Electrochemical Aspects of Ionic Liquids

ELECTROCHEMICAL ASPECTS OF IONIC LIQUIDS ffirs01.indd i 1/20/2011 10:29:12 AM ELECTROCHEMICAL ASPECTS OF IONIC LIQUI...

Author: Hiroyuki Ohno

99 downloads 1762 Views 36MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!

Report copyright / DMCA form

DOWNLOAD PDF

ELECTROCHEMICAL ASPECTS OF IONIC LIQUIDS

ffirs01.indd i

1/20/2011 10:29:12 AM

ELECTROCHEMICAL ASPECTS OF IONIC LIQUIDS Second Edition

Edited by HIROYUKI OHNO

A JOHN WILEY & SONS, INC., PUBLICATION

ffirs02.indd iii

1/20/2011 10:29:12 AM

Copyright © 2011 by John Wiley & Sons, Inc. All rights reserved Published by John Wiley & Sons, Inc., Hoboken, New Jersey Published simultaneously in Canada No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, (978) 750-8400, fax (978) 750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken, NJ 07030, (201) 748-6011, fax (201) 748-6008, or online at http://www.wiley.com/go/permission. Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts in preparing this book, they make no representations or warranties with respect to the accuracy or completeness of the contents of this book and specifically disclaim any implied warranties of merchantability or fitness for a particular purpose. No warranty may be created or extended by sales representatives or written sales materials. The advice and strategies contained herein may not be suitable for your situation. You should consult with a professional where appropriate. Neither the publisher nor author shall be liable for any loss of profit or any other commercial damages, including but not limited to special, incidental, consequential, or other damages. For general information on our other products and services or for technical support, please contact our Customer Care Department within the United States at (800) 762-2974, outside the United States at (317) 572-3993 or fax (317) 572-4002. Wiley also publishes its books in a variety of electronic formats. Some content that appears in print may not be available in electronic formats. For more information about Wiley products, visit our web site at www.wiley.com. Library of Congress Cataloging-in-Publication Data: Electrochemical aspects of ionic liquids / edited by Hiroyuki Ohno. p. cm. Includes bibliographical references and index. ISBN 978-0-470-64781-3 (cloth) 1. Ionic solutions. 2. Electrochemistry. 3. Polymerization. I. Ohno, Hiroyuki, 1953– QD562.I65E38 2011 541′.372–dc22 2010034796 Printed in Singapore. 10 9 8 7 6 5 4 3 2 1

ffirs03.indd iv

1/20/2011 10:29:13 AM

CONTENTS

PREFACE TO THE SECOND EDITION

ix

PREFACE TO THE FIRST EDITION

xi

ACKNOWLEDGMENTS FOR THE SECOND EDITION CONTRIBUTORS

1

Importance and Possibility of Ionic Liquids

xiii xv

1

Hiroyuki Ohno

2

Physical Chemistry of Ionic Liquids: Inorganic and Organic as Well as Protic and Aprotic

5

C. A. Angell, W. Xu, M. Yoshizawa-Fujita, A. Hayashi, J.-P. Belieres, P. Lucas., M. Videa, Z.-F. Zhao, K. Ueno, Y. Ansari, J. Thomson, and D. Gervasio

PART I 3

BASIC ELECTROCHEMISTRY

General Techniques

33 35

Yasushi Katayama

4

Electrochemical Windows of Room-Temperature Ionic Liquids (RTILs)

43

Hajime Matsumoto v

ftoc.indd v

1/20/2011 10:29:16 AM

vi

5

CONTENTS

Diffusion in Ionic Liquids and Correlation with Ionic Transport Behavior

65

Md. Abu Bin Hasan Susan, Akihiro Noda, and Masayoshi Watanabe

6

Ionic Conductivity

87

Hiroyuki Ohno, Masahiro Yoshizawa-Fujita, and Tomonobu Mizumo

7

Optical Waveguide Spectroscopy

95

Hiroyuki Ohno and Kyoko Fujita

8

Electrolytic Reactions

101

Toshio Fuchigami and Shinsuke Inagi

9

Electrodeposition of Metals in Ionic Liquids

129

Yasushi Katayama

PART II 10

BIOELECTROCHEMISTRY

Enzymatic Reactions

157 159

Noritaka Iwai and Tomoya Kitazume

11

Molecular Self-assembly in Ionic Liquids

169

Nobuo Kimizuka and Takuya Nakashima

12

Solubilization of Biomaterials into Ionic Liquids

183

Kyoko Fujita, Yukinobu Fukaya, and Hiroyuki Ohno

13

Redox Reaction of Proteins

193

Kyoko Fujita and Hiroyuki Ohno

PART III 14

IONIC DEVICES

Li Batteries

203 205

Hikari Sakaebe and Hajime Matsumoto

15

Photoelectrochemical Cells

221

Hajime Matsumoto

16

Fuel Cells

235

Masahiro Yoshizawa-Fujita and Hiroyuki Ohno

17

Double-Layer Capacitors

243

Makoto Ue

ftoc.indd vi

1/20/2011 10:29:16 AM

vii

CONTENTS

18

Actuators

271

Kinji Asaka

PART IV 19

FUNCTIONAL DESIGN

Novel Fluoroanion Salts

279 281

Rika Hagiwara and Kazuhiko Matsumoto

20

Neutralized Amines

293

Hiroyuki Ohno

21

Zwitterionic Liquids

301

Masahiro Yoshizawa-Fujita, Asako Narita, and Hiroyuki Ohno

22

Alkali Metal Ionic Liquids

317

Wataru Ogihara, Masahiro Yoshizawa-Fujita, and Hiroyuki Ohno

23

Polyether/Salt Hybrids

325

Tomonobu Mizumo and Hiroyuki Ohno

24

Electric Conductivity and Magnetic Ionic Liquids

337

Gunzi Saito

PART V

25

IONIC LIQUIDS IN ORDERED STRUCTURES

Ion Conduction in Organic Ionic Plastic Crystals

347 349

Maria Forsyth, Jennifer M. Pringle, and Douglas R. MacFarlane

26

Liquid Crystalline Ionic Liquids

375

Takashi Kato and Masafumi Yoshio

PART VI 27

GEL-TYPE POLYMER ELECTROLYTES

Ionic Liquid Gels

393 395

Kenji Hanabusa

28

Zwitterionic Liquid/Polymer Gels

403

Masahiro Yoshizawa-Fujita and Hiroyuki Ohno

29

Ionic Liquidized DNA

409

Naomi Nishimura and Hiroyuki Ohno

ftoc.indd vii

1/20/2011 10:29:16 AM

viii

CONTENTS

PART VII 30

POLYMERIZED IONIC LIQUIDS

Ion Conductive Polymers

417 419

Hiroyuki Ohno and Masahiro Yoshizawa-Fujita

31

Amphoteric Polymers

433

Hiroyuki Ohno, Masahiro Yoshizawa-Fujita, and Wataru Ogihara

32

Polymer Brushes

441

Masahiro Yoshizawa-Fujita and Hiroyuki Ohno

PART VIII CONCLUSION

457

33

459

Future Prospects Hiroyuki Ohno

ftoc.indd viii

APPENDIX: STRUCTURES OF ZWITTERIONS

463

INDEX

465

1/20/2011 10:29:16 AM

PREFACE TO THE SECOND EDITION

The first edition of this book was published in 2005 as the first book on the basic study and application of the ionic liquids for electrochemical aspects. At this time, there is increasing interest in ionic liquids as an electrolyte solution substituent. In particular, interests are focused on the safety of the organic ion conductive liquids. Despite the safety of ionic liquids, there is still hesitation in using these ionic liquids as an electrolyte solution. This might be caused by two major reasons, one is cost, and the other is the great possibility of the development of better ionic liquids. The former is actually important for industry, but it should also be a matter of demand. Larger demand lowers the price. The second reason is a bit serious, because there is always the possibility of finding or developing new and better ionic liquids. There should be a kind of hesitation in deciding on the industrial use of current ionic liquids, because no one can deny that there is the possibility that better ones will emerge. In any case, it should be most important to develop ionic liquids having sufficient properties for practical use. Understanding of the latest in ionic liquid science is important to provide motivation for researchers to use them. In the second edition, we considerably updated the content to catch up with the fast changes in ionic liquid science. Also, interesting new chapters have been added. In every chapter, we tried to add the latest information while keeping the number of pages as low as possible. It will be one of our great pleasures if readers find some interesting point regarding ionic liquid science that aids in their research. HIROYUKI OHNO

ix

fpref01.indd ix

1/20/2011 10:29:15 AM

PREFACE TO THE FIRST EDITION

This book introduces some basic and advanced studies on ionic liquids in the electrochemical field. Although ionic liquids are known by only a few scientists and engineers, their applications’ potential in future technologies is unlimited. There are already many reports of basic and applied studies of ionic liquids as reaction solvents, but the reaction solvent is not the only brilliant future of the ionic liquids. Electrochemistry has become a big field covering several key ideas such as energy, environment, nanotechnology, and analysis. It is hoped that the contributions on ionic liquids in this book will open other areas of study as well as to inspire future aspects in the electrochemical field. The applications of ionic liquids in this book have been narrowed to the latest results of electrochemistry. For this reason only the results on roomtemperature ionic liquids are presented, and not on high-temperature melts. The reader of this book should have some basic knowledge of electrochemistry. Those who are engaged in work or study of electrochemistry will get to know the great advantages of using ionic liquids. Some readers may find the functionally designed ionic liquids to be helpful in developing novel materials not only in electrochemistry but also in other scientific fields. This book covers a wide range of subjects involving electrochemistry. Subjects such as the solubilization of biomolecules may not seem to be necessary for electrochemistry concerning ionic liquids, but some readers will recognize the significance of solubility control of functional molecules in ionic liquids even in an electrochemical field. Many more examples and topics on ionic liquids as solvents have been summarized and published elsewhere, and the interested reader will benefit from studying the references that are provided at the end of each chapter. Hiroyuki Ohno xi

fpref02.indd xi

1/20/2011 10:29:15 AM

ACKNOWLEDGMENTS FOR THE SECOND EDITION

First of all, I would like to express my sincere thanks to all the contributors for the second edition. All authors kindly agreed to reuse their chapters and made an effort to put the latest information in every chapter. A new chapter has been added in the second edition for better reviewing in electrochemistry. Next an acknowledgment should be given to Dr. Naomi Nishimura of the Department of Biotechnology, Tokyo University of Agriculture and Technology. Naomi worked hard to help me to edit manuscripts. She was so systematic that there were no serious problems in the editing of the manuscript. Without her energetic contribution, this book would not be published by the due date. Finally I would like to thank Dr. Arza Seidel of John Wiley and Sons, Inc. for her kind support and encouragement. Hiroyuki Ohno

xiii

flast01.indd xiii

1/20/2011 10:29:13 AM

CONTRIBUTORS

C. Austen Angell, Department of Chemistry and Biochemistry, Arizona State University Younes Ansari, Department of Chemistry and Biochemistry, Arizona State University Kinji Asaka, National Institute of Advanced Industrial Science and Technology (AIST) Jean-Philippe Belieres, Department of Chemistry and Biochemistry, Arizona State University Maria Forsyth, Department of Materials Engineering, Monash University Toshio Fuchigami, Department of Electronic Chemistry, Tokyo Institute of Technology Kyoko Fujita, Department of Biotechnology, Tokyo University of Agriculture and Technology Masahiro Yoshizawa-Fujita, Department of Materials and Life Sciences, Sophia University Yukinobu Fukaya, Department of Biotechnology, Tokyo University of Agriculture and Technology Dominic Gervasio, Center for Applied Nanobioscience in the BioDesign Institute and School of Materials, Arizona State University Rika Hagiwara, Department of Fundamental Energy Science, Kyoto University xv

flast02.indd xv

1/20/2011 10:29:14 AM

xvi

CONTRIBUTORS

Kenji Hanabusa, Graduate School of Science and Technology, Shinshu University Akitoshi Hayashi, Department of Applied Chemistry, Graduate School of Engineering, Osaka Prefecture University Shinsuke Inagi, Department of Electronic Chemistry, Tokyo Institute of Technology Noritaka Iwai, Department of Bioengineering, Tokyo Institute of Technology Yasushi Katayama, Department of Applied Chemistry, Faculty of Science and Technology, Keio University Takashi Kato, Department of Chemistry and Biotechnology, School of Engineering, The University of Tokyo Nobuo Kimizuka, Department of Chemistry and Biochemistry, Graduate School of Engineering, Kyushu University Tomoya Kitazume, Department of Bioengineering, Tokyo Institute of Technology Pierre Lucas, Department of Chemistry and Biochemistry, Arizona State University Douglas R. MacFarlane, School of Chemistry, Monash University Hajime Matsumoto, Research Institute for Ubiquitous Energy Devices, National Institute of Advanced Industrial Science and Technology (AIST) Kazuhiko Matsumoto, Graduate School of Energy Science, Kyoto University Tomonobu Mizumo, Department of Applied Chemistry, Hiroshima University Takuya Nakashima, Graduate School of Materials Science, Nara Institute of Science and Technology Asako Narita, Department of Polymer Chemistry, Graduate School of Engineering, Kyoto University Naomi Nishimura, Department of Biotechnology, Tokyo University of Agriculture and Technology Akihiro Noda, Honda R & D Co., Ltd. Wataru Ogihara, Nissan Motor Co., Ltd. Hiroyuki Ohno, Department of Biotechnology, Tokyo University of Agriculture and Technology Jennifer M. Pringle, Department of Materials Engineering and School of Chemistry, Monash University Gunzi Saito, Research Institute, Meijo University

flast02.indd xvi

1/20/2011 10:29:14 AM

xvii

CONTRIBUTORS

Hikari Sakaebe, Research Institute for Ubiquitous Energy Devices, National Institute of Advanced Industrial Science and Technology (AIST) Md. Abu Bin Hasan Susan, Department of Chemistry, University of Dhaka Jeffery Thomson, Center for Applied Nanobioscience in the BioDesign Institute and School of Materials, Arizona State University Makoto Ue, Fellow, Mitsubishi Chemical Corporation Kazuhide Ueno, Department of Chemistry and Biochemistry, Arizona State University Marcelo Videa, Department of Chemistry and Biochemistry, Arizona State University Masayoshi Watanabe, Department Yokohama National University

of

Chemistry

and

Biotechnology,

Xu Wu, Department of Chemistry and Biochemistry, Arizona State University Masafumi Yoshio, Department of Chemistry and Biotechnology, School of Engineering, The University of Tokyo Zuofeng Zhao, Department of Chemistry and Biochemistry, Arizona State University

flast02.indd xvii

1/20/2011 10:29:14 AM

1 IMPORTANCE AND POSSIBILITY OF IONIC LIQUIDS Hiroyuki Ohno

1.1

IONIC LIQUIDS

Ionic liquids are salts with a very low melting temperature. Ionic liquids have been of great interest recently because of their unusual properties as liquids. Because these unique properties of ionic liquids have been mentioned in a few other books, we will not repeat them here but will summarize them in Table 1.1. Note that these are entirely different properties from those of ordinary molecular liquids. Also, every ionic liquid does not always show these properties. For electrochemical usage, the most important properties should be both nonvolatility and high ion conductivity. These are essentially the properties of advanced (and safe) electrolyte solutions that are critical to energy devices put in outdoor use. Safety is a more important issue than performance these days, and it has been taken into account in the materials developed for practical use. Thus, more developments in ionic liquids are expected in the future. The nonvolatile electrolyte solution will change the shape and performance of electronic and ionic devices. These devices will become safer and have longer operational lives. They are composed of organic ions, and these organic compounds have unlimited structural variations because of the easy preparation of many different components. So there are unlimited possibilities open to the new field of ionic liquids. The most compelling idea is that ionic liquids are “designable” or “fine-tunable.” Therefore, we can easily expect explosive developments in fields using these remarkable materials. Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

1

c01.indd 1

1/20/2011 10:26:35 AM

2

IMPORTANCE AND POSSIBILITY OF IONIC LIQUIDS

TABLE 1.1. Basic and Possible Characteristics of Organic Ionic Liquids Low melting point

• •

Nonvolatility

• •

Composed by ions

• •

Organic ions

• • •

1.2

Treated as liquid at ambient temperature Wide usable temperature range Thermal stability Flame retardancy High ion density High ion conductivity Various kinds of salts Designable Unlimited combinations

IMPORTANCE OF IONIC LIQUIDS

Ionic liquids are salts that melt at ambient temperature. The principles of physical chemistry involved in the great difference between solution properties of molecular solvents and molten salts have already been introduced and summarized in several books. Thousands of papers have already been published on their outstanding characteristics and effectiveness for a variety of fields. Thus, as mentioned, in this book, we take the most important point that these salts are composed of organic ions and explore the unlimited possibility of creating extraordinary materials using molten salts. Because ionic liquids are composed of only ions, they usually show very high ionic conductivity, nonvolatility, and flame retardancy. The organic liquids with both high ionic conductivity and flame retardancy are practical materials for use in electrochemistry. At the same time, the flame retardancy based on nonvolatility inherent in ion conductive liquids opens new possibilities in other fields as well. Because most energy devices can accidentally explode or ignite, for motor vehicles there is plenty of incentive to seek safe materials. Ionic liquids are being developed for energy devices. It is therefore important to have an understanding of the basic properties of these interesting materials. The ionic liquids are multipurpose materials, so there should be considerable (and unexpected) applications. In this book we, however, will not venture into too many other areas. Our concern will be to assess the possible uses of ionic liquids in electrochemistry and allied research areas.

1.3

POTENTIAL OF IONIC LIQUIDS

At present, most interest in ionic liquids is centered on the design of new solvents. Although the development of “new solvents” has led the development of possible applications for ionic liquids, there is more potential for development of electrochemical applications. Electrochemistry basically needs two materials: electroconductive materials and ion conductive materials. Ionic liquids open the possibility of improving

c01.indd 2

1/20/2011 10:26:35 AM

POTENTIAL OF IONIC LIQUIDS

3

ion conductive materials. The aqueous salt solution is one of the best electrolyte solutions for electrochemical studies. However, because water is volatile, it is impossible to use this at a wide temperature range or on a very small scale. Many other organic polar solvents have been used instead of water to prepare electrolyte solutions. They, however, have more or less the same drawback, depending on the characteristics. The material known to be a nonvolatile ion conductor is the polymer electrolyte. Polymers do not vaporize but decompose at higher temperatures; the vapor pressure at ambient temperature is zero. Polymer electrolytes are considered a top class of electrolytes except for the one drawback: relatively low ionic conductivity. One remarkable propertie of ionic liquids is the proton conduction at a temperature higher than 100 °C. Water-based proton conductors cannot be operated at such a high temperature because of vaporization of water. As mentioned in a later chapter, proton-conductive ionic liquids are the most expected materials. Some literature has included statements that the ionic liquids are thermally stable and never decompose. This kind of statement has led to a misunderstanding that the ionic liquids are never vaporized and are stable even when on fire. Are the ionic liquids indestructable? The answer is “no.” However, although inorganic salts are entirely stable, the thermal stability of organic salts depends largely on their structure. Because ionic liquids are organic compounds, their degradation begins at the weakest covalent bond by heating. Nevertheless, ionic liquids are stable enough at temperatures of 200 °C to 300 °C. This upper limit is high enough for ordinary use. Does it need more energy or cost to decompose ionic liquids after finishing their role? It is not difficult to design novel ionic liquids that can be decomposed at a certain temperature or by a certain trigger. It also is possible to design unique catalysts (or catalytic systems) that can decompose target ionic liquids. Some catalysts such as metal oxides or metal complexes have the potential to become excellent catalysts for the decomposition of certain ionic liquids under mild conditions. The post-treatment technologies of ionic liquids should therefore be developed along with the work on the design of ionic liquids. At the present time there has been little progress in this area. Although post-treatment technologies are beyond the scope of this book, we do attempt to give ideas on the various future developments in ionic liquid technologies as well as in electrochemistry. This book is dedicated to introducing, analyzing, and discussing ionic liquids as nonvolatile and highly ion conductive electrolyte solutions. The astute reader will find the future prospects for ionic liquids between the lines in all chapters of this book.

c01.indd 3

1/20/2011 10:26:35 AM

2 PHYSICAL CHEMISTRY OF IONIC LIQUIDS: INORGANIC AND ORGANIC AS WELL AS PROTIC AND APROTIC C. A. Angell, W. Xu, M. Yoshizawa-Fujita, A. Hayashi, J.-P. Belieres, P. Lucas, M. Videa, Z.-F. Zhao, K. Ueno, Y. Ansari, J. Thomson, and D. Gervasio

2.1

CLASSES OF IONIC LIQUIDS

Ionic liquids in their high-temperature manifestations (liquid oxides, silicates, and salts) have been studied for a long time, using sophisticated methods, and much of the physics is understood. By contrast, the low-temperature ionic liquid (IL) field (<100 °C ILs), the subject of the present volume, is still under development. The many interesting studies on transport and thermodynamic properties of the <100 °C ILs have focused mainly on characterizing new systems for potential applications [1–5]. The task of placing this behavior within the wider phenomenology of liquid and amorphous solid electrolytes as well as in the context of the liquid state in general still has a long way to go. In this chapter, we review the current state of knowledge of physical properties of ionic liquids in an attempt to place them within this larger picture. We make an effort to emphasize the special status of the protic subclass of ionic liquids because these offer a degree of freedom not encountered in other branches of the solvent-free liquid state. The first requirement of an ionic liquid is that, contrary to experience with most liquids consisting of ions, it must have a melting point that is not much Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

5

c02.indd 5

1/20/2011 10:26:43 AM

6

PHYSICAL CHEMISTRY OF IONIC LIQUIDS

higher than room temperature. The limit commonly suggested is 100 °C [1b]. Given the cohesive energy of ionic liquids (about which more will be said later on), ambient melting requires that the melting point occur at a temperature not too much higher than the glass transition temperature, Tg, which provides the natural base for liquid-like behavior. Ionic liquids nearly all melt within the range that we call the “low-temperature regime” of liquid behavior [6,7]. This means that in most cases, they will supercool readily and will exhibit “super-Arrhenius” transport behavior near and below ambient temperature— as is nearly always reported. Such liquids come in different classes. The most heavily researched class is the aprotic organic cation class [1–4,8–15]. In this cation class, the low melting point is a consequence of the problem of efficiently packing large, irregular organic cations with small inorganic anions. More on this class is given in Section 2.3. A second class [16] is one that may enjoy increased interest in the future because of the presence of one of its members in the first industrial IL process [1b], because of the new finding that its members can have aqueous solutionlike conductivities [17] and can serve as novel electrolytes for fuel cells [18], and finally, because of the evidence that these liquids, in hydrated form, can be used as tunable solvents for biomolecules, on which stability against aggregation and hydrolysis may be provided under the right tuning [19]. This class is closely related to the first but differs in that the cation has been formed by transfer of a proton from a Brønsted acid to a Brønsted base. The process is reversible if the free energy of proton transfer is not too large. When the gap across which the proton must jump to reform the original molecular liquid is small, the liquid will have a low conductivity and a high vapor pressure. These properties are not of great interest in an ionic liquid, although the liquid may be fluid. If the gap is large, as in the case of ammonium nitrate, 87 kJ/mol (from data for HNO3 + NH3 → NH4NO3), then the proton will remain largely on the cation, and for many purposes, the system is a molten salt. If the acid is a strong acid like triflic acid, HSO3CF3, or a superacid like HTFSI [16], then the transfer of the proton will be energetic, and the original acid will not be regenerated on heating before the organic cation decomposes. Such liquids will not be easily distinguishable in properties from the conventional aprotic salts in which some alkyl group, rather than a proton, has been transferred to the basic site. This is particularly true of the protic ionic liquids (PILs) recently reported by Luo et al. [20] using superbases as proton acceptors. The stability of these systems has been characterized in terms of the relation between the boiling point elevation (or excess boiling point) over the linear (or average value) of the components [21], and the excess was shown to be a linear function of the difference in pKa values determined in aqueous solutions. This relation is shown in Figure 2.1. It seems to be free of exceptions when the base is a simple amine nitrogen. The protic ionic liquids as a class [16–18] are considerably more fluid than the aprotic ionic liquids [17], most likely because of their generally reduced ionicity (see Section 2.6).

c02.indd 6

1/20/2011 10:26:43 AM

7

CLASSES OF IONIC LIQUIDS 500 450

1:1 (mol) Base-Acid ILs

400

Estimated value PA-Tf

350

ΔTb / K

300

αPic-Tf

250 200 150

EA-FA

100

PA-FA αPic-TFAc

50 0

PA-TFAc

αPic-FA 0

10

Δ pKa

20

30

Figure 2.1. Correlation of the excess boiling point (determined at the 1 : 1 composition) with the difference in aqueous solution pKa values for the component Brønsted acids and bases of the respective ionic liquids ΔpKa. The ΔTb value is determined as the difference between the measured boiling point and the value, at 1 : 1, of the linear connection between pure acid and pure base boiling points. Note the large excess boiling points extrapolated for the ionic liquids formed from the superacid HTf (open triangles). These values could not be determined experimentally because of prior decomposition. (Notation: EA = ethylammonium, PA = propylammonium, αPic = α-picolinium = 2-methylpyridinium = 2MPy, FA = formate, TFAc = trifluoroacetate, Tf = triflate = trifluoromethanesulfonate) (from Yoshizawa et al. [21]). Data for the three protic nitrates of [17] (ethylammonium nitrate, dimethylammonium nitrate, and methylammonium nitrate) fit precisely on this diagram.

The third and distinct class of ionic liquid is the one that consists entirely of inorganic entities. These are formed mostly as a consequence of the mismatch of large anions like tetrachloroaluminate or iodide with small cations like Li+. The eutectic in the system LiAlCl4–LiAlI4 system, for instance, lies at 65 °C, and the liquid is highly fluid—more fluid than most aprotic ionic liquids. The phase diagram is shown in Figure 2.2 [22]. Also in this group is the more viscous system containing silver and alkali halides [23], which exhibits an ambient temperature electrical conductivity, 10−1.4 S/cm, that is higher than that of any aprotic ionic liquid because of the highly decoupled state of the silver ions. Here, a protic subclass also may be important; in fact the first “ionic liquids” made were probably of this type, namely mixtures of ammonium salts like the NH4SCN and NH4NO3 eutectic recently shown to provide interesting

c02.indd 7

1/20/2011 10:26:43 AM

8

PHYSICAL CHEMISTRY OF IONIC LIQUIDS

250

Temperature (°C)

200

150

100 65 50

0 0 LiAlCl

20

40

60 Mol%

80

100 LiAlI4

Figure 2.2. Phase diagram of the system LiAlCl4 + LiAlI4 (from Lucas et al. [22]) showing ionic liquid domain (Tl < 100 °C) in the middiagram.

fuel cell electrolytes [24]. Hydrazinium nitrate is known to melt at 80 °C [25]. Waiting to be created here is a subclass in which the cations are derived from inorganic molecular entities, for instance, (PNCl2)3 or (SN)4. A protic salt of the latter is on record, but its melting point is unknown. We have obtained a sharply melting compound (PNCl2)3. HTFSI, from the equimolar melt of the components [26], but its conductivity is too low to class it as an ionic liquid (i.e., it is of low “ionicity” [27–31]). The nitrogen atoms of PNCl2 are known to be only weakly basic, so the salt, if formed with a stronger acid (e.g., HSbF6), although of higher ionicity, also would be of high “acidity” [32]; in fact, it would be a member of the class of superacidic ionic liquids [33] that we will describe subsequently. A fourth class may be considered, although it contains nonionic entities. This is the liquid state of various ionic solvates. In these systems, molecules usually thought of as solvent molecules are bound tightly to high field cations and have no solvent function. Such “molten solvates” have low vapor pressures at ambient temperatures and only boil at temperatures near 200 °C; for instance, LiZnBr4 · 3H2O, has Tb = 190 °C, whereas Tg = −120 °C [34]. To provide a better perspective on ionic liquids, we first make some observations on inorganic salts and the factors that make it possible to observe them as ionic liquids below 100 °C.

c02.indd 8

1/20/2011 10:26:43 AM

9

LOW-TEMPERATURE LIQUID BEHAVIOR OF IONIC MELTS

2.2

LOW-TEMPERATURE LIQUID BEHAVIOR OF IONIC MELTS

Most inorganic salts, when they melt, are found to flow and conduct electricity according to a simple Arrhenius law at all temperatures down to their melting points. For instance, unless measurements of high precision are used, the alkali halides seem to obey the Arrhenius equation, even down to the deep eutectic temperatures of their mixtures with other salts. LiCl and KCl form a eutectic mixture with a freezing point of 351 °C, approximately 300 K below either pure salt freezing point; yet the viscosity of the melt barely departs from Arrhenius behavior before freezing. To see the viscosity behavior of the highly super-Arrhenius type typical of almost any individual room-temperature molten salt (RTMS), it is necessary to avoid alkali halides altogether and examine salts that cannot form such symmetrical crystal lattices. For instance, alkali nitrates, like KNO3, do not occupy much more volume in the liquid state than KI, but they melt at much lower temperatures. However, even KNO3 exhibits an Arrhenius-type viscosity temperature dependence according to any but the most precise measurements. It is only with deep eutectics like those for the ternary systems LiNO3–NaNO3–KNO3 (TE = 143 °C) and LiNO3–NaNO2–KNO3 (TE = 125 °C) that one starts to observe clear deviations from the Arrhenius law [6a,7]. This stands in clear contrast with the behavior of the ILs (RTMS) or molten hydrates. With all <100 °C ILs, deviations from the Arrhenius law, considerably in excess of those noted for the ternary LiNaK nitrate eutectic, are found well above their melting points. There is no need, with such ILs, to invoke eutectic mixtures to extend the stable liquid range to observe the “low-temperature domain” behavior. The low-temperature domain typically is found in the temperature range at T < 2T0 [6], where T0 is the theoretical low-temperature limit to the liquid state based on the Kauzmann extrapolation [35]. At T0, extrapolations of experimental data suggest that the excess entropy of the liquid (excess over that of the crystal) would vanish and that the time scale for fluid flow would diverge [7,35]. The practical low-temperature limit to the liquid state given by the glass temperature, Tg, is considerably higher than the theoretical T0, by an amount that depends on the liquid fragility. Typically Tg/T0 = 1.2–1.3, unless the liquid is “strong” [35b]. So the low-temperature domain is entered at Tg/T ≈ 0.53–0.66. For Arrhenius behavior, which represents the “strong” liquid limit of behavior is rarely observed, T0 = OK and Tg/T0 = ∞. The upper end of this range, Tg/T = 0.66, is the number usually associated with the ratio of Tg/Tm for glassformers (the “2/3 rule”), although we have argued elsewhere [36] that this is not a rule but a tautology. The behavior of ionic liquids is familiar to workers experienced with molten hydrates. With molten hydrates, the cation size is increased effectively by the shell of water molecules shielding the central cation from its anionic neighbors such that the cation acquires a size not unlike those of cations in the typical IL. We show a selection of viscosity data for normal molten salts, molten

c02.indd 9

1/20/2011 10:26:43 AM

10

PHYSICAL CHEMISTRY OF IONIC LIQUIDS

13 SiO2

11 9

GeO2 (810)

strong

NaAlSi3O8 (1095) CaAl2SiO8 (1134) propanol (98)

STRONG

intermediate

log(viscosity/poise)

ZnCl2 (380)

7

LiCl 5.8H2O

5

Ca(NO3)2 . 4H20

3

Ca(NO3)2.8H2O

CKN (338)

fragile

o-terphenyl (247)

1

MOMNMe2EBF4 BMIBF4

-1

BMIBOB

IONIC LIQUIDS (PRESENT WORK)

FRAGILE

-3 -5 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

Tg/T

Figure 2.3. Tg-scaled Arrhenius plot showing data for molten salts ZnCl2 and calcium potassium nitrate (CKN), with data for the calcium nitrate hydrate (CaNO3 · 8H2O) and the tetrafluoroborates of quaternary ammonium (MOMNM2E, M = methyl, E = ethyl) and 1-n-butyl-3-methyl-imidazolium (BMI) cations, and the bis-oxalatoborate (BOB) of the latter cation, in relation to other liquids of varying fragility (from Xu et al. [15]).

hydrates, and ionic liquids in Figure 2.2, using a scaled Arrhenius plot to bring a wide range of data together on a single plot. A blurred distinction between “normal” and “low-temperature” domain behavior can be made by putting a vertical line at about Tg/T = 0.625. The deviations from Arrhenius behavior observed in the low-temperature regime of Figure 2.3 in most cases are well accounted for by the threeparameter Vogel–Fulcher–Tammann (VFT) equation [37] in the modified form:

η0 = η0 exp(D/[T − T0 ])

(2.1)

where η0, D, and T0 are constants, and T0, the vanishing mobility temperature, lies below the glass transition temperature if a suitable range of data (>3 orders of magnitude) are included in the data fitting. The different curvatures in Figure 2.3 then are reproduced by variations in the parameter D of equation (2.1) [6b].

c02.indd 10

1/20/2011 10:26:43 AM

11

LOW-TEMPERATURE LIQUID BEHAVIOR OF IONIC MELTS

Figure 2.3 displays behavior that is almost universal to <100 °C ILs. Note that, at their normal melting points, these liquids are almost always in the “low-temperature region” of liquid behavior (defined by easily recognized “super-Arrhenius” transport behavior)—that is, they melt within the same range that is generally difficult to access for uni-univalent inorganic molten salt systems and their mixtures. To understand how this can occur, it is helpful to give some consideration to the factors that decide at which temperature a given substance will melt (which we do in the next section). Figure 2.4, however, shows behavior for some inorganic ILs [22], which although superficially similar to that of Figure 2.3 (super-Arrhenius), is almost never found in the organic cation ILs, either protic or aprotic, or in the molten hydrates. The distinction lies in the value of the conductivity near the glass temperature Tg. The conductivity of the low-melting eutectic of Figure 2.4 can be measured easily at temperatures near and below Tg if the melt is cooled rapidly to avoid crystallization. What is interesting is that the conductivity at Tg is approximately nine orders of magnitude higher than that of the typical IL measured at its glass temperature (which requires special equipment). The explanation for this dramatic difference lies in the ability of the small cation, Li+, to slip through the crevices in the vitrified structure and to establish a

100°C

25°C

–1

Log σ (S.cm–1)

–2

–3

–4 Tg Tg –5

LiAlCl4-LiAlI4 eutectic LiAl(SCN,Cl,I)4 melt

–6

–7 20

25

30

35 10000/T (K)

40

45

50

Figure 2.4. Arrhenius plots for conductivity of lithium haloaluminate or pseudohaloaluminate melts, showing super-Arrhenius conductivity that remains high at the glass temperature because of decoupling of Li+ motion from its surroundings, in the “lowtemperature regime.” The break in the upper curve is a result of crystallization of the supercooled liquid during reheating. Note the change from curvilinear to Arrhenius behavior as T falls below Tg and the structure becomes fixed (from Lucas et al. [22]).

c02.indd 11

1/20/2011 10:26:43 AM

12

PHYSICAL CHEMISTRY OF IONIC LIQUIDS

solid-state ionic conductivity. These are known as “decoupled” systems, and they are much sought-after for solid-state ionic devices. The only equivalent in IL phenomenology of this behavior is found in some protic salts of protonated anions like H 2 PO−4 in which the proton motion apparently can be somewhat decoupled from the structural relaxation [38]. However, in these systems, the absolute conductivity is relatively low. A high degree of decoupling of protons, together with a high conductivity as in the Grotthus mechanism of aqueous solutions, is an urgent goal of research on protic ionic liquids. We discuss tests of this sort of behavior later on.

2.3

MELTING POINTS AND THE LATTICE ENERGY

What is implied by the observations in the preceding section is that the crystal lattices of substances of the IL type, like salt hydrate (and salt solvate crystals in general), become thermodynamically unstable with respect to their liquid phases (Gliq < Gcrys) at low temperatures, relative to their cohesive energies. The simplest explanation that can be given for this circumstance focuses on the difficulty that the more complex and size-mismatched ions characteristic of IL and inorganic salt hydrates find in packing closely. This causes the cohesive energy (zero in the gas phase) to be fixed at a value less negative than would be obtained if the centers of attraction could be packed together more closely. The lattice energies reflecting this end up being smaller numerically (i.e., less negative on a scale starting at zero in the reference gas phase). The idea is illustrated in Figure 2.5, which shows the Gibbs free energies of several crystalline forms of the same fictitious substance, along with that of the liquid phase, over a range of temperatures. The crystals are supposed to be nonconvertible except to the liquid and are supposed to differ only in their lattice energies (lattice energy, EL = G at T = 0 K). Figure 2.5 shows that the lowest melting point must belong to the crystal phase with the lowest lattice energy. This can be verified using data for isomers of the same substance (e.g., xylene in which all three isomers o-, m-, and phave essentially the same viscosity and boiling point) for which adequate thermodynamic data are available. We have shown elsewhere [39] how the difference in melting points of approximately 70 K between m- and p-isomers can be traced to a difference in lattice energy of ∼1 kcal/mol. Of course, the polymorph with the lowest melting point will be the one with the highest viscosity at the melting point and the one whose melting point will fall within, or closest to, the “low-temperature region” of the liquid state [40]. Most <100 °C ILs therefore are characterized by low lattice energies that approach that of the glass formed by supercooling of the liquid to the vitreous state. (A method of quantifying the lattice energies for crystalline forms of ILs of known structure has been described recently by Izgorodino and MacFarlane [41].)

c02.indd 12

1/20/2011 10:26:43 AM

13

ELECTRICAL CONDUCTIVITY AND LOW VAPOR PRESSURE

glass XtI5

gibbs energy, G

XtI4 XtI3 liquid XtI2 XtI1

Tmelt Figure 2.5. Gibbs free energies for a system in which there are multiple noninterconvertable crystalline phases, each with a different lattice energy (EL = G, at T = 0 K) but all with the same entropy. All yield a liquid with the same free energy. Arrows show access to the different crystalline phases from the liquid. The crystal with the lowest melting point must be the one with the lowest lattice energy; it will be the one that melts to a liquid with the highest viscosity and the one that therefore is the least likely to crystallize during cooling. When crystallization does not occur, the glassy state does.

This conclusion is consistent with the usual strategy for making ILs. When the salt of a given organic cation does not melt at a temperature low enough to qualify as a <100 °C IL, the problem usually is rectified by the addition of a side-group or by the replacement of an existing side-group with a larger one or one that breaks a previous symmetry [1–5,42,43]. What has been done is, of course, to interfere with the possibility of achieving low energies by efficient packing in three-dimensional (3D) order. This improves the competitive status of the disordered state; that is, it lowers the T/Tg value at the melting point.

2.4 RELATION BETWEEN ELECTRICAL CONDUCTIVITY AND LOW VAPOR PRESSURE To vaporize an aprotic ionic liquid, it is necessary to do work against electrical forces to remove an ion pair from the bulk of the liquid into the vacuum of space. If the ionic liquid were to consist mainly of ion pairs, then this work would be the same as in a strongly dipolar liquid. Fortunately, it is much larger. It is larger because much of the stabilization energy of an ionic liquid is gained by the formation of a quasilattice that is almost as efficient in minimizing the

c02.indd 13

1/20/2011 10:26:43 AM

14

PHYSICAL CHEMISTRY OF IONIC LIQUIDS

electrostatic energy of a system of ions as is the formation of a crystal lattice. After all, it is the enthalpy of vaporization that distinguishes ionic liquids from other ambient temperature liquids, not the enthalpy of fusion. Indeed, the enthalpy of fusion is not exceptional at all. The additional energy of stabilization of ionic crystals over dipolar crystals is gained by the uniform surrounding of ions of one charge by ions of the opposite charge. It is known as the Madelung energy [44]. It is rather difficult to estimate quantitatively, depending on details of crystal symmetry [44], and direct calculations for isotropic phases are not available to the authors’ knowledge. For simple molten salts, these should be related directly to the three liquid radial distribution functions available from inelastic scattering studies. In any event, the low vapor pressure of ionic liquids is a direct reflection of the work that must be done against the Madelung potential to extract an ion pair from the bulk liquid. Quantitatively, the probability p(h) of an enthalpy fluctuation h sufficient to permit an ion pair to escape from the bulk liquid into the vapor is proportional to the Boltzmann factor: p(h) ~ exp− (h/kBT ),

(2.2)

where kB is the Boltzmann constant and h is the enthalpy of extraction per molecule (ion pair), which is dominated by the Madelung energy. In an ideal quasilattice, no ion-pairs can be distinguished over statistically significant time intervals because all sites are equivalent. This is also the condition that leads to the maximum ionic conductivity for a given fluidity (at least in the absence of decoupling, see subsequent sections). This is because the presence of ion pairs lowers the conductivity by permitting diffusion, hence the fluid flow without an ionic current flow. These relations are summed up in some classical equations of electrochemistry, which were derived by considering dilute aqueous solutions in which complete dissociation into independently moving ions could be assumed. Although these solutions present a rather different physical situation from that of solvent-free ionic liquids, the laws developed for their description remain relevant to the description of the ionic liquid properties. The main difference is that the notion of “dissociation” is more obscure. In ionic liquids, the “state of dissociation” must be decided by operational criteria, as we outline in the following list. The classical equations to which we refer are as follows: 1. The Nernst–Einstein equation connecting diffusion and partial equivalent ionic conductivity λi,

λ i = RTzi Di /F 2

(2.3)

where F is the Faraday (charge per equivalent).

c02.indd 14

1/20/2011 10:26:43 AM

ELECTRICAL CONDUCTIVITY AND LOW VAPOR PRESSURE

15

2. The Stokes–Einstein equation connecting diffusivity Di of ionic species i of charge zi and radius ri, with the viscosity η of the medium in which the diffusion is occurring, Di = kBT/6πηri

(2.4)

3. The Walden rule connecting viscosity and equivalent conductivity Λ = ∑ λ i, Λη = const.

(2.5)

Equation (2.5) is particularly useful to us, and we will devote much attention to it in the following paragraphs. It should be noted [46] that the constant of equation 2.5 contains a molecular dimension (the average of ri from equation 2.4) that can affect the relation in unusual cases. The Walden rule is interpreted in the same manner as the Stokes–Einstein relation. In each case, it is supposed that the force impeding the motion of ions in the liquid is a viscous force caused by the solvent through which the ions move. It is the most appropriate for large ions moving in a solvent of small molecules. However, we will see here that, just as the Stokes–Einstein equation applies rather well to most pure nonviscous liquids [45], so does the Walden rule apply rather well to pure ionic liquids [15a]. When the units for fluidity are chosen to be reciprocal poise and when those for equivalent conductivity are Smol−1 cm2, this plot has the particularly simple form shown in Figure 2.6. To fix the position of the “ideal” Walden line in Figure 2.6, we use data for dilute aqueous KCl solutions in which the system is known to be fully dissociated and to have ions of equal mobility [46]. We have included some data for a glass-forming Ca(NO3)2 aqueous solution in which the fluidity has been measured over eight orders of magnitude [47]. For the units chosen, the ideal line runs from corner to corner of a square diagram. Figure 2.6 contains data for some salts of recent study [15] and some salts that were measured between 1983 and 1986 but remained unpublished until those publications listed in [15] (except for [15b] where the argument for unpolarizable anions like BF4− was given, and one example was reported, in proof). It also includes data for some inorganic systems of interesting types, in which the decoupling of conductivity from viscosity, which can be so large at Tg (Figure 2.6), is shown in its earliest manifestation. Because of the way Figure 2.6 reveals the various couplings and decouplings that can be encountered in ionic liquid media, we have used it as a basis for ionic liquid classification. We divide ionic liquids into ideal, “subionic” (or “poor” ionic) liquids, and “superionic” liquids. The subionic liquids still may be good conductors at ambient pressure because their fluidities are high, but the conductivity is much lower than it would be if all moving particles were cations or anions.

c02.indd 15

1/20/2011 10:26:43 AM

16

PHYSICAL CHEMISTRY OF IONIC LIQUIDS 3 Superionic glasses Good ionic liquids

2 Superionic liquids

log(Λ / Scm2mol-1)

1 0 -1

Poor ionic liquids

MOENM2E-BF4

-2

High vapor pressures

MOMNM2E-BF4 BPy-BF4

-3

BMI-BF4 LiAlCl4 -4

Non-ionic liquids

35AgCl-45AgI-20CsCl

-5 -5

-4

-3

-2

-1

0

1

2

3

log(η-1 / Poise-1)

Figure 2.6. Walden plot for tetrafluoroborate salts of various cations, showing subionic (or “supercoupled”) behavior of the salt of the methoxymethyldimethylethyl ammonium cation MOMNM2E+. By contrast, the salt of MOENM2E+ may become decoupled at low temperature. The dotted lines passing through the LiAlCl4 points and the AgCl–AgI–CsCl points are the fits of equation (2.6) to the data points. The slope, α, is inversely proportional to the log (decoupling index) of [45,49]. The plot is annotated to indicate how it may serve as a classification diagram for ionic liquids and other electrolytes. For notation, see the legends for Figures 2.1 and 2.3. (From Xu et al. [15], by permission).

Among the subionic liquids are the cases of proton transfer salts in which the ΔpKa value is small as well as certain aprotic systems in which some special interionic locking mechanism seems to operate, causing a high proportion (approximately 90%) of the diffusive motions to be of the neutral pair type. In this regime, nonaqueous lithium electrolyte solutions also fall, where the failure to fully dissociate causes the low conductivities that plague these electrolytes. It is the progressive increase, with increasing temperature, in the population of nonconducting pairs that causes the less-than-unit slope of all the electrolytes that fall on the right-hand side of the Walden line. Electrolytes that fall on the upper side of the ideal line are desirable because of the high transport numbers for the mobile ions. High transport numbers imply for the species that transports occur faster than expected by the Stokes–Einstein equation. The solutions whose conductivities are shown

c02.indd 16

1/20/2011 10:26:43 AM

ELECTRICAL CONDUCTIVITY AND LOW VAPOR PRESSURE

17

in Figure 2.4 would provide a good example of this behavior, but unfortunately, fluidity data are not available to complete the plot. Instead, we show classic data for LiAlCl4 [48] and also for the more striking case of the silver alkali halide ionic liquid for which fluidity and conductivity data were reported by McLin and Angell [49]. In this case, the silver cation can slip through the channels set up by the alkali halide sublattice [50]. These systems can retain their high conductivity in the glassy state, as illustrated in the silver–alkali halide system shown in Figure 2.4b. For such systems, a “decoupling index” has been defined from the ratio of the conductivity and fluidity relaxation times [47,51], but it equally can be defined from the exponent α in the fractional Walden rule, Ληα = const.

(2.6)

which applies to such systems. α is the slope of the line in the Walden plot and is unity for the ideal electrolyte. We note in Figure 2.6 that, although most IL systems studied lie close to the ideal Walden line, they all exhibit a slightly smaller slope, implying α < 1 as well as a smaller D parameter in equation (2.1) for conductivity than for fluidity. Because this behavior has been noted, historically, for glass-forming molten salts and liquid hydrates [37(d)], it seems to be fundamental to the physics of low-melting electrolytes. It may prove useful in interpreting the manner in which the free space, introduced during configurational excitation, is distributed in ionic liquids. This effect, which leads all except infinitely dilute solutions to have limiting high-temperature conductivities that are smaller than expected from their fluidities, is well illustrated in Figure 2.7 [17], so that comparison can be made with the high temperature limit in Figure 2.7b. Figure 2.7 is a composite representation of the transport properties of ionic liquids of different types intended to show the relation between Walden behavior and the temperature dependence of conductivity. In Figure 2.7a, we show, in this Walden representation, a different set of data from those of Figure 6—a set that emphasizes proton transfer salts (PILs). By design, this plot terminates at the universal high T limit for fluidity implied by Figure 2.3, namely at 104.5 poise. Figure 2.7b shows the behavior, in scaled Arrhenius form, of the conductivity for these systems, to be discussed later on. In principle, the conductivity of the infinitely dilute aqueous solution, in which the ions move without any influence on each other, should have the same value at 1/T = 0 as does the ideal Walden liquid at the high fluidity limit (logη = 4.5). Using data for aqueous KCl solutions at infinite dilution in the temperature range 0 °C to 100 °C given by Robinson and Stokes [46b], and applying the fact that all liquids obey an Arrhenius law for Tg/T < 0.5 (see Fig. 2.3), we see that this correspondence is almost realized. Certainly, the infinitely dilute solution (in which cation–anion friction is absent) has a much higher equivalent conductivity at the high-temperature limit than do any of the other systems.

c02.indd 17

1/20/2011 10:26:43 AM

18

PHYSICAL CHEMISTRY OF IONIC LIQUIDS (a) 35AgCl-45AgI-20CsCl LiCl-6H2O 7:3(mol)I-II [EtNH3][NO3] [BMIM][BF4] [EtNH3][HCO2] Ideal line

4

Log(Λ / S.cm2.mol–1)

3

2

superionic liquids

1 subionic liquids (high vapor pressures)

0

–1 –1

0

1 2 log[η–1(Poise–1)]

3

4

(b) lambda infinity KCl (138K) 35AgCl-45AgI-20CsCl (261K) LiCl-6H2O (138K) 7:3(mol)I-II (180K) [EtNH3][NO3] (181K) [BMIM][BF4] (188K) [EtNH3][HCO2] (148K)

4

Log(Λ / S.cm2.mol–1)

3

2

1

glass

0

–1 0

0.2

0.4

0.6

0.8

1

1.2

Tg/T

c02.indd 18

1/20/2011 10:26:43 AM

ELECTRICAL CONDUCTIVITY AND LOW VAPOR PRESSURE

19

Figure 2.7. (a) Relation of equivalent conductivity to fluidity for various protic and aprotic ionic liquids. The heavy line is the ideal Walden line. Ideally, the temperature dependence of conductivity is set by the value for fluidity because the only force impeding the motion of an ion under fixed potential gradient is the viscous friction. The position of the ideal line is fixed by data for 1 M aqueous KCl solution at ambient temperature. The data for LiCl · 6H2O fall close to it. In most charge-concentrated systems, interionic friction causes a loss of mobility, which is more important at high temperature. This results in the below-ideal slope found for all ILs, protic or aprotic, shown in the figure. A “fractional Walden rule” Ληα = const, (0 < α < 1), applies. When there is a special mechanism for conductance, the Walden plot falls above the ideal line, as for superionics, and the slope α provides a measure of the decoupling index [15]. (b) Tg-scaled Arrhenius plot to display the temperature dependence of the equivalent conductivity in relation to the temperature at which the fluidity reaches the glassy value of 10−11 poise (10−13 p for inorganic network glasses). The inorganic superionic systems have high conductivities at their glass temperatures. Subionic (associated, ion paired, etc.) systems have low conductivity at all temperatures. The ideal (ion interactionfree) behavior for conductivity is shown by the dashed line plot of infinite dilution conductivities for KCl in H2O (data from Robinson and Stokes [46]). To include these data, we assign Tg as 138 K because high-temperature viscosity fitting suggests it. The real value for water is controversial. Notation: I-[MeNH3][NO3], II-[Me2NH2][NO3] (From Xu and Angell [17] by permission of the AAAS).

Counted among these superionic electrolytes are the solutions of strong acids in water and other hydrogen-bonded solvents (e.g., glycerol, and hydrogen peroxide) in which the proton has anomalously high mobility resulting from the Grotthus mechanism [52]. An urgent search is currently in progress for protic systems that can exhibit such behavior in the ambient temperature glassy state so that all-solid fuel cells can be realized. There has been some hope that protonic superionic conductors might be found among the protic ionic liquids that recently have [17] been shown to rival aqueous solutions in their conductivity. In Figure 2.8, we show the conductivities of some of these systems in comparison with the conductivities of their nearest aprotic relatives and with aqueous lithium chloride solutions of 1 and 7.7 molar concentration. The methyl ammonium nitrate–dimethyl ammonium nitrate mixture that is so highly conducting (open triangles in Figure 2.8) is known to obey the Walden rule and indeed is known to fall close to the ideal Walden line, so its high conductance is apparently not a result of any “dry” proton mechanism. However, the extraordinary conductance found for ammonium bifluoride, which equals 7.7 M lithium chloride (LiCl · 6H2O) [53], suggests that some salts with protons in both cation and anions may have superprotonic properties (the case of dihydrogen phosphates was mentioned earlier [54]). Even without superprotonic conduction, it seems that protic ionic liquids, which are mostly neutral and noncorrosive, can serve as fuel cell electrolytes with unusual performance properties, as we describe in Section 2.6

c02.indd 19

1/20/2011 10:26:43 AM

20

PHYSICAL CHEMISTRY OF IONIC LIQUIDS

110°C

0

25°C

–1

log(σ /Scm–1)

–2

–3 [NH4][HF2] 1:6 (mol) LiCl-H2O (7.7M) 1:6 (mol) II-HNO3 1.1MLiCl-H2O 1:1 (mol) II-HNO3 1:1 (mol) II-[NH4][HF2] 4:6 (mol)V-II [EtNH3][NO3] [N1-O-2,211][NO3]

–4

–5

–6

–7 2.2

2.6

3

3.4

3.8

4.2

4.6

1000/T / K–1

Figure 2.8. Specific conductivities of protic versus aprotic ionic liquids, showing matching of concentrated lithium chloride solution conductivity by solvent-free aprotic liquids. Note that, at low temperature, the conductivity of protic nitrate in excess nitric acid solution is higher than that of the aqueous LiCl case with the same excess of solvent water. (From Xu and Angell [17] by permission of the AAAS).

after considering the cohesion factors that can be important in the determination of the IL physical properties.

2.5

COHESION AND FLUIDITY: THE TRADE-OFF

The high cohesion of molten salts associated with electrostatic forces also makes them more viscous than other liquids. The smaller the ions, the higher the cohesion. Because high cohesion means high viscosity, “good” ionic liquids need to achieve an optimum lowering of the cohesion if they are to serve as high-fluidity and high-conductivity media at low temperatures. It would seem that increasing the size of the ions would be the way to achieve high fluidities because it would decrease the coulombic attractions (which depend inversely on the distance of separation between the charge centers). Unfortunately, another effect comes into play as the particle sizes are increased. This is the van der Waals interaction that increases with the number of polarizable elec-

c02.indd 20

1/20/2011 10:26:43 AM

21

COHESION AND FLUIDITY: THE TRADE-OFF

20 0

LiBF4 LiClO4 LiClO3 LiAlCl4

-20 -40 o

Tg ( C)

Protic ILs with [NO3]Protic ILs with [HCO2][N1-O-1,211]+ [N2-O-1,211]+ [N1-O-2,211]+ [N2-O-2,211]+ [N4211]+ [P14]+ [QA][TFSI] (ref. 26) Alkali inorganic salts Fitting curve

LiNO3

KNO3

-60

[NO3]

-80

Ð

Tf

Ð

N7111 N8111

N6111

N6222

I

-100

[TFSI]

II IV

-120

III

-140 0

100

[BF4]Ð

200

[FeCl4]

N7444

N8222 N7222 N723'3'

N8444 N6444

Ð

Ð

300 3

400

500

600

-1

Vm (cm mol )

Figure 2.9. Dependence of the cohesion of salts of weakly polarizable cations and anions, assessed by Tg value, on the ambient temperature molar volume Vm, hence on interionic spacing d. A broad minimum in the ionic liquid cohesive energy is observed at a molar volume of 250 cm3 mol−1, which corresponds to an interionic separation of about 0.6 nm, assuming face-centered cubic packing of anions about cations. The lowest Tg in the plot probably should be excluded from consideration because of the nonideal Walden behavior for this IL (MOMNM 2 E + BF4− ) [15]. The line through the points is a guide to the eye. The data for open triangles are from Sun et al. [64]. For notation, I, II, etc., see Fig. 2.7. (From Xu and Angell [17] by permission of the AAAS.)

trons in the interacting particles. It is important to keep this interaction as low as possible by maintaining an unpolarizable outer surface on the anions. This is the reason for the prevalence of perfluorinated species among the anions of ionic liquids used in practice. However, perfluorinated anions are expensive to manufacture, and even perfluorinated species begin to cause fluidity decreases when their size becomes large enough [12,42]. One way of monitoring these effects is to use the glass temperature as a cohesion indicator. In Figure 2.9, we show the glass transition temperatures for numerous different cation-anion pairs having in common that the species are chosen for low polarizabilities [15]. Thus, aromatic cations are excluded, and anions either are fluorinated or, if oxyanions, the oxygen electrons are polarized strongly toward anion center species (Cl[VII], S[VI], or N[V]). The plot is made against molar volume on the assumption that this is roughly the cube of the cation–anion separation. The plot shows a broad minimum around 250 cm3 per mole, corresponding to a 4 molar ion concentration. Because anions less polarizable than the perfluorinated species represented in Figure 2.9 cannot be made, it would seem that if fluorinated species are to

c02.indd 21

1/20/2011 10:26:43 AM

22

PHYSICAL CHEMISTRY OF IONIC LIQUIDS

be avoided for economic reasons, then the number of anions that can be used in acceptable ionic liquids is rather limited. Nitrates, thiocyanates, nitrites, formates, dicyanamides, chlorosulfonates, and methanesulfonates would seem to be acceptable. Dicyanamides have been shown to have high fluidities with imidazolium and pyrrolidinium cations [43,55] but they are thermally less stable. Tetracyanoborates are known to be weakly basic, a prerequisite, but they are expensive. The size of cations for use with these anions is already limited. The aprotic systems with the minimum glass temperatures in Figure 2.6 are quaternary ammonium cations with uniformly small alkyl groups attached. If we describe the alkyl groups using Wunderlich’s “bead” notation [56] (a bead is taken as an atom or group that can participate in a change of configuration), then the methoxyethylethyldimethyl ammonium (MOENM2E) cation has eight beads. The lowest Tgs are obtained with the four-bead anion BF4−, which is perfluorinated. If fluorine is avoided, then the lowest Tgs are obtained with three-bead and two-bead anions, NO−3 , formate HCO−2 , and thiocyanate (SCN−). The formate anion is not stable, which is unfortunate, as the most conductive <100 °C IL on record is that of hydrazinium formatted by Sutter [57]. Data on salts with thiocyanate anionare reported by Pringle et al. [58]. Systems with cations smaller (and then more symmetrical) than eight beads seem to melt well above ambient temperatures, unless one uses protic cations [17]. Once protic cations are admitted, a new range of high-fluidity (low Tg) systems opens up, and the maximum ambient temperature ionic conductivity jumps sharply, as shown in Figure 2.7, largely because the Tg values go down at the same time as the charge concentration increases (to ∼10 M).

2.6 PROTON TRANSFER IONIC LIQUIDS AS NOVEL FUEL CELL ELECTROLYTES Surprisingly, it is only a recent recognition that protic ionic liquids can serve as proton transfer electrolytes in hydrogen–oxygen fuel cells. A 2003 report from Susan et al. [18a] describes the performance of a hydrogen electrode using, as the electrolyte, the salt formed by proton transfer from the acid form of bis-trifluoromethanesulfonyl imide (HTFSI) to the base imidazole, scooping our own work in this area. Our own experience [18b–d,24] has shown that, by using certain favorable protic ILs, fuel cells can be made with short-term performance superior to that of the commercially feasible phosphoric acid high-temperature fuel cell. In [18], we gave comparisons of open-circuit voltages produced, and short-circuit currents flowing, in simple U-tube type hydrogen–oxygen fuel cells using PIL electrolytes, spiral wound with platinum wire electrodes, with those produced when a phosphoric acid electrolyte is substituted in the same cell. An example from those early presentations is shown in Figure 2.10. Because the bubbling of hydrogen gas was irregular, the current flowing was subject to occasional

c02.indd 22

1/20/2011 10:26:43 AM

PROTON TRANSFER IONIC LIQUIDS AS NOVEL FUEL CELL ELECTROLYTES

23

1.3 Pt wire electrodes: same cell active surface: ~1cm2 (from cyclic voltammetry studies)

1.2 1.1

Potential (V)

1.0 0.9

[EtNH3][NO3] - 100°C H3PO4 98% - 100°C

0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1.0

2

Current Density (mA/cm )

Figure 2.10. Potential of the simple bubbler type fuel cell, under load, using ethylammonium nitrate as electrolyte. A comparison is made with the behavior of the identical cell with phosphoric acid substituted for the ethylammonium nitrate electrolyte.

large fluctuations, which are not found in more sophisticated gas diffusion cells that have since been described for the interesting case of inorganic protic electrolytes [24]. Figure 2.11 shows data from a Teflon sandwich cell for different mixed ammonium salt electrolytes compared with date for the same cell using a phosphoric acid electrolyte. Note especially in Figures 2.10 and 2.11 the high potential of the cell relative to the value obtained when the IL is replaced by 98% H3PO4 with (0.9 V vs. 0.75 V) in electrolytes containing ethylammonium nitrate as an electrolyte. The origin of this higher exchange current density at the oxygen electrode in this case and the reason for its limited current range before dropping to a lower current density is not yet clear. A performance at nearly the same low overpotential, one which is sustained over a wider range of current densities and which is actually superior to that of the same cell with phosphoric acid as electrolyte, has been reported by Thomson et al. using a PIL formed from 2-fluoropyridine and triflic acid [59]. As shown in Figure 2.12, this performance is exceeded only when 6 M triflic acid is used as electrolyte (in the same cell). Unfortunately, the performance of the cell with the 2-fluoropyridium triflate electrolyte degrades when the temperature increases above 100 °C. This electrolyte should be considered as an acid electrolyte and discussed will be in the next section. Whether acid electrolytes of the PIL variety (i.e., nonaqueous materials) will prove as corrosive to noble metal catalysts as aqueous acids are remains unclear. Meanwhile, understanding the basis for control of acidity, and independently the ionicity, in PILs is an important matter that we now address.

c02.indd 23

1/20/2011 10:26:43 AM

24

PHYSICAL CHEMISTRY OF IONIC LIQUIDS 1.3

Cell Potential (V)

1.2

H3PO4

1.1

NH4NO3 - NH4Tf (60–40) mol.%

1.0

NH4NO3 - NH4TFAc (50–50) mol.%

0.9

NH4Tf - NH4TFAc (30–70) mol.%

0.8 0.7 0.6 0.5 0.4 0.3 0.2 IR corrected

0.1 0.0 50

0

150

100 i (mA/cm2)

1.3 IR corrected

1.2 1.1 1.0

Cell Potential (V)

0.9 0.8 0.7 0.6 0.5

H3PO4 EAN - 130°C

0.4

EAN - EAH2PO4 1 Wt.% - 140°C

0.3

NH4NO3 - NH4Tf (60–40) mol.%-low current density

0.2

NH4NO3 - NH4Tf (60–40) mol.%-high current density

0.1

NH4NO3 - NH4TFAc (50–50) mol.%-low current density

0.0 1E-4

NH4NO3 - NH4TFAc (50–50) mol.%-high current density

1E-3

0.01

0.1

1

10

100

2

Log [i/mA/cm ]

Figure 2.11. Tafel plot for a gas diffusion Teflon sandwich fuel cell.

2.7 IONICITY AND ACIDITY OF PILS: THE PROTON FREE ENERGY LEVEL DIAGRAM A useful way of organizing the thinking on protic ionic liquids is introduced by Gurney in his classical monograph, Ionic Processes in Solution [60]. Gurney’s treatment was more rigorous than the present one insofar as he dealt with proton transfers in a single type of system, namely proton transfer processes

c02.indd 24

1/20/2011 10:26:44 AM

IONICITY AND ACIDITY OF PILS

25

Figure 2.12. Tafel plots for Teflon sandwich type H2/O2 gas fuel cell using E-tech electrodes and liquid electrolytes [59]. The 2-fluoropyridine-HTf average pKa places it in the acid electrolyte domain of Fig. 2.13, above the level for phosphoric acid.

in an aqueous medium, and could refer all results to a single standard state. In our case with no single solvent, every PIL is a separate system. Nevertheless, a useful (i.e., predictive) organization of data can be obtained using aqueous data as the energetic basis for placing different acid and conjugate base pair levels on the diagram. The energy level separation will be valid to the extent that the free energies of hydration to the aqueous standard state are the same for each cation and anion species. Because neither cation nor anion is strongly hydrating in most cases we are concerned with, this generally will be a reasonable approximation. This expectation has been confirmed for several cases by direct electrochemical interrogation under strictly anhydrous conditions [61]. Using both experimental aqueous pKa data and theoretically calculated values from various authors for the large negative pKa cases, we have assembled the diagram shown in Figure 2.13. This diagram proves useful for clarifying the issues of ionicity and acitity in PILs. In [32], it was shown that PILs formed by proton transfers across a gap of 0.7 eV or more are of a sufficient ionicity that the Walden rule is a good approximation to the behavior. Accordingly, it is shown that “good” ionic liquids can be produced from an enormous number of different acid + base combinations. It also is shown that, with an appropriate choice of couples, PILs with average pKa values that fall in the (i) neutral, (ii) acid, and even (iii) superacid ranges can be obtained. Basic electrolytes are less simply developed because we have found that molecular liquids containing the imine group, which have been described as “superbases” based on certain chemical processes, do not extract a proton from water to give a basic PIL as might have

c02.indd 25

1/20/2011 10:26:44 AM

26

PHYSICAL CHEMISTRY OF IONIC LIQUIDS

Super-Acids

Acid Electrolytes

Neutral Electrolytes

Occupied Vacant HSbF6 SbF6(HAlCl4) AlCl4HTFSI TFSIHSO3F SO3FpFPy pFPyH+ HTf Tf HClO4 ClO4H2SO4 HSO4HPO2F2 PO2F2NO3HNO3 CH3SO3H CH3SO3- (MS) 2-Fluoropyridine H+ 2-Fluoropyridine CF3COO- (TfAc) CF3COOH H2O H2O+ H3PO4H3PO4 1,2,3–1 H-triazole H+ 1,2,4–1 H-triazole HF FHCOOH HCOOCH3COOH CH3COOHα-MePyH+ α-MEPy hydrazine H+ hydrazine NH4+ NH3 EtNH3 EtNH3Et3NH+ Et3N (TEA) C (NH2)3+ Guanidine H2O OH-

Basic Electrolytes

NH3

NH2-

Super-Bases

H2NC (dma)

HNC (dma)

OH-

O2- (Na+)

pKa

E (eV)

–14

0.83

–10 –9

0.59 0.53

–1.3 –0.6 –0.43 –0.25 0 2.12 3 3.2 3.75 4.75 6.99 7.96 9.23 10.63 11.25

0.08 0.04 0.03 0.01 0 –0.13 –0.18 –0.19 –0.22 –0.28 –0.41 –0.47 –0.55 –0.63 –0.67

14

–0.83

28

–1.660

Figure 2.13. Proton energy level diagram, following Gurney’s scheme for aqueous solutions [60]. The occupied level is the acid form of the couple, while its conjugate base provides the vacant level. The larger the energy level gap across which the proton drops in forming the stable combination of ions for any chosen pair of levels, the more thermally stable the PIL against redissociation, i.e., the lower the vapor pressure and the higher the “ionicity.” The higher the average pKa of the couple between which the proton transfer occurs, the more acidic is the PIL. The parentheses in “(HAlCl4)” in this diagram indicate that the free acid does not exist.

been expected. However, a superacidic PIL, based on proton transfer to pentafluoropyridine, (pKa = –13) from HTFSI, has been identified positively and is being described elsewhere [33]. Separate experiments have yielded similar results when the acid used is transient HAlCl4, which is obtained from HCl bubbling over anhydrous AlCl3 [26].

c02.indd 26

1/20/2011 10:26:44 AM

27

COMPARISON OF HIGHEST-CONDUCTING ORGANIC

A case of special interest to us is that of the basic molecule guanidine and its protonated conjugate—the simple, symmetrical, and resonance-stabilized guanidinium cation, C(NH 2 )+3 . We have synthesized a collection of guanidinium salts [62] and have studied some of their binary solutions. Their high conductivities provide the material for our final section.

2.8 COMPARISON OF HIGHEST-CONDUCTING ORGANIC AND INORGANIC IONIC LIQUIDS To bring this chapter to a close we now go back to the earliest section to take one of the cases of <100 °C inorganic ILs in which the conditions for high conductivity have been optimized, and compare it with a case from the organic cation protic sub-class in which likewise the conditions for high conductivity have been more or less optimized. The inorganic case is that of a lithium salt in which the anion has been made large and strongly polarized toward the central species which is trivalent Al. The ligands of the Al are chloride in one case and iodide in the other so that the two lithium salts form a <100 °C eutectic, and the anions are large enough that the lithium ions can slip through the interanionic interstices and thus escape the repulsive forces that oppose the center of mass motion of the anionic sub-lattice. This eutectic composition can be vitrified on fast cooling and the glass transition has been observed at about −50 °C. As Figure 2.14 shows, the conductivity at 100 °C is almost 100 mScm−1. The value at ambient cannot be measured because of prior crystallization, but can be interpolated as ≈10 mScm−1. For inorganic ILs, higher conductivities at ambient can be obtained only by going to molecular cations like hydrazinium. Hydrazinium formate, for instance has a conductivity of 50 mScm−1 [29,57], a consequence of the lower cohesion (Tg ≈ −90 °C) for such liquids. The value at 100 °C can be exceeded by use of organic cations provided that they are protic in character (though the case of the tetrachloroferrate of imidazolium-type cations is competitive [15a]). Figure 2.14 shows the data for the guanidinium thiocyanate-chloride eutectic as an example of anionic liquid comparable in simplicity with the lithium aluminium chloride-iodide eutectic. Guanidinium is a three-bead cation, while the thiocyanate and chloride approximate as two- and one-bead anions. The tetrachloroferrate might well be superior both at 100 °C and certainly at ambient if crystallization can be avoided. Even with the dimethyl ammonium-methylammonium nitrate eutectic of Figure 2.8, room temperature liquids are available and both 100 mScm−1 and 10 mScm−1 at ambient are achieved. However, these are potentially explosive liquids and hence are not destined for practical applications. These high conductivity results are all correlated with high fluidities. A Grotthus mechanism might make a small contribution to conductivity in the case of guanidinium cations.

c02.indd 27

1/20/2011 10:26:44 AM

28

PHYSICAL CHEMISTRY OF IONIC LIQUIDS

−1

logσ (S/cm)

−2 100 °C

25 °C

−3

−4

−5

−6

3GdmSCN-2GdmCl eutectic LiAlCl4-LiAII4 eutectic LiAl(SCN, Cl, l)4 melt 2.5

3.0

3.5

4.0

4.5

1000/T (K)

Figure 2.14. Conductivity of the best-conducting inorganic ionic liquids from Fig. 2.3 compared with that of the best organic IL, using Arrhenius plots. The organic cation case is marginally higher despite the decoupling of the lithium ion from the anion matrix in the former case. Whether or not a Grotthus mechanism contributes in the Gdn+ case is not clear at this time. It is possible that the tetrahaloferrate salts of the guanidinium cation might be even higher-conducting than the thiocyanate-chloride mixture seen in the figure.

2.9

CONCLUDING REMARKS

We have tried to cover important aspects of the physical chemistry of the ionic liquids currently under study as well as to relate them to what is known about other types of low-melting ionic media. In concluding, we must emphasize that much of the success in their application, particularly in the green chemistry area where there is hope that they might replace the common volatile solvents of environmentally hostile character, will depend on the important chemical properties of these media. Of these we have only addressed the proton activity, that is, Bronsted acidity, classified in Figure 2.13 for the case of protic ionic liquids. Because of the preoccupation of the field with aprotic ionic liquids there has so far been no systematic application of this variable in chemical studies in ILs. Properties such as Lewis acidity and basicity, and donor and acceptor character, are all basically physical properties, though they are most commonly invoked in a more empirical manner, based on experience, as described in [1–4]. A helpful treatment of Lewis basicity in ionic liquids has recently been given by MacFarlane et al. [63].

c02.indd 28

1/20/2011 10:26:44 AM

REFERENCES

29

ACKNOWLEDGMENTS This work was supported by the National Science Foundation under the Solid State Chemistry Grant DMR 0082535 and by the DOE-LANL hightemperature fuel cell program, and most recently by the DOD Army Research Office. REFERENCES 1. (a) R. D. Rogers, K. R. Seddon, eds., Ionic Liquids: Industrial Applications to Green Chemistry, American Chemical Society, 2002. (b) K. R. Seddon, Nat. Mater., 2, 363 (2003). 2. R. D. Rogers, K. R. Seddon, S. Volkov, eds., Green Industrial Applications of Ionic Liquids, Kluwer Academic, 2003. 3. T. Welton, Chem. Rev., 99, 2071 (1999). 4. H. Ohno, ed., Ionic Liquids: The Front and Future of Material Development (in Japanese), High Tech. Info., 2003. 5. M. Gaune-Escarde, ed., Molten Salts: From Fundamentals to Applications, Kluwer Academic, 2002. 6. (a) C. A. Angell, J. Phys. Chem., 68, 1917 (1964). (b) C. A. Angell, J. Non-Cryst Sol., 13, 131 (1991). 7. C. A. Angell, in Molten Salts: From Fundamentals to Applications, M. GauneEscarde, ed., Kluwer Academic, 2002, p. 305. 8. C. L. Hussey, Adv. Molten Salt Chem., 5, 185 (1983). 9. E. I. Cooper, E. J. M. Sullivan, in Proc. 8th. Int. Symp. Molten Salts, Electrochem. Soc., 1992, p. 386. 10. J. S. Wilkes, M. J. Zawarotko, J. Chem. Soc. Chem. Commun., 965 (1992). 11. R. T. Carlin, J. S. Wilkes, in Chemistry of Nonaqueous Solutions, G. Mamantov and A. I. Popov, eds., VCH Publishers, 1994. 12. J. Sun, M. Forsyth, D. R. MacFarlane, Ionics, 3, 356 (1997). 13. P. Bonhote, A.-P. Dias, M. Armand, N. Papageorgiou, K. Kalyanasundaram, M. Graetzel, Inorg. Chem., 35, 1168 (1996). 14. (a) C. J. Bowles, D. W. Bruce, K. R. Seddon, Chem. Comm., 1625 (1996). (b) J. D. Holbrey, K. R. Seddon, J. Chem. Soc., Dalton Trans., 2133 (1999). 15. (a) W. Xu, E. I. Cooper, C. A. Angell, J. Phys. Chem. B, 107, 6170 (2003). (b) E. I. Cooper, C. A. Angell, Solid State Ionics, 9 & 10, 617 (1983). 16. (a) M. Hirao, H. Sugimoto, H. Ohno, J. Electrochem. Soc., 147, 4168 (2000). (b) M. Yoshizawa, W. Ogihara, H. Ohno, Electrochem. Solid State Lett., 4, E25 (2001). (c) H. Ohno, M. Yoshizawa, Solid State Ionics, 154, 303 (2002). 17. W. Xu, C. A. Angell, Science, 302, 422 (2003). 18. (a) A. B. H. Susan, A. Noda, S. Mitsushima, M. Watanabe, Chem. Comm., 938 (2003). (b) J.-P. Belieres, W. Xu, C. A. Angell, Abstract 982, Fall ECS meeting, Orlando, Sept. 2003. (c) M. Yoshizawa, J.-P. Belieres, W. Xu, C. A. Angell, in Abstr. Papers of the American Chemical Society, vol. 226, 2003. (d) C. A. Angell, W. Xu, J.-P. Belieres, M. Yoshizawa, Provisional patent application, No. 9138-0123.

c02.indd 29

1/20/2011 10:26:44 AM

30

PHYSICAL CHEMISTRY OF IONIC LIQUIDS

19. N. Byrne and C. A. Angell, J. Mol. Biol., 378, 707 (2008). 20. H.-M. Luo, G. A. Baker, S. L. Lee, R. M. Pagni, S. Dai, J. Phys. Chem. B, 113, 4181 (2009). 21. M. Yoshizawa, W. Xu, C. A. Angell, J. Am. Chem. Soc., 125, 15411 (2003). 22. P. Lucas, M. Videa, C. A. Angell, Inorganic “ionic liquids”: low melting inorganic salts and salt mixtures (unpublished work). 23. C.-L. Liu, H. G. K. Sundar, C. A. Angell, Mater. Res. Bull., 20, 525 (1985). 24. J.-P. Belieres, D. Gervasio, C. A. Angell, Chem. Comm. 4799 (2006). 25. E. J. Sutter, C. A. Angell, J. Phys. Chem., 75, 1826 (1971). 26. Y. Ansari, K. Ueno, C. A. Angell, submitted for publication (2010). 27. A. Noda, K. Hayamizu, M. Watanabe, J. Phys. Chem. B, 105, 4603 (2001). 28. M. Yoshizawa, W. Xu, C. A. Angell, J. Am. Chem. Soc., 125, 15411 (2003). 29. C. A. Angell, N. Byrne, J.-P. Belieres, Accounts Chem. Res. (Special Issue on Ionic Liquids), 40, 1228 (2007). 30. D. R. MacFarlane, S. A. Forsyth, E. I. Izgorodina, A. P. Abbott, G. Annat, K. Frazer, Phys. Chem. Chem. Phys., 11, 4962 (2009). 31. K. Ueno, H. Tokuda, M. Watanabe, Phys. Chem. Chem. Phys., 8, 1649 (2010). 32. J.-P. Belieres, C. A. Angell, J. Phys. Chem. B, 111, 4926 (2007). 33. Y. Ansari, K. Ueno, J.-P. Beleieres, B. R. Cherry, G. P. Holland, J. L. Yarger, C. A. Angell (to be published). 34. A. Sivaraman, H. Senapati, C. A. Angell, J. Phys. Chem. B, 103, 4159 (1999). 35. (a) W. Kauzmann, Chem. Rev., 43, 219 (1948). (b) C. A. Angell, Science, 267, 1924 (1995). 36. C. Alba, J. Fan, C. A. Angell, J. Chem. Phys., 110, 5262 (1999). 37. (a) H. Vogel, Phys. Z., 22, 645 (1921). (b) G. S. Fulcher, J. Am. Ceram. Soc., 8, 339 (1925). (c) G. Tammann, W. Hesse, Z. Anorg. Allg. Chem., 156, 245 (1926). (d) This equation is frequently referred to in current literature as the VTF equation, which is idiosyncratic in view of the publication dates of the cited authors. The idiosyncracy was introduced in the late 1960s to emphasize Tammann’s contribution to the phenomenology of glass-forming liquids; see C. A. Angell, J. Am. Ceram. Soc., 51, 117 (1968); C.A.Angell, C.T. Moynihan, in Molten Salts:Analysis and Characterisation, G. Mamantov, ed., Dekker, 1969, p. 315; and G. W. Scherer, ed., J. Am. Ceram. Soc., 75(5), 1060 (1992), and then was spread by other workers with or without attributions. 38. F. Mizuno, J.-P. Belieres, N. Kuwata, A. Pradel, M. Ribes, C. A. Angell, J. Non-Cryst. Sol., 352(42–49), 5147 (2006). 39. J. Wong, C. A. Angell, Glass: Structure by Spectroscopy, Dekker, 1976, p. 12. 40. C. Alba, L. E. Busse, C. A. Angell, J. Chem. Phys., 92, 617 (1990). 41. E. I. Izgorodina, U. L. Bernard, et al., Crystal Growth Design, 9, 4834–4839 (2009). 42. J. Sun, M. Forsyth, D. R. MacFarlane, J. Phys. Chem. B, 102, 8858 (1998). 43. D. R. MacFarlane, J. Golding, S. Forsyth, M. Forsyth, Chem. Comm., 16, 1430 (2001). 44. C. Kittel, Solid State Physics, 3rd ed., Wiley, 1967. 45. (a) F. Fujara, B. Geil, H. Sillescu, G. Fleischer, Z. Phys. B. Condens Matter, 88, 195 (1992). (b) I. Chang, H. Sillescu, J. Phys. Chem., B, 101, 8794 (1997). (c) M. T. Cicerone, M. D. Ediger, J. Chem. Phys., 103, 5684 (1995).

c02.indd 30

1/20/2011 10:26:44 AM

REFERENCES

31

46. (a) J. O’M. Bockris, Reddy Modern Electrochemistry, vol. 1, 2nd ed., Plenum Press. (b) R. A. Robinson, R. H. Stokes, Electrolyte Solutions, 2nd ed., Butterworths, 1959, p. 465. 47. (a) J. H. Ambrus, H. Dardy, C. T. Moynihan, J. Phys. Chem., 76, 3495 (1972). (b) J. H. Ambrus, C. T. Moynihan, P. B. Macedo, J. Electrochem. Soc., 119, 192 (1972). 48. C. R. Boston, L. F. Grantham, S. J. Yosim, J. Electrochem. Soc., 117, 28 (1970). 49. M. McLin, C. A. Angell, J. Phys. Chem., 92, 2083 (1988). 50. Changle Liu, H. G. K. Sundar, C. A. Angell, Solid State Ionics, 18 & 19, 442 (1986). 51. C. A. Angell, Chem. Rev., 90, 523 (1990). 52. C. J. T. von Grotthus, Annales de Chemie, 58, 54 (1806). 53. C. T. Moynihan, N. Balitactac, L. Boone, T. A. Litovitz, J. Chem. Phys., 55, 3013 (1971). 54. F. Mizuno, J.-P. Belieres, N. Kuwata, A. Pradel, M. Ribes, C. A. Angell, J. Non-Cryst. Sol., 352(42–49), 5147 (2006). 55. D. R. MacFarlane, M. Forsyth, P. C. Howlett, et al., Accounts Chem. Res., 40, 1165 (2007). 56. W. B. Wunderlich, J. Phys. Chem., 64, 1052 (1960). 57. E. J. Sutter, Hydrogen Bonding and Proton Transfer Interactions in HydrazineBased Binary Liquids and Related Systems, Ph. D. Dissertation, Purdue University, 1970. 58. J. M. Pringle, J. Golding, C. M. Forsyth, G. B. Deacon, M. Forsyth, D. R. MacFarlane, J. Mater. Chem., 12, 3475 (2002). 59. J. Thomson, P. Dunn, L. Holmes, J.-P. Belieres, C. A. Angell, D. Gervasio, ECS Trans., 13(28), 21 (2008). 60. R. W. Gurney, Ionic Processes in Solution, Dover Publications, 1962. 61. J. A. Bautista-Martinez, L. Tang, J.-P. Belieres, R. Zeller, C. A. Angell, J. Phys. Chem., 113, 12586 (2009). 62. Z.-F. Zhao, C. A. Angell, J. Phys. Chem., in press (2010). 63. D. R. MacFarlane, J. M. Pringle, et al., Chem. Commun., 18, 1905–1917 (2006). 64. J. Sun, M. Forsyth, D. R. MacFarlane, J. Phys. Chem., 102, 8858 (1998).

c02.indd 31

1/20/2011 10:26:44 AM

PART I BASIC ELECTROCHEMISTRY

c03.indd 33

1/20/2011 10:26:49 AM

3 GENERAL TECHNIQUES Yasushi Katayama

Electrochemical methods can be powerful tools. They can be used to reveal the chemical and physical properties of room-temperature ionic liquids. Most of existing electrochemical techniques [1] developed in aqueous solutions are applicable for the ionic liquids, as demonstrated in the chloroaluminate ionic liquids. However, several procedures must be observed if one is to obtain reliable data in electrochemical measurements. This section describes the procedures that are important for the ionic liquids.

3.1

EQUIPMENT

Room-temperature ionic liquids are usually prepared from a variety of organic and inorganic salts. Because both ionic liquids and halide salts are hygroscopic in most cases, they must be dried under vacuum at elevated temperatures and handled in a dry atmosphere. It is convenient to construct a vacuum line for drying, as shown in Figure 3.1. The dried materials are usually handled in a glove box filled with a dry argon or nitrogen. These materials may be handled in dry air if the existence of oxygen is allowable. However, it should be noted that the solubility of oxygen in the ionic liquids is not negligible and that the reduction of dissolved oxygen to superoxide will be observed in the absence of protic species [2–6]. The existence of oxygen is also unfavorable for the electrodeposition of base metals.

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

35

c03.indd 35

1/20/2011 10:26:49 AM

36

GENERAL TECHNIQUES

Figure 3.1. Schematic illustration of a metal vacuum line for drying materials.

3.2

PREPARATION OF ELECTROLYTES

Electrochemical methods are sensitive to the extent that it is possible to detect a trace of electroactive species in electrolyte solutions. Because of this distinctive feature, electrochemical methods have been developed and used for analytical purposes. The detection method used is known as polarography. For the electrochemical study, purification of the electrolyte solutions is therefore important. As for most aqueous and organic electrolyte solutions, there are various well-established techniques for purifying both solvents and electrolytes. In the case of room-temperature ionic liquids, it is especially important to purify the starting materials used for preparing the ionic liquids. Quaternary ammonium cations are frequently employed as the organic cations of the ionic liquids. They are usually prepared by the reaction of alkyl halides with such tertiary amines as alkylamine, pyridine, alkylimidazole, and alkylpyrolidine. The purity of the resultant ionic liquids is often dependent on the purity of these starting materials. Therefore, it is recommended that these starting materials are purified properly before the preparation of the organic cations. In addition to purification of the starting materials, it also is important to control the reaction when preparing the cations. When the reaction is exothermic, the reaction mixture becomes heated locally or wholly, resulting in the degradation of the starting materials or products. The starting materials can react even under mild conditions. So care must be taken in diluting the starting materials with aprotic solvents, stirring the reaction mixture thor-

c03.indd 36

1/20/2011 10:26:49 AM

REFERENCE ELECTRODES

37

oughly, keeping the temperature of the reaction mixture constant, adding the reactants very slowly, and so on. The organic salts are further purified by recrystallization, followed by evaporation of the solvents and vacuum drying at elevated temperatures. The compounds that provide the anions of the ionic liquids should be checked carefully for purity and further purified if necessary. In the case of chloroaluminate ionic liquids, it is essential to purify aluminum trichloride by sublimation under reduced pressure. The mixing of the organic chlorides with AlCl3 can generate heat and cause the degradation of organic cations. Some nonchloroaluminate ionic liquids, BF4−, PF6−, and CF3SO−3 ionic liquids, are prepared by the metathesis of the organic halides and silver, alkali metal, and ammonium salts of the desired anions in certain aprotic solvents [7–9]. However, contamination by the inorganic cations often is unavoidable because the solubility of the inorganic salts in the ionic liquids is not negligible. Thus, some alternative methods have been used for preparing these ionic liquids [10,11]. Some water-stable and hydrophobic ionic liquids, represented by N(CF3 SO2 )2− ionic liquids, can be prepared easily by making the equimolar amounts of an organic halide and LiN(CF3SO2)2 interact in water. The N(CF3 SO2 )−2 ionic liquid phase is separated from the aqueous phase containing lithium halide. A trace of lithium salts in the N(CF3 SO2 )−2 ionic liquid can be removed by extracting the ionic liquid into dichloromethane [8]. The removal of water from these water-stable ionic liquids can be done simply by vacuum drying at elevated temperature. It is usually difficult to remove protons from the ionic liquids that undergo hydrolysis. In the chloroaluminate ionic liquids, the evacuation coupled by the treatment with phosgene can be proposed for removing protons [12].

3.3 WORKING ELECTRODES Platinum, glasslike carbon, and tungsten often are used as inert working electrodes for the fundamental electrochemical studies in the ionic liquids. For such transient electrochemical techniques as cyclic voltammetry, chronoamperometry, and chronopotentiometry, it is safer to use the working electrode with a small active area. This is because most ionic liquids will have low conductivity, and this often causes the ohmic drop in the measured potentials by the current flowing between the working and the counter electrode. Microelectrodes may be useful for the electrochemical measurements in the case of handling low conductive media.

3.4

REFERENCE ELECTRODES

The potential of an electrode gives essential information about the phenomena being observed at the electrode. The electrode potential determines both

c03.indd 37

1/20/2011 10:26:49 AM

38

GENERAL TECHNIQUES

direction and rate of the electrode reaction. Therefore, it is necessary to measure or control the electrode potential against a reference electrode, of which the potential is stable and reproducible during a series of measurements. In aqueous solutions, there are several such well-established reference electrodes, namely Ag/AgCl, Hg/Hg2Cl2, Hg/HgO, and Hg/HgSO4. These reference electrodes may be usable also for the ionic liquids with some liquid–liquid junctions: however, contamination by water into the ionic liquids is unavoidable and not desirable in most cases. The reference electrodes based on aprotic solvents (e.g., Ag/Ag(I) in CH3CN) can be used also in the ionic liquids, although there is still a possibility of contamination by the solvents. Thus, it is always better to prepare the reference electrodes based on the ionic liquids. Figure 3.2 shows the typical reference electrodes used for the ionic liquids as well as for nonaqueous solutions. The reference electrode is basically composed of a metal electrode immersed in an inner solution. The supporting electrolyte of the inner solution should be identical to that of a test solution to minimize the contamination and the junction potential. The inner solution contains the electroactive species determining the potential of the reference electrode. Consequently the inner solution should be separated from the test solution by a glass filter, or frit, to avoid the mixing of these solutions by keeping the electrochemical junction. Porous glass also is used for this purpose. The potential of the reference electrode generally depends on both electroactive species and supporting electrolyte. Whereas it is difficult to compare the potentials measured in different ionic liquids, even if the same redox couple is employed for the reference electrodes, the redox potential of ferrocenium (Fc+)/ferrocene (Fc) often is used as a common potential standard for non-

Figure 3.2. Examples of the reference electrodes for ionic liquids.

c03.indd 38

1/20/2011 10:26:49 AM

REFERENCE ELECTRODES

39

aqueous systems because its redox potential is regarded as independent of solvents. Thus, the potential of the reference electrode should be given against the redox potential of the Fc/Fc+, also in ionic liquids. In addition, the rest potential of a metal electrode immersed in the supporting electrolyte without any redox couple (sometimes called a quasi- or pseudoreference electrode) is less reliable and should not be used for electrochemical measurements unless its potential is proved to be stable and calibrated properly with a true reference electrode. The low conductivity of the ionic liquids often causes the ohmic drop in the potential measurements. To reduce the ohmic drop, the tip of the reference electrode should be placed at the working electrode, if necessary, using a Luggin–Haber capillary. Figure 3.3 shows the potentials of the selected redox couples in some ionic liquids against Fc/Fc+. The cathodic decomposition potential of

Figure 3.3. Potentials of the selected redox couples in some ionic liquids.

c03.indd 39

1/20/2011 10:26:49 AM

40

GENERAL TECHNIQUES

1-ethyl-3-methyl-imidazolium, EMI+, is located at approximately −2.5 V in relation to Fc/Fc+ in EMICl–AlCl3, EMIBF4, and EMIN(CF3SO2)2. It seems that the difference in the potentials of Ag(I)/Ag [10,13,14] in these ionic liquids is not so significant, whereas the potential of I−/I −3 is likely to be affected by the kinds of organic cations [8,15,16]. Various redox couples have been used for the reference electrodes in the electrochemical studies in the ionic liquids. The following are some examples of the reference electrodes, which are actually used in some ionic liquids. 3.4.1

Chloroaluminate Systems

In case of chloroaluminate ionic liquids, the potential of an aluminum electrode immersed in the acidic ionic liquids that contain AlCl3 at more than 50 mol% is usually assumed as the potential standard. The electrode reaction is considered as follows: 4 Al 2 Cl 7− + 3 e− = Al + 7 AlCl −4

(3.1)

The dimeric chloroaluminate anion, Al 2Cl −7 , is present under acidic conditions. The molar fraction of AlCl3 in the acidic ionic liquid, which is used as the reference electrode, often is assigned a 66.7 or 60.0 mol%. It should be noted that the potential of this reference electrode is dependent on the mole fraction (activity) of AlCl3, as is evident in equation (3.1). This reference electrode is used not only for the acidic ionic liquids but also for basic and neutral ones in which no redox reaction between Al and monomeric chloroaluminate anion, AlCl −4 , is available at near room temperature. In addition, this Al/Al(III) reference electrode also is used for some nonchloroaluminate ionic liquids. The potential of the Al/Al(III) electrode ([AlCl3] = 60.0 mol%) is reported as −0.406 ∼ −0.400 V against Fc/Fc+ in the EMICl-AlCl3 ionic liquid [9]. 3.4.2

Chlorozincate Systems

The reference electrode similar to Al/Al(III) is used in chlorozincate ionic liquids. The electrodeposition of Zn is reported to be possible in the acidic chlorozincate melt containing ZnCl2 at more than 33.3 mol% at near room temperature [17]. The reference electrode is composed of a Zn electrode immersed in the acidic melt that contains ZnCl2, usually at 50 mol%. Although the details of the electrode reaction are not yet known, it is thought that such chlorozincate species as ZnCl −3 , Zn 2Cl 3− 7 , and (ZnCl2)n behave as electroactive species. The potential of this Zn/Zn(II) electrode has not been measured against Fc/Fc+. 3.4.3

Other Nonchloroaluminate Systems

In case of the ionic liquids having no reversible electroactive species, a certain redox couple must be selected for the reference electrode. The Al/Al(III)

c03.indd 40

1/20/2011 10:26:49 AM

REFERENCES

41

electrode is sometimes used in nonchloroaluminate ionic liquids: however, the Al/Al(III) electrode is not recommended for those who have no experience preparing chloroaluminate ionic liquids. It is possible to prepare the reference electrode composed of a platinum electrode immersed in the ionic liquid that contains both Fc+ and Fc if a soluble ferrocenium salt is available. It should be noted, however, that the volatility of Fc can lead to a change in the concentration of Fc (i.e., the electrode potential) at elevated temperatures. The redox reaction between silver and monovalent silver species, Ag(I), is one of the typical redox couples used for the reference electrodes in aqueous and nonaqueous media. The Ag/Ag(I) electrode is composed of a Ag electrode immersed in an ionic liquid that contains Ag(I) at 0.01 or 0.1 mol dm−3. Ag(I) is introduced into the ionic liquid by dissolving silver salts having the same anion as the ionic liquid. For example, AgBF4 is reported to be soluble in EMIBF4 [10], and it is probably soluble in other BF4− ionic liquids. The potential of Ag(I)/Ag ([Ag(I)] = 0.2 mol dm−3 is estimated as +0.51 V against Fc/Fc+ in EMIBF4 at room temperature [9,10]. AgNO3 and AgCl also have been used as a Ag(I) source for BF4− ionic liquids [18]. In the case of N(CF3SO2 )−2 ionic liquids, it is possible to prepare AgN(CF3SO2)2 by the reaction of excess Ag2O with HN(CF3SO2)2 aqueous solution, followed by filtration, evaporation of water, and drying as Ag 2 O + 2 HN(CF3 SO2 )2 → 2 AgN(CF3 SO2 )2 + H 2 O

(3.2)

The HN(CF3SO2)2 aqueous solution can be prepared by passing a LiN(CF3SO2)2 aqueous solution through a protonated cation-exchange resin. AgN(CF3SO2)2 is soluble in the N(CF3SO2 )2− ionic liquids. The formal potential of Ag(I)/Ag is reported as +0.49 ∼ +0.51 V against Fc/Fc+ at room temperature [14,19,20]. Of course, the anion of the silver salt does not need to be the same as the ionic liquid. For example, Ag(I) can be introduced into the N(CF3SO2 )−2 ionic liquids by dissolving AgCF3SO3. The redox reaction between iodides, I− and I −3 , also is used as the reference electrodes in nonchloroaluminate ionic liquids especially for the N(CF3SO2 )−2 ionic liquids. Platinum wire is used for the electrode and immersed in an ionic liquid that contains iodine and an iodide salt having the same cation as the ionic liquid. The concentration ratio of the iodine salt to iodide often is specified as 1 : 4, namely I − /I −3 . The potential of this [I −3 ] : [I − ] = 1 : 3 electrode is reported as −0.16 V in TMHAN(CF3SO2)2 at 50 °C [15] (TMHA+ = trimethyln-hexylammonium) and as −0.21 V in EMIN(CF3SO2)2 at room temperature [16] against Fc/Fc+.

