

NaO3S

O

O

H

C

N

CH2

OH +

(CH2)2

OCN DMAc N2 atmosphere 48–50°C 3 hours

(CH2)2 (CH2)2 O O Urethane prepolymer DMAc N2 atmosphere OH 48–50°C 24 hours H O

O

N

(CH2)2

C

O

(CH2)2

NCO

CH2

O

O

H

C

N

O

NCO

CH2

O

H

C

N

m

CH2

H

O

N m

C

O

SO3Na

Charged surface-modifying macromolecule (cSMM, MDI-DEG-HBS)

FIGURE 13.16 Synthesis and chemical structure of cSMM. (Reprinted from Mohd-Norddin, M.N.A., Ismail, A.F., Rana, D., Matsuura, T., Mustafa, A.J., and Tabe, A.M., J. Memb. Sci., 323, 404, 2008. With permission.)

Membrane Modification: Technology and Applications

CH2

OCN

O

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

441

TABLE 13.4 Characterization of Nafion 112, SPEEK, and SPEEK/cSMM Films

Polymer Film

Water Uptake (%)

CA (°) Top Surface

Methanol Permeability, τ (cm2/sec)

Proton Conductivity, σ (S/cm)

Naion 112 SPEEK SPEEK/cSMM

20 32 38

80 83.12 ± 2.36 65.82 ± 2.97

6.21 × 10−6 3.14 × 10−7 2.75 × 10−7

1.20 × 10−2 3.3 × 10−3 6.4 × 10−3

Source: Data from Mohd-Norddin, M.N.A., Ismail, A.F., Rana, D., Matsuura, T., Mustafa, A., and Tabe, A.M., J. Memb. Sci., 323, 404, 2008. With permission.

side that was in contact with air when the membrane was cast. This is most likely due to the migration of cSMM to the top surface, covering the surface with the sulfonic acid end groups that protrude vertical to the surface, while the urethane prepolymer part of cSMM is anchored to the host SPEEK. Therefore, the addition of cSMM has successfully altered the modiied SPEEK/cSMM membrane to be more hydrophilic as compared with the pristine SPEEK. Table 13.4 also shows that water uptake increased signiicantly from SPEEK to SPEEK/cSMM blend. Methanol permeability measurement is one of the fundamental tests of the membrane for DMFC application. Table 13.4 shows the methanol permeability of Naion 112, SPEEK, and SPEEK/cSMM blend membranes. It can be observed from the table that the methanol permeability of the SPEEK membrane, 3.14 × 10−7 cm2/s, is lower than that of the Naion 112 membrane, 6.21 × 10−6 cm2/s. Interestingly, the methanol permeability of the SPEEK/cSMM blend membrane, 2.75 × 10−7 cm2/s, is even lower than that of the SPEEK membrane. The decrease in methanol permeability is rather unexpected since the water uptake increased by blending cSMM. However, both methanol permeability and water uptake are determined by a delicate interplay of the membrane water type. Even though the methanol permeability generally increases with an increase in the sorbed water, it is reported that it occurs dominantly via free water inside the interconnected membrane structure channels (Kim et al. 2004b, 2009). But methanol permeation dominantly occurs via free water (freezing water) inside the interconnected membrane structure channels and insigniicantly via nonfreezing bound water associated with the ionic sites. Since the addition of cSMM reduced, the free water content decrease in methanol permeability seems natural (Mohd-Norddin et al. 2008). Another important characteristic of the membrane for DMFC application is the proton conductivity of the membrane. Table 13.3 also shows the proton conductivity at room temperature and 100% RH for Naion 112, SPEEK, and SPEEK/cSMM membranes. The proton conductivity of Naion 112 is 1.20 × 10−2 S/cm, highest of all the three membranes. The proton conductivity for SPEEK is 3.3 × 10−3 S/cm. Interestingly, the proton conductivity of the SPEEK/cSMM blend membrane is 6.4 × 10−3 S/cm, about twice as high as that of SPEEK. This result is expected from the simultaneous increase in the water uptake, since the proton conductivity relies

442

Membrane Modification: Technology and Applications

heavily on the amount of water, which facilitates the transportation of proton in the ionic domains. From the indings, cSMM is clearly a good candidate as an additive for SPEEK in improving its function as a PEM. The addition of cSMM with an additional SO3‐ charge on the surface and enhancement of the phase separation of the hydrophilic/ hydrophobic domain have enabled the proton conductivity to be increased while, at the same time, the methanol permeability was reduced.

13.6

CONCLUSION

This chapter has covered an extensive study of the development of polymeric membranes based on Naion and SPEEK for DMFC. Both Naion and SPEEK membranes have been extensively modiied in order to perform better as PEM in DMFC applications. Two types of modiication techniques were described for both Naion and SPEEK: blending with inorganic additives and surface modiication. As for Naion, this chapter reviewed many types of inorganic compounds that are able to suppress the methanol permeability but, at the same time, decrease the proton conductivity. On the development of the SPEEK and inorganic composite membrane, three approaches using different inorganic additives were discussed. These additives include clay nanocomposite with compatibilizer, boron orthophosphate, and tungstosilicic acid supported on silica–aluminum oxide. These three types of inorganic additives were able to enhance the overall performance of SPEEK in its PEM performance. This chapter also describes the modiication of Naion and SPEEK by the surface modiication technique. In particular, SPEEK blended with charged SMMs has been found to possess a good potential in enhancing the proton conductivity while suppressing the methanol permeability. Nevertheless, challenges in commercializing the SPEEK as PEM still remain, and the overall performance of the membrane, as well as the current and power density of the DMFC itself, has to be improved further.

REFERENCES Allock, H.R., Hofmann, M.A., Ambler, C.M., Lvov, S.N., Zhou, X.Y., Chalkova, E. and Weston, J. 2002. Phenyl phosponic acid functionalized poly(arloxyphosphazenes) as proton-conducting membranes for direct methanol fuel cells. J. Memb. Sci. 201: 47–54. Apanel, G. and Johnson, E. 2004. Direct methanol fuel cells – Ready to go commercial? Fuel Cell Bull. 12–17, doi: 10.1016/S1464-2859(04)00410-9. Aparicio, M., Mosa, J., Etienne, M. and Duran, A. 2005. Proton conducting methacrylate-silica sol–gel membranes containing tungstophosphoric acid. J. Power Sour. 145: 231–236. Arico, A.S., Baglio, V., Creti, P., Di Blasi, A., Antonucci, V., Brunea, J., Chapotot, A., Bozzi, A. and Schoemans, J. 2003a. Investigation of grafted ETFE-based polymer membranes as alternative electrolyte for direct methanol fuel cells. J. Power Sour. 123: 107–115. Arico, A.S., Baglio, V., Blasi, A.D. and Antonucci, V. 2003b. FTIR spectroscopic investigation of inorganic illers for composite DMFC membranes. Electrochem. Commun. 5: 862–866. Bae, B. and Kim, D. 2003. Sulfonated polystyrene grafted polypropylene composite electrolyte membranes for direct methanol fuel cells. J. Memb. Sci. 220: 75–87. Bahar, B., Hobson, A.R., Kolde, J.A. and Zuckerbrod, D. 1996. Ultrathin integral composite membrane. U.S. Patent No. 5,547,551. Washington, DC, US, Patent and Trademark Ofice.

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

443

Bauer, B., Jones, D.J., Roziere, J., Tchicaya, L., Alberti, G., Cascioala, M., Massinelli, L., Peraio, A., Besse, S. and Ramunni, E. 2000. Electrochemical characterisation of sulfonated polyetherketone membranes. J. New Mater. Electrochem. Syst. 3: 93. Bauer, F. and Porada, M.W. 2004. Microstructural characteristion of Zr-phosphate–Naion membranes for direct methanol fuel cell (DMFC) applications. J. Memb. Sci. 233: 141–149. Bouchet, R. and Siebert, E. 1999. Proton conduction in acid doped polybenzimidazole. Solid State Ionics 118: 287–299. Caretta, L., Friedrich, K.A. and Stimming, U. 2001. Fuel cell fundamentals and applications. Fuel Cells 1: 5. Cervantes-Uc, J.M., Cauich-Rodriguez, J.V., Vazquez-Torres, H., Garias-Mesias, L.F. and Paul, D.R. 2007. Thermal degradation of commercially available organoclays studied by TGA–FTIR, Thermochim. Acta 457: 92–102. Chan, C.M. 1994. Polymer Surface Modiication and Characterization. Hanser Publishers: New York, NY. Chang, J.H., Park, J.H., Park, G.G., Kim, C.S. and Park, O.O. 2003. Proton-conducting composite membrane derived from sulfonated hydrocarbon and inorganic materials. J. Power Sour. 124: 18. Chen, C.-Y., Garnica-Rodriguez, J.I., Duke, C.M., Costa, F.D.R., Dicks, L.A. and Diniz-daCosta, C.J. 2007. Naion/polyaniline/silica composite membranes for direct methanol fuel cell application. J. Power Sour. 166: 324–330. Chen, H.-L. and Porter, R.S. 1994. Observation of two stage crystalization of poly(ether ether ketone) by thermal mechanical analysis. Thermochem. Acta 243: 109–113. Choi, W.C., Kim, J.D. and Woo, F.I. 2001. Modiication of proton conducting membrane for reducing methanol crossover in a direct-methanol fuel cell. J. Power Sour. 96: 411. Chuang, S.W. Chung, S.L. and Hsu, H.C.L. 2007. Synthesis and properties of luorine containing polybenzimidazole/montmorillonite nanocomposite membranes for direct methanol fuel cell applications. J. Power Sour. 168: 172–177. Costamagna, P. and Srinivasan, S. 2001. Quantum jumps in the PEMFC science and technology from the 1960s to the year 2000; Part II. Engineering, technology development and application aspects. J. Memb. Sci. 102: 253–269. Curtin, D.E., Lousenberge, R.D., Henry, T.J., Tengeman, P.C. and Tisack, M.E. 2004. Advanced materials for improved PEMFC performance and life. J. Power Sour. 131: 41–48. Dimitrova, P., Friedrich, K.A., Stimming, U. and Vogt, B. 2002. Modiied Naion-based membranes for use in direct methanol fuel cells. Solid State Ionics 150: 115–122. Doyle, M. and Rajendran, G. 2003. Handbook of Fuel Cells, p. 351. John Wiley and Sons: Chichester. Einsla, B.R. 2005. High temperature polymers for proton exchange membrane fuel cells. Ph.D. Thesis. Virginia Polytechnic Institute and State University. Fontanella, J.J., Wintersgill, M.C., Wainright, J.S., Savinell, R.F. and Litt, M. 1998. High pressure electrical conductivity studies of acid doped polybenzimodazole. Electrochim. Acta 43: 1289–1294. Fu, Y., Li, W. and Manthiram, A. 2008. Sulfonated polysulfone with 1,3-1H-dibenzimidolebenzene additive as a membrane for direct methanol fuel cells. J. Memb. Sci. 310: 262–267. Gao, W. and Sammes, N.M. 1999. An Introduction to Electronic and Ionic Materials. World Publisher: Singapore. Gao, Y., Robertson, G.P., Guiver, M.D., Jian, X., Mikhailenko, S.D., Wang, K. and Kaliaguine, S. 2003. Sulfonation of poly(phthalazinones) with fuming sulfuric acid mixtures for proton exchange membrane materials. J. Memb. Sci. 227: 39–50. Glipa, X., Haddad, M.E., Jones, D.J. and Roziere, J. 1997. Synthesis and characterization of sulfonated polybenzimidazole: A highly conducting proton exchange polymer. Solid State Ionics 97: 323–331.

444

Membrane Modification: Technology and Applications

Gosalawit, R., Chirachanchai, S., Shishatskiy, S. and Nunes, S.P. 2008. Sulfonated montmorillonite/sulfonated poly(ether ether ketone) (SMMT/SPEEK) nanocomposite membrane for direct methanol fuel cells (DMFCs). J. Memb. Sci. 323: 337–346. Hamza, A., Pham, V.A., Matsuura, T. and Santerre, J.P. 1997. Development of membranes with low surface energy to reduce the fouling in ultrailtration applications. J. Memb. Sci. 131: 217–227. Hanley, J.M., Muzny, C.D., Ho, D.L., Glinka, C.J. and Manias, E. 2001. A SANS study of organoclay dispersions. Int. J. Thermophys. 22: 1435–1448. Hirschenhofer, J.H. 1994. Cell status. IEEE AES Systems Magazine: Fuel 9: 10–15. Hudiono, Y., Choi, S., Shu, S., Korros, W.J., Tsapatsis, M. and Nair, S. 2009. Porous layered oxide/Naion nanocomposite membranes for direct methanol fuel cell applications, Microporous Mesoporous Mater. 118: 427–434. Ismail, A.F., Othman, N.H. and Mustafa, A. 2009. Sulfonated polyether ether ketone composite membrane using tungstosilicic acid supported on silica–aluminium oxide for direct methanol fuel cell (DMFC). J. Memb. Sci. 329: 18–29. Jaafar, J., Ismail, A.F. and Matsuura, T. 2009. Preparation and barrier properties of SPEEK/ Cloisite 15A®/TAP nanocomposite membrane for DMFC application. J. Memb. Sci. 345: 119–127. Jiang, R., Kunz, R.H. and Fenton, J.M. 2005. Investigation of membrane property and fuel cell behavior with sulfonated poly(ether ether ketone) electrolyte: Temperature and relative humidity effects. J. Power Sour. 150: 120–128. Jung, D.H., Cho, S.Y., Peck, D.H., Shin, D.R. and Kim, J.S. 2003. Preparation and performance of a Naion/montmorillonite nanocomposite membrane for direct methanol fuel cell. J. Power Sour. 118: 205–211. Jung, H.-Y. and Park, J.-K. 2007. Blend membranes based on sulfonated poly(ether ether ketone) and poly(vinylidine louride). Electrochem. Acta 52: 7464–7468. Keary, C. and Moffatt, J.B. 1992. Boron phosphate: Their preparation and characterization and the existence of a high surface area, mesoporous, excess boron solid. J. Colloid Interface Sci. 154: 8. Kim, D.S., Park, H.B., Rhim, J.W. and Lee, Y.M. 2004a. Preparation and characterization of crosslinked PVA/SiO2 hybrid membranes containing sulfonic acid groups for direct methanol fuel cell applications. J. Memb. Sci. 240: 37. Kim, H.J., Kim, H.J., Shul, Y.G. and Han, H.S. 2004b. Naion–Naion/polyvinylidene luoride–Naion laminated polymer membrane for direct methanol fuel cells. J. Power Sour. 135: 66–71. Kim, T.K., Kang, M., Choi, Y.S., Kim, H.K., Lee, W., Chang, H. and Seung, D. 2007. Preparation of Naion-sulfonated clay nanocomposite membrane for direct methanol fuel cells via a ilm coating process. J. Power Sour. 165: 1–8. Kim, Y., Rana, D., Matsuura, T. and Chung, W.J. 2009. Inluence of surface modifying macromolecules on the surface properties of poly(ether sulfone) ultra-iltration membranes. J. Memb. Sci. 338: 84–91. Kreuer, K.D. 2001. On the development of proton conducting polymer membranes for hydrogen and methanol fuel cells. J. Memb. Sci. 185: 29. Kwak, S.H., Yang, T.H., Kim, C.S. and Yoon, K.H. 2003. Naion/Mordenite hybrid membrane for high-temperature operation of polymer electrolyte membrane fuel cell. Solid State Ionics 160: 309–315. Lee, C.H., Min, K.A., Park, H.B., Hong, Y.T., Jung, B.O. and Lee, Y.M. 2007. Sulfonated poly(arylene ether sulfone)– silica nanocomposite membrane for direct methanol fuel cell (DMFC). J. Memb. Sci. 303: 258. Li, L., Xu, L. and Wang, Y. 2003a. Novel proton conducting composite membranes for directmethanol fuel cell. Mater. Lett. 57: 1406–1410.

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

445

Li, L., Zhang, J. and Wang, Y. 2003b. Sulfonated poly(ether ether ketone) membranes for direct methanol fuel cell. J. Memb. Sci. 226: 159–167. Li, Q., He, R., Jensen, J.O. and Bjerrum, N.J. 2003c. Approaches and recent development of polymer electrolyte membrane for fuel cells operating above 100°C. Chem. Mater. 15: 4896–4915. Libby, B., Smyrl, W.H. and Cussler, E.L. 2003. Polymer-zeolite composite membrane for direct methanol fuel cells. AIChE J. 49(4): 991–1001. Lin, Y.F., Yen, C.Y., Ma, C.C.M., Liao, S.H., Hung, C.H. and Hsiao, Y.H. 2007. Preparation and properties of high performance nanocomposite proton exchange membrane for fuel cell. J. Power Sour. 165: 692–700. Ma, Z.Q., Cheng, P. and Zhao, T.S. 2003. A palladium-alloy deposited Naion membrane for direct methanol fuel cells. J. Memb. Sci. 215: 327–336. Malhotra, S. and Datta, R. 1997. Membrane supported non-volatile acidic electrolytes allow higher temperature operation of proton exchange membrane fuel cells. J. Electrochem. Soc. 144: L23. Matsuura, T. 1994. Synthetic Membranes and Membrane Separation Processes. CRC Press: Boca Raton, FL. Mauritz, K.A. 1998. Organic–inorganic hybrid materials: Perluorinated ionomers as sol–gel polymerization templates for inorganics alkoxides. Mater. Sci. Eng. C 6: 121–133. Mikhailenko, S.D., Zaidi, S.M.J. and Kaliaguine, S. 1998. Electrical conductivity of boron orthophosphate in presence of water. J. Chem. Soc. Faraday Trans. 94(11): 1613–1618. Mohd-Norddin, M.N.A., Ismail, A.F., Rana, D., Matsuura, T., Mustafa, A. and Tabe, A. M., 2008. Characterization and performance of proton exchange membrane for direct methanol fuel cell: Blending of SPEEK with charged surface modifying macromolecules. J. Memb. Sci. 323: 404. Nasef, M.M., Zubir, N.A., Ismail, A.F., Dahlan, K.Z.M., Saidi, H. and Khayet, M. 2005. Preparation of radiochemically pore-illed polymer electrolyte membranes for direct methanol fuel cells. J. Power Sour. 156: 200–210. Nasef, M.M., Zubir, N.A., Ismail, A.F., Khayet, M., Dahlan, K.Z.M., Saidi, H., Rohani, R., Ngah, T.I.S. and Sulaiman, N.A. 2006. PSSA Pore-illed PVDF membranes by simultaneous electron beam irradiation: Preparation and transport characteristics of protons and methanol. J. Memb. Sci. 268: 96–108. Neburchilov, V., Martin, J., Wang, H. and Zhang, J. 2007. A review of polymer electrolyte membranes for direct methanol fuel cells. J. Power Sour. 169: 221. Nunes, S., Ruffmann, B., Rikowski, E., Vetter, S. and Richau, K. 2002. Inorganic modiication of proton conductive polymer membrane for direct methanol fuel cells. J. Memb. Sci. 203: 215–225. Othman, M.H.D. 2006. The development and characterization of composite polymer/inorganic material membrane for direct methanol fuel cell application. Master Thesis, Universiti Teknologi, Malaysia. Othman, M.H.D., Ismail, A.F. and Mustafa, A. 2007. Proton conducting composite membrane from sulfonated poly(ether ether ketone) and boron orthophosphate for direct methanol fuel cell application. J. Memb. Sci. 299: 156–165. Pham, V.A. 1995. Surface modifying macromolecules for enhancement of polyethersulfone pervaporation membrane performance. M.A.Sc. Thesis, University of Ottawa. Pivovar, B.S., Wang, Y. and Cussler, E.L. 1999. Pervaporation membranes in direct methanol fuel cells. J. Memb. Sci. 154: 155–162. Qtaishat, M., Rana, D., Khayet, M. and Matsuura, T. 2009. Preparation and characterization of novel hydrophobic/hydrophilic polyetherimide composite membranes for desalination by direct contact membrane distillation. J. Memb. Sci. 327: 264–273.

446

Membrane Modification: Technology and Applications

Rana, D., Matsuura, T., Narbaitz, R.M. and Feng, C. 2005. Development and characterization of novel hydrophilic surface modifying macromolecule for polymeric membranes. J. Memb. Sci. 249: 103–112. Ren, S., Li, C., Zhao, X., Wu, Z., Wang, S., Sun, G., Xin, Q. and Yang, X. 2005a. Surface modiication of sulfonated poly (ether ether ketone) membranes using Naion solution for direct methanol fuel cells. J. Memb. Sci. 247: 59–63. Rikukawa, K. and Sanui, K. 2000. Proton conducting polymer electrolyte membrane based on hydrocarbon polymers. Prog. Polym. Sci. 25: 1463–1502. Shao, P. 2003. Pervaporation dehydration membranes based on chemically modiied poly(ether ether ketone). Ph.D. Thesis, University of Waterloo. Shao, Z.G., Wang, X. and Hsing, I.M. 2002. Composite Naion/polyvinyl alcohol membranes for direct methanol fuel cell. J. Memb. Sci. 210: 147–153. Shen, M., Roy, S., Kuhlmann, J.W., Scott, K., Lovell, K. and Horsfall, J.A. 2005. Grafted polymer electrolyte membrane for direct methanol fuel cells. J. Memb. Sci. 210: 147–153. Silva, V.S., Ruffmann, B., Vetter, S., Mendes, A., Mandeira, L.M. and Nunes, S.P. 2005. Characterization and application of composite membranes in DMFCs. Catalysis Today 104: 205. Smit, E.A., Ocampo, A.L., Espinosa-Medina, M.A. and Sebastian, P.J. 2003. A modiied Naion membrane with in situ polymerized polypyrrole for the direct methanol fuel cell. J. Power Sour. 124: 59–64. Suk, D.E., Chowdhury, G., Matsuura, T., Narbaitz, R.M., Santerre, P., Pleizier, G. and Deslandes, Y. 2002. Study of kinetics of surface migration of surface modifying macromolecules in membrane preparation. Macromolecules 35: 3017–3021. Tazi, B. and Savodogo, O. 2001. New cation exchange membranes based on Naion, silicotungstic acid and thiophene. J. New Mater. Electrochem. 185: 3–27 Tian, S.H., Shua, D., Chen, Y.L., Xiao, M. and Meng, Y.Z. 2005. Preparation and properties of novel sulfonated poly(phthalazinone ether ketone) based PEM for PEM fuel cell application. J. Power Sour. 158: 88–93. Wang, H., Neburchilov, V., Martin, J. and Zhang, J. 2007. A review of polymer electrolyte membranes for direct methanol fuel cells. J. Power Sour. 169: 221–238. Wang, Z., Ni, H., Zhao, C., Li, X., Fue, T. and Na, H. 2006. Investigation of sulfonated poly (ether ether ketone sulfone)/heteropolyacid composite membranes for high temperature fuel cell applications. Polym. Sci. B Polym. Phys. 44: 1967–1978. Watanabe, M., Uchida, H., Seki, Y. and Emori, M. 1994. The Electrochemical Society Meeting, Abstract No. 606, Pennington, NJ, PV94-2, pp. 946–947. Woo, Y., Oh, S.Y., Kang, Y.S. and Jung, B. 2003. Synthesis and characterization of sulfonated polyimide membranes for direct methanol fuel cell. J. Memb. Sci. 220: 31–45. Xing, P., Robertson, G.P., Guiver, M.D., Mikhailenko, S.D., Wang, K. and Kaliaguine, S. 2004. Synthesis and characterization of sulfonated poly(ether ether ketone) for proton exchange membranes. J. Memb. Sci. 229: 95–106. Yang, B. and Manthiram, A. 2003. Sulfonated poly(ether ether ketone) membranes for direct methanol fuel cells. Electrochem. Solid State Lett. 6: A229–A231. Yang, C., Costamagna, P., Srinivasan, S., Benziger, J. and Bocarsly, A.B. 2001. Approaches and technical challenges to high temperature operation of proton exchange membrane fuel cells. J. Power Sour. 103: 1–9. Yong, H.H., Park, H.C., Kang, Y.S., Won, J. and Kim, W.N. 2001. Zeolite-illed polyimide membrane containing 2,4,6-triaminopyrimide. J. Memb. Sci. 188: 151–163. Yoon, S.R., Hwang, G.H., Cho, W.I., Oh, I.H., Hong, S.A. and Ha, H.Y. 2002. Modiication of polymer electrolyte membranes for DMFCs using Pd ilms formed by sputtering. J. Power Sour. 106: 215–223.