REFERENCES 1. A. J. Bard, L. R. Faulkener, eds., Electrochemical Methods Fundamentals and Applications, 2nd ed., Wiley, 2001.

c03.indd 41

1/20/2011 10:26:49 AM

42

GENERAL TECHNIQUES

2. M. T. Carter, C. L. Hussey, S. K. D. Strubinger, R. A. Osteryoung, Inorg. Chem., 30, 1149 (1991). 3. I. M. AlNashef, M. L. Leonard, M. C. Kittle, M. A. Matthews, J. W. Weidner, Electrochem. Solid State Lett., 4, D16 (2001). 4. I. M. AlNashef, M. L. Leonard, M. A. Matthews, J. W. Weidner, Ind. Eng. Chem. Res., 41, 4475 (2002). 5. M. C. Buzzeo, O. V. Klymenko, J. D. Wadhawan, C. Hardacre, K. R. Seddon, R. G. Compton, J. Phys. Chem. A, 107, 8872 (2003). 6. Y. Katayama, H. Onodera, M. Yamagata, T. Miura, J. Electrochem. Soc., 151, A59 (2004). 7. J. S. Wilkes, M. J. Zaworotko, J. Chem. Soc., Chem. Commun., 965 (1992). 8. P. Bonhôte, A-P. Dias, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). 9. J. Fuller, R. T. Carlin, R. A. Osteryoung, J. Electrochem. Soc., 144, 3881 (1997). 10. Y. Katayama, S. Dan, T. Miura, T. Kishi, J. Electrochem. Soc., 148, C102 (2001). 11. J. G. Huddleston, A. E. Visser, W. M. Reichert, H. D. Willauer, G. A. Broker, R. D. Rogers, Green Chemistry, 3, 156 (2001). 12. M. A. M. Noël, P. C. Trulove, R. A. Osteryoung, Anal. Chem., 63, 2892 (1991). 13. Q. Zhu, C. L. Hussey, G. R. Stafford, J. Electrochem. Soc., 148, C88 (2001). 14. Y. Katayama, M. Yukumoto, M. Yamagata, T. Miura, T. Kishi, in Proc. 6th Int. Symp. Molten Salt Chemistry and Technology, C. Nianyi and Q. Zhiyu, eds., Shanghai University Press, 2001, p. 190. 15. K. Murase, K. Nitta, T. Hirato, Y. Awakura, J. Appl. Electrochem., 31, 1089 (2001). 16. R. Kawano, M. Watanabe, Chem. Commun., 2003, 330 (2003). 17. Y.-F. Lin, I.-W. Sun, Electrochim. Acta, 44, 2771 (1999). 18. A. Saheb, J. Janata, M. Josowicz, Electroanalysis, 18, 405 (2006). 19. G. A. Snook, A. S. Best, A. G. Pandolfo, A. F. Hollenkamp, Electrochem. Commun., 8, 1405 (2006). 20. M. Yamagata, N. Tachikawa, Y. Katayama, T. Miura, Electrochim. Acta, 52, 3317 (2007).

c03.indd 42

1/20/2011 10:26:49 AM

4 ELECTROCHEMICAL WINDOWS OF ROOM-TEMPERATURE IONIC LIQUIDS (RTILs) Hajime Matsumoto

4.1

INTRODUCTION

Room-temperature ionic liquids (RTILs) have been studied as novel electrolytes for 50 years since the discovery of the chloroaluminate systems. Recently, another system consisting of fluoroanions such as BF4− and PF6, which have good stability in air, also has been investigated extensively. In both systems, the nonvolatile, noncombustible, and heat-resistance nature of RTILs, which cannot be obtained with conventional solvents, is observed for possible applications in lithium batteries, capacitors, solar cells, and fuel cells. The nonvolatility should contribute to the long-term durability of these devices. The noncombustibility of a safe electrolyte is desired especially for the lithium battery [1]. RTILs also have been studied as an electrodeposition bath [2]. In considering RTILs as an electrolyte for these applications, it is important to know the electrochemical stability of the RTILs toward a particular electrode. For this purpose, the electrochemical potential range, starting from the point where no electrochemical reaction is observed, has been estimated using the electrochemical method. “The electrochemical window” (EW) of RTILs is a term commonly used to indicate both the potential range and the potential difference; it is calculated by subtracting the reduction potential from the oxidation potential. In the history of the development of RTILs, improvement

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

43

c04.indd 43

1/20/2011 10:26:56 AM

44

ELECTROCHEMICAL WINDOWS OF RTILs

of the EW RTILs was one of the important landmarks, especially because this meant an improvement in the cathodic limiting potential of the RTILs [2–7]. Cyclic voltammetry and linear sweep voltammetry usually are used for the EW estimation of the RTILs, as in conventional electrolytes. These methods are easy to use despite the complication often encountered in comparing the resulting voltammogram with other reported data and even for the same RTILs. Except for technical problems such as “IR drop,” this seems to be a result of the difference in the measurement conditions, including differences between the working and/or the reference electrode as well as, differences in the potential scanning rate and/or the cut-off current density that determine the cathodic and anodic limiting potentials and so on. Furthermore the quality of the RTILs affects the resulting voltammogram. For example, the presence of a small amount of water in the RTILs cannot be neglected even at 100 ppm because the electrochemical measurements are generally sensitive to impurities. Despite these difficulties, comparisons of EW data among the various RTILs provides important information to prepare electrochemically stable RTILs. In this chapter, the electrochemical windows of RTILs are reviewed and compared for different RTILs. The effects of the measurement conditions on the resulting voltammograms are an important part of the discussion. The use of “ferrocene” as an internal potential reference in the RTILs also is introduced. 4.2 EFFECTS OF THE MEASUREMENT CONDITIONS ON THE VOLTAMMETRY IN THE RTILS 4.2.1 Evaluation of the Reduction or Oxidation Potential of the RTILs by Voltammetry To estimate the EW of RTILs, one obtains the reduction and oxidation potentials of the RTILs toward a certain reference electrode (RE) as in conventional electrolyte solutions. Similar problems develop with the techniques used for finding the EW in the RTILs as in estimations of organic solvent systems. For example, the different reference electrodes make it difficult to compare obtained potential data. These issues will be discussed in Section 4.2.2. In addition, there is the important problem involved in choosing the reduction or oxidation potential of the electrolyte solutions from either cyclic voltammetry (CV) or linear sweep voltammetry (LSV). Because the oxidation or reduction reaction of cations or anions contained in the RTILs are electrochemically irreversible in general [8–10], the corresponding reduction or oxidation potential cannot be obtained specifically, unlike the case of the redox potential for an electrochemically reversible system. Figure 4.1 shows the typically observed voltammogram (LSV) for RTILs. Note that both the reduction and the oxidation current monotonically increase with the potential sweep in the cathodic and anodic directions, respectively. Because no peak is observed

c04.indd 44

1/20/2011 10:26:57 AM

EFFECTS OF THE MEASUREMENT CONDITIONS ON THE VOLTAMMETRY

45

Figure 4.1. Linear sweep voltammogram of RTILs as typically observed. Two separate voltammograms are indicated here.

even at a high current density (10 mA cm), a certain current density, which is called the cut-off current, must be selected to evaluate the reduction or oxidation potential. In this figure, the reduction or the oxidation potential—in other words, the cathodic limiting or anodic limiting potential, which are denoted as ECL and EAL, respectively—is evaluated when the reduction or oxidation current reaches 1.0 mA cm. This means that both ECL and EAL will change because of the arbitrarily selected cut-off current density. In many studies, the cut-off current density is selected from between 0.1 and 1.0 mA cm. In studies of the application in capacitors, however, the cut-off current density has been selected to be below 0.1 mA/cm [11–13]. This may be a result of the dependence of the performance of devices on the decomposition potential of the electrolyte. The comparison of different RTILs indicates that the current scale of a voltammogram reflects the current density or surface area of the working electrode. RTILs differ from conventional electrolytes by their viscosity. Even relatively low viscous RTILs possess 20 to nearly 200 mPas at 25 °C, which is two or three orders of magnitude greater than that of conventional solvents except for the F(HF)2.3 systems [14]. The conductivity of RTILs is not as low (0.1∼20 mS cm−1 at 25 °C) because of its high concentration (3∼6 mol dm−3); however, it seemed necessary to consider the compensation of the “IR drop” or “ohmic drop” especially in cases of highly viscous RTILs. However, this problem can be circumvented by selecting an appropriate size of working electrode to reduce the specific current. Table 4.1 shows rough estimates of the IR drop for a cut-off current of 1.0 mA cm through a working electrode disk of various diameters. Note from the rough estimates that the microelectrode seems to be more suitable for use for the EW even in highly viscous RTILs because the IR drop is almost negligible (<100 μV). An additional merit to the use of a microelectrode is that the mass transport property is improved

c04.indd 45

1/20/2011 10:26:57 AM

46

ELECTROCHEMICAL WINDOWS OF RTILs

TABLE 4.1. Calculated IR-drop at 1.0 mA cm−2 Flow Through a Disk Working Electrode Conductivity (mS cm−1)

Viscositya (mPas)

Resistanceb (kΩ)

0.1 1 5.0 10.0

500∼1000 100∼500 40∼100 20∼40

10 1 0.2 0.1

iR-dropc d = 32 mm 80 μV 8 μV 1.6 μV 0.8 μV

d = 1.6 mm 0.2 V 20 mV 4 mV 2 mV

d = 1.0 7.85 V 785 mV 157 mV 78.5 mV

a

Viscosity is a typical value actually observed. Roughly estimated by multiplying the distance between the working electrode and the reference electrode with the conductivity of RTILs, assuming that the distance is 1.0 cm. It is noted that the estimated value changes with the distance and both electrodes must be closed as much as possible. c d = diameter. b

with the change in the diffusion mechanism [15]. The same situation also is achieved by the enhancement of the mass transport by means of the forced convection with a rotating disk electrode (RDE) system [16]. Both methods seem more suitable to a highly viscous RTIL (>500 mPa s at 25 °C) with low conductivity. There is a brief report on the microelectrode [17–20] or the RDE [18] applied to the estimation of EW of RTILs. However, in the case of RTILs with a moderate viscosity (<100 mPa s at 20 °C), it is not necessary to consider IR drop excessively, even when working disk electrodes with a relatively small size (d < 5 mm) are used as shown in Table 4.1. The roughly estimated IR drop in such cases might not be as large (lower than tens of millivolts). 4.2.2

Reference Electrode

The configuration of the RE obtained in RTILs so far is listed in Table 4.2. Note that the configuration of the RE used in RTILs is almost the same as that in an organic solvent. In all cases, the reference electrode contains a redox system that has a stable potential. For the chloroaluminate system, the chloroaluminate anion (Al 2Cl 7−) itself is involved in the potential determining reaction. To avoid contaminating the target RTILs from the redox species in the RE, the content of the reference electrode is separated by a filtering material such as glass frit. Figure 4.2 shows the EW data for 1-ethyl-3-methylimidazolium tetrafluoroborate EMI–BF4 obtained from voltammograms of different reference electrodes [31–33]. Note that, although the range of the EW is almost the same (4.4 V), the cathodic and anodic limiting potentials are different, even in the same RTIL. One way around this problem is to use one redox compound as the internal reference. For an organic solvent system, the International Union of Pore and Applied Chemistry (IUPAC) recommends the use of a redox

c04.indd 46

1/20/2011 10:26:57 AM

47

c04.indd 47

1/20/2011 10:26:57 AM

RTILs in RE

EMI–TFSI

EMI–BF4 EMI–TaF6

13

14 15

[TMAI] = 60 mM, [I2] = 12 mM [TMAI] = 60 mM, [I2] = 12 mM [TPAI] = 60 mM, [I2] = 15 mM AgBF4 (saturated) AgTaF6 (saturated)

Al 2Cl −7 Al 2Cl −7 Al 2Cl −7 Al 2Cl −7 Al 2Cl −7 Al 2Cl −7 Al 2Cl −7 Al 2Cl −7 Al 2Cl −7 Al 2Cl −7

Redox Material

— 0.195 — —

3I − ↔ I 3− + 2e− Ag ↔ Ag+ + e− Ag ↔ Ag+ + e−

Pt wire Ag wire Ag wire

Pt wire

3I − ↔ I 3− + 2e−

0.24 0.42b 0.27b 0.40 — 0.29c — — — —

Redox Potential of Ferrocenea

—

Al + 7 AlCl −4 ↔ 4 Al 2Cl 7− + 3e− Al + 7 AlCl −4 ↔ 4 Al 2Cl 7− + 3e− Al + 7 AlCl −4 ↔ 4 Al 2Cl 7− + 3e− Al + 7 AlCl −4 ↔ 4 Al 2Cl 7− + 3e− Al + 7 AlCl −4 ↔ 4 Al 2Cl 7− + 3e− Al + 7 AlCl −4 ↔ 4 Al 2Cl −7 + 3e− Al + 7 AlCl −4 ↔ 4 Al 2Cl −7 + 3e− Al + 7 AlCl −4 ↔ 4 Al 2Cl −7 + 3e− Al + 7 AlCl −4 ↔ 4 Al 2Cl −7 + 3e− Al + 7 AlCl −4 ↔ 4 Al 2Cl 7− + 3e−

Potential Determing Reaction

3I − ↔ I 3− + 2e−

Pt wire

Al wire Al wire Al wire Al wire Al wire Al wire Al wire Al wire Al wire Al wire

Electrode Material

31 32

30

29

29

21 22 23 24 4 5 25 26 27 28

Reference

Note: BPC: N-butylpyridiniumchloride, EMIC: 1-ethyl-3-methylimidazolium-chloride; DMABPC: p-dimethylamino-N-butylpyridiniumchloride; DMEEMAC: dimethylethylethoxy-methylammonium-chrolide; DMPIC: 1,2-dimethyl3-propylimidazoliumchloride; PMIC: 1-propyl-3-methylimidazoliumchloride; TMAI: tetramethylammoniumiodide; TPAI: tetrapropyl-ammoniumiodide; TFSI: bis(trifluoromethylsulfonyl) imide. a Redox potential of ferrocene in the same RTILs that contained in RE. b These values are not obtained experimentally. They are calculated by considering the reported potential difference depending on the composition of chlorualuminate. These data are shown in the references listed in the table. c The potential difference between #1 and #5 is 0.05 V.

EMI–CH3SO3

12

Nonchloroaluminate system 11 EMI–CF3SO3

Chloroaluminate system 1 BPC + AlCl3 (N = 0.67) 2 BPC + AlCl3 (N = 0.60) 3 EMIC + AlCl3 (N = 0.67) 4 EMIC + AlCl3 (N = 0.60) 5 EMIC + AlCl3 (N = 0.44) 6 DMABPC + AlCl3 (N = 0.67) 7 DMEEMAC + AlCl3 (N = 0.40) 8 DMPIC + AlCl3 (N = 0.67) 9 DMPIC + AlCl3 (N = 0.60) 10 PMIC + AlCl3 (N = 0.60)

Number

TABLE 4.2. RE Used in RTILs

48

ELECTROCHEMICAL WINDOWS OF RTILs

Figure 4.2. Electrochemical windows of EMI–BF4 estimated with various reference electrodes. The arrow shows the position of a certain redox material such as ferrocene (the position is indicated virtually).

potential of ferrocene (dicyclopentadienyliron(II)) as the reference potential regardless of the kind of solvent molecule [34]. For RTILs, a few reports refer to the redox potential of ferrocene in RTILs. Table 4.2 shows that the redox potential of ferrocene in RTILs is identical to the content of the RE [21,24,30]. Though the validity of ferrocene has not yet been sufficiently RTIL, there are a few reports of voltammograms with a ferrocene internal reference [35–37]. This discussion will show that the reported reduction and oxidation potential of RTILs can be corrected by the use of the redox potential of ferrocene, if available. A comparison of ECl and EAL for various RTILs corrected by the use of ferrocene will be presented later. When the RTIL contained in the RE is different from the target RTIL, an experimental error can be generated because of the junction potential between these RTILs. However, the use of an internal reference such as ferrocene can remove the junction potential. This is because the redox potential measured with the RE contains the same junction potential. A quasireference electrode (QRE) also is used for voltammetry in both conventional electrolyte solutions and RTILs. Pt [18,19,38], Li [38], and Ag [12,33,39,40] metals are the most commonly used QRE in reports on the EW of RTILs. The merit of the QRE is that the junction potential, as stated previously, is negligible because these metals are immersed directly in the target RTILs. Taking the ease and convenience of a QRE into consideration, it should be noted that the potential determining reaction is unclear when only a metal is used as the QRE. Furthermore, it is possible that some by-products generated during the voltammetry can affect the potential of the QRE. To avoid this problem, the relatively high concentration of the metal cation (10∼100 mM) dissolved in a target RTIL or an electrolytic current is kept as low as possible (nA∼μA). The addition of a small amount of Ag has been shown to shift the redox potential of Ag/Ag over a few hundred mV compared with that of Ag QRE. Therefore, an internal reference such as ferrocene should be used as in conventional organic electrolytes.

c04.indd 48

1/20/2011 10:26:57 AM

EFFECTS OF THE MEASUREMENT CONDITIONS ON THE VOLTAMMETRY

4.2.3

49

Effect of Residual Water on Voltammetry

The major impurities contained in RTILs are water and oxygen, even in highly pure samples. This is because these molecules are dissolved easily into the RTILs from air. Because these molecules are electrochemically active, the removal of these molecules is essential before any voltammetric measurement. The effect of water on the EW of RTILs has been studied in RTILs containing the imidazolium cation and fluoroanions (1-butyl-3-methylimidazolium tetrafluoroborate or hexafluorophosphate). Both the ECL and EAL are reported to have shifted toward positive and negative potentials, respectively, with the addition of more than 3 wt% of water, and this has been shown to cause a decrease in the wide EW of more than 2.0 V [19]. However, there is a study of the effects of a small amount of water (10.5 mM, ca. 100 ppm) on the voltammogram of an RTIL containing N-butylpyridinium (BP) and chloroaluminate anion [41]. For glassy carbon (GC) and tungsten (W) working electrodes, no significant effects were observed regardless of the addition of water. For a platinum working electrode, a reduction current of HCl, which is generated from the reaction between water and the chloroaluminate anion, was observed at a potential much more positive than that of the cathodic limiting potential (ECL) of this system. These studies indicate the importance of removing water from the RTILs. Figure 4.3a shows a LSV with platinum as the working electrode in trimethylpropylammonium-bis(trifluoromethyl)imide (TMPA–TFSI) containing various amounts of water. The amount of water in the RTIL was controlled by vacuum drying of an air-saturated sample [42]. Before sample drying (water content was 0.59 wt%), the peak for the reduction of oxygen was observed at −1.2 V versus Fc/Fc, and the reduction wave for water was observed at below −1.5 V versus Fc/Fc. Both reduction currents decreased with the decreasing amount in the TMPA–TFSI. This fact indicates that the reduction wave was derived from oxygen and water both overlapping the reduction wave for TMPA–TFSI, which appeared after sufficient dehydration of the sample (15 ppm). Even with the presence of 58 ppm water, the reduction in current was observed, and the current may mislead the cathodic limiting potential of the sample by approximately a few hundred mV toward the positive potential. This phenomenon indicates the importance of dehydration in samples below 50 ppm. A glove box filled with an inert gas such as Ar or N2 is essential to avoid contamination by water and oxygen during the electrochemical measurements. As stated previously, GC and W are not sensitive to residual amounts of water in the chloroaluminate system [41]. For a nonchloroaluminate system the GC electrode is insensitive to water, as shown in Figure 4.3b. Fortunately, because of RTIL’s nonvolatile nature, the water can be removed easily by vacuum drying, unlike with conventional organic solvents. For example, after vacuum drying (10 Torr) at 100 °C, the water content in a hydrophobic RTIL such as EMI–TFSI will be less than 20 ppm. In general, the time

c04.indd 49

1/20/2011 10:26:57 AM

50

ELECTROCHEMICAL WINDOWS OF RTILs

Figure 4.3. Linear sweep voltammogram of trimethylpropylammonium-bis (trifluoromethylsulfonyl)imide (TMPA-TFSI) containing various amounts of water. T = 25 °C; working electrodes: (a) Pt (0.02 cm2) and (b) GC (0.00785 cm2). The scan rate is 50 mV/s−1.

needed to decrease the water content below 20 ppm will depend on the sample volume (e.g., 1 mL and 30 min) [42]. 4.2.4

Effect of Scan Rate on the Voltammogram

The scan rate is an important parameter for potential sweep methods such as CV or LSV. The current is proportional to the square root of the scan rate in all electrochemical systems—irreversible, reversible, and quasireversible systems. Figure 4.4 shows the LSV for EMI–TFSI using a GC [42]. Note in this figure that, in the relationships between the oxidation current at a certain

c04.indd 50

1/20/2011 10:26:57 AM

MUTUAL COMPARISONS OF ELECTROCHEMICAL WINDOWS OF RTILS

51

Figure 4.4. Linear sweep voltammogram of EMI–TFSI at various scan rates. The inset shows the relationships between the current density at +2.1 V versus Fc/Fc+ and the square root of the current density. T = 25 °C; working electrode: GC (0.00785 cm2).

potential and the square root of the current, the oxidation current density obeys the theoretical prediction. This indicates that the limiting potential estimated at the same current density can be shifted in the negative or positive direction by changing the scan rate. Therefore, in the case of mutual comparison among various reported voltammgrams of RTILs taken at different scan rates, the cut-off current density in each case should be changed with the relationship between the square root of the scan rate and a current density.

4.3 MUTUAL COMPARISONS OF ELECTROCHEMICAL WINDOWS OF RTILS One of the aims of this chapter is to show how the electrochemical windows compare for the various RTILs reported so far. To reduce the difference in the measurement conditions, a cathodic and anodic limiting potential is tabulated from the reported voltammogram, whose current scale can be indicated by a current density. The cut-off current density is changed by the scan rate based on the relationship between the electrolytic current and the scan rate. The basis for the selection of the cut-off current is 0.5 mA cm at 50 mV. In a case in which the maximum current density observed in a voltammogram is lower than the cut-off current density, the limiting potential is taken at the maximum current. However, when the current scale is much greater than the cut-off current density, the limiting potential is taken at a cut-off current density that is as low as possible.

c04.indd 51

1/20/2011 10:26:58 AM

52

ELECTROCHEMICAL WINDOWS OF RTILs

To show the possible effect of the internal potential reference, the redox potential of ferrocene in the RTILs is used here for the correction of the potential data estimated for the various reference electrodes. 4.3.1

Chloroaluminate Systems

Chloroaluminate systems have been studied intensively over the past few decades mainly in electrochemical research fields. Many electrochemical properties of RTILs have been sorted out using the chloroaluminate systems [43]. Chloroaluminate anions form RTILs with various organics, including not only EMI and BP but also triazolium [44] and aliphatic onium cations [7,45–47]. Other unique RTILs similar to the chloroaluminate systems such as the chlorogallate (GaCl4) [48], chloroborate (BCl4) [49], and bromoaluminate (AlBr4) systems [50] have been reported. The most significantly different feature of this system, compared with other nonchloroaluminate systems, is that the composition changed with the mole ratio between the organic chloride and AlCl3. Also, the corresponding EW changed because of the change in the potential-determining reaction. Usually, the mole fraction of AlCl3 is denoted as N = (AlCl3 / ([AlCl3] + [RCl])). The chloroaluminate systems are categorized by the N value as follows: N > 0.5, N = 0.5, and N < 0.5, which are called acidic, neutral, and basic melts, respectively. The cathodic or anodic limiting potential estimated from the reported voltammogram is shown in Table 4.3. Often, aluminum wire immersed in an acidic melt (N = 0.60 or 0.67) is used as the RE. Because the potentialdetermining reaction is aluminum deposition from the acidic melt on an Al wire, the potential scale in the voltammogram often is denoted as V versus Al/ Al. The potential differences between these two REs is not large (130∼180 mV), and it can be obtained from the reported potentiometric titration curve, if available [22–24,26]. Figure 4.5 shows the EWs of various chloroaluminate systems, excluding the potential differences in the RE. From this figure, it can be confirmed that the EW of the chloroaluminate systems depends on the composition of the melt. The ions involved in the potential-determining reaction are specified in Table 4.4. However, the anodic limiting potential is not always determined by the oxidation of AlCl4, as shown in the p-dimethylaminoN-butylpyridnium (DMABP) system. The anodic limiting potential of this system is caused by the oxidation of the DMABP cation [5]. Figure 4.6 shows the EWs of chloroaluminate systems with the correction of the potential differences derived from the melt composition and the kind of cation using the redox potential of ferrocene. This figure clearly indicates the effect of using the ferrocene internal reference. The cathodic-limiting or anodic-limiting potential is almost the same when the same cation is contained in RTILs and is independent of the measurement conditions such as the difference of melt composition, which is not considered in Figure 4.5. The potential variation shown in Figure 4.5 even for the same RTILs is almost eliminated.

c04.indd 52

1/20/2011 10:26:58 AM

53

c04.indd 53

1/20/2011 10:26:58 AM

0.60 0.52 0.50 <0.50 0.45 0.55 0.45 0.60 0.55 0.55 0.50 0.50 0.50 0.47 0.47 0.44 0.40 0.67 0.49

DMPI DMPI DMPI DMPI DMPI PMI

DMABP

EMI

Na

Cation

Pt W GC Pt W W W W W W GC W W W W W-micro W GC GC

WEb

20 7.85 70 20 7.85 7.85 7.85 78.4 78.4 28.3 70 78.4 28.3 28.3 78.4 d = 25 mm 78.4 — —

Sc (mm2)

8 9 8 8 9 10 10 3 4 4 8 4 4 4 3 4 4 6 6

REd

20 100 20 20 100 100 100 100 50 20 20 50 20 20 100 2000 50 200 200

ve (mV s−1)

0.32 0.20 0.32 0.32 0.40 0.70 0.50 0.70 0.50 0.32 0.32 0.50 0.32 0.32 0.70 50 0.50 0.14 0.14

If (mA cm−2)

TABLE 4.3. Electrochemical Windows of RTIL Based on Chloroaluminate

0.0 0.1 −2.5 −2.4 −2.3 — −2.2 −0.2 −0.5 0.0 −2.2 −1.9 −1.9 −2.0 −2.2 −2.2 −2.0 −0.5 −1.7

V vs. RE

−2.3 −2.3 −2.4 −2.4 −2.6 −2.4 −0.8 −2.0

−0.4 −0.9 −0.4

(V vs. Fc/Fc+)

ECLg

2.5 2.3 2.3 1.0 — 2.7 — 2.3 2.5 2.6 2.4 2.5 2.5 1.0 0.8 1.2 0.9 1.7 0.9

V vs. RE

2.1 2.1 0.6 0.5 0.8 0.5 1.4 0.6

2.1 2.1 2.2

(V vs. Fc/Fc+)

EALh

2.5 2.2 4.7 3.5 — — — 2.5 3.0 2.6 4.6 4.4 4.4 3.0 3.0 3.4 2.9 2.1 2.6

EW Rangei (V)

(Continued)

26 27 26 26 27 28 28 51 6 52 26 6 52 52 51 17 6 5 5

Reference

54

c04.indd 54

1/20/2011 10:26:58 AM

0.67 0.60 0.50 0.50 0.47 0.44 0.44 0.44

BP

GC W W GC W W GC Pt

WEb

— 78.4 78.4 — 78.4 77 71 49

Sc (mm2)

1 1 4 1 1 1 1 1

REd

100 100 50 100 100 50 50 50

ve (mV s−1)

0.01 0.70 0.50 0.01 0.70 0.20 0.20 0.20

If (mA cm−2) (V vs. Fc/Fc+) −0.3 −0.4 −1.5 −1.3 −1.4 −1.3 −1.3 −1.4

V vs. RE −0.1 −0.1 −1.1 −1.1 −1.2 −1.1 −1.1 −1.2

ECLg

1.9 2.4 2.5 1.4 0.7 0.9 0.8 0.7

V vs. RE 1.7 2.1 2.1 1.2 0.5 0.7 0.5 0.5

(V vs. Fc/Fc+)

EALh

2.0 2.5 3.6 2.5 1.9 2.0 1.8 1.9

EW Rangei (V)

21 51 6 21 51 41 41 41

Reference

Note: BP: N-butylpyridinium; EMI: 1-ethyl-3-methylimidazolium; DMABP: p-dimethylamino-N-butylpyridinium; DMPI: 1,2-dimethyl3-propylimidazolium; PMI: 1-propyl-3-methylimidazolium. a Mole fraction of AlCl3. b Working electrode. c Area of working electrode. d Reference electrode (see Table 4.2). e Scan rate. f Cut-off current. h Anordic limitng potential. i Range = EAL–ECL.

Na

Cation

TABLE 4.3. (Continued)

MUTUAL COMPARISONS OF ELECTROCHEMICAL WINDOWS OF RTILS

55

Figure 4.5. Electrochemical window of chloroaluminate system. The asterisk indicates the composition of the melt in RE of N = 0.60. Another case is N = 0.67. The potential difference between N = 0.60 and 0.67 (130∼180 mV) is not considered. Working electrodes: GC (glassy carbon); W (tungsten); Pt (platinum).

c04.indd 55

1/20/2011 10:26:58 AM

56

ELECTROCHEMICAL WINDOWS OF RTILs

TABLE 4.4. Ionic Species Associated with a Cathodic or Anodic Limiting Potential of Chloroaluminate Composition (N) N > 0.5 acidic N = 0.5 neutral N < 0.5 basic

Anions in Melts

Cathodic Limiting Potential

Anodic Limiting Potential

Al 2Cl −7 , AlCl 4− AlCl −4 AlC −l 4

Al 2Cl 7− Cation Cation

AlCl −4 AlCl −4 Cl−

Figure 4.6. Electrochemical windows of chloroaluminate system with a correction by the ferrocene internal reference. Original data were measured with the different REs. Solid: N = 0.67; broken: N = 0.60. Working electrode: GC (glassy carbon); W (tungsten); Pt (platinum).

4.3.2

Nonchloroaluminate System

The nonchloroaluminate system has been investigated intensively because fluoroanions are also attractive candidates for RTILs [29,53,54]. The significant difference from the chloroaluminate system is its stability in air. The number of anions, which form RTILs with various cations, demonstrate an upward trend [39,40,55–60]. The EW data for the nonchloroaluminate systems reported so far are summarized in Table 4.5. Problems resulting from the difference in the measure-

c04.indd 56

1/20/2011 10:26:59 AM

57

c04.indd 57

1/20/2011 10:26:59 AM

Anion

TFSI TFSI TFSI Methide Methide BF4 BF4 CF3SO3 CF3SO3 CH3SO3 BF4 BF4 BF4 BF4 BF4 TFSI TFSI TFSI TFSA CF3COO F(HF)2.3 TSAC TaF6 BETI DCA F(HF)2.3 F(HF)2.3 BF4 PF6 F(HF)2.3

Cation

DMEI DMPI DMPI DMPI DMPI DMFP EMP EMI EMI EMI EMI EMI EMI EMI EMI EMI EMI EMI EMI EMI EMI EMI EMI EMI EMI DMI PrMI PrMI BMI BMI

Pt GC GC GC GC GC Pt Pt Pt Pt GC Pt Pt Pt GC Pt GC GC Pt Pt GC GC Pt GC Pt GC GC GC Pt micro GC

WEa

0.78 70 77 70 70 0.785 1.77 — 0.78 — 70 20 0.785 — 0.785 0.78 77 0.785 1.77 0.78 — 0.785 — 77 1.77 — — 0.785 0.00049 —

Sb (mm2)

13 Li QRE Ag QRE Li QRE Li QRE Ag QRE Li QRE 11 13 12 6 6 13 Ag QRE Ag QRE 13 Ag QRE 13 Ag QRE 13 Pt QRE 12 15 Ag QRE Ag QRE Pt QRE Pt QRE Ag QRE Ag QRE Pt QRE

REc

50 20 20 20 20 50 50 50 50 50 100 100 50 5 50 50 20 50 100 50 10 50 — 20 100 10 10 50 100 10

Scan Rate (mV s−1)

10 1.0 0.02 1.0 0.32 0.5 0.5 1.0 10 1.0 0.7 0.7 5 1.0 0.5 10 0.02 0.5 0.1 10 0.5 0.5 0.5 0.02 0.1 0.5 0.5 0.5 0.5 0.5

Cut-off Current (mA cm−2)

— −2.0

−2.1 −2.6 −2.0 −1.5

−2.6 0.2 −1.9 0.3 0.5 −1.4 1.2 −2.2 −2.0 −2.0 −2.1 −2.0 −3.0 −2.0 — −2.1 −2.0 −2.3 −1.8 −2.1

V vs. RE

TABLE 4.5. Electrochemical Windows of RTIL Based on Nonchloroaliminate System

−1.9

−1.5 −1.9 (0.9)g

−1.9 −2.3

−2.4

(1.1)g (0.8)g −2.3

−2.5 −2.4

(V vs. Fc/Fc+)

ECLd

— 1.7

2.0 1.9 2.1 1.5

2.3 5.4 2.3 5.7 5.4 2.6 5.5 2.1 2.2 2.1 — 2.4 1.4 2.4 — 2.5 2.1 2.2 2.0 1.4

V vs. RE

2.1

1.9 2.1 (5.5)g

2.1 1.9

2.1

(5.5)g (5.4)g 2.3

— 2.0

(V vs. Fc/Fc+)

EALe

4.4 4.4 4.4 4.6 4.6 4.1 4.5 3.8 3.5 3.9 4.1 4.5 4.1 3.1 3.4 4.0 4.6 3.7 4.1

4.9 5.2 4.2 5.4 4.9 4.0 4.4 4.3 4.2 4.1

EW Rangef (V)

(Continued)

30 38 12 38 38 61 62 29 30 29 32 32 31 63 33 30 12 64 40 30 37 35 65 12 47 37 37 33 20 37

Reference

58

c04.indd 58

1/20/2011 10:26:59 AM

BF4 BF4 F(HF)2.3 F(HF)2.3 TFSI TFSA PTS TFSI TFSI TFSI TFSI TSAC TFSI TFSI TFSI TFSI

Anion

Sb (mm2)

0.785 0.002 — — 0.00049 1.77 1.77 0.785 0.785 0.785 0.785 0.785 0.785 0.785 0.785 0.785

WEa

GC Pt micro GC GC Pt micro Pt Pt GC GC GC GC GC GC GC GC GC

Ag QRE Pt QRE Pt QRE Pt QRE Ag QRE Ag QRE Ag QRE 13 13 13 13 13 13 13 13 13

REc

50 100 10 10 100 100 100 50 50 50 50 50 50 50 50 50

Scan Rate (mV s−1)

0.5 0.02 0.5 0.5 0.5 0.5 0.1 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5 0.5

Cut-off Current (mA cm−2)

−3.5 −1.9 −1.3 −3.2 −3.0 −2.7 −2.3 −2.4 −2.1 −2.3 −1.3 −2.0

−1.8

V vs. RE

−1.4 −2.1

−2.4 −2.6

−3.3 −3.1

−2.1 −2.4

(0.73)g

(V vs. Fc/Fc+)

ECLd

1.7 2.8 1.6 2.7 2.6 2.5 2.3 2.2 2.7 2.6 2.5 1.5

2.6

V vs. RE

2.4 1.4

2.2 2.1

2.5 2.5

2.1 2.1

(5.4)g

(V vs. Fc/Fc+)

EALe

4.7 4.4 4.2 4.5 5.2 4.7 2.9 5.8 5.6 5.2 4.6 4.6 4.8 4.9 3.9 3.5

EW Rangef (V) 33 19 37 37 20 40 40 40 64 64 64 35 66 66 64 42

Reference

Note: DMEI: 1,2-dimethyl-3-ethylimidazolium; DMPI: 1,2-dimethyl3-propylimidazolium; EMI: 1-ethyl-3-methylimidazolium; DMFP: 1,2-dimethyl-4fluoropyrazolium; EMP: 1-ethyl-2-methylpyrazolium; DMI: 1,3-dimethylimidazolium; PrMI: 1,propyl-3-methylimidazolium; BMI: 1-butyl-3-methylimidazolium; PeMI: 1-pentyl-3-methylimidazolium; HMI: 1-hexyl-3-methylimidazolium; TBMA: tributylmethylammonium; P13: N-methyl-N-propyl-pyrrolidinium; PP13: N-methyl-N′-propylpiperidinium; TMMMA: trimethylmethoxymethylammonium; TMPhA: trimethylphenylammonium; TeEA: tetraethylammonium; TES: triethylsulfonium; TBS: tributylsulfonium; BP: N-butylpyridinium; DMABP: p-dimethylamino-N-butylpyridinium. Anions: TFSI: bis (trifluoromethylsulfonyl) imide, Methide: tris (trifluoromethylsulfonyl) methane; TFSA: bis (trifluoromethylsulfonyl) amide (=TFSI); TSAC: 2,2,2-trifluoro-N-(trifluoromethylsulfonyl) acetamide; BETI: bis (perfluoroethylsulfonyl) imide; DCA: dicyanamide; PTS: p-toluenesulfonate. a Working electrode. b Area of working electrode. c Reference electrode (see Table 4.1). d Cut-off current. e Cathodic limiting potential. f Anordic limiting potential. g Range = EAL–ECL. h These potential data (V vs. Li/Li+) are recorrected with the potential of Li/Li+ in each RTILs.

BMI BMI PeMI HMI TBMA P13 P13 PP13 TMPA TMMMA TMPhA TeEA TES TBS BP DMABP

Cation

TABLE 4.5. (Continued)

MUTUAL COMPARISONS OF ELECTROCHEMICAL WINDOWS OF RTILS

59

ment conditions become apparent from the data obtained in the same RTILs such as EMI–BF4 and EMI–TFSI. The potential data obviously differ based on the type of RE. Furthermore, the range of the EW decreases with the decreasing current density, which is used to estimate both the cathodic and the anodic limiting potentials as stated in Section 4.2.1. For these reasons, comparison of the EW data can contain relatively large errors aside from the effects of the measurement conditions such as the scan rate and the current density used for the determination of the limiting potential. As was stated previously, ferrocene is considered to be a potential internal reference for chloroaluminate systems. Only a few studies mention the redox potential of ferrocene in nonchloroaluminate systems. Table 4.5 gives the potential data corrected with the redox potential of ferrocene in each RTIL, if available. Figure 4.7 shows the EW of the RTILs. Shown in the figure are the EWs of both chloroaluminate (solid line) and nonchloroaluminate (dotted line) systems. It is interesting to note that both the cathodic and the anodic

Figure 4.7. Electrochemical windows for RTILs using the ferrocene internal reference. Solid line: nonchloroaluminate system; dotted line: chloroaluminate system. Working electrode: GC (glassy carbon); W (tungsten); Pt (platinum).

c04.indd 59

1/20/2011 10:26:59 AM

60

ELECTROCHEMICAL WINDOWS OF RTILs

limiting potentials of the RTILs based on the same cation are the same whether the system is chloroaluminate or nonchloroaluminate. This fact suggests ferrocene to be a valid internal reference for comparing the EWs of different RTILs. In comparisons using the ferrocene internal reference, the origin of the cathodic and anodic limiting potentials of the RTILs can be described as follows: 1. The cathodic limiting potential of the RTILs basically is determined by the reduction of the cations. 2. However, if the cathodic stability of the anion is poor, then the cathodic limiting potential is determined by the anion reduction and not the cation reduction as shown for the thiosaccharinato (TSAC) anion. In comparing the cathodic and anodic limiting potentials of PP13 or TMPA–TFSI with that of TeEA–TSAC, the TSAC is found to determine not only the cathodic limiting potential but also anodic limiting potential. 3. The anodic limiting potential of RTILs basically is determined by the oxidation of anions. However, the anodic limiting potential of EMI and DMABP is determined not by the oxidation of the anions but by the cation because the anodic limiting potential of the EMI and DMABP system is independent of the anions. 4. The EW of aliphatic quaternary ammonium systems based on fluoroanions has the widest EW compared with that of aromatic systems. This result can be explained by the presence of quaternary ammonium salts, which usually are used as a supporting electrolyte in the conventional electrolyte and which have wide electrochemical windows.

4.4

CONCLUSION

This chapter described the effects of voltammetric measurement conditions on the electrochemical window. With the use of ferrocene as an internal reference, a comparison between the electrochemical windows studied so far could be carried out to a greater degree (±0.1 V). Other possible comparisons of EW data that have not been described in the text are as follows: 1. The effect of a surface passivation layer simultaneously generated with the decomposition of the RTILs, with respect to the effect of the electrode material on the cathodic or anodic limiting potential 2. A comparison of the EW data of the RTILs with that of conventional solvent systems with respect to the effect of the interaction between cations and anions on the cathodic or anodic limiting potential

c04.indd 60

1/20/2011 10:26:59 AM

REFERENCES

61

These investigatiions can help elucidate the origin of the electrochemical windows of RTILs. The purity of the RTILs always must be a major consideration when discussing the electrochemical window of RTILs. REFERENCES 1. A. Webber, G. E. Blomgren, in Advances in Lithium-Ion Batteries, W. van Schalkwijk, and B. Scrosati, eds., Kluwer Academic, 2002, p. 185. 2. F. H. Hurley, T. P. Wier Jr., J. Electrochem. Soc., 98, 203 (1951). 3. H. L. Chum, V. R. Koch, L. L. Miller, R. A. Osteryoung, J. Am. Chem. Soc., 97, 3264 (1975). 4. J. S. Wikes, J. A. Levisky, R. A. Wilson, C. L. Hussey, Inorg. Chem., 21, 1263 (1982). 5. G. T. Cheek, R. A. Osteryoung, Inorg. Chem., 21, 3581 (1982). 6. M. Lipsztajin, R. A. Osteryoung, J. Electrochem. Soc., 130, 1968 (1983). 7. G. E. Blomgren, S. D. Jones, in Molten Salts, vol. 92–16, R. J. Gale, G. Blomgren, and H. Kojima, eds., Electrochemical Society, 1992, p. 379. 8. M. Lipsztajn, R. A. Osteryoung, Electrochim. Acta, 29, 1349 (1984). 9. J. Xie, T. L. Riechel, J. Electrochem. Soc., 145, 2660 (1998). 10. L. Xiao, K. E. Johnson, J. Electrochem. Soc., 150, E307 (2003). 11. M. Ue, M. Takeda, M. Takehara, S. Mori, J. Electrochem. Soc., 144, 2684 (1997). 12. A. B. McEwen, H. L. Ngo, K. LeCompte, J. L. Goldman, J. Electrochem. Soc., 146, 1687 (1999). 13. K. Xu, M. S. Ding, T. R. Jow, J. Electrochem. Soc., 148, A267 (2001). 14. R. Hagiwara, Y. Ito, J. Fluorine. Chem., 99, 1 (1999). 15. J. Heinze, K. Borgwarth, in Electrochemical Microsystem Technologies, J. W. Schultze, T. Osaka, and M. Datta, eds., Taylor and Francis, 2002, p. 32. 16. A. J. Bard, L. R. Faulkner, Electrochemical Methods, Wiley, 1980, p. 280. 17. R. T. Carlin, R. A. Osteryoung, J. Electroanal. Chem., 252, 81 (1988). 18. P. A. Z. Suarez, V. M. Selbach, J. E. L. Dullius, S. Einloft, C. M. S. Piatnicki, D. S. Azambuja, R. F. de Souza, J. Dupont, Electrochim. Acta., 42, 2533 (1997). 19. U. Schroder, J. D. Wadhawan, R. G. Compton, F. Marken, P. A. Z. Suarez, C. S. Consorti, R. F. de Souza, J. Dupont, New J. Chem., 24, 1009 (2000). 20. B. M. Quinn, Z. Ding, R. Moulton, A. J. Bard, Langmuir, 18, 1734 (2002). 21. J. Robinson, R. A. Osteryoung, J. Am. Chem. Soc., 101, 323 (1979). 22. R. J. Gale, R. A. Osteryoung, Inorg. Chem., 18, 1603 (1979). 23. C. L. Hussey, T. B. Scheffler, J. S. Wilkes, A. A. Fannin Jr., J. Electrochem. Soc., 133, 1389 (1986). 24. C. Scordilis-Kelley, J. Fuller, R. T. Carlin, J. S. Wilkes, J. Electrochem. Soc., 139, 694 (1992). 25. J. R. Stuff, S. W. Lander, J. W. Rovang, J. S. Wilkes, J. Electrochem. Soc., 137, 1492 (1990). 26. P. R. Gifford, J. B. Palmisano, J. Electrochem. Soc., 134, 610 (1987). 27. G. E. Gray, J. Winnick, P. A. Kohl, J. Electrochem. Soc., 143, 2262 (1996).

c04.indd 61

1/20/2011 10:26:59 AM

62

ELECTROCHEMICAL WINDOWS OF RTILs

28. G. E. Gray, J. Winnick, P. A. Kohl, J. Electrochem. Soc., 143, 3820 (1996). 29. E. I. Cooper, E. J. M. O’Sullivan, in Molten Salts, vol. 92–16, R. J. Gale, G. Blomgren, and H. Kojima, eds., Electrochemical Society, 1992, p. 386. 30. P. Bonhôte, A. Dias, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). 31. Y. Katayama, S. Dan, T. Miura, T. Kishi, J. Electrochem. Soc., 148, C102 (2001). 32. J. Fuller, R. T. Carlin, R. A. Osteryoung, J. Electrochem. Soc., 144, 3881 (1997)! 33. T. Nishida, Y. Tashiro, M. Yamamoto, J. Fluorine. Chem., 120, 135 (2003). 34. A. K. Covington, Pure Appl. Chem., 57, 531 (1985). 35. H. Matsumoto, H. Kageyama, Y. Miyazaki, in Molten Salts XIII, vol. 2002-19, P. C. Trulove, H. C. DeLong, R. A. Mantz, G. R. Stafford, and M. Matsunaga, eds., Electrochemical Society, 2002, p. 1057. 36. H. Sakaebe, H. Matsumoto, Electrochem. Commun., 5, 594 (2003). 37. R. Hagiwara, K. Matsumoto, Y. Nakamori, T. Tsuda, Y. Ito, H. Matsumoto, K. Momota, J. Electrochem. Soc., 150, D195 (2003). 38. V. R. Koch, L. A. Dominey, C. Nanjundiah, M. J. Ondrechen, J. Electrochem. Soc., 143, 798 (1996). 39. J. Sun, M. Forsyth, D. R. MacFarlane, J. Phys. Chem. B, 102, 8858 (1998). 40. J. N. Barisci, G. G. Wallace, D. R. MacFarlane, R. H. Baughman, Electrochem. Commun., 6, 22 (2004). 41. S. Sahami, R. A. Osteryoung, Anal. Chem., 55, 1970 (1983). 42. H. Matsumoto, (unpublished work). 43. J. S. Wilkes, Green Chem., 4, 73 (2002). 44. B. Vestergaard, N. J. Bjerrum, I. Petrushina, H. A. Hjuler, R. W. Berg, M. Begtrup, J. Electrochem. Soc., 140, 3108 (1993). 45. S. D. Jones, G. E. Blomgren, J. Electrochem. Soc., 136, 424 (1989). 46. A. P. Abbott, C. A. Eardley, N. R. S. Farley, G. A. Griffith, A. Pratt, J. Appl. Electrochem., 31, 1345 (2001). 47. S. D. Jones, G. E. Blomgren, in Molten Salts, vol. 90-17, C. L. Hussey, S. N. Flengas, J. S. Wilkes, and Y. Ito, eds., The Electrochemical Society, 1990, p. 273. 48. S. P. Wicelinski, R. J. Gale, J. S. Wilkes, J. Electrochem. Soc., 134, 263 (1986). 49. S. D. Williams, J. P. Schoebrechts, J. C. Selkirk, G. Mamantov, J. Am. Chem. Soc., 109, 2218 (1987). 50. J. A. Boon, J. S. Wilkes, J. A. Lanning, J. Electrochem. Soc., 138, 465 (1991). 51. Z. J. Karpinski, R. A. Osteryoung, Inorg. Chem., 23, 1491 (1984). 52. T. J. Melton, J. Joyce, J. T. Maloy, J. A. Boon, J. S. Wilkes, J. Electrochem. Soc., 137, 3865 (1990). 53. E. I. Cooper, C. A. Angell, Solid State Ionics, 18, 570 (1986). 54. R. Hagiwara, Y. Ito, J. Fluorine Chem., 105, 221 (2000), and references therein. 55. D. R. MacFarlane, P. Meakin, J. Sun, N. Amini, M. Forsyth, J. Phys. Chem. B, 103, 4164 (1999). 56. D. R. MacFarlane, J. Golding, S. Forsyth, M. Forsyth, G. B. Deacon, Chem. Commun., 16, 1430 (2001). 57. D. R. MacFarlane, S. Forsyth, J. Golding, G. B. Deacon, Green Chem., 4, 444 (2002).

c04.indd 62

1/20/2011 10:26:59 AM

REFERENCES

58. 59. 60. 61.

62.

63. 64.

65. 66.

c04.indd 63

63

S. Forsyth, D. R. MacFarlane, J. Mater. Chem., 13, 2451 (2003). H. Matsumoto, H. Kageyama, Y. Miyazaki, Chem. Commun., 16, 1726 (2002). Z. B. Zhou, M. Takeda, M. Ue, J. Fluorine Chem., 123, 127 (2003). J. Caja, T. D. J. Dunstan, D. M. Ryan, V. Katovic, in Molten Salts XII, vol. 99-41, P. C. Trulove, H. C. DeLong, G. R. Stafford, and S. Deki, eds., Electrochemical Society, 2000, p. 150. J. Caja, T. D. J. Dunstan, V. Katovic, in Molten Salts XIII, vol. 2002-19, P. C. Trulove, H. C. DeLong, R. A. Mantz, G. R. Stafford, and M. Matsunaga, eds., Electrochemical Society, 2002, p. 1014. H. Nakagawa, S. Izuchi, K. Kuwana, T. Nukuda, Y. Aihara, J. Electrochem. Soc., 150, A695 (2003). H. Matsumoto, M. Yanagida, K. Tanimoto, T. Kojima, Y. Tamiya, Y. Miyazaki, in Molten Salts XII, vol. 99-41, P. C. Trulove, H. C. DeLong, G. R. Stafford, and S. Deki, eds., Electrochemical Society, 2000, p. 186. K. Matsumoto, R. Hagiwara, Y. Ito, J. Fluorine. Chem., 115, 133 (2002). H. Matsumoto, T. Matsuda, Y. Miyazaki, Chem. Lett., 12, 1430 (2000).

1/20/2011 10:26:59 AM

5 DIFFUSION IN IONIC LIQUIDS AND CORRELATION WITH IONIC TRANSPORT BEHAVIOR Md. Abu Bin Hasan Susan, Akihiro Noda, and Masayoshi Watanabe

5.1

DIFFUSION AND DIFFUSIVITY—FUNDAMENTAL ASPECTS

Diffusion, which occurrs in essentially all matter, is one of the most ubiquitous phenomena in nature. It is the process of transport of materials driven by an external force field as well as the gradients of pressure, temperature, and concentration. It is the net transport of material that occurs within a single phase in the absence of mixing either by mechanical means or by convection. The rates of different technical as well as many physical, chemical, and biological processes are influenced directly by diffusive mass transfer, and the efficiency and quality of processes are governed by diffusion [1]. The intrinsic nature of particles to perform a perpetual, irregular movement is referred to as Brownian motion, which provides the basic mechanism for diffusion [1]. Because the random fluctuations in the positions of particles in space are often translational, the kinetics of these processes can be considered comparable with the decay of a concentration gradient by translational Brownian motion. Likewise, as the orientation of any molecule undergoes similar random fluctuations in space, rotational motion also occurs. In the presence of an external field, the extent of rotational order produced depends on a balance between the magnitude and the direction of the applied field and the intensity of the rotational motion, which in turn depends on the thermal

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

65

c05.indd 65

1/20/2011 10:27:09 AM

66

CORRELATION WITH IONIC TRANSPORT BEHAVIOR

energy of the system. Removal of the field causes the rotational order [2]. The rotational and translation motion can be separated, and it may be assumed that the rotational motion occurs through a series of small displacement steps, each of which occurs rapidly [2]. The discussion in this chapter is limited to translational diffusion in isothermal, isobaric systems with no external force field gradients, and the focus is on diffusion studies under only the concentration gradient. To get a sense of the physics behind diffusion, diffusion fluxes and diffusion potentials need to be defined in terms of driving forces. Fick’s first law is used to correlate the diffusion flux density with the concentration gradient of the diffusants, which yields the diffusion coefficient as the factor of proportionality. For a system with two components and one-dimensional diffusion in the z-direction, the particle flux ji is related to the gradient of concentration, δ ci / δ z, of these particles as ji = − Di

δ ci δz

(i = 1, 2)

(5.1)

where Di is the diffusion coeffcient or diffusivity. Combining equation (5.1) with the law of matter conservation,

δ ci δ [Di (δ ci / δ z)] + =0 δt δz

(5.2)

results in Fick’s second law:

δ ci δ 2c = Di 2i δt δz

(5.3)

This partial differential equation simplifies if Di is constant and can be solved for particular initial and boundary conditions. This allows determining Di from measurements of the concentration distribution as a function of position and time.

5.2 THE PROCESSES OF DIFFUSION IN LIQUIDS Molecules in liquid, as compared with gases, are densely packed and are strongly affected by force fields of neighboring molecules, and the values of diffusivity for liquids are much smaller than those for low-pressure gases. The mass transfer by diffusion, however, depends on the concentration gradient, and it does not have to be low because, for a large concentration gradient, the transfer rates can be high [3]. In a liquid system, translational diffusion is the most fundamental form of transport and is responsible for all chemical reactions because the reacting species must collide before they can react.

c05.indd 66

1/20/2011 10:27:09 AM

THE PROCESSES OF DIFFUSION IN LIQUIDS

67

Translational diffusion can take a different form, and for a systematic description, at least the following two elementary categories need to be considered: self-diffusion and mutual diffusion [2–5]. In a liquid that is in thermodynamic equilibrium, the individual particles move with random translational motion, and they have an equal opportunity of taking up any point in the total space occupied by the liquid. The motion, characterized as a random walk of the particles, is self-diffusion. If it becomes possible to label a particle without otherwise changing its properties and to follow its motion through the unlabeled molecules, then a self-diffusion coefficient can be described in terms of concentration gradient and can be defined as in equation (5.3) [2]. A particle under such conditions can be tagged, and its trajectory can be followed over a long time. In a medium without long-range correlations, the squared displacement of the species from its original position eventually will grow linearly with time. The molecular displacements of a single particle then quantify self-diffusion. To reduce statistical errors, the selfdiffusivity, D, of a species is defined by averaging Einstein’s equation over a large enough number of molecules, N [1,4]: 1 ⎞ D=⎛ ⎝ 6N ⎠

N

1 2 rk (t ) − rk (0)

∑ lim t k =1

t →∞

(5.4)

where rk (t ) and rk (0) are the position of the kth molecule of species i at time t and 0. Because species diffuse independently of each other, only one species is considered and the index i generally is dropped in the notations. However, self-diffusion is not limited to one-component systems. The random walk of particles of each component can be observed in any composition of a mixture containing multiple components. Self-diffusivity is a measure of the translational mobility of individual molecules (or ions) driven by an internal kinetic energy. It therefore provides a molecular description of matter—liquids in this case. Self-diffusion can be combined with other data such as viscosities, electrical conductivities, and densities for evaluating and improving solvodynamic models, such as the Stokes–Einstein type, and for providing a deep insight into molecular transport in liquid systems [4]. The study of temperature dependencies of the selfdiffusion coefficient, as well as those of other physical properties, allows for calculating activation energies of the system. The temperature dependencies help to propose diffusion models and to confirm the validity of experimental and computational data on single and multicomponent systems of electrolytes and nonelectrolytes. The temperature dependencies in most cases follow Arrhenius behavior, with some obeying the type of the Vogel–Fulcher–Tamman (VFT) equation [6]. If a liquid system containing at least two components is not in thermodynamic equilibrium as a result of concentration inhomogeneities, then transport of matter occurs. This process widely is called mutual diffusion, and it denotes the macroscopically perceptible relative motion of the individual particles

c05.indd 67

1/20/2011 10:27:09 AM

68

CORRELATION WITH IONIC TRANSPORT BEHAVIOR

resulting from concentration gradients. The process is also known as chemical diffusion, interdiffusion, transport diffusion, and, in the case of systems with two components, binary diffusion. Whereas mutual diffusion characterizes a system with a single diffusion coefficient, self-diffusion gives different diffusion coefficients for all particles in the system. Self-diffusion thereby provides a more detailed description of the single chemical species. This is the molecular point of view [7], which makes the self-diffusion more significant than that of the mutual diffusion. In contrast, in practice, mutual diffusion, which involves the transport of matter in many physical and chemical processes, is far more important than self-diffusion. Moreover, mutual diffusion is cooperative by nature, and its theoretical description is complicated by nonequilibrium statistical mechanics. Not surprisingly, the theoretical basis of mutual diffusion is more complex than that of self-diffusion [8]. In addition, by definition, the measurements of mutual diffusion require mixtures of liquids, whereas self-diffusion measurements are determinable in pure liquids.

5.3

SELF-DIFFUSION AND IONIC TRANSPORT IN IONIC LIQUIDS

Regardless of the stochastic nature of its elementary steps, diffusion follows well-defined dependencies [1]. The wealth and beauty of these dependencies is particularly impressive when diffusion occurs in concentrated electrolyte solutions like ionic liquids. Owing to their structural variability and importance, ionic liquids represent a particularly attractive system for the study of self-diffusion and ionic transport behavior. Ionic liquids, as the name implies, consist entirely of ions. These are receiving an upsurge of interest for potential use in multidisciplinary areas because of, inter alia, their immeasurably low vapor pressure, high ionic conductivity, nonflammability, and greater thermal and electrochemical stability [9–30]. Research to date includes numerous attempts to prepare ionic liquids based on new anions and cations with a view to realizing unique physicochemical properties. It already has been observed that certain combinations of organic cations, such as imidazolium and pyridinium derivatives, as well as bulky and soft anions, such as PF6− , BF4− , CF3SO−3 , and (CF3SO2 )2 N −, form ionic liquids at near ambient temperature [14–16,18,22–23]. However, the fundamentals of the ionic transport properties in ionic liquids still need to be explored systematically. Conceptual understanding of ionic transport behavior in ionic liquids is complicated not only by the various physical conditions under which the diffusion phenomena may appear, but it also is complicated by the fact that the spatial temporal ranges over which both diffusion phenomena are perceived by the different experimental techniques can vary dramatically. Impedance spectroscopy, meaning the measurement of complex resistivities with ac current methods, is an important tool to study diffusion and to cor-

c05.indd 68

1/20/2011 10:27:09 AM

PGSE-NMR FOR MEASUREMENTS OF SELF-DIFFUSION COEFFICIENTS

69

relate it with ionic transport behavior. The diffusion coefficient, Dσ , obtained from conductivity measurements (vide infra) is related to the self-diffusion coefficient, D (nuclear magnetic resonance [NMR] measurements) by D = H R* Dσ . Here, HR is the Haven ratio, which can give information on the ionicity of the ionic liquids [22]. The alternate way is to compare the molar conductivity calculated from the self-diffusion coefficients ( Λ NMR ) of ionic liquids with the experimental molar conductivity calculated from the ionic conductivity and the density (Λ). The ratio of the molar conductivities ( Λ/Λ NMR), the other form of HR, is indicative of the percentage of ions, which can contribute to ionic conduction [22]. Thus, the fraction of the ion pairs or aggregates may be estimated from the Haven ratio as 1 − Λ / Λ NMR. In this chapter, we deal with the ionic diffusion coefficient, the ionic association, and the interaction between ions in ionic liquids. We employ self-diffusivity to interpret the ionic transport behavior and aim at evaluating the ionicity and the ionic states of the ionic liquids.

5.4 PULSED-GRADIENT SPIN-ECHO (PGSE)-NMR FOR MEASUREMENTS OF SELF-DIFFUSION COEFFICIENTS The pulsed-gradient spin-echo-NMR (PGSE–NMR) method is noninvasive and provides a convenient means to measure the self-diffusion coefficient independently of each ionic species in the system of an ionic liquid, provided that the components contain NMR-sensitive nuclei. In general, the method involves the attenuation of a spin-echo signal resulting from the dephasing of the nuclear spins that comes from the combination of the translational motion of the spins under the imposition of spatially well-defined gradient pulses for the measurement of the displacement of the observed spins [31]. The PGSE experiment uses a spin-echo experiment coupled with two gradient pulses. By the subsequent application of these pulses, the magnetic field is made inhomogeneous over two short time intervals. The gradient pulses create a linear gradient magnetic field that varies the precessions of the nuclei in the field. The NMR signal becomes sensitive to the differences in the locations of the nuclei (and hence of the respective atoms and molecules) between the two field gradient pulses. These differences are revealed by analyzing the attenuation of the NMR signal as a function of the intensity of the gradient pulses. Self-diffusion causes an attenuation of the spin-echo amplitude as a result of incomplete refocusing. Differences in molecular positions at two subsequent intervals, however, represent nothing else than their displacements. Figure 5.1 represents schematic illustration of the simplest pulse sequence for measuring self-diffusion [4,31]. This is a simple Hahn spin-echo pulse sequence (90 °–τ–180 °–τ–acq) with a rectangular gradient pulse inserted into each τ delay. The time interval between the leading edges of the gradient pulses is denoted by Δ. After a 90 ° radio frequency (RF) pulse, the macroscopic magnetization is rotated from the z-axis into the x-y plane. When a

c05.indd 69

1/20/2011 10:27:09 AM

70

CORRELATION WITH IONIC TRANSPORT BEHAVIOR τ

τ πy

π /2x RF pulse

δ

Gradient pulse Acquisition

g

t1

t1 + Δ

E

Figure 5.1. Pulsed-gradient spin-echo-NMR sequence for the measurement of selfdiffusion coefficient.

gradient pulse of duration δ and magnitude g is applied, the spins dephase. The spin precession is reversed after the application of a 180 ° RF pulse after a time τ. A second gradient pulse of equal duration δ and magnitude g follows to tag the spins in the same way. On one hand, if the positions of the spins remain unchanged in the sample, then the effects of the gradient pulses compensate each other, and all spins refocus. On the other hand, if the selfdiffusion causes the spins to move and the effects of the gradient pulses do not compensate, the echo amplitude is reduced. The decrease of the amplitude with the applied gradient is proportional to the movement of the spins. The echo signal attenuation, E, is related to the experimental parameters by [32]. ⎛ S ⎞ δ ln(E) = ln ⎜ = −γ 2 g 2 dDδ 2 ⎛ Δ − ⎞ ⎟ ⎝ ⎝ Sg =0 ⎠ 3⎠

(5.5)

where S is the spin-echo signal intensity and γ is the gyromagnetic ratio. The PGSE measurement involves the collection of spin-echo amplitudes as the gradient pulse strengths are varied. The value of D can be easily determined from the slope of a plot of ln(S / Sg=0 ) versus −γ 2 g 2 Dδ 2 (Δ − δ / 3). Owing to the large γ value, protons offer the best measuring conditions with respect to both the minimum concentration of the diffusants and their minimum displacements [1]. However, PGSE-NMR diffusion studies can be performed with other nuclei including 13C, 15N, and 19 F. The lower limit of observable displacements is of the order of 100 nm. With an observation time of a few 100 ms as a typical upper value, this corresponds to minimum diffusivities of about 10−10 cm2 s−1.

5.5 IONIC LIQUIDS FOR DIFFUSION STUDIES AND IONIC TRANSPORT BEHAVIOR To reveal ionic transport behavior through self-diffusion studies [22], we choose four different ionic liquids: 1-ethyl-3-methylimidazolium tetrafluoroborate

c05.indd 70

1/20/2011 10:27:09 AM

71

IONIC LIQUIDS

N +

N N+

CH3

H3CH2C BF4− EMIBF4

N +

BF4−

CH2CH2CH2CH3 BPBF4

N

H3CH2C

N+

CH3

N−(SO2CF3)2

CH2CH2CH2CH3

N−(SO2CF3)2

BPTFSI

EMITFSI

Scheme 5.1. Structures of the ionic liquids, EMIBF4, EMITFSI, BPBF4, and BPTFSI.

TABLE 5.1. Thermal Properties for EMIBF4, EMITFSI, BPBF4, and BPTFSI

EMIBF4 EMITFSI BPBF4 BPTFSI

Tg/Ka

Tc/Ka

Tm/Kb

Td/Kb

181 187 202 —

193 218 251 224

279 257 272 299

664 717 615 677

a

Onset temperatures of a heat capacity change (Tg), an exotherm peak (Tc) and an endotherm peak (Tm) during heating scan from 123 K by using differential scanning calorimetry. b Temperature of 10% weight loss during heating scan from room temperature by using thermogravimetry.

(EMIBF4), 1-ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide (EMITFSI), 1-butylpyridinium tetrafluoroborate (BPBF4), 1-butylpyridinium, and bis(trifluoromethylsulfonyl)imide (BPTFSI) (Scheme 5.1). The ionic liquids consist two different cations and anions, and they allow us to visualize the dependence of ionic transport behavior on structural variety from the comparative self-diffusion behavior. The thermal properties of the ionic liquids, as determined by differential scanning calorimetry (DSC) and thermogravimetry/differential thermal analysis (TG–DTA), are summarized in Table 5.1. The ionic liquids investigated in this study are all liquid at ambient temperature and therefore are roomtemperature ionic liquids (RTILs). All RTILs have wide liquid temperature ranges together with a high thermal stability. In particular, EMITFSI [18,21] and BPTFSI exhibit higher thermal stability than EMIBF4 and BPBF4. During cooling from 373 K to 123 K at the rate of 10 K min−1, only a heat capacity change corresponding to the glass transition can be observed for EMIBF4 and BPBF4. This indicates that the crystallization rates of both ionic liquids are

c05.indd 71

1/20/2011 10:27:09 AM

72

CORRELATION WITH IONIC TRANSPORT BEHAVIOR

TABLE 5.2. Density and Molar Concentration for EMIBF4, EMITFSI, BPBF4, and BPTFSIa

T/K 313 308 303 298 293

T/K 313 308 303 298 293

EMIBF4

EMITFSI

BPBF4

BPTFSI

ρ/gcm−3 1.266 1.271 1.275 1.279 1.283

ρ/gcm−3 1.502 1.507 1.512 1.518 1.523

ρ/gcm−3 1.208 1.212 1.216 1.220 1.224

ρ/gcm−3 1.436 1.440 1.444 1.449 1.453

EMIBF4

EMITFSI

BPBF4

BPTFSI

M/moldm−3 6.39 6.42 6.44 6.46 6.48

M/moldm−3 3.84 3.85 3.86 3.88 3.89

M/moldm−3 5.42 5.43 5.45 5.47 5.49

M/moldm−3 3.45 3.46 3.47 3.48 3.49

Formula weight/gmol−1; EMIBF4 197.97; EMITFSI 391.32; BPBF4 223.01; BPTFSI 416.36.

a

slow and that the supercooled liquids are fairly stable [19,20]. For instance, EMIBF4 does not crystallize at 213 K for more than 5 hours, and BPBF4 does not crystallize at 243 K for more than 7 hours. In contrast, an exothermic peak based on the crystallization and a heat capacity change assigned to the glass transition have been observed for EMITFSI and BPTFSI during the cooling scans, indicating relatively fast crystallization rates of these ionic liquids. In the following heating scans, the thermograms show the glass transition temperature (Tg), crystallization temperature (Tc), and melting temperature (Tm), successively, which is a common feature for most ionic liquids [21]. The ionic liquids are dense systems, with the density exhibiting well-defined temperature dependencies. Table 5.2 shows the density of the ionic liquids in a range of temperatures. The density decreases linearly with the temperature increase, regardless of the ionic liquid structures. Based on the density data and the formula weight, the molar concentration of each ionic liquid also has been calculated and listed in Table 5.2. Molar concentrations of the ionic liquids are high, and they change with structures and temperature.

5.6 SELF-DIFFUSION COEFFICIENT AND ITS CORRELATION WITH VISCOSITY The value of the self-diffusion coefficient (D) for the ionic liquids has been determined from the slope of a plot of signal attenuation, S / Sg =0 versus −γ 2 g 2 Dδ 2 (Δ − δ / 3). In these measurements, 19 F and 1H are used for the detection of the anions and the cations, respectively. Figure 5.2 gives the results for

c05.indd 72

1/20/2011 10:27:10 AM

SELF-DIFFUSION COEFFICIENT AND ITS CORRELATION WITH VISCOSITY

73

0.0 373 K 353 K 333 K 313 K

−1.0

In (E )

−2.0

303 K 293 K 283 K 273 K 273 K

−3.0 −4.0 −5.0 −6.0 −7.0 0.0

1.0

2.0

3.0

4.0

5.0

y2 g2 δ2(Δ−δ/3) / 107 cm−25

Figure 5.2. Plots of signal attenuation versus γ 2 g 2δ 2 (Δ − δ / 3) for 1H PGSE–NMR echo signals of EMIBF4. In this experiment, g is 5.78Tm−1 , δ ranges from 0 to 3.5 ms, and Δ is 50 ms.

the 1H PGSE–NMR echo signal for EMIBF4 in terms of the relationship between the left-hand term and the right-hand term in equation (5.5). Although the linearity of the plots (Figure 5.2) is consistent with the measured species in free diffusion in the time scale of Δ (50 ms), the measured species, that is, each ion, interact with each other. The presence of a strong coulombic attractive force between the anions and the cations suggests that the ions associate to form ion pairs and ion aggregates. However, a coulombic repulsive force between the like charges also exists. The equilibrium between the attractive and the repulsive interaction at each temperature, while maintaining the charge valance, relates to the chemical equilibrium for association and dissociation of the components in the ionic liquids. Since NMR measurement can detect a nucleus (i.e., 1H or 19 F) but cannot distinguish between the ion and their associated species, the PGSE-NMR diffusion coefficients obtained in this experiment are an average self-diffusion coefficient of the ions and associated ions. In addition, because the signal splitting between the ions and their aggregates cannot be observed, the rate of exchange between the ions and the ion aggregates is shorter than the NMR time scale. Figure 5.3 depicts the Arrhenius plots of the apparent self-diffusion coefficient of the cation (Dcation) and the anion (Danion) for EMIBF4 and EMITFSI

c05.indd 73

1/20/2011 10:27:10 AM

74

CORRELATION WITH IONIC TRANSPORT BEHAVIOR (a) −5.0

log (D / cm2S-1)

−5.5

EMI in EMIBF4 EMI in EMITFSI

−6.0

−6.5

−7.0

−7.5 2.6

BF4 in EMIBF4 TFSI in EMITFSI 2.8

3.0

3.2

3.4

3.6

3.8

4.0

103 T -1 / K-1 (b) −5.5

log (D / cm2S-1)

−6.0

BP IN BPTFSI TFSI in BPTFSI

−6.5

−7.0

−7.5

−8.0 2.6

BP in BPBF4 BF4 in BPBF4 2.8

3.0

3.2

3.4

3.6

3.8

4.0

103 T -1 / K-1

Figure 5.3. Arrhenius plots of self-diffusion coefficients of the anions and cations for (a) EMIBF4 and EMITFSI and (b) BPBF4 and BPTFSI.

c05.indd 74

1/20/2011 10:27:10 AM

SELF-DIFFUSION COEFFICIENT AND ITS CORRELATION WITH VISCOSITY

75

(Figure 5.3a) and for BPBF4 and BPTFSI (Figure 5.3b). The Arrhenius plots of the summation (Dcation + Danion) of the cationic and anionic diffusion coefficients also are shown in Figure 5.4. The fact that the temperature dependency of each set of the self-diffusion coefficients shows convex curved profiles implies that the ionic liquids of our interest deviate from the ideal Arrhenius behavior. Each result of the self-diffusion coefficient therefore has been fitted with the VFT equation [6]. ⎛ −B ⎞ D = D0 exp ⎜ ⎝ T − T0 ⎟⎠

(5.6)

where D0 (cm2 s−1), B (K), and T0 (K) are constants. The solid lines in Figures 5.3a and b, and Figure 5.4 are the fitted profiles of the experimental points in equation (5.6). The best-fit parameters of the VFT equation for these ionic liquids are summarized in Table 5.3. The apparent cationic transference numbers (Dcation/(Dcation + Danion)) can be used for the comparison of the self-diffusion coefficients between the anion and the cation (Figure 5.5). The cationic diffusion coefficients for EMIBF4 and BPBF4 are similar to the anionic diffusion coefficient. When the anion is replaced by more bulky TFSI, the relative diffusivity of the cations to that of

−5.0

EMITFSI EMIBF4

log ((Dcation + Danion) / cm2S-1)

−5.5

−6.0

−6.5

−7.0 BPTFSI −7.5

BPBF4

−8.0 26

28

30

32

34

36

38

40

103 T -1 / K-1

Figure 5.4. Arrhenius plots of simple summation of cationic and anionic self-diffusion coefficients (Dcation + Danion) for EMIBF4, EMITFSI, BPBF4, and BPTFSI.

c05.indd 75

1/20/2011 10:27:10 AM

76

CORRELATION WITH IONIC TRANSPORT BEHAVIOR

TABLE 5.3. VFT Equation Parameters for Apparent Self-Diffusion Coefficient Data, D = D0 exp[ - B/(T - T0 )] D0/10−4 cm2 s−1

B/102 K

T0/102 K

R2/10−1

EMIBF4 Cation Anion Cation + anion

1.6 1.6 3.3

9.7 10 9.9

1.3 1.3 1.3

9.99 9.99 9.99

EMITFSI Cation Anion Cation + anion

1.7 1.7 3.3

9.4 11 9.9

1.3 1.2 1.3

9.99 9.99 9.99

BPBF4 Cation Anion Cation + anion

6.9 7.8 15

15 15 15

1.3 1.3 1.3

9.96 9.99 9.97

BPTFSI Cation Anion Cation + anion

2.5 5.3 6.6

11 14 12

1.4 1.2 1.3

9.99 9.98 9.99

0.70

Dcation / (Dcation+Danion)

0.65

0.60

0.55

0.50

0.45

0.40 −20

0

20

40

EMITFSI EMIBF4

BPTFSI BPBF4

60

100

80

120

T / °C

Figure 5.5. Temperature dependences of apparent cationic transference numbers for EMIBF4, EMITFSI, BPBF4, and BPTFSI.

c05.indd 76

1/20/2011 10:27:10 AM

SELF-DIFFUSION COEFFICIENT AND ITS CORRELATION WITH VISCOSITY

77

the anion becomes larger, and the B parameters of the VFT equation for the cation become smaller than those of the TFSI anion (Table 5.3). The larger B values for the BP cation compared with the EMI indicates that larger activation energies (larger B values) are required for the ionic liquids with a BP cation. The differences of the D0 and T0 parameters between the cation and the anion are small, and especially, T0 parameters are around 1.3 × 102 K for all ionic liquids. These results reveal that the differences of the ionic selfdiffusion coefficient at each temperature are dependent mainly on the B parameters (i.e., the activation energies). Viscosities of the ionic liquids at different temperature have been depicted in Figure 5.6 as Arrhenius plots. Resembling the self-diffusion behavior, the temperature dependence of viscosity follows the VFT equation: ⎛ B ⎞ η = η0 exp ⎜ ⎝ T − T0 ⎟⎠

(5.7)

where η0 (mPa s), B (K), and T0 (K) are constants. Solid lines in Figure 5.6 are the fitted profile with the best-fit VFT parameters of viscosity listed in Table 5.4. The temperature dependency of viscosity for EMIBF4 and EMITFSI is in good agreement with the literature [18,21]. The temperature dependency of the viscosity (Figure 5.6) and the summation of the self-diffusion coefficient (Dcation + Danion) (Figure 5.4) interestingly

3.0

log (η / mPa s)

2.5

BPBF4 BPTFSI

2.0

1.5

1.0

EMIBF4 EMITFSI

0.5 2.6

2.8

3.0

3.2

3.4

3.6

3.8

4.0

103 T -1 / K-1

Figure 5.6. Arrhenius plots of viscosity for EMIBF4, EMITFSI, BPBF4, and BPTFSI.

c05.indd 77

1/20/2011 10:27:10 AM

78

CORRELATION WITH IONIC TRANSPORT BEHAVIOR

TABLE 5.4. VFT Equation Parameters for Viscosity Data, h = h0 exp[ B/(T - T0 )]

EMIBF4 EMITFSI BPBF4 BPTFSI

η0/10−1 mPas

B/102 K

T0/102 K

R2/10−1

2.0 1.5 2.3 1.2

7.5 8.4 7.2 8.5

1.5 1.4 1.8 1.6

9.99 9.99 9.99 9.99

show the contrasted profiles with the indication of an inverse relationship between the viscosity and the self-diffusion coefficient. This can be explained in terms of the Stokes–Einstein equation, which correlates the self-diffusion coefficient (Dcation and Danion) with viscosity (η) by the following relationship: D=

kT cπηr

(5.8)

where k is the Boltzmann’s constant, T is the absolute temperature, c is a constant (4 to 6), and r is the effective hydrodynamic or Stokes radius. Figure 5.7 shows the relationships between the self-diffusion coefficient of the anion (Danion) and the cation (Dcation) and Tη −1 for EMIBF4 and EMITFSI (Figure 5.7a) and for BPBF4 and BPTFSI (Figure 5.7b). In all cases, the straight lines passing through the origin indicate that the ionic diffusivity in the ionic liquids obeys equation (5.8). However, it should be noted that the slopes of the plots do not reflect the size of each ionic species [21]. As shown in Figure 5.7a, the EMI cations in EMITFSI and EMIBF4 indicate similar and the largest slopes in the relationships followed by the BF4 anion and the TFSI anion. If these are explained by equation (5.8), then the hydrodynamic radius of each ion follows the order TFSI > BF4 > EMI and does not reflect the calculated size of each ionic species [21]. Figure 5.7b indicates that the BP cation, TFSI anion, and BF4 anion have a similar hydrodynamic radius, regardless of the difference in the calculated ion size. Because the ionic liquids are highly concentrated electrolyte solutions (Table 5.2), the diffusion of an ion in the liquids may be dependent on that of other species. The influence of coulombic interaction between ionic species (i.e., attractive and repulsive interaction) and the equilibrium between dissociated ions and associated ions also seem to be important for the relation between the ionic size and the diffusivity. It already has been shown that the ionic charge has a strong effect on the transport properties. Hussey et al. [33] reported that the increase of an overall negative anionic charge of metal complexes increases the relative hydrodynamic radii in chloroaluminate- or bromoaluminate-ionic liquids. It is interesting that the monomeric and polymeric anions exhibit almost identical hydrodynamic radii if their overall negative charge is the same. However, the ionic liquids in our study consist of only singly charged anions and cations. The apparent cationic transference number shows that the cation diffuses a little faster than the

c05.indd 78

1/20/2011 10:27:10 AM

SELF-DIFFUSION COEFFICIENT AND ITS CORRELATION WITH VISCOSITY

79

(a) 2.5

D / 10-6 cm2S-1

2.0

1.5

1.0

EMI in EMITFSI TFSI in EMITFSI EMI in EMIBF4 BF4 in EMIBF4

0.5

0.0

0.0 0.5

1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 η-1 T / 10 mPa-1s-1K

(b) 1.6 1.4

D / 10-6 cm2S-1

1.2 1.0 0.8 0.6 0.4

BP in BPTFSI TFSI in BPTFSI BP in BPBF4 BF4 in BPBF4

0.2 0.0 0.0

0.5

1.0

1.5

2.0

2.5

3.0

3.5

η T / 10 mPa s K -1

-1 -1

−1

Figure 5.7. Relationship between Tη (T: absolute temperature, η: viscosity) and selfdiffusion coefficients of anion and cation for (a) EMIBF4 and EMITFSI and for (b) BPBF4, and BPTFSI.

c05.indd 79

1/20/2011 10:27:10 AM

80

CORRELATION WITH IONIC TRANSPORT BEHAVIOR

anion, with the exception of BPBF4. In addition, the results of the Stokes– Einstein relationship (Figure 5.7) show the cationic hydrodynamic radii to be relatively smaller than the anionic radii for both TFSI and BF4 anions in our ionic liquids. If the calculated van der Waals radius is considered, then BF4 will have the smaller value compared with that of EMI and BP. Consequently, the cation will diffuse more easily than the anion. Owing to their resonance ring structures, the delocalization of cationic charge as well as the planar structures seems to be important.

5.7 MOLAR CONDUCTIVITY AND ITS CORRELATION WITH DIFFUSION COEFFICIENT The temperature dependence of molar conductivity, calculated from ionic conductivity determined from complex impedance measurements and molar concentrations, and the VFT fitting curves are shown in Figure 5.8. The VFT equation for molar conductivity is ⎛ −B ⎞ Λ = Λ 0 exp ⎜ ⎝ T − T0 ⎟⎠

(5.9)

1.2 EMIBF4 EMITFSI

0.8

log (Λ / Scm2mol-1)

0.4 0.0 −0.4 −0.8

BPBF4 BPTFSI

−1.2 −1.6 2.6

2.8

3.0

3.2 3

10 T

3.4 -1

3.6

3.8

4.0

-1

/K

Figure 5.8. Arrhenius plots of molar conductivity (Λ, obtained experimentally from conductivity and molar concentration) for EMIBF4, EMITFSI, BPBF4, and BPTFSI.

c05.indd 80

1/20/2011 10:27:10 AM

81

MOLAR CONDUCTIVITY AND ITS CORRELATION

TABLE 5.5. VFT Equation Parameters for Molar Conductivity Data Λ0/10 Scm−2 mol−1 Λ = Λ0 exp [−B/(T − T0)] EMIBF4 EMITFSI BPBF4 BPTFSI

13 4.9 19 5.0

Λ0/10 Scm−2 mol−1 ΛNMR = Λ0 exp [−B/(T − T0)] EMIBF4 4.5 EMITFSI 4.5 BPBF4 20 BPTFSI 9.2

B/102 K

T0/102 K

R2/10−1

5.0 3.9 6.8 4.3

1.7 1.9 1.8 2.0

9.99 9.99 9.99 9.99

B/102 K

T0/102 K

R2/10−1

7.6 7.6 13 9.9

1.4 1.4 1.4 1.4

9.99 9.99 9.97 9.99

where Λ0 (Scm−2 mol−1), B (K), and T0 (K) are constants. The best-fit parameters for the VFT equation are tabulated in Table 5.5. The Nernst–Einstein equation is applied to calculate the molar conductivity (ΛNMR) from the PGSE–NMR diffusion coefficients: Λ NMR =

N Ae 2 (Dcation + Danoin ) F 2 (Dcation + Danion ) = kT RT

(5.10)

where NA is the Avogadro’s number, e is the electric charge on each ion, F is the Faraday’s constant, and R is the universal gas constant. It should be noted that equation (5.10) is derived without considering ionic association and is derived on the assumption that each ion has the activity of unity. In the temperature range studied, the Λ values are lower than the ΛNMR values at each temperature. The Λ/ΛNMR thus is plotted against temperature in Figure 5.9. The Λ/ΛNMR ratios range from 0.6 to 0.8 for EMIBF4 and BPBF4, whereas the ratio ranges from 0.3 to 0.5 for EMITFSI and BPTFSI. Depending on the ionic liquid, the Λ/ΛNMR differs, indicating the difference in the percentage of ions contributing to ionic conduction in the diffusion components. Thus, the fraction of the ion pairs or aggregations (1 − Λ/ΛNMR) also significantly differs. However, NMR cannot distinguish between the ions and the ion aggregates, and their exchange rate is faster than the NMR time scale (vide supra). In this case, the experimental results are average, and the percentage of ions or aggregations is also of the “time-averaged ions” or the “time-averaged aggregations,” in other words, the “time-averaged solvation,” as suggested by Hussey et al. [33]. We recognize that the percentage of the ions can be defined as the existence probability if NMR and conductivity measurements can be conducted on the same time scale.

c05.indd 81

1/20/2011 10:27:10 AM

82

CORRELATION WITH IONIC TRANSPORT BEHAVIOR 0.8 EMIBF4 BPBF4 0.7

Λ / ΛNMR

0.6

0.5

EMITFSI BPTFSI

0.4

0.3 −20

0

20

40

60

80

100

120

T / °C

Figure 5.9. Molar conductivity ratios (Λ/ΛNMR) plotted against temperature for EMIBF4, EMITFSI, BPBF4, and BPTFSI.

A large value of Λ/ΛNMR refers to a high ionicity of an ionic liquid and indicates that the ionic liquid consists mainly of charged cationic and anionic species, which can diffuse and contribute to the ionic conduction. The Λ/ΛNMR ratio for EMIBF4 and BPBF4 is relatively high enough to consider these ionic liquids to consist mostly of the ions. This suggests that NMR diffusion coefficients in these cases reflects the individual ionic diffusivity, and most ions contribute to the ionic conduction. In other words, the activity of each ion is close to unity. The situation might not be so straightforward and become a little complicated in EMITFSI and BPTFSI. The Λ/ΛNMR ratio is evidently smaller than those for EMIBF4 and BPBF4. The summation of the self-diffusion coefficient (Dcation + Danion) at each temperature (Figure 5.4) follows the order EMITFSI > EMIBF4 > BPTFSI > BPBF4, which reflects the magnitude of the viscosity (Figure 5.6). However, the high diffusivity for EMITFSI and BPTFSI is not reflected in their molar conductivity (Figure 5.8). This is the key factor for the small Λ/ΛNMR ratio. The viscosity of the ionic liquids seems to be influenced by many parameters, such as formula weight and shape of ions and coulombic interaction between the ionic species. A comparison of the viscosity of the ionic liquids with EMI cation and with BP cation shows that the larger cationic size and the larger cationic formula weight can cause a larger viscosity of the latter ionic liquids. Although the formula weight of the TFSI anion is much larger than that of the BF4 anion, the viscosity of the ionic liquids with

c05.indd 82

1/20/2011 10:27:10 AM

ACKNOWLEDGMENTS

83

the TFSI anion is lower than that of the BF4 anion. The consideration of anionic structures thus invalidates the effect of formula weight on the viscosity. This may be attributed to the presence of ion pairs and neutral ion aggregates in the ionic liquids with TFSI anion, which dilutes the ionic concentration and reduces coulombic interaction between the ionic species. The reduced coulombic interaction may decrease the viscosity, which in turn enhances the ionic diffusion coefficient. However, the ion pairs and neutral ion aggregates cannot contribute to the ionic conduction, and thus, the molar conductivity of the ionic liquids with the TFSI anion becomes small. Certain molecular interaction between the cations and the TFSI anion, for instance, the hydrogen bonding interaction, might play an important role in understanding the low Λ/ΛNMR ratio for the EMI ionic liquids with the TFSI anion. However, there is no evidence of hydrogen bonds involving the BP cation. Finally, the Λ/ΛNMR had no significant influence from temperature. Although the data in Figure 5.9 are scattered, the Λ/ΛNMR ratio does not vary much with temperature. This implies that the change in temperature cannot bring about a significant change in the structures of the ionic liquids.

5.8

CONCLUSIONS

Self-diffusivity, cooperatively with ionic conductivity, provides a coherent account of ionicity of ionic liquids. The PGSE-NMR method has been a convenient means to measure the self-diffusion coefficients independently of the anions and the cations in the ionic liquids. Temperature dependencies of the self-diffusion coefficient, viscosity, and ionic conductivity for the ionic liquids cannot be explained simply by the Arrhenius equation; rather, they follow the VFT equation. There is a simple correlation of the summation of the cationic and the anionic diffusion coefficients for each ionic liquid with the inverse of the viscosity. The apparent cationic transference number in ionic liquids also has been dependent on the size of the anion, although the size of each ion does not directly affect the ionic diffusion coefficient. The Λ/ΛNMR ratio of the ionic liquids varies, depending on the structures of the ionic liquids, and the extent of ionic association and/or ionic activity consequently changes.