Modification of Sulfonated Poly(Ether Ether Ketone) for DMFC Application

447

Zaidi, S.M.J. 2003. Polymer sulfonation-A versatile route to prepare proton conducting membrane material for advanced technologies. Arabian J. Sci. Eng. 28(2B): 183–194. Zaidi, S.M.J. 2005. Preparation and characterization of composite membranes using blends of SPEEK/PBI with boron phosphate. Electrochim. Acta 50: 4771. Zaidi, S.M.J. and Ahmad, M.I. 2006. Novel SPEEK/heteropolyacids loaded MCM-41 composite membranes for fuel cell applications. J. Memb. Sci. 279: 548–557. Zaidi, S.M.J., Mikhailenko, S.D., Robertson, G.P., Guiver, M.D. and Kaliaguine, S. 2000. Proton conducting composite membranes from polyether ether ketone and heteropolyacids for fuel cell applications. J. Memb. Sci. 173: 17. Zhang, H., Li, X., Zhao, C., Fu, T., Shi, Y. and Na, H. 2008. Composite membrane based on highly sulfonated PEEK and PBI: Morphology characteristics and performance. J. Memb. Sci. 308: 66–74. Zhong, S. Cui, X. Cai, H. Fu, T. Zhao, C. and Na, H. 2007. Crosslinked sulfonated poly(ether ether ketone) proton exchange membranes for direct methanol fuel cell applications. J. Power Sour. 164: 65. Zhou, X. Weston, J. Chalkova, E. Hofmann, M.A. Ambler, C.M. Allock, H.R. and Lvov, S.N. 2003. High temperature transport properties of polyphosphazene membranes for direct methanol fuel cells. Electrochem. Acta. 48: 2173–2180.

14

Nanofiltration Membrane in Textile Effluent Treatment Ahmad Fauzi Ismail and Woei Jye Lau

CONTENTS 14.1 Introduction ..................................................................................................449 14.2 Textile Industry and Its Efluent ................................................................... 450 14.3 Overview of the State-of-the-Art of NF Membrane Technology ................. 451 14.3.1 Membranes Prepared via Interfacial Polymerization Technique ..... 452 14.3.2 Membranes Prepared via a Single-Step Fabrication Process ........... 453 14.4 Advantages of NF Membrane in Textile Wastewater Treatment .................. 454 14.5 NF Membranes Fouling ................................................................................ 456 14.6 Methodologies to Overcome Limitations of NF Membranes ....................... 458 14.6.1 Membrane Surface Modiication ...................................................... 458 14.6.2 UF/MF as Pretreatment System .......................................................460 14.6.3 Variation in Process Conditions ....................................................... 461 14.6.4 Cleaning Process .............................................................................. 463 14.7 Case Studies ..................................................................................................464 14.7.1 Valencia, Spain .................................................................................464 14.7.2 Karur, India ...................................................................................... 467 14.7.3 Kasr Hellal, Tunisia ..........................................................................469 14.8 Conclusion .................................................................................................... 470 References .............................................................................................................. 471

14.1

INTRODUCTION

The textile industry is a worldwide water pollution source. Considering the enormous quantities of wastewater discharged and their compositions, the wastewater generated from dyeing and inishing operations is rated as one of the most polluting among all industrial sectors, raising environmental concerns all over the world about the elimination of color and salt from the wastewater. Due to the unique characteristics in the separation of small neutral and charged molecules from wastewater streams, the use of nanoiltration (NF) membranes in the treatment of textile wastewater has drawn much attention from membrane scientists over the last decade and has proven to be more suitable and reliable compared with the conventional treatment methods and other kinds of pressure-driven membrane processes. 449

450

Membrane Modification: Technology and Applications

Despite all the promising perspectives of NF, there is a drawback that slows down application on a large scale. The potential for NF in textile efluent treatment is noteworthy, but it is hindered by unstabilities in operation caused by membrane fouling. The negative impact of fouling is obvious as it limits water recovery rates, affecting separation eficiency as well as shortening membrane lifespan, leading to an increase in the operation cost. It should be emphasized that the fundamental understanding of the fouling phenomena resulting from different mechanisms, such as concentration polarization, dye absorption, and cake layer formation, is imperative, as it constitutes a key inluence on the feasibility of NF process application in the long run. As fouling is one of the main problems in NF membranes of textile wastewater treatment, one should be aware of the importance of fouling control methods in overcoming the limitations. Of all the practical solutions available, development of NF that is less-fouling-sensitive to dye absorption is the most sustainable one and has attracted considerable attention recently. Other possible solutions to solve fouling problems include pretreatment processes prior to NF, optimization in process conditions, and effectiveness of cleaning processes. This chapter will provide an overview on the latest development of NF fabrication technology and discuss briely the problems associated with the use of other treatment processes in the textile industry, emphasizing the main advantages of the NF process in removing dye components and dissolved salt. The decline in membrane performance over a period resulting from fouling is also reviewed and discussed to give a clear understanding of the mechanisms involved. This discussion is followed by the strategies that are available to prevent or reduce membrane fouling with an aim to develop a more sustainable and cost-effective NF. Finally, case studies of NF in textile industry conducted in different countries are also considered.

14.2

TEXTILE INDUSTRY AND ITS EFFLUENT

The textile industry is considered as one of the oldest and heaviest polluters in the world. At present, approximately half of the available water is being used for domestic purposes, and the other half is consumed by industrial and agricultural activities. In view of the increasing world population, more water will, therefore, be required for industrial processes in order to meet the need for more cloth production. In textile-reining processes, substantial amounts of water, mineral salts, and organic dyes are used for every kilogram of cotton processed. As a consequence, its daily operation generates a large amount of wastewater that contains complex contaminants. The characteristics of textile discharge vary largely depending on the kind of product being processed and chemicals involved in the particular textile factory (Hessel et al. 2007; Judd and Jefferson 2003; BTTG 1999). Therefore, identiication of the contaminant sources is a necessary step before any water-recycling technology is employed to eliminate the contaminants accordingly. Typical textile efluents contain many types of dyes, salts, detergents, solvents and, in some cases, heavy metals, depending on the particular textile process such as scouring, bleaching, dyeing, printing, or inishing. During the textile-dyeing process,

Nanofiltration Membrane in Textile Effluent Treatment

451

inorganic salt is added to the bath in order for the dye to fully penetrate the cloth and to provide the necessary dyeing action. The degree of dye ixation, however, depends on the type of iber (substrate) to be dyed, the shade of the dye, and the dyeing parameters. In general, reactive dyes demonstrate the poorest ixation rate compared with other dye classes, such as sulfur and direct dyes (Easton 1995; Othmer 1993). The technical and economic limitations of other dyes are the main reasons for making reactive dyes the most widely used dyes worldwide, though they demonstrate signiicant incomplete exhaustion during the dyeing process (Gatewood and Hall 1996). It is generally known that dyes can interfere with the light transmission in the receiving waters, causing interruption in the photosynthesis processes and endangering the aquatic life. A suggestion has been made to the effect that alternative dyestuffs should be used for dyeing processes; nevertheless, consumers, including those who complain about water pollution in rivers, demand a fastness level, brightness, and a shade gamut that can only be achieved by reactive dye. Furthermore, the use of salt at a high concentration level during dyeing processes, particularly in reactive dyeing of cotton, also raises the public awareness of the environmental impact of dyeing. Salt needs to be used during dyeing processes, but once the dyeing operation has taken place, a large amount of dissolved salt still remains in the water. This dissolved salt is now considered to be contaminating the water and requires to be removed. Otherwise, it will cause soil salinity problems in some areas and/or increase in groundwater salinity beyond the toxic limit. This, as a consequence, poses a serious danger to aquatic life due to increased osmotic pressure inside the organic cells. Not all textile-manufacturing plants use the same amount of chemicals during processing; therefore, the strategies to treat the water can differ for each plant. In general, water is treated either to meet the stringent water discharge regulations or for the purpose of reuse. It must be emphasized that wastewater treatment systems are not just processes to cope with environmental problems by minimizing the waste volume discharged, but they also recover valuable rinsing water, reducing the overall operation costs.

14.3

OVERVIEW OF THE STATE-OF-THE-ART OF NF MEMBRANE TECHNOLOGY

The irst generation of the NF membranes can be traced back to the early 1970s when most of the membranes were made of cellulose acetate (CA) and other cellulose esters. These cellulose-based membranes, however, severely limited the range of industrial applications due to their poor chemical and biological resistances coupled with insuficient water permeation. This consequently resulted in the development of a second generation of noncellulosic NF–composite membranes made of polyamide (PA) and polyurea (PU) with the aim to improve water permeability and selectivity, together with better pH and solvent stability (Schafer et al. 2003). This section does not intend to provide an exhaustive review of all the NF membranes developed to date. It simply aims to give the latest development of NF membranes in the past decade. Attention is paid to the research and development of NF prepared from two different fabrication techniques.

452

14.3.1

Membrane Modification: Technology and Applications

MEMBRANES PREPARED VIA INTERFACIAL POLYMERIZATION TECHNIQUE

The commercially available thin-ilm composite–nanoiltration (TFC–NF) membranes were mainly prepared by forming a very thin PA-active layer on a porous support membrane via an interfacial polymerization (IP) technique. The technique offers outstanding advantages, as either the top selective layer or the bottom porous substrate of the membrane can be independently modiied and optimized to enhance the water permeation rate and solute rejection while offering an excellent mechanical strength and/or compression resistance. Research has been documented in the literature reporting the methodologies used to enhance the structural properties as well as the performances of TFC–NF membranes through creating an ionic interaction between the PA-active layer and the microporous substrate and/or enhancing the degree of cross-linking between the monomers on the active top PA layer (Lau et al. 2011; Abu Seman et al. 2010; Yong et al. 2006; Song et al. 2005; Oh et al. 2005; Schafer et al. 2003). Unlike the composite membrane in the plate and frame or spiral wound conigurations, composite hollow iber membrane is experiencing a very slow growth in application, primarily due to the differences in geometry and handling of solution low. Many attempts have been made to produce composite lat membranes using IP technique, but if this technique is applied directly to the outer surface of a hollow iber microporous membrane without modiication, it is very dificult to produce a composite membrane with a defect-free PA ilm. One of the earliest mentions of the TFC hollow iber membrane was disclosed in a US patent in 1990 (Tadros and Trehu 1990). In hollow iber membranes, it is realized that the ilms, formed by IP process, do not adhere properly to the support, and pinholes can appear as a result of the droplets of the amine solution, causing low rejection rate. To reduce the number of imperfections in the PA layer, Verssimo et al. (2005) applied the intermediate organic solvent between the aqueous amine solution and the acid chloride solution with the purpose of forming a ilm on the inner surface of the support iber without pinholes. The use of intermediate organic liquid is a possible solution to the problem of pinhole formation; however, it seems too complicated for coating the lumen side of the iber, making it impractical for large-scale production. In 2007, Yang et al. applied an easier method to remove the excess piperazine (PIP) solution and droplets by lushing nitrogen gas slowly through the lumen side of the ibers. Results showed that this modiied IP procedure is able to produce composite hollow iber membranes with better stability for long-term running, mainly due to the good compatibility between the PA-active layers and the substrate membranes. From the viewpoint of large-scale industrial production, the current technology of composite hollow iber membrane preparation is still far from maturity. It still does not compete with commercially available composite membranes of lat ilm and spiral wound format, mainly due to the nonreproducibility of the results of composite hollow iber. Thus, this area is one of the subjects deserving a better level of understanding to make full use of hollow iber membranes prepared via a composite approach. In general, the IP technique is quite sophisticated and laborious because two different modifying solutions are needed, and the resulting membrane performances are greatly dependent on the fabrication conditions (Mansourpanah et al. 2009;

Nanofiltration Membrane in Textile Effluent Treatment

453

Li et al. 2009; Zhang et al. 2006; Mohammad et al. 2005; Du and Zhao 2004; Jegal et al. 2002). Furthermore, chemical stability of the membrane is a major concern, as PA in the top active layer tends to undergo ring chlorination following chlorine exposure during iltration processes. The deterioration of the top active layer will result in a drastic decrease in salt rejection, which in turn increases the maintenance cost.

14.3.2

MEMBRANES PREPARED VIA A SINGLE-STEP FABRICATION PROCESS

Over the past few years, there have been a number of studies reported on the fabrication of charged asymmetric NF membranes through a single-step process (Lau and Ismail 2009a; Ismail and Lau 2009a; Ismail and Hassan 2004, 2006, 2007; Wang et al. 2007; Wang and Chung 2006; Bowen et al. 2001, 2002). These membranes have been successfully developed with the use of charged polymeric materials as the main membrane-forming materials or through the addition of a small amount of charged additive during dope preparation. The dissociation of chemical functional groups, for example, sulfonic acid (–SO3H) and carboxylic acid group (–COOH), on the top membrane surface is the reason why these membranes demonstrate a relatively high separation rate of salt. Wang and Chung (2006) developed charged hollow iber NF membranes directly from polybenzimidazole (PBI) for chromate removal due to the self-charging ability of PBI in aqueous solution. PBI becomes self-charged because the adjacent benzene ring delocalizes the proton of the imidazole group. This speciic characteristic has also prompted the use of PBI NF as forward osmosis membrane in order to achieve both high permeation lux and greater selectivity (Wang et al. 2007). In comparison to the IP technique, the preparation of NF membranes via a single-step fabrication process is relatively simple and involves less preparation parameters while being able to exhibit the positive features of the pure component. Bowen et al. (2001) have reported the effect of a small percent of sulfonated poly(ether ether) ketone (SPEEK) as an additive for improving the performances of polysulfone (PSf)/SPEEK-blend NF membranes. The increases in salt rejection and water lux with increasing SPEEK content were attributed to the increased membrane charge density and hydrophilicity of the resultant membranes, respectively. The enhancement of water lux of ultrailtration (UF) membranes upon addition of SPEEK was later reported by Arthanareeswaran et al. (2007a,b). It is thus agreed that by incorporating charged additives such as SPEEK into the dope solutions, the membranes will possess negative charges and become more hydrophilic. As SPEEK is only miscible with polymer (i.e., PSf and polyether sulfone, PES) in n-methyl-2-pyrrolidone (NMP) solvent over a small range of polymer weight ratios, the amount added should be carefully controlled in order to minimize the defects displayed on membrane morphology (Lau and Ismail 2009a; Bowen et al. 2001). In addition to blending the charged polymers with the dope solution, the introduction of ionic coating layers on a support membrane has also been increasingly drawing attention recently. He et al. (2008a) successfully developed PES NF membranes with SPEEK as the top coating layer. The SPEEK-coated membranes were found to have higher chemical resistance and higher ion exchange capacity compared with

454

Membrane Modification: Technology and Applications

other sulfonated polymers. The development of NF membranes with various properties is also possible using different degrees of SPEEK sulfonation, which results in different values of ion exchange capacity. Ba et al. (2010) absorbed negatively charged polymers such as polyacrylic acid (PAA) and polyvinyl sulfate (PVS) onto the membrane top surface and found that the strong interactions due to the electrostatic force between the absorbed charged polymers and the membrane surface can make the coating layer more stable for use in harsh operating environments. Although TFC membranes are widely used in today’s market, several signiicant advantages of asymmetric membranes have kept them competitive. With respect to the manufacturing cost, asymmetric membrane is relatively cheaper than composite membrane. Furthermore, the surface properties of asymmetric NF are also less sensitive to chlorine disinfectants, which minimize the chances of surface deterioration such as that encountered with composite membranes.

14.4

ADVANTAGES OF NF MEMBRANE IN TEXTILE WASTEWATER TREATMENT

To date, various technologies have been employed to recover water from textile efluents. In general, these treatment technologies for dye removal can be divided into several major groups such as coagulation–locculation, adsorption, oxidation, ion exchange, and electrochemical and membrane separation. Comprehensive reviews on the performances of these treatment technologies can be found (Lau and Ismail 2009b; Senthilkumar and Muthukumar 2007; dos Santos et al. 2007; Marmagne and Coste 1996). There are advantages and disadvantages associated with each of these techniques. Nevertheless, considering the capability of NF membranes in concentrating and removing both inorganic salts and soluble dyes at a promising rejection rate, NF membranes appear as an excellent candidate over other treatment technologies. In order to make NF membranes more practical in industrial applications, the major drawback—fouling—needs to be well understood and minimized as much as possible through the development of antifouling membrane properties and/or with the installation of pretreatment processes. A conventional coagulation–locculation process is no longer a suitable treatment method in color removal mainly due to the signiicant quantities of the sludge produced and sometimes low separation eficiency with respect to the class of dyes used. The coagulation–locculation process also suffers from the cost of treating the sludge and the increasing number of restrictions concerning its disposal. Power-activated carbon (PAC) is the most commonly used technique in textile wastewater treatment; however, its direct application is limited by two main factors, which are the high cost of activated carbon and its ineffectiveness in decolorization. Alternative methods such as ozonation and oxidation also encounter other challenges as the oxidant requirements are high and expensive, causing the methods to be economically unfeasible. Moreover, they may produce undesirable side products to the treatment process. With respect to pressure-driven membrane processes, NF has been successfully proven to be more suitable and reliable than other kinds of membranes in textile

455

Nanofiltration Membrane in Textile Effluent Treatment

efluent treatment processes. Instead of removing colloidal dyes, such as disperse and vat dyes, NF is more eficient at concentrating and rejecting soluble and ionic dyes such as acid, basic, direct, and reactive dyes. It is generally agreed that microiltration (MF) and UF membranes are not appropriate for removing soluble and ionic dye simply because they result in a great variation in the dye rejection rates (Akbari et al. 2007). On the other hand, the use of reverse osmosis (RO) membranes for dyeing efluent treatment processes has also been documented in the literature (Nataraj et al. 2009; Suksaroj et al. 2005). Due to the dense structure of the membrane properties, it is found that RO is excellent at achieving almost dye-free permeate solution but with the cost of high energy consumption. Furthermore, concentration polarization phenomenon is signiicant in RO membranes; therefore, high operating pressure is required to conduct the iltration process. These problems encountered by UF and RO can be potentially overcome by NF, which is recognized as having properties common to them. NF achieves a well-balanced performance between lux and selectivity as well as offers other advantages such as relatively low operation and maintenance costs. Although the initial cost of NF setup is comparatively higher, it is outweighed by the signiicant cost savings through the recovery of salts and reuse of permeate water. Nevertheless, unlike the salts, not all the dyes can be recovered and reused in textile processes. The retained reactive dyes, for instance, in the concentrated retentate stream, is not reusable, given the fact that the dyes are hydrolyzed after exhaustion in the dye bath, resulting in changes in their chemical constituents and characteristics, as shown in Figure 14.1 (Shu et al. 2005). From N=

NaO3SOCH2CH2O2S

Na SO 3

N

H2N HO SO

N

N=

NaO3SOCH2CH2O2S

3N

a

(a)

HOCH2CH2O2S

N=

N

SO 3Na

H2N HO HOCH2CH2O2S

N

N=

SO

3 Na

(b)

FIGURE 14.1 Chemical structure of (a) Reactive Black 5, MW 959, and (b) its hydrolyzed form, MW 755. (Reprinted from Shu, L., Waite, T.D., Bliss, P.J., Fane, A., and Jegatheesan, V., Desalination, 172, 235–243, 2005. With permission.)

456

Membrane Modification: Technology and Applications

an economic viewpoint, the retained concentrated reactive dye solutions therefore cannot contribute positively to cost savings.

14.5

NF MEMBRANES FOULING

Fouling phenomena are unavoidable in all types of pressure-driven membranes. These problems may become even worse in those industries characterized by extremely polluted efluents such as in the textile-manufacturing industries. Though numerous studies have reported the viability of NF membranes in producing water with enough quality for reuse, the signiicant decline in permeation lux resulting from fouling is still a major concern to many when NF is used for textile wastewater reclamation. Since textile factories utilize different recipes of dye chemical compounds in their daily production, the wastewater composition generated can extremely luctuate. This has led to the dificulty in identifying the compounds that contribute to membrane fouling. Comprehensive studies on the nature and interactions of solute–solute and solute–membrane are thus paramount because they constitute key factors for the feasibility of NF processes. In general, the presence of a high concentration of salts in textile industrial wastewater is the initial factor affecting the reduction in water permeability of NF membrane as compared with its pure water permeability. Results reported by Gomes et al. (2005) showed that the retention of salts exhibited much higher osmotic pressure (π) than the dye molecules. Osmotic pressure caused by ionic species is found to increase considerably with concentration, but dye osmotic pressure shows little increase with concentration and becomes independent of the concentration after 3 g/l. Similar results were reported by Van der Bruggen et al. (2001), in which the permeate lux remained constant with increasing dye concentration from 0 to 10 g/l. This indicates that the initial reduction in lux caused by osmotic pressure is mainly due to the ionic concentration used in the solution. Furthermore, it should be noted that the osmotic pressure generated is varied depending on the type of salt used during the dyeing process. A comparison was made between divalent salt (Na2SO4) and monovalent salt (NaCl) and found that at the same operating conditions, the average lux of the experiments with Na2SO4 solution was much lower than those of the experiments with NaCl owing to the greater osmotic pressure resistance created (Schafer et al. 2003). Considering the better quality of sulfate salts in dyeing application, these kinds of salts are preferably used rather than monovalent salts, though the costs are relatively higher (Van der Bruggen et al. 2004). Despite the fact that dye concentrations do not contribute greatly to the osmotic pressure, a lux decline due to dye adsorption on the membrane surface and inside the pores can occur following an increase in the adsorption resistance (Rads) as a result of increasing dye concentration. Nevertheless, due to dye aggregation mechanisms in a highly concentrated dye solution, a saturated value of Rads is expected to be attained (Zollinger 1991). Normally, dye adsorption is indicated by the presence of color on the membrane after iltration and usually takes place based on the speciic interactions, such as electrostatic, var der Waals force and hydrogen bonds, between the functional groups of the dye molecules and the composite NF membrane (Gomes et al. 2005; Wang et al. 2002; Van der Brugen and Vandecasteele 2001). Similar

457

Nanofiltration Membrane in Textile Effluent Treatment

+H

(a)

3N–polyamide–COOH

+ DyeSO3–

Polyamide fiber

–

DyeSO3 + H3N–polyamide–COOH

Dye

H2N–polyamide–COO– + Dye+

H2N–polyamide–COO− + Dye

(b)

FIGURE 14.2 Reaction schemes between polyamide iber and textile dyes based on the isoelectric point (pH 4.2), (a) below the isoelectric point, and (b) above the isoelectric point. (Reprinted from Akbari, A., Remigy, J.C., and Aptel, P., Chem. Eng. Process., 41, 601–609, 2002. With permission.)

indings were also reported in the work of Akbari et al. (2002). They elaborated that due to the similarity of the chemical functionality present at the iber surface (fabric) and NF PA membrane surface, it is highly probable that reaction can occur between the textile dyes and PA iber. Figure 14.2 illustrates the interaction between PA and the dye compounds under different feed pH. As can be seen, the charge property of PA can be varied by adjusting the pH value, and either positive or negative groups of PA are able to react with dye molecules, causing possible heavy membrane fouling in a long process run. In addition to the osmotic pressure generated by the high concentration of salt used and dye adsorption, cake formation and pore blocking also contribute to membrane lux deterioration. Mo et al. (2008) observed that all class of dyes exhibited the potential to form a cake layer on top of the membrane, with thickness and hardness depending on the dye molecules used. Complete separation of dye compounds from dye aqueous solutions was possible, but it increased the degree of lux decline, owing to the accumulated dye cake layer over operation time, as shown in Figure 14.3. As soon as the cake starts to form, the cake layer will dominate

Cake layer

Acc.V 10.0 kV

Spot 2.0

Magn 2000x

Det SE

WD 7.8

Yellow 8 4hr

10 µm

FIGURE 14.3 Formation of dye cake layer on the surface of NF–PA composite membrane after 4 h of operation using aqueous dye solution containing 1000 ppm Direct Yellow 8. (Reprinted from Mo, J.H., Lee, Y.H., Kim, J., Jeong, J.Y., and Jegal, J., Dyes Pigm., 76, 429– 434, 2008. With permission.)

458

Membrane Modification: Technology and Applications

the transport process and effectively takes over the role of the membrane if there is no cleaning process employed. Depending on the particle size, other fouling mechanisms such as pore blocking and pore narrowing can also happen. For dye particles much larger than the membrane pores, it will lead to the formation of cake, whereas dye particles of size smaller than pores may deposit on the pore walls, restricting the pore size with pore narrowing. For those solutes of size similar to the membrane pores, porosity reduction may happen following membrane pore blocking. As NF membrane possesses pore sizes in the near-nanometer or subnanometer range, it is generally perceived that dye molecules are less likely to permeate through the membrane, leading to solute accumulation on the membrane surface. These retained solutes deposit gradually and the cake thickness increases with the process time. Ismail and Lau (2009b) related the color stained on the membrane surface with the cake formation and claimed that reversibility of cake formation is strongly dependent on the NF membrane properties. With increased membrane surface charge density, it was observed that cake formation caused by dye deposition was reversible, without permanent particle penetration into the membrane pore structure.