ACKNOWLEDGMENTS This research was supported in part by Grants-in-Aid for Scientific Research 14350452 and 16205024 from the Japanese Ministry of Education, Science, Sports, and Culture and by Technology Research Grant Program from the New Energy and Industrial Technology Development Organization of Japan. M.A.B.H.S. also acknowledges a postdoctoral fellowship from Japan Society for the Promotion of Science. In addition, the authors are thankful to Dr.

c05.indd 83

1/20/2011 10:27:10 AM

84

CORRELATION WITH IONIC TRANSPORT BEHAVIOR

Kikuko Hayamizu for her useful discussions and support in conducting NMR experiments. REFERENCES 1. J. Kärger, Diffusion Under Confinement, Sitzungsbericht der Sachsischen Akademie im Hirzel-Verlag, 2003. 2. H. J. V. Tyrrell, K. R. Harris, Diffusion in Liquids, Butterworths, 1984. 3. B. E. Poling, J. M. Prausnitz, J. P. O’Connell, The Properties of Gases and Liquids, McGraw-Hill, 2001. 4. P. Wasserscheid, T. Welton, eds., Ionic Liquids in Synthesis, Wiley-VCH, 2003. 5. A. L. van Geet, A. W. Adamson, J. Phys. Chem., 68, 238 (1964). 6. H. Vogel, Phys. Z., 22, 645 (1921); G. S. Fulcher, J. Am. Ceram. Soc., 8, 339 (1923). 7. H. Weingartner, in Diffusion in Condensed Matter, J. Karger, P. Heitjans, and R. Harberlandt, eds., Vieweg, 1998. 8. J. Crank, The Mathematics of Diffusion, 2nd ed., Clarendon Press, 1975. 9. J. D. Holbrey, K. R. Seddon, Clean Products Processes, 1, 223 (1999). 10. T. Welton, Chem. Rev., 99, 2071 (1999). 11. P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed., 39, 3772 (2000). 12. J. Dupont, R. F. de Souza, P. A. Z. Suarez, Chem. Rev., 102, 3667 (2002). 13. H. Ohno, ed., Ionic Liquids: The Front and Future of Material Development (in Japanese), CMC Press, 2003. 14. J. S. Wilkes, M. J. Zaworotko, J. Chem. Soc., Chem. Commun., 965 (1992). 15. J. Fuller, R. T. Carlin, H. C. De Long, D. Haworth, J. Chem. Soc., Chem. Commun., 299 (1994). 16. R. T. Carlin, H. C. De Long, J. Fuller, P. C. Trulove, J. Electrochem. Soc., 141, L73 (1994). 17. V. R. Koch, C. Nanjunduah, G. B. Appetecchi, B. Scrosati, J. Electrochem. Soc., 142, L116 (1995). 18. P. Bonhôte, A.-P. Dias, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). 19. J. Fuller, A. C. Breda, R. T. Carlin, J. Electrochem. Soc., 144, L67 (1997). 20. J. Fuller, A. C. Breda, R. T. Carlin, J. Electroanal. Chem., 459, 29 (1998). 21. A. B. McEwen, H. L. Ngo, K. LeCompte, J. L. Goldman, J. Electrochem. Soc., 146, 1687 (1999). 22. A. Noda, K. Hayamizu, M. Watanabe, J. Phys. Chem. B, 105, 4603 (2000). 23. A. Noda, M. Watanabe, Electrochim. Acta, 45, 1265 (2000). 24. H. Matsumoto, H. Kageyama, Y. Miyazaki, Chem. Commun., 1726 (2002). 25. R. Kawano, M. Watanabe, Chem. Commun., 330 (2003). 26. M. A. B. H. Susan, A. Noda, S. Mitsushima, M. Watanabe, Chem. Commun., 938 (2003). 27. A. Noda, M. A. B. H. Susan, K. Kudo, S. Mitsushima, K. Hayamizu, M. Watanabe, J. Phys. Chem. B, 107, 4024 (2003).

c05.indd 84

1/20/2011 10:27:10 AM

REFERENCES

85

28. M. A. B. H. Susan, M. Yoo, H. Nakamoto, M. Watanabe, Chem. Lett., 32, 836 (2003). 29. W. Xu, E. I. Cooper, C. A. Angell, J. Phys. Chem. B, 107, 6170 (2003). 30. M. A. B. H. Susan, H. Nakamoto, M. Yoo, M. Watanabe, Trans. MRS-J, 29, 1043 (2004). 31. W. S. Price, Concepts Magn. Reson., 9, 299 (1997); 10, 197 (1998). 32. E. O. Stejskal, J. Chem. Phys., 43, 3597 (1965). 33. L. Hussey, I.-W. Sun, S. K. D. Strubinger, P. A. Barnard, J. Electrochem. Soc., 137, 2515 (1990).

c05.indd 85

1/20/2011 10:27:10 AM

6 IONIC CONDUCTIVITY Hiroyuki Ohno, Masahiro Yoshizawa-Fujita, and Tomonobu Mizumo

Electrolyte solutions are essential for electrochemical devices. However, almost all solvents for electrolyte solutions have a crucial drawback: They are volatile solvents. Recently, ionic liquids (ILs), which are liquids that consist only of ions, have received much attention as new electrolyte materials. These ILs have two attractive features of high concentrations of ions and high mobilities of the component ions at room temperature, and many of these ILs show the ionic conductivity of more than 102 S cm−1 at room temperature [1–3]. In this chapter, we focus on the conductivity of ILs. Ionic conductivity is generally given by the equation

σi =

∑ neμ

(6.1)

where n is the carrier ion number, e is the electric charge, and μ is the mobility of carrier ions. The electrochemical properties of ILs studied are mainly their ionic conductivity, transference number, and potential window. These electrochemical properties differ in accordance with factors such as moisture absorption, experimental atmosphere, and electrode species. Therefore, the conductivity measurements must be made under consistent conditions. Often, electrochemical measurements and the construction of electrochemical cells are performed in a glove box filled with an inert gas (usually argon). Do not use nitrogen gas when metallic lithium was used as electrodes due to formation of ion-coductive Li3N. Ionic conductivity is defined as the reciprocal of proper resistance (Rb). It is obtained by calculating. Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

87

c06.indd 87

1/20/2011 10:27:12 AM

88

IONIC CONDUCTIVITY

σi =

l RbS

(6.2)

where l is the distance between the two electrodes and S is the mean area of the electrodes. Because it is almost impossible to measure the actual surface area of the applied electrode, l/S, which is known as the cell constant, is determined by measuring the conductivity of standard solutions like aqueous KCl. For ionic conductivity, S cm−1 or mS cm−1, are used as the unit of measure. The Rb based on the sample cannot be calculated correctly because the electric charge transfer resistance and the electric double layer in an electrode interface also are detected as a resistance, even if bias voltage is impressed to the measurement cell to measure the ionic conductivity. For the ionic conductivity measurement, a dc four-probe method, or the complex-impedance method, is used to separate a sample bulk and electrode interface [4]. In particular, the complex-impedance method has the advantage that it can be performed with both nonblocking electrodes (the same element for carrier ion Mn+ and metal M) and blocking electrodes (usually platinum and stainless steel were used where charge cannot be transferred between the electrode and carrier ions). The two-probe cell, in which the sample is sandwiched between two polished and washed parallel flat electrodes, is used in the ionic conductivity measurement by the complex-impedance method as shown in Figure 6.1.

(a)

(b) lead wire

lead wire

sample

sample electrode

spacer Solid

electrode Liquid

Figure 6.1. Schematic representation of cells used for ionic conductivity measurement by impedance method as an example. (a) Solid-state samples and (b) liquid-state samples. Be careful to keep the distance between the electrodes constant.

c06.indd 88

1/20/2011 10:27:12 AM

IONIC CONDUCTIVITY

89

The detected current is influenced considerably by any incomplete contact among the sample, electrode, bubbles, grain boundary, and so on. Therefore, it is necessary to be careful in preparing the contact conditions for the sample and the electrode interface in polymer systems. When the polymers are flexible films or soft paste, good contact is obtained by simply pressing. For hard or brittle polymer systems, it is difficult to get good contact by the pressing pressure. To improve the contact, it is necessary to melt the polymer material or to sputter electrode material on the sample surface. These processes must be carried out with appropriate spacers to prevent deformation of the polymer sample. When using metals, one must be aware of the strong reduction ability of metals as nonblocking electrodes. For instance, lithium metal must be handled in a completely dry atmosphere. It therefore cannot be used in an aqueous system or in samples containing active protons. In addition, because lithium nitride is produced by the reaction of lithium metal and nitrogen gas, argon gas must be used as the inert gas. Generally, highly activated metals are used on blocking electrodes to measure ionic conductivity. The impedance analyzer is used frequently under an alternating current. The reliability of the observed impedance depends on the frequency applied. It decreases with increasing frequency. Measurement frequency range is generally from 10 Hz to 1 MHz, of course, depending on the performance of the equipment used. Although approximately 10 mV amplitude voltage is desirable, it depends on the sample, and the appropriate value is chosen according to each sample. When the absolute value of impedance is large (generally around the order of GΩ), a sufficient current cannot be obtained under ordinary conditions. However, when the absolute value of the obtained impedance is small (generally under several Ω), the experimental error becomes large because of the large current. These values can be adjusted by changing the sample’s thickness. Figure 6.2 shows a typical Nyquist plot (or Cole–Cole plot) and its equivalent circuits for measurements with nonblocking electrodes [5,6]. In Figure 6.2b, where the samples show fast ion transport and the distribution of relaxation process for ion transport is small, two arcs are obtained in the Nyquist plot. The arcs show the movements of the ion and dipole moments at their frequency ranges. The left-side arc (high-frequency range) is explained by both the resistance component for ion transfer and the capacity component for the dielectric constant. The right-side arc (low-frequency range) relates to the response at the electrode surface. The resistance component responds to the charge transfer and the capacity component to the electric double layer. In the smaller frequency range, the density distribution in the sample is a result of the fully progressed redox reaction at the electrode surface. The impedance is based on the diffusion of ions. A straight line, declined 45 degrees, is depicted in the Nyquist plot (Warburg impedance). When blocking electrodes are used, a straight line is constructed perpendicularly to the real axis because the electrode surface has only a double-layer

c06.indd 89

1/20/2011 10:27:12 AM

90

IONIC CONDUCTIVITY (a)

1. Equivalent circuit for high-frequency range Rb

Cg 2. Equivalent circuit for low-frequency range Rci Rb

Cdl

(b)

Imaginary impedance Z″/ Ω

Decreasing frequency

1

2

Rb Real impedance Z′/Ω

Rb + Rci

Figure 6.2. (a) Schematic representation of equivalent circuit for an ion conductor put between a pair of nonblocking electrodes, and (b) the corresponding Nyquist plot. The high-frequency equivalent circuit (a − 1) consists of the bulk resistance of the sample (Rb) and the geometrical capacitance (Cg), whereas the low-frequency equivalent circuit (a − 2) consists of the resistance to charge transfer across the sample/electrode interface (Rct) and the double-layer capacitance (Cdl).

capacitor. In practice, the angle obtained is often less than the 90 degrees shown in Figure 6.3. This is a result of the incomplete double-layer capacitor. When redox species contact the electrode surface, the resulting Nyquist plot should be similar to plots in which electrodes are nonblocking. When the temperature is changed, the Nyquist plot also changes, as shown in Figure 6.4a, because of the change in conductivity. Figure 6.4 shows the Nyquist plots for the ionic liquid prepared by the neutralization of 1-benzyl2-methylimidazole and trifluoromethanesulfonimide (HTFSI). The temperature dependence of the ionic conductivity generally is depicted by an Arrhenius plot (Figure 6.4b). The typical temperature dependence of the ionic conductivity has been described [5]. Because most ILs including polymer systems show an upper

c06.indd 90

1/20/2011 10:27:12 AM

91

IONIC CONDUCTIVITY Rb

(a)

Cdl

Cg

espon se ical r Pract

Z″/Ω

Ideal response

(b)

Rb Z′/Ω

Figure 6.3. (a) Schematic representation of the equivalent circuit for an ion conductor put between a pair of blocking electrodes and (b) the corresponding Nyquist plot. Ideally, the sample–electrode interface consists only of the double-layer capacitance. However, the practical Nyquist plot that corresponds to this frequency region is not vertical to the real axis. The rate-limiting process of this plot is that the ion diffuses to form a double layer.

(a)

(b) −3.0

30 °C Z″/kΩ

40 °C 5.0

20 °C

50 °C

log(σi /S cm−1)

10.0

50 °C

−3.5 −4.0

40 °C

30 °C 20 °C

−4.5 −5.0

0.0

5.0 Z′/kΩ

10.0

−5.5 3.0

3.2 3.4 1000 T−1/K−1

Figure 6.4. Temperature dependence of the ionic conductivity for BzEImHTFSI. (a) Nyquist plots for various temperatures; (b) the corresponding Arrhenius plot.

c06.indd 91

1/20/2011 10:27:12 AM

92

IONIC CONDUCTIVITY

convex curvature in the Arrhenius plot, and not a straight line, the temperature dependence of the ionic conductivity is expressed by the following VogelFulcher-Tamman (VFT) equation [7]:

σ i (T ) =

A ⎡ −B ⎤ exp ⎢ ⎥ T ⎣ (T − T0 ) ⎦

(6.3)

where A and B are constants and T0 is the ideal glass transition temperature. Parameters A and B relate to the carrier ion number and the activation energy for the ion transport, respectively. Parameter T0 is usually 30 °C to 50 °C lower than that of Tg obtained by the differential scanning calorimetry (DSC) measurement. It is understood that the ions migrate in the amorphous phase and that their migration is influenced by viscosity. This is surmised by the fact that most systems show straight lines in the VFT plots. The hysteresis that may appear depends on the direction or the scan rate of temperature change. This tendency is based on the phase transition or the slow diffusion process of the sample. Remarkably, such hysteresis even can appear when a sample shows crystallization within the measurement temperature range. Therefore, the ionic conductivity measurement usually is performed at each temperature after reaching the constant value. However, ionic conductivity is measured at scan rate of 1 °C to 2 °C min−1. Therefore, when hysteresis appears during the heating or cooling process, the relationship between the phase transition and the ionic conductivity can be used in the analysis at this scan rate with the DSC measurement. It is better to use a small cell design to avoid the temperature distribution in samples. Finally, we consider the method for anisotropic ionic conductivity. Anisotropic ionic conductivity is a phenomenon that appears in a sample undergoing liquid crystalline phase or microphase separation. In the nanotech-

Glass Substrate Comb-shape Au Electrode Insulating Layer (mesogenic part) y

z

+

−

x Au

Glass Substrate

−

+

y Au

+ − Ion Conductive Layer

Figure 6.5. Schematic representation of the comb-shaped Au electrode cell, and the domain formation of the liquid crystalline-type electrolyte in Au electrodes.

c06.indd 92

1/20/2011 10:27:12 AM

REFERENCES

93

nology field, such samples are expected to yield important materials because they can raise the value of ion conduction. To measure the anisotropic ionic conductivity, one has to design a special cell. Figure 6.5 shows an example of a cell shaped like a comb [8]. In the comb-shaped cell, the ionic conductivity is measured between two teeth of the comb. As the liquid crystalline molecules orient perpendicularly on the glass substrate, the ionic conductivity is obtained along with the minor axis. The important task is to prepare the specific phase homogeneously in the measurement cell. The details will soon be presented in published papers.

REFERENCES 1. P. Bonhte, A.-P. Dias, M. Armand, N. Papageorgiou, K. Kalyanasundaram, M. Grtzel, Inorg. Chem., 35, 1168 (1996). 2. R. Hagiwara, Y. Ito, J. Fluorine Chem., 105, 221 (2000). 3. P. Wasserscheid, T. Welton, eds., Ionic Liquids in Synthesis, Wiley-VCH, 2003. 4. P. M. S. Monk, Fundamentals of Electroanalytical Chemistry, Wiley, 2001. 5. J. R. MacCallum, C. A. Vincent, eds., Polymer Electrolyte Reviews 1 and 2, Elsevier, 1987 and 1989. 6. P. R. Srensen, T. Jacobsen, Electrochim. Acta, 27, 1671 (1982). 7. (a) H. Vogel, Phys. Z., 22, 645 (1921); (b) G. S. Fulcher, J. Am. Ceram. Soc., 8, 339 (1925); (c) G. Tamman, W. Hesse, Z. Anorg. Allg. Chem., 156, 245 (1926). 8. (a) K. Ito-Akita, N. Nishina, Y. Asai, H. Ohno, T. Ohtake, Y. Takamitsu, T. Kato, Polym. Adv. Technol., 11, 529 (2000); (b) M. Yoshizawa, T. Mukai, T. Ohtake, K. Kanie, T. Kato, H. Ohno, Solid State Ionics, 154–155, 779 (2002); (c) M. Yoshio, T. Mukai, K. Kanie, M. Yoshizawa, H. Ohno, T. Kato, Adv. Mater., 14, 351 (2002).

c06.indd 93

1/20/2011 10:27:12 AM

7 OPTICAL WAVEGUIDE SPECTROSCOPY Hiroyuki Ohno and Kyoko Fujita

7.1

INTRODUCTION

Spectroelectrochemistry has proved to be a powerful technique for the study of redox processes. In the transmission experiment, spectroelectrochemistry is used to provide optical data that are complementary in nature to standard current or potential measurements [1]. Spectroelectrochemistry originally was conceived as a means to monitor, in the visible spectral region, solution electrochemical processes through absorbance changes occurring within the diffusion layer thickness adjacent to optically transparent electrodes. The power of spectroelectrochemical techniques in monitoring electrochemical processes was immediately apparent because the data obtained on the redox processes were possible in the absence of background current as a result of nonfaradaic charging and/or electroactive impurities. Furthermore, it was found that, through the choice of an appropriate wavelength, spectroelectrochemistry can add a dimension of selectivity to an electrochemical measurement. In part, because of these features, conventional transmission spectrochemistry in the visible wavelength regime is now widely used to characterize spectroscopically the redox process of diverse materials, ranging from electron transfer proteins [2], to electropolymers [3], to electrochromic organic and inorganic thin films [4], as well as to a host of organic radical-producing processes [1]. However, the standard transmission method of spectroelectrochemistry suffers from poor sensitivity.

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

95

c07.indd 95

1/20/2011 10:27:13 AM

96

OPTICAL WAVEGUIDE SPECTROSCOPY Sample dr = Z

Evanescent Wave

Intensity

n2 n1

1/e2

dp

l 2pn1

sin2

q – (n2 – n1)2

n1, n2: reflective index (n1 > n2) λ: wavelength θ: angle of the incident light dp: penetration depth of the evanescent wave

Figure 7.1. Characteristics of the evanescent wave.

However, a total internal reflection technique can be used to enhance the sensitivity at the interface [5,6]. This is because the total internal reflection of light at an interface generates an evanescent wave whose intensity falls exponentially with distance from the waveguide surface (Figure 7.1). The characteristic distance defined by this decay, corresponding to the penetration depth from the waveguide surface [7,8], is related to the wavelength of the incident light, and for visible light, it is several hundred nanometers. The evanescent wave is used to observe ultraviolet (UV)-visible absorption [9,10] as well as in ford (IR) [11] and fluorescence [12,13] spectra of molecules located within the penetration depth. This technique is useful for analyzing the effects of temperature [14], pH [15], and gas molecules [16] on the internally reflected spectral density. Optical waveguide (OWG) spectroscopy is a total internal reflection technique that is used to detect visible light absorption spectra [17–20]. Like ordinary visible spectroscopy, it gives information about both the structure and the electronic state of the molecules on a waveguide [17,20]. It also provides qualitative information about the molecules on the waveguide [19]. As the incident light propagates by repeating the total reflection in the waveguide, the evanescent wave generated at every reflection is absorbed by samples on the waveguide. The accumulation of absorbed light allows the absorption spectra of the samples to provide excellent signal-to-noise ratios. We have been using this technique to study surface-adsorbed molecules, so we can confirm that OWG spectroscopy provides reliable quantitative information about the molecules on the waveguide [19]. A combination of OWG spectroscopy with electrochemical methods can give even better information [21–25]. OWG spectroelectrochemistry is a useful technique for sensitive and simple analysis of redox reaction of molecules dissolved in a small amount of solution [22] or adsorbed on opaque electrodes [23,24]. Figure 7.2 shows the spectral change resulting from the redox reaction of Methylene blue. This OWG spectral change was observed repeatedly in situ, and it corresponded to the potential switching, because the electrochemical cell was constructed on the waveguide. The redox reaction of the molecules in the ionic liquid also was

c07.indd 96

1/20/2011 10:27:13 AM

97

ANALYSIS OF REDOX REACTIONS IN IONIC LIQUID

0.4

Absorbance

0.3 +0.5V 0.2

0.1 –0.5V

0 400

450

500

550 600 650 Wavelength (nm)

700

750

Figure 7.2. OWG spectral change of MB solution resulting from the redox reaction (W.E.: carbon; C.E.: Pt; R.E.: Ag).

analyzed easily by constructing an electrochemical cell system on the waveguide comprising working, reference, and counter electrodes [25].

7.2 ANALYSIS OF REDOX REACTIONS IN IONIC LIQUID [26] OWG spectroscopy has proved to be a highly sensitive detection method for molecules. The visible absorption spectrum has been studied with the OWG spectrophotometer (SIS-50; System Instruments Inc.) [19]. The light source used was a 150 W Xenon lamp (Hamamatsu Photonics, Hamamatsu City, Japan). The beam was modulated by a mechanical chopper and introduced into an optical fiber through a microscope object lens; the beam was focused on the edge of the waveguide. After propagating through the waveguide, the transmitted light was collected by the second microlens, and directed through the optical fiber to CCD spectrometer (Figure 7.3). An important task is to select a suitable waveguide for the OWG analysis in ionic liquid because most molten salts have high refractive indexes. The waveguide should have a higher refractive index than the refractive indexes of the ionic liquids, which are generally around 1.4 to 1.5 [27,28]. In fact, the OWG analysis in ionic liquid cannot be carried out with a quartz glass waveguide. For this purpose, an optical glass plate with n = 1.75 is used [29]. The waveguide plate was prepared by both edges being cut at 60 degrees, and then the plate was polished accurately enough for optical use. The angle of incident light was directed at 60 degrees to the surface, and then the average number of reflections in this OWG was estimated to be seven times per centimeter. In

c07.indd 97

1/20/2011 10:27:13 AM

98

OPTICAL WAVEGUIDE SPECTROSCOPY Mechanical chopper Xe-arc source PC

Optical fiber Collecting lens Collecting lens

Optical fiber Angle adjustment

Collecting lens

CCD spectrometer Angle adjustment

Waveguide

Figure 7.3. Schematic view of the OWG spectroscopy system.

(a) Top view

Ag wire (R.E.)

Pt wire (C.E.)

Working electrode

(b) Front view

Electrolyte Ag wire (R.E.)

Waveguide

Figure 7.4. Electrochemical cell system on the waveguide.

typical ILs, the OWG plate enables spectral measurements between wavelengths of 350 and 600 nm apart. The electrochemical redox reaction of molecules dissolved in ILs is detected as changes occur in the absorption spectra using the electrochemical cell system constructed on an optical glass plate (Figure 7.4) [29]. This cell system consists of a working electrode together with Pt and Ag wires as the counter and reference electrodes, respectively. The sample ionic liquid is injected directly into this electrochemical cell system, and the corresponding spectra are measured by the applied potential. Only μL range analysis is available, and it is suitable for expensive and highly valuable samples such as ILs.

REFERENCES 1. A. Bard, ed., Electroanalytical Chemistry, Dekker, 1974. 2. M. Collinson, E. F. Bowden, Anal. Chem., 64, 1470 (1992).

c07.indd 98

1/20/2011 10:27:13 AM

REFERENCES

99

3. T. A. Skotheim, R. L. Elsenbaumer, J. R. Rsynolds, eds., Handbook of Conducting Polymers, Dekker, 1998. 4. A. Ferencz, N. R. Armstrong, G. Wegner, Macromolecules, 27, 1517 (1994). 5. N. J. Harrick, ed., Internal Reflection Spectroscopy, Wiley, 1967. 6. W. N. Hansen, J. Opt. Soc. Amer., 58, 380 (1968). 7. R. M. Dickson, D. J. Norris, Y. Tzeng, W. E. Moerner, Science, 274, 966 (1996). 8. (a) X. Xu, E. Yeung, Science, 275, 1106 (1997). (b) X. Xu, E. Yeung, Science, 281, 1650 (1998). 9. S. B. Mended, L. Li, J. J. Burke, Langmuir, 12, 3374 (1996). 10. K. Kato, A. Takatsu, N. Matsuda, Chem. Lett., 31 (1999). 11. A. Bentaleb, A. Abele, Y. Haikei, P. Scheef, J. C. Voegel, Langmuir, 14, 6493 (1998). 12. T. R. E. Simpson, D. J. Revell, M. J. Cook, D.A. Russell, Langmuir, 13, 460 (1997). 13. A. N. Asanov, W. W. Wilson, P. B. Oldhan, Anal. Chem., 70, 1156 (1998). 14. L. M. Johnson, F. J. Leonberger, G. W. Pratt, Appl. Phys. Lett., 41, 134 (1982). 15. L. Yang, S. S. Saavedra, Anal. Chem., 67, 1307 (1995). 16. P. J. Skrdla, S. S. Saavedra, N. R. Armstrong, S. B. Mendes, N. Peyghambarian, Anal. Chem., 71, 1332 (1999). 17. N. Matsuda, A. Takatsu, K. Kato, Z. Shigesato, Chem. Lett., 125 (1998). 18. J. T. Bradshaw, S. B. Mendes, S. S. Saavedra, Anal. Chem., 74, 1751 (2002). 19. H. Ohno, S. Yoneyama, F. Nakamura, K. Fukuda, M. Hara, M. Shimomura, Langmuir, 18, 1661 (2002). 20. M. Mitsuishi, T. Tanuma, J. Matsui, J. Chen, T. Miyashita., Langmuir, 17, 7449 (2001). 21. (a) K. Kato, A. Takatsu, N. Matsuda, R. Azumi, M. Matsumoto, Chem. Lett., 437 (1995). (b) N. Matsuda, A. Takatsu, K. Kato, Chem. Lett., 105 (1996). 22. W. J. Doherty, C. L. Donley, N. R. Armstrong, S. S. Saavedra, Appl. Spectrosc., 56, 920 (2002). 23. H. Ohno, K. Fukuda F. Kurusu, Chem. Lett., 76 (2000). 24. K. Fujita, C. Suzuki, H. Ohno, Electrochem. Comm., 5, 47 (2003). 25. K. Fukuda, H. Ohno, Electroanalysis, 14, 605 (2002). 26. H. Ohno, C. Suzuki, K. Fukumoto, M. Yoshizawa, K. Fujita, Chem. Lett., 32, 450 (2003). 27. P. Bônhote, A. P. Dias, N. Papageprgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). 28. J. G. Huddleston, A. E. Visser, W. M. Reichart, H. D. Willauer, G. A. Broker, R. D. Rogers, Green Chem., 3, 156 (2001). 29. K. Fujita, H. Ohno, Polym. Adv. Tech., 14, 486 (2003).

c07.indd 99

1/20/2011 10:27:13 AM

8 ELECTROLYTIC REACTIONS Toshio Fuchigami and Shinsuke Inagi

Solvents have the important function of dissolving substrates in solutions and controlling desired reactions. Because many organic compounds do not dissolve in aqueous solutions, organic solvents often have been employed for organic reactions. Similar to organic reactions, organic solvents typically have been used in organic electrolytic reactions. However, water is one of the ideal electrolytic solvents. Water does have some limitations such as poor solubility of organic substrates and its competitive electrolytic reactions with substrates. Aprotic organic solvents are useful for avoiding such competitive reactions and for dissolving both organic substrates and supporting electrolytes, but most organic solvents are flammable and are not always safe from health and environmental perspectives. So far, in attempts to solve these problems, solid polymer electrolyte (SPE) electrolysis and gas-phase electrolysis are being developed using electrocatalysts. For inorganic electrolysis, solvent-free electrolysis, such as electrorefining processes using molten salts, has been commercialized already. In sharp contrast, a solvent-free organic electrode process has not yet been established. In recent years, room-temperature molten salts, namely ionic liquids, have proved to be a new class of promising solvents because of their good electroconductivity, nonflammability, thermal stability, nonvolatility, and reusability [1–3]. In addition, if a combination of cation and anion is made appropriately, then aprotic media with a wide electrochemical window can be obtained. Therefore, when ionic liquids are used as electrolytic media, organic electrolytic reactions, particularly electroorganic synthesis, should be possible without any organic solvents. Although ionic liquids already have been shown to be

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

101

c08.indd 101

1/20/2011 10:27:16 AM

102

ELECTROLYTIC REACTIONS

promising media for batteries, fuel cells, photovoltaic cells, and electroplating processes, there have still been a limited number of papers dealing with organic electrolytic reactions and electrosynthesis [4–8]. This chapter covers the applications of ionic liquids to organic electrolytic reactions, particularly electroorganic synthesis.

8.1

CELL DESIGN [9]

A proper choice of cell design is important for an electrolytic reaction. Organic electrolytic reactions are achieved on a laboratory scale by using an undivided cell. The simplest cell design is shown in Figure 8.1, but a cylindrical cell, shown in Figure 8.2, is the recommended design when anhydrous conditions or electrolysis under an inert gas atmosphere like a nitrogen gas is required. In the inert gas atmosphere, the solvent and substrate are injected with a syringe into the cell through a rubber septum. When a species reduced at a cathode is oxidized at an anode, and vice versa, a two-compartment cell, namely a divided cell like an H-type cell (Figure 8.3) with a sintered glass diaphragm or an ion-exchange membrane, should be used to prevent mixing of both anodic and cathodic solutions. To decrease the cell voltage (voltage between an anode and a cathode), the distance between both electrodes should be kept as small as possible.

Connect to Reference Electrode Salt Bridge

Thermometer

Gas Outlet

Cathode

Anode

Stirring Bar

Figure 8.1. Beaker-type cell.

c08.indd 102

1/20/2011 10:27:16 AM

103

CELL DESIGN

Inert Gas

Rubber Septum

Figure 8.2. Undivided cell for unhydrous electrolysis.

Glass Diaphragm or Ion-exchange Membran

Figure 8.3. H-type cell.

c08.indd 103

1/20/2011 10:27:16 AM

104

8.2 8.2.1

ELECTROLYTIC REACTIONS

ELECTROLYTIC METHOD Constant Current Electrolysis and Constant Potential Electrolysis [9]

It is recommended that organic electrosynthesis be carried out at a constant current at first, because the setup and operation are simple. Then the product selectivity and yield can be improved by changing the current density and the amount of electricity passed [current (A) × time (s) = electricity (C)]. However, the electrode potential changes with the consumption of the starting substrate (more positive for oxidation or more negative for reduction). Therefore, the product selectivity and the current efficiency sometimes decrease, particularly at the late stage of electrolysis. Nevertheless, a constant potential electrolysis is suitable to achieve a high selectivity and clarification of the reaction mechanisms. Moreover, by electrolysis, one can learn the number of electrons involved in the electrolytic reaction. As shown in Figure 8.1, a salt bridge is terminated either by a Luggin capillary or by a plug of porous Vycor glass placed closed to the working electrode, and an appropriate constant potential relative to a reference electrode such as aqueous saturated calomel electrode (SCE) is applied using a potentiostat. The amount of electricity passed is measured by a coulombmeter. 8.2.2

Indirect Electrolysis Using Mediators [10]

Indirect electrolysis using mediators is an alternative powerful electrolytic method. This method is suitable for the following cases: 1. Electrolysis of hardly oxidizable and/or hardly reducible substrates 2. Precise control of the desired reaction 3. Avoidance of electrode passivation (nonconducting polymer formation on the electrode) Halide ions, multiple valence metal ions (transition metal complexes), viologen, triarylamines, nitroxyl radicals, and fused arenes like naphthalene are used as mediators for this method. Mediators often act as an electrocatalyst. 8.2.3

Electropolymerization [11]

Electropolymerization is highly useful for the preparation of π-conjugated electroconducting polymers. Potential scanning polymerization is often employed in addition to an ordinary constant current and constant potential electropolymerizations. Because the polymer film is formed on the electrode surface, one can easily obtain valuable information about the propagation process from capacitance-voltage (CV) curves during the potential scanning.

c08.indd 104

1/20/2011 10:27:16 AM

105

SELECTIVE ANODIC FLUORINATION

8.3

SELECTIVE ANODIC FLUORINATION

Anodic partial fluorination, namely selective anodic fluorination of organic compounds, has much importance in the development of new types of pharmaceuticals, agrochemicals, and functional materials. However, the selective fluorination is not straightforward, and it often requires hazardous reagents. Still, a selective anodic fluorination seems to be an ideal method because it can be carried out under mild conditions without any hazardous reagents [12,13]. The fluorination is usually conducted in aprotic solvents containing HF salt ionic liquids such as Et3N · 3HF and Et4NF · 3HF. Because the discharge potential of fluoride ions is extremely high (>+2.9 V vs. SCE at Pt anode in MeCN), the fluorination proceeds via a (radical) cation intermediate as shown in Scheme 8.1. When organic solvents are used for anodic fluorination, anode passivation (the formation of a nonconducting polymer film on the anode surface that suppresses faradaic current) takes place often, which results in low efficiency. Moreover, depending on the substrates, the use of acetonitrile can yield an acetoamidation by-product. To prevent acetoamidation and anode passivation, Meurs and Eilenberg used an Et3N · 3HF ionic liquid as both a solvent and a supporting electrolyte (also a fluorine source) for the anodic fluorination of benzenes, naphthalene, olefins, furan, benzofuran, and phenanthroline. They obtained corresponding partially fluorinated products in less than 50% yields (Scheme 8.2) [14]. Fuchigami’s group and Laurent’s group have independently achieved anodic α-fluorination of α-phenylthioacetate (1) in Et3N · 3HF/MeCN [15,16]. Because anode passivation takes place during the anodic fluorination, pulse electrolysis is necessary to avoid the anode passivation. Anodic difluorination of 1 was unsuccessful in the same electrolytic solution, whereas the anodic fluorination of α-monofluorinated acetate (2) gave the α, α-dilfluorinated product 3 in moderate yield after passing an excess amount of electricity (Scheme 8.3) [17].

RH RH

-e-2e- -H+

RH • R+

-e- -H + FF-

R–F

Scheme 8.1

N

N

-2e- -H+ Et 3N•3HF (neat)

F F

F F

N

34%

N

Scheme 8.2

c08.indd 105

1/20/2011 10:27:16 AM

106

ELECTROLYTIC REACTIONS O SPh

EtO 1

O

-4e, -2H+ 6 F/mol Et3N-3HF Under Ultrasonication 65% yield -2e, -H+ -2e, -H+

Et3N-3HF/MeCN 1.6 V vs SSCE 2.5 F/mol 76% yield

O SPh

EtO F

SPh

EtO

F F 3

Et3N-3HF/MeCN 2.2 V vs SSCE 20.7 F/mol 53% yield

2

Scheme 8.3

U. S. O Et

O

SPh

Ethyl α-phenylthioacetate (1) 2 mA Stirr Still

0

1 V vs. Ag/Ag+

Sweep Rate Substrate W. Electrode

1 mV/s 10 mM Pt Plate (1×1 cm2)

2

Figure 8.4. Sonovoltammogram of ethyl α-phenylthioacetate (1) in ionic liquid, Et3N · 3HF.

To overcome the previously mentioned problems, anodic fluorination of 1 in neat Et3N · 3HF without a solvent was investigated under ultrasonication [18]. As shown in Figure 8.4, a limiting diffusion oxidation current of 1 increased markedly under ultrasonication compared with that under mechanical stirring. The visocity of the ionic fluoride salt is higher than that of ordinary organic solvents; therefore, the mass transport of substrate 1 to the anode surface from the bulk of the fluoride salt is slower than that in organic solvents. The enhanced oxidation current is attributed to the markedly promoted mass transport of 1. Namely, ultrasonication to the ionic liquid Et3N · 3HF generates cavitation, which makes microjet stream promoting mass transport of 1 to the anode surface from the bulk of the ionic liquid. The result indicates that ionic liquid Et3N · 3HF behaves just like ordinary solvents under ultrasonication.

c08.indd 106

1/20/2011 10:27:17 AM

107

SELECTIVE ANODIC FLUORINATION

Anodic fluorination of α-phenylthioacetate 1 in ionic liquid Et3N · 3HF proceeds smoothly without anode passivation under ultrasonication to provide the α-monofluorinated product 2 in high yield and with a high current efficiency. Notably, anodic difluorination of 1 can also be achieved in the same ionic liquid under ultrasonication as shown in Scheme 8.3 [18]. Thus, the current efficincy for the anodic difluorination of 1 was improved greatly as compared with that without ultrasonication. Ultrasonication also changed the stereoselectivity in addition to the increase of the yield as shown in Scheme 8.4 [18]. As described previously, Et3N · 3HF ionic liquid is highly useful for selective anodic fluorination. However, Et3N · 3HF is oxidized rather easily at around 2 V versus Ag/Ag+. Therefore, this ionic salt is not suitable for the fluorination of substrates having high oxidation potentials. Nevertheless, Momota et al. developed a new series of liquid fluoride salts, R4NF · nHF(R = Me, Et, and n-Pr, n > 3.5), which are useful in partial fluorination [19]. These electrolytes are nonviscous liquids that have high electroconductivity and anodic stability (Figure 8.5). As a result, anodic partial S

Ph

-2e, -H+

N O

O

S

F

Ph

+

N

3 F/mol Et3N-3HF

Me

S

F

Me

Ph N

O

Me

cis

trans

Under mechanical stirring: 24% yield (cis/trans = 50/50) Ultrasonication: 77% yield (cis/trans = 38/62)

Scheme 8.4 −2

−1

(CH3)4NF(HF)2

0

1

2

3

R4NF(HF)2

(C2H5)3N(HF)3 Me < Et < n-Pr

R4NF(HF)3 (C2H5)3N(HF)4 R4NF(HF)4 (C2H5)3N(HF)5 R4NF(HF)5

(C2H5)3N(HF)6

Figure 8.5. Potential windows of fluoride ionic liquids.

c08.indd 107

1/20/2011 10:27:17 AM

108

ELECTROLYTIC REACTIONS F

F F -2e

90%

Et 4 NF•4HF F

F F

2 F/mol

Scheme 8.5

CH3

CFH 2 -2e- -H+ Et4 NF•4HF

CF2 H -2e- -H+

CF2 H -2e- -H+

Et 4 NF•4HF

Et 4 NF•4HF

F

Scheme 8.6

O

-2e

F

Et 3 N•nHF 2.5 F/mol Pulse electrolysis MeOH

O

MeO O

F

F

n = 3: 13% n = 4: 73% n = 5: 91%

Scheme 8.7

fluorination of arenes such as benzene [19]; mono-, di-, and trifluorobenzenes [20]; chlorobenzene [21]; bromobenzenen [22]; toluene [23,24]; and quinolines [25] was carried out successfully at high current densities by employing these liquid fluoride salts in the absence of a solvent with good to high current efficiencies (66–90%) (Schemes 8.5 and 8.6). Nonconducting polymer films are often formed on anode surfaces when ordinary fluoride salts like Et3N · 3HF are used. However, such film formation is reduced considerably when the new supporting fluoride salts are used. Yoneda and Fukuhara also investigated the electrochemical stability (potential window) of ionic liquid Et3N · nHF (n = 3–5) by cyclic voltammetry [26]. Their anodic stability increases with an increase of HF content (n) in the fluoride salts, whereas the cathodic stability shows the reverse tendency. Thus, Et3N · 5HF is stable up to +3 V versus Ag/Ag+, but it is readily reduced at about −0.2 V to form hydrogen gas. The potential window of Et4NF · 4HF is almost the same as that of Et3N · 5HF, as shown in Figure 8.5. Thus, the selective anodic fluorination of cyclic ketones and cyclic unsaturated esters in Et3N · 5HF was carried out successfully to provide ring-opening and ring-expansion fluorinated products, respectively, as shown in Schemes 8.7 and 8.8 [26,27].

c08.indd 108

1/20/2011 10:27:17 AM

109

SELECTIVE ANODIC FLUORINATION CO2Et H 2C n

-2e

Et3 N•5HF

H 2C n+1

-20 °C

CO2Et

F F n = 0: 71% n = 1: 50%

Scheme 8.8 -2e, -H + O

2.0 M Et 4NF•4HF 2 F/mol 150 mA/cm2

F

O 80%

Scheme 8.9

O Z

O

-2e, -H + O

Z = CH2 Z=O

2.0 M Et 4NF•5HF 2 F/mol 100 mA/cm 2

Z

O

F Z = CH2 (75%) Z = O (87%)

Scheme 8.10

As mentioned previously, liquid fluoride salts like Et3N · nHF (n = 3–5) and Et4NF · 4HF have proved to be highly useful as the electrolytic media and fluoride ion source for selective anodic fluorination. However, this solvent-free method has an atom economy problem because of the use of an excessive amount of liquid salts in place of a solvent. Fuchigami et al. found that the fluorination of cyclic ethers like tetrahydrofuran, 1,4-dioxane, and 1,3-dioxolane was achieved by anodic oxidation of a mixture of a large amount of liquid cyclic ethers and a small amount of Et4NF · 4HF (only 1.5–1.7 equivalent of F− to the ether) at a high current density (150 mA cm−2) (Scheme 8.9) [28]. In this method, the ether substrate was oxidized selectively to provide the corresponding monofluorinated product in good yield and with good current efficiency. In sharp contrast, the use of organic solvents or a large amount of Et4NF · 4HF resulted in no formation or in a low yield (ca. 10%) of the desired fluorinated product. Lactone and cyclic carbonates are also fluorinated similarly to provide the monofluorinated products in good yields as shown in Scheme 8.10 [28]. It is notable that the fluorinated tetrahydrofuran is isolated easily, simply by distillation of the electrolytic ionic liquid fluoride salt after the electrolysis, with the remaining fluoride salt available for use in anodic fluorination. Thus, completely solvent-free electroorganic synthesis was demonstrated by using liquid fluoride salt, as shown in Figure 8.6.

c08.indd 109

1/20/2011 10:27:17 AM

110

ELECTROLYTIC REACTIONS

H2 Anodic fluorination

Distillation or Extraction

HF addition

Substrate

Fluoride Salt Ionic Liquids (Et4NF•nHF; n=4, 5)

+

Fluorinated Product

Figure 8.6. Procedure of solvent-free electroorganic synthesis.

TABLE 8.1. Anodic Fluorination of Phthalide under Various Conditions Run 1 2 3 4 5 6 7 8

Reaction Media

Supporting Electrolyte

Charge Passed F/mol

Yield/%a

MeCN CH2Cl2 DME — — [EMIM][OTf] [EMIM][OTf] [EMIM][OTf]

Et3N · 5HF Et3N · 5HF Et4NF · 4HF Et3N · 3HF Et3N · 5HF Et3N · 3HF Et4NF · 4HF Et3N · 5HF

6 8 8 4 8 6 4.7 4

16 23b 0 17 16b 78 70 90

Determined by 19 F- NMR. The reaction was complicated.

a

b

Moreover, Fuchigami’s group found that a combination of Et4NF · nHF (n = 4, 5) and imidazolium ionic liquids was highly effective for anodic fluorination of phthalides [29]. As shown in Table 8.1, conventional electrolysis in organic solvents resulted in low yields. Because the oxidation potential of phthalide is extremely high (2.81 V vs. SCE), the oxidation of solvents takes place preferentially or predominantly (in dimethoxyethane [DME]). Even under solvent-free conditions (in liquid fluoride salts solely), the yield is also low as a result of simultaneous oxidation of the fluoride salts during the electrolysis. In sharp contrast, when ionic liquid, 1-ethyl-3-methylimidazolium tri-

c08.indd 110

1/20/2011 10:27:17 AM

111

SELECTIVE ANODIC FLUORINATION

F O

- 2e, -H+ , FO

[EMIMF][2.3HF] O

6F/mol

O 31%

Scheme 8.11

OTf -2e, -H O O

+

Et3N•5HF /[EMIM][OTf]

O O

H4F 5

F O O

A activated cation Scheme 8.12

fluoromethanesulofonate ([EMIM][OTf]) is used, the yield increased markedly. The combination of [EMIM][OTf] and Et3N · 5HF is the best choice to provide excellent yield. However, the use of 1-ethyl-3-methylimidazolium fluoride · 2.3HF ([EMIMF][2.3HF]) gave an unsatisfactory yield (Scheme 8.11). From these results, it seems that the double ionic liquid system consisting of [EMIM][OTf] and a liquid fluoride salt like Et3N · 5HF enhances not only the nucleophilicity of F− but also the electrophilicity of a cationic intermediate of phthalide, as shown in Scheme 8.12. After generation of the cationic intermediate from phthalide, the cation should have TfO− as the counter anion to form the activated cation A, which readily reacts with F− to provide the fluorinated phthalide in good yield. Fuchigami et al. also found that imidazolium ionic liquid and dichloromethane show similar solvent effects on anodic fluorodesulfurization. As shown in Table 8.2, anodic fluorodesulfurization of 3-phenylthiophthalide (4) takes place exclusively in [EMIM][OTf] and CH2Cl2 to afford 5, whereas αfluorination proceeds predominantly in DME to provide 6. This suggests that [EMIM][OTf] destabilizes the radical cation intermediate B anodically generated from 4 similarly to CH2Cl2 as shown in Scheme 8.13. This anodic fluorodesulfurization can be performed by the reuse of the imidazolium and fluoride ionic liquids. It has been shown that anodically fluorinated product selectivity is greatly affected by fluoride ionic liquids, as shown in Scheme 8.14 [30]. The electrolysis of a phenylthioglycoside derivative 7 in Et3N · 4HF provides fluorodesulfurization product 8 exclusively, whereas that in Et3N · 3HF affords α-fluorinated product 9 along with 8. Because Et3N · 3HF contains a considerable amount

c08.indd 111

1/20/2011 10:27:17 AM

112

ELECTROLYTIC REACTIONS

TABLE 8.2. Solvent Effect on Anodic Fluorination of 3-Phenylthiophthalide (4) Run

Reaction Media

1 2 3 4 5 6 7

Supporting Electrolyte

MeCN CH2Cl2 THF DME [emim][OTf] [emim][OTf] [emim][BF4]

Et3N Et3N Et3N Et3N Et3N Et3N Et3N

· · · · · · ·

Yield %a

Charge Passed F/mol

3HF 3HF 3HF 3HF 3HF 5HF 5HF

4 4 4 3 4 4 3.5

5

6

44 86 6 9 44 83 65

— — 22 72 — Trace —

a

Determined by 19F-NMR.

SPh

SPh

-e

SPh -e

O

O

O

O 4

O H+

O

B F-

PhS

Fin CH2Cl2 [emim][OTf]

in DME THF F

F SPh

O 5

O

O

6 O

Scheme 8.13

OAc -e,-•SPh

O

FO

OAc AcO AcO

H3C

OAc AcO AcO

O

AcO AcO

F

AcO 8 (b only)

O SPh AcO 7

O

OAc +

-2e,-H

AcO AcO

O O H3C

OAc SPh O

F-

AcO AcO

O AcO

SPh F

9 (a, b) in Et3N•4HF, Et4NF•4HF: 8 (58-62 %) in Et3N•3HF: 8 (12 %), 9 (26 %)

Scheme 8.14

c08.indd 112

1/20/2011 10:27:17 AM

113

SELECTIVE ANODIC FLUORINATION

of free Et3N, Et3N seems to accelerate the deprotonation of the radical cation intermediate resulting in the formation of α-fluorinated product 9. In a solvent-free system, the nucleophilicity of fluoride ions of ionic liquid HF salts is rather low, which results in a decrease of fluorination yields. Fuchigami et al. successfully carried out the solvent-free electrochemical fluorination of α-(2-pyrimidylthio)acetate and O-ethyl benzothioate in the presence of ether-containing additives such as DME and poly(ethylene glycol) (PEG) [31]. As shown in Scheme 8.15, the addition of PEG additives (ca. 3%) into the reaction system highly improved the yield because of its anodic stability and its coordination ability to counter cations of fluoride ions. Moreover, highly efficient difluorodesulfurization of O-ethyl benzothioate was achieved by using this unique effect as shown in Scheme 8.16 [32]. As mentioned previously, severe passivation of the anode occurs often even in HF-based ionic liquid electrolytic systems without any organic solvents. Fuchigami et al. developed a novel indirect electrochemical fluorination system employing a task-specific ionic liquid with an iodoarene moiety as the mediator in a HF-based ionic liquid [33]. The mediator improved the reaction efficiency for a variety of electrochemical fluorination (Schemes 8.17 and 8.18)

N N

-2e,

O

S

Et4NF-5HF

N

Additive

Yield (%)

DME Dioxane PEG [Mn ~ 200]

35 58 5 60

F

N

-H+

O

S

O

O

Scheme 8.15

S

Additive

Yield (%)

DME PEG [Mn ~ 200]

22 18 80

F F -2e

O

Et4NF-3HF

O

Scheme 8.16

Et3N-3HF Mediator (10 mol %)

N N

Mediator:

S

Undivided cell 4 F/mol 5 mA/cm2

EWG

Tf2NI

O

N

N

N N

F S

EWG

EWG = COOEt, 87% (without mediator: 31%) EWG = CN, 72%

Scheme 8.17

c08.indd 113

1/20/2011 10:27:17 AM

114

ELECTROLYTIC REACTIONS

Et3N-5HF Mediator (10 mol %)

X EWG

Undivided cell 3 F/mol 5 mA/cm2

F X EWG X = H, 42~55% X = Cl, Br, 69~79%

Scheme 8.18

Figure 8.7. Reuse of mediator and HF-based ionic liquid for anodic fluorination.

Figure 8.8. Electrochemical fluorination of polyfluorene derivative.

and remained intact in the ionic liquids after the extraction process to be reused for subsequent runs as shown in Figure 8.7. Although an electrochemical substitution reaction to the polymer main chain or the side chain has been difficult, Fuchigami et al. accomplished a successive electrochemical fluorination of conducting polymer film fixed on an electrode. As shown in Figure 8.8, the electrochemical difluorodesulfurization

c08.indd 114

1/20/2011 10:27:17 AM

115

ELECTROCHEMICAL POLYMERIZATION

S

S

C10H21 N -2e, -2SAr

n

S

S

C10H21 N

Et4NF-5HF 24 F/mol

F

n-m

F

C10H21 N

m

Conversion: 50%

Scheme 8.19

of a poly(dialkylfluorene-alt-disulfanylfluorene) was carried out successfully in ionic liquid, Et4NF-5HF, to produce a novel conducting polymer with a difluorofluorene unit almost quantitatively [34,35]. The use of ionic liquid as an electrolytic solution effectively prevented the detachment of the polymer film from the electrode. The replacement of the dialkylfluorene unit with a more electrondonating unit, alkylcarbazole, resulted in the decrease of conversion (Scheme 8.19) [36]. In this case, the localization of cation charge on the alkylcarbazole unit probably prevented the desired oxidative fluorodesulfurization. 8.4

ELECTROCHEMICAL POLYMERIZATION

Electrochemical synthesis of electroconducting polymers such as polyarene [37–40], polypyrrole [41–43], polythiophene [44], and polyaniline [45,46] has been carried out in moisture-sensitive chloroaluminate ionic liquids. However, the polymer films are decomposed rapidly by the corrosive products like HCl generated by hydrolysis of the ionic liquids. In addition, the treatment of the chloroalminate ionic liquids requires a special equipment such as a glove box. Nevertheless, Mattes’s group and Fuchigami’s group independently achieved the electrooxidative polymerization of pyrrole, thiophene, and aniline in different moisture-stable imidazolium ionic liquids [47–49]. Mattes’s group used 1-butyl-3-methylimidazolium tetrafluoroborate and hexafluorophosphate ([BMIM)[BF4] and [BMIM][PF6]) for electropolymerization, and they found that π-conjugated polymers thus obtained are highly stable and that they can be electrochemical cycles in the ionic liquids up to 1 million cycles. In addition, because the polymers have cycle switching speeds as fast as 100 ms, they can be used as electrochemical actuators, electrochromic windows, and numeric displays. Fuchigami et al. employed [EMIM][OTf] for the electropolymerization. They found that the polymerization of pyrrole in the ionic liquid proceeds much faster than that in conventional media like aqueous and acetonitrile solutions containing 0.1 M [EMIM][OTf] as a supporting electrolyte, as shown in Figure 8.9 [48,49]. It is known that in the radical–radical coupling, continued oxidation of oligomers and polymer deposition in the electrooxidative polymerization are affected favorably because the reaction products are accumulated in the vicinity of the electrode surface under slow diffusion conditions, and consequently, the polymerization rate is increased. It is reasonable

c08.indd 115

1/20/2011 10:27:17 AM

116

ELECTROLYTIC REACTIONS

(a)

2 mA cm-2

(b)

2 mA cm-2

(c)

2 mA cm-2

-0.8

Ptential / V vs. SCE

+1.2

Figure 8.9. Cyclic voltammograms in the course of polymerization of 0.1 M pyrrole for 20 potential scan cycles in (a) 0.1 M [EMIM][OTf] + H2O, (b) 0.1 M [EMIM] [OTf] + MeCN, and (c) neat [EMIM][OTf].

to assume that the polymerization rate in [EMIM][OTf] is higher than that in the conventional media, because neat [EMIM][OTf] (viscosity: 42.7 cP) has a higher viscosity than the others. Electropolymerization of thiophene also is accelerated in this ionic liquid, whereas that of aniline is decelerated. It is notable that the morphology and some physical properties of polypyrrole and polythiophene films are affected greatly by the use of the ionic liquid [EMIM] [OTf] as the electrolytic medium. As shown in Figure 8.10, grains are

c08.indd 116

1/20/2011 10:27:18 AM

ELECTROCHEMICAL POLYMERIZATION

117

(a)

(b)

(c)

Figure 8.10. SEM photographs of polypyrrole films polymerized in (a) 0.1 M [EMIM] [OTf] + H2O, (b) 0.1 M [EMIM][OTf] + MeCN, and (c) neat [EMIM][OTf].

c08.indd 117

1/20/2011 10:27:18 AM

118

ELECTROLYTIC REACTIONS

TABLE 8.3. Properties of Polypyrrole and Polythiophene Films Prepared in Various Media Polymer

Polypyrrole Polypyrrole Polypyrrole Polythiophene Polythiophene

Media

H2O CH3CN EMICF3SO3 CH3CN EMICF3SO3

Roughness Electrochemical Capacity/C cm−3 Factora/ Dimensional 3.4 0.48 0.29 8.6 3.3

77 190 250 9 45

Electroconductivity/ S cm−1

Doping Level (%)b

1.4 × 10−7 1.1 × 10−6 7.2 × 10−2 4.1 × 10−8 1.9 × 10−5

22 29 42 — —

a

The roughness factor is the standard deviation of the film thickness. The doping level was determined by elemental analysis.

b

observed in the polypyrrole films prepared in aqueous and acetonitrile solutions, although the sizes of grain differ (Figure 8.10a, b). In sharp contrast, the surface of the polypyrrole film prepared in neat [EMIM][OTf] is so smooth that no grains are observed in the photograph (Figure 8.10c). As summarized in Table 8.3, both the electrochemical capacity and the electroconductivity are markedly increased when the polypyrrole and polythiophene films are prepared in the ionic liquid. This may be attributable to the extremely high concentration of anions as the dopants, which results in a much higher doping level. As described earlier, the polymer films prepared in the ionic liquid have a higher electrochemical density and highly regulated morphological structures. Therefore, their possible uses are as high-performance electrochemical capacitors, ion-sieving films, ion-selective electrodes, matrices for hosting catalyst particles, and so on. Figure 8.11 shows scanning election microscopy (SEM) photographs of surfaces of palladium-polypyrrole composite films. The polypyrrole matrix of the composite films (Figure 8.11a, b) was prepared in aqueous electrolyte and in neat [EMIM][OTf], respectively, and the pale-grey particles observed in the photographs are attributed to the presence of palladium. It is evident that aggregated palladium clusters of a few micrometers are deposited locally on grains of polypyrrole prepared in the aqueous solution, whereas submicron clusters are highly dispersed on the whole surface of the polypyrrole matrix prepared in [EMIM][OTf]. This finding may be attributed to the homogeneous polarization of the matrix in the palladium deposition process because a polypyrrole matrix prepared in [EMIM][OTf] has a smooth surface. In addition, such a composite film can be used for several catalytic and electrocatalytic reactions. The utility of the ionic liquid as a recyclable medium for the polymerization was also demonstrated. More than 90% of [EMIM][OTf] after the polymerization was easily recovered simply by extracting the remaining pyrrole monomer with chloroform. The recovered [EMIM][OTf] could be reused five times without significant loss of reactivity for the polymerization.

c08.indd 118

1/20/2011 10:27:18 AM

ELECTROCHEMICAL FIXATION OF CO2

119

(a)

(b)

Figure 8.11. SEM photographs of polypyrrole films polymerized in (a) 0.1 M [EMIM] [OTf] + MeCN and (b) neat [EMIM][OTf].

Anodic polymerization of 3-(p-fluorophenyl)thiophene was carried out in [EMIM][NTf2] as shown in Scheme 8.20, and the resulting polymer film was also smooth [50]. Furthermore, the electrosynthesis of poly(3,4ethylenedioxythiophene) (PEDOT) (Scheme 8.21) and polyphenylene in ionic liquids was reported [51,52].

8.5

ELECTROCHEMICAL FIXATION OF CO2

A quaternary ammonium salt Et4N+ · p-TsO− can be changed to a melt with good conductivity (1 × 10−2 S cm−1) at 140 °C. Approximately 30 years ago, Weinberg successfully used this melt for electrochemical synthesis of an αamino acid derivative from an imine and CO2 using a mercury cathode, as

c08.indd 119

1/20/2011 10:27:18 AM

120

ELECTROLYTIC REACTIONS

F

F

-ne S

Me N

FPT

Tf2NN Et

S

n

PFPT

Me

Scheme 8.20

O

O

-ne [EMIM][Tf2N]

S EDOT

O

O

S

n

PEDOT

Scheme 8.21

PhHC NPh

2e, CO2, 2H+ Et4N+•p-TsO- (melt) 140 °C

PhHC NHPh COOH 60%

Scheme 8.22

O O

e +

R

O

CO2 (1 atm) - 2.4 V vs. Ag / AgCl ionic liquid, rt Cu (-) - Mg or Al (+)

O

R R = Me, Cl, Ph 54 ~ 92% yield 32 ~ 87 % Current Efficiency

Scheme 8.23

shown in Scheme 8.22 [53]. The reaction proceeds via the radical anionic intermediate of the imine. Deng et al. found that CO2 was reduced at a cupper cathode at −2.4 V versus Ag/AgCl [54]. They successfully prepared cyclic carbonates by reducing CO2 in the presence of epoxides in various ionic liquids like [EMIM][BF4], [BMIM] [PF6] and by reducing N-butylpyridinium tetrafluoroborate [BPy][BF4] using a Cu cathode and an Mg or Al anode as shown in Scheme 8.23. This electrolysis

c08.indd 120

1/20/2011 10:27:19 AM

121

ELECTROREDUCTION OF CARBONYL COMPOUNDS

O Bn

H

N Me

electrolysis

Bn

[bmim][BF4]

N Me

OEt

divided cell CO2 (0.1 MPa), 55 oC -2.4 V vs. Ag/AgCl then EtI

Scheme 8.24 Me Ph

Cl

Me

2e, CO2 [MeEt2NCH2CH2OMe][NTf2] Pt cathode - Mg anode 43 oC, 8 MPa

Ph

COOH

Scheme 8.25

is a nonfaradaic reaction in which a small amount of electricity engaged in the electroreduction of CO2 generates a catalytic species responsible for the addition of CO2 to epoxides to form cyclic carbonates. Tascedda et al. also reported the electrochemical reaction of epoxides with CO2 in the presence of a nickel complex providing the corresponding cyclic carbonates [55]. Electrochemical synthesis of carbamates was also achieved by the electrolysis of a solution of CO2 and amine in [BMIM][BF4] followed by the addition of an alkylating agent as shown in Scheme 8.24 [56]. Electrochemical carboxylation of α-chloroethylbenzene was carried out successfully in ionic liquid like N,N-diethyl-N-methyl-N-(2-methoxyethyl) ammonium bis(trifluoromethanesulfonyl)amide compressed with CO2 as shown in Scheme 8.25 [57]. The current efficiency increased with temperature and pressure because mass transfer of the substrate was enhanced with pressure and temperature.

8.6

ELECTROREDUCTION OF CARBONYL COMPOUNDS

The cathodic reduction of carbonyl compounds like bezaldehyde and acetophenone was investigated in ionic liquids, and notably, their dimerization proceeded predominantly [58,59]. For acetophenone, the corresponding pinacol was formed as a diastereomeric mixture, and the diastereoselectivity is greatly affected by the ionic liquids used, as shown in Scheme 8.26 [59]. Lu et al. reported the electroreduction of benzoylformic acid at a glassy carbon cathode and a Pt anode in ionic liquid [EMIM][Br] at 80 °C giving mandelic acid in high yield and with moderate current efficiency as illustrated in Scheme 8.27 [60]. In this case, no coupling product was formed.

c08.indd 121

1/20/2011 10:27:20 AM

122

ELECTROLYTIC REACTIONS O Ph

HO

2e (1.0 F/mol) Me

Ph

divided cell

Ph OH

[bmim][NTf2]

92%, 6% d.e.

[Et3BuN][NTf2]

82%, 58% d.e.

[Me3BuN][NTf2] 80%, 54% d.e.

Scheme 8.26

PhCOCOOH

2e [EMIM][Br] - 1.3 V vs SCE GC (-) - Pt (+) 80

PhCH(OH)COOH 91 % yield (57 % Current Efficiency)

Scheme 8.27

Villagrán et al. demonstrated that the electroreducton of Nmethylphthalimide was promoted under ultrasonication resulting in higher conversion and current efficiency [61]. This is also a result of the facilitated mass transport of the substrate under ultrasonication.

8.7 ELECTROREDUCTIVE COUPLING USING METAL COMPLEX CATALYSTS Sweeny and Peters carried out cyclic voltammetry of nickel(II) salen at a glassy carbon electrode in [BMIM][BF4]. They found that nickel(II) salen exhibits one-electron, quasireversible reduction to nickel(I) salen and that the latter species serves as a catalyst for cleavage of carbon-halogen bonds in iodoethane and 1,1,2-trichlorotrifluoroethane (FreonR 113) [62]. They also showed the diffusion coefficient for nickel(II) salen in the ionic liquid at room temperature is more than 500 times smaller (1.0 × 10−5 cm2 s−1) than that in a typical organic solvent–electrolyte system such as dimethylformamide (DMF) containg 0.1 M Et4N · BF4. This is because of the high viscosity of the ionic liquid. Mellah et al. reported the electrocatalytic homo-coupling of PhBr and PhCH2Br using a NiCl(bpy) complex in [BMIM][NTf2] as shown in Scheme 8.28 [63]. Barhdadi et al. reported that direct electroreductive coupling of organic halides in [Octyl-MIM][BF4] was unsuccessful because of its high viscosity. However, they successfully carried out the reaction in the ionic liquid containing DMF (10% v/v) using a Ni cathode and a Mg or Al anode [64]. They

c08.indd 122

1/20/2011 10:27:20 AM

123

ELECTROREDUCTIVE COUPLING USING METAL COMPLEX CATALYSTS

2e

2 Ph(CH2)xBr

Ph(CH2CH2)xPh

cat. NiCl2(bpy) [BMIM][Tf2N] - 1.4 V vs Ag / Ag+ 2 F / mol

x = 0 : 35 % 1 : 75 %

Scheme 8.28

R PhBr

+

RCH = CHY

2e Y cat. NiBr2(bpy) [Oxtyl-MIM][BF4] / DMF ( 9 : 1 v / v) Stainless (+) - Ni (-)

Ph R = H, Y = COMe : 58 % R = H, Y = COOBu : 61 % R = Y = COOMe : 41 %

Scheme 8.29

Br

[BMIM][BF4]

Br 2 Co(I)salen

2 Co(II)salen

68%

2e

Scheme 8.30

performed homo-coupling of PhBr using NiBr2(bpy) (5% vs. PhBr) in the same electrolytic solution. Then they obtained 100% current efficiency and 82% yield, using Ni, SUS, and Fe sacrificial anodes. They also achieved Nicatalyzed electroreductive coupling of aryl halides and activated olefins, as shown in Scheme 8.29. One nitrogen atom bearing a free electron pair of imidazolium cation seems to interact with the electrogenerated low-valent Ni-species, preventing the precipitation of nickel metal. They also achieved the nickel-catalyzed electroreductive coupling of benzyl chloride with acetic anhydride in [BMIM][BF4] to provide benzyl methyl ketone in moderate yield (51%) [65]. Interestingly, Pd nanoparticles generated cathodically in ionic liquid were shown to be a highly effective ligand-free catalyst for the coupling of aryl halides [66]. The electroreductive dehalogenation of vic-dihalides using a Co(II)salen complex in [BMIM][BF4] also was achieved, as shown in Scheme 8.30 [67]. The product isolation is much easier compared with the similar dehalogenation in

c08.indd 123

1/20/2011 10:27:20 AM

124

ELECTROLYTIC REACTIONS

BF4e Me N H

N Bu

Me N

[BMIM][BF4]

1 H2 N Bu + 2

H

O Ph

HO H

NO2

+ MeNO2

NO2 +

Me N

N Bu

Ph

Ph 81%

17%

Scheme 8.31

MeO

-6e, -6H+

MeO

3 MeO

MeO

3

Scheme 8.32

ordinary molecular solvents, and the recyclability of the catalyst/solvent system was demonstrated, although the catalytic reactivity decreased. N-Heterocyclic carbenes can be generated by cathodic reduction of imidazolium-based ionic liquids [68–70]. Feroci et al. reported that the resulting carbenes are stable bases that can catalyze the Henry reaction as shown in Scheme 8.31 [70].

8.8 ANODIC OXIDATION OF ALCOHOLS AND AROMATIC COMPOUNDS Tempo-mediated anodic oxidation of alcohols to aldehydes and ketones was carried out successfully in ionic liquids [71,72]. High viscosity of the ionic liquids limiting the mass transport of the catalyst was readily overcome by the addition of base and alcohols, and high yields and current efficiencies were achieved [71]. Mellah et al. reported anodic oxidation of various aromatic compounds like anisole, mesitylene, and fused aromatic compounds like naphthalene and anthracene in various ionic liquids such as [BMIM][PF6] and [BMIM][NTf2] to provide the corresponding dimers in moderate to good yields. Under similar conditions, the anodic oxidation of 1,2-dimethoxybenzene leads to the corresponding trimer as shown in Scheme 8.32 [73].

c08.indd 124

1/20/2011 10:27:20 AM

REFERENCES

8.9

125

OTHER DEVELOPMENTS

Fry found that 1,4- and 1,2-dinitrobenzenes exhibit two one-electron waves in acetonitrile but exhibit a single two-electron wave in the ionic liquid [BMIM] [BF4] [74]. He explained that the latter effect is ascribed to a strong ion pairing between the imidazolium cation and the dinitrobenzene dianion. However, it was clarified that this interesting effect is attributable to the hydrogen bonding between the hydrogen at the 2-position of the imidazolium cation and the dinitrobenzene dianion [75].

8.10

CONCLUSION

Organic electrochemistry is an inherently hybrid and interdisciplinary area. Electroorganic synthesis has been recognized as an environmentally friendly process since the 1970s. However, new methodologies must be developed to achieve modern electrosynthsis as a green and sustainable chemistry. As studies on electrolytic reactions and electroorganic synthesis in ionic liquids continue, it is hoped that this chapter will inspire not only electrochemists but also organic chemists to turn to the new and exciting area of organic electrochemistry in ionic liquids.

REFERENCES 1. T. Welton, Chem. Rev., 99, 207 (1999). 2. P. Wassersheid, W. Keim, Angew. Chem. Int. Ed., 39, 3772 (2000). 3. R. D. Rogers, K. R. Seddon, S. Volkov, eds., Green Industrial Applications of Ionic Liquids, Kluwer Academic, 2002. 4. H. Ishii, T. Fuchigami, Electrochemistry, 70, 46 (2002). 5. T. Fuchigami, M. Atobe, Mater. Integr., 16, 20 (2003). 6. T. Fuchigami, J. Fluorine Chem., 128, 311 (2007). 7. P. Hapiot, C. Lagrost, Chem. Rev., 108, 2238 (2008). 8. J. Yoshida, K. Kataoka, R. Horcajada, A. Nagaki, Chem. Rev., 108, 2265 (2008). 9. M. R. Rifi, F. H. Covitz, Introduction to Organic Electrochemistry, Dekker, 1974. 10. J. Simonet, J.-F. Pilard, in Organic Electrochemistry, 4th ed., H. Lund, and O. Hammerich, eds., Dekker, 2000. 11. J. Heinze, in Topics in Current Chemistry: Electrochemistry IV, E. Steckhan, ed., Springer, 1990, p. 1. 12. T. Fuchigami, Advances in Electron Transfer Chemistry, vol. 6, P. S. Mariano, ed., JAI Press, 1999, p. 41. 13. T. Fuchigami, in Organic Electrochemistry, 4th ed., H. Lund, and O. Hammerich, eds., Dekker, 2000.

c08.indd 125

1/20/2011 10:27:20 AM

126

ELECTROLYTIC REACTIONS

14. 15. 16. 17. 18. 19. 20.

J. H. H. Meurs, W. Eilenberg, Tetrahedron, 47, 705 (1991). T. Fuchigami, M. Shimojo, A. Konno, K. Nakagawa, J. Org. Chem., 55, 6074 (1990). T. Brigaud, E. Laurent, Tetrahedron Lett., 31, 2287 (1990). T. Fuchigami, A. Konno, M. Shimojo, J. Org. Chem., 59, 3459 (1994). T. Sunaga, M. Atobe, S. Inagi, T. Fuchigami, Chem. Commun., 956 (2009). K. Momota, M. Morita, Y. Matsuda, Electrochim. Acta, 38, 1123 (1993). K. Momota, T. Yonezawa, Y. Hayakawa, K. Kato, M. Morita, Y. Matsuda, J. Appl. Electrochem., 25, 651 (1995). K. Momota, H. Horio, K. Kato, M. Morita, Y. Matsuda, Electrochim. Acta, 40, 233 (1995). K. Momota, H. Horio, K. Kato, M. Morita, Y. Matsuda, Denki Kagaku (presently, Electrochemistry), 62, 1106 (1994). K. Momota, K. Mukai, K. Kato, M. Morita, Electrochim. Acta, 43, 2503 (1998). K. Momota, K. Mukai, K. Kato, M. Morita, J. Fluorine Chem., 87, 173 (1998). K. Saeki, M. Tmomizitsu, Y. Kawazoe, K. Momota, H. Kimoto, Chem. Pharm. Bull., 44, 2254 (1996). S.-Q. Chen, T. Hatakeyama, T. Fukuhara, S. Hara, N. Yoneda, Electrochim. Acta, 42, 1951 (1997) S. Hara, S. Q. Chen, T. Hoshio, T. Fukuhara, N. Yoneda, Tetrahedron Lett., 37, 8511 (1996). M. Hasegawa, H. Ishii, T. Fuchigami, Tetrahedron Lett., 43, 1502 (2002). M. Hasegawa, H. Ishii, T. Fuchigami, Green Chem., 5, 512 (2003). M. Hasegawa, T. Fuchigami, Electrochim. Acta, 49, 3367 (2004). S. Inagi, T. Sawamura, T. Fuchigami, Electrochem. Commun., 10, 1158 (2008). T. Sawamura, S. Inagi, T. Fuchigami, J. Electrochem. Soc., 156, E26 (2009). T. Sawamura, S. Kuribayashi, S. Inagi, T. Fuchigami, Org. Lett., 12, 644 (2010). S. Inagi, S. Hayashi, T. Fuchigami, Chem. Commun., 1718 (2009). S. Hayashi, S. Inagi, T. Fuchigami, Macromolecules, 42, 3755 (2009). S. Hayashi, S. Inagi, T. Fuchigami, Electrochemistry, 78, 114 (2010). D. C. Trivedi, Chem. Commun., 544 (1989). L. M. Goldenberg, G. E. Pelekh, V. I. Krinichnyi, O. S. Roshchupkina, A. F. Zueva, R. N. Lyubovskaya, O. N. Efimov, Synth. Met., 36, 217 (1990). V. M. Kobryanskii, S. A. Arnautov, Makromol. Chem., 193, 455 (1992). V. M. Kobryanskii, S. A. Arnautov, Synth. Met., 55, 1371 (1993). P. G. Pickup, R. A. Osteryoung, J. Am. Chem. Soc., 106, 2294 (1984). P. G. Pickup, R. A. Osteryoung, J. Electroanal. Chem., 195, 271 (1985). T. A. Zawodzinski, Jr., L. Janiszewska, R. A. Osteryoung, J. Electroanal. Chem., 255, 111 (1988). L. Janiszewska, R. A. Osteryoung, J. Electrochem. Soc., 134, 2787 (1987). N. Koura, H. Ejiri, K. Takeishi, Denki Kagaku, 59, 74 (1991). N. Koura, H. Ejiri, K. Takeishi, J. Electrochem. Soc., 140, 602 (1993). W. Lu, A. G. Fadeev, B. Qi, E. Smela, B. R. Mattes, J. Ding, G. M. Spinks, J. Mazurkiewicz, D. Zhou, G. G. Wallace, D. R. MacFarlane, S. A. Forsyth, M. Forsyth, Science, 297, 983 (2002).

21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47.

c08.indd 126

1/20/2011 10:27:20 AM

REFERENCES

127

48. K. Sekiguchi, M. Atobe, T. Fuchigami, Electrochem. Commun., 4, 881 (2002). 49. K. Sekiguchi, M. Atobe, T. Fuchigami, J. Electroanal. Chem., 557, 1 (2003). 50. E. Naudin, H. A. Ho, S. Branchaud, L. Breau, D. Belanger, J. Phys. Chem. B, 106, 10585 (2002). 51. H. Randriamahazaka, C. Plesse, D. Chevrot, Electrochem. Commun., 6, 299 (2004). 52. S. Z. E. Abedin, N. Borissenko, F. Endres, Electrochem. Commun., 6, 422 (2004). 53. N. L. Weinberg, A. K. Hoffmann, T. B. Reddy, Tetrahedron Lett., 12, 2271 (1971). 54. H. Yang, Y. Gu, Y. Deng, F. Shi, Chem. Commun., 274 (2002). 55. P. Tascedda, M. Weidemann, E. Dinjus, E. Dunach, Appl. Organomet. Chem., 15, 141 (2001). 56. M. Feroci, M. Orsini, L. Rossi, G. Sotgiu, A. Inesi, J. Org. Chem., 72, 200 (2007). 57. Y. Hiejima, M. Hayashi, A. Uda, S. Oya, H. Kondo, H. Senboku, K. Takahashi, Phys. Chem. Chem. Phys., 12, 1953 (2010). 58. A. P. Doherty, C. A. Brooks, Electrochim. Acta, 49, 3821 (2004). 59. C. Lagrost, P. Hapiot, M. Vaultier, Green Chem., 7, 468 (2005). 60. J.-X. Lu, Q. Sun, M.-Y. He, Chin. J. Chem., 21, 1229 (2003). 61. C. Villagrán, C. E. Banks, W. R. Pitner, C. Hardacre, R. G. Compton, Ultrason. Sonochem., 12, 423 (2005). 62. B. K. Sweeny, D. G. Peters, Electrochem. Commun., 3, 712 (2001). 63. M. Mellah, S. Gmouh, M. Vaultier, V. Jouikov, Electrochem. Commun., 5, 591 (2003). 64. R. Barhdadi, C. Courtinard, J. Y. Nedelec, M. Troupel, Chem. Commun., 1434 (2003). 65. R. Barhdadi, H. Maekawa, C. Comminges, M. Troupel, Bull. Chem. Soc. Jpn., 82, 1510 (2009). 66. L. Duran Pachon, C. J. Elsevier, G. Rothenberg, Adv. Synth. Catal., 348, 1705 (2006). 67. Y. Shen, M. Atobe, T. Tajima, T. Fuchigami, Electrochemistry, 72, 849 (2004). 68. S. Chowdhury, R. S. Mohan, J. L. Scott, Tetrahedron, 63, 2363 (2007). 69. M. Feroci, I. Chiarotto, M. Orsini, G. Sotgiu, A. Inesi, Adv. Synth. Catal., 350, 1355 (2008). 70. M. Feroci, M. N. Elinson, L. Rossi, A. Inesi, Electrochem. Commun., 11, 1523 (2009). 71. R. Barhdadi, C. Comminges, A. P. Doherty, J.–Y. Nédélec, S. O’Toole, M. Troupel, J. Appl. Electrochem., 37, 723 (2007). 72. M. Kuroboshi, J. Fujisawa, H. Tanaka, Electrochemistry, 72, 846 (2004). 73. M. Mellah, J. Zeitouny, S. Gmouh, M. Vaultier, V. Jouikov, Electrochem. Commun., 7, 869 (2005). 74. A. J. Fry, J. Electroanal. Chem., 546, 35 (2003). 75. T. Iwasa, M. Atobe, T. Fuchigami, (unpublished work).

c08.indd 127

1/20/2011 10:27:20 AM

9 ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS Yasushi Katayama

Electrodeposition of metals and alloys in condensed media, including electroand electrolessplating, has been an essential technology for the advanced electronics industry besides its conventional industrial uses in electrorefining, decoration, corrosion protection, and other functional surface finishing. Most practical wet processes are based on aqueous media, mainly because of the knowledge and experience accumulated over many years. However, the metals that can be deposited in aqueous media are limited by the electrochemical potential window of water. Thus, such nonaqueous media as aprotic organic electrolytes and high-temperature molten salts have been investigated as the alternative electrolytes for the electrodeposition of base metals. Room-temperature ionic liquids (RTILs) are the promising electrolytes for the electrodeposition of various metals because they have the merits of both organic electrolytes and high-temperature molten salts. Ionic liquids can be used in a wide temperature range, so temperatures can be elevated to accelerate such phenomena as nucleation, surface diffusion, and crystallization associated with the electrodeposition of metals. In addition, the process can be safely constructed because ionic liquids are neither flammable nor volatile if they are kept below the thermal decomposition temperature of the organic cations. Numberous studies on the electrodeposition of metals have been reported for chloroaluminate ionic liquids, but so far, only a few studies have explored nonchloroaluminate ionic liquids. Because chloroaluminate ionic liquids are highly hygroscopic and, upon undergoing hydrolysis, produce toxic and Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

129

c09.indd 129

1/20/2011 10:27:53 AM

130

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

corrosive hydrogen chloride, these ionic liquids are not suited for practical use. However, various chemical substances, especially metal chlorides, can be dissolved by controlling the Lewis acidity or basicity of the chloroaluminate ionic liquids, and these systems are worth investigating. Nonchloroaluminate ionic liquids, especially those stable against moisture, have potential for practical use. In developing ionic liquids for practical applications, it will be necessary to accumulate much fundamental data on the solubility of metal compounds, redox potentials of metals, and mass transport of dissolved species. Several studies describe the electrodeposition of metals and alloys in the “ionic liquids” diluted with certain aprotic organic solvents. It is important to recognize that, although from a practical point of view, the addition of organic solvents improves the mobility of the electroactive species as well as the morphology of the electrodeposits, these electrolytes should not be classified equally as ionic liquids but be treated as separate entities. The addition of the organic solvents not only weakens the important properties of ionic liquids, such as nonflammability and nonvolatility, but at the same time, it also significantly alters the environment of the species’ participating electrode reactions. Therefore, this chapter will not deal with these electrolytes. There are several studies on the electrochemistry and electrodeposition of metals in the ionic liquids [1,2]. The following sections review examples pertaining to the electrodeposition of metals in ionic liquids based on organic cations. The organic cations in these sections are abbreviated as follows: EMI BMI DMPI BP MP BTMA TMHA TEA 9.1

1-Ethyl-3-methylimmidazolium 1-Butyl-3-methylimidazolium 1,2-Dimethyl-3-propylimidazolim N-Butylpyridinium Methylpyridinium Benzyltrimethylammonium Trimethylhexylammonium Tetraethylammonim

CHLOROALUMINATE IONIC LIQUIDS

The electrodeposition of metals and alloys has been investigated extensively in the chloroaluminate ionic liquids. Many kinds of metal salts, mostly chlorides, can be dissolved in ionic liquids with their Lewis acidity or basicity controlled by changing the composition of AlCl3. For acidic ionic liquids that contain AlCl3 at more than 50 mol%, a dimeric chloroaluminate anion, Al 2 Cl −7 , acts as a Lewis acid and withdraws a chloride ion from a metal chloride, MCln to form a cationic species, MCl mn−+m, when the Lewis acidity of MCln is weaker than that of Al 2 Cl −7 : MCl n + m Al 2 Cl 7− = MCl nm−+m + 2 m AlCl 4−

c09.indd 130

(9.1)

1/20/2011 10:27:53 AM

CHLOROALUMINATE IONIC LIQUIDS

131

The cationic species, MCl mn−+m, is reduced, possibly to its metallic state under cathodic polarization. In the acidic ionic liquids, the electrodeposition of Al occurs by the following reduction of Al 2 Cl −7 : 4 Al 2 Cl 7− + 3 e− = Al + 7 AlCl −4 + 3 Cl −

(9.2)

Many studies report the electrodeposition of Al in acidic chloroaluminate ionic liquids [3–9]. When the reduction potential of MCl mn−+m is more positive than that of Al 2 Cl −7 , it is possible to obtain the electrodeposit of pure metal. Alloys of the metal with Al are obtained if the reduction potential of MCl mn−+m is more negative than that of Al 2 Cl −7 . Moreover, the underpotential deposition of Al sometimes leads to the formation of Al–M alloys, even if the reduction potential of MCl mn−+m is more positive than that of Al 2 Cl −7 . For basic ionic liquids that contain AlCl3 at less than 50 mol%, the metal salt acts as a Lewis acid by accepting the chloride anions in the ionic liquids to form a chlorocomplex anion: MCl n + m Cl − = MCl mn−+m

(9.3)

Polymeric chlorocomplex anions also are known for some metals. Because the electrodeposition of Al does not occur within the electrochemical potential window of known alkylated imidazolium and pyridinium cations under the basic condition of room temperature, an electrodeposit of the pure metal can be obtained. It is possible to prepare the neutral ionic liquids by mixing the equimolar amounts of the organic chloride and AlCl3. When a metal salt is added to a neutral ionic liquid, the metal salt will act as either a Lewis base or acid to affect the Lewis acidity or basicity of the ionic liquid. Based on this observation, excess LiCl, NaCl, or HCl is added to maintain the neutrality of the ionic liquid. Figure 9.1 shows the metals that can be electrodeposited as pure metals or alloys in chloroaluminate ionic liquids. In the subsequent discussion, the potentials are given in relation to the Al/Al(III) electrode. This electrode consists of Al immersed in an acidic ionic liquid of 66.7 or 60.0 mol% AlCl3. The formal potentials of some redox couples in ionic liquids are illustrated in Figure 9.2 and are summarized in Tables 9.1 and 9.2. 9.1.1

Electrodeposition of Main Group Elements

The electrodeposition of alkaline and alkaline earth metals has been investigated in the neutral (buffered) ionic liquids for the purpose of applying the ionic liquids to electrolytes of rechargeable batteries using these metals as anodes. However, the reduction potentials of alkaline metals (Li, Na, K, Rb, and Cs) were more negative than those of alkylated imidazolium cations [36]. Reversible deposition and dissolution of some alkaline metals have been

c09.indd 131

1/20/2011 10:27:53 AM

132

c09.indd 132

1/20/2011 10:27:53 AM

Ra

Fr

An

Ln

Y

Actinide

Lanthanide

Ba

A

Sr

A

Ca

Cs

Rb

K

Sc

Ac

Th

Ce

La A

Db

Ta

A

Nb

V

Rf

Hf

Zr

A

Ti

Pa

Pr

Sg

W

Mo

A

Cr

U

Nd

Bh

Re

Tc

Mn

Np

Pm

Hs

Os

Ru

A

Fe

Pu

Sm

Mt

Ir

Rh

A

Co

Am

Eu

Cm

Gd

Uuu

B

Uun

Au

A

A

Ag

AB

Cu

Pt

B

Pd

A

Ni

Bk

Tb

Uub

AB

Hg

B

Cd

A

Zn

12

Cf

Dy

B

Tl

B

In

A

Ga

B

13/III

Es

Ho

A

Pb

AB

Sn

Ge

Si

C

14/IV

Fm

Er

A

Bi

AB

Sb

As

P

N

15/V

Md

Tm

Po

B

Te

Se

S

O

16/VI

No

Yb

At

I

Br

Cl

F

17/VII

18/VIII

Lr

Lu

Rn

Xe

Kr

Ar

A

11

Al

10

A

9

Mg

8

Na

7

Ne

6

Be

5

Li

4

He

3

H

2

Figure 9.1. Grayed elements can be deposited in the acidic (A) and/or basic (B) chloroaluminate ionic liquids. The hatched elements are deposited only as a constituent of alloys.

7

6

5

4

3

2

1

1

133

CHLOROALUMINATE IONIC LIQUIDS

–3

–3

Cs+ + e– = Cs Rb+ + e– = Rb K+ + e– = K Na+ + e– = Na Li+ + e– = Li

–2

–2

Decmp. of EMI+

CdCl42– + 2e– = Cd + 4Cl– InCl52– + 3e– = In + 5Cl– –1

–1

TlCl43– + e– = Tl + 4Cl–

Potential / V vs. Al/Al(III)

Sn(II) + 2e– = Sn HgCl42– + 2e– = Hg + 4Cl– 4Al2Cl7– + 3e– = Al + 7AlCl4–

PdCl42– + 2e– = Pd + 4Cl–

Zn(II) + 2e– = Zn 0

Pb(II) + 2e– = Pb

0

TeCl62– + 4e– = Te + 6Cl– TlCl63– + 3e– = TlCl43– + 2Cl–

Ga(I) + e– = Ga Sn(II) + 2e– = Sn

AuCl2– + e– = Au

Ga(III) + 2e– = Ga(I)

AuCl4– + 2e– = AuCl2– + 2Cl–

Co(II) + 2e– = Co 1

Cu(I) + e– = Cu

1

Ag(I) + e– = Ag Bi53+ + 3e– = 5Bi Cl2 + 2e– = Cl–

Hg22+ + 2e– = 2Hg 2Hg2+ + 2e– = Hg22+ 2

Cu(II) + e– = Cu(I)

2

2Al2Cl7– + Cl2 + 2e– = 4AlCl4–

3

3 acidic

basic or neutral

Figure 9.2. The formal potentials of some redox couples in the chloroaluminate ionic liquids.

reported in the presence of such additives as HCl [31–37] and SOCl2 [38,39], as these are considered to produce a passivation film on the deposited alkaline metals. The anodic dissolution of magnesium, Mg, has been reported as possible in the basic or buffered ionic liquids that consist of EMI+ and DMPI+ [46].

c09.indd 133

1/20/2011 10:27:53 AM

134

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

TABLE 9.1. Formal Potentials of Certain Redox Couples in Acidic Chloroaluminate Ionic Liquids Redox Couple Ga(III) + 2 e− = Ga(I) Ga(I) + e− = Ga Sn(II) + 2 e− = Sn Pb(II) + 2 e− = Pb SbCl +2 + 3 e− = Sb + 2 Cl − Bi 35+ + 3 e− = 5 Bi Fe(III) + e− = Fe(II) Fe(II) + 2 e− = Fe Co(II) + 2 e− = Co Ni(II) + 2 e− = Ni Cu(I) + e− = Cu

Cu(II) + e− = Cu(I) Zn(II) + 2 e− = Zn Ag(I) + e− = Ag 2 Hg 2 + + 2 e− = Hg 22 + Hg 22 + + 2 e− = 2 Hg

E0′/Va

AlCl3 (mol%)

Cation

Temperature (°C)

Reference

0.655 0.437 0.55 0.400 0.389** 0.925 2.036* 0.773* 0.71 0.894* 0.800* 1.017* 0.784 0.837 0.843 0.777 1.825 1.851 0.322 0.844 1.21 1.093

60.0 60.0 66.7 66.7 — 66.7 66.7 66.7 60.0 66.7 60.0 66.7 66.7 60.0 66.7 66.7 66.7 66.7 60.0 66.7 66.7 66.7

EMI EMI EMI EMI BP BP BP BP EMI BP BP BP BP EMI EMI MP BP MP EMI EMI EMI EMI

30 30 40 40 40 25 40 40 — 36 40 40 40 40 40 30 40 30 40 25 40 40

10 10 11 12 13 14 15 15 16 17 18 18 19 20 20 21 19 21 22 23 24 24

a

The potential is represented against Al/Al(III) on molar scale except those couples noted with * and ** (see text).