14.6 METHODOLOGIES TO OVERCOME LIMITATIONS OF NF MEMBRANES Through the appropriate ways of fouling control, it is possible to reduce/minimize the membrane fouling propensity to a certain extent and enhance the permeate yield. These strategies in general can be divided into two main categories in which the irst category consists of the use of antifouling resistant membrane, adequate feed pretreatment, and optimized operational conditions. The second category involves membrane remediation through cleaning processes.

14.6.1

MEMBRANE SURFACE MODIFICATION

Among all the techniques used to overcome membrane limitations, development of a less-fouling-sensitive membrane is the most sustainable solution to minimize the fouling tendency of NF during a treatment process. Hydrophilic groups and/or charged functional groups can be introduced onto membrane surface via coating, blending, or grafting with the aim of producing a membrane capable of withstanding the varying chemical composition of textile efluents and exhibiting less fouling sensitivity through decreased dye deposition. Recent results showed that membrane fouling caused by reactive dye accumulation was alleviated after the introduction of a small percentage of charged SPEEK to PES membrane (Ismail and Lau 2009b). Figure 14.4 shows the outer surface of membranes blended with different SPEEK concentrations after a chemical cleaning process. Prior to the cleaning process, all these membranes were used for treating a synthetic dyeing solution that consisted of Reactive Black 5 (RB 5), polyvinyl alcohol (PVA), NaCl, and Na2SO4. With the introduction of negatively charged sulfonic acid groups by SPEEK, it is interesting to notice that the development of the

Nanofiltration Membrane in Textile Effluent Treatment

459

(a)

(b)

(c)

FIGURE 14.4 Direct observation of outer surfaces of PES NF hollow iber membranes after chemical cleaning, (a) pure PES membrane, (b) PES membrane blended with 2 wt% SPEEK, and (c) PES membrane blended with 4 wt% SPEEK. (Reprinted from Ismail, A.F. and Lau, W.J., Desalin. Water Treat., 6, 281–288, 2009b. With permission.)

membrane with surface charge similar to that of the dye foulants may help mitigate the fouling and shows promising potential in industrial applications. In view of different charge properties of textile dyes, Akbari et al. (2006) developed NF membranes with surface charge identical with that of dye molecules using UV-irradiation technique. Negatively charged and positively charged NF membranes were prepared using monomers of sodium p-styrene sulfonate (NaSS) and [2-(acryloyloxy)-ethyl] trimethyl ammonium chloride (AC), respectively. Through these surface modiications, it was found that membranes can be successfully used for the treatment of dye efluents, containing either anionic dyes or cationic dyes, limiting the osmotic pressure and/or fouling of the membrane and providing stable water permeation. Similar observations were also documented in the work of Buonomenna et al. (2007), in which positively charged polyvinylidene luoride (PVDF) membranes developed through a plasma surface modiication technique showed complete retention of the positively charged dye due to the strong repulsion forces between the surface-modiied membrane and the charged components. In addition to modifying the surface charge properties, the surface morphology (e.g., roughness) and hydrophilicity of the membrane can also be modiied to become less susceptible to dye adsorption. The afinity of organic components to NF membranes can cause fouling of hydrophobic materials. Thus, blending of original polymers with hydrophilic polymers is likely to improve the hydrophilic characteristics of the developed membrane, minimizing the fouling tendency. To study the effects of surface roughness on dye adsorption, Yu et al. (2010) recently developed two types of NF membranes through the phase inversion process and IP technique. It was found that the composite membrane displayed signiicantly higher surface roughness as compared with the asymmetric CA membrane and is more likely to promote dye adsorption on the membrane surfaces, leading to a low permeate lux, especially at high concentrations of dye. The distinct high roughness of the composite membrane

460

Membrane Modification: Technology and Applications

is due to the presence of interfacially polymerized aromatic PA at the membrane surface (Petersen 1993).

14.6.2 UF/MF AS PRETREATMENT SYSTEM Many different integrated treatment processes have been investigated for color and salt removal, all of which are capable of minimizing membrane fouling to a certain degree. Recently, the utilization of microporous membrane prior to the NF iltration process has been attracting great attention among membrane scientists in textile wastewater treatment (Fersi et al. 2009; Uzal et al. 2009; GozalvezZafrilla et al. 2008; Capar et al. 2006a,b; Barredo-Damas et al. 2006; Marcucci et al. 2003). This pretreatment removes the process damage due to larger particles in the feed stream and reduces membrane fouling. In addition to fouling reduction, other signiicant advantages offered by the pretreatment membrane process include the recovery of unspent auxiliary chemicals (e.g., sizing agent—PVA), high-molecular-weight insoluble dyes (e.g., indigo, disperse), and caustic compounds (Porter 1990; Majewska et al. 1989; Gaddis et al. 1989). In addition, water is reused for certain feed processes such as rinsing and washing when salinity is not a major problem (Rott and Minke 1999). The hybrid system using UF/MF membrane technology can be lexibly adjusted with respect to the system operating conditions to accommodate changes in the feed stream. Membrane–membrane hybrid systems are recognized as treatment processes of high eficiency, which can be competitive in comparison to traditional methods of wastewater treatment. Preliminary removal of ine suspended solids and colloids from textile efluents is fundamental to prevent severe fouling and module damage of downstream NF, which in turn guarantees a good and constant performance of the NF system. Fersi and coworkers (2009) investigated the eficiency of UF and NF membranes against fouling resistance for the treatment of industrial efluent using a resistance-in-series model. Results indicated that lux decline in UF and NF is mainly caused by the concentration polarization, as well as reversible and irreversible fouling. The fouling phenomenon in NF could be the result of the cake layer formation and/or pore blocking. Without suitable pretreatment, the situation is expected to become worse due to the cake compression with a decrease in cake porosity against operation time. Most recently, Alcaina-Miranda et al. (2009) highlighted the effectiveness of UF as a pretreatment stage in controlling NF fouling. They reported that the permeate lux of two commercial NF membranes remained practically constant during the studied period. The small percent difference between the initial clean water lux and the wastewater lux in NF could be explained by the fact that particles that caused fouling were eficiently removed in the previous UF stage. On the other hand, with the use of MF followed by NF, Marcucci et al. (2003) observed that signiicant lux decline was only recorded in the NF system after 60 h of operation. The NF lux was reported to decrease from ~500 l/h to ~450 l/h and to ~400 l/h at 80 and 100 h, respectively. However, the decline in water lux of the NF was able to be recovered through a simple chemical cleaning procedure. After cleaning, a lux that was close to the initial value of the water lux was

Nanofiltration Membrane in Textile Effluent Treatment

461

reported, indicating that fouling was not too severe and reversible with the use of MF at the pretreatment stage. In certain cases, the adsorption of acid dye onto particulates could result in the easy separation of dye molecules even with the use of the MF process. This would offer signiicant beneit in fouling minimization at the NF stage. According to Capar et al. (2006b), separation using an MF membrane with a pore size of 0.2 μm is good enough in eliminating color and turbidity of speciic acid dye bath wastewater but does not reduce the parameters associated with chemical oxygen demand (COD) and total solids. Van der Bruggen et al. (2004) also proposed the use of MF membrane for an integrated wastewater treatment system. They reported that inorganic MF membranes, particularly sintered stainless-steel membranes, are a robust pretreatment process and advantageous in view of their extremely high porosity ranging between 65% and 85%. However, in another study, Capar et al. (2006a) reported that UF pretreatment processes using membranes with either tight or loose structure are not effective in improving the lux decline of an NF process and recommended that single-stage NF is suficient in water reclamation of printing efluents. These contradictory indings may be due to the characteristics of the samples used in the work of Capar et al. (2006b), which had relatively less intense color, with a high ixation degree of acid dyes, and/or difference in the evaluation technique used in determining the lux decline. The evaluation technique used by Capar et al. (2006b) in the determination of lux decline is dependent on the volume reduction factor (VRF), which is different compared with the technique of other researchers who evaluated lux decline as a function of operation time. Briely, most researchers agreed with the realistic solution offered by microporous membrane as a pretreatment process in an integrated approach for the effective long-term treatment of textile waste streams. The solution addresses the fouling problem by enhancing the lifespan of the NF membrane without compromising its selectivity and permeability.

14.6.3 VARIATION IN PROCESS CONDITIONS Optimized process conditions are found to be effective in the reduction of fouling caused by concentration polarization. This problem has found some resolution in the process arrangement by lowing the feed liquid parallel to the membrane surface rather than perpendicular to it. Previous studies showed that the hydrodynamics of cross-low velocity (CFV) is the crucial operational characteristic in controlling the concentration polarization phenomenon as cross-low removes deposited material and limits the thickness of the boundary layer adjacent to the membrane surface. In view of the volume and quality of the efluent discharged from textile industries, the cross-low iltration method has been one of the most important developments in the iltration industry as it offers signiicant advantage in reducing the need for frequent membrane cleaning (He et al. 2008b; Petrinic et al. 2007; Al-Amoudi and Lovitt 2007; Koyuncu 2002; Akbari et al. 2002). In addition to minimizing the instabilities in the permeate yield, Akbari et al. (2002) observed that greater dye rejection is obtained when the stirring velocity in the membranes induces turbulence. However, the increased rejection rates were insigniicant for the dye solutions containing direct red and disperse blue. This was

462

Membrane Modification: Technology and Applications

because excellent color removal (>99%) could be easily obtained under conditions without stirring, mainly due to the insolubility and partial solubility of these dyes in water. On the other hand, Yu et al. (2010) observed that the permeate lux and dye removal of both types of PA–NF and CA–NF membranes are enhanced with increasing CFV from 0.1 to 1.1 m/sec, as shown in Figure 14.5. This revealed that the effect of CFV limiting dye adsorption and concentration polarization is pronounced with the enhancement of the equilibrium permeate lux and dye removal of NF membranes. However, no research has reported the optimized CFV in the NF treatment process as it is strongly dependent on module and system design.

Equilibrium permeate flux (1/m2 h)

40

36

32

28 24 20 0.1

PA–NF CA–NF 0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

1.1

Cross-flow velocity (m/s)

(a) 100.0

Dye removal rate (%)

99.8

99.6

99.4 PA–NF CA–NF 99.2 0.1

(b)

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0 1.1

Cross-flow velocity (m/s)

FIGURE 14.5 (a) Equilibrium lux and (b) dye removal rate of PA–NF (◾) and CA–NF (•) membranes at different CFV. Test conditions employed were initial water permeability = 50 l/m2 h, RB 5 concentration = 1000 mg/l, NaCl concentration = 10,000 mg/l, and pH 6.8 ± 0.2. (Reprinted from Yu, S., Liu, M., Ma, M., Qi, M., Lü, Z., and Gao, C., J. Memb. Sci., 350, 83–91, 2010. With permission.)

Nanofiltration Membrane in Textile Effluent Treatment

463

To further reduce the concentration polarization in NF of different module designs, the concept of bubble sparging is applicable (Verberk and van Dijk 2006; Cui et al. 2003). Bubble motion generates transient shear patterns at the vicinity of the membrane feed side and improves the mass transfer of molecules, which in turn increases the water lux. Compared with an NF system with only water cross-low, the energy consumption of an air-sparged NF system is comparatively lower because air bubbles are able to partially replace water low, creating an additional turbulence in liquid slugs. Interestingly, in addition to enhanced water lux, Verberk and van Dijk (2006) also observed an increase in salt rejection rate when air was injected into water low. They attributed the pronounced increase in rejection to a decrease in the value of concentration polarization. With the use of high-resolution optical microscopy, Pal et al. (2006) found that there is a direct consequence of dye particles moving inside the membrane and depositing on the membrane surface as a result of increases in the operating pressure. Appreciable increases in penetration inside the membrane surface and deposition over the membrane surface were reported with increasing convective forces, causing a lux reduction over time. With increasing feed pressure of the dye solution from 0.75 to 2.5 MPa, Yu et al. (2010) observed a decrease in permeate lux of NF membranes and attributed the results to the increase in the concentration polarization and dye adsorption on the membrane surface. Undeniably, increasing pressure promotes water molecule permeation through the membrane, but it also results in solute particles moving toward the membrane surface, making the deposited layer more compact. Therefore, it must be pointed out here that the operating pressure should be carefully optimized to lessen the impacts of dye penetration/deposition on the water lux. The solute concentration cannot be modiied by the experimenter during the iltration process as efluent is discharged directly from the process plant of the industry. It is, however, still important to understand the impact of solute concentration on membrane permeability and separation performances. Theoretically, a highly concentrated solution would cause obvious lux reduction in the membrane system, mainly due to signiicant concentration polarization. According to Yu et al. (2010), the electrostatic repulsive interaction between the composite NF and the charged solutes is diminished with increasing concentration of feed dye, leading to decreased water permeability. At this point, the charge property of composite NF is no longer responsible for fouling reduction as the increased feed dye concentration promotes dye adsorption on the membrane surface. In addition to decreasing water lux, electrolyte retention of charged NF membranes was also found to decrease with increasing salt concentration (Ismail and Lau 2009b; Van der Bruggen et al. 2001). This decrease in separation eficiency can be explained by the fact that salt concentration plays an important role in reducing the electrostatic repulsion of charged NF.

14.6.4 CLEANING PROCESS There are a wide variety of cleaning mixtures and protocols recommended by either membrane scientists or membrane manufacturers in the open literature, but none of them are versatile for the range of industrial applications. The membrane-cleaning

464

Membrane Modification: Technology and Applications

protocol that is applied in principle is dependent on the properties, design and process, and feed characteristics of the membrane. The cost should also be taken into account for cleaning chemicals and energy consumption during the cleaning process. Generally, chemical cleaning is the most commonly used strategy in NF membrane processing textile colored wastewater as physical cleaning has been found not suitable for membranes of dense structure (Ismail and Lau 2009b; Qin et al. 2007; Capar et al. 2006c; Sungpet et al. 2004; Marcucci et al. 2003). In certain cases, this approach has the ability to completely recover the initial membrane permeability and requires relatively less energy consumption compared with physical methods. Table 14.1 presents various chemical cleaning strategies used by membrane scientists in NF for textile wastewater treatment and the outcomes after the cleaning procedure. Acid washing is typically used for the removal of inorganic polluting agents (scaling) while alkaline cleaning is for the removal of organic polluting agents (fouling). However, a thorough study on cleaning agents is required because these agents may damage particular membrane properties, shortening the membrane lifetime. It appears from the literature that not much work has been published on this speciic area and it is worthy of further investigation. The success of chemical cleaning methods also depends on many other factors such as pH, temperature, concentration of cleaning solution, duration of cleaning and its frequency, and operation conditions such as pressure and CFV. A comprehensive review on these factors affecting the cleaning eficiency of NF can be found in the work of Al-Amoudi and Lovitt (2007). In summary, it is observed that proper cleaning procedure is critical to the performance of all membrane systems. No chemical cleaning procedure can avoid causing performance decline as a result of degrading membrane properties, but the impacts of cleaning can be minimized by using well-chosen cleaning agents through a trial-and-error approach (Van der Bruggen et al. 2008).

14.7

CASE STUDIES

In this section, three case studies conducted in different countries are considered. In practice, the huge diversity in the textile industrial processes prevents any generalization of any single membrane treatment unit with respect to the limitations of the individual unit. The choice of the most effective and reliable integrated treatment processes depends on the dyestuffs and dyeing methods used during production as well as the quantities of efluents generated from operation. Central to discussion is the issue of the eficiency and reliability of the NF, which is incorporated with other unit processes to produce puriied water for reuse. Its reuse nevertheless strongly depends on the inal quality of the water, which will be a function of textile production process.

14.7.1

VALENCIA, SPAIN

Recently, Gozalvez-Zafrilla et al. (2008) investigated the performance of commercial TFC–NF membranes as a secondary treatment process after an industrial wastewater plant (i.e., biological process) of a textile factory situated in Valencia, Spain.

465

Nanofiltration Membrane in Textile Effluent Treatment

TABLE 14.1 Various Chemical Cleaning Strategies Applied to NF Membrane Treatment for Textile Effluent NF Membrane Properties In-house made PES/SPEEK membranes with pore radius 20.0

6.3 0.66

8.2 >20.00

7.6 >20.00

7.8 >20.00

13,770 213 702 192

15,396 — — 163

364 (97.6) 14 26 3 (98.2)

29,356 600 1,535 317

28,594 (2.6) 180 (70.0) 351 (77.1) 107 (66.2)

54,734 1,320 3,290 518

115

154

2 (98.7)

307

88 (71.3)

499

1,419 5,715 3,900 84 97 122

1,486 6,268 4,146 86 98 194

38 (97.4) 166 (97.4) 140 (96.6) 10 (88.4) 95 36

2,254 12,443 13,000 230 98 317

154 (93.17) 14,931 12,000 (7.7) 300 98 229

5,644 22,674 22,000 415 98 419

a

Source: Data from Ranganathan, K., Karunagaran, K., and Sharma, D.C. Resour. Conserv. Recycl., 50, 306–318, 2007. With permission. a The values within brackets represent the rejection rate (%) of certain parameters.

overcome the drawback in membrane iltration for wastewater treatment, with the minimization of the level of secondary pollution in the retentate stream. These indings showed that the EC process might be another eficient pretreatment method for the NF processing of textile wastewater, provided the optimal conditions of the EC process are satisied.

14.8

CONCLUSION

The process of wastewater reclamation and reuse has become an attractive option to fully utilize discharged efluent and minimize fresh water consumption in the textile industry. Though many methods have been extensively researched and reported in the literature to deal with textile efluent, the use of the green technology of NF membrane emerges as an outstanding candidate with a relatively small carbon footprint. The effort devoted to NF membrane development over the past four decades has resulted in tremendous advances in membrane performance with respect to permeability, selectivity, and chemical and thermal resistances. As a consequence, there is a great potential for using membrane-based iltration technologies

471

Nanofiltration Membrane in Textile Effluent Treatment

TABLE 14.4 Characteristics of Textile Effluent and Treated Waters Using Hybrid EC/NF Process Parameter Conductivity (μS/cm) Turbidity (NTU) COD (mg/l) bColor Cl− (mg/l) SO42− (mg/l) Ca2+ (mg/l) K+ (mg/l) Mg2+ (mg/l) Na+ (mg/l)

a

Effluent Treated by EC/NF

Untreated Effluent

VRF = 1.09

VRF = 1.32

VRF = 1.66

15,190 252 4,800 1,480 3,740 3,503 599 236 162 3,518

883 0.481 80 1.94 189 0 333 12 0 353

1405 0.145 51 6.49 268 0 203 9 0 452

1872 0.237 240 1.05 358 0 60 18 0 644

Source: Data from Aouni, A., Fersi, C., Ali, M.B.S., and Dhahbi, M., J. Hazard. Mater., 168, 868–874, 2009. With permission. a The hybrid process was operated in optimum conditions of EC (current density: 40 mA/cm2; electrolysis time: 60 min) and at pressure of 10 bar in NF process. b Integral of the absorbance curve in the whole visible range (400–800 nm).

for wastewater treatment. As original textile efluent is characterized as color-intensive and highly saline in addition to the presence of various components, control of membrane fouling will undoubtedly remain a high-priority research. Among the fouling control strategies available, the development of less-fouling-sensitive membrane materials might offer interesting sustainable alternatives. Given the increasing stringency of environmental legislation worldwide, the poor availability of good water sources, and continuing decreased membrane-manufacturing costs through technological innovation, it is inevitable that the use of NF membranes in textile industries will become more widespread.

REFERENCES Abu Seman, M.N., Khayet, M. and Hilal, N. 2010. Nanoiltration thin-ilm composite polyester polyethersulfone-based membranes prepared by interfacial polymerization. J. Memb. Sci. 348: 109–116. Akbari, A., Remigy, J.C. and Aptel, P. 2002. Treatment of textile dye efluent using polyamidebased nanoiltration membrane. Chem. Eng. Process. 41: 601–609. Akbari, A., Desclaux, S., Rouch, J.C., Aptel, P. and Remigy, J.C. 2006. New UV-photografted nanoiltration membranes for the treatment of colored textile dye efluents. J. Memb. Sci. 286: 342–350. Akbari, A., Desclaux, S., Rouch, J.C. and Remigy, J.C. 2007. Application of nanoiltration hollow ibre membranes developed by photografting to treatment of anionic dye solutions. J. Memb. Sci. 297: 243–252.

472

Membrane Modification: Technology and Applications

Al-Amoudi, A. and Lovitt, R.W. 2007. Fouling strategies and the cleaning system of NF membranes and factors affecting cleaning eficiency. J. Memb. Sci. 303: 4–28. Alcaina-Miranda, M.I., Barredo-Damas, S., Des-Pia, A., Iborra-Clar, M.I., Iborra-Clar, A. and Mendoza-Roca, J.A. 2009. Nanoiltration as a inal step towards textile wastewater reclamation. Desalination 240: 290–297. Aouni, A., Fersi, C., Ali, M.B.S. and Dhahbi, M. 2009. Treatment of textile wastewater by a hybrid electrocoagulation/nanoiltration process. J. Hazard. Mater. 168: 868–874. Arthanareeswaran, G., Mohan, D. and Raajenthiren, M. 2007a. Preparation and performance of polysulfone-sulfonated poly (ether ether ketone) blend ultrailtration membranes. Part I. Appl. Surf. Sci. 253: 8705–8712. Arthanareeswaran, G., Thanikaivelan, P., Jaya, N., Mohan, D. and Raajenthiren, M. 2007b. Removal of chromium from aqueous solution using cellulose acetate and sulfonated poly (ether ether ketone) blend ultrailtration membranes. J. Hazard. Mater. B. 139: 44–49. Ba, C., Ladner, D.A. and Economy, J. 2010. Using polyelectrolyte coatings to improve fouling resistance of a positively charged nanoiltration membrane. J. Memb. Sci. 347: 250–259. Barredo-Damas, S., Alcaina-Miranda, M.I., Iborra-Clar, M.I., Bes-Pia, A., Mendoza-Roca, J.A. and Iborra-Clar, A. 2006. Study of the UF process as pretreatment of NF membranes for textile wastewater reuse. Desalination 200: 745–747. Bowen, W.R., Doneva, T.A. and Yin, H.B. 2001. Polysulfone-sulfonated poly(ether ether) ketone blend membranes: Systematic synthesis and characterization. J. Memb. Sci. 181: 253–263. Bowen, W.R., Doneva, T.A. and Yin, H.B. 2002. Separation of humic acid from a model surface water with PSU/SPEEK blend UF/NF membranes. J. Memb. Sci. 206: 417–429. BTTG. 1999. Report 5: Waste Minimization and Best Practice. British Textile Technology Group. Buonomenna, M.G., Lopez, L.C., Favia, P., d’Agostino, R., Gordano, A. and Drioli, E. 2007. New PVDF membranes: The effect of plasma surface modiication on retention in nanoiltration of aqueous solution containing organic compounds. Water Res. 41: 4309–4316. Capar, G., Yetis, U. and Yilmaz, L. 2006a. Reclamation of printing efluents of a carpet manufacturing industry by membrane processes. J. Memb. Sci. 277: 120–128. Capar, G., Yilmaz, L. and Yetis, U. 2006b. Membrane based strategies for the pre-treatment of acid dye bath wastewaters. J. Hazard. Mater. B 135: 423–430. Capar, G., Yilmaz, L. and Yetis, U. 2006c. Reclamation of acid dye bath wastewater: Effect of pH on nanoiltration performance. J. Memb. Sci. 281: 560–569. Cui, Z.F., Chang, S. and Fane, A.G. 2003. The use of gas bubbling to enhance membrane processes – A review. J. Memb. Sci. 221: 1–35. dos Santos, A.B., Cervantes, F.J. and van Lier, J.B. 2007. Review paper on current technologies for decolourisation of textile wastewaters: Perspectives for anaerobic technology. Bioresour. Technol. 98: 2369–2385. Du, R.H. and Zhao, J.S. 2004. Properties of poly (N,N-dimethylaminoethyl methacrylate)/ polysulfone positively charged composite nanoiltration membrane. J. Memb. Sci. 239: 83–188. Easton, J.R. 1995. The dye maker’s view. In: Colour in Dyehouse Efluent, ed. P. Cooper, pp. 6–21. The Alden Press: Oxford. Fersi, C., Gzara, L. and Dhahbi, M. 2009. Flux decline study for textile wastewater treatment by membrane processes. Desalination 244: 321–332. Gaddis, J., Spencer, H. and Jernigan, D. 1989. Caustic recovery and recycling at a textile dyeing and inishing plant. Natl. Res. Counc. Can. Report NRCC 29895. Adv. Reverse Osmosis and Ultrailtration, pp. 347–355. Gatewood, B.M. and Hall, J. 1996. Evaluation of after treatments for reusing reactive dyes. Text. Chem. Color. 28: 38–42.