However, the deposition of Mg is impossible because of the instability of metallic Mg against these organic cations, whereas the formation of Al–Mg alloys containing Mg up to 2.2 at% has been observed in the acidic EMICl– AlCl3 ionic liquid [41]. Calcium and strontium dichlorides, CaCl2 and SrCl2, are soluble in an acidic EMICl–AlCl3 ionic liquid, and the codeposition of calcium and strontium with bismuth and copper has been examined in the acidic ionic liquid [42]. The electrodeposition of gallium, Ga, has been investigated in an acidic EMICl–AlCl3 ionic liquid [10]. A gallium species can be introduced by the anodic dissolution of metallic Ga or dissolution of GaCl3. Metallic Ga can be obtained by reducing monovalent gallium species, Ga(I), which can be produced by the following reduction of trivalent gallium species, Ga(III): Ga(III) + 2 e− = Ga(I)

c09.indd 134

(9.4)

1/20/2011 10:27:53 AM

135

CHLOROALUMINATE IONIC LIQUIDS

TABLE 9.2. Formal Potentials of Certain Redox Couples in Basic Chloroaluminate Ionic Liquids Redox Couple InCl 25 − + 3 e− = In + 5 Cl − TiCl 63− + 3 e− = TiCl 34− + 2 Cl − TiCl 34− + e− = Tl + 4 Cl − Sn(II) + 2 e− = Sn Sb Cl −4 + 3 e− = Sb + 4 Cl − TeCl 62 − + 4 e− = Te + 6 Cl − Te + 2 e− = Te(− II) Cu(I) + e− = Cu Cu(II) + e− = Cu(I) Pd Cl 24 − + 2 e− = Pd + 4 Cl Au Cl −2 + e− = Au Au Cl −4 + 2 e− = Au Cl −2 + 2 Cl − Hg Cl 24 − + 2 e− = Hg + 4 Cl −

E0′/Va

AlCl3 (mol%)

Cation

Temperature (°C)

Reference

−1.009 −1.096 −0.025 −0.005 −0.965 −0.900 −0.85 −0.523* −0.013 0.077 −1.030 −1.036 −0.647 0.046 −0.230 −0.110 0.310 0.374 −0.370

49.0 44.0 40.0 44.4 40.0 44.4 44.4 — 44.4 49.0 44.4 49.0 42.9 42.9 44.4 49.0 44.4 44.4 44.4

DMPI DMPI EMI EMI EMI EMI EMI BP EMI EMI EMI EMI BP BP EMI EMI EMI EMI EMI

27 27 30 30 30 30 40 40 30 30 30 30 40 40 40 40 40 40 40

25 25 26 26 26 26 11 13 27 27 27 27 19 19 28 28 29 29 24

a

The potentials are given against Al/Al(III) on the molar scale. The formal potentials are given at the specified AlCl3 compositions except for Sb(III)/Sb (marked with *).

and Ga(I) + 2 e− = Ga

(9.5)

The formal potential of Ga(III)/Ga(I) and Ga(I)/Ga is reported as 0.655 and 0.437 V, respectively. The electrodeposition of indium, In, has been reported in a basic DMPICl– AlCl3 ionic liquid [25]. Indium trichloride, InCl3, dissolves in the basic ionic liquid and forms a trivalent indium chlorocomplex anion, InCl 2− 5 , which can be reduced to metallic In by the following three-electron transfer reaction: InCl 25− + 3 e− = In + 5 Cl −

(9.6)

The formal potentials of In(III)/In in the basic ionic liquids of 49.0 and 44.4 mol% AlCl3 are reported as −1.009 and −1.096 V, respectively. A monovalent indium species, In(I), is not stable in this melt forming metallic In and InCl 2− 5 as shown in the following disproportionation reaction:

c09.indd 135

1/20/2011 10:27:54 AM

136

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

3InCl + 2 Cl − = 2In + InCl 25−

(9.7)

Metallic thallium, Tl, can be electrodeposited in a basic EMICl–AlCl3 ionic liquid [26]. It is possible to dissolve thallium trichloride, TlCl3, to form a trivalent thallium chlorocomplex anion, TlCl 63− , which is reducible to a monovalent thallium chlorocomplex anion, TlCl 3− 4 : TlCl 63− + 3 e− = TlCl 34− + 2 Cl −

(9.8)

The formal potentials of Tl(III)/Tl(I) in the basic ionic liquids of 40.0 and 44.4 mol% AlCl3 are reported as −0.025 and −0.005 V, respectively. Thallium metal is obtained by another reduction of TlCl 3− 4 : TlCl 34− + e− = Tl + 4 Cl −

(9.9)

The formal potentials of Tl(I)/Tl in the basic ionic liquids of 40.0 and 44.4 mol% AlCl3 are reported as −0.965 and −0.900 V, respectively. The electrodeposition of tin, Sn, has been reported in both basic and acidic EMICl–AlCl3 ionic liquid [11]. A divalent tin species, Sn(II), can be introduced by the anodic dissolution of metallic tin. The introduction of a tetravalent tin species, Sn(IV), is also possible by dissolving tin tetrachloride, SnCl4. However, the evaporation of SnCl4 occurs in the case of an acidic ionic liquid. The irreversible reduction of Sn(IV) to Sn(II) occurs at around 0.91 and −0.9 V in the acidic and basic ionic liquids, respectively. The electrodeposition of metallic Sn is possible by the following reduction of Sn(II): Sn(II) + 2 e− = Sn

(9.10)

The formal potentials of Sn(II)/Sn in the acidic (66.7 mol% AlCl3) and basic (44.4 mol% AlCl3) ionic liquids are reported as 0.55 and −0.85 V, respectively. The electrodeposition of lead, Pb, has been investigated in an acidic EMICl– AlCl3 ionic liquid [12]. A divalent lead species, Pb(II), can be introduced by dissolving lead dichloride, PbCl2, or the anodic dissolution of metallic Pb. Metallic Pb is obtained by the following reduction of Pb(II): Pb(II) + 2 e− = Pb

(9.11)

The formal potential of Pb(II)/Pb is reported as 0.400 V. The electrodeposition of antimony, Sb, has been reported in both acidic and basic BPCl–AlCl3 ionic liquid [13,43]. Antimony trichloride, SbCl3, is soluble in both acidic and basic ionic liquids. In an acidic ionic liquid, a cationic trivalent antimony species, SbCl +2 , is formed and can be reduced to the following metallic state: SbCl +2 + 3 e− = Sb + 2 Cl −

c09.indd 136

(9.12)

1/20/2011 10:27:54 AM

CHLOROALUMINATE IONIC LIQUIDS

137

The formal potential is dependent on the concentration of Cl− and is given as follows: E 0′ = Eeq −

RT 2 RT ln[SbCl 2+ ] + ln[Cl − ] 3F 3F

(9.13)

The experimental E 0′ value is reported as 0.389 V. However, a trivalent antimony chlorocomplex anion, SbCl −4 , is dominant in the basic ionic liquid and is also reducible to metallic Sb at more negative potential: SbCl −4 + 3 e− = Sb + 4 Cl −

(9.14)

The formal potential for this equilibrium is given by the following: E 0′ = Eeq −

RT 4 RT ln[SbCl 4− ] + ln[Cl − ] 3F 3F

(9.15)

and is reported as −0.523 V. The electrodeposition of bismuth, Bi, has been investigated in an acidic BPCl–AlCl3 ionic liquid [14]. Bismuth trichloride, BiCl3, is soluble in the acidic ionic liquid forming a trivalent bismuth species, Bi(III), which can be reduced 3+ to a cluster cation, Bi 3+ 5 , at 0.95 V. Metallic Bi can be obtained by reducing Bi 5 + via an intermediate species, Bi 5 : Bi 35+ + 2 e− = Bi +5

(9.16)

Bi +5 + e− = 5 Bi

(9.17)

and

The formal potential of Bi 3+ 5 / Bi is reported as 0.925 V. The electrodeposition of tellurium, Te, has been reported in a basic EMICl– AlCl3 ionic liquid [27]. Tellurium tetrachloride, TeCl4, is soluble in the basic ionic liquid to form a tetravalent tellurium chlorocomplex anion, TeCl 62− , which can be reduced to metallic Te by the following four-electron transfer reaction: TeCl 62− + 4 e− = Te + 6 Cl −

(9.18)

The formal potentials of Te(IV)/Te in the basic ionic liquids of 44.4 and 49.0 mol% AlCl3 are reported as −0.013 and 0.077 V, respectively. Metallic Te is reduced to Te(−II), which is soluble in the ionic liquid: Te + 2 e− = Te(− II)

c09.indd 137

(9.19)

1/20/2011 10:27:54 AM

138

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

The formal potentials of Te/Te(−II) in the basic ionic liquids of 44.4 and 49.0 mol% AlCl3 are reported as −1.030 and −1.036 V, respectively. It is also possible to obtain metallic Te by oxidizing Te(−II). 9.1.2

Electrodeposition of Transition Metals

The electrochemical behavior of titanium species has been investigated in an acidic EMICl–AlCl3 ionic liquid [44]. Titanium dichloride, TiCl2, dissolves up to 60 and 170 mmol dm−3 at 80 °C in the acidic ionic liquids of 60.0 and 66.7 mol% AlCl3, respectively. The divalent titanium species, Ti(II), is less stable in the weakly acidic ionic liquid, leading to the disproportionation into insoluble titanium trichloride and metallic Ti. Metallic Ti is impossible to obtain probably because of the large overpotential of the electrodeposition of metallic Ti whose reduction potential is expected to be more positive than that of Al. However, Al–Ti alloys containing Ti up to approximately 19 at% can be obtained by the codeposition of Ti and Al. The electrodeposition of chromium, Cr, has been investigated in an acidic BPCl–AlCl3 ionic liquid [45]. Chromium dichloride, CrCl2, dissolves in the ionic liquid up to 0.31 mol dm−3 at 80 °C. It is reported that the reduction of a divalent chromium species, Cr(II), occurs on a platinum electrode at a potential close to that of the bulk electrodeposition of Al to give the electrodeposit containing Cr at 94 at%. The Al content in the deposit increases with the decrease in the deposition potential. Al–Cr alloys containing Cr at 0 to 94 at% have been prepared by controlling the deposition potential. In a basic EMICl– or DMPICl–AlCl3 ionic liquid, CrCl2 dissolves in the basic ionic liquid and forms a divalent chromium chlorocomplex anion, CrCl 2− 4 , which is not, in this case, reduced to metallic state within the electrochemical potential window of the ionic liquid [46]. The electrodeposition of iron, Fe, has been investigated in BPCl– and EMICl–AlCl3 ionic liquids [47–50]. In the basic ionic liquids, iron dichloride, FeCl2, dissolves and forms a divalent iron chlorocomplex anion, FeCl4 [2– 15,25–27,30–50], which is not reducible to metallic Fe within the electrochemical potential window of the EMICl–AlCl3 ionic liquid [49]. In the neutral and neutral-like ionic liquids, metallic Fe can be obtained by reducing a divalent iron species, but the dissolved iron species in these ionic liquids are not identified [48,49]. In the acidic ionic liquids, FeCl2 dissolves and forms a divalent iron species, which is reducible to metallic Fe at 0.3 to 0.4 V prior to the electrodeposition of Al [47,49]. The formal potential of Fe(II)/Fe in the acidic BPCl–AlCl3 ionic liquid is reported as 0.773 V on the mole fraction scale [15]. The formation of Al–Fe alloys containing Fe at 0 to 61 at% also has been reported in an acidic EMICl–AlCl3 ionic liquid [50]. The electrodeposition of cobalt, Co, has been studied extensively in acidic ionic liquids [16,17,50–59]. A divalent cobalt species, Co(II), can be introduced into the ionic liquids by dissolving cobalt dichloride, CoCl2 or the anodic dissolution of Co metal. The formal potentials of Co(II)/Co in the acidic EMICl–

c09.indd 138

1/20/2011 10:27:54 AM

CHLOROALUMINATE IONIC LIQUIDS

139

AlCl3 and BPCl–AlCl3 ionic liquids are reported as 0.71 [16] and 0.894 [17] V (on the mole fraction scale), respectively. The electrodeposition of Al–Co alloys containing Co at 33 to 50 at% occurs at a potential more negative than the formal potential of Co(II)/Co. The nucleation and growth processes of Co and Al–Co alloys are studied on a glassy carbon electrode in EMICl–AlCl3 ionic liquid [16] and on a single crystal Au electrode in the BMICl–AlCl3 ionic liquid [53,54]. In a basic BPCl–AlCl3 ionic liquid, CoCl2 dissolves and forms a divalent cobalt chlorocomplex anion, CoCl 2− 4 , which is not reducible to metallic Co [17]. The electrodeposition of nickel, Ni, has been studied in acidic ionic liquids [18,49,50,54–56]. A divalent nickel species, Ni(II), can be introduced into the ionic liquids by dissolving nickel dichloride, NiCl2, or through the the anodic dissolution of Ni metal. The reduction of Ni(II) occurs at around 0.38 V, producing metallic nickel in the acidic EMICl–AlCl3 ionic liquid [55]. The formal potential of Ni(II)/Ni in an acidic BPCl–AlCl3 ionic liquid is reported as 0.800 V on the mole fraction scale [18]. The electrodeposition of Al–Ni alloys occurs at the potential more negative than the deposition potential of metallic Ni. The nucleation and growth processes of Ni and Al–Ni alloys are studied on a glassy carbon electrode in the EMICl–AlCl3 ionic liquid [55] and a single crystal Au electrode in the BMICl–AlCl3 ionic liquid [54]. In a neutral buffered EMICl–AlCl3 ionic liquid, the anodic oxidation of Ni metal results in NiCl2, which is insoluble in this ionic liquid [49]. In a basic BPCl–AlCl3 ionic liquid, NiCl2 dissolves and forms a divalent nickel chlorocomplex anion, NiCl 2− 4 , which cannot be reduced to metallic Ni [18]. The electrodeposition of copper, Cu, has been studied in MPCl– [21], BPCl– [19], EMICl– [20,42,49,50,57–59], and BMICl–AlCl3 [60] ionic liquids. In acidic ionic liquids, it is possible to introduce monovalent and divalent copper species, Cu(I) and Cu(II), by dissolving copper chloride and dichloride, respectively, as well as by the anodic dissolution of metallic copper. Metallic Cu can be obtained by the reduction of Cu(I). The formal potentials of Cu(I)/Cu in an acidic MPCl–, BPCl–, and EMICl–AlCl3 are reported as 0.777 [21], 0.784 [19], and 0.837 [20] V, respectively. Cu(I) also can be obtained by the reduction of Cu(II), of which the formal potentials in the acidic MPCl– and BPCl–AlCl3 are reported as 1.851 [21] and 1.825 [19] V, respectively. The nucleation and growth process of Cu are investigated on a single crystal Au electrode [60]. The formation of Al–Cu alloys containing Al up to 43 at% occurs at the potential more negative than the formal potential of Cu(I)/Cu. However, it is reported that the displacement of Al with Cu occurs when the Al–Cu alloy is immersed in the ionic liquid containing Cu(I) [59]. In basic ionic liquids, it is possible to introduce the monovalent copper chlorocomplex anions, CuCl2 and CuCl 3− 4 , by dissolving CuCl and by the anodic dissolution of Cu metal. CuCl2 also dissolves in the basic ionic liquids and forms a divalent copper chlorocomplex anion, CuCl 2− 4 [19]. The formal potentials of Cu(I)/Cu and Cu(II)/Cu(I) in the BPCl–AlCl3 ionic liquid are reported as −0.647 and 0.046 V, respectively [19]. The electrodeposition of copper is also possible in a basic EMICl–AlCl ionic liquid [58].

c09.indd 139

1/20/2011 10:27:54 AM

140

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

The electrochemical study on zinc, Zn, has been reported in EMICl– and BMICl–AlCl3 ionic liquids [22,41–63]. A divalent zinc species, Zn(II), can be introduced by the anodic dissolution of Zn metal in the acidic ionic liquid. The electrodeposition of metallic Zn is possible in the acidic ionic liquid. The formal potential of Zn(II)/Zn is reported as 0.322 V in the EMICl–AlCl3 ionic liquid [22]. The nucleation and growth processes have been studied on some polycrystalline metal electrodes in the EMICl–AlCl3 ionic liquid [22] as well as on a single crystal gold electrode in the BMICl–AlCl3 ionic liquid [43]. The formation of Al–Zn alloys was not reported. The electrodeposition of Zn has been investigated in the neutral EMICl–AlCl3 ionic liquid mixed with the 1 : 1 mixture of EMICl and ZnCl2 [62]. The divalent zinc species in this ionic liquid is considered to be ZnCl −3 , which can be reduced to metallic Zn at around −0.8 V on a glassy carbon electrode. In a basic EMICl–AlCl3 ionic liquid, the anodic dissolution of Zn metal results in the formation of a divalent zinc chlorocomplex anion, ZnCl −4 , which cannot be reduced within the electrochemical potential window of the ionic liquid [41]. The electrodeposition of Al–Nb alloys has been reported in a BPCl–AlCl3 ionic liquid containing NbCl5 or Nb3Cl8 [64]. A pentavalent niobium species is reduced chemically by Al powder prior to the electrodeposition of Al–Nb alloys. The Al–Nb alloys containing Nb up to 29.3 wt% have been obtained at 90 °C to 140 °C. The electrodeposition of palladium, Pd, has been studied in a EMICl–AlCl3 ionic liquid [28,65]. Palladium dichloride, PdCl2, is soluble in the basic ionic liquid but is less soluble in the neutral and acidic ionic liquids. In the basic ionic liquid, the divalent palladium species is considered to be PdCl 2− 4 , which is reducible to metallic Pd as follows: PdCl 24− + 2 e− = Pd + 4 Cl −

(9.20)

The formal potential of Pd(II)/Pd in the basic ionic liquids of 44.4 and 49.0 mol% AlCl3 are reported as −0.230 and −0.110 V, respectively. The electrodeposition of Pd in the acidic ionic liquid also is reported as occurring at 1.3–1.6 V. The electrodeposition of silver, Ag, has been investigated in acidic EMICl– and BMICl–AlCl3 ionic liquids [23,50,65–68]. Silver chloride, AgCl, is soluble in the acidic ionic liquids. The formal potential of Ag(I)/Ag in the EMICl– AlCl3 ionic liquid is reported as 0.844 V [68]. The nucleation and growth processes of metallic silver have been studied on some polycrystalline metal electrodes in the EMICl–AlCl3 ionic liquid [66] and the highly oriented pyrolytic graphite (HOPG) electrode in the BMICl–AlCl3 ionic liquid [67,68]. The electrodeposition of Al–Ag alloys is reported in the acidic EMICl–AlCl3 ionic liquid [23,50]. AgCl is soluble also in a basic EMICl–AlCl3 ionic liquid to form AgCl 3− 4 as a predominant species [69]. However, there is no description of the deposition of metallic Ag in the basic ionic liquid.

c09.indd 140

1/20/2011 10:27:54 AM

CHLOROALUMINATE IONIC LIQUIDS

141

The electrodeposition of cadmium, Cd, is possible in the basic and neutral EMICl–AlCl3 ionic liquids [61,62]. In the basic ionic liquid, cadmium dichloride, CdCl2, dissolves and forms a divalent cadmium chlorocomplex anion, CdCl 2− 4 [70], which is reducible to metallic Cd at around −1.2 V. In a neutral buffered ionic liquid, the electrodeposition of Cd is observed at around −1.0 V, which is slightly more positive than in the basic ionic liquid probably because of the formation of more reducible species, CdCl −3 [70a]. The electrodeposition of lanthanum, La, has been attempted in neutral and acidic EMICl–AlCl3 ionic liquids [71,72]. Lanthanum trichloride, LaCl3, is soluble in the acidic ionic liquid. The solubility of LaCl3 in the acidic ionic liquid containing AlCl3 at 66.7 mol% is reported as 45 mmol kg−1 at 25 °C. The deposition of metallic La is not confirmed, whereas the reduction of a trivalent lanthanum species is suggested in the presence of SOCl2. The electrodeposition of Al–La alloys also has been examined in the acidic ionic liquid where the content of La in the Al–La alloys is less than 0.1 at%. The electrodeposition of platinum, Pt, has been reported in an acidic BTMACl–AlCl3 ionic liquid [9]. A divalent platinum species, PtCl 62− , is introduced by dissolving either (BTMA)2PtCl6 or (TEA)2PtCl6. The deposition of metallic Pt occurs at a potential more positive (∼100 mV) than that of the bulk deposition of Al. The formation of Al–Pt alloys containing Pt up to 13 at% also is reported. The electrodeposition of gold, Au, has been studied in a basic EMICl–AlCl3 ionic liquid [29]. Gold trichlolride, AuCl3, is soluble in an ionic liquid and forms a trivalent gold chlorocomplex anion, AuCl −4 , which is reducible to a monovalent gold chlorocomplex anion, AuCl −2 , electrochemically: AuCl −4 + 2 e− = AuCl −2 + 2 Cl −

(9.21)

It is also possible to introduce AuCl −2 by the anodic dissolution of Au metal. Metallic Au can be obtained by the following reduction of AuCl −2 : AuCl −2 + e− = Au + 2 Cl −

(9.22)

The formal potentials of Au(III)/Au(I) and Au(I)/Au are reported as 0.374 and 0.310 V, respectively. The electrodeposition of mercury, Hg, has been studied in both basic and acidic EMICl–AlCl3 ionic liquids [24]. Mercury dichloride, HgCl2, is soluble in the acidic ionic liquid to form a divalent mercury species, Hg2+, which can be reduced irreversibly as follows to a cluster cation, Hg 2+ 2 : 2 Hg 2+ + 2 e− = Hg 22+

(9.23)

It is also possible to introduce Hg 2+ 2 by dissolving Hg2Cl2. Metallic Hg can be obtained by the following reduction of Hg 2+ 2 :

c09.indd 141

1/20/2011 10:27:54 AM

142

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

Hg 22+ + 2 e− = 2 Hg

(9.24)

The formal potentials of Hg 2+ /Hg 22+ and Hg 2+ 2 /Hg are reported as 1.21 and 1.093 V, respectively. In a basic ionic liquid, HgCl2 dissolves and forms a diva2− lent mercury chlorocomplex anion, HgCl 2− 4 . The reduction of HgCl 4 results in the following deposition of metallic Hg: HgCl 24− + 2 e− = Hg + 4 Cl −

(9.25)

The formal potential of Hg(II)/Hg in the basic ionic liquid is reported as −0.370 V.

9.2

NONCHLOROALUMINATE IONIC LIQUIDS

9.2.1 Tetrafluoroborate (BF4- ) and Hexafluorophosphate (PF6- ) Ionic Liquids The electrodeposition of several metals and alloys has been investigated in tetrafluoroborate ionic liquids. In contrast to the chloroaluminate ionic liquids, the tetrafluoroborate ionic liquids are considerably more stable against moisture and are expected to be applicable to practical use. Moreover, the codeposition of the metals derived from the ionic liquids does not occur because the tetrafluoroborate anion is not reducible within the cathodic potential limits of known organic cations. However, the introduction of metal ions is not straightforward because the ionic liquids are neutral in the concept of Lewis acid and base. Some metal ions can be introduced simply by dissolving their tetrafluoroborate salts. However, this method is limited to a few metals, for which the anhydrous tetrafluoroborate salts are available. It is difficult to adjust the Lewis acidity or basicity by the addition of BF3 or other binary fluorides. Thus, organic chlorides usually are added as a Lewis base to dissolve metal chlorides, which act as Lewis acids to form their chlorocomplex anions, the same as those observed in basic chloroaluminate ionic liquids. The hexafluorophosphate ionic liquids are similar to the tetrafluoroborate one’s ionic liquids, but the former is not completely miscible with water. Thus, the PF6− ionic liquids sometimes are regarded as hydrophobic. However, it should be noted that the PF6− anion can undergo hydrolysis when it come into contact with an acidic aqueous solution. There is no common reference electrode in these ionic liquids, but the Al/ Al(III) electrode is used in EMICl–EMIBF4 ionic liquids. 9.2.1.1 The Electrodeposition of Main Group Metals. The electrodeposition of germanium, Ge, has been studied in a BMIPF6 ionic liquid [54,73–75]. Germanium halides, GeCl4, GeBr4, and GeI4, dissolve slowly in the ionic liquid. It is presumed that the dissolved tetravalent germanium species is reduced to

c09.indd 142

1/20/2011 10:27:54 AM

NONCHLOROALUMINATE IONIC LIQUIDS

143

metallic Ge via an intermediate divalent germanium species. The electrodeposition process of nanoscaled Ge has been investigated by in situ scanning tunneling microscopy (STM). The electrodeposition of antimony [76] and indium-antimony [77] alloys has been reported in a basic EMICl–EMIBF4 ionic liquid. Antimony trichloride, SbCl3, dissolves in the ionic liquid and forms SbCl −4 , the same as in the basic chloroaluminate ionic liquid. Metallic Sb can be obtained by the cathodic reduction of SbCl −4 , as shown in equation. (9.14). The formal potential of Sb(III)/Sb is reported as −0.27 V versus Al/Al(III) in the ionic liquid containing Cl− at 0.11 M. In addition, the oxidation of SbCl −4 leads to the formation of a pentavalent antimony species, SbCl 6−: SbCl 6− + 2 e− = SbCl 4− + 2 Cl −

(9.26)

The formal potential of Sb(V)/Sb(III) is reported as 0.60 V versus Al/Al(III) in the ionic liquid containing Cl− at 0.11 M. However, a trivalent indium species, InCl 2− 5 , can be introduced by dissolving indium trichloride, InCl3. The reduction of InCl 2− 5 occurs at around −0.9 V versus Al/Al(III) in the ionic liquid containing Cl− at 0.35 M. The electrodeposition of InSb alloys has been reported as produced from the ionic liquid containing both Sb(III) and In(III). 9.2.1.2 Electrodeposition of Transition Metals. The electrodeposition of copper, Cu, has been reported in a basic EMICl–EMIBF4 ionic liquid [78]. A monovalent copper chlorocomplex anion, CuCl −2 , is introduced by the anodic dissolution of Cu metal. The electrodeposition of metallic Cu is possible by reducing CuCl −2 as follows: CuCl −2 + e− = Cu + 2 Cl −

(9.27)

The formal potential of Cu(I)/Cu is reported as −0.633 V versus Al/Al(III). The oxidation of CuCl −2 leads to the following formation of a divalent copper chlorocomplex anion, CuCl 2− 4 : CuCl 24− + e− = Cu + 2 Cl −

(9.28)

The formal potential of Cu(II)/Cu(I) is reported as 0.192–0.202 V versus Al/ Al(III). The electrodeposition of silver, Ag, has been reported in a neutral EMIBF4 ionic liquid [79]. Silver tetrafluoroborate, AgBF4, dissolves in the ionic liquid up to about 0.2 M at room temperature and forms a monovalent silver species. The reduction of the monovalent silver species occurs at around 1 V versus Al/Al(III). AgCl dissolves in a basic EMICl–EMIBF4 ionic liquid by forming a monovalent silver complex anion, AgCl 3− 4 , which can be reduced to Ag metal [80] although the reduction of AgCl 3− has not been reported in basic chloro4 aluminate ionic liquids.

c09.indd 143

1/20/2011 10:27:54 AM

144

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

The electrodeposition of cadmium, Cd, has been studied in a basic EMICl– EMIBF4 ionic liquid [81]. Cadmium dichloride, CdCl2, dissolves in the ionic liquid and forms a divalent cadmium chlorocomplex anion, CdCl 2− 4 , which can be reduced to metallic Cd at around −1.1 V versus Al/Al(III): CdCl 24− + 2 e− = Cd + 4 Cl −

(9.29)

The electrodeposition of some palladium alloys (Pd–Au [82], Pd–Ag [80], Pd–In [83] and Pd–Cu [84]) has been reported in a basic EMICl–EMIBF4 ionic liquid. PdCl2 dissolves in the ionic liquid to form a divalent palladium complex anion, PdCl 2− 4 . Au(I) can be introduced by the anodic oxidation of Au metal to form AuCl −2 . A Pd–Au solid solution can be obtained by the cathodic reduction of Pd(II) and Au(I) above 80 °C. A Pd–Ag alloy can be obtained by the cathodic reduction of Pd(II) and Ag(I) from 35 °C to 120 °C. CuCl2 dissolves by forming CuCl 2− 4 , which can be reduced to Cu(0) via Cu(I). The potentiostatic cathodic reduction of Pd(II) and Cu(II) gives the Pd–Cu alloy on an indium-tin oxide (ITO) electrode. A Pd–In alloy can be obtained by the potentiostatic cathodic reduction of Pd(II) and In(III). Sodium tetrachloroaurate(III), NaAuCl4, is soluble in a neutral EMIBF4 and BMIBF4 [85]. The cathodic reduction of AuCl −4 leads to the deposition of nanoscaled Au metal via a monovalent gold complex, AuCl −2 , which is likely to disproportionate to Au and AuCl −4 : 3AuCl −2 = AuCl −4 + 2Au + 2Cl −

(9.30)

Electrodeposition and anodic dissolution of Au has been investigated with an electrochemical quartz crystal microbalance (EQCM) in a basic EMICl– EMIBF4 ionic liquid [86]. The redox reaction between Au and AuCl −2 is reported to be reversible with a current efficiency of almost 100%. 9.2.2

Chlorozincate Ionic Liquids

It has been reported that the mixture of EMICl and zinc dichloride, ZnCl2, forms an ionic liquid, which melts at around room temperature and is less reactive against water than the chloroaluminate ionic liquids. Becuase ZnCl2 acts as a Lewis acid, it is possible to adjust the Lewis acidity or basicity by changing the ratio of ZnCl2 to EMICl as well as the chloroaluminate ionic liquids. The electrochemical potential window of this ionic liquid is dependent on the molar fraction of ZnCl2 [87]. In the basic ionic liquid, which contains ZnCl2 at less than 33.3 mol%, the predominant ionic species are EMI+, − ZnCl 2− 4 , and Cl . Thus, the cathodic and anodic limits are determined by the reduction of EMI+ and the oxidation of Cl−, respectively. The deposition of metallic zinc does not occur in the basic condition because ZnCl 2− 4 is not reducible at a potential more positive than the cathodic decomposition of EMI+. In an acidic ionic liquid that contains ZnCl2 at more than 33.3 mol%, the pre-

c09.indd 144

1/20/2011 10:27:55 AM

NONCHLOROALUMINATE IONIC LIQUIDS

145

− − − dominant ionic species are EMI+, ZnCl 2− 4 , ZnCl 3 , Zn 2 Cl 5 , and Zn 3 Cl 7 . The cathodic limit is determined by the electrodeposition of metallic zinc as a − − result of the reduction of the chlorozincate anions, ZnCl 2− 4 , ZnCl 3 , Zn 2 Cl 5 , and − Zn 3 Cl 7 . The electrodeposition of zinc has been studied in an acidic EMICl– ZnCl2 ionic liquid that contains ZnCl2 at 50 mol% [82,88]. The anodic limit is considered to be the chlorine evolution derived from the oxidation of the chlorozincate anions. The dissolution of metal chlorides is expected to be the same as in the chloroaluminate ionic liquids. The electrode potentials usually are measured against a Zn/Zn(II) reference electrode, which consists of a zinc electrode immersed in the EMICl–ZnCl2 ionic liquid containing ZnCl2 at 50 mol%.

9.2.2.1 Electrodeposition of Main Group Elements and Transition Metals. A mixture of BMICl–SbCl3 (55 : 45 in mol%) is soluble in an EMICl–ZnCl2 (55 : 45 in mol%) ionic liquid [89]. The electrodeposition of Sb and SbZn is possible by the cathodic reduction of SbCl −4 and ZnCl −4 on a gold electrode at 50 °C. The electrodeposition of tellurium, Te, has been investigated in an acidic EMICl–ZnCl2 ionic liquid [90]. Tellurium tetrachloride, TeCl4, is soluble in the ionic liquid. The electrodeposition of metallic tellurium occurs at around 1 V versus Zn/Zn(II). The formation of the Zn–Te alloys also occurs at more negative potentials. The continued reduction of metallic tellurium to Te(−II) is reported as well just as in the chloroaluminate ionic liquid. The electrodeposition of iron, Fe, has been reported in an acidic EMICl– ZnCl2 ionic liquid [91]. Iron dichloride, FeCl2, is soluble in the ionic liquid that contains ZnCl2 at 40 mol%. The dissolved iron species is reducible at around 0.1 V versus Zn/Zn(II), resulting in the electrodeposition of metallic iron. Zn–Fe alloys are obtained in this ionic liquid. The electrodeposition of cobalt, Co, has been investigated in an acidic EMICl–ZnCl2 ionic liquid [92,93]. Cobalt dichloride, CoCl2, dissolves in the ionic liquid at 80 °C. The electrodeposition of metallic cobalt is observed at around 0.2 V versus Zn/Zn(II). The formation of Zn–Co alloys also is reported at more negative potentials. The electrodeposition of Zn–Co alloys has been reported in BPCl– and EMICl–CoCl2–ZnCl2 ionic liquids [94,95]. It was reported that Zn–Co–Dy alloys can be obtained from an EMICl–ZnCl2 ionic liquid containing both CoCl2 and DyCl3 [93]. The electrodeposition of Zn–Ni alloys has been reported in a EMICl– ZnCl2–NiCl2 ionic liquid [96]. The Zn–Ni alloys containing Ni at 12.3 to 98.6 mol% were obtained by changing the deposition potential as well as the composition of the ionic liquid. The electrodeposition of Ni metal and Zn–Ni alloy also has been investigated in a EMICl–ZnCl2 (64.0 : 36.0 in mol%) ionic liquid containing NiCl2 [97]. In this ionic liquid, a divalent nickel species, Ni(II), is expected to be formed in the presence of such chloride ion acceptors as ZnCl −3 , Zn 2Cl −5 , and Zn 3Cl −7 . Ni metal and Zn–Ni alloy can be deposited on a tungsten electrode by potentiostatic cathodic reduction at 150 °C.

c09.indd 145

1/20/2011 10:27:55 AM

146

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

The electrodeposition of copper, Cu, has been investigated in an acidic EMICl–ZnCl2 ionic liquid [98]. A monovalent copper species, Cu(I), is introduced by dissolving copper chloride, CuCl, or by the anodic dissolution of Cu metal. The electrodeposition of metallic cupper is possible by reducing Cu(I) at 0.5 ∼ 0.6 V versus Zn/Zn(II). The formation of Zn–Cu alloys also was reported. The electrodeposition of cadmium, Cd, has been studied in an acidic EMICl–ZnCl2 ionic liquid [99]. Cadmium dichloride, CdCl2, dissolves in the ionc liquid and forms a divalent cadmium species, which is reducible to metallic Cd prior to the bulk deposition of metallic Zn. The electrodeposition of Zn-Cd alloys also was reported. PtCl2 dissolves in a Lewis acidic EMICl–ZnCl2 (60.0 : 40.0 in mol%) ionic liquid [100] to form a divalent platinum complex, PtCl 2− 4 . Cathodic reduction at a tungsten electrode in this ionic liquid leads to the deposition of Pt–Zn alloy. 9.2.3

Bis(trifluoromethylsulfonyl)amide ( N(CF3 SO2 )2- ) Ionic Liquids

Most ionic liquids melting at low temperature consist of alkylated imidazolium cations, and there is no good explanation as to why the melting point and viscosity become lower in the presence of these imidazolium cations. However, the cathodic stability of the imidazolium cations is not sufficient for the electrodeposition of the base metals, which have not been obtained in roomtemperature ionic liquids. It has been known that many kinds of room-temperature ionic liquids can be prepared by combining various quaternary ammonium cations with bis(trifluoromethylsulfonyl)imide, N(CF3 SO2 )−2 [101–105]. Furthermore, these N(CF3 SO2 )−2 ionic liquids are not only stable against moisture but also are immiscible with water, whereas a certain amount of water dissolves in the N(CF3 SO2 )−2 ionic liquids. Therefore, these N(CF3 SO2 )−2 ionic liquids are promising candidates as supporting electrolytes for practical electrodeposition processes. 9.2.3.1 Electrodeposition of Main Group Elements. The electrodeposition of lithium, Li, has been studied in some N(CF3 SO2 )−2 ionic liquids consisting of the aliphatic and alicyclic ammonium cations [106–113]. LiN(CF3SO2)2 often is added to the ionic liquids as a Li(I) source. Because these ammonium cations are more stable against lithium metal than alkylated imidazolium cations, the reversible deposition and dissolution of Li metal are reported to be possible in these ionic liquids. However, it has been suggested that some surface film (often called “solid-electrolyte interphase [SEI] film”) is formed as a result of the cathodic decomposition of organic cations [108,109]. The formal potential of Li(I)/Li is in the range of −3.1–−3.2 V versus Fc/Fc+ in ionic liquids [113]. The electrodeposition of lithium and potassium also has been reported in the N(CF3 SO2 )−2 ionic liquids comprising tributylhexylphophonium and tributylhexylammonium [112].

c09.indd 146

1/20/2011 10:27:55 AM

NONCHLOROALUMINATE IONIC LIQUIDS

147

The deposition and dissolution of magnesium, Mg, has been reported in a BMPN(CF3SO2)2 ionic liquid containing a Grignard reagent, phenylmagnesium chloride, at 150 °C [114]. The electrodeposition of aluminum, Al, has been reported from some N(CF3 SO2 )−2 ionic liquids with the addition of AlCl3 [115,116]. The biphasic behavior is observed at higher AlCl3 concentrations. The cathodic reduction of aluminum species forming at a certain AlCl3 concentration leads to the deposition of Al metal. The aluminum species in the ionic liquids have been studied by spectroscopy and theoretical calculation [117,118]. SiCl4 is soluble in a BMPN(CF3SO2)2 ionic liquid [119,120]. The cathodic reduction of the silicon species leads to the deposition of nanoscale silicon on HOPG and gold. Although the deposited silicon has not been indentified as crystalline silicon by X-ray diffraction (XRD), the change of the surface morphology during the cathodic reduction has been observed by in situ STM [120]. The electrochemical reduction of GaCl3 has been investigated on a gold electrode in a BMPN(CF3SO2)2 ionic liquid. [121] The electrochemical and electroless deposition of Ga has been indicated by STM and X-ray photoelectron spectroscopy (XPS) analyses. The electrodeposition of germanium, Ge, has been reported from a BMPN(CF3SO2)2 ionic liquid [122]. GeCl4 is soluble in the ionic liquid and is reducible to Ge(0) via Ge(II). It is possible to obtain SixGe1−x from the ionic liquid containing both Ge(II) and Si(IV). The deposition of selenium, Se, and indium, In, has been reported in a BMPN(CF3SO2)2 ionic liquid [123]. SeCl4 is soluble in the ionic liquid and can be reduced to Se with several reduction steps. The deposit obtained at 100 °C has been confirmed as crystalline Se by XRD. InCl3 can be used as an In source in the ionic liquid. The reduction of In(III) is possible on a platinum electrode at 25 °C. The electrochemical behavior of tin, Sn, has been reported in a BMPN(CF3SO2)2 ionic liquid at 25 °C [124–126]. Sn(II) can be introduced by anodic oxidation of Sn metal [124,125] and by an addition of Sn[N(CF3SO2)2]2 or SnCl2 [126]. The formal potential of Sn(II)/Sn is −0.14 V versus Fc/Fc+. The granular deposit can be obtained by cathodic reduction at 25 °C and has been confirmed as Sn metal by XRD. The deposition of Sn metal also has been examined by EQCM [125]. The deposition and stripping of cesium, Cs, have been studied on a mercury electrode in a tributylmethylammonim N(CF3SO2)2 ionic liquid [127]. Cs(I) is introduced by anodic oxidation of Cs liquid metal at 110 °C. The redox reaction of Cs(I)/Cs(Hg) is reported to be reversible. The formal potential of Cs(I)/ Cs(Hg) is −2.273 V versus Fc/Fc+ at 30 °C. 9.2.3.2 Electrodeposition of Transition Metals. The electrochemical reduction of TiCl4 in a BMIN(CF3SO2)2 [128] and TiBr4 in a BMPN(CF3SO2)2 [129] has been reported. The change of a Au electrode surface has been observed

c09.indd 147

1/20/2011 10:27:55 AM

148

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

during the cathodic reduction of Ti(IV) in the BMIN(CF3SO2)2 ionic liquid by STM [128]. However, electrodeposition of Ti metal has not been confirmed in the BMPN(CF3SO2)2 ionic liquid containing TiBr4 [129]. The electrodeposition of manganese, Mn, has been investigated in the N(CF3 SO2 )−2 ionic liquids comprising tributylmethylammonium and BMP+ [130,131]. Mn(II) can be introduced by anodic oxidation of Mn metal. Amorphous Mn deposits have been obtained and characterized by scanning electron microscope (SEM) and energy dispersive spectroscopy (EDX). The electrochemical behavior of Fe(II) has been investigated in a BMPN(CF3SO2)2 ionic liquid at 25 °C [132,133]. Fe[N(CF3SO2)2]2 can be prepared by reacting Fe powder with HN(CF3SO2)2 aqueous solution under nitrogen atmosphere. The redox reaction of Fe(III)/Fe(II) is observed at 1.5 V versus Fc/Fc+. The cathodic reduction of Fe(II) and the anodic oxidation of Fe are observed in the ionic liquid. However, the overpotential for the cathodic reduction of Fe(II) to Fe is as large as 1 V. The electrodeposit obtained at 25 °C is reported to be X-ray amorphous. The electrodeposition of cobalt, Co, has been reported in a BMPN(CF3SO2)2 ionic liquid [134,135]. Co[N(CF3SO2)2]2 can be prepared by the following reaction of CoCO3 and HN(CF3SO2)2 aqueous solution: CoCO3 + 2 HN(CF3SO2 )2 → Co[N(CF3SO2 )2 ]2 + H 2O + CO2

(9.31)

Co[N(CF3SO2)2]2 is soluble in the ionic liquid. Co(II) is considered octahedrally coordinated by N(CF3 SO2 )−2 from spectroscopy [136]. The deposition of Co has been confirmed by the cathodic reduction of Co(II) at 200 °C. However, the electrodeposits obtained at lower temperatures are X-ray amorphous. The change of the overpotential for the reduction of Co(II) is reported to be influenced by the addition of a small amount of acetone. The electrodeposition of nickel, Ni, has been reported in a BMPN(CF3SO2)2 ionic liquid containing Ni[N(CF3SO2)2]2 [137]. Ni[N(CF3SO2)2]2 can be prepared by reacting NiCO3·Ni(OH)2·4H2O with HN(CF3SO2)2 aqueous solution. Ni[N(CF3SO2)2]2 is soluble in the ionic liquid to give a octahedrally coordinated nickel complex [136]. The electrodeposition of Ni metal is possible through a cathodic reduction of Ni(II). The electrodeposit obtained above 70 °C has been identified as Ni metal by XRD. The electrodeposition of copper, Cu, has been reported in a TMHAN(CF3SO2)2 ionic liquid [138,139]. A divalent copper species, Cu(II), is introduced by dissolving Cu[N(CF3SO2)2]2, which is prepared by reacting CuO with HN(CF3SO2)2. The reduction of Cu(II) to a monovalent copper species, Cu(I), occurs at around 0.6 V versus I − / I −3 . The formation of Cu(I) also is observed with the proportionation reaction of Cu(II) and Cu metal. Metallic Cu is obtained through the reduction of Cu(I) at around −0.2 V versus I − / I −3 . The electrodeposition of Cu also has been reported in a BMPN(CF3SO2)2 ionic liquid [123,140]. Cu(I) is introduced by the anodic dissolution of Cu metal. The elec-

c09.indd 148

1/20/2011 10:27:55 AM

NONCHLOROALUMINATE IONIC LIQUIDS

149

trodeposits obtained by the cathodic reduction of Cu(I) have been examined by SEM and EDX. The codeposition of Cu and Mn also has been attempted in the ionic liquid. The formation of Cu–Sn alloy has been reported through contact deposition on Cu in a TMHAN(CF3SO2)2 ionic liquid containing Sn[N(CF3SO2)2]2 [141,142]. The electrodeposition of palladium, Pd, has been studied in a BMPN(CF3SO2)2 ionic liquid [143,144]. PdX2 can be dissolved in the ionic liquid with an addition of BMPX by forming PdX 2− 4 (X = Cl or Br). The electrodeposits obtained at 60 °C give XRD peaks corresponding to Pd metal. The electrodeposition of silver,Ag, has been investigated in a BMPN(CF3SO2)2 ionic liquid containing AgN(CF3SO2)2, which can be prepared by reacting Ag2O with HN(CF3SO2)2 aqueous solution [145,146]. The deposition and dissolution of Ag are confirmed to be reversible by EQCM. The redox reaction of Ag(I)/Ag often has been used for a reference electrode in N(CF3SO2 )−2 ionic liquids. The potential of Ag(I)/Ag is reported to be about 0.44 V versus Fc/Fc+ [133–135,147]. The electrochemical reduction of lanthanum (La), samarium (Sm), and europium (Eu) has been attempted in a butyltrimethylammonium N(CF3SO2)2 ionic liquid [148]. Ln[N(CF3SO2)2]3·(H2O)3 (Ln = La, Sm, and Eu) is soluble in the ionic liquid. The cathodic reduction from Ln(III) to Ln(II) is observed for Sm and Eu. An electrodeposit has been obtained by the potentiostatic cathodic reduction of Eu(III). However, the deposit has not been confirmed as Eu metal. The redox reaction of Ln(III)/Ln(II) has been studied in a BMPN(CF3SO2)2 ionic liquid [149]. However, the deposition of Ln(0) has not been confirmed within the electrochemical potential window of the ionic liquid. The electrodeposition of tantalum, Ta, has been reported in some N(CF3SO2 )−2 ionic liquids containing TaF5 [150–152]. Ta(V) can be reduced above 100 °C. Tantalum metal has been deposited at 200 °C. The change of the morphology as well as the mass of a gold electrode has been examined by STM and EQCM. The electrochemical reduction of Th(IV) has been investigated in a butyltrimethylammonium N(CF3SO2)2 ionic liquid [153] Th[N(CF3SO2)2]4·[HN(CF 3SO2)2]·(H2O)2 is dissolved in the ionic liquid as a Th source. The reduction of Th(IV) has been observed by cyclic voltammetry. However, the deposition of Th metal has not been confirmed. 9.2.4

Dicyanoamide (N(CN)2- ) Ionic Liquids

Many metal chlorides can be dissolved in dicyanoamide, N(CN)−2 , ionic liquids because the complexing ability of N(CN)−2 is stronger than chloride. Thus, the use of N(CN)−2 ionic liquids is more advantageous for the electrodeposition of metals than the use of such ionic liquids comprising BF4− , PF6− , and N(CF3SO2 )2− . There have been several reports on the electrodeposition of metals in the N(CN)−2 ionic liquid.

c09.indd 149

1/20/2011 10:27:55 AM

150

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

The electrodeposition of manganese, Mn, has been reported in a BMPN(CN)2 ionic liquid [154]. Mn(II) can be introduced either by dissolving MnCl2 or by anodic oxidation of Mn metal. The electrodeposit obtained by the cathodic reduction of Mn(II) is reported to be X-ray amorphous. The electrodeposition of nickel, Ni, has been studied in a EMIN(CN)2 ionic liquid [155]. NiCl2 is soluble in the ionic liquid to form Ni(CN)2− 4 . The reduction of Ni(II) is possible at 28 °C to give Ni metal, which has been identified by XRD. The electrodeposition of Cu has been examined in a EMIN(CN)2 ionic liquid [156]. CuCl is soluble in the ionic liquid. The redox reactions of Cu(II)/ Cu(I) and Cu(I)/Cu(0) are observed at 40 °C. The deposition of Cu metal has been confirmed by XRD. Codeposition of Cu and Ni also has been examined in the ionic liquid containing both CuCl and NiCl2 [157]. The electrodeposition of thin ruthenium, Ru, film has been reported in a BMIN(CN)2 ionic liquid [158]. RuCl3 is soluble in the ionic liquid. The deposition of Ru film with a thickness of about 5 nm has been confirmed by STM. The deposit is considered to be Ru metal by XPS analysis.

9.2.5

Other Ionic Liquids

Some ionic liquids comprising organic halides and metal halides have been studied for the electrodeposition of the metals and alloys. Most of these ionic liquids are hygroscopic and unstable against water, the same as the chloroaluminate ionic liquids. Moreover, some of these ionic liquids are viscous to the extent that they need elevating temperature and/or an addition of cosolvents. The electrodeposition of Ga [159], Ga–As [160], In–Sb [161], Sn [162], and Nb–Sn [163–165] has been reported in these ionic liquids.

REFERENCES 1. F. Endres, A. P. Abbott, D. R. MacRarlane, eds., Electrodeposition from Ionic Liquids, Wiley-VCH Verlag (2008). 2. G. Mamantov, A. I. Popov, eds., Chemistry of Nonaqueous Solvents Recent Advances, VCH, (1994). 3. R. T. Carlin, W. Crawford, M. Bersch, J. Electrochem. Soc., 139, 2720 (1992). 4. Q. Liao, W. R. Pinter, G. Stewart, C. L. Hussey, G. R. Stafford, J. Electrchem. Soc., 144, 936 (1997). 5. Y. Zhao, T. J. VanderNoot, Electrochim. Acta, 42, 1639 (1997). 6. C. A. Zell, F. Endres, W. Freyland, Phys. Chem. Chem. Phys., 1, 697 (1999). 7. J.-J. Lee, I.-T. Bae, D. A. Scherson, B. Miller, K. A. Wheeler, J. Electrochem. Soc., 147, 562 (2000). 8. J.-J. Lee, B. Miller, X. Shi, R. Kalish, K. A. Wheeler, J. Electrochem. Soc., 147, 3370 (2000).

c09.indd 150

1/20/2011 10:27:55 AM

REFERENCES

151

9. X. A. P. Abbot, C. A. Eardley, N. R. S. Farley, G. A. Griffith, A. Pratt, J. Appl. Electrochem., 31, 1345 (2001). 10. P.-Y. Chen, Y.-F. Lin, I.-W. Sun, J. Electrochem. Soc., 146, 3290 (1999). 11. X.-H. Xu, C. L. Hussey, J. Electrochem. Soc., 140, 618 (1993). 12. C. L. Hussey, X. Xu, J. Electrochem. Soc., 138, 1886 (1991). 13. D. A. Habboush, R. A. Osteryoung, Inorg. Chem., 23, 1726 (1984). 14. L. Heerman, W. D’Olieslager, J. Electrochem. Soc., 138, 1372 (1991). 15. T. M. Laher, C. L. Hussey, Inorg. Chem., 21, 4079 (1982). 16. R. T. Carlin, P. C. Trulove, H. C. De Long, J. Electrochem. Soc., 143, 2747 (1996). 17. C. L. Hussey, T. M. Laher, Inorg. Chem., 20, 4201 (1981). 18. R. J. Gale, B. Gilbert, R. A. Osteryoung, Inorg. Chem., 18, 2723 (1979). 19. C. Nanjunsiah, R. A. Osteryoung, J. Electrochem. Soc., 130, 1312 (1983). 20. B. J. Tierney, W. R. Pitner, J. A. Mitchell, C. L. Hussey, G. R. Stafford, J. Electrochem. Soc., 145, 3110 (1998). 21. C. L. Hussey, L. A. King, R. A. Carpio, J. Electrochem. Soc., 126, 1029 (1979). 22. W. R. Pitner, C. L. Hussey, J. Electrochem. Soc., 144, 3095 (1997). 23. Q. Zhu, C. L. Hussey, G. R. Stafford, J. Electrochem. Soc., 148(2), C88 (2001). 24. X.-H. Xu, C. L. Hussey, J. Electrochem. Soc., 140, 1226 (1993). 25. J. S.-Y. Liu, I.-W. Sun, J. Electrochem. Soc., 144, 140 (1997). 26. E. G.-S. Jeng, I.-W. Sun, J. Electrochem. Soc., 145, 1196 (1998). 27. E. G.-S. Jeng, I.-W. Sun, J. Electrochem. Soc., 144, 2369 (1997). 28. I.-W. Sun, C. L. Hussey, J. Electroanal. Chem., 274, 325 (1989). 29. X.-H. Xu, C. L. Hussey, J. Electrochem. Soc., 139, 3103 (1992). 30. X. C. Scordilis-Kelley, J. Fuller, R. T. Carlin, J. Electrochem. Soc., 139, 694 (1992). 31. T. L. Riechel, J. S. Wilkes, J. Electrochem. Soc., 139, 977 (1992). 32. C. Scordilis-Kelley, R. T. Carlin, J. Electrochem. Soc., 140, 1606 (1993). 33. C. Scordilis-Kelley, R. T. Carlin, J. Electrochem. Soc., 141, 873 (1994). 34. G. E. Gray, P. A. Kohl, J. Winnick, J. Electrochem., Soc., 142, 3636 (1995). 35. D. M. Ryan, E. R. Schumacher, T. L. Riechel, J. Electrochem. Soc., 143, 908 (1996). 36. G. E. Grey, J. Winnick, P. A. Kohl, J. Electrochem. Soc., 143, 2262 (1996). 37. G. E. Grey, J. Winnick, P. A. Kohl, J. Electrochem. Soc., 143, 3820 (1996). 38. J. Fuller, R. A. Osteryoung, R. T. Carlin, J. Electrochem. Soc., 142, 3632 (1995). 39. J. Fuller, R. T. Carlin, R. A. Osteryoung, J. Electrochem. Soc., 143, L145 (1996). 40. J. Fuller, R. T. Carlin, R. A. Osteryoung, P. Koranaios, R. Mantz, J. Electrochem. Soc., 145, 24 (1998). 41. M. Morimitsu, N. Tanaka, M. Matsunaga, Chem. Lett., 29, 1028 (2000). 42. Y. S. Fung, W. B. Zhang, J. Appl. Electrochem., 27, 857 (1997). 43. M. Lipsztajn, R. A. Osteryoung, Inorg. Chem., 24, 3492 (1985). 44. T. Tsuda, C. L. Hussey, G. R. Stafford, J. E. Bonevich, J. Electrochem. Soc., 150(4), C234 (2003). 45. M. R. Ali, A. Nishikata, T. Tsuru, Electrochim. Acta, 42, 2347 (1997). 46. J. S.-Y. Liu, P.-Y. Chen, I.-W. Sun, J. Electrochem. Soc., 144, 2388 (1997).

c09.indd 151

1/20/2011 10:27:55 AM

152

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

47. C. Nanjundiah, K. Shimizu, R. A. Osteryoung, J. Electrochem. Soc., 129, 2474 (1982). 48. M. Lipsztajn, R. A. Osteryoung, Inorg. Chem., 24, 716 (1985). 49. S. Pye, J. Winnick, P. A. Kohl, J. Electrochem. Soc., 144, 1933 (1997). 50. R. T. Carlin, H. C. De Long, J. Fuller, P. C. Trulove, J. Electrochem. Soc., 145, 1598 (1998). 51. J. A. Mitchell, W. R. Pitner, C. L. Hussey, G. R. Stafford, J. Electrochem. Soc., 143, 3448 (1996). 52. M. R. Ali, A. Nishikata, T. Tsuru, Electrochim. Acta, 42, 1819 (1997). 53. C. A. Zell, W. Freyland, Langmuir, 19, 7445 (2003). 54. W. Freyland, C. A. Zell, S.-Z. El Abedin, F. Endres, Electrochim. Acta, 48, 3053 (2003). 55. W. R. Pitner, C. L. Hussey, G. R. Stafford, J. Electrochem. Soc., 143, 130 (1996). 56. M. R. Ali, A. Nishikata, T. Tsuru, J. Electroanal. Chem., 513, 111 (2001). 57. T. M. Laher, C. L. Hussey, Inorg. Chem., 22, 3247 (1983). 58. J.-J. Lee, B. Miller, X. Shi, R. Kalish, K. A. Wheeler, J. Electrochem. Soc., 148(3), C183 (2001). 59. Q. Zhu, C. L. Hussey, J. Electrochem. Soc., 148(5), C395 (2001). 60. F. Endres, A. Schweizer, Phys. Chem. Chem. Phys., 2, 5455 (2000). 61. C. J. Dymek Jr., G. F. Reynolds, J. S. Wilkes, J. Electrochem. Soc., 134, 1658 (1987). 62. Y.-F. Lin, I.-W. Sun, J. Electrochem. Soc., 146, 1054 (1999). 63. J. Dogel, W. Freyland, Phys. Chem. Chem. Phys., 5, 2484 (2003). 64. N. Koura, T. Kato, E. Yumoto, Hyomen Gijutsu, 45, 805 (1994). 65. H. C. De Long, J. S. Wilkes, R. T. Carlin, J. Electrochem. Soc., 141, 1000 (1994). 66. X.-H. Xu, C. L. Hussey, J. Electrochem. Soc., 139, 1295 (1992). 67. F. Endres, W. Freyland, Ber. Bunsenges. Phys. Chem., 101, 1075 (1997). 68. F. Endres, W. Freyland, J. Phys. Chem. B, 102, 10229 (1998). 69. T. M. Laher, C. L. Hussey, Inorg. Chem., 22, 1279 (1983). 70. M. A. M. Noel, R. A. Osteryoung, J. Electroanal. Chem., 284, 413 (1990). 70a. M. A. M. Noel, R. A. Osteryoung, J. Electroanal. Chem., 293, 139 (1990). 71. T. Tsuda, T. Nohira, Y. Ito, Electrochim. Acta, 46, 1891 (2001). 72. T. Tsuda, T. Nohira, Y. Ito, Electrochim. Acta, 47, 2817 (2002). 73. F. Endres, C. Schrodt, Phys. Chem. Chem. Phys., 2, 5517 (2000). 74. F. Endres, Phys. Chem. Chem. Phys., 3, 3165 (2001). 75. F. Endres, S. Z. El Abedin, Phys. Chem. Chem. Phys., 4, 1640 (2002). 76. M.-H. Yang, I.-W. Sun, J. Appl. Electrochem., 33, 1077 (2003). 77. M.-H. Yang, M.-C. Yang, I.-W. Sun, J. Electrochem. Soc., 150(8), C544 (2003). 78. P.-Y. Chen, I.-W. Sun, Electrochim. Acta, 45, 441 (1999). 79. Y. Katayama, S. Dan, T. Miura, T. Kishi, J. Electrochem. Soc., 148(2), C102 (2001). 80. C.-C. Tai, F.-Y. Su, I.-W. Sun, Electrochim. Acta., 50, 5504 (2005). 81. P.-Y. Chen, I.-W. Sun, Electrochim. Acta, 45, 3163 (2000). 82. F.-Y. Su, J.-F. Huang, I.-W. Sun, J. Electrochem. Soc., 151(12), C811 (2004).

c09.indd 152

1/20/2011 10:27:55 AM

REFERENCES

153

83. S.-I. Hsiu, C.-C. Tai, I.-W. Sun, Electrochim. Acta., 51, 2607 (2006). 84. L.-S. Jou, J.-K. Chang, T.-J. Twhang, I.-W. Sun, J. Electrochem. Soc., 156(6), D193 (2009). 85. Au T. Oyama, T. Okajima, T. Ohsaka, J. Electrochem. Soc., 154(6), D322 (2007). 86. T. Oyama, T. Okajima, T. Ohsaka, S. Yamaguchi, N. Oyama, Bull. Chem. Soc. Jpn., 8, 726 (2008). 87. S.-I. Hsiu, J.-F. Huang, I.-W. Sun, C.-H. Yuan, J. Shiea, Electrochim. Acta, 47, 4367 (2002). 88. Y.-F. Lin, I.-W. Sun, Electrochim. Acta, 44, 2771 (1999). 89. O. Mann, W. Freyland, Electrochim. Acta, 53, 518 (2007). 90. M.-C. Lin, P.-Y. Chen, I.-W. Sun, J. Electrochem. Soc., 148(10), C653 (2001). 91. J.-F. Huang, I.-W. Sun, J. Electrochem. Soc., 151(1), C8 (2004). 92. P.-Y. Chen, I.-W. Sun, Electrochim. Acta, 46, 1169 (2001). 93. H.-Y. Hsu, C.-C. Yang, Z. Naturforsch., 58b, 139 (2003). 94. N. Koura, T. Endo, Y. Idemoto, J. Non-Cryst. Solids, 205–207, 650 (1996). 95. N. Koura, S. Matsumoto, Y. Idemoto, Hyomen Gijutsu, 49, 1215 (1998). 96. N. Koura, Y. Suzuki, Y. Idemoto, T. Kato, F. Matsumoto, Surf. Coat. Technol., 169, 120 (2003). 97. S.-P. Gou, I.-W. Sun, Electrochim. Acta, 53, 2538 (2008). 98. P.-Y. Chen, M.-C. Lin, I.-W. Sun, J. Electrochem. Soc., 147, 3350 (2000). 99. J.-F. Huang, I.-W. Sun, J. Electrochem. Soc., 149(9), E348 (2002). 100. J.-F. Huang, I.-W. Sun, Electrochim. Acta, 49, 3251 (2004). 101. V. R. Koch, C. Nanjundiah, G. Battista Appetecchi, B. Scrosati, J. Electrochem. Soc., 142, L116 (1995). 102. P. Bonhôte, A.-P. Dias, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). 103. D. R. MacFarlane, J. Sun, J. Golding, P. Meakin, M. Forsyth, Electrochim. Acta, 45, 1271 (2000). 104. H. Matsumoto, M. Yanagida, K. Tanimoto, M. Nomura, Y. Kitagawa, Y. Miyazaki, Chem. Lett., 2000, 922. 105. J. Sun, D. R. MacFalane, M. Forsyth, Electrochim. Acta, 48, 1707 (2003). 106. H. Matsumoto, M. Yanagida, K. Tanimoto, T. Kojima, Y. Tamiya, Y. Miyazaki, in Molten Salts XII, vol. 99–41, P. C. Trulove, H. C. De Long, G. R. Stafford, and S. Deki, eds., Electrochemical Society, 1999, p. 186. 107. H. Sakaebe, H. Matsumoto, Electrochem. Commun., 5, 594 (2003). 108. Y. Katayama, T. Morita, M. Yamagata, T. Miura, Electrochemistry, 71, 186 (2003). 109. P. C. Howlett, D. R. MacFarlane, A. F. Hollenkamp, Electrochem. SolidState Lett., 7, A97 (2004). 110. R. Wibowo, S. E. Ward Jones, R. G. Compton, J. Phys. Chem. B, 113, 12293 (2009). 111. L. H. S. Gasparotto, N. Borisenko, N. Bocchi, S. Zein El Abedin, F. Endres, Phys. Chem. Chem. Phys., 11, 11140 (2009). 112. J. A. Vega, J. Zhou, P. A. Kohl, J. Electrochem. Soc., 156(4), A253 (2009). 113. A. Lewandowski, A. Swiderska-Mocek, J. Appl. Electrochem., 40, 515 (2010).

c09.indd 153

1/20/2011 10:27:55 AM

154

ELECTRODEPOSITION OF METALS IN IONIC LIQUIDS

114. G. T. Cheek, W. E. O’Grady, S. Zein El Abedin, E. M. Moustafa, F. Endres, J. Electrochem. Soc., 155(1), D91 (2008). 115. S. Zein El Abedin, E. M. Moustafa, R. Hempelmann, H. Natter, F. Endres, Phys. Chem. Chem. Phys.,7, 1535 (2006). 116. E. M. Moustafa, S. Zein El Abedin, A. Shkurankov, E. Zschippang, A. Y. Saad, A. Bund, F. Endres, J. Phys. Chem. B, 111, 4693 (2007). 117. P. Eiden, Q.-X. Liu, S. Zein El Abedin, F. Endres, I. Krossing, Chem. Eur. J., 15, 3426 (2009). 118. N. M. Rocher, E. I. Izgorodina, T. Rüther, M. Forsyth, D. R. MacFarlane, T. Rodopoulos, M. D. Horne, A. M. Bond, Chem. Eur. J., 15, 3435 (2009). 119. S. Zein El Abedin, N. Borissenko, F. Endres, Electrochem. Commun., 6, 510 (2004). 120. N. Borisenko, S. Zein El Abedin, F. Endres, J. Phys. Chem. B, 110, 6250 (2006). 121. L. H. S. Gasparotto, N. Borisenko, O. Höfft, R. Al-Salman, W. Maus-Friedrichs, N. Bocchi, S. Zein El Abedin, F. Endres, Electrtochem. Acta, 55, 218 (2009). 122. R. Al-Salman, S. Zein El Abedin, F. Endres, Phys. Chem. Chem. Phys., 10, 4650 (2008). 123. S. Zein El Abedin, A. Y. Saad, H. K. Farag, N. Borisenko, Q.-X. Liu, F. Endres, Electrochem. Acta, 52, 2746 (2007). 124. N. Tachikawa, N. Serizawa, Y. Katayama, T. Miura, Electrochim. Acta, 53, 6530 (2008). 125. N. Serizawa, N. Tachikawa, Y. Katayama, T. Miura, Electrochemistry, 77, 630 (2009). 126. B. C. M. Martindale, S. E. Ward Jones, R. G. Compton, Phys. Chem. Chem. Phys., 12, 1827 (2010). 127. P.-Y. Chen, C. L. Hussey, Electrochim. Acta, 49, 5125 (2004). 128. I. Mukhopadhyay, C. L. Aravinda, D. Borissov, W. Freyland, Electrochim. Acta, 50, 1275 (2005). 129. Y. Katayama, K. Ogawa, T. Miura, Electrochemistry, 73, 576 (2005). 130. P.-Y. Chen, C. L. Hussey, Electrochim. Acta, 52, 1857 (2007). 131. J.-K. Chang, W.-T. Tsai, P.-Y. Chen, C.-H. Huang, F.-H. Yeh, I.-W. Sun, Electrochem. Solid-State Lett., 10(1), A9 (2007). 132. M. Yamagata, N. Tachikawa, Y. Katayama, T. Miura, Electochemistry, 73, 564 (2005). 133. M. Yamagata, N. Tachikawa, Y. Katayama, T. Miura, Electrochim. Acta, 52, 3317 (2007). 134. R. Fukui, Y. Katayama, T. Miura, Electrochemistry, 73, 567 (2005). 135. Y. Katayama, R. Fukui, T. Miura, J. Electrochem. Soc., 154(10), D534 (2007). 136. K. Fujii, T. Nonaka, Y. Akimoto, Y. Umebayashi, S. Ishiguro, Anal. Sci., 24, 1377 (2008). 137. Y.-L. Zhu, Y. Kozuma, Y. Katayama, T. Miura, Electrochim. Acta, 54, 7502 (2009). 138. K. Murase, K. Nitta, T. Hirato, Y. Awakura, J. Appl. Electrochem., 31, 1089 (2001). 139. T. Katase, K. Murase, T. Hirato, Y. Awakura, J. Appl. Electrochem., 37, 339 (2007). 140. P.-Y. Chen, M.-J. Deng, D.-X. Zhuang, Electrochim. Acta, 54, 6935 (2009). 141. T. Katase, R. Kurosaki, K. Murase, T. Hirato, Y. Awakura, Electrochem. Solid-State Lett., 9(4), C69 (2006).

c09.indd 154

1/20/2011 10:27:55 AM

REFERENCES

155

142. K. Murase, R. Kurosaki, T. Katase, H. Sugimura, T. Hirato, Y. Awakura, J. Electrochem. Soc., 154(11), D612 (2007). 143. Y. Bando, Y. Katayama, T. Miura, Electrochim. Acta, 53, 87 (2007). 144. Y. Katayama, Y. Bando, T. Miura, Trans. Inst. Met. Finish., 86, 205 (2008). 145. E. I. Rogers, D. S. Silvester, S. E. Ward Jones, L. Aldous, C. Hardacre, A. J. Russell, S. G. Davies, R. G. Compton, J. Phys. Chem. C, 111, 13957 (2007). 146. N. Serizawa, Y. Katayama, T. Miura, J. Electrochem. Soc., 156(11), D503 (2009). 147. G. A. Snook, A. S. Best, A. G. Pandolfo, A. F. Hollenkamp, Electrochem. Commun., 8, 1405 (2006). 148. A. I. Bhatt, I. May, V. A. Volkovich, D. Collison, M. Helliwell, I. B. Polovov, R. G. Lewin, Inorg. Chem., 44, 4934 (2005). 149. M. Yamagata, Y. Katayama, T. Miura, J. Electrochem. Soc., 153(1), E5 (2006). 150. N. Borisenko, A. Ispas, E. Zschippang, Q. Liu, S. Zein El Abedin, A. Bund, F. Endres, Electrochim. Acta, 54, 1519 (2009). 151. S. Zein El Abedin, H. K. Farag, E. M. Moustafa, U. Welz-Biermann, F. Endres, Phys. Chem. Chem. Phys., 7, 2333 (2005). 152. A. Ispas, B. Adolphi, A. Bund, F. Endres, Phys. Chem. Chem. Phys., 12, 1793 (2010). 153. A. I. Bhatt, N. W. Duffy, D. Collison, I. May, R. G. Lewin, Inorg. Chem., 45, 1677 (2006). 154. D.-X. Zhuang, M.-J. Deng, P.-Y. Chen, I.-W. Sun, J. Electrochem. Soc., 155(9), D575 (2008). 155. M.-J. Deng, I.-W. Sun, P.-Y. Chen, J.-K. Chang, W.-T. Tsai, Electrochim. Acta, 53, 5812 (2008). 156. T.-I. Leong, I.-W. Sun, M.-J. Deng, C.-M. Wu, P.-Y. Chen, J. Electrochem. Soc., 155(4), F55 (2008). 157. M.-J. Deng, P.-C. Lin, I.-W. Sun, P.-Y. Chen, J.-K. Chang, Electrochemistry, 77, 582 (2009). 158. O. Mann, W. Freyland, O. Raz, Y. Ein-Eli, Chem. Phys. Lett., 460, 178 (2008). 159. S. P. Wicelinski, R. J. Gale, J. S. Wilkes, J. Electrochem. Soc., 134, 262 (1987). 160. M. K. Carpenter, M. W. Verbrugge, J. Electrochem. Soc., 137, 123 (1990). 161. M. K. Carpenter, M. W. Verbrugge, J. Mater. Res., 9, 2584 (1994). 162. G. Lin, N. Koura, Electrochemistry, 65, 149 (1997). 163. N. Koura, G. Ling, H. Ito, Hyomen Gijutsu, 46, 1162 (1995). 164. G. Ling, N. Koura, Hyomen Gijutsu, 48, 454 (1997). 165. N. Koura, T. Umebayashi, Y. Idemoto, G. Ling, Electrochemistry, 67, 684 (1999). 166. M. A. M. Noel, R. A. Osteryoung, J. Electroanal. Chem., 293, 139 (1990).

c09.indd 155

1/20/2011 10:27:55 AM

PART II BIOELECTROCHEMISTRY

c10.indd 157

1/20/2011 10:27:57 AM

10 ENZYMATIC REACTIONS Noritaka Iwai and Tomoya Kitazume

Ionic liquids are having an important impact on organic reactions, and various kinds of reactions in ionic liquids have been reported until now [1]. The reaction using biocatalysts in ionic liquids also has been developed because biocatalysts are renewable natural catalysts with high enantio-, regio-, and chemo-selectivities. Furthermore, there is a growing interest in the development of the greener chemical processes that are required both now and in the future.

10.1 GEOTRICHUM CANDIDUM-CATALYZED SYNTHESIS OF OPTICALLY ACTIVE ALCOHOLS IN AN IONIC LIQUID In general, syntheses of chiral materials as valuable building blocks in organic syntheses have been developed by various kinds of chemical and/or biological methods [2]. In particular, in the synthesis of optically active secondary alcohols, the following methods have been known for the synthesis of optically active alcohols: optical resolution of alcohols [3], asymmetric reduction of ketones [3], and deracemization of racemic alcohols by chemical and/or biological tools in organic solvents and/or water medium [4]. Among these methods, deracemization is the most promising because it permits 100% conversion of the racemate to the corresponding enantiomerically pure material. Nakamura et al. have reported enzymatic deracemization of racemic 1-arylethanols, 1-pyridylethanols, and β-hydroxyesters in water medium to the corresponding enantiomerically pure alcohol [5]. As an other example,

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

159

c10.indd 159

1/20/2011 10:27:57 AM

160

ENZYMATIC REACTIONS

two biocatalysts were used for the oxidation of (S)-mandelic acid to give benzoylformate by Alcaligenes bronchisepticus, and then the produced benzoylformate was reduced to give (R)-mandelic acid (85% yield; >99% ee) by Streptococcus faecalis. Furthermore, deracemization of 3-pentyn-2-ol with Nocardia fusca to produce (R)-isomer also has been reported. Recently, tandem biocatalytic oxidation and reduction using two biocatalysts have been reported to give the (S)-enantiomer of secondary alcohols. Based on those results, obviously, it remains the general method for the biocatalytic synthesis of optically active secondary alcohols, which slows the use of various types of aryl-, alkyl- and heterocyclic derivatives.

10.2

OPTICAL RESOLUTION BY BIOCATALYTIC OXIDATION

In our recent report for the utility of ionic liquid in biocatalytic reduction [6], we revealed that water was not miscible with [bmim][PF6], whereas [emim] [BF4] is completely miscible with water and that the stabilizing effect of the enzyme by keeping water around it and using the polymer improves the yield of (S)-secondary alcohols in an ionic liquid [emim][BF4]. For the deracemization of secondary alcohols in an ionic liquid, we used the immobilized cell produced from Geotrichum candidum and a water-absorbing polymer (BL100) at first. In the immobilized cell system holding water and an enzyme, the enantioselective oxidation accelerated in an ionic liquid shown in Table 10.1, yielding the optically active (R)-secondary alcohols that were the same stereoisomer obtaind from the water medium [5]. This method is a reasonably general route to obtain the optically active secondary alcohols in an ionic liquid.

10.3 DERACEMIZATION OF RACEMIC SECONDARY ALCOHOLS WITH CHEMOENZYMATIC METHOD From the results shown in Table 10.1, we have found that it is necessary to reduce the produced ketone to the corresponding racemic alcohol to complete the deracemization in an ionic liquid containing water for the stabilizing effect of the enzyme. To realize this idea, the deracemization of 1-phenylethanol was

OH R

immobilized Geotrichum candidum [emim][BF4]

O R

OH +

R (R)

Scheme 10.1. Asymmetric oxidation of racemic alcohols by immobilized Geotrichum candidum.

c10.indd 160

1/20/2011 10:27:57 AM

161

c10.indd 161

1/20/2011 10:27:57 AM

43

51

52

52

Yield of Ketone (%)

39

42

48

49

Yield of Alcohol (%)

97

95

94

97

% ee of (R)-Alcohola

S

O

OH

OH

OH

OH

Substrate

39

42

30

41

Yield of Ketone (%)

46

55

67

48

Yield of Alcohol (%)

>99

81

41

97

% ee of (R)-Alcohola

Reaction conditions: 24 hours at 30 °C. a Determined by gas chromatograph (GC) analysis with chiral column (Chirasil-DEX CB or CP-Cyclodextrin-B-2,3,6-M-19. The absolute configuration was determined to be R for all samples examined by comparing the GC retention times.

C9H19

OH

OH

OH

OH

Substrate

TABLE 10.1. Substrate Specificity of the Oxidation System in [emim][BF4] by Immobilized G. Candidum Cell on Water-Absorbing Polymer

162

ENZYMATIC REACTIONS

examined first as a model reaction by the system Geotrichum candidumreductive reagent (NaBH4 and/or NaBH3CN) in an ionic liquid to test the improvement for the yield of deracemization and/or the optical purity of secondary alcohol in the various ratios of reductive reagent, wet cell, and 2 (N-morpholino) ethanesulfonic acid (MES) buffer solution to create the most convenient reaction condition. From the results of entries 6 and 7 shown in Table 10.2, deracemization proceeded drastically to give (R)-2-phenylethanol with a high optical purity (99% ee). However, NaBH3CN as a reductive reagent (entry 3) did not improve the reduction of ketone. Obviously, in other alcohols (entries 9 and 10), deracemization is not complete in the absence of an ionic liquid. Furthermore, an important factor for deracemization with a high optical purity is the ratio of MES buffer soution and ionic liquid shown in Table 10.2. To aim for the construction of the wide substrate specificity with the excellent enantioselectivity for a valuable synthetic practical route using an ionic liquid as media, we examined the deracemization with the complex system derived from Geotrichum candidum NBRC 5767 and NaBH4 in a mixture of MES buffer and ionic liquid [emim][BF4]. Various types of secondary alcohols such as 1-aryl ethanol, 4-phenyl-2-butanol, (E)-3-octen-2-ol, 2-octanol, 1-(2furyl)ethanol, 1-(thiophen-2-yl)ethanol, and/or 2-naphtylethanol were used as substrates, and it was found that all were deracemizated with moderate to excellent yield and that the enantioselectivity was excellent for all substrates tested from the results shown in Table 10.3. In 1-phenylpropanol and 3-octanol, becauese we found that the first step (oxidation process) does not proceed smoothly based on the results shown in Table 10.3 (entries 14, 15, and 17), the reaction condition such as an excess of Geotrichum candidum and/or extention time was used to obtain the corresponding chiral alcohol. In our previous study [6], we reported that the reduction of ketones with Geotrichum candidum produced (S)-secondary alcohols. In this system, NaBH4 in the mixture of [emim][BF4]-MES buffer reduced the ketones to the corresponding alcohol more quickly than the ketone reduction with Geotrichum candidum. O

OH R

R

NBRC 5767

R (R)

Geotrichum candidum NAD +

OH +

NADH

resting cell OH R

O

ionic liquid

X

R

NaBH4 in MES buffer solution-[emim][BF4]

Scheme 10.2. Deracemization process in MES buffer solution-[emim][BF4].

c10.indd 162

1/20/2011 10:27:57 AM

163

c10.indd 163

1/20/2011 10:27:57 AM

2-octanol 6-methyl-5hepten-2-ol

1-phenylethanol

Substrate

NaBH4 or NaBH3CN NaBH4 (3.2 eq.) NaBH4 (3.0 eq.) NaBH3CN (3.7 eq.) NaBH4 (3.0 eq.) NaBH4 (3.1 eq.) NaBH4 (6.4 eq.) NaBH4 (9.0 eq.) none NaBH4 (9.0 eq.) NaBH4 (9.0 eq.)

Ionic Liquid (3 ml)

[emim][BF4] [emim][BF4]b [emim][BF4] [emim][BF4] [emim][BF4] [emim][BF4] [emim][BF4] [emim][BF4] none none

24 24 24 24 48 48 24 24 24 24

Reaction Time (hour) 0.25 0.25 0.25 0.50 0.50 0.50 0.50 0.25 0.50 0.50

Wet Cell (g) (1.5, H2O) 1.5 1.5 3.0 3.0 3.0 3.0 (1.5, H2O) 6.0 6.0

MES Buffer ml 4 10 31 20 31 0 1 38 5 22

Ketone

89 81 60 65 60 95 96 57 36 10

Alcohol

Yield (%)

19 12 51 98 92 99 99 64 74 86

% ee of Alcohol

Reaction conditions: 24 hours at 30 °C. Determined by GC analysis with chiral column (Chirasil-DEX CB or CP-Cyclodextrin-B-2,3,6-M-19. The absolute configuration was determined to be R for all samples examined by comparing the GC retention times with those of the authentic samples in the references. b Immobilized system prepared from water-absorbing polymer was used.

a

1 2 3 4 5 6 7 8 9 10

Entry

TABLE 10.2. Effect of Reaction Conditions on the Deracemizationa

TABLE 10.3. Deracemization of Secondary Alcohols Entry

Substrate

NaBH4

Wet cell (g)

Yield (%) Ketone

Alcohol

% ee of Alcohola

OH 1

NaBH4 (9.0 eq.)

0.50

1

96

99

NaBH4 (9.0 eq.)

0.50

5

87

95

NaBH4 (9.0 eq.)

0.50

15

79

96

NaBH4 (9.0 eq.) NaBH4 (9.0 eq.)b

0.50 0.75

0 0

54 58

>99 >99

NaBH4 (9.1 eq.)

0.50

8

68

97

NaBH4 (9.0 eq.) NaBH4 (9.0 eq.)b NaBH4 (9.0 eq.)c

0.50 0.75 1.00

10 10 14

67 74 85

89 99 98

NaBH4 (9.0 eq.)b

0.75

21

56

>99

NaBH4 (9.0 eq.) NaBH4 (9.1 eq.)

0.50 0.50

18 16

60 66

97 >99

NaBH4 (8.9 eq.)

0.50

12

73

>99

OH 2

OH 3 4 5 6 7 8 9

OH Bu OH OH C6H13 OH

10 11 12

O

13

S

OH

OH

14 15 16

OH

NaBH4 (9.0 eq.) NaBH4 (9.0 eq.) NaBH4 (9.0 eq.)

0.50 6.00d 6.00e

4 17 32

96 33 49

19 66 >99

17 18

OH

NaBH4 (9.0 eq.) NaBH4 (9.0 eq.)

0.50 3.00f

3 9

90 41

16 97

C5H11

Reaction conditions: 24 hours at 30 °C in the mixture of MES buffer (3 mL) and [emim][BF4] (3 mL). a Determined by GC analysis with chiral column (Chirasil-DEX CB or CP-Cyclodextrin-B2,3,6-M-19. The absolute configuration was determined to be R for all samples examined by comparing the GC retention times with those of the authentic samples in the references. b MES buffer solution (4.5 mL), [emim][BF4] (4.5 mL). c MES buffer solution (6.0 mL), [emim][BF4] (6.0 mL). d MES buffer solution (36.0 mL), [emim][BF4] (36.0 mL). e MES buffer solution (18.0 mL), [emim][BF4] (18.0 mL), 51 hours. f MES buffer solution (18.0 mL), [emim][BF4] (18.0 mL), 24 hours.

c10.indd 164

1/20/2011 10:27:57 AM

IMPROVEMENT OF FLUORINATION REACTION

165

10.4 IMPROVEMENT OF FLUORINATION REACTION WITH IMMOBILIZED FLUORINASE In the fluorine chemistry, the fluorination process with fluorinase (5′-fluoro-5′deoxyadenosine synthase), which was found from Streptomyces cattleya, has had an important impact on the biological synthesis to introduce the fluorine atom on the carbon atom [7]. However, it is not easy to handle in applications because of the low reactivity and stability of fluorinase [8]. In view of the development of more advanced processing systems, the challenge to develop practical processes, reaction media, and/or conditions is one of the most important issues in biological methodology for the synthesis of fluorinated materials. In the fluorination reaction with fluorinase (200 mL, 1 mg/mL)-Tris-HCl buffer solution (pH 8.0) in the presence of KF (1 mM in H2O, 2 mmol), the conversion yield from S-adenosyl-L-methionine (SAM; 50 nmol) to 5′-fluoro5′-deoxyadenosine (5′-FDA) is 5.7%. To improve the conversion yield in this reaction, we have examined the fluorination reaction with fluorinase in an ionic liquid instead of in a Tris-HCl buffer solution based on our reports for the utility of ionic liquid in organic reactions [9]. However, the fluorination reaction in ionic liquids did not proceed in the absence of water. In our recent papers [6], we reported the stabilizing effect of the enzyme by keeping water around it using a water-absorbing polymer as well as the utility of ionic liquid in the case of an immobilized biocatalyst. Therefore, in attempting to increase the conversion yield from SAM to 5′-FDA, we used the immobilized fluorinase (immobilized fluorinase was prepared from fluorinase in Tris-HCl buffer solution [200 mL, 1 mg/mL], KF 2 mmol, SAM 50 nmol, and water-absorbing polymer 20 mg) in ionic liquid (800 mL) or in Tris-HCl buffer solution (800 mL). After the fluorination reaction was carried at 37 °C, 5′-FDA was extracted with diethyl ether (10 × 10 mL). The product peak and yields were determined by high-performance liquid chromatography (HPLC) analyses according to a previous report [7]. As it is well known that an ionic liquid can be reused as the reaction medium in the same reaction, we examined its reuse in the previous above reaction system. After removing 5′-FDA from the reaction mixture with diethyl ether (10 × 10 mL), the remaining diethyl ether in the reaction mixture was removed under dynamic vacuum, and the residue (from second to third cycles) was incubated for 1 hour at 37 °C as shown in Table 10.4. From the results shown in Table 10.4, we found that some ionic liquids are convenient to improve the total conversion yields to increase more than those of nonimmobilized fluorinase as well as that the ratio (fluorinase, SAM, KF, ionic liquid, and Tris-HCl buffer solution) is an important factor as shown in Table 10.3 (entries 4, 5, and 14). Furthermore, in ionic liquids (entries 11–13), it is impossible to detect 5′-FDA in the mixture solution of the ionic liquid and the product of HPLC. In the next step, we uesd twice the amount of fluorinase (400 mg) and KF (4 mmol) for the preparation of immobilized fluorinase. The fluorination

c10.indd 165

1/20/2011 10:27:57 AM

166

ENZYMATIC REACTIONS

TABLE 10.4. Utility of an Ionic Liquid with Fluorinase (200 μg) Entry

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

ILe or Buffer

[emim][BF4] [amim][BF4] [bmim][BF4] [DEME][BF4] [DEME][BF4]b [hexylmim][BF4] [bmim][PF6] [octylmim][PF6] [emim][OTf] PP13 TFSA [Bu3P(C12)][BF4] [bmim][TFSA] [DEMD][TFSA] Tris-HCL buffer

Conveersion Yield (%)a

Total

Ratioc

First

Second

Third

1.7 1.6 2.8 2.9 3.2 4.9 2.7 5.5 1.3 1.8 —d —d —d 3.4

1.0 1.2 2.1 1.8 1.1 3.3 0.3 —d 0.9 —d

0.7 0.6 0.6 1.2 2.0 0.9 0.1 —d 0.4

3.4 3.4 5.5 5.9 6.3 9.1 3.1 5.5 2.6 1.8

87 87 141 151 162 233 79 141 67 46

0.5

—d

3.9

100

Reation Condition: immobilized cell was prepared from fluorinase (1 mg/ml, 200 μl), 2 μmol KF, SAM 50 nmol, and water-absorbing polymer 20mg, 37 °C, 6 hours, Tris-HCL buffer or ionic liquid (800 μl). b Reaction condition: immobilized cell was prepared from fluorinase 40 μg (1 mg/ml, 400 μl), 4 μmol KF, SAM 50 nmol, and water-absorbing polymer 20 mg, 37 °C, 6 hours, ionic liquid (2 ml). c Ratio of total conversion yield of ionic liquid/Tris-HCL buffer. d 5′-FDA was not isolated. e [emim][BF4]: 1-ethy-3-methylimidazolium terafluoroborate; [amim][BF4]: 1-ally-3methylimidazolium terafluoroborate; [bmim] [BF4]: 1-buty-3-methylimidazolium terafluoroborate; [DEME][BF4]; N,N-dithyl-N-methy-N-(2-methoxyethyl) ammonium terafluoroborate; [hexylmim][BF4]: 1-hexyl-3-methylimidazolium terafluoroborate; [bmim][PF6] 1-butyl-3methylimidazolium hexafluorophosphate; [octylmim][PF6]: 1-octyl-3-methyl-imidazolium hexafluorophosphate; [emim][OTf]: 1-ethyl-3-methylimidazolium trifluoromethane-sulfonate; PP13 TFSA:N-methyl-N-propylpiperidinium bis(trifluoromethanesulfonyl)amide; [Bu3P(C12)][BF4]: tributyl(dodecyl)phosphoniim tetrafluoroborate; [bmim][TFSA]: 1-butyl-3-methylimidazolium bis(trifluoromethanesulfonyl)amide; [DEME][TFSA];N,N-deethy-N-methyl-N-(2-methoxyethyl) ammonium bis(trifluoromethansesulfonyl)amide. a

reaction using one of the immobilized fluorinases prepared from fluorinase in Tris-HCl buffer solution (400 mL, 1 mg/mL), KF 4 mmol, SAM 50 nmol, and water-absorbing polymer 20 mg, was carried in an ionic liquid (2 mL) and/or a Tris-HCl buffer solution (2 mL) at 37 °C for 1 hour as shown in Figure 10.1. In the previous immobilized fluorinase system, it is possible to recycle the reaction system to increase the conversion yield of the fluorination reaction up to 9.4%, as shown in Table 10.5. In the fluorination with the immobilized fluorinase in ionic liquids [octylmim][PF6] and [hexylmim][PF6], we found that the conversion yield was increased up to 2.4 and/or 1.6 times more than that of the Tris-HCl buffer solution (pH 8.0). Furthermore, we found that the ionic

c10.indd 166

1/20/2011 10:27:57 AM

167

IMPROVEMENT OF FLUORINATION REACTION

5'-FDA was extracted with diethyl ether +H

O3N

O

+S Me N N

O HO

NH2 N N

OH

Ffluorinase cell ionic liquid

S-adenosyl-L-methionine (SAM)

N N

F

NH2 N N

O HO

fluorinase in 50 mM Tris-HCl buffer (400 µl, 1 mg/mL) SAM 50 nmol KF 4 µmol

recovered reaction system removed diethyl ether

OH

water-absorbing 5Õ-fluoro-5Õ-deoxyadenosine polymer 20 mg (5'-FDA)

buffer solution (2 mL) or ionic liquid (2 mL)

second cycle

Figure 10.1. fluorination reaction with immobilized fluorinase in solvent.

TABLE 10.5. Utility of an Ionic Liquid in Fluorination with Fluorinase (400 μg) Entry

15 16 17 18 19 20

Conversion Yield (%)a

IL or Buffer

[DEME][BF4] [hexylmim][BF4] [octylmim][BF4]d [octylmim][PF6] [hexylmim][PF6]d Tris-HCl buffer

First

Second

Third

Fourth

Fifth

2.3 4.7 4.5 19.8 12.9 2.5

2.5 3.2 3.8 2.6 2.7 3.5

1.4 1.8 2.6 0.4 0.3 2.7

1.0 1.3 1.1 —c —c 0.5

1.3 1.5 0.7 —c —c 0.2

Total

Ratiob

8.5 12.5 12.7 22.8 15.9 9.4

90 133 135 243 163 100

a

Reaction condition: immobilized cell was prepared from fluorinase in Tris-HCl buffer solution (1 mg/ml, 400 μl), 4 μmol KF, SAM 50 nmol, and water-absorbing polymer 20 mg, 37 °C, 1 hour, ionic liquid (2 ml). b Ratio of total conversion yield of ionic liquid/Tris-HCl buffer. c 5′-FDA was not isolated. d [octylmim][BF4]: 1-octyl-3-methylimidazolium tetrafluoroborate; [hexylmim][PF6]: 1-hexyl-3methylimidazolium hexafluorophosphate.

liquids ([hexylmim][BF4], [octylmim][BF4], [octylmim][PF6], and/or [hexylmim][PF6], entries 16–19) are convenient for the improvement of the total conversion yields to increase up to 2.2–4 times more than that of nonimmobilized fluorinase. In this system, it seems that the transformation of 5′-FDA from the immobilized fluorinase to the ionic liquid layer, such as the immiscible hydrophobic ionic liquids ([octylmim][PF6] and [hexylmim][PF6]), is the driving force in increasing the conversion yield. Furthermore, in the use of more miscible ionic liquids or Tris-buffer solution, it seems that the transformation of SAM from the immobilized fluorinase to the ionic liquid layer is easier than that of the immiscible hydrophobic ionic liquids. Therefore, the higher concentration of SAM in the immiscible hydrophobic ionic liquids may result in a higher conversion yield in the first run.

c10.indd 167

1/20/2011 10:27:58 AM

168

ENZYMATIC REACTIONS

REFERENCES 1. (a) S. V. Malhotra, eds., ACS Symposium Series 950, ACS, (2007). (b) S. Chowdhury, R. S. Mohan, J. L. Scott, Tetrahedron, 63, 2363 (2007). 2. (a) T. Hayashi, K. Tomioka, O. Yonemitsu, eds., Asymmetric Synthesis, Kodansha and Gordon and Breach Science Publishers, (1998). (b) K. K. Laali, ed., Advances in Organic Synthesis, Bentham Science (2006). (c) F. van Rantwijk, R. Sheldon, Chem. Rev., 107, 2757 (2007). 3. (a) T. Itoh, E. Akasaki , K. Kubo, S. Shirakami, Chem. Lett., 30, 262 (2001). (b) S. H. Schöfer, N. Kaftzik, P. Wasserscheid, U. Kragl, Chem. Commun., 5, 425 (2001). (c) J. Howarth, P. James, J. F. Dai, Tetrahedron Lett., 42, 7517 (2001). (d) M. Eckstein, M. V. Filho, A. Liese, U. Kragl, Chem. Commun., 8, 1084 (2005). 4. (a) R. Azerad, D. Buisson, Curr. Opin. Biotechnol., 11, 565 (2000). (b) A. Berkessel, M. L. Sebastian-Ibarz, T. N. Müller, Angew. Chem. Int. Ed., 45, 6567 (2006). 5. K. Nakamura, M. Fujii, Y. Ida, Tetrahedron: Asymmetry, 12, 3147 (2001). 6. T. Matsuda, Y. Yamagishi, S. Koguchi, N. Iwai, T. Kitazume, Tetrahedron Lett., 47, 4619–4622 (2006). 7. (a) D. O’Hagan, C. Schaffrath, S. Cobb, J. T. G. Hamilton, C. D. Murphy, Nature, 416, 279 (2002). (b) D. O’Hagan, R. J. M. Goss, A. Meddour, J. Courtieu, J. Am. Chem. Soc., 125, 379 (2003). (c) H. Deng, S. L. Cobb, A. D. Gee, A. Lockhart, M. Laurent, R. P. McGlinchey, D. O’Hagan, M. Onega, Chem. Commun., 652 (2006). (d) S. L. Cobb, H. Deng, A. R. McEwan, J. H. Naismith, D. O’Hagan, D. A. Robinson, Org. Biomol. Chem., 4, 1458–1460 (2006). (e) M. Sanada, T. Miyano, S. Iwadare, J. M. Williamson, B. H. Arison, J. L. Smith, A. W. Douglas, J. M. Liesch, E. Inamine, J. Antibioti. 39, 259 (1986). (b) T. Tamura, M. Wada, N. Esaki, K. Soda, J. Bacteriol., 177, 2265 (1995). 8. C. Dong, F. Huang, H. deng, C. Schaffrath, J. B. Spencer, D. O’Hagan, J. H. Naismith, Nature, 427, 561–565 (2004). 9. (a) T. Kitazume, K. Kasai, Green Chem., 3, 30 (2000). (b) T. Kitazume, H. Nagura, S. Koguchi, J. Fluorine Chem., 125, 79 (2004). (c) S. Koguchi, T. Kitazume, Tetrahedron Lett., 47, 2797 (2006).

c10.indd 168

1/20/2011 10:27:58 AM

11 MOLECULAR SELF-ASSEMBLY IN IONIC LIQUIDS Nobuo Kimizuka and Takuya Nakashima

Room-temperature ionic liquids are considered environmentally benign solvents, which gathered much interest and lead to a wide range of applications including organic chemical reactions [1], separations [2], and electrochemical applications [3,4]. They take advantage of the following unique properties: low vapor pressure, high ionic conductivity, and limited miscibility with water and common organic solvents [5]. Some recent studies also highlight their use for inorganic synthesis [6,7] and molecular self-assembly [8,9]. Ionic liquids as media for molecular assemblies are particularly important because they provide new self-assembling materials. Historically, the design of amphiphiles assumed the use of water as solvent, and they have been designed by connecting hydrophilic and hydrophobic moieties. Meanwhile the formation of ordered molecular self-assemblies such as bilayer membranes has been extended to organic solvents, by introducing the general concept of solvophilic-solvophobic amphiphilicity[10]. For gels formed in water or in organic solvents, the molecular design required for gelator molecules seems to be more facile compared with those for bilayer membrane-forming amphiphiles in which stable solvation of lipophilic units need to be satisfied to keep the assembly dispersed in the corresponding solvents. The use of ionic liquids as solvent allows us to acquire new knowledge on ionophilic-ionophobic functionalities, which can be done by systematically designing molecules that behave amphiphilically in ionic liquid and developing new ionic liquids that are capable of solvating ordered molecular assemblies. In 2001, we reported that ether-containing ionic liquids that carry bromide counter anions can dissolve carbohydrates, Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

169

c11.indd 169

1/20/2011 10:27:59 AM

170

MOLECULAR SELF-ASSEMBLY IN IONIC LIQUIDS

glycosylated proteins, and glycolipid bilayer membranes [8]. Our report touched off the succeeding studies on the dissolution of carbohydrates [11] and self-assembly in ionic liquids, including gelation of ionic liquids by molecular assemblies. In this chapter, we first give an overview of molecular assemblies in water and in organic media and then introduce molecular self-assemblies in ionic liquids. It complements reviews on these molecularly organized sytems [10,12].

11.1 MOLECULAR SELF-ASSEMBLIES IN AQUEOUS AND IN ORGANIC MEDIA Aqueous molecular assemblies such as micelles and bilayer membranes are formed by the self-assembly of amphiphilic compounds (Figure 11.1a, b) [10]. Aqueous micelles have been used for a variety of applications in surfactant industry, including emulsification, washing, and extraction processes [13]. Bilayer membranes are basic structural components of biomembranes, and their structures are maintained even in dilute aqueous media. This is in contrast to micelles that show dynamic equilibrium between aggregates and monomeric species. Thus, bilayers are more stable and sophisticated selfassemblies, and they require a suitable molecular design of the constituent amphiphiles [10]. Bilayer membranes and vesicles have wide-ranging applications, as exemplified by drug delivery [14], sensors [15], and bilayer-templated material synthesis [16]. The basic difference between the micelle-forming surfactants and bilayerforming amphiphiles lies on the chemical structure. Typical micelle-forming

(a)

(b)

(c)

(d) CH2 CH3(CH2)15 N CH3 CH3 Br CTAB

CH3(CH2)n-1 CH3(CH2)n-1

CH3 N CH3 Br

+

2CnN ( n = 12~ )

Figure 11.1. Molecular structures and schematic illustrations of micelle and bilayer membranes: a micelle, b bilayer membrane and vesicle, c CTAB, and d 2CnN+.

c11.indd 170

1/20/2011 10:27:59 AM

FORMATION OF MICELLES AND LIQUID CRYSTALS

171

surfactants have a single alkyl chain (hydrophobic group) and a hydrophilic group, as exemplified by cetyltrimethylammonium bromide (CTAB; Figure 11.1c). In contrast, double-chained amphiphiles 2CnN+ (Figure 11.1d; typically, n = 12–18) spontaneously give bilayer membranes when dispersed in water [10]. Although the formation of amphiphilic molecular assemblies in water is facilitated by hydrophobic interactions, micelles and bilayers are available even in organic solvents. For example, single-chain alkylammonium salts form micelles in polar solvents such as hydrazine [17], formamide [18], and glycerol [19]. The ability to form a hydrogen bonded network is not a mandatory characteristic of the solvents, because micelles and bilayer membranes also are formed in nonpolar, hydrocarbon solvents [10,20,21]. These findings clearly indicate that self-assembly in solution is a general phenomenon observed for amphiphilic molecules consisting of solvophobic and solvophilic moieties. Bilayer membranes also were formed in water-alcohol mixtures [22] and in ethanol [23]. In these polar media, enhancement of the intermolecular interactions is required to maintain the bilayer structure. The formation of mesoscopic supramolecular assemblies in chloroform by amphiphilic complementary hydrogen bond networks clearly shows that solvophobic interactions are rationally designable by combining hydrogen bonds and aromatic stacking [24]. Clearly, the formation of molecular assemblies is not limited to aqueous media. The interesting question is how to design solvophilic and solvophobic modules in ionic liquids.

11.2 EARLY STUDIES OF THE FORMATION OF MICELLES AND LIQUID CRYSTALS IN IONIC LIQUIDS The formation of micelles in ionic liquid was observed more than 30 years ago. At that time, the ionic liquid was referred to as a low-melting fused salt [25– 27]. Bloom and Reinsborough determined the critical micelle concentration (CMC) of CTAB in pyridinium chloride (melting point, 146 °C) by means of surface tension studies (CMC, 66 mM at 155 °C) [25]. Evans et al. reported the formation of micelles from CTAB in N-ethylammonium nitrate (EAN, melting point, 12 °C). N-ethylammonium nitrate was used because it has three hydrogen bonding sites for each solvent molecule, and it has the ability, like water, to form such three-dimensional structures in the liquid state. The CMC of CTAB micelles in ethylammonium nitrate EAN was determined to be 2.2 mM at 55 °C, which is 7–10 times larger than what is observed in aqueous solution [26]. This is attributable to the considerably lower interfacial tension for oilfused salt (ca. 20 mN m−1) compared with that observed for oil-water (ca. 50 mN m−1). Thus, the ethylammonium cation is a better solvent than water for hydrocarbons. Evidence for micelle formation also was given by light-scattering experiments. The micellar aggregation number and a hydrodynamic radius of hexadecylpyridinium bromide were determined as approximately 26 and

c11.indd 171

1/20/2011 10:27:59 AM

172

MOLECULAR SELF-ASSEMBLY IN IONIC LIQUIDS

22 ± 3 Å, respectively [27]. For spherical micelles, these numbers correspond to surface areas of 100 Å2, which is considerably larger than the corresponding values of 50 – 60 Å2 observed in an aqueous solution. These values are consistent with either a small classical spherical micelle containing only surfactant or a spherical mixed micelle containing surfactant and ethylammonium ions as a cosurfactant. Evans et al. also showed that the 1 : 1 mixture of EAN and β, γ-distearoyl-phosphotidylcholine (DSPC) gives a smectic A texture in the temperature range of 57.3 °C to 100 °C [28]. This is the first notice of lyotropic lamellar liquid crystals formed in the ionic medium. Recently, these earlier studies of micelle and liquid crystalline formation in EAN were critically reviewed by Drummond, which emphasized the difference between aqueous system [9]. However, prior to our report [8], no report existed on the selfassembly of developed nanoassemblies that are stably dispersed in ionic liquids. In the next section, the formation of bilayer membranes and vesicles in ionic liquids is discussed.

11.3 SUGAR-PHILIC IONIC LIQUIDS: DISSOLUTION OF CARBOHYDRATES AND FORMATION OF GLYCOLIPID BILAYER MEMBRANES, IONOGELS We developed ionic liquids that molecularly dissolve carbohydrates and polysaccharides by using ether-containing N, N′-dialkylimizalolium derivatives, (Scheme 11.1) [8]. The ether linkages were introduced in the alkyl chain moieties in anticipation of their role as a hydrogen bond acceptor for hydrodxyl or ammonium groups. These compounds become solid upon freezedrying, but they liquefy at room temperature when exposed to air, probably because of the absorption of water from the atmosphere. These ionic liquids were miscible with water, and the content of water in these ionic liquids was 2.6 wt% for 1 and 2.5 wt% for 2, as determined by the Karl Fischer method. To compare the solvent properties of these ionic liquids with conventional ionic liquids, the compound 3 in Scheme 11.2 was saturated with water (water content, 2.5 wt%)

N + N

BrO n

1;n=2 2;n=1

Scheme 11.1

N + N

PF6-

3

Scheme 11.2

c11.indd 172

1/20/2011 10:27:59 AM

173

SUGAR-PHILIC IONIC LIQUIDS

and used as a reference. Interestingly, when heated, the ether-containing ionic liquids 1 and 2 homogeneously dissolve carbohydrates such as β-d-glucose (solubility, 450 mg mL−1), α-cyclodextrin (350 mg mL−1), amylose (30 mg mL−1), and agarose (10–20 mg mL−1). The observed solubility is not ascribed to the water molecules slightly coexisting in these ionic liquids, because amylose, which is a coiled polysaccharide of α-1,4-glucosyl units and only slightly soluble in pure water (water solubility below 0.5 mg ml−1)[29], can be dissolved. A highly glycosylated protein, glucose oxidase (GOD), was also soluble in these ionic liquids (concentration, 1 mg mL−1), whereas other proteins such as cytochrome c, myoglobin, hemoglobin, and catalase were insoluble. The “sugar-philic” property observed for the ether-containing ionic liquids 1 and 2 could be attributed to the presence of bromide anion because the solubility of sugar derivatives was diminished when the bromide was replaced by bis-trifluoromethanesulfonylamide (TFSA). It seems that bromide ions effectively solvate the hydroxyl groups by hydrogen bonding. The dissolution of cellulose in chloride- or bromide-containing ionic liquids upon heating also was reported by Rogers et al. [11]. Interestingly, when the heated agarose solution of 1 was cooled to room temperature, the ionic liquid gelatinized. We have coined a term “ionogel” to express this physical state of gelation because the similar terms “hydrogel” and “organogel” have been commonly used to refer the solvents being gelatinized [8]. To investigate the possibility of ordered molecular assemblies forming in ionic liquids, the solubility of glycolipids (Schemes 11.3 and 11.4) were tested. l-Glutamate derivatives 4 (n = 12, 16) were employed because they are insoluble in water, even at a concentration of 1 mM. It was found that these waterinsoluble glycolipids can be dispersed in ionic liquids 1 and 2 as microcrystalline aggregates (Figure 11.2a). Gel-to-liquid crystalline phase transition is one of H OH OH O N CH3(CH2)n-1 O C CH OH CH2 O OH OH CH3(CH2)n-1 O C CH2 O 4 ( n = 12, 16 )

Scheme 11.3 H OH OH H O N CH3(CH2)11O(CH2)3 N C CH OH CH2 O OH OH CH3(CH2)11O(CH2)3 N C CH2 H O 5

Scheme 11.4

c11.indd 173

1/20/2011 10:27:59 AM

174

MOLECULAR SELF-ASSEMBLY IN IONIC LIQUIDS

(a)

20 mm (b)

20 mm (c)

20 mm

Figure 11.2. Dark-field optical micrographs of glycolipids dispersed in ionic liquid 1: a 4 (n = 12), b 5 (20 °C), and c 5 (50 °C). [4] = [5] = 10 mM.

c11.indd 174

1/20/2011 10:27:59 AM

SUGAR-PHILIC IONIC LIQUIDS

175

Figure 11.3. Picture of self-assembling ionogel formed from 5 in 1 (10 mM).

the basic aggregate characteristics of bilayer membranes and that of 4 (n = 16) was investigated by (DSC). The endothermic peaks (Tc) were observed at 47 °C (ΔH, 40.6 kJ mol−1; ΔS, 127.0 J K−1 mol−1) and at 66 °C (ΔH, 32.4 kJ mol−1; ΔS, 95.7 J K−1 mol−1) in ionic liquid 1, and in ionic liquid 2, a peak was observed at 43 °C (ΔH, 33.0 kJ mol−1; ΔS, 104.3 J K−1 mol−1). These ΔH and ΔS values are in the same order as those observed for the gel-to-liquid crystal phase transition of aqueous bilayer membranes, and therefore, glycolipids 4 become dispersed in ionic liquids 1 and 2 with the basic structure of bilayer membrane. Surprisingly, when glycolipid 5, which contains ether-linkages in alkylchains and three amide bonds, was dissolved in ionic liquids in 1 and 2, physical gelation occurred (concentrations above 10 mM, Figure 11.3). The developed fibrous structures whose length extends to several hundred micrometers were abundantly present in dark-field optical microscopy (Figure 11.2b). The DSC measurement of 5 in ionic liquid 1 showed an endothermic peak at 40 °C, at which temperature the ionogel was dissolved. This is the first case of physical gelation occurring in ionic liquids because of the self-assembly of low-molecular weight compounds [8]. To our surprise, this phase transition was accompanied by a reversible morphological transformation of the fibrous aggregate into vesicles (Figure 11.2c; with diameters of 3–5 μm) [8]. These findings indicate that the ether-containing ionic liquids behave as water-like solvents for carbohydrates and molecular assemblies. The ether-linkages seem to have enhanced the net polarity of the ionic liquids and thus to facilitate the self-assembly.

c11.indd 175

1/20/2011 10:27:59 AM

176

11.4

MOLECULAR SELF-ASSEMBLY IN IONIC LIQUIDS

CHARGED BILAYER MEMBRANES IN IONIC LIQUIDS

Ionic surfactants and amphiphiles intrinsically possess solvophilic (or ionophilic) groups. We therefore were interested in learning whether the simple dialkylammonium bromides (2CnN+, n = 12, 14; Figure 11.1d) that belong to the original family of synthetic bilayer membranes [10] form a bilayer in ionic liquids. Amphiphiles 2CnN+ were dispersed in three ionic liquids (Schemes 11.1–11.2) by ultrasonication (concentrations, 10 mM). Although 2C12N+, as homogeneously dispersed in the conventional ionic liquid 3, 2C14N+ was insoluble even at the lower concentration of 1 mM. In the DSC measurement, the 2C12N+ in ionic liquid 3 gave no endothermic peak (Figure 11.4c). Because no aggregate structure was observed under dark-field optical microscopy, it seems that 2C12N+ must have dispersed without forming bilayer membranes. Unlike the conventional ionic liquid 3, both amphiphiles were dispersed homogeneously in the ether-containing ionic liquids and gave translucent dispersions. The 2C12N+ showed an endothermic peak at 52 °C (ΔH, 18.0 kJ mol−1; ΔS,

Tc

gel phase

liquid crystalline phase

(a)

Endothermic

(b)

52 °C

72 °C

(c)

30

40

50

60

70

80

T / °C

Figure 11.4. Schematic illustration of gel-to-liquid crystalline phase transition and DSC thermograms: a 2C12N+ dispersed in 1, b 2C14N+ dispersed in 1, c 2C12N+ dispersed in 3. Similar endothermic peaks also were observed for 2CnN+ (n = 12, 14) in 2. [2CnN+] = 10 mM.

c11.indd 176

1/20/2011 10:27:59 AM

GENERALIZATION OF MOLECULAR SELF-ASSEMBLIES

177

55.3 J K−1 mol−1, Figure 11.4a) in ionic liquid 1 and similarly at 51 °C (ΔH, 12.5 kJ mol−1; ΔS, 37.2 J K−1 mol−1) in 2. The longer alkyl-chained compound 2C14N+ gave endothermic peaks at higher temperatures (72 °C in 1, ΔH, 23.8 kJ mol−1, ΔS, 69.2 J K−1 mol−1, Figure 11.4b; 71 °C in 2, ΔH, 22.1 kJ mol−1, ΔS, 64.4 J K−1 mol−1). These endothermic peaks are considerably higher than those observed for the corresponding aqueous bilayers (n = 12; Tc, below 0 °C, n = 14; Tc, 16 °C) [30]. The gel-to-liquid crystalline phase transition temperature of charged bilayer membranes is affected by the balance of electrostatic repulsions and attractive interactions such as van der Waals forces. In ionic liquids, it is expected that the electrostatic repulsion between ammonium head groups are highly screened. This leads to the increased association interactions in the totally ionized solvent that simultaneously enhances the thermal stability of gel(crystalline)-state bilayers. In dark-field optical microscopy, the 2C12N+ that dispersed in 1 showed microcrystalline aggregates at room temperature, which is consistent with its multilayered structure (Figure 11.5a). Remarkably, these aggregates were reversibly transformed into vesicles upon heating above Tc (Figure 11.5b) [31]. Similar morphological changes also were observed for 2C12N+ dispersed in ionic liquid 2 and for the 2C14N+ dispersed in ionic liquids 1 (Figure 11.5c) and 2. These observations confirm that the ammonium amphiphiles 2CnN+ form bilayer membranes in ether-containing ionic liquids [31]. Because these bilayers did not form in ionic liquid 3, the ether linkages and bromide counter ions in 1 and 2 again must play some important role. Because ether linkages are ionophilic [32], they should display a high affinity to the ammonium groups. The bromide anion in ionic liquids 1 and 2 also secured stable solvation of the ion pairs unlike the hydrophobic PF6− anions, which lowered their solubility.

11.5 CONTROL OF IONOPHILIC–IONOPHOBIC INTERACTIONS AND GENERALIZATION OF MOLECULAR SELF-ASSEMBLIES IN IONIC LIQUIDS It is now clear that the molecular assembly in ionic liquids mainly is governed by the following factors: (i) the ionophilic-ionophobic balance of the constituent molecules and (ii) the chemical structure of the ionic liquids. The increase in intermolecular interactions in bilayer-forming amphiphiles makes their selfassembly possible even in conventional ionic liquids. In aprotic organic media, hydrogen bonding occurs efficiently. The multiple-hydrogen bond-forming bilayers [33,34] and mesoscopic-sized supramolecular assemblies [24] are stably dispersed in the organic media. The cationic glutamate amphiphile (Scheme 11.7) has three amide linkages, and it forms fibrous nanoassemblies when dispersed in chloroform [34]. The amphiphiles, 8 and 9 (Schemes 11.7 and 11.8) are dispersed in ionic liquids 6 and 7 (Schemes 11.5 and 11.6), which contain TFSA anions. Although the diester-type amphiphile 9 did not form aggregates in ionic liquid 7, vesicles were formed in the ether-containing ionic

c11.indd 177

1/20/2011 10:27:59 AM

178

MOLECULAR SELF-ASSEMBLY IN IONIC LIQUIDS

(a)

(b)

(c)

Figure 11.5. Dark-field optical micrographs of 2CnN+ dispersed in ionic liquids: a 2C12N+ in 1, at room -temperature, b 2C12N+ in 1, at 60 °C, c 2C14N+ in 1, at 80 °C. Scale bar: 20 μm. [2CnN+] = 10 mM.

c11.indd 178

1/20/2011 10:28:00 AM

GENERALIZATION OF MOLECULAR SELF-ASSEMBLIES

N + N

179

TFSAO

6 TFSA- = N(SO2CF3)2-

Scheme 11.5

TFSAN + N 7 TFSA- = N(SO2CF3)2-

Scheme 11.6

H O H O CH3(CH2)11 N C CH N C (CH2)10 N (CH2)2OH CH2 BrCH3(CH2)11 N C CH2 8 H O

Scheme 11.7

O H O CH3(CH2)11 O C CH N C (CH2)10 N (CH2)2OH CH2 BrCH3(CH2)11 O C CH2 9 O

Scheme 11.8

liquid 6 (above Tc). This is consistent with the molecular-assembly-supporting property of ether-containing ionic liquids. In contrast, the amide-enriched compound 8 afforded self-assembling ionogels in both ionic liquids 6 and 7 (concentrations, 10 mM) [35]. Apparently, enhancement of intermolecular interactions by introducing multiple hydrogen bonding drives self-assembly even in the conventional ionic liquids. It is noteworthy that amphiphile 8 does not form hydrogels when dispersed in pure water. Physical gelation of liquids requires cross linking of the nanofibrous assemblies dispersed in the solvents, which would be suppressed in water because of electrostatic repulsive forces operating between the fibrous nanoassemblies [36]. The formation of ionogels that form from 8 must be a consequence of the suppressed electrostatic repulsive forces in ionic environments (the concentration of ionic liquids is ca. 7 M).

c11.indd 179

1/20/2011 10:28:00 AM

180

MOLECULAR SELF-ASSEMBLY IN IONIC LIQUIDS

In summary, the formation of self-assemblies in ionic liquids is a general phenomenon that depends on the molecular structure of amphiphiles, ionic liquids, and their combinations. The ether-introduced ionic liquids facilitate the formation of molecular self-assemblies. However, stable self-assemblies also can form in conventional ionic liquids if the intermolecular interactions are suitably designed.

11.6

SUMMARY, UPDATES AND OUTLOOK

In this chapter, we introduced molecular self-assembly, which is becoming growing research area in the chemistry of ionic liquids. Our own earlier reports [8, 31] had touched off a burst of research on molecular assembly in ionic liquids and ionogels. The solubilization of developed nanoassemblies including vesicles was extended further by a metal-fluorous surfactant complex, amphiphilic polymers, and block-copolymers. Hao et al. demonstrated the vesicle formation from Zn2+-fluorous surfactant complex in 1-butyl-3methylimidazolium (BMIm) BF4 and BMIm PF6 [37]. The nonionic polymer surfactant, Brij 76, showed a tentative liquid crystal phase progression of nanofibers to gel to vesicles upon heating in BMIm BF4 [38]. The self-assembling structures of polybutadiene-poly(ethylene oxide) (PB-PEO) block copolymers also were studied in ionic liquids by Lodge et al. [39]. BMIm PF6 was shown to be a good solvent for PEO block as well as a nonsolvent for PB block, leading to strong segregation of PB block from well-solvated PEO block. The self-assembling structures readily varied from spheres to wormlike micelles and then to bilayer vesicles depending on the length of the PEO block [39]. Supramolecular complexation of polymers via noncovalent interactions in ionic liquids has also been demonstrated to form thin films [40] and gels [41]. Polyelectrolytes multilayer films were prepared by a layer-by-layer method from ionic liquid solutions [40], although electrostatic interactions between oppositely charged polyelectrolytes are effectively screened in the ionic media. The formation of self-assembling ionogels and liquid crystalline mesophases are now studied extensively and will be reviewed further in Chapters 27 and 28, respectively. The formation of molecular assemblies in ionic liquids concurrently introduces supramolecular interfaces. Interfaces in ionic liquid would play important roles to develop interfacial materials chemistry. We have shown that the introduction of microscopic interfaces expands the potential of ionic liquids toward interfacial materials synthesis. By conducting an interfacial sol-gel reaction of Ti(OBu)4 dissolved in toluene microdroplets dispersed in ionic liquids, hollow TiO2 microcapsules were facilely obtained [42]. In this interfacial synthesis, ionic liquids act not only as solvents but also as stabilizers to keep the formed TiO2 microcapsules stably dispersed in the ionic reaction media. Ultra-thin gold nanosheets also were prepared by using an ionicwater interface as a template for photoreducing gold ions [43]. Other recent

c11.indd 180

1/20/2011 10:28:00 AM

REFERENCES

181

reports on microporous lamellar metaloxide nanocrystals [44] also confirm the potential of interfacial materials chemistry in ionic liquids. The use of selfassemblies in ionic liquids toward interfacial materials chemistry also will be possible. For example, semiconductor ZnS nanoparticles were prepared successfully templating the vesicles formed by Zn2+-fluorous surfactant in ionic liquids [37]. The strategy of self-assembly could be applied to open up developments in molecular-based nanomaterials. We believe that the combination of ionic liquids and biomolecules as well as organic molecular self-assemblies and inorganic nanomaterials can lead to new dimensions in materials science.

REFERENCES 1. V. I. Parvulescu, C. Hardacre, Chem. Rev., 107, 2615 (2007). 2. X. Han, D. W. Armstrong, Acc. Chem. Res., 40, 1079 (2007). 3. D. R. MacFarlane, J. M. Pringle, P. C. Howlett, M. Forsyth, Chem. Phys. Phys. Chem., 12, 1659 (2010). 4. M. Armand, F. Endres, D. R. MacFarlane, H. Ohno, B. Scrosati, Nat. Mater., 8, 621 (2009). 5. K. R. Seddon, A. Stark, M.-J. Torres, Pure Appl. Chem., 72, 2275 (2000). 6. T. Nakashima, N. Kimizuka, J. Am. Chem. Soc., 125, 6386 (2003). 7. Z. Ma, J. H. Yu, S. Dai, Adv. Mater., 22, 261 (2010). 8. N. Kimizuka, T. Nakashima, Langmuir, 17, 6759 (2001). 9. T. L. Greaves, C. J. Drummond, Chem. Soc. Rev., 37, 1709 (2008). 10. T. Kunitake, Angew. Chem. Int. Ed., 31, 709 (1992). 11. R. P. Swatloski, S. K. Spear, J. D. Holbrey, R. D. Rogers, J. Am. Chem. Soc., 124, 4974 (2002). 12. (a) N. Kimizuka, Curr. Opin. Chem. Biol., 7, 702 (2003). (b) N. Kimizuka, in Supramolecular Design for Biological Applications, N. Yui, ed., CRC Press, 2002, p. 373. (c) N. Kimizuka, in Supramolecular Polymers, 2nd ed., A. Ciferri ed., Taylor & Francis, 2005. 13. H. Hoffmann, C. Elbert, Angew. Chem. Int. Ed. Engl., 27, 902 (1988). 14. (a) A. D. Miller, Angew. Chem. Int. Ed., 37, 1768 (1998). (b) I. S. Zuhorn, D. Hoekstra, J. Membr Biol., 189, 167 (2002). 15. S. Okada, S. Peng, W. Spevak, D. Charych, Acc. Chem. Res., 31, 229 (1998). 16. (a) K. Sakata, T. Kunitake, Chem. Commun., 504 (1990). (b) S. Asakuma, H. Okada, T. Kunitake, J. Am. Chem. Soc., 113, 1749 (1991). (c) T. Soejima, T. Hasegawa, M-a Morikawa, N. Kimizuka, Chem. Lett., 38, 688 (2009). 17. D. F. Evans, Langmuir, 4, 3 (1988). 18. I. Rico, A. Lattes, J. Phys. Chem., 90, 5870 (1986). 19. A. Ray, Nature, 231, 313 (1971). 20. M. P. Turberg, J. E. Brady, J. Am. Chem. Soc., 110, 7797 (1988). 21. Y. Ishikawa, H. Kuwahara, T. Kunitake, J. Am. Chem. Soc., 116, 5579 (1994).

c11.indd 181

1/20/2011 10:28:00 AM

182

MOLECULAR SELF-ASSEMBLY IN IONIC LIQUIDS

22. N. Kimizuka, T. Wakiyama, H. Miyauchi, T. Yoshimi. M. Tokuhiro, T. Kunitake, J. Am. Chem. Soc., 118, 5808 (1996). 23. N. Kimizuka, M. Tokuhiro, H. Miyauchi, T. Wakiyama, T. Kunitake, Chem. Lett., 26, 1049 (1997). 24. N. Kimizuka, T. Kawasaki, K. Hirata, T. Kunitake, J. Am. Chem. Soc., 117, 6360 (1995). 25. H. Bloom, V. C. Reinsborough, Aust. J. Chem., 21, 1525 (1968). 26. D. F. Evans, A. Yamaguchi, R. Roman, E. Z. Casassa, J. Colloid Interface Sci., 88, 89 (1982). 27. D. F. Evans, A. Yamaguchi, G. J. Wei, V. A. Bloomfield, J. Phys. Chem., 87, 3537 (1983). 28. D. F. Evans, E. W. Kaler, W. J. Benton, J. Phys. Chem., 87, 533 (1983). 29. J. Putaux, A. Buléon, H. Chanzy, Macromolecules, 33, 6416 (2000). 30. Y. Okahata, R. Ando, T. Kunitake, Ber. Bunsen-Ges. Phys. Chem., 85, 789 (1981). 31. T. Nakashima, N. Kimizuka, Chem. Lett., 31, 1018 (2002). 32. N. Kimizuka, T. Wakiyama, T. Kunitake, Chem. Lett., 25, 521 (1996). 33. (a) H. Ihara, H. Hachisako, C. Hirayama, K. Yamada, J. Chem. Soc. Chem. Commun., 1244 (1992). (b) N. Yamada, E. Koyama, M. Kaneko, H. Seki, H. Ohtsu, T. Furuse, Chem. Lett., 387 (1995). 34. N. Kimizuka, M. Shimizu, S. Fujikawa, K. Fujimura, M. Sano, T. Kunitake, Chem. Lett., 27, 967 (1998). 35. T. Nakashima, N. Kimizuka, Ph.D thesis, Graduate School of Engineering, Kyushu University (2003). 36. T. Nakashima, N. Kimizuka, Adv. Mater., 14, 1113 (2002). 37. J. C. Hao, A. X. Song, J. Z. Wang, X. Chen, W. C. Zhuang, F. Shi, F. Zhou, W. M. Liu, Eur. J. Chem., 11, 3936 (2005). 38. J. Tang, D. Li, C. Sun, L. Zheng, J. Li, Colloids Surf., A, 273, 24 (2006). 39. (a) Y. He, Z. Li, P. Simone, T. P. Lodge, J. Am. Chem. Soc., 128, 2745 (2006). (b) P. M. Simone, T. P. Lodge, Macromolecules, 41, 1753 (2008). 40. T. Nakashima, J. Zhu, M. Qin, S. Ho, N. A. Kotov, Nanoscale, 2, 2084 (2010). 41. A. Noro, Y. Matsushita, T. P. Lodge, Macromolecules, 41, 5839 (2008). 42. T. Nakashima, N. Kimizuka, J. Am. Chem. Soc., 125, 6386 (2003). 43. T. Soejima, N. Kimizuka, Chem. Lett., 34, 1234 (2005). 44. M. Antonietti, D. B. Kuang, B. Smarsly, Z. Yong, Angew. Chem., Int. Ed., 43, 4988 (2004).

c11.indd 182

1/20/2011 10:28:00 AM

12 SOLUBILIZATION OF BIOMATERIALS INTO IONIC LIQUIDS Kyoko Fujita, Yukinobu Fukaya, and Hiroyuki Ohno

Ionic liquids (ILs) are expected to be unique solvents for electrochemical reactions because of their extraordinary properties. ILs can dissolve a wide variety of molecules and materials and are excellent solvents for the bioelectrochemistry field, but great care must be taken with the solubilization procedure because most biomaterials such as proteins and enzymes are characterized by higher ordered structures that have an origin in their numerous functions. Therefore, consideration must be given to the structural changes of biomaterials when they are dissolved in ILs. Not all ILs are good solvents for proteins, however. There is the interesting example of lipase. Lipase is soluble in both aqueous and organic solvents, so it can be easily solubilized in ILs. Certain lipases even become dispersed or dissolved in some ILs. Because lipase is a stable enzyme, it catalyzes the hydrolysis of lipids. Enzymatic activity is reported to be maintained in ILs [1]. There is not much published on the solubilization of biomaterials in ILs. In the present chapter, we introduce a procedure to use in solubilizing biomaterials in ILs for energy conversion. First, we consider the preparation of the ILs as solvents for cellulose. Then we proposed hydrated ILs as solvents for proteins as well as the chemical modification of biomaterials suitable for dissolution. We have found this procedure helpful when we tried to construct biofule cells or other biomolecular energy conversion systems using ILs.

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

183

c12.indd 183

1/20/2011 10:28:01 AM

184

12.1

SOLUBILIZATION OF BIOMATERIALS INTO IONIC LIQUIDS

POLAR IONIC LIQUIDS AS SOLVENTS FOR CELLULOSE

New energy production from renewable biomass is one of the most important tasks for the future. It is out of the question to treat an edible biomass as a starting material for energy production [2], and hence, we should develop processes with inedible biomass for that purpose. Because component polysaccharides of such biomasses exist as supramolecular complexes in nature, it is almost impossible to extract and solubilize polysaccharides from the biomass under mild conditions. To obtain energies from these biomasses, it should be necessary to minimize the energy spent for the process. Here, we describe the first step, namely the extraction of polysaccharides from biomass under mild condition. As a pioneering work of cellulose dissolution, Rogers et al. have reported that alkylimidazolium chloride salts are potential solvents for cellulose [2]. Despite the ability to dissolve the cellulose of these chloride salts, severe problems must be solved, such as chloride salts inherently showing both high melting temperature and high viscosity. However, a series of alkylimidazolium carboxylates such as acetate and formate salts are obtained as roomtemperature ILs with relatively low viscosity [3]. Despite the possibility of carboxylate-based ILs, however, the carboxylate series are not thermally stable enough that limits their use condition as solvents [4]. However, based on our fundamental studies on the relation between ion structures and their physicochemical properties, we have prepared novel polar ILs, namely a series of alkylphosphate-derivative ILs. These phosphate-derivative ILs displayed a higher hydrogen bond basicity than previously reported polar ionic liquids, namely 1-alkyl-3-methylimidazolium salts with chloride and carboxylate anions [5]. Furthermore, 1-ethyl-3-methylimidazolium methylphosphonate is also a good candidate for cellulose solvent. For example, this IL can dissolve cellulose owing to their strong hydrogen bonding basicity. A 2.0 wt% cellulose powder dissolved completely in this IL at room temperature (25 °C) within 3 hours, and 4.0 wt% cellulose powder dissolved within 5 hours. With the aid of this IL, we attempted to extract polysaccharides from naturally obtained biomass. Figure 12.1 shows the solubility of cellulose in three different ILs as a function of temperature under stirring for 30 minutes. These three ILs are potential candidates as a solvent for cellulose. Then, we describe the polysaccharide extraction from bran as a typical example for biomass extraction. The extraction of polysaccharides has been analyzed at ambient temperature under stirring. These phosphate derivative ILs have extracted a considerable amount of polysaccharides. Especially, 1-ethyl-3-methylimidazolium methylphosphonate extracted more than 32% of polysaccharides from added bran powder within 30 minutes. In conclusion, a series of phosphate derivative ILs, especially 1-ethyl-3-methylimidazolium methylphosphonate, was polar ionic liquid with relatively low viscosity, and they dissolved cellulose and extracted polysaccharides at high concentration from bran under mild conditions [6]. Extraction of cellulose should be fol-

c12.indd 184

1/20/2011 10:28:01 AM

NOVEL SOLVENT FOR DISSOLUTION OF NATIVE PROTEIN

185

Figure 12.1. Cellulose solubility for 30 minutes, depending on the IL structure and temperature.

lowed by the hydrolysis of the cellulose to produce glucose, cellobiose, and other oligosaccharides. These should be converted into electric energy by the biofuel cells mainly comprising ILs.

12.2 NOVEL SOLVENT FOR DISSOLUTION OF NATIVE PROTEIN: HYDRATED IL Polarity of ILs is an important factor for the dissolution of a biopolymer such as cellulose. Highly polar IL, such as 1-ethyl-3-methylimidazolium methylphosphonate, enabled the dissolution of cellulose without heating [6]. Such highly polar ILs also dissolve proteins; however, the process accompanied denaturation of proteins. The hydrogen bondings in the molecule were broken by the dissolution, which should be the factor to change the higher ordered structure. For the use of proteins in IL, it is important to keep their higherordered structure. Apart from the case of lipase, there was no suitable ILs as solvent for proteins. Thus, there are strong requests to prepare ILs that have a suitable affinity with (and no other influence to induce denaturation of) general proteins. We proposed a new IL system as one method to dissolve proteins, the so-called “hydrated ionic liquid.”

c12.indd 185

1/20/2011 10:28:01 AM

186

SOLUBILIZATION OF BIOMATERIALS INTO IONIC LIQUIDS

OH

N

N

HO

choline

C4mpyr

O

O

P

P

OH

O

O OC4H9

C4H9O

dhp

N

bmim

N

TMA

S O

methanesulfonate

O

O

tetrafluoroborate

Cl-

chloride

O

acetate

S

saccharinate

HO

O

O

BF4¯

N O

dbp

O N

O

lactate

O

N

C

N

C

N

dca

Figure 12.2. Structure of cations and anions for ILs examined in protein solubilization.

Concentrated salt solution has been believed to be an unsuitable solvent for protein because protein solubility turned worse after increasing the salt concentration in buffer solution. To prepare a hydrated IL as a solvent for proteins, some cations and anions were examined as shown in Figure 12.2. To prepare ILs after the following cations are used such as the well-known 1-butyl-3-methylimidazoloim (bmim), N-methyl-N-butyl pyrrolidinium (C4mpyr), and tetra methyl ammonium (TMA) and choline, which are both common in biological systems. The anions used in this study were hydrogen phosphate (dhp), methanesulfonate (MeSO4), dicyanamide (dca), acetate, lactate, chloride, tetrafluorobrate, dicyanamide, and saccharinate. Dibutyl phosphate (dbp) also was chosen because it has a similar structure to phosphate, but it lacks a hydroxide group as to investigate the contribution of hydrogen bonding. Some ILs prepared by the combination of these cations and anions were solid at room temperature. Except for TMA and dhp, all of these salts form a colorless liquid by adding of 20 wt% of water. Among these ILs containing 20 wt% water, the solubility of cytochrome c (cyt. c) was examined as a model protein. ILs containing anions such as saccharinate, chloride, dca, or tetrafluoroborate are not good solvents for cyt. c even after adding 20 wt% water. This result revealed the fact that cyt. c is difficult to solubilize in most ILs even in the presence of some water [7]. However, other ILs solubilized significant amounts of cyt. c (about 3 mM or 37 mg/mL), forming a dark red homogeneous solution as shown in Figure 12.3. The solubility of cyt. c made no difference in ILs prepared with C4mpyr, bmim, or choline cations. As the anions, it was strongly suggested to have the oxo acid residue for a sufficient solubility of cyt c The “hydrated ionic liquid” mainly consisted of “bound water” and therefore, it showed IL-like properties. Most ILs, however, cannot dissolve proteins even after adding a small amount of water. After adding 20 wt% water, a combination of choline cation and dhp anion was evaluated to be the most suitable salt for cyt c by ultraviolet-visible, circular dichroism,

c12.indd 186

1/20/2011 10:28:01 AM

187

SOLUBILIZATION METHOD OF PROTEINS IN IL

(a)

(b)

(c)

Figure 12.3. Photographs of cyt. c in (a) phosphate buffer, (b) C4mpyr dca with 20 wt% water, and (c) choline dhp with 20 wt% water.

and resonance Raman spectroscopy. Hydrated choline dhp showed its potential as a novel solvent to keep higher-ordered structure, activity, and thermal stability of cyt c. In the case of hydrated IL, one of the effects of the component ion on the protein activity after dissolution was shown as kosmotropicity, which is a similar relation with the Hofmeister series [8]. It was revealed that the secondary structure and intramolecular interactions at the heme crevice were strongly influenced by the kosmotropicity of the component ions [9]. Furthermore, significant thermal stability and long-term stability of the dissolved cyt c was observed in a hydrated choline dhp. It is important to evaluate hydrated ILs with their kosmotropicity for the development of new protein solvents.

12.3 SOLUBILIZATION METHOD OF PROTEINS IN IL: POLYETHER MODIFICATION The construction of a biosensor, a biochip, or a bioreactor frequently requires the fixation of corresponding biomacromolecules, especially proteins, on an electrode without denaturation. Protein functions have been studied only in an aqueous medium. However, the aqueous medium is not the best matrix in practical applications. The available temperature range of an aqueous system is not sufficiently wide, and it is scarcely possible to maintain the long-term stability of the proteins in an aqueous medium. It is therefore important to find a nonaqueous medium in which these biomolecules can play their respective roles without denaturation. In the previous section, the possibility and effectiveness of protein solubilization by hydrated IL has been taken up. A hydrated IL realized the direct dissolution of cyt c retaining the original structure and the activity; however, there are only a few ion species to construct a

c12.indd 187

1/20/2011 10:28:01 AM

188

SOLUBILIZATION OF BIOMATERIALS INTO IONIC LIQUIDS

δ– PEO CH2CH2

H2C O

n

O

H2C O

+

O CH 2 δ+ O CH2

C CH2 H2 Figure 12.4. Scheme of solvated ion by poly(ethylene oxide) (PEO).

hydrated IL as protein a solvent. General ILs cannot be used as a solvent for proteins even after the addition of water. For the use of general ILs as a protein solvent for the wide application, chemical modification on the proteins turned to be useful as an alternative method. We already have reported that salt-containing polyethers are excellent nonaqueous media to use in creating an electrochemical reaction field for biomolecules [10]. Polyethers, especially poly(ethylene oxide) (PEO), are a linear and polar polymer with a repeating unit similar to that of a water molecule (Figure 12.4). In our studies, redox reactions of different molecules were conducted in liquid and solid polyethers without using any water [11,12]. We observed the redox reactions of different PEO-modified proteins dissolved in other organic solvents as well as in PEO [11–13]. PEO modification already had proved to be an excellent way to make proteins soluble in nonaqueous solvents [14]. There are, nevertheless, some reports of enzymatic activities [15,16] and structure [17] determination of PEO-modified enzymes in homogeneous solutions of organic solvents. Although native heme proteins are insoluble and redox inactive in dehydrated PEO oligomers, these heme proteins have proved to be soluble and show redox activity even in PEO oligomers when they are chemically modified with PEO [18]. Furthermore, it has been revealed that PEO-modified proteins, dissolved in solvent PEO, continue to function over a wide temperature range [19] and show long-term stability [20]; these characteristics cannot be obtained in an aqueous medium. Because there is an effective interaction between the cation and the ether oxygen unit of PEO through the ion–dipole interaction, many different salts are soluble in PEO [21]. The affinity between PEO and ions is a strong indication that PEOs are soluble in ILs, as they indeed are soluble in numerous ILs. We measured solubility of PEOs in 1-ethyl-3-methylimidazolium bis(trifluoromethanesulfonyl)imide (emim TFSI, 1, Figure 12.5), which is a typical ionic liquid. PEO was mixed with this ionic liquid, and stirred for 12 hours. Most PEOs with different average molecular weight were found to be soluble in the IL. Even for a solid PEO, after the dissolution into the IL, there is no phase separation and reliquation of the PEO. This compelled us to examine

c12.indd 188

1/20/2011 10:28:01 AM

189

SOLUBILIZATION METHOD OF PROTEINS IN IL

N

N

O N O S CF3 F3C S O

O

Figure 12.5. emim TFSI, 1.

the solubilization of PEO-modified heme proteins in ILs for which we expected no denaturation. We modified cytochrome c, which is a typical heme protein, with PEO chains (PEO-cyt. c). We synthesized the PEO-cyt. c the same way as in our previous article [10]. The terminal-activated PEO (a-PEO) was prepared from these PEO monomethyl ethers. The ratio of a-PEO to cyt. c was changed in the feed to prepare the PEO-cyt. c with different degrees of PEO modification. The modification degree was determined by titration of the remaining amino groups of cyt. c with 2,4,6-trinitrobenzene sulfonic acid [22]. The cyt. c was modified with 16.7 chains of PEO150 (PEO150-cyt. c(16.7)), 14.3 chains of PEO550 (PEO550-cyt. c(14.3)), 15.4 chains of PEO1000 (PEO1000-cyt. c(15.4)), 6.1 or 13.5 chains of PEO2000 (PEO2000-cyt. c (6.1 or 13.5)), and 9.7 chains of PEO5000 (PEO5000-cyt. c(9.7)) to compare the effects of both the PEO chain length and the degree of modification on solubilization in 1. Native cyt. c, or the prepared PEO-cyt. c, was mixed with 1, and stirred slowly to reach the final concentration of 0.1 mM. Native cyt. c was insoluble in 1, but cyt. c modified with PEO150(16.7), PEO550(14.3), PEO1000(15.4), or PEO2000(6.1) was partly soluble. However, when cyt. c was modified with PEO2000(13.5) or PEO5000(9.7), it dissolved well in 1. Figure 12.6 shows the photographs after mixing. Figure 12.7 shows the optical want guide (OWG) spectra of the IL containing native cyt. c (a), PEO150-cyt. c(16.7) (b), and PEO2000-cyt. c(13.5) (c) [23]. (See Chapter 7 for more on the OWG spectroscopy.) Because native cyt. c was insoluble in 1, no absorption was observed. However, a Soret band was clearly observed at 408 nm in PEO2000-cyt. c(13.5) in 1. The appearance of this band varied depending on the IL species. It turned out that at least 10 chains of PEOs with an averaged molecular weight of more than 2000 were required for the dissolution of cyt. c in 1. From the absorption spectrum, we learned that the dissolved PEO-cyt. c contained heme with a ferric ion (Fe3+) and that the vicinity of the heme remained unchanged. Our results showed that modified PEO chains can be used to dissolve proteins into ILs. Goto et al. also had reported the effect of modification of comb-shaped poly(ethylene glycol) to solubilize the protein into ILs [24]. PEO modification thus was confirmed to be an effective way to dissolve biomolecules in ILs. Some other factors served to regulate the solubility of the prepared PEO-modified molecules in our series of IL studies. The details will be reported elsewhere.

c12.indd 189

1/20/2011 10:28:01 AM

190

SOLUBILIZATION OF BIOMATERIALS INTO IONIC LIQUIDS (a)

(b)

Figure 12.6. Solubility of native cyt. c (a) and PEO2000 modified cyt. c (b) into 1.

(b)

(c)

0.10

0.08

0.08

0.08

0.06 0.04

Absorbance

0.10

0.10

Absorbance

Absorbance

(a)

0.06 0.04 0.02

0.02

400

500

600

Wavelength (nm)

0.04 0.02

0.00

0.00

0.06

400

500

600

Wavelength (nm)

0.00 400

500

600

Wavelength (nm)

Figure 12.7. OWG spectra of native cyt. c (a), PEO150-cyt. c (b), and PEO2000-cyt. c (c) in 1.

12.4

NONAQUEOUS BIOFUEL CELLS

Recently, it is a keen and important task to produce energy from biomass. For this process, an inedible biomass should be used to substitute for foods. Cellulose is the most abundant and inedible biorenewable material [25]. However, natural cellulose is insoluble in water and most general organic

c12.indd 190

1/20/2011 10:28:01 AM

REFERENCES

191

molecular liquids because it consists of polydispersed linear glucose polymer chains; furthermore, a supramolecular complex structure is stabilized by many hydrogen bondings [26]. Degradation of cellulose in ILs should open a new strategy to build sustainable energy conversion systems such as biofuel cells. The following steps should be realized in ILs for the energy conversion of biomass. The first step is cellulose treatment and dissolution. The second step is depolymerization of cellulose. The third step is generation of electrons from produced mono- or oligosaccharides. In an early stage, Rogers et al. reported that 1-buthyl-3-methylimidazolium chloride, [bmim]Cl, possibly can dissolve cellulose at high temperature [2]. Furthermore, the dissolution in a mild condition with ILs has been reported [4,5]. For the second step, cellulose after solubilization was expected to be enzymatically converted into glucose in the IL. However, cellulase irreversibly lost the activity in [bmim]Cl [27]. Additional inventions are necessary to design better ILs for enzymatic depolymerization. At present, there is no potential ILs to satisfy their requirements for the previews three steps. In the near future, better ILs should be designed for nonaqueous biofuel cells.

REFERENCES 1. J. L. Kaar, A. M. Jesionowski, J. A. Berberich, R. Moulton, A. J. Russell, J. Am. Chem. Soc., 125, 4125 (2003). 2. R. P. Swatloski, S. K. Spear, J. D. Holbrey, R. D. Rogers, J. Am. Chem. Soc., 124, 4974 (2002). 3. Y. Fukaya, A. Sugimoto, H. Ohno, Biomacromolecules, 7, 3295 (2006). 4. H. Ohno, Y. Fukaya, Chem. Lett., 38, 2 (2009). 5. Y. Fukaya, K. Hayashi, M. Wada, H. Ohno, Green Chem., 10, 44 (2008). 6. M. Abe, Y. Fukaya, H. Ohno, Green Chem., in press. 7. K. Fujita, D. R. MacFarlane, M. Forsyth, Chem. Commun., 4804 (2005). 8. Zhao, H. Biophysi. Chem., 122, 157 (2006). 9. K. Fujita, D.R. MacFarlane, M. Forsyth, M. Yoshizawa-Fujita, K. Murata, N. Nakamura, H. Ohno, Biomacromolecules, 8, 2080 (2007). 10. H. Ohno, T. Tsukuda, J. Electroanal.Chem., 341, 137 (1992). 11. (a) G. Shi, H. Ohno, J. Electroanal. Chem., 314, 59 (1991). (b) H. Ohno, H. Yoshihara, Solid State Ionics, 80, 251 (1995). 12. H. Ohno, T. Tsukuda, J. Electroanal. Chem., 367, 189 (1994). 13. (a) H. Ohno, F. Kurusu, Chem. Lett., 693 (1996), (b) N. Nakamura, Y. Nakamura, R. Tanimura, N. Y. Kawahara, H. Ohno, Deligeer, S. Suzuki, Electrochim. Acta, 46, 1605 (2001). 14. G.-A. Humberto, V. Brenda, S.-R. Gloria, V.-D. Rafael, Bioconjugate Chem., 13, 1336 (2002). 15. K. Takahashi, T. Yoshimoto, A. Ajima, Y. Tamaura, Y. Inada, Enzyme, 32, 235 (1984). 16. A. Matsushima, M. Okada, Y. Inada, FEBS Lett., 178, 275 (1984).

c12.indd 191

1/20/2011 10:28:01 AM

192

SOLUBILIZATION OF BIOMATERIALS INTO IONIC LIQUIDS

17. P. A. Mabrouk, J. Am. Chem. Soc., 117, 2141 (1995). 18. H. Ohno, N. Yamaguchi, M. Watanabe, Polymer. Adv. Technol., 4, 133 (1993). 19. (a) H. Ohno, N. Yamaguchi, Bioconjugate Chem., 5, 379 (1994). (b) N. Y. Kawahara, H. Ohno, Bioconjugate Chem., 8, 643 (1997). 20. F. Kurusu, H. Ohno, Electrochim. Acta, 45, 2911 (2000). 21. H. Ohno, Electrochim. Acta., 37, 1649 (1992). 22. A. F. S. A. Habeeb, Anal. Biochem., 14, 328 (1966). 23. H. Ohno, C. Suzuki, K. Fukumoto, M. Yoshizawa, K. Fujita, Chem. Lett., 450 (2003). 24. K. Nakashima, T. Maruyama, N. Kamiya, M. Goto, Chem. Comm., 4297 (2005). 25. Kirk-Othmer, Encyclopedia of Chemical Technology, 4th ed. Wiley, 1993. 26. V. L. Finkenstadt, R. P. Millane, Macromolecules, 31, 7776 (1998). 27. M. B. Turner, S. K. Spear, J. G. Huddleston, J. D. Holbrey, R. D. Rogers, Green Chem., 5, 443 (2003).

c12.indd 192

1/20/2011 10:28:01 AM

13 REDOX REACTION OF PROTEINS Kyoko Fujita and Hiroyuki Ohno

13.1

PEO MODIFIED PROTEIN

There are some studies on the direct electron transfer between an electrode and heme proteins that suggest the applications of functional proteins in vitro [1]. Almost all of these studies have been investigated in aqueous media [2,3]. However, the electrochemistry of proteins can be captured in a variety of media, including organic solvents, polymers, and even ionic liquids. The redox reaction of molecules in ionic liquids (ILs) is usually analyzed by cyclic voltammetry (CV) measurement [4,5]. However, the redox response of PEO2000-cyt. c in ILs cannot be detected with ordinary CV measurements. The serious drawback has been that the large background current of the ILs masks the faradaic component. Also, the high viscosity of these media has been shown to induce low diffusion coefficients and low electrochemical currents [6,7]. Therefore, it is necessary to carry out the voltammetric experiments at relatively high concentrations of the electroactive substance, typically 10 mM, to get sufficient signal-to-noise ratios [8]. Spectroelectrochemistry is useful in such cases (see Chapter 7) because data can be obtained on the redox processes in the absence of a background current [9–12]. Because the absorption spectrum is the result of the integrated changes, it is easy to detect reactions that are difficult to analyze with ordinary electrochemical techniques. Specroelectrochemistry, optical wave guile (OWG) spectroscopy (see Chapter 7), has proved to be particularly useful for sensitive and simple spectral analysis of redox reactions of molecules in a small amount of solution on opaque

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

193

c13.indd 193

1/20/2011 10:28:02 AM

194

REDOX REACTION OF PROTEINS

electrodes [13]. The available μL range analysis makes it more suitable for expensive and highly valuable samples such as ILs. In this chapter, the redox reactions of proteins dissolved in ILs are analyzed with the introduction of OWG spectroscopy. As shown in Chapter 12, poly(ethylene oxide) (PEO) modification to cytochrome c (cyt. c) can effectively solubilize proteins without changing the heme vicinity in 1-ethyl-3-methylimidazolium bis(trifluoromethanesulfonyl)imide, 1—one of the typical ionic liquids. In particular, cyt. c modified with PEO2000(13.5) or PEO5000(9.7) dissolved well in 1. The redox activity of dissolved PEO2000-cyt. c(13.5) in 1 was analyzed by OWG spectroelectrochemistry [10]. The electrochemical cell system was constructed using the waveguide as shown previously in Chapter 7. Carbon plate (3 × 40 mm) was used as the working electrode, and Pt and Ag wires (0.5 mm in diameter) were used as the counter and reference electrode, respectively. PEO2000-cyt. c(13.5) was dissolved in 1 (0.1 mM), and 150 μL of this solution was directly introduced into the cell system. The OWG spectra were analyzed by applying the potential (−800 ∼ +800 mV vs. Ag, sweep rate: 5 mV/s−1). However, the spectral change because of the redox reaction of PEO-cyt. c in 1 was scarcely evident. Because the resistance of the solution was small as a result of the excellent ionic conductivity of the IL as a solvent, this could not be caused by the solvent’s resistance. This was thought to be caused by the lack of small ions, despite the extremely high density of ions in ILs. During the redox reaction, globular proteins containing active enter the inner domain, where small and suitably sized ion species are essential for the electron transfer activity in the ILs. Because KCl is known to be a good electrolyte for the electrochemical reaction of heme proteins in PEOs [14], we added KCl as the supporting electrolyte into 1. The solubility of KCl in 1 was, however, low (i.e., less than 0.1 M). PEO2000-cyt. c (13.5) (0.1 mM) then was dissolved in the KCl-saturated 1 in a similar analysis to that mentioned earlier. The spectrum was the same as that without KCl (Figure 13.1, solid line), indicating an oxidized form. When the negative potential (−0.8 V vs. Ag) was applied to the spectrum, the Soret band peak showed a red shift from 408 to 414 nm based on the reduction (Figure 13.1, dotted line). After that, the λmax returned to 408 nm by applying a positive potential (+0.5 V vs. Ag). This spectral change resolting from the redox reaction of PEO-cyt. c in 1 was observed repeatedly in situ and corresponded to the potential switching. From these absorbance changes, the repetition of the redox response of PEO-cyt. c also was observed in situ followed by the potential sweep. Figure 13.2 shows the absorbance change at 414 nm induced by the potential sweeping (sweep rate, 5 mV/s; potential sweep range, between −0.8 and +0.8 V vs. Ag). The stable absorbance change resulting from the redox reaction of PEO-cyt. c was observed without attenuation. Figure 13.3 shows the effect of the given potential on the absorbance change at 414 nm of PEO-cyt. c dissolved in 1 (a) and in KCl-saturated 1 (b). The KCl addition clearly affects the redox reaction of PEO-cyt. c in 1. No absorbance

c13.indd 194

1/20/2011 10:28:02 AM

195

PEO MODIFIED PROTEIN

0.10

Absorbance

0.08

0.06

0.04

0.02

0.00 400 500 Wavelength (nm)

600

Figure 13.1. OWG spectra of oxidized (solid) and reduced (dotted) PEO2000-cyt. c in KCl-saturated 1.

0.08

–0.8

Absorbance at 414 nm

–0.4 0.07

–0.2 0

0.065

0.2

0.06

0.4 0.055

Potential (V, Ag)

–0.6

0.075

0.6

0.05 0

500

0.8 1000 1500 2000 2500 3000 3500 4000 Time (sec)

Figure 13.2. Spectral change at 414 nm of PEO2000-cyt. c as a result of the potential applied in 1 containing 0.02 M KCl.

c13.indd 195

1/20/2011 10:28:02 AM

196

REDOX REACTION OF PROTEINS 0.08 (b)

Absorbance at 414 nm

0.07

0.06

0.05

0.04 (a) 0.03 0.8

0.4

0

–0.4

–0.8

Potential (V)

Figure 13.3. Effect of the given potential on the absorbance change at 414 nm of PEO2000-cyt. c dissolved in (a) 1 only, and (b) KCl-saturated 1.

change was confirmed along with the potential sweep, when PEO-cyt. c was dissolved in 1 without a salt addition (Figure 13.3a). Nevertheless, the redox reaction of PEO-cyt. c was improved by the addition of KCl in 1 (Figure 13.3a). Thus, PEO modification makes it possible to introduce cyt. c into 1, which consists of only oversized ions, but it is difficult to realize the electron transfer reaction of cyt. c under this condition in which the cyt. c are surrounded by relatively huge ions. It became evident that in the huge ionic system the presence of microions is essential to the electron transfer reaction of proteins with the mounted active center. Through the analyses of various salts, the Cl− ion was determined to be the most effective ion to use in improving the redox reaction of PEO-cyt. c dissolved in 1. From the absorbance change at certain wavelengths as the function of the given potential, we could obtain the electrochemical characteristics such as the redox potential. From the absorbance change as the function of the applied potential in Figure 13.3b, the oxidation-reduction potential of PEO-cyt. c dissolved in 1/KCl-mixed electrolyte solution was estimated to be about −0.3 V versus Ag. This redox reaction was the same even after we kept the solution in a dry desiccator for several months. The extreme stability is understood to be a result of the lack of water molecules.

c13.indd 196

1/20/2011 10:28:03 AM

NATIVE PROTEIN IN HYDRATED IL

197

Figure 13.4. Scheme of solvated PEO-modified protein in ionic liquid containing added salts.

Our study confirmed that a suitable-sized electrolyte is essential for the smooth electron transfer reaction of heme proteins in ILs, as shown in Figure 13.4.

13.2

NATIVE PROTEIN IN HYDRATED IL

As shown in Chapter 12, hydrated IL that consists of appropriate cation and anion dissolved native proteins retaining the higher-ordered structure and the activity. In this section, we would like to show the possibility of hydrated IL as a medium for the electron transfer reaction of proteins. Especially, the electron transfer reaction related to biocatalytic oxidation of biomass will lead to the construction of biofuel cell. After degradation of cellulose, not only glucose but also some other oligosaccharides were produced. These oligosaccharides also should be converted into electric energy by biofuel cell. Thus, we investigated the enzymatic oxidation of cellobiose in the hydrated ILs as the third step for the energy conversion of biomass as descried in Chapter 12. Hydrated choline hydrogen phosphate (dhp) (Hy[ch][dhp]) was used as a medium. Cellobiose dehydrogenase (EC 1. 1. 99. 18; CDH) was chosen as one of the target enzymes to harvest electrons from cellobiose. Cellobiose dehydrogenase (CDH) consists of two domains such as flavin (FAD) and the other consisted of a type-b heme. First, the electron transfer reaction of the FAD domain was examined by 2,6-dichloroindophenol sodium (DCIP). DCIP is used as a kind of indicator for the electron transfer reaction and receives two electrons directly from the FAD domain and reduced to DCIPH2 (Figure 13.5, type [A]). Generally, the electron transfer to DCIP from

c13.indd 197

1/20/2011 10:28:03 AM

198

REDOX REACTION OF PROTEINS CDH lactose

FAD 2e

cellobiono lactone

2 cyt c(Fe2+)

heme (Fe2+) e

-

e- [B]

heme (Fe3+)

-

FADH2

2 cyt c(Fe3+)

2eDCIPH2

DCIP

[A]

Figure 13.5. Schematic representation of the pathway of electron transfer of cellobiose, CDH, and cyt. c or DCIP.

Absorbance

0.6

0.4

0.2

0.0 400

500

600

700

800

Wavelength (nm)

Figure 13.6. Visible absorption spectra of DCIP in hydrated IL with CDH before (oxidized form; dotted line) and after (reduced form; solid line) mixing with cellobiose.

the FAD domain takes precedence over that occurring to the heme from the FAD domain because the electron transfer between FAD and DCIP is faster than that between FAD and heme. The spectral change of DCIP before and after mixing with cellobiose solution in the Hy[ch][dhp] is shown in Figure 13.6. The absorbance around 600 nm, based on DCIP, completely vanished after receiving electrons from CDH. It was confirmed that the FAD domain is active and can carry out the electron transfer reaction in Hy[ch][dhp]. Cyt. c is one well-known electron acceptor of CDH in vitro. The cyt. c was used as an indicator to confirm the intraelectron transfer reaction of CDH. The electron transfer path is proposed as shown in Figure 13.5, type [B], when cyt. c works as an electron acceptor of CDH. Two electrons are transferred at the FAD domain in the oxidation reaction of cellobiose. These electrons must be delivered one by one from FAD to the heme via an internal electron transfer process. Then, these electrons are transferred to the cyt. c from the heme

c13.indd 198

1/20/2011 10:28:03 AM

199

NATIVE PROTEIN IN HYDRATED IL

Absorbance

0.6 0.4 0.2 0.0 400

500 600 Wavelength (nm)

700

Figure 13.7. Absorption spectra of cyt. c in hydrated IL with cellobiose before (oxidized from; dotted line) and after (reduced form; solid line) mixing with CDH.

domain. When the inter- and intraelectron transfer reactions of CDH are progressing, it leads to cyt. c reduction in the end. That is, the cyt. c reduction indicates that definite activity occurs for both domains of the CDH. No spectral change was observed without CDH or cellobiose when the absorption spectrum of cyt. c was measured. However, as shown in Figure 13.7, a typical reduction of cyt. c was found after mixing cellobiose with CDH in Hy[ch][dhp]. This means that the electron was transferred from cellobiose to cyt. c successfully in Hy[ch][dhp] via two domains of CDH [15]. Next, we investigated the immobilization of CDH on an electrode and analyzed the enzymatic reaction by the electrochemical method. The working electrode was prepared by using a carbon nanoparticle on a carbon paper to increase the surface area. After drying, the CDH solution was casted gently on the constructed carbon electrode. The electron transfer reaction was investigated in the Hy[ch][dhp] by adding the cellobiose solution (final water concentration was up to 35 wt%). When cellobiose solution was added to Hy[ch] [dhp], catalytic currents were observed at a scan rate of 0.2 mV/s. No catalytic reaction was observed when water was added. These results suggest that CDH electrochemically catalyzed the oxidation of cellobiose through direct electron transfer between the active site of CDH (flavin and heme) and the electrode in Hy[ch][dhp]. The catalytic oxidative current began to increase and reached maximum at +400 mV versus Ag/AgCl. This result agrees well with that in buffer solution. However, because the observed current was small, the amperometric investigation was carried out to evaluate the reaction in detail. The current change was measured under an applying potential of +400 mV. The catalytic current increased when the cellobiose solution was added. However, no current change was observed by water addition. When cellobiose solution was added several times, a stair-like current increase reflected the cellobiose concentration (Figure 13.8). A long-term stability of bioelectroactivity of CDH immobilized

c13.indd 199

1/20/2011 10:28:03 AM

200

REDOX REACTION OF PROTEINS

-1.30

Current /µA

-1.35

-1.40

-1.45

-1.50 0

500 1000 Time (sec)

1500

Figure 13.8. Current change of CDH-modified carbon electrode in Hy[ch][dhp] applying 400 mV; 63 μM cellobiose solution was added at arrow point.

on the electrode also has been examined. The solution became turbid after 1 week, and no catalytic current was observed when a CDH-immobilized carbon electrode was soaked in a buffer solution. The immobilized CDH was denatured during storage in buffer solution and led to a loss of the activity. However, when CDH-immobilized electrode was stored in Hy[ch][dhp], the solution kept clearness, no turbidity was noticed, and the catalytic current was observed continuously. Furthermore, it was revealed that CDH immobilized on the electrode was still active after preservation for 3 weeks at room temperature in Hy[ch][dhp]. It may contribute to proposing a new strategy to construct a high-performance biofuel cell.

REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9.

c13.indd 200

F. A. Armstrong, H. A. O. Hill, N. J. Walton, Acc. Chem. Res., 21, 407 (1988). P. Yeh, T. Kuwana, Chem. Lett., 1145 (1977). K. Niki, T. Yagi, H. Inokuchi, K. Kimura, J. Am. Chem. Soc., 101, 3335 (1979). D. L. Compton, J. A. Laszlo, J. Electroanal. Chem., 520, 71 (2002). J. D. Wadhawan, A. J. Wain, A. N. Kirkham, D. J. Walton, B. Wood, R. R. France, S. D. Bull, R. G. Compton, J. Am. Chem. Soc., 125, 11418 (2003). T. Welton, Chem. Rev., 99, 2071 (1999). P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed., 39, 3772 (2000). C. Lagrost, D. Carrie¯, M. Vaultier, P. Hapiot, J. Phys. Chem. A, 107, 745 (2003). H. Ohno, C. Suzuki, K. Fukumoto, M. Yoshizawa, K. Fujita, Chem. Lett., 32, 450 (2003).

1/20/2011 10:28:03 AM

REFERENCES

201

10. 11. 12. 13. 14.

K. Fujita, H. Ohno, Polymer. Adv. Technol., 14, 486 (2003). N. Matsuda, A. Takatsu, K. Kato, Y. Shigesato, Chem. Lett., 125 (1998). J. T. Bradshaw, S. B. Mendes, S. S. Saavedra, Anal. Chem., 74, 1751 (2002). K. Fujita, C. Suzuki, H. Ohno, Electrochem. Comm., 5, 47 (2003). (a) H. Ohno, N. Yamaguchi, Bioconjugate Chem., 5, 379 (1994). (b) F. Kurusu, H. Ohno, Electrochim. Acta, 45, 2911 (2000). 15. (a) K. Fujita, D. R. MacFarlane, Forsyth, M. Chem. Comm., 4804 (2005). (b) K. Fujita, D. R. MacFarlane, M. Forsyth, M. Yoshizawa-Fujita, K. Murata, N. Nakamura, H. Ohno, Biomacromolecules, 8, 2080 (2007).

c13.indd 201

1/20/2011 10:28:03 AM

PART III IONIC DEVICES

c14.indd 203

1/20/2011 10:28:04 AM

14 LI BATTERIES Hikari Sakaebe and Hajime Matsumoto

14.1

INTRODUCTION

“Ionic liquid”(IL) or, classically, “room temperature molten salt” is an interesting series of materials as a novel electrolyte system for electrochemical devices. This contains anions and cations being in a liquid state around room temperature without any solvent. The combination of anionic and cationic species gives a lot of variations in the properties of ionic liquid (e.g., viscosity, conductivity, electrochemical stability and so on) which enables the desirable design of the ionic liquid. Moreover, because of the nonvolatile and flame-resistant nature, this kind of material seems especially interesting for the lithium-ion batteries, the thermal stability improvement of which has been investigated for a long time. In this chapter, several efforts that have been made for battery application will be introduced.

14.2 SAFETY ASPECTS OF LI-ION BATTERY CONCERNING THE ADVANTAGE OF USE OF IONIC LIQUIDS The exothermic reaction at the first stage of the thermal runaway of the Li-ion battery with organic electrolyte, involving the electrolyte decomposition and the electrolyte–electrode reaction, accelerates the continued heat evolution [1]. The organic electrolyte usually contains volatile solvents (e.g., propylenecarbonate, dimethylcarbonate, dimethoxyethane, and so on). These solvents would push up the inner pressure of the battery in increasing temperature,

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

205

c14.indd 205

1/20/2011 10:28:04 AM

206

LI BATTERIES

which may lead to the explosion of the battery. Once a battery was opened to the air, these organic electrolytes even can work as a fuel because of their flammability. If the battery could stay unopened in elevated temperature, then the positive electrode material at the charged state decomposes around 200 °C to provide oxygen inside of the battery with an organic solvent. In any case, it seems essentially important to replace a volatile and flammable electrolyte with different systems. For this purpose, a solid-state electrolyte is ultimately advantageous. However, the manufacturing of the solid-state battery requires a big breakthrough as maintaining close contact between the electrode and the electrolye solid throughout the charge-discharge process is difficult and the electrolyte conductivity is still unsufficient, although great efforts have been made. It would take a long time to commercialize the solid-state battery. However, a certain liquid that is nonvolatile and flame-resistant, (i.e., ILs) is a promising candidate for the electrolyte of a battery with more safety because of its interesting properties it. The basic and general properties of ILs are not discussed here in detail, but some interesting properties for battery application are provided in the following list. 1. Showing a liquid nature in wide temperature range including ambient temperature 2. Thermally stable, nonvolatile, flame-resistant 3. Wide electrochemical window 4. Higher conductivity among the novel electrolyte 5. Chemically stable 14.3 SOME EXAMPLES FOR APPLICATION TO LI-ION BATTERIES Several kinds of ionic liquids have been investigated for the electrolyte of Li-ion batteries with various ingenuities, depending on the drawbacks and advantages of each ionic liquid. The application techniques that have been reported could be classified roughly into the following categories: 1. IL+ supporting electrolyte + additives 2. IL+ supporting electrolyte + alternative negative electrode with higher potential 3. IL(Neat) + supporting electrolyte Hereafter, IL would be classified into the following categories, and the outline of the reported battery application is summarized from 14.3.1–14.3.3: 1. IL with chloroaluminate anions 2. IL with alkylimidazolium cation-fluorinated anion combination 3. IL with non-imidazolium cation-fluorinated anion combination

c14.indd 206

1/20/2011 10:28:04 AM

SOME EXAMPLES FOR APPLICATION TO LI-ION BATTERIES

14.3.1

207

IL with Chloroaluminate Anions

IL first was reported as a new bath for electrolysis about 50 years ago [2a]. A mixture of organic bromide and aluminum chloride showed the melting point to be lower than temperature, which led to the discovery of the mixture of organic chloride and aluminum chloride that remains liquid in a wide temperature range. This attracted more researchers because of the lower viscosity, higher conductivity among the ILs, and relatively wide electrochemical window. However, this series of IL is extremely sensitive to water and is corrosive to certain kinds of metals. We could find sevaral examples of battery application of chloroaluminatecontaining ILs, owing to the longest history among all ILs. The corrosive nature of this kind of IL sometimes was difficult to apply to the conventional batteries, and these examples include the ideas for novel battery systems, which resulted in batteries such as the chloroaluminate acid-base concentration cell [3], Al-chloride cell [4], Cd-bromide cell [5], Al-polyaniline cell [6], Mg-metal vanadate cell [7] and so on. Application to Li-ion batteries still seems tough mainly because the cathodic stability of the chloroaluminate-containing electrolyte against lithium has not been sufficient. In these days, the use of alternative negative electrodes (e.g., Li–Al alloy-layered oxides cell [8–10], Li–Sn alloy negative electrode [11]) or additives (e.g., SOCl2 [12], H2O [13]) seems to have enhanced the potential of ILs of the chloroaluminate family. A detailed description can be found in the articles by Webber and Blomgren [14], and two recent representative works on the application to lithium batteries are introduced in the following: Li/AlCl3–EMIC–LiCl–SOCl2/Graphite [12] EMIC: 1-ethyl-3-methylimidazolium chloride (IL+ supporting electrolyte + additives) It has been difficult to find literature concerning successful Li+ intercalationdeintercalation into/from graphitized carbon electrode. This seems to be not only because alkylimidazolium-containing melt decomposes at 1 V above Li/ Li+ but also because these melts are thought to have difficulty in forming a solid electrolyte interface (SEI) layer on the carbon electrode surface like when using organic electrolyte. Moreover, imidazolium cation intercalates (probably irreversibly) into the interlayer of carbon [15,16], all of which implies the complicated reactions in the reduction of melts on a graphitized carbon electrode. However, in the presence of SOCl2 in the chloroaluminatecontaining IL, charge-discharge properties of the graphitized carbon electrode could be improved with a discharge capacity of 300 mAh g−1 as long as the electrode was fabricated without a binder (Figure 14.1). The coulombic efficiency of each cycle was more than 90% except for the first cycle of c.a. 40%. Irreversible capacity observed in the initial charge was thought to be used in the cathodic decomposition of SOCl2, which leads to the SEI formation.

c14.indd 207

1/20/2011 10:28:04 AM

208

LI BATTERIES 3.0

E/V vs Li/Li+

2.5 2.0

1st

10th

2nd

20th

5th

30th

1.5 1.0 0.5 0.0 0

100

200

300

400

500

600

700

Capacity/mAh×g-1

Figure 14.1. Reduction/oxidation cycles obtained with the binder-free graphite electrode (artificial graphite, KS-25, 1.8 mg on Mo) in the Li-Cl saturated AlCl3(60 mol%)EMIC(40 mol%) melt containing 0.11 mol dm−3 SOCl2. First cycle: at 0.5C (0.28 mA cm−2) and rest of the cycles: at 0.2C (after reference 12).

This group recently has fabricated another binder-free electrode by the electrophoretic deposition, and the electrode showed higher efficiency in all across the cycles [17]. Li–Al/MeEtImCl–AlCl3–LiCl–C6H5SO2Cl/LiCoO2 [9] MeEtImCl:1-methyl-3-ethylimidazolium chloride (IL+ supporting electrolyte + alternative negative electrode with higher potential, also with additives) Li-Al whose electrode potential is ∼0.3 V higher than Li/Li+ was applied as a negative electrode in a melt MeEtImCl–AlCl3–LiCl containing C6H5SO2Cl as an additive. Lithium metal deposited on Al has been left shiny thanks to the additive’s effect, and the authors concluded that Li–Al would have been stably used in this cell configuration. The charge–discharge curves of the cell at the first cycle is shown in Figure 14.2. During the first charge, a plateau around 3.5 V and an excess charge capacity was observed. The succeeding discharge curve has lost the plateau with a coulombic efficiency of less than 50%. Similar to the example of Li/AlCl3–EMIC–LiCl–SOCl2/Graphite [12], the irreversible capacity at the initial charge could be used to form the SEI layer on the surface of the negative electrode. 14.3.2

IL with Alkylimidazolium Cation–Fluorinated Anion Combination

A certain kind of fluorinated anion formed an IL with a melting point lower than room temperature and sufficient anodic stability. Moreover, the sensitiv-

c14.indd 208

1/20/2011 10:28:04 AM

209

SOME EXAMPLES FOR APPLICATION TO LI-ION BATTERIES 5.0 4.5

Charging

Discharging

Cell voltage (V)

4.0 3.5 3.0 2.5 2.0 1.5 50

0

100

150

200

250

300

400

350

Capacity (mAh)

Current density/mA cm–2

Figure 14.2. The variation of the cell voltage of the LiAl/LiCoO2 battery using the RTMS as the electrolyte media with the addition of C6H5SOCl2 (0.05 mol kg−1) at the first cycle. Positive electrode: LiCoO2 85 wt%, graphite 10 wt%, Teflon 5 wt.%. RTMS: MeEtImCl/AlCl3/LiCl = 1.0/1.2/0.15. Current density: 1.0 mA cm−2 (after reference 9).

O

10

N+

N+

N

0 anodic decomposition of TFSI N −N

+

−10

+

N

BP

N

SO2CF3 SO2CF3

EMI

−4.0

−3.0

−2.0

−1.0

0.0

1.0

2.0

3.0

4.0

Potential/V vs l−/l3−

Figure 14.3. Linear sweep voltammogram of glassy carbon in various ionic liquid with TFSI anion at 25 °C. sweep rate: 50 mv/sc (after reference 18).

ity against water decreased [2b]. The imide anion greatly improved the chemical stability in the air, which enabled the easy handling of ILs. An EMI cation is prevailing through the IL of nonchloroaluminate family because usually it provides the wieldy electrolyte with a lower viscosity. Concerning the electrochemical stability, an EMI-containing IL generally falls short of cathodic stability. As shown in Figure 14.3 [18], cathodic decomposition of the neat IL breaks out in a positive potential region against Li/Li+, meaning the

c14.indd 209

1/20/2011 10:28:04 AM

210

LI BATTERIES

Cell voltage / [V]

3.0 2.5 2.0 1.5

1st cycle 2nd cycle 3rd cycle

1.0 0.5 0.0

0

20

40

60

(2) LiEMIBF4 80

100

120

140

160

180

Specific capacity of Li[Li1/3Ti5/3]O4/[mAhg−1]

Figure 14.4. The discharge curves of the cell Li4Ti5O12/EMIBF4-LiBF4/LiCoO2 (after reference 19).

application to Li battery systems essentially would be difficult. As is the case with chloroaluminate-containing IL, electrolyte design or negative electrode material needs to be considered for lithium battery applications. One example for a successful attempt at a battery with a good cycleability and long life is outlined in the following: Li4Ti5O12/EMIBF4–LiBF4/LiCoO2 [19] EMIBF4: 1-ethyl-3-methylimidazolium tetrafluoroborate (IL+ supporting electrolyte + alternative anode with higher potential) One effective solution for the application of EMI-containing IL to a lithium battery system could be the use of an alternative negative electrode material. Several kinds of Li alloy and metal oxide work over 0.2–2.0 V versus Li/Li+ and help to overcome the drawbacks of cathodic stability. In this case, a spinel phase of Li4Ti5O12 working at 1.5 V versus Li/Li+ was used as a negative electrode. The structural change of Li4Ti5O12 was studied intensively by Ohzuku et al., and Li4Ti5O12 was called “zero-strain insertion material,” as an almost negligible volume change of the unit cell of the crystal structure was observed during the charge–discharge process [20]. These crystallographic properties could prevent the electrode composite from the degradation in the cycling and Li4Ti5O12 makes the battery cycle with a long life. Figure 14.4 shows the initial charge–discharge curves of the Li4Ti5O12/EMIBF4–LiBF4/LiCoO2 cell. The cell voltage that was decreased around 2 V which reduced the energy density of the battery; however, the capacity decay was extremely small after 50 cycles. 14.3.3

IL with Nonimidazolium Cation-Fluorinated Anion Combination

It has been difficult to find the “neat” ILs that worked in Li batteries among the ILs introduced in both 14.3.1 and 14.3.2. “Neat” in this chapter is defined

c14.indd 210

1/20/2011 10:28:04 AM

211

SOME EXAMPLES FOR APPLICATION TO LI-ION BATTERIES 4.5 4.3

Voltage (V)

4.1

0.66 mAh

l = 0.02mA Charge

Discharge

3.9 3.7

ocv

ccv

3.5 3.3 3.1 2.9 2.7 2.5 25

35

45 Time (hrs)

55

Figure 14.5. Charge/discharge curves for the cell Li/EMPBF4+0.8 molar LiAsF6/ LiMn2O4 at room temperature (after reference 22).

as the state of ILs containing only a supporting electrolyte and without any additive. The nonimidazolium system, quaternary asymmetric ammonium system, and pyrazolium system could be used as examples of a “neat” application of IL to Li batteries. Li/DMFP(or EMP)BF4-LiBF4 or LiAsF6/LiMn2O4 [21, 22] DMFPBF4: 1,2-dimethyl-4-fluoropyrazolium tetrafluoroborate, EMPBF4:1-ethyl-2methylpyrazolium tetrafluoroborate. DMFPBF4 and EMPBF4 were known to have an excellent thermal stability, and then they seem to have potential as an electrolyte of batteries with more safety. However, cathodic decomposition of these salts was set in 1.0–1.3 V above Li/Li+. When using the electrolyte of DMFPBF4–LiBF4, 75% of the theoretical capacity was obtained in the initial cycles, and the capacity was quickly faded. When replacing LiBF4 with LiAsF6, use was low (c.a. 30%), but coulombic efficiency was improved to 100%. The authors concluded that the SEI layer on the Li metal would have been improved to have a good conductivity by containing arsenate compounds [21]. Figure 14.5 is a charge–discharge curve of a Li/ EMPBF4–LiAsF6/LiMn2O4 cell in the way of the cycling. In this cell configuration, coulombic efficiency seems close to 100%, but the use of the cathode was less than 60% [22]. The authors suggested that this could occur because the cell resistance was too large with a high resistance of electrolyte and resistive product on Li. Li/LiTFSI + PP13-TFSI/LiCoO2 [23] (PP13:N-methyl-N-propylpiperidinium, Fig. 14.6a)

c14.indd 211

1/20/2011 10:28:04 AM

212

LI BATTERIES (a)

(b)

N+

(c)

N+

CF3

O

O CF3 S S N− O O

Figure 14.6. Typical cations and anion for ionic liquid with annmonium cation: (a) N-methyl-N-propylpiperidinium cation (PP13), (b) trimethylpropylammonium cation (TMPA), and (c) bis(trifluoromethane sulfonyl)imide anion (TFSI).

The authors believed that using a simple composition in battery materials as long as possible could be an important factor to increasing the reliability of the battery. So the “neat” IL without any volatile additives has been of interest. In the past, a certain kind of ammonium cation was known to form the IL, and Matsumoto, found the novel IL with aliphatic and asymmetric quaternary ammonium cations. The series of IL was characterized as the most stable cathodically, and the reversible plating/stripping of the Li metal was observed in that media [24]. In Figure 14.3, linear sweep voltammograms of several classes of the IL with a bis(trifluoromethane sulfonyl)imide anion (TFSI, Figure 14.6c) were shown, and trimethylpropylammonium cation (TMPA)TFSI was stable enough around the Li/Li+ potential. When including EMI– TFSI, several ILs were applied to the electrolyte of the Li/LiCoO2 cell, and the charge–discharge properties were evaluated [23,25]. As a result, it could be concluded that sufficient cathodic stability should be the minimum requirement for the acceptable battery performance. In addition, it was realized that PP13-TFSI, which is one of the most stable ILs cathodically, showed the excellent performance as an electrolyte in a Li/LiCoO2 cell [23,25]. Figure 14.7 shows the charge–discharge curves of the Li/LiTFSI + PP13–TFSI/LiCoO2 cell. The obtained capacity was close to the theoretical capacity (Figure 14.7a), and the coulombic efficiency of each cycle was more than 97% for 30 cycles (Figure 14.7b). For the first cycle, the efficiency was a little bit lower (93%) which implied that no significant decomposition occurred. The physical properties of the ILs studied in this work are listed in Table 14.1 [13,18,25,26]. An ammonium series of ILs are less advantageous in kinetic properties than EMI-containing ILs. The conductivity of PP13-TFSI was 1.5 ms cm−1 at 25 °C, and it decreased to 0.5 ms cm−1 when 0.4 mol dm−3 LiTFSI was dissolved in PP13–TFSI. Continued investigation seems necessary for the commercialization of an ammonium system through improving the conductivity and viscosity. Although it has a drawback in conductivity, a thinner electrode enables the cycling at the higher rate. Figure 14.8 is the result of a charge–discharge cycling test of Li/LiTFSI + PP13–TFSI/LiCoO2 cell at 0.5C current rate. The electrode loading was c.a. 1 mg for the cell shown in Figures 14.7 and 14.8. The retention of the capacity at 0.5C against the one at 0.1C ([fisrt discharge capacity at 0.5C] × 100/[first discharge capacity at 0.1C]) was 80%. The cell construction affects the cycling test results, and it is extremely

c14.indd 212

1/20/2011 10:28:05 AM

213

SOME EXAMPLES FOR APPLICATION TO LI-ION BATTERIES (a) 4.5 15

Cell Voltage / V

4.2

5 2 1

3.9 3.6

Li/LiTFSI in PP13-TFSI/LiCoO2

3.3

0.1C 3.2 - 4.2 V(CC mode)

1 15 5

2

3 20

0

40

60

80

100

120

140

160

Specific Capacity / mAh g –1 (b)

Coulombic Efficiency / %

100 95 90 85 80 75

Li/LiTFSI in PP13-TFSI/LiCoO2

70

0.1C 3.2 - 4.2 V(CC mode)

65 60

1

3

5

7

9

11

13

15

17

19

21

23

25

27

Cycle Number

Figure 14.7. Galvanostatic charge-discharge curves for initial 25 cycles (a) and cycle properties (b) of Li/0.4 mol dm−3 LiTFSI in PP13-TFSI/LiCoO2 cell at C/10 current rate, 3.2–4.2 V (after reference 25).

TABLE 14.1. Physical Properties of the ILs ILs

EMI-TFSI TMPA-TFSI P13-TFSI PP13-TFSI

c14.indd 213

Melting Point (°C)

Density at 20 °C (g cm−1)

Viscosity at 25 °C (mPa s)

Conductivity at 25 °C (mS cm−1)

References

−15 22 12 8.7

1.52 1.44 1.45 —

34 72 63 117

10.8 3.27 1.40 1.51

13 18 26 25

1/20/2011 10:28:05 AM

214

LI BATTERIES 4.5

40

Specific Capacity / mAh g –1

50

30 20

10

4.2 1 3.9 3.6 Li/LiTFSI in PP13-TFSI/LiCoO 2

3.3

0.5C 3.2 - 4.2 V(CC mode)

3 0

20

40

60

80

100

120

Cell Voltage / V

Figure 14.8. Galvanostatic charge-discharge curves for initial 50 cycles of Li/0.4 mol dm−3 LiTFSI in PP13-TFSI/LiCoO2 cell at C/2 current rate, 3.2–4.2 V (after reference 25).

difficult to compare the absolute value of the results between the PP13–TFSI and other ILs; however, the cell using PP13–TFSI as an electrolyte base showed the well-balanced performance in cell voltage, capacity use, coulombic efficiency of the each cycles, cycling properties, and so on. It is known that the Al current collector undergoes corrosion in the organic electrolyte containing a LiTFSI supporting electrolyte [27,28]. The TFSI anion is rather stable, releasing little free fluoride anions. When using LiPF6 instead, more F− exists in the organic electrolyte, which helps to form the protective surface film on Al. However, it could be found in the literature that suggested that Al would have dissolved into an organic electrolyte as a species [Al(TFSI)x]3−x [28]. Several TFSI anions are contained in the PP13–TFSI electrolyte, and there was concern that the Al current collector also would be corroded in the IL electrolyte. Then, the electrochemical measurement was carried out with Li/0.4 mol dm−3 LiTFSI + PP13–TFSI/Al cell and Li/0.4 mol dm−3 LiTFSI + EC–DMC/Al cell (EC: ethylenecarbonate, DMC: dimethylcarbonate), and the surface morphology was compared [29]. A cyclic voltammogram between Open circuit voltage (OCV) and 5.5 V showed that the current density passed through the cell containing IL was two-figures smaller than that containing EC–DMC. This result indicated that the Al corrosion in an ILcontaining electrolyte would not be significant, and an SEM image also supported this result. On the Al oxidized in organic electrolyte (Figure 14.9a), a lot of pit was formed as a result of the corrosion, whereas the corrosion seems not to have taken place on Al oxidized in PP13–TFSI electrolyte. In the severe condition applied (e.g., oxidation to 5.5 V and storage for 1 month at 60 °C), oxidation to 6.5 V (Figure 14.9b) no pit was observed on the Al surface [29]. The Al current collector, which is cheap, easy-handling, light-weighted, and stable in a higher voltage region, is a key component for the commercial-

c14.indd 214

1/20/2011 10:28:05 AM

SOME EXAMPLES FOR APPLICATION TO LI-ION BATTERIES

215

(a)

(b)

Figure 14.9. Scanning electron micrographs for the surface of Al oxydized (a) up to 5.5 V in 0.4 mol dm−3 LiTFSI/EC-DMC and (b) up to 6.5 V in 0.4 mol dm−3 LiTFSI/PP13TFSI. Al surface was scratched in Ar atmosphere by emery paper before electrochemical reaction (after reference 29).

ization of the Li-ion battery, and study on the corrosion in ILs should be continued. The availability of ILs seems to depend on the kinetic nature and on the electrode compatibility. Several electrode materials have been cycled in PP13– TFSI–based electrolyte. Concerning the carbon electrode, which was in practical use, it is difficult to obtain the full capacity using PP13–TFSI-based “neat” electrolyte. For example, the charge–discharge properties of MCMB2800 are shown in Figure 14.10 [30]. Constant current (CC) charging provided only 10% capacity of the theoretical one. Constant current–constant voltage (CCCV) charging at 5 mV increased the capacity to 20% of the theoretical capacity.

c14.indd 215

1/20/2011 10:28:05 AM

216

LI BATTERIES 3 MCMB in 0.4M LiTFSI/PP13-TFSI

Cell Voltage / V

2.5

1st

2 5-10th

disch.

1.5

12th discharge (charged by CCCV mode)

1 0.5

1st charge 0 0

50

100

150

200

Specific Capacity/mAh g–1

Figure 14.10. Galvanostatic charge-discharge curves for initial 12 cycles of Li/0.4 mol dm−3 LiTFSI in PP13-TFSI/ MCMB2800 cell at C/10 current rate (CC charge for initial 11 cycles and CCCV charge at 5 mV for the 12th cycle), 0–1.5 V (after reference 30).

This is partly because the inner resistance of the cell was high (∼20 Ω), and the wettability of the electrode against the IL also seems responsible. Moreover, it is implied that the SEI favorable for Li+ intercalation could not be formed properly. However, so-called hard carbons showed more capacity (more than 50% of the capacity obtained in organic electrolyte). These results tell us that the exploration of the alternative electrode material will lead us to the suitable cell configuration using ILs.

14.4

RECENT TOPICS AFTER THE FIRST EDITION PUBLISHED

After the publication of the first edition of this book, remarkable progress has been made on the research and development of ionic liquids for lithium batteries. One piece of important progress was that the battery performance using neat ionic liquids was dramatically improved through the research and development of new anionic species for ionic liquids. The other development was that the thermal stability assessments of ionic liquids coexisting with lithium battery components such as a cathode and/or an anode material at the charged state have just started. In the last section of this chapter, these recent topics after the first edition was published are briefly described. 14.4.1 Improvement of a Lithium Battery Performance with Neat Ionic Liquids Containing New Perfluoroanion From the earlier reports on ILs based on BF4− [19] and TFSA−, [25,31] the rate performance of a lithium battery using ILs as an electrolyte media could not

c14.indd 216

1/20/2011 10:28:05 AM

RECENT TOPICS AFTER THE FIRST EDITION PUBLISHED

217

compare with those using conventional organic solvents. This situation was totally changed by the first reports on new ILs based on FSA− ([(FSO2)2N]−) [32]. The rate property of a Li/LiCoO2 using FSA-ILs was 10 times faster than that of TFSA-ILs. One important point of this result was that the viscosity of FSA-ILs were only about half of the corresponding TFSA-ILs, which suggests that the transport property of Li+ in bulk liquid phase is not the only factor to control the rate property. Recent results on the anion dependence of the rate property of Li/LiCoO2 using N,N-diethyl-N-methyl-N-(2-methoxymethyl) ammonium (DEME+) clearly indicated that the rate property of Li/LiCoO2 depends on the charge transfer resistance on a cathode and an anode, which reflected the electrochemical kinetics on the Li+ intercalation/deintercalation process and Li metal deposition/dissolution process, respectively. The interaction between Li+ and an anion must affect such electrochemical processes [33]. However, other unique properties were observed in FSA-ILs. The Li metal plating and stripping observed [32] even in 1-ethyl-3-methylimidazolium (C2mim+)[FSA], and carbon anodes could be operable in FSA-ILs [34] without any additives. The reason why Li+ intercalates into graphite in FSA-ILs is still unclear. However, it must be noted that this recent progress has been brought through the discovery of new ionic liquids containing new anionic species. 14.4.2 Thermal Stability Assessment of Ionic Liquids in a Battery Environment One of the reasons for investigating ionic liquids as a lithium battery electrolyte is their good thermal stability and less volatility compared with conventional organic liquids. However, it seems important to know that the ionic liquids are also stable in a lithium battery system, especially at a high temperature when the thermal runaway process in a lithium battery was started. There are few reports on the thermal stability assessment of ionic liquids in a lithium battery environment adopting an accelerating rate calorimetry (ARC) [35] and a differential scanning calorimetry (DSC) [36,37]. In many cases, a pressure-tight sealed pan containing target ionic liquids as well as cathode and/ or anode active material were tested instead of an actual lithium battery such as an 18,650 cell. These results indicate that the thermal stability of ionic liquids depends on the kind of active materials that were coexisting with the target ionic liquids in the same sample pan. The thermal stability of typical ionic liquids containing TFSA, BF4, and PF6 was almost the same as in conventional organic solvents when a graphite and delithiated LiNi1/ 3Mn1/3Co1/3O2 coexisted [37]. A SEI film was formed in precycling in a conventional electrolyte in this case, and the thermal decomposition of SEI may affect the experimental results. However, TFSA-ILs exhibited a distinctive thermal stability compared with conventional organic solvents when delithiated LiCoO2 and lithium metal coexisted [36]. In case of other anode materials such as a silicon and a lithium titanate, FSA-ILs did not show sufficient thermal stability; however, TFSA-ILs (except for EMI[TFSA])

c14.indd 217

1/20/2011 10:28:05 AM

218

LI BATTERIES

exhibited better thermal stability compared with a conventional organic solvents. Considering these results, the stability of ionic liquids in a current lithium battery system containing a graphite anode might not be expected to be a thermally stable electrolyte. However, as shown in the thermal stability of ionic liquids for new active materials such as a silicon and a lithium titanate, ionic liquids will have the chance to be used as a key material for constructing safe lithium battery systems by combining new cathode and/or an anode materials.

14.5

SUMMARY AND FUTURE VIEW

Through many works for the exploitation of new IL materials, new systems have been found, and the cell performance with ILs has become closer to or better than that with conventional organic electrolyte. This makes cell fabrication easier as a ready-made electrode could be used without modification and more researchers are attracted. The kinetic properties of asymmetric quaternary ammonium-based IL especially has been improved, with their cathodic stability remaining almost unchanged. It seems worthwhile to investigate them as an electrolyte for Li-ion batteries because of the thermal stability in the battery environment. At the same time, including other classes of ILs seems “more” important to make the best use of them by tuning them with the additives and battery system for practical use. For this achievement, a lot of work should be done first to obtain the general and unified view of the basic behavior of ILs. In sum, it can be said that these interesting findings have brought ILs closer to the practical application for the battery, attracting a lot of researcher interest. At this stage, we have a sign to balance the electrochemical stability of IL as well as the conductivity compatible to the organic electrolyte. Cations and anions, other than those introduced here, with different structures are now being reported. And the modification of IL properties with combinations of those ions makes more every progress day. We will soon find a favorable way to use the IL-containing battery and will optimize the battery configuration for it to popularize the “ionic-liquid batteries.”

REFERENCES 1. J. Yamaki, Y. Baba, N. Katayama, H. Takatsuji, M. Egashira, S. Okada, J. Power. Sources, 119–121, 789 (2003). 2. (a)M. Matsunaga, Electrochemistry, 70, 126 (2002). (b)R. Hagiwara, Electrochemistry, 70, 130 (2002). 3. C. J. Dymek, Jr., J. L. Williams, D. J. Groeger, J. Electrochem. Soc., 131, 2887 (1984). 4. P. R. Gifford, J. B. Palmisano, J. Electrochem. Soc., 135, 650 (1988). 5. C. J. Dymek, G. F. Reynolds, J. S. Wilkes, J. Electrochem. Soc., 134, 1658 (1987).

c14.indd 218

1/20/2011 10:28:05 AM

REFERENCES

6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19.

20. 21. 22.

23. 24. 25. 26. 27. 28. 29. 30.