Nanofiltration Membrane in Textile Effluent Treatment

473

Gomes, A.C., Goncalves, I.C. and de Pinho, M.N. 2005. The role of adsorption on nanoiltration of azo dye. J. Memb. Sci. 255: 157–165. Gozalvez-Zafrilla, J.M., Sanz-Escribano, D., Lora-Garcia, J. and Leon Hidalgo, M.C. 2008. Nanoiltration of secondary efluent for wastewater reuse in the textile industry. Desalination 222: 272–279. He, T., Frank, H., Mulder, M.H.V. and Wessling, M. 2008a. Preparation and characterization of nanoiltration membranes by coating polyethersulfone hollow ibers with sulfonated poly(ether ether ketone) (SPEEK). J. Memb. Sci. 307: 62–72. He, Y., Li, G., Wang, H., Zhao, J. and Huang, Q. 2008b. Effects of operating conditions on separation performance of reactive dye solution with membrane process. J. Memb. Sci. 321: 183–189. Hessel, C., Allegre, C., Maisseu, M., Charbit, F. and Moulin, P. 2007. Guidelines and legislation for dye house efluents. J. Environ. Manage. 83: 171–180. Ismail, A.F. and Hassan, A.R. 2004. The deduction of ine structural details of asymmetric nanoiltration membranes using theoretical models. J. Memb. Sci. 231: 25–36. Ismail, A.F. and Hassan, A.R. 2006. Formation and characterization of asymmetric nanoiltration membrane: Effect of shear rate and polymer concentration. J. Memb. Sci. 270: 57–72. Ismail, A.F. and Hassan, A.R. 2007. Effect of additive contents on the performances and structural properties of asymmetric polyethersulfone (PES) nanoiltration membranes. Sep. Purif. Technol. 55: 98–109. Ismail, A.F. and Lau, W.J. 2009a. Theoretical studies on structural and electrical properties of PES/SPEEK blend nanoiltration membrane. AIChE J. 55: 2081–2093. Ismail, A.F. and Lau, W.J. 2009b. Inluence of feed conditions on the rejection of salt and dye in aqueous solution by different characteristics of hollow iber nanoiltration membranes. Desalin. Water Treat. 6: 281–288. Jegal, J., Min, S.G. and Lee, K.H. 2002. Factors affecting the interfacial polymerization of polyamide layers for the formation of polyamide composite membranes. J. Appl. Polym. Sci. 86: 2781–2787. Judd, S. and Jefferson, B. 2003. Membranes for Industrial Wastewater Recovery and Reuse. Elsevier: Oxford. Koyuncu, I. 2002. Reactive dye removal in dye/salt mixtures by nanoiltration membranes containing vinylsulphone dyes: Effects of feed concentration and cross low velocity. Desalination 143: 243–253. Lau, W.J. and Ismail, A.F. 2009a. Theoretical studies on the morphological and electrical properties of blended PES/SPEEK nanoiltration membranes using different sulfonation degree of SPEEK. J. Memb. Sci. 334: 30–42. Lau, W.J. and Ismail, A.F. 2009b. Polymeric nanoiltration membrane for textile dyeing wastewater treatment: Preparation, performance evaluation, transport modelling, and fouling control – A review. Desalination 245: 321–348. Lau, W.J., Ismail, A.F., Misdan, N. and Kassim, M.A. 2011. A recent progress in thin ilm composite membrane: A review. Desalination doi:10.1016/j.desal.2011.04.004. Li, L., Zhang, S. and Zhang, X. 2009. Preparation and characterization of poly(piperazineamide) composite nanoiltration membrane by interfacial polymerization of 3,3′,5,5′-biphenyl tetraacyl chloride and piperazine. J. Memb. Sci. 335: 133–139. Majewska, N.K., Winnicki, T. and Wisniewski, J. 1989. Effects of low conditions on ultrailtration eficiency of dye solutions and textile efluents. Desalination 71: 127–135. Mansourpanah, Y., Madaeni, S.S. and Rahimpour, A. 2009. Fabrication and development of interfacial polymerized thin-ilm composite nanoiltration membrane using different surfactants in organic phase; study of morphology and performance. J. Memb. Sci. 343: 219–228. Marcucci, M., Ciabatti, I., Matteucci, A. and Vernaglione, G. 2003. Membrane technologies applied to textile wastewater treatment. Ann. N.Y. Acad. Sci. 984: 53–64.

474

Membrane Modification: Technology and Applications

Marmagne, O. and Coste, C. 1996. Color removal from textile plant efluents. Am. Dyest. Rep. 85: 15–21 Mo, J.H., Lee, Y.H., Kim, J., Jeong, J.Y. and Jegal, J. 2008. Treatment of dye aqueous solutions using nanoiltration polyamide composite membranes for the dye wastewater reuse. Dyes Pigm. 76: 429–434. Mohammad, A.W., Hilal, N. and Seman, M.N.A. 2005. Interfacially polymerized nanoiltration membranes: Atomic force microscopy and salts rejection studies. J. Appl. Polym. Sci. 96: 605–612. Nataraj, S.K., Hosamani, K.M. and Aminabhavi, T.M. 2009. Nanoiltration and reverse osmosis thin ilm composite membrane module for the removal of dye and salts from the simulated mixtures. Desalination 249: 12–17. Norit Group. 2010. India expanding opportunities in the water sector. Knowhow 11: 20–21. Oh, N.W., Jegal, J. and Lee, K.H. 2005. Preparation and characterization of nanoiltration composite membranes using polyacrylonitrile (PAN). II. Preparation and characterization of polyamide composite membranes. J. Appl. Polym. Sci. 80: 2729–2736. Othmer, K. 1993. Encyclopedia of Chemical Technology, Vol. 8. John Wiley & Sons: New York. Pal, S., Ghosha, A., Ghosh, T.B., De, S. and DasGupta, S. 2006. Optical quantiication of fouling during nanoiltration of dyes. Sep. Purif. Technol. 52: 372–379. Petersen, R.J. 1993. Composite reverse osmosis and nanoiltration membranes. J. Memb. Sci. 83: 81–150. Petrinic, I., Andersen, N.P.R., Sostor-Turk, S. and Marechal, A.M.L. 2007. The removal of reactive dye printing compounds using nanoiltration. Dyes Pigm. 74: 512–518. Porter, J.J. 1990. Membrane iltration techniques used for recovery of dyes, chemicals and energy. Text. Chem. Color. 22: 21–25. Qin, J.J., Oo, M.H. and Kekre, K.A. 2007. Nanoiltration for recovering wastewater from a speciic dyeing facility. Sep. Purif. Technol. 56: 199–203. Ranganathan, K., Karunagaran, K. and Sharma, D.C. 2007. Recycling of wastewaters of textile dyeing industries using advanced treatment technology and cost analysis – Case studies. Resour. Conserv. Recycl. 50: 306–318. Rott, U. and Minke, R. 1999. Overview of wastewater treatment and recycling in the textile processing industry. Water Sci. Technol. 40: 37–144. Schafer, A.I., Fane, A.G. and Waite, T.D. 2003. Nanoiltration: Principles and Applications. Elsevier: Britain. Senthilkumar, M. and Muthukumar, M. 2007. Studies on the possibility of recycling reactive dye bath efluent after decolouration using ozone. Dyes Pigm. 72: 251–255. Shu, L., Waite, T.D., Bliss, P.J., Fane, A. and Jegatheesan, V. 2005. Nanoiltration for the possible reuse of water and recovery of sodium chloride salt from textile efluent. Desalination 172: 235–243. Song, Y.J., Liu, F. and Sun, B.H. 2005. Preparation, characterization and application of thin ilm composite nanoiltration membranes. J. Appl. Polym. Sci. 95: 1251–1261. Suksaroj, C., Heran, M., Allegre, C. and Persin, F. 2005. Treatment of textile plant efluent by nanoiltration and/or reverse osmosis for water reuse. Desalination 178: 333–341. Sungpet, A., Jiraratananon, R. and Luangsowan, P. 2004. Treatment of efluents from textilerinsing operations by thermally stable nanoiltration membranes. Desalination 160: 75–81. Tadros, S.E. and Trehu, Y.M. 1990. Coating process for composite reverse osmosis membranes. U.S. Patent No 4980061. Uzal, N., Yilmaz, L. and Yetis, U. 2009. Microiltration/ultrailtration as pretreatment for reclamation of rinsing waters of indigo dyeing. Desalination 240: 198–208. Van der Brugen, B. and Vandecasteele, C. 2001. Flux decline during nanoiltration of organic components in aqueous solution. Environ. Sci. Technol. 35: 3535–3540.

Nanofiltration Membrane in Textile Effluent Treatment

475

Van der Bruggen, B., Daems, B., Wilms, D. and Vandecasteele, C. 2001. Mechanisms of retention and lux decline for the nanoiltration of dye baths from the textile industry. Sep. Purif. Technol. 22–23: 519–528. Van der Bruggen, B., Curcio, E. and Drioli, E. 2004. Process intensiication in the textile industry: The role of membrane technology. J. Environ. Manage. 73: 267–274. Van der Bruggen, B., Manttari, M. and Nystrom, M. 2008. Drawbacks of applying nanoiltration and how to avoid them: A review. Sep. Purif. Technol. 63: 251–263. Verberk, J.Q.J.C. and van Dijk, J.C. 2006. Air sparging in capillary nanoiltration. J. Memb. Sci. 284: 339–351. Verssimo, S., Peinemann, K.V. and Bordado, J. 2005. New composite hollow iber membrane for nanoiltration. Desalination 184: 1–11. Wang, K.Y. and Chung, T.S. 2006. Fabrication of polybenzimidazole (PBI) nanoiltration hollow iber membranes for removal of chromate. J. Memb. Sci. 281: 307–315. Wang, X.L., Wang, W.N. and Wang, D.X. 2002. Experimental investigation on separation performance of nanoiltration membranes for inorganic electrolyte solutions. Desalination 145: 115–122. Wang, K.Y., Chung, T.S. and Qin, J.J. 2007. Polybenzimidazole (PBI) nanoiltration hollow iber membranes applied in forward osmosis process. J. Memb. Sci. 300: 6–12. Yang, F., Zhang, S., Yang, D. and Jian, X. 2007. Preparation and characterization of polypiperazine amide/PPESK hollow iber composite nanoiltration membrane. J. Memb. Sci. 301: 85–92. Yong, Z., Yu, S., Liu, M. and Gao, C. 2006. Polyamide thin ilm composite membrane prepared from m-phenylenediamine and m-phenylenediamine-5-sulfonic acid. J. Memb. Sci. 270: 162–168. Yu, S., Liu, M., Ma, M., Qi, M., Lü, Z. and Gao, C. 2010. Impacts of membrane properties on reactive dye removal from dye/salt mixtures by asymmetric cellulose acetate and composite polyamide nanoiltration membranes. J. Memb. Sci. 350: 83–91. Zhang, Y.F., Xiao, C.F., Liu, E.H., Du, Q.Y., Wang, X. and Yu, H.L. 2006. Investigations on the structures and performances of a polypiperazine amide/polysulfone composite membrane. Desalination 191: 291–295. Zollinger, H. 1991. Color Chemistry, 2nd edn. VCH: Weinheim.