31.

c14.indd 219

219

N. Koura, H. Ejiri, K. Takeishi, J. Electrochem. Soc., 140, 602 (1993). P. Novak, W. Scheifele, F. Joho, O. Haas, J. Electrochem. Soc., 142, 2544 (1995). J. Dvynck, R. Messina, J. Pingarron, J. Electrochem. Soc., 131, 2274 (1984). Y. S. Fung, R. Q. Zhou, J. Power. Sources 81, 891 (1999). K. Ui, N. Koura, Y. Idemoto, K. Iizuka, Denki Kagaku., 65, 161 (1997). Y. S. Fung, D. R. Zhu, J. Electrochem. Soc., 149, A319 (2002). N. Koura, K. Etoh, Y. Idemoto, F. Matsumoto, Chem. Lett., 12, 1320 (2001). J. Fuller, R. T. Carlin, R. A. Osteryoung, J. Electrochem. Soc., 144, 3881 (1997). A. Webber, G. E. Blomgren, Advances in Lithium-Ion Battery, W. van Schlkwijk and B. Scrosati, eds., Kluwer Academic/Plenum Publishers, 2002, p.185. R. T. Carlin, J. Fuller, W. K. Kuhn, M. J. Lysaght, P. C. Trulove, J. Appl. Electrochem., 26, 1147 (1996). T. E. Sutto, D. M. Fox, P. C. Trulove, H. C. De Long, Meeting Abstract of 203rd ECS Meeting, vol 2003-1, Paris, 1161, 2003. K. Ui, T. Minami, Y. Idemoto, N. Koura, Meeting Abstract of the 44th Battery Symp. in Japan, 3D08, 2003. H. Matsumoto, M. Yanagida, K. Tanimoto, T. Kojima, Y. Tamiya, Y. Miyazaki, Molten Salts XII, P. C. Trulove et al., eds. Electrochem. Soc., 2000, p. 186. (a) H. Nakagawa, S. Izuchi, S. Sano, K. Takeuchi, K. Yamamoto, H. Arai, Meeting Abstract of the 41th Battery Symp. in Japan, 3C18, 2000. (b) H. Nakagawa, S. Izuchi, K. Kuwana, T. Nukuda, Y. Aihara, J. Electrochem. Soc., 150, A695 (2003). T. Ohzuku, A. Ueda, N. Yamamoto, J. Electrochem. Soc., 142, 1430 (1995). J. Caja, T. D. Dunstan, D. M. Ryan, V. Katovic, Molten Salts XII, P. C. Trulove et al., eds., Electrochem. Soc., 2000, p. 150. J. Caja, T. D. Dunstan, V. Katovic, Molten Salts XIII, H. C. Delong, R. W. Bradshaw, M. Matsunaga, G. R. Stafford, and P. C. Trulove, eds., Electrochem. Soc., 2002, p. 1014. H. Sakaebe, H. Matsumoto, H. Kobayashi, Y. Miyazaki, Meeting Abstract of the 42th Battery Symp. in Japan, 2B04, 2001. H. Matsumoto, M. Yanagida, K. Tanimoto, M. Nomura, Y. Kitagawa, Y. Miyazaki, Chem. Lett., 8, 922 (2000). H. Sakaebe, H. Matsumoto, Electrochem. Commun., 5, 594 (2003). D. R. MacFarlane, P. Meakin, J. Sun, N. Amini, M. Forsyth, J. Phys. Chem. B, 103, 4164 (1999). K. Kanamura, T. Umegaki, S. Shiraishi, M. Ohashi, Z. Takehara, J. Electrochem. Soc., 149, A185 (2002). X. Wang, E. Yasukawa, S. Mori, Electrochim. Acta, 45, 2677 (2000). H. Sakaebe, H. Matsumoto, Meeting Abstract of the 43th Battery Symp. in Japan, 1B16, 2002. H. Sakaebe, H. Matsumoto, K. Tatsumi, in New Trends in Intercalation Compounds for Energy Storage and Conversion, C. Julien, ed., The Electrochemical Society Proceedings Series PV2003-20, Pennington, NJ, P.57 (2003). B. Garcia, S. Lavallee, G. Perron, C. Michot, M. Armand, Electrochim. Acta, 49, 4583 (2004).

1/20/2011 10:28:05 AM

220

LI BATTERIES

32. H. Matsumoto, H. Sakaebe, K. Tatsumi, M. Kikuta, E. Ishiko, M. Kono, J. Power Sources, 160, 1308 (2006). 33. H. Matsumoto, H. Sakaebe, K. Tatsumi, ECS Trans., 16(35), 59 (2009). 34. M. Ishikawa, T. Sugimoto, M. Kikuta, E. Ishiko, M. Kono, J. Power Sources, 162, 658 (2006). 35. Y. Wang, K. Zaghib, A. Guerfi, F. F. C. Bazito, R. M. Torresi, J. R. Dahn, Electrochim. Acta, 52, 6346 (2007). 36. H. Sakaebe, H. Matsumoto, K. Tatsumi, Electrochim. Acta, 53, 1048 (2007). 37. M. Ue, H. Tokuda, T. Kawai, M. Yanagidate, Y. Otake, ECS Trans., 16 (35) 171 (2009).

c14.indd 220

1/20/2011 10:28:05 AM

15 PHOTOELECTROCHEMICAL CELLS Hajime Matsumoto

15.1

INTRODUCTION

Photoelectrochemical solar energy conversions have been intensively studied since the discovery of the Honda–Fujishima effect [1]. Especially, the solar energy conversion to electricity with the use of a photoelectrochemical (PEC) cell has attracted much attention because of the possibility of reducing the fabrication cost compared with conventional solid-state semiconductor devices. The PEC cells investigated so far are divided into two systems based on the kind of adopted photoelectrode. One is the system that uses the n- or p-type semiconductor electrode as the photoanode or photocathode [2], respectively (Figure 15.1, right). The other is the system that uses a dye-loaded semiconductor (Figure 15.1, left) [3]. The reason for the cost reduction of these PEC cells compared with solid-state solar cells is that a much higher quality of semiconductor material is not needed for their construction [4]. For solid-state systems, both p-type and n-type semiconductors are needed to form a schottky-junction, which is necessary to separate the photogenerated electrons and holes. Both types of semiconductors are prepared from highly pure semiconductors with a severely controlled doping process. However, for PEC cells, such a junction can be achieved with a semiconductor electrode immersed in an electrolyte with a redox compound [4]. Though PEC cells possibily can reduce the fabrication costs, problems commonly encountered in a practical application of electrochemical devices, which are derived from the use of volatile solvents, must be resolved. The use of a liquid electrolyte containing organic solvents seems essential for such devices

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

221

c15.indd 221

1/20/2011 10:28:06 AM

222

c15.indd 222

1/20/2011 10:28:06 AM

2

1

ECB

e−

e− Ox

5

Ox e

−

Red Eredox

diffusion

Red

DYE

3

5

e−

4 2

e−

6

7

h+

e−

e−

Red diffusion

Ox

e−

Eredox

Red

VOC = ECB-Eredox 5

Ox

EVB

e

−

ECB

3

5

e−

4

e−

2

e−

Sun Light

Figure 15.1. Schematic illustration of two type of photoelectrochemical cells. (left) a dye sensitized solar cell, (b) a conventional electrochemical solar cell. 1. a dye loaded porous wide gap semiconductor; 2. transparent conducting grass; 3. an electrolyte containing a redox couple; 4. Pt or carbon counter electrode; 5. a sealing material; 6. back contact; 7. a semiconductor electrode (n-type). Photogeneration of electron is occurred at a gray zone.

Sun Light

e−

Energy level

PARAMETERS FOR THE PERFORMANCE EVALUATION OF PEC CELLS

223

not only to achieve fast mobility of the active materials but also to maintain good contact between the electrolyte and an electrode with a high specific surface area. The room temperature ionic liquid (RTIL) is one possible candidate to resolve the problems with the use of conventional solvents because of its unique properties such as nonvolatility and noncombustibility as described in Part III of this book. The major problem with the use of RTILs in electrochemical energy devices is their relatively high viscosity, which limits the mobility of the electrochemically active material in RTILs such as Li+. This problem also might limit the performance of PEC cells using ionic liquids (ILs). Schematic illustrations of two typical PEC systems are shown in Figure 15.1. A redox compound such as ferrocene/ferricinium and iodide/triiodide is contained in an electrolyte to provide carrier transport (Figure 15.1). The mobility of these materials limits the current density output of the PEC cells. PEC cells constructed using a dye-loaded semiconductor electrode, which are called “dye sensitized solar cells” (Figure 15.1 left), have been intensively investigated since the high conversion efficiency of more than 7% achieved in 1991 by Regan and Grätzel [5]. The key for such a high efficiency is the development of a novel dye, which can efficiently absorb sunlight and the mesoporous titanium oxide electrode. This type of cell is called the “Grätzel cell.” The highest efficiency of the systems reaches more than 10% at AM 1.5 [6]. His colleague also discovered moisture-stable ionic liquids based on the bis(trifluormethylsulfonyl)imide anion for the first time [7] as well as a low temperature melting iodide [8]. However, a conventional PEC cell using a semiconductor electrode using the chloroaluminate system was reported in 1980 by Singh et al. [9]. Based on these reports, the problem derived from the relatively highly viscous RTILs compared with the conventional solvents seems to be overcome by using redox hopping or the so-called Grotthus mechanism [8]. In this chapter, PEC systems using RTILs will be introduced, focusing on the relationship between the viscosity of the RTILs and the short-circuit photocurrent to show the possibility of overcoming the problem derived from the viscosity of RTILs in an electrochemical device application and then recent reports on quasi solid-state dye sensitized solar cell (DSSCs) using RTILs.

15.2 PARAMETERS FOR THE PERFORMANCE EVALUATION OF PEC CELLS Figure 15.2 shows a typically observed current–voltage characteristic curve for PEC cells [4]. This figure can be obtained by linear sweep voltammetry using a conventional electrochemical apparatus. Two major parameters to estimate the PEC cell can be obtained as follows: (i) a short-circuit photocurrent, which is denoted as JSC, and (ii) an open-circuit photovoltage, which is denoted as VOC. JSC is generally related to the mobility of the redox couple in an electrolyte. VOC is basically related to the energy difference between the energy level

c15.indd 223

1/20/2011 10:28:07 AM

224

PHOTOELECTROCHEMICAL CELLS 10.0 JSC WMAX[W cm−2]

Photocurrent /mA cm2

8.0

6.0 WMAX FF = JSC n VOC

4.0 η= 2.0

WMAX light intensity

sunlight ca. 0.1 W cm–2 [AM 1.5] 0.0 0.0

0.2

0.4

0.6

VOC 0.8

1.0

Voltage/V

Figure 15.2. Typically observed voltage–current characteristic curve for PEC cell. JSC: a short-circuit photocurrent, VOC: a open-circuit photovoltage, FF: fill factor, η: conversion efficiency. WMAX: a maximum output.

of the semiconductor electrode and the redox potential using a redox couple as shown in Figure 15.2. A fill factor (FF) is reflected in the overall performance of a PEC cell. The conversion efficiency is the most important parameter to judge the practicality of the PEC cells. This review especially focuses on JSC to discuss the relationship between JSC and the viscosity of the RTILs.

15.3 15.3.1

IONIC LIQUIDS USED AS AN IN-VOLATILE SOLVENT N-Butylpyridinium Chloroaluminate Systems

In 1980, Singh et al. reported the first example of a photoelectrochemical cell using an RTIL. They employed a chloroaluminate melt based on the N-butylpyridinium cation (BP) as a solvent and the III-V semiconductors such as n-GaAs [9–12] or n-InP [13] as the photoanode. Ferrocene and ferricinium chloride were used as the redox couple. The chloroaluminate-based RTIL used here by prepared by mixing equimolar amounts of AlCl3 and BP-Cl. The viscosity of the RTIL is 21 mPas at 25 °C [14]. The current–voltage characteristic, which indicates cell performance, of the PEC cell is shown in Figure 15.3 [10]. At that time, the short lifetime of the PEC cells resulting from photocorrosion of the semiconductor electrode was the major problem [15]. Though the performance of the PEC cell using RTILs under 1 sun (100 mW cm−2) shown in Figure 15.3 was not that high (η = 1.7%)

c15.indd 224

1/20/2011 10:28:07 AM

225

IONIC LIQUIDS USED AS AN IN-VOLATILE SOLVENT 5.0

CURRENT DENSITY (mA /cm2)

4.0

3.0

2.0

1.0

0 0

100

200 300 400 500 600 700 VOLTAGE (mV)

Figure 15.3. Current–voltage characteristics of n-GaAs in BP–AlCl4 containing 0.2 M ferrocene and 0.02 M ferricinium chloride. (triangle) single crystal n-GaAs, (circle) polycrystalline n-GaAs. The light intensity in these experiments was 100 mW/cm2 [10]. (Reprinted by permission of the publisher, The Electrochemical Society). TABLE 15.1. Photovoltaic Characteristics of n-GaAs in BP-AlCl4 Ionic Liquid Containing 0.2 M Ferrocene and 20 mM Ferricenium Chloride Under Tungsten-Halogen Lamp Illumination (100 mW cm−2) Viscosity15)/ mPas BP-AlCl4 BP-AlCl4 + benzene (50 vol%) BP-AlCl4 (85 °C)

JSC/mA cm−2

VOC/mV

Fill Factor

Conversion Efficiency/%

20.9 5.0

4.9 6.7

611 625

0.39 0.40

1.2 1.7

<20.9

5.7

395

0.28

0.6

compared with that of organic solvents, the photocorrosion of n-GaAs could be suppressed by using chloroaluminate-based RTILs for 720 hours of irradiation [10]. The addition of benzene (50 vol%), which increased JSC about 1.3 times, indicated that the higher viscosity of the RTIL compared with molecular liquids limited the cell performance, as shown in Table 15.1. However, the current increase was not as high as expected from the decrease in viscosity (21 → 5 mPas). In this case, the chloride anion contained in the chloroaluminate melts became attached to the n-GaAs surface and acted as a surface

c15.indd 225

1/20/2011 10:28:07 AM

226

PHOTOELECTROCHEMICAL CELLS

recombination center, which is the main cause of the decreased conversion efficiency [11]. 15.3.2

Grätzel Cell with an Ionic-Liquid-Based on Iodide

Papageorgiou et al. reported the first example of the RTIL-based dye sensitized solar cell systems (Grätzel cell [16]). They found that 1-hexyl-3methylimidazolium iodide (HMI-I) melts at room temperature [8]. However, the viscosity of HMI-I is high at more than 1000 mPas at room temperature; therefore, JSC of DSSC using HMI-I (0.75 mA cm−2 at the irradiation of 120,000 Lux [= 1 sun]) is much lower than that using an organic solvent (more than 15 mA cm−2 at 1 sun). This indicates that the slow diffusion of the iodide/ triiodide redox couple in HMI-I limits the overall photocurrent. However, as stated subsequently, the JSC of 0.75 mA cm−2 is unexpectedly high considering its high viscosity. They investigated the diffusion coefficient of triiodide in pure HMI-I or in a mixture of HMI-I and relatively low RTIL such as 1-butyl-3methyl-imidazolium triflate or an organic solvent. Through the discussion by applying the Walden product and/or Einstein–Stokes equation to the experimentally obtained diffusion coefficient and the dynamic viscosity, they concluded that a Grotthus-type charge carrier transfer mechanism between triiodide and iodide existed in pure HMI-I, which is the reason for the unexpectedly high JSC even in a highly viscous RTIL. They also showed the longterm stability of the DSSC system using an RTIL as shown in Figure 15.4. More than 2000 hours of exposure in light (AM 1) did not seriously damage the DSSC systems. Furthermore, at less than 5000 lx (= 0.04 sun), the DSSC systems using RTILs continuously operated for more than 230 days. This result first demonstrated that the nonvolatile nature of RTILs indeed could undergo long-term durability for use in electrochemical devices [8]. 15.3.3

Grätzel Cell Containing a Low Viscousity Ionic Liquid

The hydrofluoric acid-based molten salts such as quaternary ammonium cations and F(HF)n− anions have a low viscosity and are used as an electrochemical fluorination bath [17]. Hagiwara et al. reported new RTILs with a low viscosity based on the 1-ethyl-3-methylimidazolium cation (EMI) and F(HF)2.3 anion, whose viscosity is 4.8 mPas at 25 °C) [18,19], and the application in the Grätzel cell [20]. The current–voltage characteristics of DSSC using the low viscous RTILs are shown in Figure 15.5 [20,21]. Though the additives such as t-butylpyridine, which improves the performance of the DSSC cells [20], was not contained for the RTILs, this figure clearly shows that JSC significantly depends on the viscosity of liquid media. Table 15.2 shows the parameters taken from Figure 15.5 and the diffusion coefficient roughly estimated by the cyclic voltammogram of the iodide redox in each liquid. From these results, the JSC clearly decreased with the decreasing diffusion coefficient of the iodide redox in liquid

c15.indd 226

1/20/2011 10:28:07 AM

227

Current /μAcm−2

IONIC LIQUIDS USED AS AN IN-VOLATILE SOLVENT

0 200 400 600 800 1000 1200 1600 1800 2000 2200 2400 600 800

im e/

h

15.0

na

tio nt

10.0

Illu

mi

5.0

0.0 0

200

400 Voltage/mV

Figure 15.4. Current–voltage characteristics of quasi-transparent, nanocrystalline, dyesensitized solar cells operating with molten salts as electrolyte and redox mediator (90% 1-ethyl-3-methylimidazolioumtriflate, 10% 1-hexyl-3-methylimidazoliumidodide, 5 mM I2) measured under 0.020 AM1, irradiation aged under AM1, as a function of illumination time [8]. (Reprinted by permission of the publisher, The Electrochemical Society).

Figure 15.5. Photocurrent–voltage characteristics of a dye-sensitized solar cell. Electrolyte composition: (ionic liquids) 0.9 M of DMHI-I and 30 mM of I2, (acetonitrile) 0.8 M of DMHI-I, 0.1 M LiI, 50 mM I2, 0.2 M t-butylpyridine. Temperature: 25 °C. Area: 1.0 cm2. AM 1.5 (100 mW cm−2) [21]. (Reprinted by permission of the publisher, The Electrochemical Society of Japan).

c15.indd 227

1/20/2011 10:28:07 AM

228

c15.indd 228

1/20/2011 10:28:07 AM

1800 34 4.8 0.4

(4.6) 0.9 0.9 0.9

d

[DMHI-I]a/M 50 30 30 50

[I2]/mM 0.8 0.9 5.8 13.7

Jsc/mAcm−2 — 0.50 0.65 0.70

Voc/V — 0.80 0.56 0.59

Fill Factor

b

1,2-dimethyl-3-hexylimidazoliumiodide. Conversion efficiency. c Diffusion coefficient of triiodide ion calculated from the reduction peak current of cyclic voltammogram. d Concentration of 1-hexyl-3-methylimidazoliumiodide. e The electrolyte contained 0.1 M of LiI, and 0.2 M of t-butylpyridine.

a

HMI-I EMI-TFSI EMI-F(HF)2.3 Acetonitrilee

Viscosity/mPa s — 0.36 2.1 5.7

η/%

0.9 7.6 43 110

D(I3−)c/10−7 cm 2 s−1

TABLE 15.2. Photovoltaic Characteristics of Dye Sensitized Solar Cell in Various Ionic Liquids at 25 °C (AM 1.5, 100 mW cm−2)

8 20 20 20

Ref.

IONIC LIQUIDS USED AS AN IN-VOLATILE SOLVENT

229

media. To improve JSC, which is almost the same as increasing the conversion efficiency, much less viscous RTILs than EMI-F(HF)2.3 must be prepared. However, the preparation of much less viscous RTILs as low as acetonitrile (0.4 mPas at 25 °C) seems difficult considering the strong coulombic interaction between a cation and an anion. 15.3.4 The Relationships Between JSC and Viscosity of Various Liquid Media Figure 15.6 shows the relationship between JSC of the Grätzel cell and the viscosity of various liquid media under the same conditions [21]. JSC of the DSSC systems simply depends on the viscosity of the liquid media. However, as described previously, JSC of the iodide-based RTILs is much higher even for a high viscosity of more than 200 mPas at 25 °C. This figure also indicates that the amount of iodine (I2) influences the improvement of JSC. This implies that the apparent diffusion coefficient of the iodide redox increased with the increasing amount of triiodide even for a high viscosity. Kawano et al. explained in detail the diffusion mechanism of the iodide/ triiodide redox couple by using a microelectrode technique and the Dahms– Ruff equation, which can explain the physical diffusion accompanying an

Figure 15.6. Relationship between Jsc and viscosity of the redox medium containing 0.9 M DMHI-I, 50 mM iodine. (organic solvent). AN: acetonitrile, MPN: methoxypropionitrile, NMO: N-methyloxazolidinone, TMS: sulfolane, (ionic liquid): EMI: 1-ethyl-3methylimidazolium, TMPA: trimethylpropylammonium, THA: tetrahexylammonium, HMI-I: 1-hexyl-3-methyl-imidazolium-iodide, TFSI: bis(trifluoromethyl-sulfonyl)imide [21]. (Reprinted by permission of the publisher, The Electrochemical Society of Japan).

c15.indd 229

1/20/2011 10:28:07 AM

230

PHOTOELECTROCHEMICAL CELLS

exchange reaction [22]. As a result, in the case of high concentrations of iodide and triiodide in a RTIL as the solvent, the diffusion coefficient derived from the exchange reaction expressed by I − + I −3 → I −3 + I − become significant and superior to a simple physical diffusion mechanism. This indicates the existence of a Grotthus-like mechanism, which enhances the apparent diffusion coefficient of the iodide redox in RTILs based on an iodide. From the result of the iodide-based RTILs as stated previously, the problem resulting from the relatively high viscosity of the RTILs for an actual application might be reduced by adopting a Grotthus-like hopping mechanism at least for an output current density such as JSC. However, the conversion efficiency of a DSSC using RTILs does not reach the level of an organic solvent-based system. Kubo et al. reported in detail the effect of using an of iodide-based RTIL on the mechanism of the charge transfer process involving the photocurrent generation process of a DSSC system such as electron diffusion in the TiO2 electrode—a charge recombination process [23]. A high concentration of cationic species, which is inevitable with the use of ILs, decreases the diffusion constant of the electrons in TiO2, and a relatively high concentration of triiodide, which is necessary to increase JSC, increases the recombination loss between the photogenerated electrons and triiodide (Table 15.2). Furthermore, the triiodide acts as a filter because of its relatively high absorption coefficient in the visible light region [23,24]. All these phenomena are causes for the relatively low conversion efficiency of a DSSC using iodide-based RTILs compared with that using an organic solvent electrolyte (>10%).

15.4 QUASI SOLID-STATE DSSC SYSTEM USING RTILS BASED ON IODIDE Solidification of a liquid electrolyte while maintaining their performance is one of the important targets to produce a DSSC system for actual devices. Recently, various reports on the solidification of a liquid electrolyte using iodide-based RTILs and a gelator [23,25], cross-linked gel [26–28], fluoropolymer-based gel [29], and silica nanoparticles [30,31] have been reported. Surprisingly, the conversion efficiency of these cells is almost the same before and after the solidification of the electrolyte. Especially, the conversion efficiency of 7.0% recently was achieved with the use of silicon nanoparticles as a gelator instead of an organic gelator [30]. These results indicate that iodide RTILs might be a key material for the solidification of an electrolyte for DSSC systems, and novel techniques can overcome the shortcomings of using of RTILs. Kubo et al. reported that the quasisolidification with a gelator is valid for long-term stability at 85 °C as shown in Figure 15.7. The reason for the decrease in the normalized efficiency in an iodide-based RTIL without a gelator is because of to the decrease in the triiodide and adsorbed dye onto TiO2 [23,25].

c15.indd 230

1/20/2011 10:28:07 AM

231

RECENT PROGRESS 1.2

Normalized Efficiency

1.0

0.8 0.6 0.4

0.2

0.0 0

200

400

600

800

1000

Time / h

Figure 15.7. Time-course change under dry heat test of the normalized photoconversion efficiency of the solar cells stored at 85 °C; with a gelled molten salt electrolyte (solid line), a molten salt electrolyte (long dashed line), a gelled organic solvent electrolyte (short dashed line), a organic solvent electrolyte (dotted line) [25]. (Reprinted by permission of the publisher, The Royal Society of Chemistry).

However, CuI is a p-type semiconductor, which can be used as a hole transport layer for DSSC systems. The performance of a DSSC with the use of microcrystalline CuI is much improved by combining the CuI and ILs [32,33]. These reports show that the ILs serve as a protecting reagent for CuI from degradation and maintain the particle size small enough to allow effective contact with a dye-loaded oxide electrode.

15.5

RECENT PROGRESS

Since the first edition of this book was published at 2005, remarkable progress on the use of ionic liquids electrolyte for dye sensitized solar cells were reported by Professor Grätzel’s group. A new low viscosity ionic liquid based on tetracyanoborate anion first was reported as a nonvolatile electrolyte for the dye sensitized solar cells [34]. The relatively high conversion efficiency (7.0%) under AM 1.5 full sunlight was achieved when Z-907Na dye was used. Another cyanide-type ionic liquid such as thiocyanate [35], selenocyanate [36], and tricyanomehide [37] were not used successfully because of their instability toward prolonged thermal stress and light soaking. These results suggested that a resistance to light of ionic liquids, which is not a serious factor for other electrochemical devices such as a lithium battery and an electrical double layer

c15.indd 231

1/25/2011 3:29:00 PM

232

PHOTOELECTROCHEMICAL CELLS

capacitor should be considered. As stated in Section 15.3.2, the viscosity of a simple iodide such as 1-hexyl-3-methylimidaolium iodide was large, which did not allow achieving high conversion efficiency as a result of the low diffusion coefficient of iodide and triiodide; however, the same group also recently reported that these iodide melts could be used successfully even in a neat iodide. They discovered that a eutectic melt containing three different iodides such as 1-ethyl-3-methylimidazolium, 1-allyl-3-methylimidazolium, and 1,3-dimethylimidazolium iodide exhibited relatively high ionic conductivity. The best conversion efficiency of a DSC cell with using such eutectic melts and the tetracyanoborate ionic liquids was 8.2% [38]. Recently, Wang et al. reported that the performance of DSSCs with the use of MK-2 dye [39], which do not contain a Ru metal center, did not remarkably degrade even with the use of ionic liquids electrolyte instead of a conventional organic solvent [40]. Another interesting point was that the lifetime of an electron in a TiO2 layer injected from the MK-2 dye was longer than that of a typical Ru complex sensitizer such as N719. This indicates that an appropriate molecular design of a sensitizer will be necessary to improve the DSSC system with the use of ionic liquids. 15.6

SUMMARY

In this chapter, PEC systems using RTILs were introduced, focusing on the relationship between the viscosity of the RTILs and the short-circuit photocurrent. A Grotthus-like mechanism observed for a high concentration of iodide/ triiodide redox in RTILs indeed cover the shortcomings of the relatively high viscosity of the RTILs. This fact seems to be a key for application in an actual electrochemical power device. Continued improvement of the performance of a DSSC using neat ionic liquids will be achieved not only with the modification of ionic liquids electrolytes but also with the other cell components such as a sensitizer.

REFERENCES 1. K. Honda, A. Fujishima, Nature, 238, 37 (1972). 2. (a) H. Gerischer, Solar Energy Convresion, B. O. Seraphin, ed., Springer-Verlag, 1979, pp. 115. (b) R. Memming, Semiconductor Electrochemistry, Wiley-VCH, 2001. 3. H. Tsubomura, M. Matsumura, Y. Nomura, T. Amamiya, Nature, 261, 402 (1977). 4. Y. V. Pleskov, Y. Y. Gurevich, Semiconductor Photoelectrochemistry, Consultants Bureau, 1986 5. B. O’Regan, M. Grätzel, Nature, 353, 737 (1991). 6. M. Grätzel, J. Photochem. Photobio. Chem., 145, 4 (2003). 7. P. Bonhôte, A. P. Dias, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996).

c15.indd 232

1/20/2011 10:28:07 AM

REFERENCES

233

8. N. Papageorgiou, Y. Athanassov, M. Armand, P. Bonhôte, H. Pettersson, A. Azam, M. Grätzel, J. Electrochem. Soc., 143, 3099 (1996). 9. P. Singh, K. Rajeshwar, J. DuBow, R. Job, J. Am. Chem. Soc., 102, 4676 (1980). 10. P. Singh, R. Singh, K. Rajeshwar, J. DuBow, J. Electrochem. Soc., 128, 1145 (1981). 11. P. Singh, K. Rajeshwar, J. Electrochem. Soc., 128, 1724 (1981). 12. K. Rajeshwar, P. Singh, R. Thapar, J. Electrochem. Soc., 128, 1750 (1981). 13. R. Thapar, J. DuBow, K. Rajeshwar, J. Electrochem. Soc., 129, 1145 (1982). 14. J. Robinson, R. C. Bugle, H. L. Chum, D. Koran, R. A. Osteryoung, J. Am. Chem. Soc., 101, 3776 (1979). 15. Y. V. Pleskov, Y. Y. Gurevich, Semiconductor Photoelectrochemistry, Consultants Bureau, 1986, pp: 243. 16. M. K. Nazeeruddin, A. Kay, I. Rodico, R. Humphry-Baker, E. Muller, P. Liska, N. Vlachopoulos, M. Grätzel, J. Am. Chem. Soc., 115, 6382 (1993). 17. A. Tasaka, T. Yachi, T. Makino, K. Hamano, T. Kimura, K. Momota, J. Fluorine Chem., 97, 253 (1999). 18. R. Hagiwara, K. Matsumoto, Y. Nakamori, T. Tsuda, Y. Ito, H. Matsumoto, K. Momota, J. Electrochem. Soc., 150, D195 (2003). 19. R. Hagiwara, T. Hirashige, T. Tsuda, Y. Ito, J. Fluorine Chem., 99, 1 (1999). 20. H. Matsumoto, T. Matsuda, T. Tsuda, R. Hagiwara, Y. Ito, Y. Miyazaki, Chem. Lett., 26 (2001). 21. H. Matsumoto, T. Matsuda, Electrochemisry, 70, 190 (2002). 22. R. Kawano, M. Watanabe, Chem. Commun., 3, 330 (2003). 23. W. Kubo, S. Kambe, S. Nakade, T. Kitamura, K. Hanabusa, Y. Wada, S. Yanagida, J. Phys. Chem. B, 107, 4374 (2003). 24. T. Matsuda, H. Matsumoto, Electrochemistry, 70, 446 (2002). 25. W. Kubo, T. Kitamura, K. Hanabusa, Y. Wada, S. Yanagida, Chem. Commun., 4, 374 (2002). 26. Y. Shibata, T. Kato, T. Kado, R. Shiratuchi, W. Takashima, K. Kaneto, S. Hayase, Chem. Commun., 2730 (2003). 27. S. Murai, S. Mikoshiba, H. Sumino, S. Hayase, J. Photochem. Photobio. Chem., 148, 33 (2002). 28. S. Murai, S. Mikoshiba, H. Sumino, T. Kato, S. Hayase, Chem. Commun., 13, 1534 (2003). 29. P. Wang, S. M. Zakeeruddin, I. Exnar, M. Gräzel, Chem. Commun., 2972 (2002). 30. P. Wang, S. M. Zakeeruddin, P. Comte, I. Exnar, M. Grätzel, J. Am. Chem. Soc., 125, 1166 (2003). 31. E. Stathatos, P. Lianos, S. M. Zakeeruddin, P. Liska, M. Grätzel, Chem. Mater., 15, 1825 (2003). 32. A. Konno, G. R. A. Kumara, R. Hata, K. Tennakone, Electrochemistry, 70, 432 (2002). 33. Q.-B. Meng, A. Konno, A. K. Takahashi, X.-T. Zhang, I. Sutanto, T. N. Rao, O. Sato, A. Fujishima, Langmuir, 19, 3572 (2003). 34. D. Kuang, P. Wang, S. Ito, S. M. Zakeeruddin, M. Grätzel, J. Am. Chem. Soc.,128, 7732 (2006).

c15.indd 233

1/20/2011 10:28:07 AM

234

PHOTOELECTROCHEMICAL CELLS

35. P. Wang, S. M. Zakeeruddin, R. Humphry-Baker, M. Grätzel, Chem. Mater., 16, 2694 (2004). 36. P. Wang, S. M. Zakeeruddin, J.-E. Moser, R. Humphry-Baker, M. Grätzel, J. Am. Chem. Soc., 126, 7164 (2004). 37. P. Wang, B. Wenger, R. Humphry-Baker, J.-E. Moser, J. Teuscher, W. Kantlehner, J. Mezger, E. V. Stoyanov, S. M. Zakeeruddin, M. Grätzel, J. Am. Chem. Soc., 127, 6850 (2005). 38. Y. Bai, Y. Cao, J. Zhang, M. Wang, R. Li, P. Wang, S. M. Zakeeruddin, M. Grätzel, Nat. Mater., 7, 626 (2008). 39. N. Koumura, Z.-S. Wang, S. Mori, M. Miyashita, E. Suzuki, K. Hara, J. Am. Chem. Soc., 128, 14256 (2006). 40. Z.-S. Wang, N. Koumura, Y. Cui, M. Miyashita, S. Mori, K. Hara, Chem. Mater., 21, 2810 (2009).

c15.indd 234

1/20/2011 10:28:07 AM

16 FUEL CELLS Masahiro Yoshizawa-Fujita and Hiroyuki Ohno

We face multiple crises in the world, including burgeoning energy needs and a continuing discharge of global warming gases. Technologies for improving these environmental problems are being studied. In particular, a need to develop ecofriendly energy sources is receiving much attention worldwide. The focus is on a device that produces electricity through the redox reaction of molecules, which is called a fuel cell. From such a fuel cell, there would be zero emission in terms of greenhouse gases. Several motor vehicle manufacturers have undertaken the development of proton membrane fuel cells to power electric vehicles [1–4]. However, improved electrolyte materials have to be developed to obtain high-performance fuel cells. There are many problems with the present proton-transport materials in the emerging field of electric vehicle fuel cell applications. A high proton conductivity and high thermal stability are two of the critical requirements for fuel cell electrolytes. A major drawback is that in the proton membrane fuel cell the upper limit of available temperature is around 130 °C because the system needs to be hydrated to transport protons. Ionic liquids (ILs) are being considered more and more as alternatives for conventional electrolyte materials [5–7]. ILs offer the unique features of nonvolatility and nonflammability even in a liquid state. Systems that show ionic conductivity of more than 10−2 S cm−1 at room temperature have been reported close to the level required for fuel cell applications [8–10]. However, this value is based on the IL itself, and they do not include target ions such as the proton. This is a critical subject of research to make the present system viable.

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

235

c16.indd 235

1/20/2011 10:28:08 AM

236

FUEL CELLS

Proton-conducting membranes based on a perfluorinated ionomer and IL were prepared by Doyle et al. [11]. Traditional proton-conducting membranes such as Nafion (DuPont) suffer from the volatility of water at more than 100 °C. Proton conductivity therefore decreases immediately above 130 °C. It was thought that ILs would work as a thermal stable solvent in a transporting proton because ILs are nonvolatile liquids. Indeed, the membranes containing 1-butyl-3-methylimidazolium trifluoromethanesulfonate have been demonstrated to have an ionic conductivity of more than 0.1 S cm−1 at 180 °C. Although the primary charge carrier in these membranes is not known, the high ionic conductivity has been demonstrated to be in the temperature range of 100– 200 °C when using ILs. Recently, the morphology of proton-conducting membranes containing various ILs has been studied by small-angle X-ray scattering and tapping mode atomic force microscopy by Sekhon et al. [12]. Their results strongly depended on the hydrophilic/hydrophobic nature of ILs and indicated that Nafion membranes containing hydrophilic ILs showed larger ionic clusters and obvious microphase separation. The size and the connectivity of the ionic clusters affect the ionic conductivity of proton-conducting membranes. Nafion membranes containing hydrophilic ILs exhibited a higher ionic conductivity than those of the membranes containing hydrophobic ones. Recently, ILs have been studied vigorously for proton transfer in fuel cell electrolytes under water-free conditions [13–27]. These ILs also are called “Brønsted acid–base ILs” [15]. Brønsted acid–base ILs can be obtained by a simple coupling of a wide variety of tertiary amines with various acids. Table 16.1 gives a summary of the structures of tertiary amines and acids, thermal properties, and ionic conductivities for Brønsted acid–base ILs [15,16d]. Note that there are many systems showing Tm below 100 °C and a high ionic conductivity of more than 10−2 S cm−1 at 120 °C. These have activated protons, so they can be used as proton conductors for the fuel cell electrolytes over a wide temperature range. Angell et al. found a simple relation between the boiling point (the temperature at which the total vapor pressure reaches one atm) and the difference in pKa value (ΔpKa) for the acid and base determined in dilute aqueous solutions [14]. For ΔpKa values greater than 10, the boiling point elevation becomes so high (>300 °C) that preemptive decomposition prevents any accurate measurement. The completeness of proton transfer in such cases is suggested by the ionic liquid-like values of the Walden products. There was a period when imidazole received considerable interest as a proton conductive material for fuel cell applications. Imidazole, which is a self-dissociating compound, shows high proton conductivity over its Tm without any acid doping [28]. This behavior occurs because imidazole has both proton donor and proton accepter groups within one molecule [29]. Recently, a simple combination of imidazole and various acids has been reported as a novel proton conductive material [15,30]. For the specific mixing ratio of imidazole and acid, this combination showed a higher ionic conductivity than the values of pure imidazole and pure acid. It was thought that the maximum conductivity would be obtained when the acid–base mixing ratio is 1 : 1. However, the

c16.indd 236

1/20/2011 10:28:08 AM

237

FUEL CELL

TABLE 16.1. Structures, Thermal Properties, and Ionic Conductivities for Tertiary Amines Neutralized with an Equimolar Amount of Acids Abbreviation

Structure

[dmea] [TFSI]

N+

[dema] [TFSI]

N+

F3C H

O

H

N

N

+

H

N+

F3C H

[dema][TfO] N+

[EPp][TfO]

O

+

N

[dema] [MeO]

+ N

[dmea] [HSO4]

N+

[dema] [HSO4]

+ N

HO

+

HO

c16.indd 237

H

41.6

5.6

−13.1

4.3

−76.2

52

2.6

−72.2

10.6

2.22

−62.1

17.4

1.63

−117

O – O

– O

−75.9

3.8

−82.1

2.0

−53.7

0.88

O –

S

2.4

–

O

S

S

O

O

73

CF3

O

[EPp][HSO4] N

–

–

S

O

HO

4.1

O

O

H

H

24

O

– O

S

H3C

H

−67

O

O

+

N

S

O

S

H

4.6

O

H3C

[tea][MeO]

O

S

O

H

–

–

F3 C

66

O

O

H

−42

O

O

O F3C

CF3

S

O O

S

Tm (°C)

O

–

N

S

CF3

S

O O

F3 C

[dmea][TfO]

N

S

O

[Im][TFSI]

–

O O

F3C H

N

S

σi × 102 (S cm−1) at 120 °C

Tg (°C)

O O

1/20/2011 10:28:08 AM

238

FUEL CELLS (a) H+

N

N H N

N H N

N H N

N H

H+

(b) H

N

N H

+ N

H

N N

H

N

H

N +

N

N +

H

H

N H

Figure 16.1. Models of proton transfer process via (a) Grotthuss mechanism and (b) Vehicle mechanism.

imidazole–1,1,1-trifluro-N-(trifluoromethylsulfonyl)methanesulfoneamide (HTFSI) mixture showed the maximum conductivity at the imidazole excess region. There are two typical mechanisms of proton conduction—namely proton hopping (Grotthuss mechanism) and matrix transport (vehicle mechanism)—in such a system. Figure 16.1 shows models of the proton transfer process via the Grotthuss mechanism and the vehicle mechanism. It seems that, because the Grotthuss mechanism is dominant during the proton conduction as the imidazole ratio is increased, proton conductivity also should increase in the imidazole–HTFSI mixed system. Although this property is similar to that in a water molecule, which is a well-known proton conductor, imidazole seems to poison the Pt electrode [31]. This is a serious problem for fuel cell applications. To overcome such problems, imidazole analogues are being reported [32,33]. Polymer film electrolytes have been prepared by polymerization of the imidazole–acid mixtures, as shown in Figure 16.2. The copolymer was synthesized by imidazole derivatives and phosphoric acid derivatives that have a polymerizable group [34]. Their ionic conductivity was low, 10−8–10−9 S cm−1 at even 100 °C, which is even lower than that of monomer systems. This effect can be explained by the suppression of molecular motion. To overcome this effect, a flexible spacer was tethered between the imidazole and the polymerizable group [35–38]. Both groups showed an ionic conductivity of more than 10−5 S cm−1 at 100 °C, indicating that such a system can effectively maintain high proton conductivity. This result is consistent with that of IL-type polymer brushes tethering the imidazolium salt with a flexible spacer [39]. The long flexible spacer (i.e., to maintain low Tg), seems to be crucial in maintaining high ionic conductivity for polymer systems and to be even more influential than the density of imidazole moieties. Other approaches for the proton conductive ILs mentioned in this book should be helpful for this research too. Excellent proton conductive films can be obtained using these strategies.

c16.indd 238

1/20/2011 10:28:09 AM

239

FUEL CELL N

y P O OH N OH

x

HN

N

O

O

N H

n

NH

N N

O

O

N H

n

NH

HN x

O O Si O O HN

O n

N

O

O

N

Si O

Si O Si O Si

O

NH N

Figure 16.2. Structures of polymeric imidazole derivatives.

The evaluation of ILs as electrolytes for fuel cells has been done already. Watanabe et al. tested the power generation of the hydrogen/oxygen fuel cell by using a platinum electrode for the imidazole–HTFSI mixture system, as shown in Figure 16.3 [16a]. Because the oxygen reduction at the cathode was slow, early on the polarization became large. This seems to have been the first attempt to apply ILs for power generation in the absence of water molecules. Recently, Belieres et al. reported that the fuel cell performance of lowmelting inorganic salts with protonated cations (e.g., ammonium) was comparable with that of the phosphoric acid fuel cell operating in the temperature range 100–200 °C [18]. Nakamoto and Watanabe also found that the fuel cell performance of [dema][TfO] is much better than that obtained using phosphoric acid under the same conditions [16d]. Because [dema][TfO] had suitable properties as a fuel cell electrolyte at medium temperature, polymer

c16.indd 239

1/20/2011 10:28:09 AM

240

FUEL CELLS

E/V vs RHE

0.8 0.6 0.4 0.2 0.0 0.0

0.2

0.6

0.4 i/mA cm–2

Figure 16.3. Fuel cell characteristics for imidazole/HTFSI neutral salt at 130 °C. Scan rate is 1 mV s−1. W.E. is a Pt-wire in O2 atmosphere, C.E. is a Pt-wire in H2 atmosphere, and R.E. is a Pt-wire in H2 atmosphere. (Reprinted by permission of the Publisher, The Royal Society of Chemistry).

1.0

Cell voltage / V

0.8

100 80

0.6 60 0.4 40 0.2 0.0

20

0

100

200

300

Power density / mW cm–2

120 30 °C 80 °C

0 400

Current density / mA cm–2

Figure 16.4. Polarization curves of a H2/O2 fuel cell using a composite membrane of sulfonated polyimide and [dema][TfO] without humidification at gas use ratios of 30% for H2 and 15% for O2: (■, □) operation at 30 °C, (•, ○) operation at 80 °C. Reprinted by permission of the publisher, Elsevier.)

electrolyte membranes for nonhumidified fuel cells were prepared by using [dema][TfO] and sulfonated polyimides [40]. The composite membranes had enough thermal stability beyond 300 °C as well as an ionic conductivity of more than 10−2 S cm−1 at 120 °C under nonhumidified conditions. Figure 16.4 shows the performance of a H2/O2 fuel cell operated at 30 °C and 80 °C using a composite membrane without humidification. The composite membrane showed a maximum power density of 100 mW cm−2 and a limit current density of 240 mA cm−2 without humidification at 80 °C.

c16.indd 240

1/20/2011 10:28:09 AM

REFERENCES

241

These promising features of protic ILs are useful not only for fuel cell applications but also for a lot of applications such as solvents for organic synthesis reactions and biological applications [24]. It is readily anticipated that protic ILs will find uses in many more kinds of applications. REFERENCES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12.

13.

14. 15. 16.

17. 18. 19. 20. 21. 22.

c16.indd 241

B. C. H. Steele, A. Heinzel, Nature, 414, 345 (2001). K. D. Kreuer, ChemPhysChem, 3, 771 (2002). K. D. Kreuer, J. Memb. Sci., 185, 29 (2001). K. D. Kreuer, Chem. Mater., 8, 610 (1996). H. Ohno, ed., Ionic Liquids: The Front and Future of Material Developments, CMC, 2003. H. Ohno, ed., Ionic Liquid II: Marvelous Developments and Colorful Near Future, CMC, 2006. M. Armand, F. Endres, D. R. MacFarlane, H. Ohno, B. Scrosati, Nat. Mater., 8, 621 (2009). P. Bonhôte, A.-P. Dias, M. Armand, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). R. Hagiwara, Y. Ito, J. Fluorine Chem., 105, 221 (2000). W. Xu, C. A. Angell, Science, 302, 422 (2003). M. Doyle, S. K. Choi, G. Proulx, J. Electrochem. Soc., 147, 34 (2000). (a) S. S. Sekhon, J.-S. Park, E. Cho, Y.-G. Yoon, C.-S. Kim, W.-Y. Lee, Macromolecules, 42, 2054 (2009). (b) S. S. Sekhon, J.-S. Park, J.-S. Baek, S.-D. Yim, T.-H. Yang, C.-S. Kim, Chem. Mater., 22, 803 (2010). (a) M. Hirao, H. Sugimoto, H. Ohno, J. Electrochem. Soc., 147, 4168 (2000). (b) M. Yoshizawa, W. Ogihara, H. Ohno, Electrochem. Solid-State Lett., 4, E25 (2001). (c) H. Ohno, M. Yoshizawa, Solid State Ionics, 154–155, 303 (2003). M. Yoshizawa, W. Xu, C. A. Angell, J. Am. Chem. Soc., 125, 15411 (2003). A. Noda, M. A. B. H. Susan, K. Kudo, S. Mitsushima, K. Hayamizu, M. Watanabe, J. Phys. Chem. B, 107, 4024 (2003). (a) M. A. B. H. Susan, A. Noda, S. Mitsushima, M. Watanabe, Chem. Commun., 938 (2003). (b) M. A. B. H. Susan, M. Yoo, H. Nakamoto, M. Watanabe, Chem. Lett., 32, 836 (2003). (c) H. Nakamoto, A. Noda, K. Hayamizu, S. Hayashi, H. Hamaguchi, M. Watanabe, J. Phys. Chem. C, 111, 1541 (2007). (d) H. Nakamoto, M. Watanabe, Chem. Commun., 2539 (2007). (e) T. Yasuda, A. Ogawa, M. Kanno, K. Mori, K. Sakakibara, M. Watanabe, Chem. Lett., 38, 692 (2009). M. Yamada, I. Honma, Electrochim. Acta, 48, 2411 (2003). J.-P. Belieres, D. Gervasio, C. A. Angell, Chem. Commun., 4799 (2006). W. Ogihara, H. Kosukegawa, H. Ohno, Chem. Commun., 3637 (2006). J.-P. Belieres, C. A. Angell, J. Phys. Chem. B, 111, 4926 (2007). C. Iojoiu, P. Judeinstein, J.-Y. Sanchez, Electrochim. Acta, 53, 1395 (2007). R. Arvai, F. Toulgoat, M. Médebielle, B. Langlois, F. Alloin, C. Iojoiu, J.-Y. Sanchez, J. Fluorine Chem., 129, 1029 (2008).

1/20/2011 10:28:09 AM

242

FUEL CELLS

23. M. Anouti, M. Caillon-Caravanier, Y. Dridi, H. Galiano, D. Lemordant, J. Phys. Chem. B, 112, 13335 (2008). 24. T. L. Greaves, C. J. Drummond, Chem. Rev., 108, 206 (2008). 25. C. Iojoiu, M. Martinez, M. Hanna, Y. Molmeret, L. Cointeaux, J.-C. Leprêtre, N. E. Kissi, J. Guindet, P. Judeinstein, J.-Y. Sanchez, Polymer. Adv. Technol., 19, 1406 (2008). 26. M. Hanna, J.-C. Lepretre, J.-Y. Sanchez, Chem. Biochem. Eng. Q., 23, 107 (2009). 27. C. Brigouleix, M. Anouti, J. Jacquemin, M. Caillon-Caravanier, H. Galiano, D. Lemordant, J. Phys. Chem. B, 114, 1757 (2010). 28. K. D. Kreuer, Solid State Ionics, 94, 55 (1997). 29. W. Munch, K. D. Kreuer, W. Silvestri, J. Maier, G. Seifert, Solid State Ionics, 145, 437 (2001). 30. K. D. Kreuer, A. Fuchs, M. Ise, M. Spaeth, J. Maier, Electrochim. Acta, 43, 1281 (1998). 31. C. Yang, P. Costamagna, S. Srinivasan, J. Benziger, A. B. Bocarsly, J. Power Sources, 103, 1 (2001). 32. W. Q. Deng, V. Molinero, W. A. Goddard III, J. Am. Chem. Soc., 126, 15644 (2004). 33. (a) Z. Zhou, S. Li, Y. Zhang, M. Liu, W. Li, J. Am. Chem. Soc., 127, 10824 (2005). (b) S. Li, Z. Zhou, Y. Zhang, M. Liu, W. Li, Chem. Mater., 17, 5884 (2005). 34. A. Bozkurt, W. H. Meyer, J. Gutmann, G. Wegner, Solid State Ionics, 164, 169 (2003). 35. M. Schuster, W. H. Meyer, G. Wegner, H. G. Herz, M. Ise, M. Schuster, K. D. Kreuer, J. Maier, Solid State Ionics, 145, 85 (2001). 36. J. C. Persson, P. Jannasch, Chem. Mater., 15, 3044 (2003). 37. H. G. Herz, K. D. Kreuer, J. Maier, G. Scharfenberger, M. F. H. Schuster, W. H. Meyer, Electrochim. Acta, 48, 2165 (2003). 38. M. F. H. Schuster, W. H. Meyer, M. Schuster, K. D. Kreuer, Chem. Mater., 16, 329 (2004). 39. (a) M. Yoshizawa, H. Ohno, Chem. Lett., 889 (1999). (b) M. Yoshizawa, H. Ohno, Electrochim. Acta, 46, 1723 (2001). (c) S. Washiro, M. Yoshizawa, H. Nakajima, H. Ohno, Polymer, 45, 1577 (2004). 40. S.-Y. Lee, T. Yasuda, M. Watanabe, J. Power Sources, 195, 5909 (2010).

c16.indd 242

1/20/2011 10:28:10 AM

17 DOUBLE-LAYER CAPACITORS Makoto Ue

17.1

INTRODUCTION

Ionic liquids are being considered for use as liquid electrolytes for electrochemical energy storage devices because they can increase device safety with their nonvolatile and nonflammable properties [1–4]. The double-layer capacitor (DLC) has shown the most promise as an electrochemical energy storage device for hybrid electric vehicles and hybrid fuel cell vehicles is because it has a higher pulse power capability than conventional rechargeable batteries [5–7]. In this chapter, the technology of applying ionic liquids to the doublelayer capacitor as a liquid electrolyte is reviewed based mainly on the results of our laboratory. 17.2

OUTLINE OF DOUBLE-LAYER CAPACITORS [8]

When an electrode (electronic conductor) comes into contact with an electrolyte (ionic conductor), it shows some potential and attracts ions with the opposite sign, forming an “electrical double-layer” at the electrode/electrolyte interface, as shown in Figure 17.1a. Increasing its electrode potential causes additional adsorption of ions (charging current), and eventually, an electrode reaction (electrolytic current, Faradaic current) occurs. The “polarizable electrode” is an electrode in which charge transfer does not occur when the electrode potential is varied. A typical example of the ideal polarizable electrode is Hg, which shows a “double-layer capacitance” of 10∼20 μF cm−2 in aqueous electrolyte solutions. Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

243

c17.indd 243

1/20/2011 10:28:18 AM

244

DOUBLE-LAYER CAPACITORS

Because the double-layer capacitance is dependent on electrode potential V, the minimum values of differential double-layer capacitance Cd were adopted, defined as follows, where Q is the charge: Cd =

dQ dV

(17.1)

Formally, the capacitance C of a plane capacitor with plates of equal area S in parallel configuration, separated by a distance d, is given by equation (17.2), where ε0 and εr are the permittivity of the vacuum and the relative permittivity of the dielectric material, respectively: C=

ε rε0S d

(17.2)

The relative permittivity εr of the double layer is calculated to be 3∼7 if the Helmholtz model is considered, where the concept of a double-layer corresponds to a model consisting of two array layers of opposite charges separated by a small distance d (e.g., 0.3 nm). Therefore, the use of porous materials with high surface areas serving as electrodes enables capacitors to be made with high capacitance. For example, the use of an activated carbon (10 μF cm−2) with the specific surface area of 1000 m2 g−1 gives a capacitance as high as 100 F g−1. It should be noted that the capacitance measured in a unit cell corresponds to a quarter of the capacitance per unit weight or volume of the single electrode (F g−1 or F cm−3) because the cell consists of the two single electrodes (half capacitance and double weight), as shown in Figure 17.1b [9,10]. In this case, the capacitance of the unit cell becomes around 25 F g−1 if the weight of other components is neglected. (a)

(b) electrolyte

electrode

activated carbon electrode electrolyte current collector separator

+ + Helmholz plane + + + + +

potential

+

+

+ + + + + + + + - + + + + + + + + - + + + - + - + +

-

Figure 17.1. An electrical double-layer model (a) and a double-layer capacitor (b).

c17.indd 244

1/20/2011 10:28:19 AM

REQUIREMENTS FOR ELECTROLYTE MATERIALS

245

The double-layer capacitor is one of the “electrochemical capacitors” showing intermediate performances between conventional capacitors and rechargeable batteries from the viewpoint of energy and power densities. Although the terms “supercapacitor” and “ultracapacitor” are often used for double-layer capacitors in a sense that they have higher capacitance than conventional capacitors (ceramic, film, aluminum electrolytic, or tantalum electrolytic capacitors), these terms are not to be used because they are the trademarks of certain companies’ products.

17.3

REQUIREMENTS FOR ELECTROLYTE MATERIALS [8]

The selection of the electrolyte material is dependent on the electrode type. In principle, any ionic conductor can be used as the electrolyte material of a double-layer capacitor as long as the current caused by the electrochemical reactions (Faradaic current) does not flow in the desired potential region. The electrolyte merely works as an ion source to form a double layer. The energy W and power P of the electrochemical capacitor when it is discharged at a constant current I from initial voltage Vi to final voltage Vf are written as follows: W = 21 C (Vi2 − Vf2 ) = 21 C[(V0 − IR)2 − Vf2 ] P = 21 I (Vi + Vf ) = 21 I (V0 + Vf − IR)

(17.3) (17.4)

where C, V0, and R are the capacitance, the open circuit voltage, and the internal resistance, respectively. Both C and V0 must be large and R must be small for the capacitors to have high energy and high power. The following performances are required for the electrolyte materials: a. High double-layer capacitance. The capacitance C of a capacitor is proportional to the capacitance of an electrode, which is dependent on the type of electrolyte material chosen. An electrolyte material showing a high double-layer capacitance Cd for a given electrode is desired. b. High decomposition voltage. The maximum operational voltage V0 of a capacitor is governed by the decomposition voltage Vs of an electrolyte material. An electrolyte material with a wide stable potential region (electrochemical window) for a given electrode is desired. c. High electrolytic conductivity. The internal resistance R of a capacitor leads to energy loss caused by the voltage drop during charge–discharge cycling. Although it includes not only the resistance of an electrolyte but also those from other components, an electrolyte material with a high electrolytic conductivity κ is desired.

c17.indd 245

1/20/2011 10:28:19 AM

246

DOUBLE-LAYER CAPACITORS

d. Wide operational temperature range. The electrochemical capacitor must operate in the temperature range ΔT of at least −25 °C to 70 °C, and its performances must be closely related to the temperature characteristics of the electrolyte material. An electrolyte material satisfying the above requirements a–c at a wide temperature range is desired. e. High safety. One advantage of the double-layer capacitor is that it uses environmentally friendly materials in contrast to the rechargeable batteries that use heavy metals. An electrolyte material with high safety and low environmental impact is desired. For commercial double-layer capacitors using activated carbon electrodes, a nonaqueous solution such as 0.5–1 mol dm−3 (≡M) Et4NBF4 in propylene carbonate (PC) or an aqueous solution such as 3.7–4.5 M (30∼35 wt%) H2SO4 is used. The advantages of the nonaqueous liquid electrolytes are as follows: • • •

High decomposition voltage Wide operational temperature range Noncorrosive property can use components of low-cost metals such as Al. (The coin or cylindrical type cell can be fabricated.)

The disadvantages are as follows: • • • •

Low electrolytic conductivity Need of tight closure to isolate from atmospheric moisture High environmental impact High cost

Based on these properties, the double-layer capacitor comprising a pair of the activated carbon electrodes and the nonaqueous liquid electrolyte is the most favorable from the viewpoint of energy density. Ionic liquids are a kind of nonaqueous liquid electrolytes.

17.4

FUNDAMENTAL PROPERTIES OF IONIC LIQUIDS

Typical ionic liquids consist of a cation and an anion represented in Figure 17.2 [11,12]. However, as described in the previous section, a fundamental electrochemical property such as high electrolytic conductivity is required for ionic liquids in their use as a liquid electrolyte. Many combinations of heterocyclic cations and various anions have been examined, and the ionic compounds containing the 1-ethyl-3-methylimidazolium cation (EMI+) generally have shown the highest conductivity. No cation better than EMI+ has been found.

c17.indd 246

1/20/2011 10:28:19 AM

247

FUNDAMENTAL PROPERTIES OF IONIC LIQUIDS

+

N Me

N

N Et

Bu

BMI+

PF6−

CF3SO3−

+

N

N

Me

EMI +

BF4−

+

Me

N Oct

MOI +

(CF3SO2)2N−

Figure 17.2. Popular ions for ionic liquids.

To apply ionic liquids as an electrolyte for double-layer capacitors, they should have a wide operational temperature range and high safety. Although ionic liquids have favorable properties such as nonvolatility and nonflammability from the viewpoint of thermal stability and safety, they have the disadvantage that, because of their higher viscosity, they rapidly lose their electrolytic conductivity as the temperature decreases. The fundamental properties of EMI salts include the melting point Tmp, density d, viscosity η, electrolytic conductivity κ, and electrochemical window Ered–Eox; these properties are collected in Table 17.1. Because the electrochemical windows were measured by different working and reference electrodes as well as different sweep rates, it is not easy to compare them with each other. Therefore, the original data obtained by platinum or glassy carbon (GC) electrodes were converted to values based on Li/Li+ in PC by taking, for convenience, into account the effect of the sweep rate and the cutoff current density [35] and using the reduction potential of EMI+ as a probe. The converted values, using a relation depicted in Figure 17.3, include some deviations (∼0.2 V) because most reference electrodes suffer from the “solvent effect” of ionic liquids and each because EMI salt shows a different equilibrium potential. The number of ionic liquids with a melting point lower than 25 °C is limited.

17.4.1

Electrolytic Conductivity

The temperature dependence of electrolytic conductivity of several EMI salts is given in Figure 17.4 and compared with those of an aqueous solution (4.5 M H2SO4/H2O) and a nonaqueous propylene carbonate solution (1 M Et3MeNBF4/

c17.indd 247

1/20/2011 10:28:19 AM

248

DOUBLE-LAYER CAPACITORS

TABLE 17.1. Physicochemical Properties of EMI Salts Salt

EMIAlCl4 EMIAl2Cl7 EMIF · HF EMIF · 2.3HF EMINO2 EMINO3 EMIBF4 EMIAlF4 EMIPF6 EMIAsF6 EMISbF6 EMINbF6 EMITaF6 EMIWF7 EMICH3CO2 EMICF3CO2 EMIC3F7CO2 EMICH3SO3 EMICF3SO3 EMIC4F9SO3 EMI(CF3CO) (CF3SO2)N EMI(CF3SO2)2N EMI(CF3SO2) (C2F5SO2)N EMI(C2F5SO2)2N EMI(CF3SO2)3C EMI(CN)2N EMI(CN)3C

Tmp

η

d −3

κ

Ered −1

(°C)

(g cm )

(mPa s)

(mS cm )

8 nc 51 −90 55 38 15 45 62 53 10 −1 2 −15 −45 −14 nc 39 −10 28 −2

1.29 1.39 1.26 1.14 1.27a 1.28a 1.28

18 14

22.6 14.5

−15 nc −1 39 −12 −11

1.56 1.78 1.85 1.67 2.17 2.27

5

32

100

13.6

Eox

(V vs. Li/Li in PCc)

Ref.

+

1.0d 3.0d

5.5d 5.5d

1.5

5.5

1.0

5.5

13–15 13–15 16 17,18 19 19 19–21 22 23,24 21,24 21 21,25 21,25 21 19,26 26 26 13 13 26 27,28

6.2 8.5 7.1 3.2 2.8a 9.6a 2.7a 2.7 9.3

1.8 2.1 1.1 3.2d

5.8 5.7 5.7 5.7d

1.29b 1.45b 1.25 1.38

67 49 51 171 162a 35a 105a 160 43

1.1

4.6

1.2 1.0

4.8 5.3

1.46

25

9.8

1.1d

5.2d

1.52 1.55

28 49

8.4 4.2

1.0

5.6

24,26 29

1.08a 1.11a

61 181 17b 18b

3.4 1.7 27a 18a

1.0 1.0d 1.0 1.0

5.7 5.9d 4.9 4.4

24,30 31,32 33,34 34

Note: nc: not crystallized, 25 °C(a20 °C, b22 °C), cPt, 1 mA cm−2, 50 mV s−1 (dGC).

PC) [36–38]. Note that most ionic liquids (Figure 17.4a) show inferior conductivity compared with their aqueous and nonaqueous counterparts (Figure 17.4b) at low temperatures because of their higher viscosities and because their conductivities at −20 °C are less than 1 mS cm−1. EMIF · 2.3HF is the only exception. It shows high conductivity at 25 °C that is comparable with that of a neutral aqueous solution (1 M KCl/H2O), and it shows superior conductivity even at −40 °C to the nonaqueous counterpart (Figure 17.4b) [17,18].

c17.indd 248

1/20/2011 10:28:19 AM

FUNDAMENTAL PROPERTIES OF IONIC LIQUIDS

249

Figure 17.3. Equilibrium potentials of reference electrodes in EMIBF4, PC, and H2O.

17.4.2

Decomposition Voltage

It has been proven that EMI+ not only has an inferior cathodic stability to Et3MeN+ and but also an inferior anodic stability to BF4−, from the comparison between the electrochemical window of EMIBF4 (Figure 17.5a) and those of 0.5 M EMIBF4/PC and 0.5 M Et3MeNBF4/PC (Figure 17.5b) [36]. This narrower electrochemical window of EMI+ indicates that the EMI+ is weak not only for reduction but also oxidation because of the π-electron conjugated system. Although most ionic liquids showed the same cathodic limit governed by the reduction of EMI+, EMIF · 2.3HF showed a bit higher reduction potential, which is probably a result of acidic protons. EMIAl2Cl7, EMISbF6, EMINbF6, and EMIWF7 showed higher reduction potentials because of the reduction of their center atoms. Anodic limits usually were governed by the oxidation of EMI+; however, some organic anion salts such as EMIRCO2 (R = CH3, CF3, etc.), EMICH3SO3, EMI(CF3CO)(CF3SO2)N, and EMI(CN)nX (X = N, C) showed lower anodic limits because of the oxidation of their anions. It also was shown that the ionization energies of the MFn− type anion (inorganic fluoroanions) are larger than those of the (CF3SO2)nX− anion (organic fluoroanions) by ab initio molecular orbital and density functional theories [35]. The following anodic stability order observed in PC solutions [39] was applicable to ionic liquid systems: SbF6− > AsF6− ≥ PF6− > BF4− > (CF3 SO2 )3 C − ≥ (CF3 SO2 )2 N − > (CF3 SO2 )O− (17.5)

c17.indd 249

1/20/2011 10:28:19 AM

250

DOUBLE-LAYER CAPACITORS (a) 15 EMIBF4 EMITaF6 EMICF3SO3 EMI(CF3SO2)2N EMI(C2F5SO2)2N

κ / mS cm–1

10

5

0 –40

–30

–20

–10

0

10

20

30

T / °C (b) 103 4.5 M H2SO4/H2O

κ / mS cm–1

102 EMIF 2.3HF

101

1 M Et3MeNBF4/PC EMIBF4

100

–40

–30

–20

–10

0

10

20

30

T / °C

Figure 17.4. Temperature dependence of electrolytic conductivities of ionic liquids (a) compared with conventional aqueous and nonaqueous electrolyte solutions (b). (Reproduced by permission from The Electrochemical Society, Inc.)

17.4.3

Double-Layer Capacitance

Two reports determined the double-layer capacitance of ionic liquids [31,40]. By an electrocapillary curve measurement using dropping mercury electrode (DME), the integral double-layer capacitances Cdint of ionic liquids were smaller than those of aqueous solutions and were larger than those of

c17.indd 250

1/20/2011 10:28:20 AM

251

FUNDAMENTAL PROPERTIES OF IONIC LIQUIDS (a) 1.0

I / mA cm–2

0.5 EMIBF4 0

–0.5 Pt electrode 5 mV s–1 –1.0 –4

–3

–2

–1

0

1

2

3

4

V / V vs. I–3 / I– (b) 1.0

I / mA cm–2

0.5 0.5 M EMIBF4/PC 0 0.5 M Et3MeNBF4/PC

–0.5 Pt electrode 5 mV s–1 –1.0 –4

–3

–2

–1

0

1

2

3

4

V / V vs. Ag / Ag +

Figure 17.5. Linear sweep voltammograms for EMIBF4 (a), EMIBF4/PC, and Et3MeNBF4/PC (b).

nonaqueous solutions, as summarized in Table 17.2 [31]. This behavior can be explained by the thinner double layer being caused by the higher ionic concentration than that of nonaqueous solutions. However, the correlation between the double-layer capacitance and anion size [41] observed in PC solutions [8] is not clear. It also was shown that the double-layer capacitance of

c17.indd 251

1/20/2011 10:28:20 AM

252

DOUBLE-LAYER CAPACITORS

TABLE 17.2. Double-Layer Capacitance on Hg Electrolyte

Cdint (μF cm−2)

EMIBF4 EMICF3SO3 EMI(CF3SO2)3C EMI(CF3SO2)2N EMI(CF3SO2)2N EMI(CF3SO2)2N 1.5 M EMI(CF3SO2)2N/PC 1 M Et4NBF4/PC 0.1 M KCl/H2O 3 M H2SO4/H2O

10.6 12.4 10.6 11.7 12.0a 11.4b 9.1 7.0 15.1 14.6

a

GC. SpectraCarb 2220 yarn.

b

the ionic liquid was not dependent on the choice of electrode from among DME, GC, and activated carbon fiber [31].

17.5 PERFORMANCES OF IONIC LIQUIDS IN DOUBLE-LAYER CAPACITORS A double-layer capacitor consisting of activated carbon cloth electrodes (SpectraCarb 2220 yarn) and an ionic liquid (EMI(CF3SO2)2N) was proposed by Covalent Associates, Inc. (Corvallis, OR) [42]. A moderate capacitance around 25 F g−1 was obtained for 1000 charge–discharge cycles at 2.0 V, but its capacitance faded rapidly at 3.0 V down to 40% after 3500 cycles, as shown in the top panel of Table 17.3. After this study, the research group began to use ionic liquids as the supporting electrolyte salt in the organic solvents, which was probably because of the bad low-temperature characteristics [24,43–48]. EMIAlCl4 also was tested for a double-layer capacitor using carbon black, but the charge–discharge efficiency was low (52∼32%) because the electrolyte decomposed at a high cutoff voltage (3.5∼4.0 V), the capacity was also low, as shown in the lower panel of Table 17.3, because of an inappropriate selection of carbon material [49]. To learn about the potentiality of ionic liquids, we have examined the performances of double-layer capacitors comprising of a pair of activated carbon electrodes and each ionic liquid listed in Table 17.1, which we compared with the conventional nonaqueous electrolyte (1 M Et3MeNBF4/PC [8]) and the aqueous electrolyte (4.5 M H2SO4/H2O) [36–38]. EMIF · 2.3HF, EMIBF4, EMITaF6, EMICF3SO3, EMI(CF3SO2)2N, and EMI(C2F5SO2)2N were selected because they have melting points below 25 °C and enough electrolytic conductivity and electrochemical window. EMIAlCl4 was excluded because of its moisture sensitivity.

c17.indd 252

1/20/2011 10:28:20 AM

PERFORMANCES OF IONIC LIQUIDS IN DOUBLE-LAYER CAPACITORS

253

TABLE 17.3. Preceding Examples of Double-Layer Capacitors Using Ionic Liquids ΔV

System

i (mA, cm−2)

(a) Carbon fiber (BET surface area = 2000 cm2 g−1)

2.0

3.0 (3C rate)

EMI(CF3SO2)2N 39 mg for each electrode

3.0

3.0 (1.6C rate)

4.0

7.9

(b) Carbon black (BET surface area > 2000 cm2 g−1)

C (F g−1) Cycle# 5 1000 5000 5 1000 3500

25.0 25.0 20.8 30.5 24.3 12.4

T (°C) 25 52 82

3.8 6.3 7.5

EMIAlCl4 136 mg for active components

17.5.1

Preparation of Ionic Liquids

EMIBF4, EMICF3SO3, EMI(CF3SO2)2N, and EMI(C2F5SO2)2N were prepared by the neutralization of EMIOCO2CH3 (1-ethyl-3-methylimidazolium methylcarbonate) with the following corresponding acids: HBF4, CF3SO3H, (CF3SO2)2NH, and (C2F5SO2)2NH, respectively. A methanol solution of EMIOCO2CH3 was prepared by the reaction between 1-ethylimidazole and dimethyl carbonate in methanol at 145 °C for 13 hours [50]. EMIF · 2.3HF [17,18] and EMITaF6 [37] were prepared according to methods described elsewhere. 17.5.2

Fabrication of Double-Layer Capacitors

Coconut shell charcoal (average pore diameter 2.0 nm, surface area 1700 m2 g−1, average particle size 10 μm) 80 wt%, acetylene black conductor 10 wt%, and polytetrafluoroethylene binder 10 wt% were mixed, ground, and pressed at 6 MPa to form a disk composite 10 mm in diameter and 0.55 mm thick. A pair of these disk composite electrodes were dried at 300 °C below 7.5 × 10−3 Pa for 3 hours. Afterward, they were cooled in an argon atmosphere, and an electrolyte was immersed in them under reduced pressure. By sandwiching a nonwoven polypropylene separator with the two identical immersed electrodes, a 2032 coin cell was assembled with a stainless spacer, as depicted in Figure 17.6. The assembled cell was charged for t = 50 minutes by CC–CV mode from V = 0 to a given voltage (0.8, 2.0, or 2.8 V) at a constant current I = 5 mA and was discharged by CC mode to 0 V at a constant current I = 5 mA and a given temperature T as shown in Figure 17.6b. The capacitance C was calculated as

c17.indd 253

1/20/2011 10:28:20 AM

254

DOUBLE-LAYER CAPACITORS

Figure 17.6. Components and their configuration in a model cell (a) (Reproduced by permission from The Electrochemical Society, Inc.) and a charge–discharge curve (b).

TABLE 17.4. Capacitances and Internal Resistances of Double-Layer Capacitors Using Various Electrolytes κ (mS cm−1)

Electrolyte

EMIBF4 EMITaF6 EMIF · 2.3HF EMICF3SO3 EMI(CF3SO2)2N EMI(C2F5SO2)2N 1 M Et3MeNBF4/PC 4.5 M H2SO4/H2O

13.6 7.1 100 9.3 8.4 3.4 13 848

R/Ω

C (F cm−3)

R/Ω

C (F cm−3)

R/Ω

0.8 V

25 °C

2.8 V

70 °C

2.8 V

19.8

14.1 13.6 24.7a 14.6 13.6 10.7 13.1 —

43 71 43a 37 39 80 24 —

16.0 15.9 36.2a 14.0 16.1 15.7 13.3 —

31 42 53a 53 19 24 18 —

η (mPa s)

C (F cm−3)

25 °C

25 °C

32 51 4.9 43 28 61 3.5 2.5

6.8 11.1

6.6

8.1 23.2

11.8 4.6

a

2.0 V.

C = 2WV−2, where energy output W was calculated from the discharge curve as W = ∫ I Vd t . The volumetric capacitance was obtained by dividing the previews capacitance by the total volume of a pair of the disk electrodes. The internal resistance R was calculated from IR drop on the discharge curve. All data were the average values of the same three cells. 17.5.3

Initial Characteristics

Table 17.4 summarizes the performances of the double-layer capacitors using various electrolytes. To begin, the charging voltage was set to 0.8 V to make a comparison with the aqueous system. The double-layer capacitor using

c17.indd 254

1/20/2011 10:28:20 AM

PERFORMANCES OF IONIC LIQUIDS IN DOUBLE-LAYER CAPACITORS

255

EMIF · 2.3HF showed an intermediate capacitance and internal resistance between the aqueous (4.5 M H2SO4/H2O) and the nonaqueous (1 M Et3MeNBF4/ PC) electrolytes. These results reflect the differences in double-layer capacitance and electrolytic conductivity of the electrolytes themselves. It is evident that a low viscosity helps to increase the capacitance and to decrease the internal resistance because the capacitance of EMIBF4 was slightly smaller than that of the 1 M Et3MeNBF4/PC and because the internal resistance of EMIBF4 is larger than that of 1 M Et3MeNBF4/PC, even though EMIBF4 has higher double-layer capacitance (Table 17.2, Et3MeNBF4 is equivalent to Et4NBF4) and the similar electrolytic conductivity compared with 1 M Et3MeNBF4/PC. Then the charging voltage was increased to 2.8 V (2.0 V for EMIF · 2.3HF as discussed later). The capacitances of the cells were around 13 F cm−3 or 19 F g−1, based on the total volume or weight of the electrodes at 25 °C, and no remarkable difference was observed between the ionic liquids and the 1 M Et3MeNBF4/PC, except EMIF · 2.3HF showed almost two times the capacitance of those of other ionic liquids (Table 17.4). All ionic liquids showed a higher internal resistances than the 1 M Et3MeNBF4/PC. The capacitance and internal resistance generally increased and decreased, respectively, when the temperature was increased to 75 °C. The EMICF3SO3 showed the opposite tendency because of the electrolyte’s decomposition. The high internal resistance of the EMIF · 2.3HF system is also an indication of the electrolyte’s decomposition. It was reported by other groups that ionic liquids can show a higher capacitance than Et4NBF4/PC solutions [51,52]. This observation is attributed to the “ion-sieving effect” of the activated carbons because EMI+ (Vc = 116 Å3) is smaller than Et4N+ (159) [41]. The smaller anion also affords higher capacitance in the following order [51]: EMIBF4− (Va = 49 Å 3 ) > CF3SO−3 (80) > (CF3SO2 )2 N − (199). Figure 17.7 shows the change of the capacitances of the double-layer capacitors using various electrolytes, when the operating temperature was varied from 25 °C to −25 °C. The capacitance of all ionic liquids decreased rapidly compared with the 1 M Et3MeNBF4/PC because of the increase of internal resistance reflecting the temperature dependence of electrolytic conductivity in Figure 17.4. We have learned that this behavior is a fatal disadvantage of ionic liquids, and that EMIF · 2.3HF is the one exception that affords enough capacitance even at −25 °C, which reflects its high electrolytic conductivity. 17.5.4

Life Characteristics

Unlike for rechargeable batteries, a cycling test is not so important for doublelayer capacitors because the deterioration mostly occurs at maximum operating voltage. Figure 17.8 shows the deterioration over time when 3 V (or 2 V for EMIF · 2.3HF) is continuously applied to the cells at 70 °C. The capacitance generally has deteriorated over time. The EMIBF4 shows good life, similar to

c17.indd 255

1/20/2011 10:28:20 AM

256

DOUBLE-LAYER CAPACITORS (a) 15

EMIBF4 EMITaF6 EMICF3SO3 EMI(CF3SO2)2N EMI(C2F5SO2)2N

C / F cm–3

10

5

0 –30

–20

–10

0

10

20

30

20

30

T / °C (b) 30 EMIF 2.3HF (2.0 V) 4.5 M H2SO4/H2O (0.8 V) 1 M Et3 MeNBF4/PC (2.8 V) EMIBF4 (2.8 V) C / F cm–3

20

10

0 –30

–20

–10

0

10

T / °C

Figure 17.7. Temperature dependence of the capacitances of double-layer capacitors using ionic liquids (a) compared with conventional aqueous and nonaqueous electrolyte solutions (b). (Reproduced by permission from The Electrochemical Society, Inc.)

the 1 M Et3MeNBF4/PC. The EMITaF6 is not as stable as the EMIBF4 but is more stable than any of the organic fluoroanion compounds, EMICF3SO3, EMI(CF3SO2)2N, and EMI(C2F5SO2)2N. This behavior can be attributed to the inferior anodic stability of organic fluoranions [35]. The EMIF · 2.3HF did not have enough stability even at 2 V.

c17.indd 256

1/20/2011 10:28:20 AM

PERFORMANCES OF IONIC LIQUIDS IN DOUBLE-LAYER CAPACITORS

257

Figure 17.8. Life test of double-layer capacitors using ionic liquids at 3 V, 70 °C compared with a conventional nonaqueous electrolyte solution.

17.5.5 Voltage Dependence of Capacitance The most important problem for the evaluation of new electrolytes is to determine the charging voltage because a voltage that is too high will shorten the life and a voltage that is too low will not represent the maximum potential of the materials. The voltage behaviors of the double-layer capacitors using EMIF · 2.3HF, EMIBF4, and 1 M Et3MeNBF4/PC were examined by monitoring the charge–discharge curves at 70 °C, as the applied voltages were increased from 0.8 V to a given voltage by a 0.1 V step after each charge–discharge cycling. When the voltage increase in the charge curves became sluggish, the result was an increase in internal resistance and a decrease in capacitance, as shown in Figure 17.9, accompanied by a decomposition of the electrolyte. The approximate decomposition voltages at 70 °C were estimated to be 1.5, 3.0, and 3.2 V for EMIF · 2.3HF, EMIBF4, and 1 M Et3MeNBF4/PC, respectively. The capacitance of the double-layer capacitors was dependent on the applied voltage, particularly, for EMIF · 2.3HF. 17.5.6

Energy Density

The capacitance density and energy density at 70 °C, 25 °C, and −25 °C were estimated based on the experimental results in which practical charge voltages were taken into account. As listed in Table 17.5, EMIBF4 affords a higher energy density than the 1 M Et3MeNBF4/PC over 25 °C, but it fails at −25 °C. The energy density for EMIF · 2.3HF is situated between the aqueous and the PC systems.

c17.indd 257

1/20/2011 10:28:20 AM

258

DOUBLE-LAYER CAPACITORS

Figure 17.9. Voltage dependence of charge–discharge curves (a) and capacitance of double-layer capacitors when applied voltages were increased from 0.8 V by a 0.1 V step at 70 °C. (Reproduced by permission from The Electrochemical Society, Inc.)

17.6

RECENT PROGRESS

The main papers on the application of ionic liquids to double-layer capacitors published after 2003 are listed in Table 17.6. The electrolyte system to add organic solvents into ionic liquids was excluded here because it belongs to the conventional organic electrolyte system, which is practically important to take advantages of high solubility of ionic liquids to organic solvents.

c17.indd 258

1/20/2011 10:28:20 AM

259

RECENT PROGRESS

TABLE 17.5. Comparison of Energy Densities of Double-Layer Capacitors Using Various Electrolytes Electrolyte

EMIBF4 1 M Et3MeNBF4/PC EMIF · 2.3HFa 4.5 M H2SO4/H2O

C/F cm−3

V/V

2.8 2.8 1.5 0.8

W/Wh l−1

70 °C

25 °C

−25 °C

70 °C

25 °C

−25 °C

16.0 13.3 27.2

14.1 13.1 18.5 23.2

0 8.7 9.2 7.9

17.4 14.5 8.5

15.3 14.3 5.8 2.1

0 9.5 2.9 0.7

a

assumed to be 75% capacitance at 2.0 V.

Progress was made in this field by adopting new ionic liquids, which contain newly synthesized anions or cations listed in Table 17.7. Data of ionic liquids based on BF4− and (CF3SO2)2N− also were added for reference. 1. Introduction of perfluoroalkyltrifluoroborate anions Mitsubishi Chemical Corporation (Ibaraki, Japan) has applied perfluoroalkyltrifluoroborate C n F2 n +1BF3− salts as an electrolyte salt for the first time. We envisaged that an increase in anion size as well as an introduction of structural asymmetry would enhance the dissociation ability of BF4− salts [83]. New ionic liquids EMICnF2n+1BF3 (n = 1∼4) showed lower melting points than that of EMIBF4. Particularly, EMICF3BF3 and EMIC2F5BF3 showed a higher electrolytic conductivity than EMIBF4 at low temperature as well as comparable temperature dependence with 1 M Et3MeNBF4/PC down to −20 °C as shown in Figure 17.10 [56,84]. Ionic liquids based on CF3 BF3− and C 2 F5 BF3− generally showed a lower viscosity and a higher electrolytic conductivity than the counterparts based on BF4− and (CF3SO2)2N− [84]. A double-layer capacitor using EMIC2F5BF3 exhibited excellent low-temperature characteristics and gave comparable capacitance with 1 M Et3MeNBF4/PC at −25 °C as shown in Figure 17.11a [56]. This result was a technological breakthrough that removed the limitation of using ionic liquids at low temperature, encouraging the search for new chemical structures. Unfortunately, the double-layer capacitor using EMIC2F5BF3 had a short load life at 3.0 V, 70 °C because of its inferior anodic stability as shown in Figure 17.11b [56]. The reason why EMIBF4 showed a better life than 1 M Et3MeNBF4/ PC irrespective of its lower decomposition voltage is considered to be that CO2 gas generation from PC decomposition limited the cell life. Low gas generation is an advantage of ionic liquids [85]. 2. Introduction of trialkylalkoxyammonium cations Matsumoto in National Institute of Advanced Industrial Science and Technology (Osaka, Japan) has investigated aliphatic quaternary ammonium salts to improve the intrinsic narrower electrochemical window of

c17.indd 259

1/20/2011 10:28:21 AM

260

DOUBLE-LAYER CAPACITORS

TABLE 17.6. List of Main Papers on Double-Layer Capacitors Based on Ionic Liquids System AC/IL–PAN/ AC AC/IL/AC

CNT/IL/CNT

AC/IL/AC

Ionic Liquids (IL)

Comment

EMIBF4, PY14NTf2, BMIPF6 EMIBF4, BMIBF4, BMIPF6, EMINTf2, PY14NTf2 EMIN(CN)2, EMINTf2, BMIPF6, PY13NTf2 EMIC2F5BF3

Required sulfolane as a plasticizer Comparable capacitance with organic solution

Poznań U. Tech. Poznań U. Tech.

53

Some loss of capacitance

U. Wollongong

55

Mitsubishi Chemical Nisshinbo Industries Poznań U. Tech.

56

Shinshu U.

59

Yamaguchi U.

60

Yokohama Nat. U. Poznań U. Tech. Shanxi Institute U. Paul Sabatier Kansai U.

61

63

U. Bologna

66

ADA Technologies U. Bologna

67

ACF/IL/ACF

DEMEBF4, DEMENTf2

ACF/IL/ACF

EMIBF4 + LiBF4

CNT/IL/CNT

EMINTf2

AC/IL–PAN/ AC AC/IL/AC

EMIOTf

AC/IL/AC

PY14NTf2

AC/IL/AC

EMIBF4

XGG/IL/XGC

EMINTf2, PY14NTf2 EMIBF4, EMINTf2

Good low temp. characteristics Excellent durability over 100 °C Comparable capacitance with organic solution Higher specific capacitance at low current Expanded cathodic stability by LiBF4 Higher capacitance per unit surface area Required sulfolane as a plasticizer Well cycled at between 0 and 2.0 V High voltage (3.5 V) at 60 °C Activation of an IL by BaTiO3 addition High voltage (3.4 or 3.7 V) at 60 °C Excellent cycle life

ACF/IL/ACF CNT/IL/CNT

PY14NTf2, PY14OTf, PY1(2O1)NTf2 EMIN(FSO2)2 EMINTf2

High voltage (3.9 V) at 20 °C Excellent rate capability Superior performances

AC/IL/AC

PY1(2O1)NTf2

Operate between −30 and 60 °C

AC/IL/AC AC/IL/–PAN/ AC

AC/IL–(PVdF– HFP)/AC AC/IL/AC

DEMEBF4, (DEMENTf2) (EMI)3PO4

EMICF3CO2

Institute

Kansai U. ADA Technologies U. Bologna

Ref.

54

57 58

62

64 65

68 69 70 71

Note: AC: activated carbon; PAN: polyacrylonitrile; Tf: CF3SO2; CNT: carbon nanotube; ACF: activated carbon fiber; XGC: xerogel carbon; PVdF-HFP: poly (vinylidene fluoride-hexafluoropropylene).

c17.indd 260

1/20/2011 10:28:21 AM

261

c17.indd 261

1/20/2011 10:28:21 AM

Ionic liquid

−18 —

17 5

(CF3SO2)2N−

(FSO2)2N− (FSO2)(CF3SO2)N−

— 8

(CF3SO2)2N− BF4−

7

15 −20 1

BF4− CF3 BF3− C 2 F5 BF3−

PO2 F2−

−13 −28 −16

(FSO2)2N− (FSO2)(CF3SO2)N− (CF3SO2)2N−

Tmp/°C

TABLE 17.7. Physicochemical Properties of New Ionic Liquids

1.34 1.38

1.44 1.51

1.42 1.20

1.31

1.28 1.35 1.42

1.44 1.49 1.52

d/g cm−3

40 43

72 50

69 426

35

32 26 27

18 23 33

η/mPa s

8.2 6.8

3.3 4.7

2.6 1.3

12

13.6 14.8 12.2

15.4 12.6 8.3

κ/mS cm−1

0.4 0.1

0.0 0.4

0.0 −0.1

0.7

0.9 0.8 0.8

0.8 0.8 0.8

5.7 5.8

5.8 5.8

5.6 5.5

5.0

5.5 5.4 5.5

5.4 5.5 5.5

Ered Eox V vs. Li/Li + in PC*

72,74 74 (Continued)

78 78

77 77

76

75 75

72,74 73,74 72,74

Ref.

262

c17.indd 262

1/20/2011 10:28:21 AM

25 °C, * GC, 1 mA cm−2, 50 mV s−1.

Ionic liquid

TABLE 17.7. (Continued)

(CF3SO2)2N−

(FSO2)2N−

(CF3SO2)2N− BF4− CF3 BF3− C 2 F5 BF3−

(CF3SO2)2N− (FSO2)(CF3SO2)N−

1.48 1.25 1.31 1.38 1.45 1.39 1.37 1.46

−55 — — −36

1.45

d/g cm−3

−21 −18 −4 26

12 21

Tmp/°C

24 20 25 30

40 100 46 37

63 44

η/mPa s

15.7 15.3 12.4 7.1

5.5 6.8 7.7 6.6

3.6 7.0

κ/mS cm−1

1.1 1.1 1.1

0.0 0.3

0.0 0.1

5.7 5.9 5.8

5.5 5.5

5.5 5.9

Ered Eox + V vs. Li/Li in PC*

81 81 81 82

80 80 80 80

74,79 74

Ref.

263

κ

RECENT PROGRESS

Figure 17.10. Temperature dependence of electrolytic conductivity of new ionic liquids.

EMI salts and found that the incorporation of an alkoxy group in tetraalkylammonium cations decreased viscosity and increased electrolytic conductivity [78]. For example, the electrolytic conductivities (mS cm−1) at 25 °C of (CF3SO2)2N− salts are as follows (see Table 17.7): Trimetylpropylammonium (N1113): 3.3 versus Trimetyl(methoxymethyl) ammonium (N111(1O1)): 4.7 Methylpropylpyrrolidinium (PY13): 3.6 versus Methyl(methoxymethyl) pyrrolidinium (PY1(1O1)): 5.5 Along this line, Nisshinbo Industries, Inc. (Chiba, Japan) developed new ionic liquids with a diethylmethyl(2-methoxyethyl)ammonium (DEME+ : N122(2O1)) cation [57]. A double-layer capacitor using DEMEBF4 gave excellent thermal stability and worked at 150 °C, where those using 1 M Et4NBF4/PC and EMIBF4 failed as shown in Figure 17.12 [57,86]. It became evident that one advantage to using ionic liquid electrolytes is their high thermal stability, which allows for high-temperature applications. Unfortunately, the double-layer capacitor using DEMEBF4 showed a 15% capacitance decrease at 4 mA cm−2, 25 °C because of its low electrolytic conductivity (1.3 mS cm−1) [57]. 3. Introduction of bis(fluorosulfonyl)amide anion Dai-ichi Kogyo Seiyaku Co., Ltd. (Kyoto, Japan) developed new ionic liquids with a bis(fluorosulfonyl)amide (FSA−) anion [72], and Ishikawa in Kansai University (Osaka, Japan) jointly investigated the performance of a double-layer capacitor using EMI(FSO2)2N in comparison with that using EMI(CF3SO2)2N [69]. The double-layer capacitor using

c17.indd 263

1/20/2011 10:28:21 AM

264

DOUBLE-LAYER CAPACITORS (a)

Δ

(b)

Figure 17.11. Low-temperature characteristics (a) and load life at high temperature (b) of double-layer capacitors.

EMI(FSO2)2N exhibited excellent charge-discharge characteristics at high current densities such as 50 mA cm−2 as shown in Figure 17.13 [69,86] because EMI(FSO2)2N has a high electrolytic conductivity (15.4 mS cm−1), which is almost equivalent to that of 2 M Et3MeNBF4/PC (15.3 mS cm−1). Taking account of the fact that EMI(FSO2)2N has a lower melting point (−13 °C) than EMIC2F5BF3 (1 °C), the double-layer capacitor using

c17.indd 264

1/20/2011 10:28:21 AM

RECENT PROGRESS

265

Figure 17.12. High-temperature characteristics of double-layer capacitors.

Figure 17.13. Rate capability of double-layer capacitors.

EMI(FSO2)2N would show excellent low-temperature characteristics. Unfortunately, the double-layer capacitor using EMI(FSO2)2N showed a high self-discharge at 2.5 V by unknown origin [69]. 4. Other research Lewandowski in Poznań University of Technology (Poznan, Poland) used polymer electrolytes including ionic liquids to construct

c17.indd 265

1/20/2011 10:28:21 AM

266

DOUBLE-LAYER CAPACITORS

double-layer capacitors, which removed the disadvantage of liquid electrolytes with immobilizing ionic liquids with polymers (PAN and PEO) [53,58,62]. Mastragostino in the University of Bologna (Bologna, Italy) fabricated double-layer capacitors using ionic liquids based on butylmethylpyrrolidinium (PY14) cation and examined their high-voltage performance at 60 °C [66,68]. By using PY1(2O1)(CF3SO2)2N based on a new cation, methyl(methoxyethyl)pyrrolidinium (PY1(2O1)), the operation temperature of double-layer capacitors was extended down to 20 °C (V = 0∼3.9 V, I = 5∼50 mA cm−2) and −30 °C (0.4 mA cm−2) [68,71]. 17.7

CONCLUDING REMARKS

The performance of double-layer capacitors using ionic liquids comes close to that with organic electrolytes by developing new ionic liquids. However, there is no ionic liquid yet that satisfies both the high electrolytic conductivity and the wide electrochemical window over the wide temperature range. Instead of the conventional ions such as EMI+, PY14+ , BF4−, (CF3SO2)2N−, new ions such as + + PY1(1O1) , S122 , CF3 BF3−, C 2 F5 BF3−, (FSO2)2N−, and (FSO2)(CF3SO2)N− will be examined extensively and their optimized combination will produce a promising ionic liquid for practical application. For example, PY1(1O1)(FSO2)2N and PY1(1O1)(FSO2)2N are interesting compounds to synthesize among them. The author expects continued progress in this field.

REFERENCES 1. J. S. Wilkes, in Ionic Liquids in Synthesis, P. Wasserscheid and T. Welton, eds., WileyVCH, 2003, p. 1. 2. J. S. Wilkes, in Green Industrial Applications of Ionic Liquids, R. D. Rogers, K. R. Seddon, and S. Volkov, eds., Kluwer Academic, 2002, p. 295. 3. A. Weber, G. E. Blomgren, in Advances in Lithium-Ion Batteries, W. A. van Schalkwijk, and B. Scrosati, eds., Kluwer Academic, 2002, p. 185. 4. H. Matsumoto, M. Ue, M. Watanabe, in Ionic Liquids, (in Japanese), H. Ohno, ed., CMC, 2003, p. 208. 5. B. E. Conway, Electrochemical Supercapacitors: Scientific Fundamentals and Technological Applications, Kluwer Academic, 1999. 6. A. Nishno, K. Naoi, eds., Technologies and Materials for EDLC, (in Japanese), CMC, 1998. 7. A. Nishino, K. Naoi, eds., Technologies and Materials for EDLC II, (in Japanese), CMC, 2003. 8. M. Ue, Curr. Top. Electrochem., 7, 49 (2000). 9. M. Ue, M. Takeda, Electrochemistry, 70, 194 (2002) (in Japanese). 10. M. Ue, in Ref. 4, p. 232 and Ref. 7, p. 295 (in Japanese). 11. M. Ue, Mater. Integration, 16, 40 (2003) (in Japanese).

c17.indd 266

1/20/2011 10:28:22 AM

REFERENCES

267

12. M. Ue, Mater. Stage, 3 (7), 19 (2003) (in Japanese). 13. E. I. Cooper, E. J. M. O’Sullivan, in Proc. 8th Int. Symp. Ionic Liquids, R. J. Gale, G. Blomgren, and H. Kojima, eds., PV92-16, Electrochemical Society, 1992, p. 386. 14. J. S. Wilkes, J. A. Levisky, R. A. Wilson, C. L. Hussey, Inorg. Chem., 21, 1263 (1982). 15. P. R. Gifford, J. B. Palmisano, J. Electrochem. Soc., 134, 610 (1987). 16. K. Matsumoto, T. Tsuda, R. Hagiwara, Y. Ito, O. Tamada, Solid St. Sci., 4, 23 (2002). 17. R. Hagiwara, T. Hirashige, T. Tsuda, Y. Ito, J. Fluorine Chem., 99, 1 (1999). 18. R. Hagiwara, T. Hirashige, T. Tsuda, Y. Ito, J. Electrochem. Soc., 149, D1 (2002). 19. J. S. Wilkes, M. J. Zaworotko, J. Chem. Soc., Chem. Commun., 1992, 965. 20. A. Noda, K. Hayamizu, M. Watanabe, J. Phys. Chem. B, 105, 4603 (2001). 21. K. Matusmoto, R. Hagiwara, R. Yoshida, Y. Ito, Z. Mazej, P. Benkič, B. Žemva, O. Tamada, H. Yoshino, S. Matsubara, J. Chem. Soc., Dalton Trans., 2004, 144. 22. (a) M. Ue, M. Takeda, T. Takahashi, M. Takehara, unpublished results. (b) M. Takehara, M. Ue, M. Takeda, World Patent, No. WO02101773 (2002). 23. J. Fuller, R. T. Carlin, H. C. De Long, D. Haworth, J. Chem. Soc., Chem., Commun., 1994, 299. 24. A. B. McEwen, H. L. Ngo, K. LeCompte, J. L. Goldman, J. Electrochem. Soc., 146, 1687 (1999). 25. K. Matsumoto, R. Hagiwara, Y. Ito, J. Fluorine Chem., 115, 133 (2002). 26. P. Bonhôte, A.-P. Dias, N. Papageorgiou, K. Kalyanasundaram, M. Gräzel, Inorg. Chem., 35, 1168 (1996). 27. H. Matsumoto, H. Kageyama, Y. Miyazaki, J. Chem. Soc., Chem. Commun., 2002, 1726. 28. H. Matsumoto, H. Kageyama, Y. Miyazaki, in Molten Salts XIII, H. C. Delong, R. W. Bradshaw, M. Matsunaga, G. R. Stafford, and P. C. Trulove, eds., PV2002-19, Electrochemical Society, 2002, p. 1057. 29. H. Matusmoto, private communication (2003). 30. J. L. Goldman, A. B. McEwen, in Selected Battery Topics, W. R. Cieslak, ed., PV9815, Electrochemical Society, 1999, p. 507. 31. C. Nanjundiah, S. F. McDevitt, V. R. Koch, J. Electrochem. Soc., 144, 3392 (1997). 32. A. B. McEwen, J. L. Goldman, D. Wasel, L. Hargens, in Molten Salts XII, H. C. De Long, S. Deki, G. R. Stafford, and P. C. Trulove, eds., PV99-41, Electrochemical Society, 2000, p. 222. 33. D. R. MacFarlane, S. A. Forsyth, J. Golding, G. B. Deacon, Green Chem., 4, 444 (2002). 34. Y. Yoshida, K. Muroi, A. Otsuka, G. Saito, M. Takahashi, T. Yoko, Inorg. Chem., 43, 1458 (2004). 35. M. Ue, A. Murakami, S. Nakamura, J. Electrochem. Soc., 149, A1572 (2002). 36. M. Ue, M. Takeda, J. Korean Electrochem. Soc., 5, 192 (2002). 37. M. Ue, M. Takeda, T. Takahashi, M. Takehara, Electrochem. Solid-State Lett., 5, A119 (2002). 38. M. Ue, M. Takeda, A. Toriumi, A. Kominato, R. Hagiwara, Y. Ito, J. Electrochem. Soc., 150, A499 (2003). 39. M. Ue, M. Takeda, M. Takehara, S. Mori, J. Electrochem. Soc., 144, 2684 (1997).

c17.indd 267

1/20/2011 10:28:22 AM

268

DOUBLE-LAYER CAPACITORS

40. R. J. Gale, R. A. Osteryoung, Electrochim. Acta, 25, 1527 (1980). 41. M. Ue, A. Murakami, S. Nakamura, J. Electrochem. Soc., 149, A1385 (2002). 42. (a) V. R. Koch, C. Nanjundiah, J. L. Goldman, in Proc. 5th Int. Sem. Double Layer Capacitors and Similar Energy Storage Devices, Florida Educational Seminars, 1995. (b) V. R. Koch, C. Nanjundiah, R. T. Carlin, U.S. Patent, No. 5827602, 1998. 43. C. Nanjundiah, J. L. Goldman, S. F. McDevitt, V. R. Koch, in Electrochemical Capacitors II, F. M. Delnick, D. Ingersoll, X. Andrieu, and K. Naoi, eds., PV96-25, Electrochemical Society, 1997, p. 301. 44. A. B. McEwen, R. Chadha, T. Blakley, V. R. Koch, in Electrochemical Capacitors II, F. M. Delnick, D. Ingersoll, X. Andrieu, and K. Naoi, eds., PV96-25, Electrochemical Society, 1997, p. 313. 45. A. B. McEwen, S. F. McDevitt, V. R. Koch, J. Electrochem. Soc., 144, L84 (1997). 46. A. B. McEwen, J. L. Goldman, T. Blakley, W. F. Averill, V. R. Koch, in Batteries for Portable Applications and Electric Vehicles, C. F. Holmes, and A. R. Landgrebe, eds., PV97-18, Electrochemical Society, 1997, p. 602. 47. H. L. Ngo, A. B. McEwen, in Selected Battery Topics, W. R. Cieslak, ed., PV98-15, Electrochemical Society, 1999, p. 683. 48. J. L. Goldman, A. B. McEwen, Electrochem. Solid State Lett., 2, 501 (1999). 49. J. Caja, T. D. J. Dunstan, G. Mamantov, in Molten Salts XI, P. C. Trulove, H. C. De Long, G. R. Stafford, and S. Deki, eds., PV98-11, Electrochemical Society, 1998, p. 333. 50. (a) M. Ue, in Ionic Liquids, (in Japanese), H. Ohno, ed., CMC, 2003, p. 9. (b) S. Mori, K. Ida, M. Ue, U.S. Patent, No. 4892944, (1998). 51. T. Kaneko, M. Watanabe, Abstracts of 43rd Battery Symp. in Japan, 2002, p. 412. 52. S. Shiraishi, N. Nishina, A. Oya, R. Hagiwara, Electrochemistry, 73, 593 (2005). 53. A. Lewandowski, A. Swiderska, Solid State Ionics, 161, 243 (2003). 54. A. Lewandowski, M. Galiński, J. Phys. Chem. Solid, 65, 281 (2004). 55. J. N. Barisci, G. G. Wallace, D. R. MacFarlane, R. H. Baughman, Electrochem. Commun., 6, 22 (2004). 56. Z.-B. Zhou, M. Takeda, M. Ue, J. Fluorine Chem., 125, 471 (2004). 57. T. Sato, G. Masuda, K. Takagi, Electrochim. Acta, 49, 3603 (2004). 58. A. Lewandowski, A. Swiderska, Pol. J. Chem., 78, 1371 (2004). 59. Y.-J. Kim, Y. Matsuzawa, S. Ozaki, K. C. Park, C. Kim, M. Endo, H. Yoshida, G. Masuda, T. Sato, M. S. Dresselhaus, J. Electrochem. Soc., 152, A710 (2005). 60. I. Murayama, N. Yoshimoto, M. Egashira, M. Morita, Y. Kobayashi, M. Ishikawa, Electrochemistry, 73, 600 (2005). 61. T. Katakabe, T. Kaneko, M. Watanabe, T. Fukushima, T. Aida, J. Electrochem. Soc., 152, A1913 (2005). 62. A. Lewandowski, A. Swiderska, Appl. Phys. A, 82, 579 (2006). 63. G.-H. Sun, K.-X. Li, Q. Li, H. Fan, J.-Y. Gu, Dianhuaxue (Electrochemistry), 13, 44 (2007) (in Chinese). 64. A. Balducci, R. Dugas, P. L. Taberna, P. Simon, D. Plée, M. Mastragostino, S. Passerini, J. Power Sources, 165, 922 (2007). 65. Y. Nagao, Y. Nakayama, H. Oda, M. Ishikawa, J. Power Sources, 166, 595 (2007).

c17.indd 268

1/20/2011 10:28:22 AM

REFERENCES

269

66. M. Lazzari, F. Soavi, M. Mastragostino, J. Power Sources, 178, 490 (2008). 67. W. Lu, K. Henry, C. Turchi, J. Pellegrino, J. Electrochem. Soc., 155, A361 (2008). 68. C. Arbizzani, M. Biso, D. Cericola, M. Lazzari, F. Soavi, M. Mastragostino, J. Power Sources, 185, 1575 (2008). 69. N. Handa, T. Sugimoto, M. Yamagata, M. Kikuta, M. Kono, M. Ishikawa, J. Power Sources, 185, 1585 (2008). 70. W. Lu, L. Qu, K. Henry, L. Dai, J. Power Sources, 189, 1270 (2009). 71. M. Lazzari, F. Soavi, M. Mastragostino, J. Electrochem. Soc., 156, A661 (2009). 72. H. Matsumoto, H. Sakaebe, K. Tatsumi, M. Kikuta, E. Ishiko, M. Kono, J. Power Sources, 160, 1308 (2006). 73. H. Matsumoto, N. Terasawa, T. Umecky, S. Tsuzuki, H. Sakaebe, K. Asaka, K. Tatsumi, Chem. Lett., 37, 1020 (2008). 74. H. Matsumoto, N. Terasawa, S. Tsuzuki, T. Umecky, H. Sakaebe, K. Tatsumi, The 49th Battery Symp. in Japan, 1G23, 2008. 75. Z.-B. Zhou, H. Matsumoto, K. Tatsumi, Chem. Eur. J., 10, 6581 (2004). 76. K. Matsumoto, R. Hagiwara, Inorg. Chem., 48, 7350 (2009). 77. Z.-B. Zhou, H. Matsumoto, K. Tatsumi, Chem. Eur. J., 11, 752 (2005). 78. H. Matsumoto, M. Yanagida, K. Tanimoto, M. Nomura, Y. Kitagawa, Y. Miyazaki, Chem. Lett., 922 (2000). 79. D. R. MacFarlane, P. Meakin, J. Sun, N. Amini, M. Forsyth, J. Phys. Chem. B, 103, 4164 (1999). 80. Z.-B. Zhou, H. Matsumoto, K. Tatsumi, Chem. Eur. J., 12, 2196 (2006). 81. H.-B. Hana, J. Niea, K. Liua, W.-K. Li, W.-F. Fenga, M. Armand, H. Matsumoto, Z.-B. Zhou, Electrochim. Acta, 55, 1221 (2010). 82. H. Matsumoto, T. Matsuda, Y. Miyazaki, Chem. Lett., 1430 (2000). 83. Z.-B. Zhou, M. Takeda, M. Ue, J. Fluorine Chem., 123, 127 (2003). 84. M. Ue, in Ionic Liquids II (in Japanese), H. Ohno, ed., CMC, 2005, p. 172. 85. M. Ue, H. Tokuda, T. Kawai, M. Yanagidate, Y. Otake, ECS Trans., 16(35), 173 (2009). 86. M. Ue, in Technologies and Materials for Supercapacitors IV, (in Japanese), A. Nishino, and K. Naoi, eds., CMC, 2010, p. 114.

c17.indd 269

1/20/2011 10:28:22 AM

18 ACTUATORS Kinji Asaka

Recently, much attention has focused on electroactive polymer (EAP) actuators based on electrochemical (EC) processes such as conducting polymers (CPs), carbon nanotubes (CNTs), ionic polymer-metal composites (IPMCs) that can directly transform electric energy into mechanical work because these EAP actuators have many advantages as compared with conventional actuators such as light, flexible, low-driven voltage, high-power density, and so on. [1]. The EAP actuator system consists of the EAP electrode, an electrolyte solution, and the counter electrode. The actuation of the EAP is based on EC processes such as redox reactions of the EAP or charging/discharging of the double layer between the EAP and the electrolyte. Hence, the electrolyte is essential for the EAP actuators as other EC devices. Ionic liquids (ILs) are known to be nonvolatile and show high ionic conductivities and wide potential windows. Hence, ILs are considered ideal electrolytes for the EAP actuators that can operate in air. CPs are electronically conducting organic materials featuring conjugated structures such as polyaniline, polypyrrole, polythiophene, and so on, the chemical structures of which are shown in Figure 18.1. Electrochemically changing the oxidation state leads to the addition or removal of charge from the polymer backbone as well as a flux of ions to balance charges. This ion flux, often accompanied by solvent, induces swelling or contraction of the material as shown in Figure 18.2. In addition, the conformation changes of the polymeric chain also take place. By these processes, the input electric energy changes into the mechanical work [2]. Several authors have tried to use ILs as electrolytes for CP actuators that can work in air [3–8]. The actuator

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

271

c18.indd 271

1/20/2011 10:28:24 AM

272

ACTUATORS (a)

(b) N H

(c) N H

S n

N H

n

n

Figure 18.1. Chemical structures of polypyrrole (a), polythiophene (b), and polyaniline (c).

Figure 18.2. Schematic drawing of the conducting polymer actuator system example.

performance was significantly improved in the ILs compared with that of a conventional organic electrolyte [3,4]. Solid-state actuators were developed using polymer electrolyte gels using ILs [5–7]. Ionic polymer-metal composites (IPMCs), which consist of ionic polymer plated with metal electrodes, are another promising type of EAP actuator [9]. When applying a voltage, the counter cation moves the cathode side with a dragging solvent, which results in the pressure gradient for the ionic gel polymer as shown in Figure 18.3. Ionic polymers usually used for the IPMCs are fluorinated ion-exchange resins such as perfluorosulfonic acid or perfluorocarboxylic acid polymers (Figure 18.4). The IPMC actuator can work when absorption of the solvent is fully absorbed in the fluorinated exchange resins. Usually, water is their solvent. Several reports change the solvent for the fluorinated ion-exchange resins to ILs to make the IPMC actuators operable in air for a long time [10–13]. The discovery of carbon nanotubes (CNTs) has stimulated intense research activities, leading to remarkable scientific and technological advances in the development of nanostrutured materials. The CNT consists of a rolled-up

c18.indd 272

1/20/2011 10:28:24 AM

273

ACTUATORS

Figure 18.3. Schematic drawing of the structure and electromechanical response mechanism of the IPMC.

-(CF2-CF2)n-(CF-CF2)O CF2 CF-CF3 O (CF2)y X

Figure 18.4. Chemical structure of prefluorinated ion-exchange polymer. X = SO−3 : perfluorosulfonic acid polymer, X = COO− perfluorocarboxylic acid polymer.

graphene sheet with hexagonally arranged SP2-hybridized carbon atoms. Because CNTs provide extraordinary mechanical, electrical, and thermal properties, numerous potential applications are expected. Baughman et al. [14] reported that single-walled carbon nanotube (SWNT) paper sheets show an electrochemical response when applying a low voltage against the counter electrode in aqueous electrolyte solutions. They proposed the “charge injection” model in which the actuation is attributed to the dimensional changes in

c18.indd 273

1/20/2011 10:28:24 AM

274

ACTUATORS

Figure 18.5. Schematic drawing of the bucky gel actuator.

covalently bonded directions caused by the charge injection, which originates from quantum and double-layer electrostatic effects. Contrary to various advantages of the CNTs for the applications, one major drawback is that CNTs are hard to process because of their poor dispersibility. Fukushima et al. [15] reported imidazolium-ion-based ionic liquids as a new class of dispersants of CNTs. They found that a suspension comprising singlewalled carbon nanotubes and an imidazolium-based ionic liquid was converted to a gel after mixing by grinding the suspension in an agate mortar. The gel was called bucky gel, which has both electronic and ionic conduction. It is easy to process into any shape. By using a bucky gel, we developed a dry actuator (bucky gel actuator), which simply can be fabricated by casting. Figure 18.5 shows the schematic drawing of the bucky gel actuator, which adopts a bimorph configuration with a polymer-supported internal ionic liquid electrolyte layer, sandwiched by polymer-supported bucky gel electrode layers [16]. The actuator can be operated in air at voltages lower than 3 V. The actuator can be readily fabricated by layer-by-layer casting of the electrode and by electrolyte components in gelatinous PVdF(HFP)-containing solvent. In a typical example, the bucky gel electrode layers include 13 wt% SWNTs, 54 wt% BMIBF4, and 33 wt% PVdF(HFP), and the electrolyte layer contains 67 wt% of BMIBF4 and 33 wt% of PVdF(HFP). The performance of the actuator has been improved greatly by developing the bucky gel electrode layer with a high content of SWNTs that are well dispersed in the electrode layer [17]. ILs are critical components for determining the performance of the bucky gel actuator. Seven kinds of ionic liquids, as summarized in Table 18.1, were studied for

c18.indd 274

1/20/2011 10:28:24 AM

275

ACTUATORS

TABLE 18.1. Ionic Liquids Used for Testing the Performance of the Bucky Gel Actuator [18] Ionic Liquid

Cation

Anion R = C2H5 R = C4H9 R = C6H13 R = C8H17

EMIBF4 BMIBF4 HMIBF4 OMIBF4

BF4−

EMITFSI

(CF3SO2)2N−

A-3

BF4−

A-4

(CF3SO2)2N−

TABLE 18.2. Double-Layer Capacitance of the Bucky Gel Electrode, the Conductivity of Ionic Gel Electrolyte Layer, and the Generated Strain of the Bucky Gel Actuator Containing Each IL Ionic Liquid EMIBF4 BMIBF4 HMIBF4 OMIBF4 EMITFSI A-3 A-4

CSWNT (Fg−1)

κ (mScm−1)

εo (%)

45.5 45.3 44.8 44.6 49.2 35.4 41.6

1.71 0.312 0.134 0.014 1.33 0.393 0.729

0.53 0.45 0.48 0.60 0.21 0.17 0.13

testing the frequency response of the bucky gel actuators [18]. Table 18.2 summarizes the double-layer capacitance of the bucky gel electrode, the conductivity of ionic gel electrolyte layer, and the generated strain of the bucky gel actuators containing each IL. The response of the actuator is affected by the ionic conductivity of the ionic gel. The static-generated strain εo depends on the spcies of the ILs. The performance of the bucky gel actuator depends on the ionic conductivity and sizes of the ILs. From these results, the electromechanical response of the bucky gel actuator is speculated as shown in Figure 18.6. When the voltage is applied, the dimensional changes of both electrode layers because of the transfer of ions in addition to the charge injection give the synchronous effect to the bending motion of the bucky gel actuator. Kiyohara and Asaka [19] studied the system comprising of porous electrode and electrolytes, using the Monte Carlo simulation in the constant voltage ensemble. It was found that that the osmotic stress is induced in the porous

c18.indd 275

1/20/2011 10:28:24 AM

276

ACTUATORS

Figure 18.6. Schematic drawing of the actuation mechanism of the bucky gel actuator.

electrode when voltage is applied. It also was found that the induced osmotic stress depends on the ionic sizes and the pore dimension of the porous electrode. The simulation result can be applied to the experimental resutls of the bucky gel actuator. One disadvantage of the bucky gel actuator is that the supporting polymer, essential for the actuator strip to possess a sufficient mechanical robustness, lowers the electrical conductivity as well as the capacitance of the electrode layers. A second-generation SWNT actuator has been developed comprising polymer-free nanotube electrodes, which has a much better performance than the previously reported “dry” actuator, including the first generation design [20] as shown in Figure 18.7. This achievement was possible because of the interesting finding that millimeter-long “super-growth” carbon nanotubes (SG-SWNTs), produced by the water-assisted modified CVD method [21], associate together tightly with ionic liquids, affording a free-standing sheet with superb conductivity. The SG-SWNT actuator can respond to an applied sinusoidal voltage with a frequency of more than 100 Hz. Biso and Ricci [22] developed the bucky gel actuator using multiwalled carbon nanotubes. A strain of 1% was achieved, which was comparable with the SWNT bucky gel actuator. Saito et al. [23] developed a photocurable ionic gel for fabricating the EAP actuator using electrodes comprising of activated carbon and acetylene black as well as ionic gel that consisted of EMITFSI and

c18.indd 276

1/20/2011 10:28:24 AM

277

REFERENCES

(a)

(b)

Figure 18.7. (a) High conductive electrode film comprising SG–SWNT and EMITFSI. (b) Bending response of the actuator comprising two SG–SWNT/EMITFSI electrodes sandwiching an EMITFSI/PVdF(HFP)-ionic gel layer in response to 3 V applied voltage by two AAA batteries.

PVdF(HFP). They fabricated a microgripper using two ionic gel actuators and demonstrated the grasp of a microobject in air. In this chapter, the EAP actuators using ILs as electrolytes are introduced. ILs are an essential component for the EAP actuators based on EC processes such as CPs, IPMCs, and CNTs, which can be operated in air. The performance of the actuator depends on the ILs such as their conductivity, mobility, ionic sizes, and so on. To develop a higher performance of the EAP actuators, ILs that are optimized for the EAP actuators should be developed in the near future.

REFERENCES 1. Y. Bar-Cohen, ed., Electroactive Polymer (EAP) Actuators as Artificial Muscles: Reality, Potential, and Challenges, 2nd ed., SPIE Press, 2004. 2. (a) E. Smela, Adv. Mater., 15, 481 (2003). (b) J. D. Madden, in Electroactive Polymers for Robotic Applications, K. J. Kim, and S. Tadokoro, eds., Springer, 2007, p. 121. (c) T. F. Otero, in Handbook of Conducting Polymers, 3rd Ed., T. A. Skotheim, and J. R. Reynolds, eds., CRC Press, 2007. (d) G. M. Spinks, G. Allici, S. McGovern, B. Xi, G. G. Wallace, in Biomedical Applications of Electroactive Polymer Actuators, F. Carpi, and E. Smela, eds., Wiley 2009, p. 195. 3. W. Lu, A. G. Fedeev, B. Qi, E. Smela, B. R. Mattes, J. Ding, G. M. Spinks, J. Mazurkiewicz, D. Zhou, G. G. Wallace, D. R. MacFarlane, S. A. Forsyth, M. Forsyth, Science, 297, 983 (2002). 4. J. Ding, D. Zhou, G. Spinks, G. Wallace, S. Forsyth, M. Forsyth, D. MacFarlane, Chem. Mater., 15, 2392 (2003).

c18.indd 277

1/20/2011 10:28:25 AM

278

ACTUATORS

5. D. Zhou, G. M. Spinks, G. G. Wallace, C. Tiyapiboonchaiya, D. R. MacFarlance, M. Forsyth, J. Sun, Electrochim. Acta, 48, 2355 (2003). 6. (a) C. Plesse, F. Vidal, H. Randriamahazaka, D. Teyssie, C. Chevrot, Polymer, 46, 7771 (2005). (b) F. Vidal, C. Plesse, G. Palaprat, A. Khaddar, J. Citerin, D. Teyssie, C. Chevrot, Synth. Met., 156, 1299 (2006). 7. (a) M. S. Cho, H. J. Seo, J. D. Nam, H. R. Choi, J. C. Koo, K. G. Song, Y. Lee, Sensor Actuator B chem., 119, 621 (2006). (b) M. Cho, H. Soo, J. Nam, H. Choi, J. Koo, Y. Lee, Sensor and Actuator B chem., 128, 70 (2007). 8. K. Yamato, K. Tominaga, Y. Takashima, K. Kaneto, Synth, Met. 159, 839 (2009). 9. (a) M. Shahinpoor, K. J Kim, Smart Mater. Struct., 10, 819 (2001). (b) K. Asaka, K. Oguro, in Biomedical Applications of Electroactive Polymer Actuators, F. Carpi and E. Smela, eds., Wiley 2009, p. 103. 10. M. D. Bennett, D. J. Leo, Sensor Actuator A chem., 115, 79 (2004). 11. B. J. Akle, M. D. Bennett, D. J. Leo, Sensor Actuator A chem., 126, 173 (2006). 12. J. Wang, C. Xu, M. Taya, Y. Kuga, Smart Mater. Struct., 16, S214 (2007). 13. J. -W. Lee, Y. -T. Yoo, Sensor Actuator B chem., 137, 539 (2009). 14. R. H. Baughman, C. Cui, A. A. Zakhidov, Z. Iqbal, J. N. Barisci, G. M. Spinks, G. G. Wallace, A. Mazzoldi, D. DeRossi, A. G. Rinzler, O. Jaschinski, S. Roth, M. Keertesz, Science, 284, 1340 (1999). 15. (a) T. Fukushima, A. Kosaka, Y. Ishimura, T. Yamamoto, T. Takigawa, N. Ishii, T. Aida, Science, 300, 2072 (2003). (b) T. Fukushima, Polymer J., 38, 743 (2006). (c) T. Fukushima, T. Aida, Chem. Eur. J., 13, 5048, 2007. 16. T. Fukushima, K. Asaka, A. Kosaka, T. Aida, Angew. Chem. Int. Ed., 44, 2410 (2005). 17. K. Mukai, K. Asaka, K. Kiyohara, T. Sugino, I. Takeuchi, T. Fukushima, T. Aida, Electrochim Acta, 53, 5555 (2008). 18. (a) I. Takeuchi, K. Asaka, K. Kiyohara, T. Sugino, N. Terasawa, K. Mukai, T. Fukushima, T. Aida, Electrochim. Acta, 54, 1762 (2009). (b) I. Takeuchi, K. Asaka, K. Kiyohara, T. Sugino, N. Terasawa, K. Mukai, S. Shiraishi, Carbon, 47, 1373 (2009). 19. (a) K. Kiyohara, K. Asaka, J. Chem. Phys., 126, 214704 (2007), (b) K. Kiyohara, K. Asaka, J. Phys. Chem. C, 111, 15903 (2007). (c) K. Kiyohara, T. Sugino, K. Asaka, J. Chem. Phys., 132, 144705 (2010). 20. K. Mukai, K. Asaka, T. Sugino, K. Kiyohara, I. Takeuchi, N. Terasawa, D. N. Futaba, K. Hata, T. Fukushima, T. Aida, Adv. Mater., 21, 1582 (2009). 21. K. Hata, D. N. Futaba, K. Mizuno, T. Namai, M. Yumura, S. Iijima, Science, 306, 1363, 2004. 22. M. Biso, D. Rissi, Phys. Status Solidi B, 246, 2820 (2009). 23. S. Sato, Y. Katoh, H. Kokubo, M. Watanabe, S. Mauro, J. Micromech. Microeng., 19, 035005 (2009).

c18.indd 278

1/20/2011 10:28:25 AM

PART IV FUNCTIONAL DESIGN

c19.indd 279

1/20/2011 10:28:28 AM

19 NOVEL FLUOROANION SALTS Rika Hagiwara and Kazuhiko Matsumoto

Since the first report of 1-ethyl-3-methylimidazolium fluorohydrogenate, EMIm(FH)2.3F [1], a series of fluorohydrogenate ionic liquids (ILs) have been synthesized and characterized by several techniques. Fluorohydrogenate ILs contain fluorohydrogenate anions formulated as (FH)nF− (Figure 19.1) and exhibit a low viscosity and high conductivity (e.g., 102 mS cm−1 at 298 K for EMIm(FH)2.3F). Physical, chemical, and structural properties [2–11] as well as applications as electrolytes [12–15] and reaction media [16–18] have been studied because of their interesting properties. Recent progress in chemistry and electrochemistry of some fluorohydrogenate ILs will be briefly reviewed in this section. Fluorohydrogenate ILs are prepared by metathesis of starting chloride and anhydrous hydrogen fluoride. The following scheme is an example of 1-alkyl3-methylimidazolium fluorohydrogenate, RMIm(FH)2.3F (R = n-alkyl group) [2]: (19.1) The IL with high purity is obtained by the elimination of volatile by-product HCl with excess HF under vacuum. The use of bromide or iodide precursor is preferably avoided because of the difficulty in complete elimination of HBr or HI compared with that of HCl from the resulting IL. Another solventless route was reported recently for the synthesis of fluorohydrogenate ILs [19]. This method does not involve halide methathesis and the products are Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

281

c19.indd 281

1/20/2011 10:28:28 AM

282

NOVEL FLUOROANION SALTS

Figure 19.1. Structures of FHF−, (FH)2F−, and (FH)3F−.

Figure 19.2. Organic cations that form room-temperature ILs with fluorohydrogenate anions. (a) Dialkylimidazolium, (b) alkylpyridinium, (c) dialkylpyrrolidinium, (d) dialkylpiperidinium, (e) dialkylmorpholinium, and (f) tetraalkylphosphonium.

prepared via a carboxylated-dialkylimidazolium zwitterion. The noninteger HF composition, 2.3, occurs for most vacuum-stable fluorohydrogenate ILs at room temperature regardless of the cationic structure. The organic cations that have been found to form room-temperature ILs with (FH)nF− are summarized in Figure 19.2 [2,5–7]. Infrared spectroscopy is the best way to identify (FH)nF−, where the ratio of FHF−, (FH)2F−, and (FH)3F− in fluorohydrogenate ILs depends on the HF composition n [20–23]. For example, RMIm(FH)2.3F contains (FH)2F− and (FH)3F− in the ratio of 7 : 3. Because the 1H- and 19F-NMR spectra give single signals and do not distinguish the hydrogen and fluorine atoms in these two anions even at 233 K, HF is thought to exchange among the anions in shorter periods than the nuclear magnetic resonance (NMR) resolution time scale [2]. No free neutral HF is detectable in the fluorohydrogenate ILs with an n value of lower than 2.3 by spectroscopic methods. There is still no clear explanation for this “magic number,” 2.3. Because of the absence of free HF, these ILs do not etch borosilicate glass container if they are thoroughly dried.

c19.indd 282

1/20/2011 10:28:28 AM

NOVEL FLUOROANION SALTS

283

The (FH)nF− anions are crystallographically characterized in numerous inorganic and organic salts, including dialkylimidazolium salts. The bifluoride salt EMImFHF, prepared by the thermal decomposition of EMIm(FH)2.3F, is a colorless crystalline solid at room temperature [24] and melts at 324 K. In this salt, cations form a pillar by making hydrogen bonds between the ring proton and π electrons of the adjacent cation rings, as well as with the coplanar bifluoride ions. The bifluoride ions also form a pillar along the same direction as that of the cations, sandwiching the alkyl side-chains of the cations as dielectric spacers. As a result, these cations and anions form an interesting layered structure in which each sheet is two-dimensionally connected by hydrogen bonds between the cations and anions. The bent (FH)2F− and trigonal planar (FH)3F− were observed in their 1,3-dimethylimidazolium (DMIm) salts, in which one-dimensional ordering of cation or anion occurs again in both cases (Figure 19.3) [25]. High-energy X-ray diffraction and molecular dynamics studies on fluorohydrogenate ILs also indicated an ordering of the cations and anions in their liquid state [8–10,26,27]. The reaction of EMImFHF with a stoichiometric amount of anhydrous HF gives EMIm(FH)nF (1 < n ≤ 2.3) with negligibly small HF dissociation pressures at room temperature. The phase diagram of EMImF–HF was constructed in this composition range [3]. The melting point of EMIm(FH)nF is lower than 298 K at n > 1.3 and is lower than 223 K at n > 1.7. The viscosity, and therefore

Figure 19.3. Crystal structures of (a) DMIm(FH)2F and (b) DMIm(FH)3F.

c19.indd 283

1/20/2011 10:28:29 AM

284

NOVEL FLUOROANION SALTS

TABLE 19.1. Physical Properties of Selected Fluorohydrogenate ILs Saltsa DMIm(FH)2.3F [2] EMIm(FH)2.3F [2] BMIm(FH)2.3F [2] AMIm(FH)2.3F [6] EMPyr(FH)2.3F [5] AMPyr(FH)2.3F [6] MOMMPyr(FH)2.3F [13] EMPip(FH)2.3F [5] AMPip(FH)2.3F [6] BPy(FH)2.3F [6] APy(FH)2.3F [6] AMMor(FH)2.3F [6] P4441(FH)2.3F [7]

Tm/K

Tg/K

Densityb/ g cm−3

Viscosityb/ cP

Conductivityb/ mS cm−1

272 208 — — — 201 — 217, 237 — — 203 — 249

— 148 154 — — — — — — — — — 169

1.17 1.13 1.08 1.11 1.07 1.05 1.13 1.07 1.06 1.09 1.12 1.16 0.969

5.1 4.9 19.6 5.5 9.9 8.5 7.9 24.2 16.7 29.7 5.3 34.6 36

110 100 33 90 75 78 75 37 63 25 82 35 6.0

a DMIm: 1,3-dimethylimidazolium; EMIm: 1-ethyl-3-methylimidazolium; BMIm: 1-butyl-3methylimidazolium; AMIm: 1-allyl-3-methylimidazolium; EMPyr: 1-ethyl-1-methylpyrrolidinium; AMPyr: 1-allyl-1-methylpyrrolidinium; MOMMPyr: 1-methoxymethyl-1-methylpyrrolidinium; EMPip: 1-ethyl-1-methylpiperidinium; AMPip: 1-allyl-1-methylpiperidinium; BPy: 1butylpyridinium; APy: 1-allylpyridinium; AMMor: 1-allyl-1-methylmorphlinium; P4441: tri-nbutyl-methylphosphonium. b Density, viscosity and conductivity are the values at 298 K.

the conductivity, are significantly affected by n. The conductivity of EMIm(FH)nF monotonously decreases with decreasing n. The electrochemical window is approximately 3 V, not sensitive to n. The physical properties of selected fluorohydrogenate ILs are shown in Table 19.1. Fluorohydrogenate ILs based on RMIm+, in general, exhibit higher conductivities than those based on nonaromatic heterocyclic ammonium or tetraalkylphosphonium cation [2,5–7]. Figure 19.4 shows a Walden plot (plot of reciprocal molar conductivity against viscosity) for the alkylimidazolium ILs including fluorohydrogenate salts (hatched points). Logarithmic scales are used for both axes because the values are distributed over two orders of magnitude. The linear relation observed between these parameters suggests that ionic conduction follows Walden’s rule. Namely, exceptionally high conductivities of fluorohydrogenate ILs are explained by low viscosities of them without introducing an ion hopping mechanism. A pulsed gradient spin-echo NMR measurement for RMIm(FH)2.3F revealed that the self-diffusion coefficient of (FH)nF− is larger than that of RMIm+ in the fluorohydrogenate ILs, whereas the self-diffusion coefficient of the cation is larger than that of the anion in other ILs such as EMImBF4 [4]. The electrochemical windows of the imidazolium-based RMIm(FH)2.3F are 3–3.5 V (Figure 19.5), with the extension of the alkyl side-chain of the cation

c19.indd 284

1/20/2011 10:28:29 AM

285

NOVEL FLUOROANION SALTS

log(l–1 / S–1 cm–2 mol)

1 26 25 23 24

0.5

21 17 22 20 12 19 16 1413 18 10 15 11 7 6 8 9

0

–0.5

4 5 –1

3 2

1 –1.5 0.5

1

1.5

2

2.5

3

log(η/ cP) Figure 19.4. Walden plot of room temperature ILs comprising alkylimidazolium cation and fluoroanions. (1) EMIm(FH)2.3F, (2) DMIm(FH)2.3F, (3) PrMIm(FH)2.3F, (4) BMIm(FH)2.3F, (5) PeMIm(FH)2.3F, (6) HMIm(FH)2.3F, (7) 1,3-diethylimidazolium bis(trifluoromethylsulfonyl)amide, (8) EMImBF4, (9) DMImN(SO2CF3)2, (10) 1-ethyl3,4-dimethylimidazolium bis(trifluoromethylsulfonyl)amide, (11) 1,3-dimethyl-4methylimidazolium bis(trifluoromethylsulfonyl)amide, (12) EMImOCOCF3, (13) 1,3-diethylimidazolium triflate, (14) 1,3-diethylimidazolium trifluoromethylcarboxylate, (15) 1-ethyl-3,4-dimethylimidazolium triflate, (16) BMImN(SO2CF3)2, (17) 1etoxymethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide, (18) 1-ethyl-2,3dimethylimidazolium bis(trifluoromethylsulfonyl)amide, (19) BMImOCOCF3, (20) 1isobutyl-3-methylimidazolium bis(trifluoromethylsulfonyl)amide, (21) BMImOCOCF3, (22) 1-butyl-3-ethylimidazolium trifluoromethylcarboxylate, (23) BMImOCOCF2CF2CF3, (24) 1-(2,2,2-trifluoromethyl)-3-methylimidazolium bis(trifluoromethylsulfonyl)amide, (25) 1-butyl-3-ethylimidazolium perfluorobutylsulfate, and (26) BMImOSO2CF2CF2CF2CF3. Viscosities and conductivities of the salts are summarized in the reference [28].

slightly increasing the window mainly by the shift of cathodic limit toward negative potentials [2]. In comparison with RMIm(FH)2.3F, the electrochemical windows of nonaromatic cation-based fluorohydrogenate ILs are significantly extended mainly by the shift of the anodic limit toward positive potentials, resulting in the electrochemical window of more than 5 V on a glassy carbon electrode (Figure 19.5) [5–7]. Some fluorohydrogenate ILs were studied as electrolytes in electrochemical devices. Electrochemical capacitors using fluorohydrogenate ILs show a large voltage dependence of capacitance as shown in Figure 19.6, resulting in a

c19.indd 285

1/20/2011 10:28:29 AM

286

NOVEL FLUOROANION SALTS

Figure 19.5. Electrochemical windows of (a) EMIm(FH)2.3F, (b) EMPip(FH)2.3F, and (c) EMPyr(FH)2.3F. W.E.: GC disk, C.E.: Pt plate, Scanning rate: 10 mV s−1. Potentials are referenced to the potential of ferrocene/ferrocenium redox couple in each salt.

Figure 19.6. Voltage dependence of capacitance for EMIm(FH)2.3F. (a) 298 K, (b) 273 K, (c) 263 K, (d) 253 K, (e) 243 K, and (f) 233 K.

significantly high capacitance at high charging voltages (e.g., 178 F g−1 for DMIm(FH)2.3F at 298 K compared with 42 F g−1 for EMImBF4 in the same cell) [12,13]. Charge–discharge tests were performed successfully at low temperatures, although the capacitance decreases with a decrease in temperature (Figure 19.6). Fluorohydrogenate ILs are applied also as an electrolyte in fuel

c19.indd 286

1/20/2011 10:28:29 AM

287

NOVEL FLUOROANION SALTS

cells because the (FH)nF− anion can work as a hydrogen carrier in the electrolyte [15]. Such fuel cells do not require humidification at elevated temperatures and enables operation at temperatures higher than 373 K without losing electrolyte solvent by evaporation. Cell performance was studied using composite electrolytes of fluorohydrogenate ILs with polymers [14]. The output power of 50 mWcm−2 has been obtained recently with the cell test at 120 °C under unhumidified condition using EMPyr(FH)1.7F-HEMA composite electrolyte [29]. Fluorohydrogenate ILs are good (FH)nF− sources in various reactions because of their easy handling. For example, the ET-based cation radical salt α’-(ET)2(FH)2F (ET: bis(ethylenedithio)tetrathiafulvalene) was electrocrystallized from tetrahydrofuran/EMIm(FH)2.3F. This salt shows semiconducting behavior with an activation energy of 0.05–0.08 eV [30]. The fluorination of organic compounds with fluorohydrogenate ILs as mild HF sources is another interesting application [17]. Halofluorination of alkenes, ring opening fluorination of epoxides, and deprotection of silylether were demonstrated [16–18]. Mesophases derived from ionic liquids attract more and more attention as functional materials including electrolytes. Ionic plastic crystals based on alkylpyrrolidinium cation and fluorohydrogenate anions exhibit high conductivities around room temperature without additives (101 mS cm−1) [31]. An X-ray diffraction study revealed that free rotation of the component ions results in the formation of a NaCl-type lattice. Ionic liquid crystals are observed for RMIm(FH)nF when the carbon number of the alkyl group R is greater than or equal to 10 [32]. Polarized optical microscopy and X-ray diffraction suggest the formation of a smectic A2 liquid crystalline mesophase. Anisotropic ion conduction was confirmed for the liquid crystalline phase as shown in Figure 19.7. For C12MIm(FH)2F, the ionic conductivity parallel to the smectic layer is more than 10 times larger than that perpendicular to the layer. Metathesis of a starting halide and a salt containing the corresponding anion is the most popular and simple method to prepare ILs. Complete elimination of the reaction by-product is often difficult for hydrophilic ILs because of the involatile nature of ILs, unlike hydrophobic ILs that can be washed with water. Fluorohydrogenate ILs are synthesized with high purities by the reaction described in equation (19.1) and so act as good fluorobases against Lewis acidic fluorides or oxofluorides to give room-temperature IL fluorometallates with high purities (equation 19.2) [33,34]. Volatile HF is again easily eliminated from the reaction product after the reaction.

R

N +N

Me

(FH)2.3F–

MFn –HF

R

N+N

MFn+1– Me

(19.2)

This Lewis acid–base reaction is usually highly exothermic, and precautions are necessary. Some physical properties of 1-ethyl-3-methylimidazolium salts

c19.indd 287

1/20/2011 10:28:29 AM

288

NOVEL FLUOROANION SALTS

Conductivity / mS cm–1

102

101

100

10–1 2.9

3.0

3.1 3.2 T –1 × 1000 / K–1

3.3

3.4

Figure 19.7. Arrhenius plots of ionic conductivity for the ionic liquid crystal phase of 1-dodecyl-3-methylimidazolium fluorohydrogenate (C12MIm(FH)2F). □: Ionic conductivity perpenducular to the smectic layer, and ○: ionic conductivity parallel to the smectic layer.

TABLE 19.2. Selected Physical Properties of EMIm Salts Based on Fluorometallate Anions Tm/K

Densitya/g cm−3

Conductivitya/ mS cm−1

Viscositya/cP

288 333 326 283 272 275 258 348 253 — 284

1.27 1.56 1.78 1.85 1.67 2.17 2.27 — 2.25 1.76 2.43

13.0 — — 6.2 8.5 7.1 3.2 — 3.0 5.1 7.9

34 — — 67 49 51 171 — 105 86 59

EMImBF4[35] EMImPF6[36] EMImAsF6[34] EMImSbF6[34] RMImNbF6[34] EMImTaF6[34] EMImWF7[34] EMImVOF4[37] EMImWOF5[38] EMImMoOF5[37] EMImUF6[39] a

These values are at 298 K.

prepared by this method are shown in Table 19.2. EMIm cation is stable against SbF6− ; however, SbF5 oxidizes the cation. Moreover, if excess SbF5 is supplied, then contamination of the salt of dimeric anion, Sb2 F11− , occurs. Therefore, conventional metathesis might be exceptionally superior to synthesize EMImSbF6 from the viewpoint of purity. Hexafluorometallate salts of

c19.indd 288

1/20/2011 10:28:29 AM

NOVEL FLUOROANION SALTS

289

Figure 19.8. Crystal packings of (a) EMImAsF6 and (b) EMImNbF6.

group 5 transition metal (Nb and Ta) were synthesized with the reaction of EMIm(FH)2.3F and solid pentafluorides. Crystal structures of the 1-ethyl-3methylimidazolium hexafluorometallate (PF6, AsF6, SbF6, NbF6 and TaF6) salts are divided into the following groups: EMImPF6-type structure (PF6, AsF6, and SbF6 salts) and EMImNbF6-type structure (NbF6 and TaF6 salts) (Figure 19.8), where the SbF6, NbF6, and TaF6 salts are liquids at room temperature [34,36,40,41]. Almost the same atomic sizes of Nb and Ta give similar physical properties of those ILs listed in Table 19.2, except for their densities, which of course are a result of the difference of mass. EMImWF7 is the first example in which heptacoordinated tungsten species are present in the liquid state, and gives EMImWOF5 by hydrolysis [38]. The low melting salts of the oxofluorocomplex anions, VOF4− , MoOF5− , and WOF5− , were also prepared by the acid–base reactions of the fluorohydrogenates and oxofluorides, VOF3, MoOF4, and WOF4, respectively [37]. X-ray single-crystal diffraction revealed that VOF4− forms a dimeric anion in EMImVOF4 (Figure 19.9) and that this dimer is preserved in the liquid state. Hydrolysis of the butylpyridinium salt of MoOF5− gives a new dinuclear anion Mo2O4 F62− with two Mo–O–Mo bridgings (Figure 19.9). The reactions of fluorohydrogenate ILs with UF5 yield ILs based on UF6− [39]. The green-colored room-temperature IL, EMImUF6, has a melting point of 284 K. One possible application of this IL is isotope separation by electrophoresis.

c19.indd 289

1/20/2011 10:28:29 AM

290

NOVEL FLUOROANION SALTS

Figure 19.9. Structures of the (VOF4− )2 and Mo 2O4 F62− anions determined in EMImVOF4 and (BPy)2Mo2O4F6 (BPy = N-butylpyridinium), respectively.

REFERENCES 1. R. Hagiwara, T. Hirashige, T. Tsuda, Y. Ito, J. Fluorine Chem., 99, 1 (1999). 2. R. Hagiwara, K. Matsumoto, Y. Nakamori, T. Tsuda, Y. Ito, H. Matsumoto, K. Momota, J. Electrochem. Soc., 150, D195 (2003). 3. R. Hagiwara, Y. Nakamori, K. Matsumoto, Y. Ito, J. Phys. Chem. B, 109, 5445 (2005). 4. Y. Saito, K. Hirai, K. Matsumoto, R. Hagiwara, Y. Minamizaki, J. Phys. Chem. B, 109, 2942 (2005). 5. K. Matsumoto, R. Hagiwara, Y. Ito, Electrochem. Solid-State Lett., 7, E41 (2004). 6. M. Yamagata, S. Konno, K. Matsumoto, R. Hagiwara, Electrochem. Solid-State Lett., 12, F9 (2009). 7. S. Kanematsu, K. Matsumoto, R. Hagiwara, Electrochem. Commun., 11, 1312 (2009). 8. M. Salanne, C. Simon, P. Turq, J. Phys. Chem. B, 110, 3504 (2006). 9. M. Salanne, C. Simon, P. Turq, J. New Mater. Electrochem. Syst., 9, 291 (2006). 10. B. L. Bhargava, S. Balasubramanian, J. Phys. Chem. B, 112, 7566 (2008). 11. K. Matsumoto, R. Hagiwara, Electrochemistry, 73, 730 (2005). 12. M. Ue, M. Takeda, A. Toriumi, A. Kominato, R. Hagiwara, Y. Ito, J. Electrochem. Soc., 150, A499 (2003). 13. A. Senda, K. Matsumoto, T. Nohira, R. Hagiwara, J. Power Sourc., 195, 4414 (2010). 14. J. S. Lee, T. Nohira, R. Hagiwara, J. Power Sourc., 171, 535 (2007). 15. R. Hagiwara, T. Nohira, K. Matsumoto, Y. Tamba, Electrochem. Solid-State Lett., 8, A231 (2005). 16. H. Yoshino, S. Matsubara, K. Oshima, K. Matsumoto, R. Hagiwara, Y. Ito, J. Fluorine Chem., 125, 455 (2004). 17. H. Yoshino, K. Matsumoto, R. Hagiwara, Y. Ito, K. Oshima, S. Matsubara, J. Fluorine Chem., 127, 29 (2006). 18. H. Yoshino, K. Nomura, S. Matsubara, K. Oshima, K. Matsumoto, R. Hagiwara, Y. Ito, J. Fluorine Chem., 125, 1127 (2004).

c19.indd 290

1/20/2011 10:28:30 AM

REFERENCES

19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

c19.indd 291

291

C. Rijksen, R. D. Rogers, J. Org. Chem., 73, 5582 (2008). W. D. Chandler, K. E. Johnson, J. L. E. Campbell, Inorg. Chem., 34, 4943 (1995). J. H. Clark, J. Emsley, R. E. Overill, J. Chem. Soc., Dalton Trans., 1219 (1981). I. Gennick, K. M. Harmon, M. M. Potvin, Inorg. Chem., 16, 2033 (1977). T. von Rosenvinge, M. Parrinello, M. L. Klein, J. Chem. Phys., 107, 8012 (1997). K. Matsumoto, T. Tsuda, R. Hagiwara, Y. Ito, O. Tamada, Solid State Sci., 4, 23 (2002). T. Enomoto, K. Matsumoto, R. Hagiwara, J. Phys. Chem. C., submitted. R. Hagiwara, K. Matsumoto, T. Tsuda, Y. Ito, S. Kohara, K. Suzuya, H. Matsumoto, Y. Miyazaki, J. Non-Cryst. Solids, 312, 414 (2002). K. Matsumoto, R. Hagiwara, Y. Ito, S. Kohara, K. Suzuya, Nucl. Instrum. Meth. Phys. Res., 199, 29 (2003). R. Hagiwara, Electrochemistry, 70, 130 (2002). J. Xu, T. Nohira, R. Hagiwara, Meeting Abstract of 77th Electrochemical Socity of Japan Meeting, 222 (2010). Y. Yoshida, M. Sakata, G. Saito, K. Matsumoto, R. Hagiwara, Chem. Lett., 36, 226 (2007). R. Taniki, K. Matsumoto, R. Hagiwara, Meeting Abstract of 77th Electrochemical Socity of Japan Meeting, 345 (2010). F. Xu, K. Matsumoto, R. Hagiwara, Chem. Eur. J., 16, 12970 (2010). K. Matsumoto, R. Hagiwara, Y. Ito, J. Fluorine Chem., 115, 133 (2002). K. Matsumoto, R. Hagiwara, R. Yoshida, Y. Ito, Z. Mazej, P. Benkic, B. Zemva, O. Tamada, H. Yoshino, S. Matsubara, Dalton Trans., 144 (2004). J. Ohtsuki, K. Matsumoto, R. Hagiwara, Electrochemistry, 77, 624 (2009). J. Fuller, R. T. Carlin, H. C. Delong, D. Haworth, J. Chem. Soc., Chem. Commun., 299 (1994). T. Kanatani, K. Matsumoto, R. Hagiwara, Eur. J. Inorg. Chem., 1049 (2010). K. Matsumoto, R. Hagiwara, J. Fluorine Chem., 126, 1095 (2005). T. Kanatani, K. Matsumoto, R. Hagiwara, Chem. Lett., 38, 714 (2009). K. Matsumoto, R. Hagiwara, Z. Mazej, P. Benkic, B. Zemva, Solid St Sci., 8, 1250 (2006). T. Kanatani, R. Ueno, K. Matsumoto, T. Nohira, R. Hagiwara, J. Fluorine Chem., 130, 979 (2009).

1/20/2011 10:28:30 AM

20 NEUTRALIZED AMINES Hiroyuki Ohno

20.1

REQUIREMENT OF EASY PREPARATION

It is important to select a suitable ion species for salts with excellent properties. However, the ions must be with counter ions that are oppositely charged in ordinary conditions, which makes it impossible to directly mix suitable cations and anions. Therefore, the salt should be prepared by a certain method, for example, by mixing two salts that must contain the suitable cation and anion. Then the target salts are isolated from the mixture. Because the solubility (or miscibility) of different salts in molten salts must be taken into account, it is usually difficult to isolate the target salts. Of course, some salts can easily be separated. For example, hydrophobic salts such as some bis(trifluoromethanesulfonyl)imide (TFSI) salts can be separated from hydrophilic salts by washing the salt with water. Some other systems are known to separate with fairly good yield. It should be noted here that there is no systematic method to prepare any kind of salts from a mixture. This is still a serious problem of ionic liquid preparation, especially for the scientists who wish to create (or find) new ionic liquids. Salt synthesis is a helpful development for solving this problem. Acid esters are foremost among the salts used for this purpose. Tertiary amines are reacted with acid esters to introduce an alkyl group onto the nitrogen atom and to prepare the onium cation. The reacted acid then remains the counter anion. This method is well covered in several other books on ionic liquids [1,2]. Several acid esters are known to be effective for direct preparation. Recently,

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

293

c20.indd 293

1/20/2011 10:28:31 AM

294

NEUTRALIZED AMINES

even TFSI methyl ester was reported as an effective reagent for the preparation of an ionic liquid containing TFSI anion [3]. Although there are still many anions not suitable for this method, a strong interest has been voiced for an easy preparation method of the various ionic liquids. 20.2

NEUTRALIZATION METHOD

It is well known that amines can be neutralized with acids to generate salts. This neutralization process is useful in preparing organic salts with low melting points. In 1914, Walden reported [4] neutralized ethylamine with nitric acid to prepare a salt that had a melting point of 12 °C. This seems to have been the first ionic liquid prepared by a neutralization method [5]. As shown in equation (20.1), there is a slight difference between onium cations prepared by quaternization (a) and neutralization (b) when tertiary amines are used as a starting material: (a)

X-

R2X

R1

N

N

N R1

++

MZ

N

N

R1

R2

HZ

(b)

N

++

++

ZN

+ R2

MX

(20.1)

ZNH

R1

This neutralization reaction involves one step and generates nothing. The tertiary amine is dissolved in a suitable solvent like ethyl alcohol. Acid is mixed slowly into the ethyl alcohol solution. At the initial stage, it is better not to heat the solution to avoid esterification between ethyl alcohol and the acid. The reaction mixture is evaporated after suitable mixing at room temperature or on an ice bath. The concentrated solution is dried further using an evaporator. Then dehydrated diethyl ether is added to this solution to wash the salts. The phase-separated liquid or solid is further dried in vacuo for 2 days. The solvent and reaction conditions may vary, depending on the components. The objective is to obtain equimolar mixing, so any excess component is eliminated by evaporation. Because this method can be performed easily under mild conditions, it is the prefered way to prepare polymerizable ionic liquids, meaning ions containing vinyl groups. These polymerizable units are sensitive to heating, and undesired polymerization can occur during preparation of the monomer. Some examples can be found in Chapters 30–32. It should be noted here that water is also useful to neutralize amines with acids, and the resulting salts are easy to purify by heating them under reduced pressure. However, it also is noted that there is a certain vapor pressure if the neutralization was under equilibrium. In such a weak salt formation, there will be no product by heating under reduced pressure for a long time. This also inspired us to design a “distillable” ionic liquid.

c20.indd 294

1/20/2011 10:28:31 AM

295

NEUTRALIZATION METHOD

TABLE 20.1. Thermal Properties of Amines Neutralized with HBF4 Sample Number

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

Starting Amines

2-Methyl-1-pyrroline 1-Ethyl-2-phenylindole 1,2-Dimethylindole 1-Ethylcarbazole 2,4-Lutidine 2,3-Lutidine 3,4-Lutidine 2,6-Lutidine N,N′-Dimethylcyclohexylamine N,N′-Dimethylcyclo-hexan methylamine 1-Methylindole 1-Methylpyrrole 1-Methylpyrazole 1-Methylbenzimidazole 2,3-Dimethylindole 2-Methylindole Pyrrole Carbazole 1-Methylimidazole 1-Ethylpiperidine 1-Methylpyrrolidine

Neutralized Amines Tg (°C)

Tm (°C)

σi (S cm−1) at 50 °C

σi (S cm−1) at 25 °C

−94.3 −73.9 −74.8 −68.0 −44.8 — −33.3 −10.9 — −18.4

17.1 29.8 24.5 — 34.1 59.4 45.9 104.6 89.0 143.8

2.7 × 10−2 1.6 × 10−2 1.1 × 10−2 5.1 × 10−3 5.9 × 10−4 8.0 × 10−5 5.3 × 10−5 1.8 × 10−7 9.8 × 10−8 1.3 × 10−8

1.6 8.9 4.3 2.2 2.3 5.9 3.6 1.6 7.3

— −15.9 −109.3 — — — 0.1 59.9 — −77.2 —

— — −5.9 99.9 — 131.0 — — 36.9 164 −31.9

2.0 × 10−9 1.7 × 10−9 3.5 × 10−2 1.3 × 10−8 2.0 × 10−9 1.6 × 10−9 3.4 × 10−9 1.3 × 10−9 2.8 × 10−3 5.9 × 10−4 2.5 × 10−2

— 1.3 × 10−2 1.9 × 10−2 — — — 1.4 × 10−9 — 3.1 × 10−4 9.4 × 10−5 1.6 × 10−2

× 10−2 × 10−3 × 10−3 × 10−3 × 10−4 × 10−6 × 10−6 × 10−8 × 10−9 —

Table 20.1 shows the characteristics of several amines neutralized with HBF4. The BF4 salts cannot be prepared by the acid ester method and are water soluble. Also, pure BF4 salts are extremely difficult to prepare. However, it is easy to get salts by neutralizing the tertiary amines with HBF4. The examples mentioned here are prepared by neutralization with HBF4; it is also possible to do so with other acids. It is also important to mention here that the obtained neutralized amines are protic ionic liquids. These ionic liquids are expected to act as proton conductors or as a proton generator. This will be mentioned in later chapters, too. The amines neutralized with HBF4 are stable. They show no weight loss up to 250 °C, and the decomposition temperature depends on the starting tertiary amines. This is a remarkable improvement in stability. The melting point (Tm) of neutralized amines is not a simple function of the molar mass of the starting amines, however. The Tm of the starting amines can affect the Tm of the salts. Figure 20.1 shows the relation between Tm of the starting amines and corresponding salts. If the Tm of the starting amine dominates the Tm of the salt, then a straight line (or parallel lines) is obtained as shown in Figure 20.1. There

c20.indd 295

1/20/2011 10:28:31 AM

296

NEUTRALIZED AMINES 200 20

Tm of salts / °C

150

10

16

8

100

14 9

6 7 19

50

0

3

2

13 21

-50 –100

–50

0 50 Tm of amines / °C

100

Figure 20.1. Comparison of melting points for starting tertiary amines and ionic liquids obtained by their neutralization.

seems to be some relation between these two Tms. Therefore, for the preparation of neutralized amines with a low Tm, it is better to use starting amines with a low Tm. It should be mentioned that there are nevertheless some exceptions. Among these salts, there are two interesting cases—cases 2 and 3 in Figure 20.1. These two salts showed Tms lower than that of the starting amines. Generally, the Tm of an amine increases considerably after neutralization. The extraordinary drop in Tm evident in cases 2 and 3 cannot be attributed to the starting amine, indole. Furthermore, although these two salts are formed by indole derivatives, other indole derivatives, such as 11, 15, and 16, do not show such an unusual characteristic. A comparison of cases 3 and 16 shows the only difference to be in the N-substituted methyl group. The introduction of a methyl group lowered the Tm by more than 100 °C. In contrast, the introduction of an alkyl group is not always effective in lowering the Tm. The structure– property relationships of organic ionic liquids are more complicated. A study on these relationships should provide helpful information for their functional design.

20.3 EFFECT OF ION SPECIES ON THE PROPERTIES OF NEUTRALIZED AMINES An ion species has a considerable effect on the characteristics of salts as well as on ordinary ionic liquids. The series of amines shown in Table 20.1 was neutralized with HBF4. The salts, of course, showed considerable Tm differ-

c20.indd 296

1/20/2011 10:28:31 AM

IONIC CONDUCTIVITY

297

ences upon being treated by the acids. For example, the Tm of N-methylimidazole neutralized with HBF4 (case 19) is 37 °C, whereas that of N-methylimidazole neutralized with hydrochloric acid or nitric acid is around 70 °C. That of triflate salt showed a Tg at 83.5 °C. Clearly, it is important to select a suitable acid in preparing neutralized amines with relatively low Tms. Although the ability of acids to give excellent salts is roughly understood, there are many exceptions. For example, if instead of N-methylimidazole, N-ethylimidazole is neutralized with HBF4, then there is no change in Tm, whereas when it is neutralized with nitric acid, the Tm is 31 °C. Triflate salts show a Tm at 8 °C, whereas hydrochloric acid salts show a Tm at 57 °C. These and several other exceptions prevent us from establishing a reliable rule on the structure–property relationship of neutralized amines. Because the neutralization is a simple process and is not difficult to assess, many case studies have been carried out to attempt to summarize the basic findings that might lead to a generalization about the amine– salt relationship.

20.4

IONIC CONDUCTIVITY

Some neutralized amines are obtained as liquid. Analyses of their ionic conductivity results have provided some interesting results. Figure 20.2 shows the Arrhenius plots of the ionic conductivity of neutralized salts shown in Table 20.1. Generally, ionic conductivity is governed by the mobility of dissociated carrier ions. Tg is used to evaluate the mobility of the matrix. A matrix with low Tg usually has a high segmental motion at ambient temperature. Note that Tgs of neutralized amines are low—even below −80 °C. Accordingly, these salts generally show a high ionic conductivity. Some obtained salts show excellent ionic conductivity around 10−2 S cm−1 at 25 °C. Figure 20.2 provides the Arrhenius plots of ionic conductivity for these neutralized salts. The ionic conductivity of the salts of cases 1, 2, 3, 4, 13, and 21 is high, which can be expected from their Tm (and Tg, data not shown here). Among these salts, case 21 shows a high ionic conductivity over 10−2 S cm−1 at room temperature because of its low melting point. These salts keep their ionic conductivity high at low temperatures, as is shown in Figure 20.2. Some salts, such as case 5, show a considerable conductivity drop at Tm because of a phase change. Despite the moderate Tm at room temperature, a drop of ionic conductivity caused by a phase change was not observed for the other salts. This can be explained by a supercooling phenomenon that takes place below the melting points of salts. Figure 20.3 shows the relationships between ionic conductivity and the Tg or Tm of a series of neutralized amines. The salts with a lower Tgs and Tms show a higher ionic conductivity. The Tm is actually just one factor that governs the motion of ions, so care must be taken in describing this kind of relationship. Because the ionic conductivity plotted in this figure was measured at 50 °C, it dropped at around 50 °C. Otherwise, there is no particular meaning to the 50 °C. The deviations in the plots can be understood to be a result of the effect

c20.indd 297

1/20/2011 10:28:31 AM

298

NEUTRALIZED AMINES -1

1

13

-2

2 21 3

Ionic conductivity log(σi / S cm-1)

-3

4 19

-4

20 -5

5 6

-6

7

-7

14 10

-8

15

18 8

16 11

17

3.0

3.1

3.2

3.3

9

12

-9 3.4

3.5

3.6

1000 T-1 / K-1

Figure 20.2. Arrhenius plots of the ionic conductivity for neutralized amines shown in Table 20.1. -1

log(σ i / S cm-1) at 50 oC

-2

1 13

-3

20

Tg

13

21

4

2

-4 -5

1

3

20

5

2 19

3 5

6

7

7

Tm

-6 8

-7 -8

10

-9

12

-100

-50

8

9

10

14 17

16

18

0

50

100

150

o

Temperature / C

Figure 20.3. Relation between thermal behavior and the ionic conductivity for neutralized amines.

c20.indd 298

1/20/2011 10:28:31 AM

CONCLUSION

299

of supercooling. In the relationship between Tg and ionic conductivity, a clear decrease of ionic conductivity was observed in these neutralized salts with the increasing Tg. This is a clear indication that the ionic conductivity of ionic liquids is mainly governed by the glass transition temperature. A detailed discussion on the effect of Tg on the ionic conductivity can be found in a published paper [6].

20.5 MODEL SYSTEMS FOR ORDINARY QUATERNIZED ONIUM SALTS Although neutralized amines are useful as models of onium type salts, they are not classified among typical ionic liquids. We have already observed that there is a similarity between the thermal properties and the ionic conductivities of neutralized amines and quaternized onium salts [6]. Here, we investigate the relationship of ionic conductivity between neutralized and alkylated imidazolium ionic liquids with BF4− . Recall that the ionic conductivity of case 19 in Figure 20.2 was high (2.8 × 10−3 S cm−1 at 50 °C). Likewise, alkylated imidazolium salts have high ionic conductivities at temperatures above their melting points. The melting point and ionic conductivity of 1-ethyl-3methylimidazolium tetrafluoroborate, a typical room-temperature ionic liquid, is 15 °C and 1.4 × 10−2 S cm−1 at 25 °C [7]. To test this relationship, we prepared a series of dialkylated imidazolium salts and compared their ionic conductivity with that of neutralized amines that have the same structure except for the lack of one alkyl chain. We found a certain relation in their ionic conductivity, which indicated that the neutralized amines are good models for quaternized onium salts of the same structure. Therefore, alkylated amine salts, such as those of cases 1, 2, 3, 4, 13, and 21, should form ionic liquids and show a high ionic conductivity at room temperature. Already alkylated pyrrolidinium salts with (not BF4− but) TFSI− as the counter anion have been reported as roomtemperature molten salts with a high ionic conductivity [8]. Our method proved to be convenient for preparing a series of neutralized amines that allowed us to select suitable amines. If these amines are quaternized with alkylating agents, then they will form ionic liquids that have excellent properties.

20.6

CONCLUSION

Stable salts can be synthesized that have low melting points for electrochemical and other applications. The neutralization method introduced in this chapter is an easy way to prepare pure salts and to use them in learning about the structure–property relationships of salts. It has been shown that excellent onium salts can be obtained from the results of neutralized amines. Besides serving as sources of quaternized onium salts, neutralized amines have

c20.indd 299

1/20/2011 10:28:32 AM

300

NEUTRALIZED AMINES

excellent properties. Some of these are unique and not retained by ordinary onium salts. More vigorous studies on neutralized amines can be expected to be published in the near future including studies on protic properties.

REFERENCES 1. R. D. Rogers, K. R. Seddon, eds., Ionic Liquids—Industrial Applications for Green Chemistry, ACS Symposium Ser., 2002. 2. P. Wasserscheid, T. Welton, eds., Ionic Liquids in Synthesis, Wiley-VCH, 2003. 3. J. Zhang, G. Robert, D. D. DesMarteau, Chem. Comm., 2334 (2003). 4. P. Walden, Bull. Acad. Imper. Sci., 1800 (1914). 5. T. Welton, Chem. Rev., 99, 2071 (1999). 6. M. Hirao, H. Sugimoto, H. Ohno, J. Electrochem. Soc., 147, 4168 (2000). 7. J. Fuller, R. T. Carlin, R. A. Osteryoung, J. Electrochem. Soc., 144, 3881 (1997). 8. D. R. MacFarlane, P. Meakin, J. Sun, N. Amini, M. Forsyth, J. Phys. Chem. B, 103, 4146 (1999).

c20.indd 300

1/20/2011 10:28:32 AM

21 ZWITTERIONIC LIQUIDS Masahiro Yoshizawa-Fujita, Asako Narita, and Hiroyuki Ohno

Molten salts at room temperature, so-called ionic liquids [1,2], are attracting the attention of many researchers because of their excellent properties, such as a high ion content, liquid-state over a wide temperature range, low viscosity, nonvolatility, nonflammability, and high ionic conductivity. The current literature on these unique salts can be divided into the following areas of research: neoteric solvents as environmentally benign reaction media [3–8] and electrolyte solutions for electrochemical applications, for example, in lithium batteries [9–13], fuel cells [14–17], and solar cells [18–27]. Most electrochemical systems require target carrier ions such as the lithium cation, the proton, or the iodide for the construction of corresponding cells. In other words, a matrix that predominantly transports these target ions is required for electrochemical applications. Although ionic liquids show excellent ionic conductivity of more than 10−2 S cm−1 at room temperature, this derives from the migration of component ions [28–30]. These ions are mostly useless as target ions. Even when target ions were added to the ionic liquid, the ionic liquid component ions still migrated along the potential gradient. Continued addition of other salts induced increases in both Tg and viscosity, so the ionic conductivity was considerably reduced [31,32]. In recent years, some designs of ionic liquids in which the component ions cannot migrate along the potential gradient have been tried [33–35]. One such design is the zwitterion structure, in which both the cation and the anion are tethered [33]. Because a zwitterion has both a cation and an anion, it is expected not to migrate even under the potential gradient. Zwitterions are

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

301

c21.indd 301

1/20/2011 10:28:34 AM

302

ZWITTERIONIC LIQUIDS

O S R

O

O

N

N

N

N

N

N

SO3

O R

N

N

+

S O

O

N R´

R

O

SO3

R

SO3

SO3

N O

R'

Scheme 21.1. Synthesis of zwitterions with a sulfonate anion.

gradually getting attention as ion conductive matrices [33,36,37]. Although the zwitterions serve as an insulator, they show ionic conductivity values of 10−4– 10−7 S cm−1 at room temperature when alkali metal salts or super strong acids are added. Of course, their ionic conductivity depends on the structure and properties of both cation and anion in the zwitterion as well as those of the added salts. The analysis of the properties of zwitterions is still less developed. Previously, major efforts have been made to investigate ion conductive poly(zwitterion)s [38–42]. Details on the polymerized zwitterions can be found in Chapter 31. This chapter covers the relations between structures of zwitterions and both their thermal properties and their ionic conductivity after alkali metal salts are added. Structures of zwitterions are presented in the Appendix at the back of the book. Schemes 21.1–21.5 show the synthetic routes of zwitterions with different anion structures [43–49]. As shown in Scheme 21.1, zwitterions are obtained by one-step reactions of a tertiary amine and an alkylsultone. This reaction gives no by-products, so the purification is simple and there is no contamination of the zwitterions by small ions. This one-pot procedure is useful for the preparation of pure zwitterions as well as for the ionic liquids prepared by neutralization of tertiary amines and organic acids [29]. In addition, the zwitterions with various counter anions, such as carboxylate, dicyanoethenolate, sulfonylamide, and organoborate, have been prepared, as shown in Schemes 21.2–21.5. It is important to change the anion structure of zwitterions in studying the relationship between the zwitterion structure and its properties as well as for the optimization of such properties. Table 21.1 summarizes melting points (Tm) of zwitterions that have a sulfonate group as the counter anion. Most zwitterions melt above 100 °C, indicating an increase in the Tm after the tethering of cation with anion. Some decomposed before melting, namely zwitterions 1 and 2. A considerable increase in Tm would be caused by a decrease in the motion of each ion by

c21.indd 302

1/20/2011 10:28:34 AM

303

ZWITTERIONIC LIQUIDS

Scheme 21.2. Synthesis of zwitterions with a carboxylate anion.

O

R

N

O

CN

O

CN

O

CN

R

N

N

CN

O

CN

N +

CN R

O

CN

N

N

O

CN O

Scheme 21.3. Synthesis of zwitterions with a dicyanoethenolate anion.

tethering. Moreover, the intermolecular electrostatic interaction between oppositely charged groups was observed in a single crystal of 4-(1-butyl-1Himidazol-3-ium-3-yl)butane-1-sulfonate, as shown in Figure 21.1 [54]. When a sulfonate group acts as the counter anion, zwitterions 4, 5, and 14 showed relatively low Tm compared with those of other zwitterions. Because zwitterion 4 has a branched spacer structure, it can obstruct the intermolecular ionic association. Zwitterion 5 showed a lower Tm of 161 °C than those of ammonium zwitterions without an ether bond. This low value is obviously the effect of the ether bond because most ammonium zwitterions decomposed

c21.indd 303

1/20/2011 10:28:34 AM

304

ZWITTERIONIC LIQUIDS

3

R

Br

N

N

X n

H2N Cl

O

CF3

S

O

X = CO, SO2 Br R

O

N

N

X n

N

OH

ion exchange

R

X n

N

S

R

N

N

X n

OH R

CF3

NH

N

NH O

-2H2O

NH

N

O

N

R

R

CF3 NH

O

N

N

S

N

R

Cl

O

N

O S

2 CF3

R

N

N

Scheme 21.4. Synthesis of zwitterions with a sulfonylamide anion.

R

Br

N

N

O

B

Li Br R

N

N

O

B

F

F

F

F F

TFSI

+ LiTFSI

R

N

N

F B O + Li

F

F

F F

Scheme 21.5. Synthesis of zwitterions with an organoborate anion.

c21.indd 304

1/20/2011 10:28:35 AM

305

ZWITTERIONIC LIQUIDS

TABLE 21.1. Melting Points of Zwitterions with Sulfonate Number

Structure

Tm (°C) a

1 2 3 4 5 O

N

Reference

271.1 284.7a 293.5a 146.9 161

43 43 43 43 50

197

50

183

50

SO3

5 6 O

N

SO3

6 7 O

N

SO3

7 8 9 10 11 12 13 14 15 16 17 18 19

169–170 274.3 279.1a 216.5 178 175 158 190 240 227 248 224

51 43 43 43 52 52 52 53 53 53 53 53

a

Decomposition temperature.

around 300 °C before melting. In the case of zwitterion 14, the butylimidazolium cation also has a long side-chain, which can prevent the zwitterion from dense packing. However, zwitterions 16 and 18, which have a methyl group at the 2-position on the imidazolium cation ring, show higher Tm than zwitterions 15 and 17, respectively. Such a tendency agrees with that of simple ionic liquids, namely the steric effect of the 2-methyl group. Figures 21.2 and 21.3 show the differential scanning calorimetry (DSC) charts of zwitterions 12 and 13. In the first heating scan, only Tm was observed in both systems. In the second heating for zwitterion 12, the glass transition temperature (Tg), crystallization point (Tc), and Tm were found at 22 °C, 91 °C, and 177 °C, respectively. Moreover, zwitterion 13 also shows Tg, Tc, and Tm at

c21.indd 305

1/20/2011 10:28:35 AM

306

ZWITTERIONIC LIQUIDS

c a 0

b

Figure 21.1. Packing diagram, viewed along the a axis, showing the herring-bone structure and the cationic group.

14 °C, 134 °C, and 175 °C, respectively in the second heating process. According to these data, zwitterions can easily become amorphous. When the Tg of a zwitterion is around 20 °C, it is still solid at room temperature. It is thus necessary to analyze the effective and controllable factors that lead to a lowering of the Tm of zwitterions. The effect of anion structure on the thermal properties of the zwitterions is a good place to begin. Table 21.2 summarizes the Tm of zwitterions with a carboxylate anion. To analyze the effect of the anion structure on the Tm of zwitterions, we compared those with sulfonate anions and those with carboxylate anions. Note that the Tm values of zwitterions 12 and 21 are almost the same. However, zwitterion 22 shows a lower Tm than that of 13. However, one cannot conclude that the carboxylate group is more suitable for lowering the Tm of zwitterions from only these data. Table 21.2 shows that zwitterions that have a longer spacer length between the cation and the anion have a lower Tm (within the range of spacer length studied). This is likely the result of compensation for the increase in the flexibility of ions and the decrease in the ion density. Zwitterion 25, with a branched spacer structure, shows a lower Tm than that of zwitterion 20 and those of zwitterions that have a sulfonate group as a counter anion. It can be concluded that, because of the steric hindrance, the branched spacer is effective in lowering the Tm of zwitterions.

c21.indd 306

1/20/2011 10:28:35 AM

307

ZWITTERIONIC LIQUIDS

endo. ←

→exo.

(a)

Tm –100

–50

0

50

100

150

200

(b) Tc

endo. ←

→exo.

Tg

Tm –100

–50

0

50

100

150

200

Temperature / °C

Figure 21.2. DSC charts of zwitterion 12. (a) First heating; (b) second heating. (Reproduced with permission from CSIRO Publishing.)

Tables 21.3 and 21.4 summarize the reported Tm values of zwitterions comprising the onium cation and the dicyanoethenolate anion. Zwitterions with a dicyanoethenolate anion show a lower Tm than zwitterions that have a sulfonate anion. For instance, 1 and 2 decomposed before melting at 271 °C and 285 °C, whereas zwitterions 38 and 39 showed Tm’s at 187 °C and 148 °C. In addition, the Tm of 9 was 274 °C, whereas that of 32 was 174 °C. In particular, zwitterion 40 showed the lowest Tm of 70 °C among the systems listed in all five tables. Because a negative charge is delocalized by the electron-withdrawing effect of the cyano group, the interaction force between the cation and the dicyanoethenolate anion is weaker than that in the case of the sulfonate and the carboxylate anion. Therefore, the dicyanoethenolate structure is effective in lowering the Tm of the zwitterions. However, it should be mentioned here

c21.indd 307

1/20/2011 10:28:35 AM

308

ZWITTERIONIC LIQUIDS

endo. ←

→exo.

(a)

Tm –50

0

50

100

150

200

Temperature / °C (b) Tc

endo. ←

→exo.

Tg

Tm

–50

0

50

100

150

200

Temperature / °C

Figure 21.3. DSC charts of zwitterion 14. (a) First heating; (b) second heating. TABLE 21.2. Melting Points of Zwitterions Having Carboxylatea Number 20 21 22 23 24 25 26

Structure

Tm (°C) 250 171 165 130 134 168 212b,c

a

(Reproduced by permission of CSIRO publishing). Decomposition temperature. c Holbrey et al. [55]. b

c21.indd 308

1/20/2011 10:28:35 AM

309

ZWITTERIONIC LIQUIDS

TABLE 21.3. Melting Points of Zwitterions with Dicyanoethenolate Number

Structure

27 28 29 30 31 32 33 34 35 36 37

Tm (°C)

Reference

186–187 113 201–202 174 168–170 173–174 242–246 213–217 239–240 >300 183–185

44 44 45 45 45 45 44 44 44 44 44

TABLE 21.4. Melting Points of Zwitterions with Dicyanoethenolate Number 38 39 40 41 42 43 44 45 46 47 48

Structure

Tm (°C)

Reference

186.9 148.4 70 94–95 143–144 115 136–137 144–145 129.9 118–119 111.4

43 43 45 44 44 44 44 44 44 44 44

that the starting material, dicyanoketene cyclic acetal, is obtained from highly toxic reagents like potassium cyanide. It is well known that organic salts with sulfonylamide anions (e.g., bis(trifluoromethylsulfonyl)amide (TFSI) anion) show low melting points and low viscosity values [28,56]. A symmetric anion site and an asymmetric anion site were introduced into zwitterions for lowering the Tm of the zwitterions [47]. Table 21.5 summarizes the the Tm of the zwitterions with sulfonylamide anion sites. Zwitterions 49 and 50 melted at 87 °C and 104 °C, respectively, which were significantly lower than those of zwitterions 14 and 23. Such a tendency should be based on the delocalization of the negative charge by the electron-withdrawing effect of the trifluoromethylsulfonyl group as well as of the cyano group for the dicyanoethenolate anion.

c21.indd 309

1/20/2011 10:28:35 AM

310

ZWITTERIONIC LIQUIDS

TABLE 21.5. Melting Points of Zwitterions with Sulfonylamide Number

Structure O

O

49 N

N

N 49 O

50 N

S O

S O

S O

Reference

87

47

104

47

CF3

O N

Tm (°C)

CF3

50

TABLE 21.6. Thermal Behavior of Zwitterion 13 with an Equimolar Amount of LiX*

Neat zwitterion 13 (2nd heating) +LiTFSI +LiBETI +LiCF3SO3 +LiBF4 +LiClO4

Tm (°C)

Tg (°C)

175 — — — — —

18 −37 −5 19 5 24

Zwitterions containing the polymerizable group (8, 15–19, 27, 28, 33–37, 41–48) are important monomers for the preparation of zwitterionic polymers. Characteristics of zwitterionic polymers will be discussed later. These zwitterions are nonconductive materials that turn conductive when suitable salts are added, so the thermal behaviors of zwitterions become important after the salt addition. Table 21.6 shows the thermal behavior of zwitterion 13 containing an equimolar amount of lithium salts. Five kinds of lithium salts [LiTFSI, lithium bis(perfluoroethylsulfonyl)amide LiBETI, LiCF3SO3, LiBF4, and LiClO4] were compared to study the effect of added salt species on the thermal behavior of zwitterions. Zwitterion 13 showed a Tg at 18 °C in the second heating without a salt addition. When lithium salts were added to the zwitterion, the Tg remained the same despite the added salt species. The Tg of zwitterion 13 after mixing with an equimolar amount of LiTFSI was −37 °C (see Figure 21.4). This was the lowest Tg among the mixtures of zwitterion 13 with five lithium salts. The Tg of the mixture of 13 and lithium salts showed the following sequence of anions: TFSI − < BETI − < BF4− < CF3 SO3− < ClO4−. LiTFSI also gave the lowest Tg for an analogous zwitterion 12 [33]. It is known that the TFSI anion has a plasticizing effect on the polyether matrix [57]. This also lowers the Tg in zwitterions.

c21.indd 310

1/20/2011 10:28:35 AM

311

endo. ←

→exo.

ZWITTERIONIC LIQUIDS

–100

–50

0

50

100

150

200

Temperature / °C

Figure 21.4. DSC chart of zwitterion 13 with an equimolar amount of LiTFSI*.

In Figure 21.5, the ionic conductivity of zwitterions with and without lithium salts was measured over a wide temperature range. It already has been confirmed that zwitterions including an equimolar amount of LiTFSI decompose at around 300 °C [58]. Therefore, these zwitterionic liquids are all good nonvolatile solvents that are capable of dissolving salts. This feature is important for the electrochemical applications. Figure 21.5 shows the temperature dependence of the ionic conductivity for zwitterion 13 with and without an equimolar amount of lithium salts. Neat zwitterion 13 shows a low ionic conductivity value of about 10−6 S cm−1 at even 200 °C, as found from the alternating current (AC) impedance measurement because there are no mobile ions in the system. However, zwitterion 13, which is mixed with an equimolar amount of lithium salt, shows much higher ionic conductivity than that of the neat zwitterion. The higher conductivity is caused by the increase of carrier ions from the addition of lithium salts. The ionic conductivities of BETI− and CF3 SO3− are about 10−4 S cm−1 and 10−5 S cm−1 at 100 °C. Among these lithium salts, the zwitterion mixture including LiTFSI shows the highest ionic conductivity of 8.9 × 10−4 S cm−1 at 100 °C, reflecting the lowest Tg of −37 °C. This result agrees with that of zwitterion 12, which contains various lithium salts [33]. For all lithium salts, the ionic conductivity of zwitterion with and without lithium salts traces an upper convex curve. This behavior explained by the ion transport in zwitterions being led by the motion of the zwitterion itself. Moreover, there is a good relationship between the ionic conductivity and Tg for various zwitterions, including lithium salts (see Figure 21.6). The ionic conductivity of zwitterions containing an equimolar of lithium salt thus increases with the decreasing Tg. The Tg was found to govern the ionic conductivity of zwitterions containing LiTFSI. + The lithium transference number (t Li ) in zwitterion 13 containing an equimolar amount of LiTFSI is 0.56 [59], which is higher than the numbers in

c21.indd 311

1/20/2011 10:28:35 AM

312

ZWITTERIONIC LIQUIDS 100 °C

-1

Ionic conductivity log(σi / S cm )

-2

-3 + LiTFSI

-4

-5

-6 neat

-7 2.0

+ LiCF3SO3

2.2

2.4

2.6

2.8 -1

3.0

+ LiBETI

3.2

3.4

-1

1000 T / K

Figure 21.5. Temperature dependence of the ionic conductivity for zwitterion 13 with an equimolar amount of LiX*. *(Reproduced with permission from CSIRO publishing)

–3

log(σi / S cm–1) at 50 °C

–4

–5

–6

–7

–40

–20

0

20

Tg / °C

Figure 21.6. Relation between Tg and the ionic conductivity for zwitterions containing an equimolar amount of LiTFSI.

c21.indd 312

1/20/2011 10:28:35 AM

REFERENCES

313

typical ionic liquids [60]. This number suggests that zwitterions can transport lithium ions the best. This result seems to be based on the imidazolium cation and the TFSI anion interacting in the ionic liquid domain, which then provides + the ion conductive path for the lithium cation. However, the t Li in zwitterion 21 containing carboxylate is 0.14 [47], which indicates that it is necessary to introduce a counter anion structure with a high degree of dissociation to + + in the zwitterion. In fact, the t Li of zwitterion 50 and that obtain the high t Li of the zwitterion with the organoborate anion (R = Et) shown in Scheme 21.5 were 0.55 and 0.69, respectively [47,49]. Zwitterions can promote lithium ion transport in polymer electrolytes [61−63]; that is, zwitterions act as solvents or ion dissociators. The high dipole moment of zwitterions promotes their ability to solvate a variety of inorganic salts [43]. Furthermore, when a zwitterion was added to the ionic liquid/lithium salt mixture [64] or 1 M LiPF6 in organic solvents [65] as a salt dissociator, the coulombic efficiency of the electrolyte systems was significantly improved. These results clearly show that zwitterions can act as excellent target ion conductors and can promote target ion transport in electrolyte solutions. Although zwitterions mainly are considered for their novel ion conductive matrix in this chapter, they are being used not only as solvents and catalysts for organic reactions [66] but also as organogelators [67]. Zwitterions have been screened as solvent/catalysts for several classical acid-promoted organic reactions such as the Fischer esterification, alcohol dehydrodimerization, and pinacol/benzopinacole rearrangement. The zwitterion containing an equimolar trifluoromethane sulfonic acid is liquid at room temperature. Because they can work as solvent/catalysts, as shown in the reactions discussed in this chapter, zwitterionic liquids should open the door to a whole new area of applications. REFERENCES 1. J. S. Wilkes, M. J. Zaworotko, J. Chem. Soc., Chem. Comm., 965 (1992). 2. E. I. Cooper, J. M. O’Sullivan, in Proc. 8th Int. Symp. Ionic Liquids, vol. 92-16, R. J. Gale, G. Blomgren, and H. Kojima, eds., Electrochemical Society, 1992, p. 386. 3. P. Wasserscheid, T. Welton, eds., Ionic Liquids in Synthesis, Wiley-VCH, 2003. 4. T. Welton, Chem. Rev., 99, 2071 (1999). 5. P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed., 39, 3772 (2000). 6. J. Dupont, R. F. de Souza, P. A. Z. Suarez, Chem. Rev., 102, 3667 (2002). 7. J. G. Huddleston, H. D. Willauer, R. P. Swatloski, A. E. Visser, R. D. Rogers, Chem Comm., 1765 (1998). 8. (a) J. D. Holbrey, K. R. Seddon, Clean Prod. Proc., 1, 223 (1999). (b) K. R. Seddon, J. Chem. Tech. Biotechnol., 68, 351 (1997). 9. (a) H. Sakaebe, H. Matsumoto, Electrochem. Comm., 5, 594 (2003). (b) H. Sakaebe, H. Matsumoto, K. Tatsumi, J. Power Sourc, 146, 693 (2005). (c) H. Matsumoto, H. Sakaebe, K. Tatsumi, J. Power Sourc, 146, 45 (2005). (d) H. Sakaebe, H. Matsumoto, K. Tatsumi, Electrochim. Acta, 53, 1048 (2007).

c21.indd 313

1/20/2011 10:28:35 AM

314

ZWITTERIONIC LIQUIDS

10. (a) P. C. Howlett, D. R. MacFarlane, A. F. Hollenkamp, Electrochem. Solid-State Lett., 7, A97 (2004). (b) P. C. Howlett, N. Brack, A. F. Hollenkamp, D. R. MacFarlane, M. Forsyth, J. Electrochem. Soc., 153, A595 (2006). 11. (a) S. Seki, Y. Kobayashi, H. Miyashiro, Y. Ohno, Y. Mita, A. Usami, N. Terada, M. Watanabe, Electrochem. Solid-State Lett., 8, A577 (2005). (b) S. Seki, Y. Kobayashi, H. Miyashiro, Y. Ohno, A. Usami, Y. Mita, N. Kihira, M. Watanabe, N. Terada, J. Phys. Chem. B, 110, 10228 (2006). (c) S. Seki, Y. Kobayashi, H. Miyashiro, Y. Ohno, A. Usami, Y. Mita, M. Watanabe, N. Terada, Chem. Comm., 544 (2006). 12. (a) K. Tsunashima, F. Yonekawa, M. Sugiya, Electrochem. Solid-State Lett., 12, A54 (2009). (b) K. Tsunashima, F. Yonekawa, M. Sugiya, Chem. Lett., 37, 314 (2008). (c) K. Tsunashima, M. Sugiya, Electrochem. Solid-State Lett., 11, A17 (2008). 13. A. I. Bhatt, A. S. Best, J. Huang, A. F. Hollenkamp, J. Electrochem. Soc. 157, A66 (2010). 14. (a) A. Noda, M. A. B. H. Susan, K. Kudo, S. Mitsushima, K. Hayamizu, M. Watanabe, J. Phys. Chem. B, 107, 4024 (2003). (b) M. A. B. H. Susan, A. Noda, S. Mitsushima, M. Watanabe, Chem. Comm., 938 (2003). 15. (a) E. Cho, J. -S. Park, S. S. Sekhon, G. -G. Park, T. -H. Yang, W. -Y. Lee, C. -S. Kim, S. -B. Park, J. Electrochem. Soc., 156, B197 (2009). (b) S. S. Sekhon, J. -S. Park, E. Cho, Y. -G. Yoon, C. -S. Kim, W. -Y. Lee, Macromolecules, 42, 2054 (2009). (c) S. S. Sekhon, J. -S. Park, J. -S. Baek, S. -D. Yim, T. -H. Yang, C. -S. Kim, Chem. Mater., 22, 803 (2010). 16. (a) A. Fernicola, M. A. Navarra, S. Panero, J. Appl. Electrochem., 38, 993 (2008). (b) A. Fernicola, S. Panero, B. Scrosati, J. Power Sourc, 178, 591 (2008). 17. (a) K. Kudo, S. Mitsushima, N. Kamiya, K. Ota, Electrochemistry, 73, 668 (2005). (b) K. Kudo, S. Mitsushima, N. Kamiya, K. Ota, Electrochemistry, 73, 272 (2005). 18. N. Papageogiou, Y. Athanassov, M. Armand, P. Bonhôte, H. Pettersson, A. Azam, M. Grätzel, J. Electrochem. Soc., 143, 3099 (1996). 19. (a) R. Kawano, M. Watanabe, Chem. Comm., 330 (2003). (b) R. Kawano, H. Matsui, C. Matsuyama, A. Sato, M. A. B. H. Susan, N. Tanabe, M. Watanabe, J. Photochem. Photobio. A, 164, 87 (2004). (c) R. Kawano, M. Watanabe, Chem. Comm., 2107 (2005). 20. W. Kubo, T. Kitamura, K. Hanabusa, Y. Wada, S. Yanagida, Chem. Comm., 374 (2002). 21. (a) S. Murai, S. Mikoshiba, H. Sumino, T. Kato, S. Hayase, Chem. Comm., 1534 (2003). (b) Y. Shibata, T. Kato, T. Kado, R. Shiratuchi, W. Takashima, K. Kaneto, S. Hayase, Chem. Comm., 2730 (2003). 22. P. Wang, B. Wenger, R. Humphry-Baker, J. -E. Moser, J. Teuscher, W. Kantlehner, J. Mezger, E. V. Stoyanov, S. M. Zakeeruddin, M. Grätzel, J. Am. Chem. Soc., 127, 6850 (2005). 23. J. Chen, D. L. Officer, J. M. Pringle, D. R. MacFarlane, C. O. Too, G. G. Wallace, Electrochem. Solid-State Lett., 8, A528 (2005). 24. N. Mohmeyer, D. Kuang, P. Wang, H. -W. Schmidt, S. M. Zakeeruddin, M. Grätzel, J. Mater. Chem., 16, 2978 (2006). 25. D. Kuang, S. Uchida, R. Humphry-Baker, S. M. Zakeeruddin, M. Grätzel, Angew. Chem. Int. Ed., 47, 1923 (2008).

c21.indd 314

1/20/2011 10:28:35 AM

REFERENCES

315

26. Y. Bai, Y. Cao, J. Zhang, M. Wang, R. Li, P. Wang, S. M. Zakeeruddin, M. Grätzel, Nat. Mater., 7, 626 (2008). 27. P. Chen, J. Brillet, H. Bala, P. Wang, S. M. Zakeeruddin, M. Grätzel, J. Mater. Chem., 19, 5325 (2009). 28. P. Bonhôte, A.-P. Dias, M. Armand, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). 29. (a) M. Hirao, H. Sugimoto, H. Ohno, J. Electrochem. Soc., 147, 4168 (2000). (b) M. Yoshizawa, W. Ogihara, H. Ohno, Electrochem. Solid-State Lett., 4, E25 (2001). (c) H. Ohno, M. Yoshizawa, Solid State Ionics, 154–155, 303 (2002). 30. R. Hagiwara, Y. Ito, J. Fluorine Chem., 105, 221 (2000). 31. A. Hayashi, M. Yoshizawa, C. A. Angell, F. Mizuno, T. Minami, M. Tatsumisago, Electrochem. Solid-State Lett., 6, E19 (2003). 32. H. Nakagawa, S. Izuchi, K. Kuwana, T. Nukuda, Y. Aihara, J. Electrochem. Soc., 150, A695 (2003). 33. M. Yoshizawa, M. Hirao, K. Ito-Akita, H. Ohno, J. Mater. Chem., 11, 1057 (2001). 34. (a) M. Yoshizawa, H. Ohno, Ionics, 8, 267 (2002). (b) W. Ogihara, M. Yoshizawa, H. Ohno, Chem. Lett., 31, 880 (2002). 35. (a) H. Ohno, K. Ito, Chem. Lett., 27, 751 (1998). (b) M. Yoshizawa, H. Ohno, Chem. Lett., 28, 889 (1999). (c) M. Hirao, K. Ito-Akita, H. Ohno, Polymer. Adv. Technol., 11, 534 (2000). (d) H. Ohno, Electrochim. Acta, 46, 1407 (2001). (e) M. Yoshizawa, H. Ohno, Electrochim. Acta, 46, 1723 (2001). (f) M. Yoshizawa, W. Ogihara, H. Ohno, Polymer. Adv. Technol., 13, 589 (2002). 36. (a) M. Yoshizawa, H. Ohno, Chem. Comm., 1828 (2004). (b) T. Mizumo, T. Watanabe, H. Ohno, Polymer. J., 40, 1099 (2008). 37. H. S. Park, S. R. Kwon, Y. M. Jung, H. S. Kim, H. J. Lee, W. H. Hong, Chem. Comm., 6388 (2009). 38. J. Cardoso, A. Huanosta, O. Manero, Macromolecules, 24, 2890 (1991). 39. M. Galin, E. Marchal, A. Mathis, J.-C. Galin, Polymer. Adv. Technol., 8, 75 (1997). 40. M. Galin, A. Mathis, J.-C. Galin, Polymer. Adv. Technol., 12, 574 (2001). 41. A. Narita, W. Shibayama, N. Matsumi, H. Ohno, Polymer. Bull., 57, 109 (2006). 42. R. H. Brown, A. J. Duncan, J. -H. Choi, J. K. Park, T. Wu, D. J. Leo, K. I. Winey, R. B. Moore, T. E. Long, Macromolecules, 43, 790 (2010). 43. M. Galin, A. Chapoton, J.-C. Galin, J. Chem. Soc. Perkin Trans., 2, 545 (1993). 44. M.-L. Pujol-Fortin, J.-C. Galin, Macromolecules, 24, 4523 (1991). 45. W. J. Middleton, V. A. Engelhardt, J. Am. Chem. Soc., 80, 2788 (1958). 46. Y. Chevalier, P. L. Perchec, J. Phys. Chem., 94, 1768 (1990). 47. A. Narita, W. Shibayama, H. Ohno, J. Mater. Chem., 16, 1475 (2006). 48. Y. R. Mirzaei, H. Xue, J. M. Shreeve, Inorg. Chem., 43, 361 (2004). 49. A. Narita, W. Shibayama, K. Sakamoto, T. Mizumo, N. Matsumi, H. Ohno, Chem. Comm., 1926 (2006). 50. M. Yoshizawa, H. Ohno, Chem. Lett., 33, 1594 (2004). 51. R. S. Armentrout, C. L. McCormick, Macromolecules, 33, 419 (2000). 52. M. Yoshizawa, A. Narita, H. Ohno, Aust. J. Chem., 57, 139 (2004).

c21.indd 315

1/20/2011 10:28:35 AM

316

ZWITTERIONIC LIQUIDS

53. J. C. Salamone, W. Volksen, S. C. Israel, A. P. Olson, D. C. Raia, Polymer, 18, 1058 (1997). 54. J. M. Pringle, C. M. Forsyth, M. Forsyth, D. R. MacFarlane, Acta Cryst., E59, o1759 (2003). 55. J. D. Holbrey, W. M. Reichert, I. Tkatchenko, E. Bouajila, O. Walter, I. Tommasi, R. D. Rogers, Chem. Comm., 28 (2003). 56. H. Matsumoto, H. Kageyama, Y. Miyazaki, Chem. Comm., 1726 (2002). 57. S. Besner, A. Vallée, G. Bouchard, J. Prud’homme, Macromolecules, 25, 6480 (1992). 58. H. Ohno, M. Yoshizawa, W. Ogihara, Electrochim. Acta, 48, 2079 (2003). 59. M. Yoshizawa, Design and Characteristics of Ionic Liquids Featuring Target Ion Transport, Ph. D. Thesis, Tokyo University of Agriculture and Technology (2002). 60. J. -H. Shin, W. A. Henderson, S. Passerini, Electrochem. Comm., 5, 1016 (2003). 61. C. Tiyapiboonchaiya, J. M. Pringle, J. Sun, N. Byrne, P. C. Howlett, D. R. MacFarlane, M. Forsyth, Nat. Mater., 3, 29 (2004). 62. N. Byrne, D. R. MacFarlane, M. Forsyth, Electrochim. Acta, 50, 3917 (2005). 63. A. Narita, W. Shibayama, M. Tamada, H. Ohno, Polymer. Bull., 57, 115 (2006). 64. N. Byrne, P. C. Howlett, D. R. MacFarlane, M. Forsyth, Adv. Mater., 17, 2497 (2005). 65. D. Q. Nguyen, J. Hwang, J. S. Lee, H. Kim, H. Lee, M. Cheong, B. Lee, H. S. Kim, Electrochem. Comm., 9, 109 (2007). 66. A. C. Cole, J. L. Jensen, I. Ntai, K. L. Tran, K. J. Weaver, D. C. Forbes, J. H. Davis Jr., J. Am. Chem. Soc., 124, 5962 (2002). 67. C. Wang, A. Robertson, R. G. Weiss, Langmuir, 19, 1036 (2003).

c21.indd 316

1/20/2011 10:28:36 AM

22 ALKALI METAL IONIC LIQUIDS Wataru Ogihara, Masahiro Yoshizawa-Fujita, and Hiroyuki Ohno

Ionic liquids (ILs) inherently have high ionic conductivity. However, their serious drawback is that they provide migration of only component ions. Their structure and combination of ions can form low-temperature molten salts, but it is extremely difficult to prepare ILs comprising the desired (target) carrier ions such as Li+. Thus, target carrier ions such as Li+ or H+ must be added to general ILs. Other upper limits must be observed for the solubilization of most additives in common ILs. The addition of the target carrier ions raises the viscosity of ILs and, thus, decreases their ionic conductivity. Inorganic salts such as LiCl containing useful carrier ions at high temperatures show extremely high ionic conductivity. In this chapter, we report on our experiments with ILs that contain useful carrier ions such as Li+ and H+.