15

Future Prospects Nidal Hilal, Mohamed Khayet, and Chris J. Wright

In the preceding chapters, we have demonstrated how important membrane modiication is for the optimization of membrane separation processes. The versatility of the membranes required for different processes and applications has meant that the technology for membrane modiication must be equally versatile to improve the membrane performance. Hence, the irst part of the book presents a comprehensive review to identify the possibilities of modiication processes that can be exploited to meet industrial challenges. The second part of the book, which discusses the application of membrane modiication technologies, has identiied the strategies that can be pursued for industrial implementation of a membrane separation. We hope that this will serve as an encouragement to the readers who have the challenge of optimizing membrane separation within their area of industry. Looking to the future, we are very optimistic that the adoption of membrane technologies in different areas will continue to grow. Thus, the importance for membrane modiication strategies will also increase as existing industrial membrane systems require optimization for new processes and membranes are tailored for new industrial sectors. As can be seen throughout this book, an integral part of membrane optimization that guides the modiication strategy is the issue of membrane fouling. It would be naive to believe that through polymer chemistry, a membrane that solves all the problems of membrane fouling would be developed or a single modiication process can be developed to tailor bespoke membranes for each application within separation processes. Thus, the future prospect for membrane modiication will be in the form of established research strategies. New membrane processes require a lead in time prior to industrial large-scale implementation. This lead in time consists of a research protocol of laboratory-scale and then pilot-scale studies, which optimize the separation process through consideration of the materials and economics involved. Prior knowledge from experience and literature will guide the choice of membrane material and the appropriate modiication process for a given application. As prior knowledge improves, as exempliied in this book, the lead in time will be reduced. However, the often large expense of commissioning membrane plant will mean prudent continued use of experimental laboratory-scale studies to monitor the outcomes of membrane modiication. We will now look to the separate chapters of the book to identify the future prospects of the speciic areas of membrane modiication and its application. In Chapter 1, membrane modiication in the nuclear industry is discussed. In this challenging industry where the membrane feeds dictate a high degree of physical and chemical robustness of the membrane, it is clear that a lot of research and development is still 477

478

Membrane Modification: Technology and Applications

required. However, there have been some recent improvements in the use of ceramic membranes where chemisorbed alcohol on ceramic membrane improved separation with a high stability against acids, down to pH 1. This will prove to be a useful modiication that can be further applied to the wastes originating from nuclear fuel reprocessing. Another promising area is the use of nanostructured materials and nanoparticles embedded in the membrane matrix. Indeed, this is emerging as a very useful modiication process in other membrane systems and applications. The use of impedance spectroscopy (IS) in the characterization of modiied membranes is introduced in Chapter 2. IS is a nondestructive technique for electrical characterization of solid and liquid systems. IS allows the determination of the electrical properties of heterogeneous systems formed by a series array of layers with different electrical and/or structural properties such as membrane/electrolyte systems. This permits a separate evaluation of the electrical contribution of each layer by using impedance plots and equivalent circuits as models, where the different circuit elements are related to the structural/transport properties of the systems. Chapter 2 demonstrates the “current state of the art” and how it has been applied to membrane systems to provide information on the processes related to the interface, external electrolyte solution layer, and bulk membrane by analyzing the impedance plots and using equivalent circuits as models. IS measurements are associated with membrane material changes to provide a greater understanding of the modiication process. IS is one of many advanced characterization techniques that have been applied to the monitoring of membrane modiication. As with all these instruments, which seem to be constantly growing in capabilities, with improved sampling rates and greater experimental rigor, there will be a corresponding improvement in their application to membrane systems. For IS, this will provide a greater understanding of the electrical properties during membrane modiication with improved data resolution in both spatial and temporal terms. Chapter 3 presents a review of polymer modiication techniques that have been used for reducing membrane fouling. A range of surface modiication methods including photochemical and redox grafting, plasma treatment, physical coating with hydrophilic polymers, chemical reactions on the membrane surface, and immobilization of nanoparticles have been used over the last few decades to reduce fouling with different degrees of success. Thus, improvements on these techniques continue to be explored. Chapter 3 highlights that the UV/redox initiating graft polymerization and physical coating of membranes with hydrophilic polymer layers are promising techniques as they have the advantages of simplicity, low cost of operation, mild reaction conditions, and a possibility of incorporation into the end stages of a membrane-manufacturing process. Another issue identiied in Chapter 3 that needs to be resolved is the lack of studies on the estimation of antifouling properties of surface-modiied membranes with real multicomponent feed streams. Future research and development needs targeted tailoring of membrane properties via surface modiication in combination with optimization of operation conditions and feed stream pretreatment, to reduce membrane fouling and to render the membrane resistant to different types of fouling. This is essential for the achievement of effective and long-term membrane performance. Recent advances in nanotechnology that enable the improved production and control of nanoparticle fabrication have had and will have signiicant impact on

Future Prospects

479

membrane modiication technology. This has been identiied by many of the authors of this book and is the topic of Chapter 4. There will be a continued adoption of the use of nanoparticles for the improvement of membrane functionality and mechanical strength. However, the ratio of nanoparticles and polymeric materials requires optimization in order to produce more cost-competitive membranes with higher performance in the future. A current issue that is the focus of debate is the fate and toxicology of nanoparticles. The growing use of nanoparticles in novel materials has raised concerns. Nanotoxicology studies are yet to deinitively report on the potential of nanoparticles to cross natural barriers, such as cell membranes and those found in the lungs, to cause harm. Thus, care must be taken when using nanoparticles for membrane modiication. If nanoparticle membrane modiication is considered because of the advantages it offers, then the choice of material and its retainment in the membrane system must be investigated. For instance, silica nanoparticles are more suitable to be incorporated into membranes, which will be used in drinking water application because silica exhibits lower toxicity and is environmentally inert. Chapter 5 discusses the use of interfacial polymerization (IP) and photografting techniques to improve antifouling properties and performance of nanoiltration (NF) membranes. The UV-photografting technique is suggested to be superior to IP for membrane surface modiication because, for the membranes that have been tested, no substantial reduction of membrane permeability was detected. Among the UV-photografting–modiied membranes, N-vinylpyrrolidone–grafted membranes with UV at moderate levels and for a short irradiation time exhibit a better NF performance with higher permeability and low irreversible fouling at both neutral and acidic pH conditions. Such an approach should be adopted by other researchers looking to the UV-photografting technique to modify membranes for different applications. Membrane distillation (MD) is emerging as an important membrane separation process that has been investigated widely for many applications including seawater desalination, food processing, and removal of volatile organic compounds from water. To avoid membrane pore wetting, which is the principal problem of MD conigurations, membrane modiication will have an important part to play in its optimization for different applications. Chapter 6 focuses on the use of novel composite hydrophobic/hydrophilic membranes made speciically for MD and successfully tested for desalination by direct contact membrane distillation (DCMD). This was achieved through blending different types of surface-modifying macromolecules into hydrophilic polymers. The chapter demonstrates that for future modiication of MD membranes, it is important to consider the characteristics of the top skin layer (the hydrophobic layer), which highly inluence the DCMD permeate lux. These characteristics are, in particular, the liquid entry pressure of water, pore size, porosity, and the cross-sectional structure of the membrane. Plasma modiication of membrane surfaces discussed in Chapter 7 is a useful technique that will continue to be an essential part of many membrane modiication strategies. In a very short time, by using very small quantities of reactants, novel membranes can be prepared. By applying this method, it is possible to create a whole range of new membranes using only one type of porous substrate. The bulk properties of the membrane that infer advantage to the process remain

480

Membrane Modification: Technology and Applications

unchanged, while the membrane surfaces can be designed and optimized for a speciic application. Another set of techniques, reviewed in Chapter 8, that have been proven to be a useful part of the membrane modiication tool box are electrospun nanoibers, nanostructured membranes, and their modiication. The properties of electrospun polymer nanoibers, which include very large surface area to volume ratio, excellent mechanical properties, and a highly open porous structure, make them ideal candidates for membrane materials. The surface chemical and physical properties of the nanoiber membrane play a crucial role in their application. However, most of the polymer nanoibers do not possess the required speciic functional groups and therefore must be functionalized for successful applications. Thus, as their application grows, the use of different novel modifying techniques is necessary. Surface modiication techniques of nanoibers include plasma treatment, wet chemical methods, surface graft polymerization, layer-by-layer assembly, and co-electrospinning of surface-active agents and polymers. At this point, we should identify that as in many ields of polymer chemistry, the application of “smart” materials will become smarter. This will also be the case for membrane surfaces, which have the potential to undertake multiple tasks beyond separation that could include chemical transformation, chemical delivery, and process monitoring. Nanoibers and their modiication to improve their application described in Chapter 8 are a good example. To increase nanoiber functionality, a variety of bioactive molecules including anticancer drugs, enzymes, cytokines, and polysaccharides as well as nanoparticles have been entrapped within the interior or physically immobilized on the surfaces of nanoibers. The surfaces of electrospun nanoibers have also been modiied by immobilizing cell-speciic bioactive ligands to enhance cell adhesion and cell proliferation by mimicking the morphology and biological function of the extracellular matrix. Such approaches will also be developed in other membrane systems and processes. Chapter 9 demonstrates how membrane modiication based on the incorporation of inorganic/organic particles and multilayer membranes has a promising future in pervaporation (PV). This separation process continues to emerge as a feasible separation technology for many different and diverse applications from puriication of solvents to wastewater treatment. As a proven method of separation at low temperatures and pressures, further development and modiication of membranes are likely to be undertaken to improve PV application in food processing. In addition, the modiication of PV membranes to improve the removal of a variety of organic compounds to clean wastewater streams also holds much promise. Many membrane separation applications have already beneitted from membrane modiication strategies. Chapter 10 describes how bespoke polymeric membranes have been used to improve the crystallization of biomolecules. Membrane crystallization allows through a careful control of the process parameters the production of crystals with controlled shape, size, size distribution, and polymorphism. Further research is required to provide comprehensive understanding of the complex relationships between membrane process parameters and crystal structure. The control of product polymorphism will continue to be important in the pharmaceutical industry, which, as the range of drugs and their speciicity increase, will require improved

Future Prospects

481

prediction and control of crystallization. The speciicity and high resolution required by biologically active materials mean that membrane modiication to improve the range of nanostructured hydrophobic membranes to support crystallization operations will be essential. Chapter 11 discusses how chemical cross-linking has been used to improve the application of polymeric membranes in gas separation applications. This chapter elegantly demonstrates how for future application of membrane modiication processes, such as chemical cross-linking, prior knowledge is important. Selecting the most suitable cross-linking or modiication technique can be dificult, but a detailed literature review together with good knowledge of the application of the cross-linked membrane will narrow the cross-linking method and polymer selection signiicantly. For gas separation, further research is required to maintain gas permeability after the cross-linking modiication, in order to make this modiication technique attractive for the preparation of membranes for gas separations. Chapters 12 and 13 concentrate on fuel cell applications. Chapter 12 demonstrates how grafting with high-energy radiation is a convenient method to prepare polymer electrolyte membranes for improved fuel cell operation. The use of simultaneous electron beam irradiation was found to simplify the process and reduce the reaction time together with a reduction in the monomer consumption. The use of the preirradiation method in a single-step mode allows quicker preparation of polymer electrolyte membranes with improved properties and reduced cost. However, the chemical stability of the prepared membranes was found to be an issue and therefore an important future development of fuel cell application is an improved strategy for high-energy radiation grafting. Chapter 13 presents an extensive study of the development of polymeric membranes based on Naion and sulfonated poly (ether ether ketone) (SPEEK) for direct methanol fuel cell (DMFC). Both Naion and SPEEK membranes have been extensively modiied to increase their performance in DMFCs. The chapter demonstrates the potential of membrane modiication in this ield. In particular, SPEEK blended with charged surface-modifying macromolecules has been found to possess a good potential in enhancing the proton conductivity while suppressing the methanol permeability. However, challenges in commercializing the SPEEK as a polyelectrolyte membrane still remain and the overall performance of the membrane as well as the current and power density of the DMFC itself have to be improved further. The book inally examines the use of modiied membranes to improve wastewater treatment in the key industry of textile manufacture. Chapter 14 discusses how the textile industry was an early adopter of membrane technology but still strives to improve membrane application through membrane modiication. NF is emerging as a dominant technology for the future of this industry as environmental legislation becomes ever more stringent, with poor availability of water sources, and NF membrane-manufacturing costs reduce through technological innovation. However, fouling continues to be a problem. A future priority for membrane modiication strategies in the textile industry is the development of fouling-resistant membranes and materials. To conclude, membrane processes have spread through a host of industries, so membrane modiication strategies have evolved to improve eficiencies based on

482

Membrane Modification: Technology and Applications

membrane productivity, quality of permeate, costs, energy consumption, and fouling resistance. These modiication strategies continue to evolve as technological advances occur in related areas such as polymer chemistry, nanoparticle production, bioactive composites, advanced characterization techniques, and online process monitoring. In addition, as economics of membrane manufacture and modiication improve through innovation, a larger number of processes will adopt the technology. The future appears bright for the membrane technologist.