22.1 ALKALI METAL IONIC LIQUID [1] Most ILs generally consist of monovalent cations (imidazolium cation or pyridinium cation) and monovalent anions (tetrafluoroborate anion or bis(trifluoromethanesulfonyl)imide anion). It is difficult to form an IL with multivalent ions owing to their strong electrostatic interaction, which causes an elevation of melting point (Tm) or an increase of viscosity. However, we tried to lower the Tm of salts containing divalent anions with aid of organic cations. Many salts consist of a divalent anion, an organic cation, and an alkali metal cation. Alkali metal ionic liquids (AMILs) consist of an organic cation

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

317

c22.indd 317

1/20/2011 10:28:36 AM

318

ALKALI METAL IONIC LIQUIDS

CH3CH2

M+

N + NH M+ 2−

SO4 1-M

+

SO42−

N CH3

H 2-M

M+ = H+, Li+, Na+, K+, Rb+, Cs+ etc.

Scheme 22.1. Structure of AMILs.

TABLE 22.1. Tg of AMILs M=

1-M Tg (°C)

2-M Tg (°C)

H Li Na K Rb Cs

−75 −60 −68 −67 −55 −59

−96 −78 −82 −85 −85 −84

and various kinds of target carrier cations that interact with the sulfate anion (Scheme 22.1). It has been confirmed that the series of N-ethylimidazolium salts (1-M) and 1-methylpyrrolidinium salts (2-M) become liquid at room temperature. It has been recognized that the N-ethylimidazolium cation can provide an ionic liquid at room temperature with a lot of monovalent anions [2]. Table 22.1 shows the glass transition temperature (Tg) of these AMILs. The obtained AMILs had no Tm and showed only Tg. These liquids showed good thermal stability like other ILs. However, AMILs were highly viscous compared with general ILs such as EMImTFSI or EMImBF4. This high viscosity is thought to be the result of a strong interaction among the divalent anions. Even the lowest viscosity of 754 cP for 2-K at 40 °C is far larger than that of general ILs. It is well known that Tg and viscosity are the factors that govern ionic conductivity [3]. A structural redesign is therefore necessary to improve the thermal properties and viscosity. The ionic conductivity of AMILs is nearly equal regardless of the alkali metal ion species. The value of 1-M is around 10−4 S cm−1 and that of 2-M is around 10−3 S cm−1. Figure 22.1 shows the ionic conductivity of two AMILs containing Li+ as the carrier ion. As the figure shows, the ionic conductivity of AMILs mainly is affected by the structure of the organic cation. This difference is reflected in the Tg. Figure 22.2 shows the relationship between the ionic conductivity and the Tg of all AMILs as well as the difference in properties resulting from the organic cation structure. The Li+ migration in AMILs is confirmed by pulsed field gradient nuclear magnetic resonance (PFG-NMR) measurement of the two salts containing Li+.

c22.indd 318

1/20/2011 10:28:37 AM

319

ALKALI METAL IONIC LIQUID -1.0

-2.0

log(σi / S cm−1)

2-Li -3.0

-4.0 1-Li -5.0

-6.0 3.0

3.1

3.2

3.3 –1

1000T / K

3.4

3.5

3.6

–1

Figure 22.1. Temperature dependence of the ionic conductivity of 1-Li and 2-Li. -2.0

log(σi / S cm–1) at 30 °C

-2.5

-3.0

2-M

-3.5

-4.0

1-M

-4.5

-5.0 -100

-90

-80

-70

-60

-50

Tg / °C

Figure 22.2. Relationship between the ionic conductivity and the Tg of 1-M and 2-M.

However, its migration is less diffusive than that of the organic cation for each salt. The 2-Li ions have a 7Li diffusion coefficient, which is 10 times larger than that of 1-Li, reflecting the difference of ionic conductivity. This result suggests the 2-Li to be superior to 1-Li as a lithium ion conductor. To study the use of this system as an ion conductive material, it is necessary to consider some other

c22.indd 319

1/20/2011 10:28:37 AM

320

ALKALI METAL IONIC LIQUIDS

n

CH2CF2

m

CF2CF CF3

m:n = 85:15 P(VdF-HFP)

n

CH3 n

O

NH

CH2SO3Li

O

OCH2CH2CH2SO3Li

CH3 PAMPSLi

PAC3SLi

Scheme 22.2. Structure of host polymers.

characteristics, such as the identification of the carrier ion species and its dissociative behavior in a matrix.

22.2

GELATION OF ALKALI METAL IONIC LIQUID [4]

In some fields, solid (or film) electrolytes are preferred to the liquid electrolytes. This is especially the case in industrial applications. Accordingly, gel-type ion conductive polymers are prepared by mixing AMILs with several polymers. AMILs with Li+ can be mixed with a host of polymers, as shown in Scheme 22.2, to study their gelation behavior. Although the hydrophilic polymers, PAC3SLi and PAMPSLi, are soluble in AMILs, P(VdF–HFP) is insoluble in these liquids. The compatibility observed for the AMILs with particular polymers is understood to be a result of their hydrophobicity. The gels based on AMILs are prepared by changing the amount of polymers from 5 to 20 wt%. Figure 22.3 shows the ionic conductivity of the obtained gel in each host polymer fraction. These gels show a similar ionic conductivity to that of bulk AMILs. Even with 20 wt%, the gel based on 2-Li had an ionic conductivity greater than 10−4 S cm−1 at room temperature. The ionic conductivity for each gel is also different by one order of magnitude, depending on the structure of the organic cation. This result implies that the component ions of AMILs migrate in the gel matrix as carrier ions. To study the mobility of Li+ on the gel matrix, we measured the 7Li diffusion coefficient before and after gelation of the 2-Li (with 5 wt% PAMPSLi) using PFG-NMR. The 7Li diffusion coefficient in the gel was almost unchanged from the value before gelation, despite the decrease in the ionic conductivity with gelation. Because, as this result suggests, the mobility of Li+ is not suppressed by gelation, this gel system is favorable for Li+ conductive materials.

22.3 TRIPLE ION-TYPE IMIDAZOLIUM SALT [5] The idea of developing a zwitterionic liquid (ZIL) for alkali metal transport has led to the consideration of a new cation conductive material. As shown in Chapter 21, an increase in cation conduction occurs when LiTFSI is added to

c22.indd 320

1/20/2011 10:28:37 AM

321

TRIPLE ION-TYPE IMIDAZOLIUM SALT

log(σi / S cm–1) at 30 °C

-2.0

-3.0

2-Li

-4.0 1-Li -5.0

-6.0 0

5

10

15

20

Polymer fraction / wt.%

Figure 22.3. Ionic conductivities at 30 °C plotted as a function of polymer fraction. Square plot: PAMPSLi-based gel; circle plot: PAC3SLi-based gel.

ZILs. An equimolar mixture of ZIL and LiTFSI may give us a new model, namely an imidazolium cation containing two tethered anions. This novel system, called triple ion type imidazolium salt, consists of three charges. Scheme 22.3 shows the structure of such triple ion type imidazolium salts. These salts are prepared by reacting an imidazole analogue with an alkyl sultone (see Chapter 21). Besides the imidazolium cation with two tethered sulfonate anions, this salt has a target carrier cation (an alkali metal cation as the counter ion). This system is expected to become a novel single ion conductive material because the alkali metal cation is more mobile than a triple ion. Table 22.2 shows thermal properties of eight kinds of triple ion type imidazolium salts. The Tm of the lithium salt of compound A is 287 °C, and it is almost 100 °C higher than that of the corresponding ZIL. This finding can be regarded as a decrease of freedom for the component ion because of the increasing number of immobilized anions. To lower the Tm of lithium, several approaches were attempted. When the alkali metal ion of compound A was Na+, the Tm of the corresponding salt was 254 °C, which is about 30 °C lower than that of the lithium salt. This difference is probably caused by the weaker electrostatic interaction between Na+ and the sulfonate group because the surface charge density of Na+ is smaller than that of Li+. However, the introduction of the alkyl group into a position on the imidazolium ring is another factor that can improve thermal properties. The salts forming the methyl group on the imidazolium ring (compounds B, C, and D) show only a Tg at around ambient temperature. This result suggests that the presence of the alkyl group

c22.indd 321

1/20/2011 10:28:37 AM

322

ALKALI METAL IONIC LIQUIDS CH2CH2CH2−SO3− N + N

CH2CH2CH2−SO3− N + N

M+ CH2CH2CH2−SO3−

CH2CH2CH2−SO3−

A

B

CH2CH2CH2−SO3− CH3

M+

CH3

CH2CH2CH2−SO3− CH3

N + N

M+

N + N

CH2CH2CH2−SO3−

M+

CH2CH3 CH2CH2CH2−SO3−

C

D CH2CH2CH2CH2−SO3− N + N

Li+ CH2CH2CH2CH2−SO3− E

Scheme 22.3. Structure of triple ion type imidazolium salt (M+ = Li+ or Na+).

TABLE 22.2. Thermal Properties of Triple Ion Type Imidazolium Salts M+ = Li+

A B C D E

M+ = Na+

Tm (°C)

Tg (°C)

Tm (°C)

Tg (°C)

287 — — — —

— 42 40 60 30

254 — —

— 20 23

Note: — Not detected.

suppresses the crystallization of the system. The same result is confirmed for compound E, which has a longer spacer between the imidazolium cation and the sulfonate anion. However, both the Tg and the Tm of these triple ion type salts are still high. The thermal properties of these ions should be improved by molecular design. Figure 22.4 shows the ionic conductivity of a triple ion type imidazolium salt. Among these salts, B and C containing Na+ show the highest ionic conductivity of around 10−3 S cm−1 at 200 °C. However, the ionic conductivity at

c22.indd 322

1/20/2011 10:28:37 AM

323

CONCLUSION Temperature / °C 200

160

120

-3

log (σi / S cm−1)

-4

-5

-6

-7

-8 2.0

2.2

2.4

2.6

2.8

1000T-1/K-1

Figure 22.4. Temperature dependence of the ionic conductivity for triple ion type imidazolium salts. ▲: BNa; ▽: CNa; □: BLi; ○: ANa; : ELi; ■: ALi; ◇: DLi.

room temperature of each salt was a low value because of their high Tm and Tg. As in the case of the ZIL, the ionic conductivity can be greatly improved by certain additive salt species because the ZIL has no carrier ions. Salt addition into this system also can improve the ionic conductivity, but this approach provides bi-ionic conductive properties, which conflicts with the design concept of this chapter. The details of our studies on the addition of salts will be published separately. Other than the compounds mentioned in this chapter, some studies have been vigorously changing anion structure, spacer structure, and spacer length to investigate improvements to the characteristics of imidazolium salts. Roomtemperature triple ion type molten salt is expected to be obtained in the near future with the accumulation of these studies.

22.4

CONCLUSION

In this chapter, two methods and their possible applications were discussed and a few examples showed their relevance for the molecular design of ILs. Although the direction ILs will take is difficult to predict, at the present time, the ILs designed could be highly functional because many IL are being studied using organic ions that have unlimited ion species and combinations.

c22.indd 323

1/20/2011 10:28:37 AM

324

ALKALI METAL IONIC LIQUIDS

REFERENCES 1. W. Ogihara, M. Yoshizawa, H. Ohno, Chem. Lett., 31, 880 (2002). 2. H. Ohno, M. Yoshizawa, Solid State Ionics, 154–155, 303 (2002). 3. (a) P. Bonhôte, A. P. Dias, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). (b) M. Hirao, H. Sugimoto, H. Ohno, J. Electrochem. Soc., 147, 4168 (2000). 4. W. Ogihara, J. Sun, M. Forsyth, D. R. MacFarlane, M. Yoshizawa, H. Ohno, Electrochim. Acta, 49, 1797 (2004). 5. M. Yoshizawa, H. Ohno, Ionics, 8, 267 (2002).

c22.indd 324

1/20/2011 10:28:37 AM

23 POLYETHER/SALT HYBRIDS Tomonobu Mizumo and Hiroyuki Ohno

23.1 ANOTHER IONIC LIQUID Ambient temperature molten salt can be obtained by several methods. One effective way to obtain a room-temperature molten salt is by the introduction of polyether chains to ions. The term “polyether/salt hybrid” is used in this chapter as a common name for polyether oligomers with anionic or cationic charge(s) on the chain (Figure 23.1). Polyethers, such as poly(ethylene oxide) (PEO), are known as representative ion conductive polymers [1]. Polyether/ salt hybrids have been studied as a kind of room-temperature molten salt apart from the development of onium-type ionic liquids [2]. The preparation of ionic liquids consisting of metal ions has been one of the important goals in this research field. Polyether/salt hybrid derivatives give one such solution for this task. The ion conductive property of a semicrystalline PEO/alkali metal salt complex was first reported by Wright in 1975 [3]. Since then, numerous PEObased polymer electrolytes have been reported for application in solid-state batteries [4]. Polyether derivatives can induce the dissociation of a variety of salts because of their large dipole moment. The dissociated ions can migrate in the polyether matrices assisted by the segmental motion. Generally, the PEO/alkali metal salt complex shows an ionic conductivity of more than 10−5 S cm−1 above its melting point (Tm) at around 40 °C to 60 °C. However, because the ionic conductivity at the crystalline state is lower than 10−8 S cm−1, many researchers have examined amorphous and rubbery systems. In these studies, some researchers have observed that PEO oligomers with salt groups

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

325

c23.indd 325

1/20/2011 10:28:39 AM

326

POLYETHER/SALT HYBRIDS (a) O Li+ CF3

−

S N CH2

Li+ O CH2

O

CH2

n

N − S CF3

CH2

O

O

(b) Br−

Br− N + N

CH2

CH2

O

CH2

n

CH2

N + N

Figure 23.1. Representative chemical structures of polyether/salt hybrids. (a) Cation conductor and (b) anion conductor.

Figure 23.2. A polyether/salt hybrid can be obtained as molten salt in appropriate conditions.

became an amorphous fluid showing no Tm [5,6]. These salts that have flexible tails can be categorized as an ionic liquid. In this chapter, we introduce this unique material.

23.2

DESIGN OF POLYETHER/SALT HYBRIDS

Although numerous salts with organic chains have been synthesized to date, only a few salts were obtained as liquid or glass. In the case of the polyether/ salt hybrid, the crystalline phase of both the salt and the polyether disappear under appropriate conditions (Figure 23.2). The polyether moiety in the polyether/salt hybrid is assumed to inhibit the crystallization of the salts by its flexible segmental motion. Because polyethers help dissociate the bound salt through their ion–dipole interactions, the polyether/salt hybrid inherently shows ionic conductivity. Polyether/salt hybrids show the following important features:

c23.indd 326

1/20/2011 10:28:39 AM

IMPROVEMENT OF IONIC CONDUCTIVITY

327

1. Can be employed with a large variety of organic salts. As mentioned in other chapters, most ionic liquids consist of only a few kinds of ions. It has been shown experimentally that larger and asymmetrically substituted ions with delocalized charge are favored to form ionic liquids with low Tm’s. However, numerous ions remain that are recognized as unsuitable for ionic liquid formation, despite their useful features [7]. In contrast, the methodology of the polyether/salt hybrid can be applied to a variety of salts including metal salts, such as lithium carboxylate and lithium sulfonate. Generally, the introduction of the polyether enables any salt to melt at ambient temperatures. 2. Selective charge transfer. Ionic liquids show excellent ion conductive properties, but the ionic species that constitute the matrix will migrate along the potential gradient. Therefore, many ionic liquids are not useful. Ionic liquids that can selectively transport specific ions, such as the lithium ion, nevertheless are needed for energy devices [8]. Against these, in the bulk polyether/salt hybrids, only one ionic species can readily migrate because the counter ion is anchored to the bulky polyether chain, which has a small diffusion coefficient. These features of polyether/salt hybrids depend on the structure of both the polyether and the salt moiety. In particular, it should be noted that the salt concentration of this system increases with the shortening of the polyether chain length. In the low molecular weight region, the polyether/salt hybrids tend to show a higher glass transition temperature (Tg) and become stiff glass because the salt characteristics become dominant. In contrast, in the higher molecular weight region, the polyether feature is dominant. To obtain a liquidtype polyether/salt hybrid, the polyether chain length must be set in the appropriate region. The appropriate PEO chain length has been shown empirically to be roughly Mn = 400 − 1000 when the linear PEO/salt hybrids have salt moieties on both ends of the chain [2,5,6,9–11]. When only one charge is fixed on the PEO chain, the crystalline phase tends to appear in the region lower than 750. Furthermore, we examined several types of polyether/salt hybrid, as shown in Figure 23.3. We found that there are suitable polyether chain lengths depending on the type (Figure 23.3) and the salt structure. 23.3

IMPROVEMENT OF IONIC CONDUCTIVITY

PEO oligomers with carboxylic acids at their chain ends are commercially available. The PEO/carboxylate hybrids are obtained simply by neutralizing PEO-carboxylic acids with their corresponding bases [5]. Because polyether has a high affinity with salts, the polyether/salt hybrids should be prepared carefully to avoid contamination or generation of low molecular weight salts. After removing water or any other solvents, the PEO/carboxylate hybrids can be obtained as a viscous liquid with a low Tg below −50 °C. However, the bulk

c23.indd 327

1/20/2011 10:28:40 AM

328

POLYETHER/SALT HYBRIDS

Figure 23.3. Schematic representation for several types of polyether/salt hybrids containing anionic charge(s) on the chains.

ionic conductivity of the PEO/carboxylate hybrids is lower than 10−7 S cm−1 at room temperature because of the poor dissociation of the carboxylate. To improve the ionic conductivity, we investigated the structural design of both the polyether and the salt moiety. Because ionic conductivity is the function of the number of carrier ions and their mobility (equation 23.1), the dissociation constant of the salt has a great influence on the ionic conductivity, which is expressed as follows:

σ=

∑ neμ

(23.1)

where n, e, and μ are the number, charge, and mobility of carrier ions, respectively. We examined the ion conductive properties of PEO/salt hybrids that have a variety of salts such as sulfonate (−SO−3 M + , where M+ represents alkali metal ion) [2,6], benzene sulfonate (−C 6 H 4 − SO−3 M + ) [9], thiolate (–S−M+) [10], sulfonamide salt (–N−–SO2–R M+) [11,12], and sulfonimide salt (–SO2–N−– SO2–R M+) [13] groups. The ionic conductivity was improved with the increase of the salt dissociation constant [14], whereas the maximum ionic conductivity was observed to depend on the PEO chain length in every case. Among these salts, the trifluoromethylsulfonylimide salt (–SO2–N−–SO2CF3 M+) gave a high dissociation constant because of the delocalized anionic charge [15]. The ionic

c23.indd 328

1/20/2011 10:28:40 AM

IMPROVEMENT OF IONIC CONDUCTIVITY

329

Figure 23.4. Temperature dependence of the ionic conductivity for a series of PEO/ salt hybrids. The number of ethylene oxide repeating unit is 23 for (b)–(d), and is 12 for (a). (b)–(d) have salt units on both α, and ω chain ends, whereas the α chain end of (a) is capped by a methoxy group.

conductivity of α-methoxyethyl PEO (Mn = 550) containing this salt structure was 7.0 × 10−5 S cm−1 at 30 °C (see Figure 23.4). Although the introduction of the highly dissociative salt group proved to be effective in increasing the number of carrier ions, the increase of the salt fraction significantly elevated the Tg. The characteristics of salt turned dominant by increasing the salt fraction, as mentioned previously. Because the carrier ions can migrate faster in less viscous media, it is also important to optimize the polyether chain length and their structures by considering several factors such as Tm, Tg, and the number of carrier ions. Generally, PEO/lithium salt hybrids show a maximum ionic conductivity in the salt concentration ([Li+]/[ethylene oxide unit]) of around 10 mol%. The introduction of a noncharged group also can enhance their ionic conductivity. In linear-type PEO/ salt hybrids, the ionic conductivity is improved by capping one of the chain ends with a methoxy group. The methoxy-capped system shows lower Tg and higher ionic conductivity than those of the PEO/salt hybrids with salt groups on both chain ends, even though both systems contain nearly equal salt concentrations. Poly(propylene oxide) (PPO) also has been investigated as an interesting alternative to PEO because of its amorphous property [12,13]. PPO/salt hybrids show no Tm regardless of the chain length and are readily obtained as liquid or glass. The PPO-based system also is favorable for fast ion migration.

c23.indd 329

1/20/2011 10:28:40 AM

330

POLYETHER/SALT HYBRIDS

However, the ionic conductivity of the PPO/salt hybrid is generally lower than that of the corresponding PEO system. The lower ionic conductivity of the PPO/salt hybrid is ascribed to the lower polarity of the propylene oxide unit [16]. To improve the ionic conductivity of PPO systems, highly dissociative salt groups, such as sulfonimide salt groups, have been employed. The α-n-butoxy PPO/lithium trifluoromethylsulfonylimide hybrid has shown an ionic conductivity of 3.3 × 10−6 S cm−1 at 30 °C [13]. It is also noteworthy that the ionic conductivity of a simple mixture of PEO and alkali metal salt was improved 10–100 times by employing larger cations such as sodium or potassium ions [17], whereas numerous lithium cation conductors are known to date. The improved ionic conductivity is attributed to the weaker ion–dipole interaction, because larger ions have lower charge densities. Higher carrier ion mobility could be realized by the control of such interaction forces.

23.4

POLYMERIZED PEO/SALT HYBRIDS

So far, many high molecular weight polymer electrolytes consisting of PEO (side-chain and/or main chain) and salt units have been evaluated for application in rechargeable batteries [18]. In particular, polyanionic electrolytes show ion transference numbers of almost unity, whereas the lithium ion transference number of conventional PEO/salt mixtures is no more than 0.3. However, the ionic conductivity of these lithium ion conductors is generally lower than 10−8 S cm−1 at ambient temperatures because of the poor dissociation constant of salt moieties such as carboxylate, sulfonate, and phosphate. To improve the ionic conductivity, polyanions with a highly dissociative salt structure such as −R f − SO−3 M + [19], –N−–SO2CF3 M+ [20], and –SO2–N−–SO2CF3 M+ [21] recently have been prepared. Sisca et al. prepared comb-type copolymers consising of a polysiloxane backbone with alkyl-trifluoromethylsulfonamide and oligo(ethylene oxide) side-chains (Figure 23.5). The flexible polysiloxane backbone was effective in avoiding the elevation of Tg of the system. The obtained polymers showed a low Tg of −67 °C ([Li+]/[ethylene oxide unit] was 4.5 mol%) and relatively high ionic conductivity of 1.20 × 10−6 S cm−1 at 25 °C. O CF3

S

Li+ N−

OCH2CH2

O

(CH3 )3Si

O

Si

O

Si

O x y

CH3

3

OCH3

Si (CH3 )3

CH3

Figure 23.5. Polyanionic electrolyte.

c23.indd 330

1/20/2011 10:28:40 AM

331

POLYETHER/ALMINATE OR BORATE SALT HYBRIDS

However, we have reported the preparation of gel-type polymer electrolytes using polyether/salt hybrid oligomers [12]. A network-type PPO film was swollen with the PPO/trifluoromethylsulfonamide lithium salt hybrid. The obtained gel containing 50 wt% PPO/salt hybrid showed a low Tg of −62 °C. The PPO/salt hybrid behaved not only as an added salt but also as a plasticizer. The ionic conductivity was more than 10−6 S cm−1 at room temperature. Furthermore, this gel electrolyte showed a favorable lithium ion transference number of 0.7–0.8.

23.5

POLYETHER/ALMINATE OR BORATE SALT HYBRIDS

Recently, polyether derivatives with inorganic salt moiety such as alminate or borate salt units have been vigorously investigated. Early works were developed by Shriver’s and Fujinami’s groups [22]. Doan et al. first examined the network-type PEO derivatives containing sodium tetraalkoxyalminate groups. The ionic conductivity of this system remains at around 10−8 S cm−1 at room temperature. However, the ionic conductivity is considerably improved to 2.3 × 10−5 S cm−1 at 25 °C by employing siloxyalminate instead of alkoxyalminate (Figure 23.6a). Although the lithium ion transference number of 0.71 suggests that oligomeric derivatives remained in this system, the improved ionic conductivity can be ascribed to the effective charge delocalization of siloxyaluminate via Si–O pπ-dπ system. Recently, Fujinami and Buzoujima have prepared alminate salts containing two PEO tails and two electron withdrawing unit such as −CO2CF3, −SO3CF3, and −OC6F5 [23]. These compounds were obtained as molten salts. Among them, a remarkably high ionic conductivity of 10−4 S cm−1 at ambient temperature was realized when the –CF3SO3 group was introduced (Figure 23.6b). In another approach, Lewis acidic boron containing PEO derivatives was investigated as an “anion-trapping” polymer electrolyte [24]. Because Lewis acidic boron compounds have a vacant p-electron orbital, the mobile anions

(b)

(a) CH3

R Si

Al O Si

R=

CH3

O

Si O

R

R

CH3 CH2

O Li+ Al O Si R CH3 OCH2CH2

3

CF3

O O

Li+ CF3

CH2CH2O

S O

S O O

O Al

3

CH 3

Li+ O CH2CH2O

3

CH 3

OCH3 3

Figure 23.6. Polyethers containing lithium aluminate groups.

c23.indd 331

1/20/2011 10:28:40 AM

332

POLYETHER/SALT HYBRIDS

(a)

(b) OCH2CH2

n

CH2CH2O

O B O

O B

O

n

B O B O

O CH2CH2O

n

CH2CH2O

n

m

m

(c)

(d) O

O Li+

O H

O

O B O CH2CH2

OCH2CH2

OH n

B O

m

CH2CH2O

n

m

(e) CH3 B

O

CH2

CH

B

O n

Figure 23.7. Polyethers containing boron atoms.

in the polymer electrolytes effectively should be trapped [25]. Mehta and Fujinami et al. reported that a boroxine-ring-based network polymer/LiSO3CF3 mixture showed a high lithium ion transference number of more than 0.7 (Figure 23.7a). Angell et al. prepared anion-trapping polymers with the dehydrative condensation reaction between phenylboric acid and PEO (Figure 23.7b). They pointed out that polyelectrolytes containing borate groups simply can be prepared by adding a variety of salts or bases to these acidic boroncontaining polyethers. Because borate anions tend to coordinate weakly to cations, organoboron-based anionic species are favorable to develop high ionic conductivities [26]. Xu and Angell prepared poly(lithium mono-oxalato borate), a new type of polyanionic electrolytes (Figure 23.7c) [27]. The obtained polymer showed the ionic conductivity of more than 10−5 S cm−1 at ambient temperatures. Finally, well-defined boron-containing polymers by dehydrocoupling reactions have been prepared for a hydroboration reaction (Figure 23.7d) [28,29]. These polymers can be obtained readily in mild conditions without generating water or other by-products. This synthetic route is preferred because it avoids the hydrolysis and disproportionation of boron-containing polymers. Recently,

c23.indd 332

1/20/2011 10:28:40 AM

REFERENCES

333

we examined a facile preparation of anion-trapping oligomers by the one-pot reaction between 9-borabicyclo[3.3.1]nonane (9-BBN) and PPO [30]. The 9-BBN is a common, commercially available hydroboration reagent [31]. The obtained oligomer/LiSO3CF3 mixture showed ionic conductivity of more than 10−6 S cm−1 at room temperature and high lithium-ion transference numbers of more than 0.7; both values are comparable with the conventional PPO/salt hybrids (Figure 23.7e).

23.6 ANION CONDUCTIVE POLYETHER/SALT HYBRID As mentioned previously, cation-conductive polymer electrolytes have been earnestly researched for practical use. However, polyethers are originally suitable for anion conduction, because the ion–dipole interaction is firmer for cations rather than for anions. We have investigated the preparation and ion conductive properties of PEOs with onium salt moieties such as imidazolium, pyridinium, and triethylammonium salts on chain ends (Figure 23.5b) [2,32]. These oligomers were readily obtained as viscous liquid with a low Tg, even when the PEO chain length was shorter than 400. The ionic conductivity of each system was higher than that of the corresponding cation conductors. Both low Tg and high ionic conductivity were attributed to a weak anion–dipole interaction. In a halogenic ion conductor, the ionic conductivity was improved to 10−5 S cm−1 by employing smaller anions such as Cl−. The ionic conductivity also was improved by employing bulky and charge delocalized anions such as SCN− and CF3 SO3− . At the same time, significantly low Tg at about −70 °C was observed. These results suggest that smaller anions can migrate faster than bulky ions in PEO matrices, whereas the bulky ions are effective both to increase the number of carriers and to avoid the ion–dipole interactions.

23.7

CONCLUSION

Polyether/salt hybrids have considerable possibility as new molten salts. The most advantageous point is that any salt can be prepared as a liquid by introducing a polyether chain to the ions. Some drawbacks such as relatively low conductivity, high viscosity, and moderate thermal stability of ether bond remain to be improved.

REFERENCES 1. (a) F. M. Gray, ed., Solid Polymer Electrolytes, VCH, 1991. (b) D. Baril, C. Michot, M. Armand, Solid State Ionics, 94, 35 (1997). 2. (a) K. Ito, N. Nishina, Y. Tominaga, H. Ohno, Solid State Ionics, 86–88, 325 (1996). (b) K. Ito, H. Ohno, Electrochim. Acta, 43, 1247 (1998).

c23.indd 333

1/20/2011 10:28:40 AM

334

POLYETHER/SALT HYBRIDS

3. (a) D. E. Fenton, J. M. Parker, P. V. Wright, Polymer, 14, 589 (1973). (b) P. V. Wright, Br. Polymer J., 7, 319 (1975). (c) P. V. Wright, Electrochim. Acta, 43, 1137 (1998). 4. J. M. Tarascon, M. Armand, Nature, 414, 359 (2001). 5. H. Ohno, K. Ito, Polymer, 36, 891 (1995). 6. K. Xu, C. A. Angell, Electrochim. Acta, 40, 2401 (1995). 7. W. Ogihara, M. Yoshizawa, H. Ohno, Chem. Lett., 880, 2002. 8. (a) M. Yoshizawa, M. Hirao, K. I. Akita, H. Ohno, J. Mater. Chem., 11, 1057 (2001). (b) M. Yoshizawa, A. Narita, H. Ohno, Aust. J. Chem., 57, 139 (2004). 9. (a) K. Ito, H. Ohno, Solid State Ionics, 79, 300 (1995). (b) K. Ito, Y. Tominaga, H. Ohno, Electrochim. Acta, 42, 1561 (1997). 10. K. Kato, K. I. Akita, H. Ohno, J. Solid State Electrochem, 4, 141 (2000). 11. (a) Y. Tominaga, K. Ito, H. Ohno, Polymer, 38, 1949 (1997). (b) Y. Tominaga, H. Ohno, Solid State Ionics, 124, 323 (1999). (c) Y. Tominaga, H. Ohno, Chem. Lett., 955 (1998). (d) Y. Tominaga, H. Ohno, Electrochim. Acta, 45, 3081 (2000). 12. Y. Tominaga, T. Mizumo, H. Ohno, Polymer Adv. Technol., 11, 524 (2000). 13. T. Mizumo, H. Ohno, Polymer, 45, 861 (2004). 14. H. Ohno, H. Kawanabe, Polymer Adv. Technol., 7, 754 (1996). 15. (a) J. Foropoulos, D. D. Desmarteau, Inorg. Chem., 23, 3720 (1984). (b) A. Webber, J. Electrochem. Soc., 138, 2586 (1991). (c) P. Johansson, P. Jaccobsson, Electrochim. Acta, 46, 1545 (2001). 16. (a) D. Teeters, S. L. Stewart, L. Svoboda, Solid State Ionics, 28–30, 1054 (1988). (b) M. Begin, C. Vachon, C. Labreche, B. Goulet, J. Prud’homme, Macromolecules, 31, 96 (1998). (c) A. Ferry, M. Tian, Macromolecules, 30, 1214 (1997). 17. (a) G. Zouu, I. M. Khan, J. Smid, Macromolecules, 26, 2202 (1993). (b) H. Ohno, K. Ito, Polymer Adv. Technol., 2, 97 (1991). 18. (a) E. Tsuchida, N. Kobayashi, H. Ohno, Macromolecules, 21, 96 (1986). (b) M. Watanabe, S. Nagano, K. Sanui, N. Ogata, Solid State Ionics, 28–30, 911 (1988). (c) H. Derand, B. Wesslen, B. E. Mellander, Electrochim. Acta, 43, 1525 (1998). 19. (a) D. Benrabah, S. Sylla, F. Alloin, J. Y. Sanchez, M. Armand, Electrochim. Acta, 40, 2259 (1995). (b) J. F. Snyder, J. C. Hutchison, M. A. Ratner, D. F. Shriver, Chem. Mater., 15, 4223 (2003). 20. D. P. Siska, D. F. Shriver, Chem. Mater., 13, 4698 (2001). 21. (a) A. E. Feiring, S. K. Choi, M. Doyle, E. R. Wonchoba, Macromolecules, 33, 9262 (2000). (b) O. E. Geiculescu, J. Yang, H. Blau, R. Bauley-Walsh, S. E. Creager, W. T. Pennington, D. D. Desmarteau, Solid State Ionics, 148, 173 (2002). (c) M. A. Hofmann, C. M. Ambler, A. E. Maher, E. Chalkova, X. Y. Zhou, S. N. Lvov, H. R. Allcock, Macromolecules, 35, 6490 (2002). 22. (a) K. E. Doan, M. A. Ratner, D. F. Shriver, Chem. Mater., 3, 418 (1991). (b) T. Fujinami, A. Tokimune, M. A. Mehta, D. F. Shriver, G. C. Rawsky, Chem. Mater., 9, 2236 (1997). (c) T. Fujinami, M. A. Mehta, K. Sugie, M. Mori, Electrochim. Acta, 45, 1181 (2000). 23. T. Fujinami, Y. Buzoujima, J. Power Sourc, 119–121, 438 (2003). 24. (a) S. S. Zhang, C. A. Angell, J. Electrochem. Soc., 143, 4047 (1996). (b) H. S. Lee, X. Q. Yang, C. L. Xiang, J. M. McBreen, L. S. Choi, J. Electrochem. Soc., 145, 2813 (1998). (c) J. McBreen, H. S. Lee, X. Q. Yang, X. Sun, J. Power Sourc, 89, 163 (2000). (d) H. S. Lee, X. Q. Yang, X. Sun, J. McBreen, J. Power Sourc, 97–98, 566 (2001).

c23.indd 334

1/20/2011 10:28:40 AM

REFERENCES

335

25. (a) M. A. Mehta, T. Fujinami, Chem. Lett., 915, 1997. (b) M. A. Mehta, T. Fujinami, Solid State Ionics, 113–115, 187 (1998). (c) M. A. Mehta, T. Fukonami, S. Inoue, K. Matsushita, T. Miwa, T. Inoue, Electrochim. Acta, 45, 1175 (2000). (d) X. Sun, C. A. Angell, Electrochim. Acta, 46, 1467 (2001). (e) W. Xu, X. Sun, C. A. Angell, Electrochim. Acta, 48, 2255 (2003). (f) T. Hirakimoto, M. Nishiura, M. Watanabe, Electrochim. Acta, 46, 1609 (2001). (g) S. Tabata, T. Hirakimoto, M. Nishiura, M. Watanabe, Electrochim. Acta, 48, 2105 (2003). 26. (a) W. Xu, C. A. Angell, Electrochem. Solid State Lett., 4, E1 (2001). (b) H. Yamaguchi, H. Takahashi, M. Kato, J. Arai, J. Electrochem. Soc., 150, A312 (2003). 27. (a) W. Xu, C. A. Angell, Solid State Ionics, 147, 295 (2002). (b) W. Xu, D. Williams, C. A. Angell, Chem. Mater., 14, 401 (2002). (c) W. Xu, L. Wang, C. A. Angell, Electrochim. Acta, 48, 2037 (2003). 28. (a) N. Matsumi, K. Sugai, H. Ohno, Macromolecules, 35, 5731 (2002). (b) N. Matsumi, K. Sugai, H. Ohno, Macromolecules, 36, 2321 (2003). 29. (a) Y. Chujo, I. Tomita, Y. Hashiguchi, H. Tanigawa, E. Ihara, T. Saegusa, Macromolecules, 24, 345 (1991). (b) Y. Chujo, I. Tomita, N. Murata, H. Mauermann, T. Saegusa, Macromolecules, 25, 27 (1992). 30. T. Mizumo, K. Sakamoto, N. Matsumi, H. Ohno, Chem. Lett., 33, 396 (2004). 31. (a) E. F. Knights, H. C. Brown, J. Am. Chem. Soc., 90, 5280 (1968). (b) H. C. Brown, R. Liotta, G. W. Kramer, J. Am. Chem. Soc., 101, 2966 (1979). (c) H. C. Brown, J. C. Chen, J. Org. Chem., 46, 3978 (1981). 32. M. Yoshizawa, K. I. Akita, H. Ohno, Electrochim. Acta, 45, 1617 (2000).

c23.indd 335

1/20/2011 10:28:40 AM

24 ELECTRIC CONDUCTIVITY AND MAGNETIC IONIC LIQUIDS Gunzi Saito

In the 1950s, an increase of electric conductivity by about 102–104 times was observed in the melting of insulating aromatic hydrocarbons such as naphthalene [1]. The increase of electric conductivity was ascribed not to the increase of carrier concentration or to the contribution of ionic conduction of the molecular ions but to the increase of carrier mobility [1]. The first highly conductive organic semiconductor, developed by Akamatu, et al. in the 1950s, was perylene • bromide, in which the cation radical perylene molecules contributed to the electronic conduction. This compound was classified as a cation radical salt [2]. The first organic metal tetrathiafulvalene (TTF, the main molecules are depicted in Figure 24.1) • tetracyanoquinodimethane (TCNQ) was developed in the 1970s [3]. The TTF molecule is an electron donor (D), whereas the TCNQ molecule is an electron acceptor (A). The complex is classified as an ionic DA charge transfer complex with a partial charge transfer state TTF0.59+ • TCNQ0.59−, and both components contribute to the electronic conduction. The first organic superconductor (TMTSF)2PF6 (TMTSF: tetramethyltetraselenafulvalene) and the 10 K class organic superconductor κ-bis(ethylenedithio),[BEDT]-TTF2–Cu(NCS)2 developed in the 1980s are cation radical salts [4,5]. The component molecules or ions in these conducting crystals are associated with the ionic and van der Waals interactions in addition to the metallic bond, and in general, the crystals are mechanically fragile and thermally not durable. During heating, they either decompose or provide an unstable molten complex that decomposes or dissociates into its component

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

337

c24.indd 337

1/20/2011 10:28:43 AM

338

ELECTRIC CONDUCTIVITY AND MAGNETIC IONIC LIQUIDS S

S

Me

Se

Se

Me

S

S

S

S

S

S

Me

Se

Se

Me

S

S

S

S

TTF

BEDT-TTF

TMTSF

NC

CN

S

S

S

S

NC

CN

S

S

S

S

C7TET-TTF

TCNQ S

S

S

S

S

S

S

S

Me N

TTC7-TTF

N Et

EMI

Figure 24.1. Chemicals in this chapter.

molecules, resulting in the rapid decrease of electric conductivity. However, although the neutral DA charge transfer complexes have rather poor electric conductivity, some of them provide stable melts, and a rapid increase of conductivity has been observed during the melting [6]. So far, no liquid organic metals of the charge transfer type have been developed. The organic semiconductive condenser (OSCON) is based on the melting phenomenon of the solid complex, N-n-butylisoquinolinium (TCNQ)2. Generally, the use of molecules that have long alkyl chains makes it difficult to grow single crystals, but the obtained solid complex has both a low melting point and a thermally stable melting state. The solidified complex N-nbutylisoquinolinium (TCNQ)2 (anion radical salt in which TCNQ0.5− contributes the electric conduction) after melting (melting point = 215 °C) keeps high conductivity (room temperature conductivity σRT = 0.3 S cm−1). The frequency dependence of the impedance of the alumina condenser including the aforementioned solidified complex as an electrolyte is considerably better than that of the tantalum condenser in the frequency region of 10 to about 1000 kHz [7]. In this chapter, the neutral DA charge transfer complex that exhibits a rapid increase of conductivity during the melt, the cation radical salts of TTF derivatives with a low melting point, the anion radical salts of TCNQ with imidazolium cations with a low melting point, and 1-ethyl-3-methylimidazolium (EMI) salts containing complexes with paramagnetic metals will be described.

24.1

NEUTRAL DA CHARGE TRANSFER COMPLEX

Neutral DA charge transfer complexes Dδ+ • Aδ− (δ ≈ 0) have a lower melting point than those of ionic DA complexes (δ ≥ 0.5), because the Madelung

c24.indd 338

1/20/2011 10:28:43 AM

NEUTRAL DA CHARGE TRANSFER COMPLEX

339

energy in the neutral DA makes a much smaller contribution to the lattice energy. For the neutral DA complex, (i) the concentration of conduction carriers is negligible at low temperatures because δ ≈ 0 and (ii) there is no conduction path with a small activation energy because the lattice is made of alternating stacks comprising D and A molecules with different potential energies. Consequently, the neutral DA complex has a small n (carrier concentration) and μ (carrier mobility), resulting in poor conductivity, σ = neμ, in the solid state. Many studies on the transport properties of the charge transfer solids have been reported, whereas for the liquid state, no such studies have been done. Only a few studies on the optical characterization of the molten charge transfer complexes have been reported [8]. We observed a rapid increase of direct correct (DC) conductivity up to 10−4 S cm−1 greater than about 60 °C in a 1 : 1 neutral charge transfer DA complex (σRT < 10−9 S cm−1) comprising donor C7TET-TTF (melting point = 12 °C) and acceptor TCNQ molecules. At about 60 °C, the complex became a fluid that coexisted with the microcrystals of TCNQ. Concomitantly, the spin concentration exhibited an increase by about 100 times, which corresponded to about 10−2% of the total complexes in terms of election paramagnetic resonance (EPR) measurement. The fluid also presented a charge transfer band similar to that observed in the solution’s spectrum. The contribution of ionic conductivity was not large in the fluid because the conductivity was kept nearly constant for a long period of time under constant DC voltage. Based on these observations, the mechanism and design of liquid organic semiconductors (liquid electronic conductors) can be described as follows. The simultaneous occurrence of the rapid increase of conductivity and spin concentration upon the melting of solid complex (Dδ+ • Aδ−) indicates that the carriers (D•+ and A•−) are generated in the following sequence: The molten complex (Dδ+ • Aδ−) dissociates into liquid neutral donors and neutral TCNQ molecules. These have an equilibrium with the radical species (D•+ and A•−), and the radicals migrate in the liquid. However, the migration of the ionic species may be different from conventional ionic conduction because the ionic conductivity was a minor contribution. Because the C7TET-TTF molecules have long alkyl chains, it is possible that they assemble in the melt to form a self-assembled matrix in which the πmoieties and alkyl chains of C7TET-TTF are arranged as depicted in Figure 24.2. The electron transfer is surmised to be as shown by equation (24.1). The donor molecules in the following donor array are close to those observed in a segregated column in a highly conductive charge transfer solid: Di + DDD → DDi + DD → DDDi + D → DDDDi +

(24.1)

To obtain a conductive melt based on a charge transfer complex, it is advantageous to use the neutral DA charge transfer solid because the Madelung energy is not dominant in the lattice energy. Although a high concentration of carrier density is not present in a neutral complex, it is known that, for the

c24.indd 339

1/20/2011 10:28:43 AM

340

ELECTRIC CONDUCTIVITY AND MAGNETIC IONIC LIQUIDS

D

D

D ⋅+

D

D

D

Figure 24.2. Schematic representation of a self-assembled array that consists of D•+ and D (C7TET-TTF) molecules. Each molecule consists of a π-electron part (ellipse) and two alkyl chains (wavy lines).

extrinsic semiconductor, some carrier doping remarkably enhances the conductivity [9]. The high conductivity in liquid crystal (1.2 × 10−3 S cm−1 at 110 °C) resulting from the viologen derivative has been interpreted in terms of the interaction of ionic conduction of the viologen dication, the cation radical, and the iodine species [10].

24.2

CATION RADICAL SALTS WITH LOW MELTING POINT

Recently, cation radical salts with a low melting point have been prepared using the TTC7–TTF molecule (melting point = 44–46 °C), in which the nheptylthio groups attached to both sides of TTF skeleton [11]. Mixing neutral TTC7–TTF with radical salt (TTC 7 − TTF 2+ )(Br3− )2 in a suitable ratio resulted in mixed valence salts (TTC 7 − TTF n+ )(Br3− )n, and their characteristics varied depending on the dopant concentration. For 1 < n < 2, the resulting compounds were purple crystallites, whereas dark brown viscous liquids were obtained for 0 < n < 1. The conductivity at room temperature was in the range of 10−5 to 10−7 S cm−1 for the liquid form (the maximum value was 8 × 10−6 S cm−1), and the ion aggregations led to a slight increase of conductivity (maximum value was 2 × 10−5 S cm−1). It is thus apparent that the conduction is entirely between the TTC7–TTFn+ molecules instead of the Br3− anions and is therefore by way of the electrons (holes). When neutral TTC7–TTF and nitrosonium NOBF4 were reacted in hexane, viscous liquids (TTC 7 − TTF n+ )(BF4− )n were obtained. As the doped concentration increased, the color changed from initially dark brown to deep green and

c24.indd 340

1/20/2011 10:28:43 AM

TCNQ ANION RADICAL SALTS WITH LOW MELTING POINT

341

then to blue. Although the conductivity at room temperature increased as the mixing ratio of the starting materials approached unity (n = 1; the maximum value was 4 × 10−7 S cm−1), the values were much lower than those of the tribromide salts.

24.3 TCNQ ANION RADICAL SALTS WITH LOW MELTING POINT The series of TCNQ complexes is well known to be conductive organic anion radical salts. Among them, only those with a partial charge transfer state, in which the average charge of TCNQ is a noninteger, exhibit high conductivity. In general, TCNQ molecules form a one-dimensional column, which exhibits the physical phenomena characteristic of low-dimensional electron systems (e.g., Peierls and spin-Peierls transitions). Because most melting (or decomposition) points of TCNQ complexes are higher than 200 °C, the studies of TCNQ complexes mainly have been carried out in the solid state. For example, of the TCNQ complexes with a low melting point, a mixed crystal of propylisoquinolinium, methyl-isoquinolinium, TCNQ, and Me-TCNQ (melting point = 80 °C) was reported [12]. The average charge of the TCNQ derivative molecules in the mixed crystal was −0.5, and its conductivity at room temperature was 5 × 10−5 S cm−1, which was much lower than that of other 1 : 2 TCNQ salts (e.g., [N-n-butyl-isoquinolinium][TCNQ]2: σRT = 3 × 10−1 S cm−1, melting point = 215 °C [7]). This mixed crystal shows semiconductive behavior, and the activation energies (Ea) in the solid and liquid states were 0.14 eV and 0.11 eV, respectively, and the conductivity increased twice by melting. In the ionic liquids based on the low-symmetric imidazolium cations, inorganic closed-shell anions often have been used as counter anions. The replacement of the inorganic anion with a TCNQ anion radical should lead to the development of liquid conductors in which ionic and electronic conductions occur simultaneously. TCNQ salts of 1-alkyl-3-methylimidazolium (alkyl = methyl, ethyl, n-propyl, n-butyl, and n-hexyl) have been reported [12], and 1 : 1 and 2 : 3 stoichiometry salts, in which the average charge of TCNQ molecules were −1 and −0.67, respectively, were obtained for each cation. The conductivities at room temperature were 10−6–10−9 S cm−1 and 10−1–10−5 S cm−1 for 1 : 1 and 2 : 3 salts, respectively (see Table 24.1), and both liquids exhibited semiconductive behavior. For the 1 : 1 EMI salt, dimers of the TCNQ molecules formed a one-dimensional column [14]. In Table 24.2, the melting points of the 1 : 1 salts are presented; those of the 2 : 3 salts were not reported [13]. We obtained 1 : 2 salts of 1-alkyl-3-methylimidazolium (alkyl = ethyl and n-butyl) [15], and their decomposition points also are presented in the table. All salts are solid at room temperature and have melting points higher than 100 °C. The 1 : 2 salts have higher melting (or decomposition) points than those of the corresponding 1 : 1 salts, owing to their higher cohesive energy. A decrease in the melting point was observed when TCNQ derivatives (R2–TCNQ) were used instead of the pristine TCNQ. In the 1 : 1 salts of EMI, the melting points

c24.indd 341

1/20/2011 10:28:43 AM

342

ELECTRIC CONDUCTIVITY AND MAGNETIC IONIC LIQUIDS

TABLE 24.1. List of Conductivities at Room Temperature (σRT/S cm−1) of TCNQ Anion Radical Salts of 1-Alkyl(R1)-3Methylimidazolium and TCNQ R1

1 : 1a

2 : 3a −8

8 × 10 4 × 10−8 3 × 10−9 3 × 10−7 7 × 10−6

CH3 C2H5 n-C3H7 n-C4H9 n-C6H13

7 2 2 2 4

× × × × ×

1 :2 −5

10 10−3 10−2 10−1 10−2

5 × 10−4 2 × 10−1

a

From Šorm et al. [13].

TABLE 24.2. List of Melting Points (°C) of TCNQ Anion Radical Salts with Imidazolium Cations. Left Panel: Salts of 1-Alkyl(R1)-3-Methylimidazolium and TCNQ; Right Panel: Salts 1-Ethyl-3-Methylimidazolium and Substituted TCNQs (R2-TCNQ) R1 CH3 C2H5 n-C3H7 n-C4H9 n-C6H13

1 : 1a 174–177 143–146 159–161 154–156 124–126

1 : 2b 225–226 190–194

R2

1 :1

H CH3 CF3 n-C10H21

139.9–140.2 128.8–129.4 144.3–144.7 ∼40

a

From Šorm et al. [13]. Decomposition point.

b

decreased in the following order R2 = CF3, H, CH3, and n-C10H21 (see Table 24.2, right). The conducting behavior in the liquid state has not been reported for these TCNQ salts. We found that the conductivity of the 1 : 1 salt of EMI–TCNQ increased by four orders of magnitude by melting and decreased by three orders of magnitude by freezing (Figure 24.3) [15]. The arrows in the figure show the hysteresis of the temperature dependence of conductivity. The conducting behavior was semiconductive with Ea = 0.41 and 0.38 eV for the solid and liquid states, respectively. The value of conductivity in the liquid state (∼10−2 S cm−1) was quite high compared to those of ordinary monovalent TCNQ salts, namely <10−4 S cm−1. It seems that the high conductivity in the liquid state was realized by the increase of conducting carriers by the dissociation of dimers ( (TCNQi − )2 2 TCNQi − ) and the contribution of ionic conductivity in addition to electronic conductivity of TCNQ•−. This salt kept the supercooled state down to around 100 °C, which was lower than its melting point by 40 °C.

c24.indd 342

1/20/2011 10:28:43 AM

343

EMI SALTS CONTAINING COMPLEXES WITH PARAMAGNETIC METALS

Temperature / °C 50 100

10–9

heat

Conductivity / S.cm–1

10–8 10–7

150

ing

20~77 °C Ea = 0.41 eV

10–6 10–5 cooling

10–4 10–3 10–2 10–1 0.0035

166~103 °C Ea = 0.38 eV 0.003 0.0025 Temperature–1 / K–1

Figure 24.3. Temperature dependence of electronic conductivity of EMI · TCNQ (1 : 1).

TABLE 24.3. EMI-Based RT Ionic Liquids Containing Paramagnetic Metals Compound [EMI][FeCl4]

Central Metal

Melting Point

References

Fe(III)

18 °C

16–18

24.4 EMI SALTS CONTAINING COMPLEXES WITH PARAMAGNETIC METALS Although numerous reports have appeared describing catalytic, spectrochemical, and electrochemical investigations of halometalates in N,N′dialkylimidazolium (e.g., EMI) chloride/AlCl3 ionic liquid, there are only a few room temperature ionic liquids comprising N,N′-dialkylimidazolium cation and complexes with paramagnetic metals, for which their physical properties were investigated. Table 24.3 shows the EMI-based room temperature ionic liquid containing paramagnetic metals [EMI][FeCl4] [16–18], and Table 24.4 the structurally characterized EMI-based salts containing paramagnetic metals [18–24]. All metal complexes listed in the tables are based on halometallates, although some EMI-based salts with diamagnetic AgI(CN)2 [25], Cd II2 (SCN )6 [26], and PbII(O2CMe)4 [27] anions as well as 1-butyl-3-methylimidazolium

c24.indd 343

1/20/2011 10:28:43 AM

344

ELECTRIC CONDUCTIVITY AND MAGNETIC IONIC LIQUIDS

TABLE 24.4. Structurally Characterized EMI-Based Salts Containing Paramagnetic Metals Anion

Central Metal

[VOCl4]2− [Zr6MnCl18]5− [FeCl4]2− [Zr6FeCl18]4− [CoCl4]2− [NiCl4]2− [Ru2Br9]3− [IrCl6]2−

V(IV) Mn(II)a Fe(II) Fe(III)a Co(II) Ni(II) Ru(III) Ir(IV)

Melting Point (°C)

References

99

19 20,21 18 20 22 22 23 24

86 100–102 92–93 220b

a

Valence state for starting material. Decomposition temperature.

b

(BMI)-based salts with paramagnetic MII(dmit)2 (M = Ni and Pd) [28] anions have been structurally reported. This implies that EMI-based salts with paramagnetic metals are at a primitive level of development. One of the most pronounced features is that the salts with monoanionic complexes have a lower melting point, and this marked difference can be explained by the interionic coulomb interactions. Physical properties of [EMI] [FeCl4] recently have been investigated by our group [17,18], and it seems that its ionic conductivity (1.8 × 10−2 S cm−1 at 20 °C) is comparable with that of the well-known [EMI] [AlCl4] (2.3 × 10−2 S cm−1 at 25 °C [29]). From the alternating current (AC) and DC susceptibility measurements, we learned that the salt has an uncorrelated high-spin system (S = 5/2) at around room temperature regardless of its form, indicating the first conductive-paramagnetic bifunctional ionic liquid. Also, it passes through an antiferromagnetic transition at 4.2 K (spin-flop field is ca. 8 kOe). The transition temperature is considerably lower than the value of 9.75 K observed for neat FeCl3 [30], possibly because of the presence of EMI cations. However, the FeIIICl4 salt magnetically diluted by diamagnetic GaIIICl4 anions exhibits no magnetic transition down 1.9 K and follows the Curie-Weiss law. An EMI-based salt with dicationic irons, [EMI]2[FeIICl4], also was prepared and characterized by our group [18]. The powder X-ray diffraction data were collected using a synchrotron radiation on the beam line BL02B2 at SPring-8, and the crystal structure was refined by the Rietveld method. It seems that the salt is isomorphous to the reported [EMI]2[MIICl4] (M = Co and Ni) salts [22], in which each EMI cation has hydrogen bonds with three tetrahedral MCl4 anions. The unit cell volume (3929.03 Å3) is apparently larger than those of the two isomorphous salts (3890 and 3871 Å3), which may be related to its lower melting point than those of the CoIICl4 and NiIICl4 salts. An estimated magnetic moment at room temperature (5.22μB) indicates the uncorrelated high-spin system (S = 2), as observed for the CoIICl4 and NiIICl4 salts, and

c24.indd 344

1/20/2011 10:28:43 AM

REFERENCES

345

the spin configurations are caused largely by the tetrahedral geometry of anions. The intercalation of ionic liquids containing paramagnetic anions into materials with layer (e.g., graphite) or porous (e.g., zeolite) structures as well as research concerning transport properties under the magnetic field are both promising areas for future exploration. Notes: After submittion of this work and Reference 17, a work on the paramagnetic behavior of [BMI] [FeCl4] was submitted and recently published [31].

REFERENCES 1. J. Kommandeur, G. J. Korinek, W. G. Schneider, Can. J. Chem., 36, 513 (1958). 2. H. Akamatu, H. Inokuchi, Y. Matsunaga, Nature, 173, 168 (1954). 3. J. P. Ferraris, D. O. Cowan, V. Walatka Jr., J. H. Perlstein, J. Am. Chem. Soc., 95, 948 (1973). 4. D. Jerome, K. Bechgaard, Sci. Am., 247, 50 (1982). 5. T. Ishiguro, K. Yamaji, G. Saito, Organic Superconductors, 2nd ed., Springer, 1998. 6. A. Otsuka, G. Saito, T. Nakamura, M. Matsumoto, Y. Kawabata, K. Honda, M. Goto, M. Kurahashi, Synth. Met., 27, B575 (1988). 7. (a) S. Niwa, Synth. Met., 18, 665 (1987). (b) I. Isa, Nikkei New Materials, Jan. 30, 48 (1989). (c) S. Niwa, Y. Taketani, J. Power Sourc., 60, 165 (1996). 8. (a) J. Aihara, A. Sasaki, Y. Matsunaga, Bull. Chem. Soc. Jpn., 43, 3323 (1970). (b) J. Aihara, Bull. Chem. Soc. Jpn., 48, 1031 (1975). 9. C. Kittel, Introduction to Solid State Physics, 7th ed., Wiley, 1996. 10. I. Tabushi, K. Yamamura, K. Kominami, J. Am. Chem. Soc., 108, 6409 (1986). 11. Y. Yoshida, A. Otsuka, G. Saito, In press. 12. V. A. Starodub, E. M. Gluzman, K. I. Pokhodnya, M. Ya. Valakh, Theor. Exp. Chem., 29, 240 (1993). 13. M. Šorm, S. Nešpurek, M. Procházka, I. Koropecký, Collect Czech. Chem. C., 48, 103 (1983). 14. M. C. Grossel, P. B. Hitchcock, K. R. Seddon, T. Welton, S. C. Weston, Chem. Mater., 6, 1106 (1994). 15. K. Nishimura, G. Saito, Synth. Met. In press. 16. Y. Katayama, I. Konishiike, T. Miura, T. Kishi, J. Power Sourc., 109, 327 (2002). 17. Y. Yoshida, J. Fujii, K. Muroi, A. Otsuka, G. Saito, M. Takahashi, T. Yoko, Synth. Met. In press. 18. Y. Yoshida, A. Otsuka, G. Saito, S. Natsume, E. Nishibori, M. Takata, M. Sakata, M. Takahashi, T. Yoko, submitted for publication. 19. P. B. Hitchcock, R. J. Lewis, T. Welton, Polyhedron, 12, 2039 (1993). 20. C. E. Runyan, Jr., T. Hughbanks, J. Am. Chem. Soc., 116, 7909 (1994). 21. D. Sun, T. Hughbanks, Inorg. Chem., 39, 1964 (2000). 22. P. B. Hitchcock, K. R. Seddon, T. Welton, J. Chem. Soc., Dalton Trans., 2639 (1993).

c24.indd 345

1/20/2011 10:28:43 AM

346

ELECTRIC CONDUCTIVITY AND MAGNETIC IONIC LIQUIDS

23. D. Appleby, R. I. Crisp, P. B. Hitchcock, C. L. Hussey, T. A. Ryan, J. R. Sanders, K. R. Seddon, J. E. Turp, J. A. Zora, J. Chem. Soc., Chem. Commun., 483 (1986). 24. M. Hasan, I. V. Kozhevnikov, M. R. H. Siddiqui, C. Femoni, A. Steiner, N. Winterton, Inorg. Chem., 40, 795 (2001). 25. Y. Yoshida, K. Muroi, A. Otsuka, G. Saito, M. Takahashi, T. Yoko, Inorg. Chem., 43, 1458 (2004). 26. F. Liu, W. Chen, X. You, J. Solid State Chem., 169, 199 (2002). 27. J. T. Hamill, C. Hardacre, M. Nieuwenhuyzen, K. R. Seddon, S. A. Thompson, B. Ellis, Chem. Comm., 1929 (2000). 28. (a) W. Xu, D. Zhang, C. Yang, X. Jin, Y. Li, D. Zhu, Synth. Met., 122, 409 (2001) (b) C. Jia, D. Zhang, W. Xu, X. Shao, D. Zhu, Synth. Met., 137, 1345 (2003). 29. A. A. Fannin, D. A. Floreani, L. A. King, J. S. Landers, B. J. Piersma, D. J. Stech, R. L. Vaughn, J. S. Wilkes, J. L. Williams, J. Phys. Chem., 88, 2614 (1984). 30. E. R. Jones, O. B. Morton, L. Cathey, T. Auel, E. L. Amma, J. Chem. Phys., 50, 4755 (1969). 31. S. Hayashi, H. Hamaguchi, Chem. Lett., 1590 (2004).

c24.indd 346

1/20/2011 10:28:43 AM

PART V IONIC LIQUIDS IN ORDERED STRUCTURES

c25.indd 347

1/20/2011 10:28:45 AM

25 ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS Maria Forsyth, Jennifer M. Pringle, and Douglas R. MacFarlane

25.1

INTRODUCTION

A plastic crystal can be defined as a material that consists of a regular threedimensional (3D) crystalline lattice such that the long-range position of species is well defined, but orientational and/or rotational disorder exists with respect to the molecular species or molecular ions. These materials thus possess more degrees of freedom and hence display higher entropy than a fully ordered solid [1–3]. As a result of this disorder, a significant degree of motion is available that leads to a high degree of plasticity (i.e., ready deformation under an applied stress) and high diffusivity, which, in the case of ionic substances, leads to ionically conducting phases [4–10]. Fast ion conduction in plastic crystal phases of inorganic materials including Li2SO4 [11–14] and Na3PO4 [15,16] have been of interest for some time; these materials conduct lithium ions and 3− sodium ions, respectively, via a paddle wheel motion of the SO2− 4 and PO4 anions. Plastic crystal phases in organic materials have been known since the time of Timmermans [1], and these phases are often reached via a solid–solid transition below the final melting point of the crystal. These transitions often represent the onset of rotational motions of the molecules within the crystalline lattice, and the resultant phases are sometimes referred to as rotator phases [17–22]. Timmermans proposed a general rule that plastic phases have low

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

349

c25.indd 349

1/20/2011 10:28:45 AM

350

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS

final entropy of fusion (ΔSf < 20 J K−1 mol−1) because the rotational component of the entropy of fusion of the fully ordered phase is already present in the plastic phase. The bulk of the early work on plastic crystals was on a wide range of organic molecular compounds [2] such as cyclohexane, succinonitrile [23–25], and pentaglycerine [26]. Only more recently has it been discovered that ionic compounds can show plastic crystal behavior. One such family of compounds discovered in our laboratories are those based on substituted pyrrolidinium cations, with anions including bis(trifluoromethanesulfonyl)amide (NTf2), BF4, PF6, SCN and dicyanamide (N(CN)2), shown subsequently. These compounds usually display a series of submelting phase transitions and conduction of one or both ions in the higher temperature solid phases [5,27–30]. O

Me

N +

CF3 S R

[Cnmpyr]

O

PF6–

S CF3

N O – O NTf2

BF4–

SCN– – C N C

N N

Another interesting effect discovered in our laboratories is that the addition of low levels of a second component, or “dopant” ion, can lead to significant increases in the ionic conductivity [5,31–34]. Typically these dopant species, for example, Li+, OH−, and H+, are much smaller than the organic ions of the matrix, and because the relaxation times characterizing the motion of these ions are more rapid than those of the bulk matrix itself, these materials may represent a new class of fast ion conductor. The dopant ion effect can be used to design materials for specific applications, for example, Li+ for lithium batteries, H+/OH− for fuel cells, or other specific sensor applications. Finally, we have recently discovered that this dopant effect can also be applied to molecular plastic crystals such as succinonitrile [35]. Such materials have the added advantage that the ionic conductivity is purely a result of the dopant ions and not of the solvent matrix itself.

25.2

PLASTIC CRYSTAL PHASES—BACKGROUND

Since the initial work by Timmermans more than 40 years ago [1], plastic crystals have been studied extensively [10,17–20,22,23,36–73]. The entropy of fusion criterion developed by Timmermans is not always obeyed in ionic plastic crystal materials, in which the salt possesses some internal degrees of freedom that only become activated at the melting point. However, in most cases where plastic crystal behavior is observed, the final entropy of fusion represents only a fraction of the total entropy change in the material from its lowest energy, low-temperature phase to the final melting. Figure 25.1 shows typical thermal analysis traces for a family of pyrrolidinium compounds. One

c25.indd 350

1/20/2011 10:28:45 AM

351

PLASTIC CRYSTAL PHASES—BACKGROUND phase II phase I phase IV

melt

phase III

[C1mpyr][SCN]

endo phase I

phase II

Phase IV

[C1mpyr][N(CN)2 ]

phase II

Phase III

phase I

[C1 mpyr][BF4 ]

exo Phase II

-80 -60 -40 -20

0

Phase I

20

40

60

[C1 mpyr][NTf2 ]

80 100 120 140 160 180

temperature / °C Figure 25.1. Differential scanning calorimetry (DSC) traces for different N,Ndimethylpyrrolidinium ([C1mpyr]) species. The y-axis has been adjusted to allow comparison between species. Thermal transitions are recorded as peaks (endothermic) or troughs (exothermic) and indicate changes of phase. The structurally similar dicyanamide and thiocyanate species both show broad phase transitions in comparison to the NTf2 species.

mechanism whereby the entropy of the solid can increase, either suddenly at a phase transition or gradually over a range of temperatures, is through rotational disorder, (i.e., the onset of rotator motions around one or more molecular axes). Another related alternative suggested is that molecules can find themselves in different, energetically equivalent positions between which they can interchange; the positions are related by rotational symmetry operators associated with the individual species (orientational disorder) [2]. The onset of rotator motions in the phases below the melting point then increases the entropy of the solid state closer to that of the liquid, resulting in a lower ΔSf compared with that for a fully ordered, static crystal. It is proposed that with the onset of these rotations comes the formation of vacancy defects allowing high diffusivity and, ultimately, high “plasticity” (i.e., easy deformation under an applied stress). These materials are often observed to flow under their own weight (creep) [2]. Many salts comprising quaternary alkylammonium cations, in combination with anions of high symmetry, have been found to display rotator phases and/or plastic crystal phases, as determined by thermal analysis and nuclear magnetic resonance (NMR) measurements [18,19,36,53–60]. High ionic

c25.indd 351

1/20/2011 10:28:45 AM

352

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS

conductivities have been measured in some of these systems, which is attributed to vacancy migration. Tansho et al. [9] have investigated plastic phases in NH4Cl, NH4Br, and NH4NO3, which all display isotropic rotation of the ammonium cation in the plastic phase and show cation diffusion via either a direct or indirect (interstitial intermediate) vacancy migration process. More complex species such as (CH3NH3)2ZnBr4 have also been shown to possess ionic plastic crystal phases at elevated temperatures [74]. The relationship between the ion conduction and defects in the [C1mpyr] [NTf2] species has been studied using positron annihilation lifetime spectroscopy (PALS) [75,76]. It is believed that rotation in a plastic crystal is less hindered in the vicinity of defects and that this, therefore, enhances the rotational motion through the material. The mechanical flexibility of plastic crystals also may be explained by the diffusion of vacancies and dislocations facilitating crystalline plane slip. Defects have also been suggested to be associated with ionic diffusion through the material, occurring through either vacancies or through extended defects such as dislocations or grain boundaries. The number and size of defects in a plastic crystal can be probed by PALS analysis. Analysis of [C1mpyr][NTf2] shows a close correlation between the ionic conduction and the defects in the plastic crystal. In the subambient phase III of this salt, the vacancies measured are smaller than the ions, and therefore, the motion of the ions is concluded to be predominantly rotational. In phase II, the vacancies are larger, but the critical volume (the minimum volume required to accommodate the mobile ions) is larger than either one vacancy or a cation– anion divacancy, and thus, it is postulated that ions diffuse through extended defects such as dislocations via a pipe diffusion mechanism. This highlights the fact that it is both the size of the vacancies and the connectivity between these vacancies that can influence the conduction mechanism in an organic ionic plastic crystal [77]. The crystal structures of this plastic crystal in phases III and IV indicate that the packing of the ions is almost identical in both phases [78], but in phase III, all ions are disordered. Both the cations and the anions in phase III seem to possess rotational disorder, and the conformations of the ions are different. The disorder in the NTf2 anions is a C2 ↔ C2 disordering mode (rather than a mixture of C1 and C2 conformations as one might expect), and this new mode of disorder also has been observed in other NTf2 salts [78,79], implying that it is a common mode of disorder for solid-state materials of this anion.

25.3 SYNTHESIS AND THERMAL PROPERTIES OF IONIC PLASTIC CRYSTAL ELECTROLYTES The compounds prepared and characterized in our laboratories can also be described as salts of quaternized nitrogen cations combined with a variety of anions. The pyrrolidinium family of cations, in particular, shows rich phase behavior in the solid state. The anions themselves often possess some degree

c25.indd 352

1/20/2011 10:28:45 AM

PROPERTIES OF IONIC PLASTIC CRYSTAL ELECTROLYTES

353

of rotational symmetry and the degree of plasticity is certainly dependent on the nature of the anion as well as the cation, which will be discussed later. The compounds are prepared by a metathesis reaction from the corresponding iodide, bromide, or chloride salt. The compounds are typically low melting point salts (in many cases, molten at room temperature), which range from being hydrophobic (NTf2, PF6) to completely water soluble and hygroscopic (N(CN)2, BF4). The purity of the samples is established via solution NMR, electrospray mass spectroscopy, and chemical analytical analysis. Figures 25.1 and 25.2 present the thermal analysis traces for a range of typical plastic crystal materials based on the pyrrolidinium cation. The highest temperature phase is denoted as phase I, with subsequent lower temperature phases denoted as phases II, III and so on. It is shown that the nature of the anion plays a significant role in the thermal behavior of these compounds. For example, [C1mpyr][NTf2] has a barely detectable solid–solid transition at 20 °C, followed by a melt at above 130 °C, whereas [C1mpyr][BF4] has a large transition from phase IV to phase III at −43 °C, a small transition from phase III to phase II at 61 °C, followed by phase II to phase I at 112 °C, and finally, decomposes at 340 °C. [C1mpyr][N(CN)2] has a similarly rich thermal behavior. Comparison of Figures 25.1 and 25.2 also demonstrates that the substituent

Figure 25.2. DSC traces for the N-methyl-N-ethylpyrrolidinium ([C2mpyr]) species. The y-axis has been adjusted to allow comparison between species. As observed for the [C1mpyr] series, the dicyanamide and thiocyanate species show broad phase transitions in comparison to the sharp NTf2 salt. The significant differences observed between the species highlights the strong influence of the anion in determining the phase behavior.

c25.indd 353

1/20/2011 10:28:45 AM

354

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS

Phase I->Liquid Endo

Phase II->I

– 42

– 41

– 40

– 39

Phase III->II

Tg

Exo

Phase II->Phase III – 40

– 20

0

20 T /°C

40

60

80

Figure 25.3. DSC trace of N-methyl-N-butylpyrrolidinium hexafluorophosphate ([C4mpyr][PF6]) showing the low-temperature glass transition, the exothermic crystallization, and the subsequent endothermic transitions up to the melt.

on the pyrrolidinium ring has a strong influence on the solid-state behavior of the salts. In many organic ionic plastic crystals, thermal analysis reveals metastable behavior, in which rapid quenching can trap the material in higher temperature phases, and glass transition temperatures often can be observed. The [C4mpyr][PF6] system (Figure 25.3) is one example in which quenching from ambient temperature leads to the formation of a phase II glass, where the rotationally disordered state found in phase II is quenched into an arrested state and subsequently undergoes a transition (glass transition) at −47 °C, and it is at this temperature that onset of the rotational motions of phase II occurs. Given that phase II is metastable at this temperature, the sample undergoes a cold crystallization to the stable phase III state before subsequently transforming to equilibrium phase II at higher temperatures. This behavior can lead to confusion when interpreting thermal analysis (and conductivity) traces of plastic crystals; one needs to be confident that the behavior is reproducible equilibrium behavior as opposed to metastable behavior. Another interesting feature of plastic crystal thermal behavior is that the shape of the transitions is often nonsymmetric and strongly resembles a lambda-like transition. Such transitions result from the cooperative nature of the phenomenon leading to the phase transition; for example, rotational motion initially may occur in only some fraction of the molecules/ions in the crystal lattice, but with increasing temperature, a larger fraction of molecules undergo the same motions, assisted by those already rotating, until at some culminating temperature the entire sample is undergoing identical motions. Figure 25.4 shows an example of this in the thermal analysis behavior of

c25.indd 354

1/20/2011 10:28:46 AM

355

CONDUCTIVITY IN ORGANIC IONIC PLASTIC CRYSTALS

Figure 25.4. Plot of the DSC trace (solid line) and conductivity (scatter graph) as a function of temperature for tetramethylammonium dicyanamide. The changes in conductivity with temperature mirror the thermal transitions, with an increase up to phase III and a dramatic increase on melting. As the temperature increases, there is a greater degree of internal motion within the material, and consequently, the conductivity increases.

[Me4N][N(CN)2]. This behavior is consistent with NMR second moment analyses (Figure 25.5) [80] and with heat capacity data of a related compound, [Et4N][N(CN)2] (Figure 25.6) [81]. The broad melting associated with many organic ionic plastic crystal compounds investigated in our laboratories is suggested to result from the large fraction of vacancies present in phase I of the most plastic materials, which can lead to an impurity effect in which the impurities are essentially the vacancy defects. This is consistent with the variable shape of the final melt that is often observed for the most plastic samples. Doping of lithium salts into organic ionic plastic crystals usually leaves the solid–solid phases unchanged but decreases and broadens the final melting point of the compound, consistent with liquidus behavior. For higher dopant concentrations, extra thermal transitions often appear, which is consistent with eutectic melting and the presence of a new lithium-rich phase. Figure 25.7 shows the change in thermal behavior of [C1mpyr][N(CN)2] when doped with Li N(CN)2. The appearance of a second, isothermal peak at about 32 °C, which is present even at 2 wt% of LiN(CN)2, is suggestive of a relatively low solubility limit of this lithium salt in [C1mpyr][N(CN)2]. The final melting peak shifts to lower temperatures, as is expected when a second component (or impurity) is mixed into a pure system. 25.4

CONDUCTIVITY IN ORGANIC IONIC PLASTIC CRYSTALS

The conductivity behavior of several different dimethylpyrrolidinium salts is illustrated in Figure 25.8. In all cases, high conductivity is observed in phase I,

c25.indd 355

1/20/2011 10:28:46 AM

356

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS (a)

138K

182K

207K

240K

280K

328K

370K

390K

-120

-90

-60

-30

0

30

60

90

120

Frequency (kHz)

Figure 25.5. (a). The change in single-pulse static 1H NMR spectra with temperature for tetramethylammonium dicyanamide. The spectra show line narrowing as the temperature increases, indicating the onset of various rotational modes within the material. Multiple peaks within each spectrum indicate multiple modes of rotation.

the highest being displayed by the BF4 salt, which shows conductivity in excess of 5 × 10−3 S cm−1 at 110 °C. Some salts show a distinct step in conductivity at one or more of solid–solid phase transition temperatures. This behavior also was observed in [Et4N][N(CN)2], presented in Figure 25.6, and may be associated with an increased number of vacancy defects, as discussed later. The nature of the cation also can have a strong influence on the plastic behavior and conductivity of the salt, as revealed by a comparison of a range of imidazolium ([Cnmim]), pyrrolidinium ([Cnmpyr]), and tetraalkylammonium ([R4N]) bis(trifluoromethanesulfonyl)amide (NTf2) salts [82]. The small [Cnmpyr] salts exhibit significantly higher conductivities than the [R4N][NTf2] salts studied, and it is proposed that this high conductivity is caused by the rotation of the cation, which does not occur for the [R4N] salts studied. [R4N]

c25.indd 356

1/20/2011 10:28:46 AM

357

CONDUCTIVITY IN ORGANIC IONIC PLASTIC CRYSTALS (b)

percentage of rotation

100% Rigid

80%

Isotropic rotation

60% 40%

Me group ratcheting

20%

Me group spinning

0% 120

160

200

240

280

320

360

400

Temperature / K Figure 25.5 (b). The relative contributions of the predominant modes of cation rotation within the tetramethylammonium dicyanamide salt across a range of temperatures. The static state decreases rapidly with increasing temperature, as does the methyl group spinning. Isotropic tumbling begins to a small degree at 240 K and increases dramatically above 315 K. These rotational transitions are unusual for the tetramethylammonium cation.

(c)

18 Me group ratcheting

15 FWHM (kHz)

Isotropic rotation 12 9 6 3 0 120

160

200

240

280

320

360

400

T (K) Figure 25.5 (c). Temperature dependence of peak width for the two highest modes of cation rotation within tetramethylammonium dicyanamide. The data show that the onset of rotations is not sudden, but occurs over a broad temperature range. As the temperature increases, the number of species undergoing isotropic rotations increases up to the solid–solid phase transition.

c25.indd 357

1/20/2011 10:28:46 AM

358

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS

0.320

Volume / (cm3)

0.315 0.310 0.305 0.300 0.295 240

260

280

300

320

340

360 4000 3500

–2 –3

3000

–4

2500

–5

2000

–6

conductivity 1500 heat capacity

–7

1000

–8 phase II

–9

phase I

melt

– 10 240

260

280 300 320 temperature / K

340

Heat capacity / (J mol–1 K–1)

log conductivity / (S cm–1)

–1

500 0 360

Figure 25.6. The correlation among changes in volume (top), conductivity, and thermal behavior with temperature for tetraethylammonium dicyanamide. There is an extremely large volume expansion on moving from phase II to phase I indicating considerable vacancy formation. This is mirrored by the simultaneous increase in conductivity.

salts that are plastic and exhibit significant ion conduction can be achieved when combined with smaller, more symmetrical anions such as Cl, Br, or SCN [8,71,83], but these salts have much higher melting points than the NTf2 analogues. However, the link between the rotator motions of the [Cnmpyr] cation and the conductivity mechanism has still not been fully elucidated. It is hypothesized that ion transport results from lattice defects and that the thermodynamic cost of formation of these lattice defects is lower if there is ion rotation. However, the conductivity mechanism may change with even small changes

c25.indd 358

1/20/2011 10:28:46 AM

359

CONDUCTIVITY IN ORGANIC IONIC PLASTIC CRYSTALS

[C1mpyr][N(CN)2] + 8% Li N(CN)2 [C1mpyr][N(CN)2 ] + 5% Li N(CN)2

[C1mpyr][N(CN)2 ] + 2% Li N(CN)2

pure [C1 mpyr][N(CN)2] -20

0

20

40

60

80

100

120

140

160

180

200

temperature / °C Figure 25.7. The change in thermal behavior of [C1mpyr][N(CN)2] with increasing lithium concentration. As the concentration increases, the melting transition gets broader and weaker and an additional transition appears at around 30 °C. This is a result of the eutectic melting and the formation of a new phase rich in lithium.

in cation structure; the activation energies and the temperature dependence of the conductivity of the [C1mpyr][NTf2] and [C2mpyr][NTf2] salts across the phase changes is different [77,82]. The effect of doping organic ionic plastic crystals with lithium salts containing the same anion has been investigated. The addition of dopants can result in conductivity increases of several orders of magnitude and can produce materials suitable for use in devices such as lithium batteries, as discussed later. For example, an increase in the conductivity of phases I and II of [C2mpyr] [NTf2] of up to three orders of magnitude has been observed when the material is doped with Li NTf2 [5,32]. Similar enhancements of ion conductivity are observed when Li N(CN)2 is doped into the plastic crystal phase of [C1mpyr] [N(CN)2] (Figure 25.9). Because the anion is common to both the additive and the matrix, the lithium can be thought of as the cation dopant. The effect of the dopant on the conductivity is dramatic given the small amounts of lithium involved. The addition of LiI to [C1mpyr][I] results in an almost three orders of magnitude increase in ionic conductivity, which results from the relatively mobile lithium ions as well as an increase in the mobility of the matrix cations [34]. Although the material is not particularly plastic, the doping behavior of this

c25.indd 359

1/20/2011 10:28:46 AM

360

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS

log10 conductivity (S cm-1)

-2 -3

[C1 mpyr][BF4 ] [C1mpyr][PF6 ] [C1mpyr][N(CN)2] [C1mpyr][NTf2 ]

-4 -5 -6 -7 -8 -9 300

350

400

temperature / K Figure 25.8. Comparison of the conductivity of different N,N-dimethylpyrrolidinium species. The significant differences in conductivity for the four species again highlight the importance of anion type in determining the physical properties of the material. The conductivity of [C1mpyr][PF6] is considerably worse than the BF4 species, as is the NTf2 species, until the melt at 400 K.

Figure 25.9. The effect of lithium doping on the conductivity of [C1mpyr][N(CN)2]. The addition of even small amounts (2%) of LiN(CN)2 results in a dramatic increase in the conductivity of the material.

c25.indd 360

1/20/2011 10:28:46 AM

CONDUCTIVITY IN ORGANIC IONIC PLASTIC CRYSTALS

361

salt is interesting as it demonstrates the first direct evidence of solid solution behavior in the pyrrolidinium family of salts. It is important to note, however, that there is a maximum solid solubility of LiI in [C1mpyr][I], and above this concentration, a second, low-conductivity phase is produced, and this is likely to also occur in other lithium-doped plastic crystal electrolyte systems. The existence of a solid solution in the LiI in [C1mpyr][I] system was evidenced by the fact that the phase transitions of the material were shifted to lower temperatures with doping. A change in transition temperatures also is observed when the plastic crystal [C2mpyr][BF4] is modified with supercritical CO2 [84]. This increases the ionic conductivity and also stabilizes the most ordered phase to a lower temperature. In contrast, the addition of a miscible polymer to the plastic crystal [C3mpyr][PF6] significantly increases the mechanical strength of the material, and decreases the conductivity, but without affecting its phase behavior [85]. It is interesting to observe that just 2 wt% of poly(vinyl pyrrolidone) (PVP) is capable of suppressing plastic behavior and reducing the ionic conductivity of an organic ionic plastic crystals; optical microscopy comparisons of [C3mpyr][BF4]-based systems suggest that the crystal habitat changes substantially with polymer additions, although an explanation of this phenomenon is yet to be established [77,86]. Recently, the use of nanoparticles such as TiO2 and SiO2 as additives for plastic crystals has received increasing attention. The addition of nanosized TiO2 or SiO2 to [C2mpyr][NTf2] increases the ionic conductivity by one or two orders of magnitude, respectively [87,88]. Both types of filler show an optimum concentration of about 10 wt% in the plastic crystal. It is postulated that the conductivity enhancement is associated with an increase in the number of mobile defects, as well as an increase in the plastic flow of the material (which increases the mobility in the matrix). However, although both lithium doping and addition of SiO2 nanoparticles to the plastic crystal result in significant conductivity enhancements, the use of these two additives together actually results in a decrease in conductivity [89]. This is probably a result of either a decrease in the lithium and/or anion transport or from disruption of the grain boundary regions. It is possible to address this via the use of lithiumfunctionalized silica nanoparticles, which improve the conductivity of the plastic crystal system, but the enhancement is not as substantial as with the addition of unsubstituted nanoparticles [90]. Analysis of these nanocomposites indicates that the functionalized nanoparticles cause an increase in the size and number of defects in the plastic crystal matrix, with a concomitant increase in both the number of mobile cations and anions as well as their mobility through the material. As the higher mobility results from an increase in defect size and number, we postulate that the conductivity enhancement is a result of “strain-induced defects” [91,92]. However, more studies are needed to investigate the influence of nanoparticle fillers fully on OIPCs and to allow the promising properties of these nanocomposite electrolyte materials to be fully exploited [77].

c25.indd 361

1/20/2011 10:28:47 AM

362

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS

Figure 25.10. Space-filled model of the N,N-dimethylpyrrolidinium cation. The 3D model shows the spherical nature of the cation and gives an indication as to how easy rotation around the axis should be given the lack of steric hindrance.

25.5 TRANSPORT MECHANISMS IN PLASTIC CRYSTAL PHASES The NMR data presented in Figure 25.5 indicates that in the high-temperature phases of [Me4N][N(CN)2], rotational motions of some or all ions take place. Figure 25.10 shows how such rotations are certainly possible in the case of, for example, dimethyl pyrrolidinium salts. In this case, the rotation around the two-fold axis seems to involve little steric hindrance. Such rotations cause the crystal lattice symmetry to become of a higher order because the rotational asymmetry disappears. As a result, the X-ray diffraction pattern exhibits fewer reflections in the higher temperature phases, as exemplified by Figure 25.11. However, this is not always observed; the X-ray diffraction pattern of [C1mpyr] [BF4] shows an unusual transition to a more complex XRD pattern with heating into phase I [93]. The conductivity of this material increases up to the phase II to I transition, when it plateaus, and 1H NMR analysis indicates that at these higher temperatures the material exhibits lower cation mobility (or shorter interionic cation distances), whereas Raman spectroscopy also suggests a change in the environment or mobility of the cation in phase I. This indication of a significant structural change in phase I is consistent with the XRD analysis, which shows changes upon heating from a simple spectrum in phase II to a more complex pattern in phase I (Figure 25.11b). In the situation in which full isotropic motion of both ions occurs, a simple cubic symmetry can be observed, as in phase 1 of [Me4N][N(CN)2]. The onset

c25.indd 362

1/20/2011 10:28:47 AM

363

TRANSPORT MECHANISMS IN PLASTIC CRYSTAL PHASES (a)

(b)

120oC

I

105oC 100oC

II

95oC 75oC 65oC

III

60oC 25oC

16

18

20

22

24

26

28

2θ Figure 25.11. (a) The X-ray diffraction pattern of [Me4N][N(CN)2] shows the change in lattice parameters on moving between phases. As the temperature increases, there is a loss of some peaks and the appearance of new ones, indicating a change in unit cell. Generally, the diffraction patterns become simpler on moving to lower phases, culminating in only one peak for the phase I species. Leaving the material for 2 days does not seem to have any effect on the diffraction pattern and hence on the structure of the salt. (b) In contrast, the X-ray powder diffraction of [C1mpyr][BF4] as a function of temperature shows an unusual transition to a more complex pattern with increasing temperature into phase I [93].

c25.indd 363

1/20/2011 10:28:47 AM

364

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS

of such rotations previously have been linked to phase changes into what are known as “rotator phases” [2,53], which in turn has been associated with the formation of vacancies, presumably because of the expansion of the lattice and the subsequent decrease in the strength of intermolecular/interionic interactions. Interestingly, in most pyrrolidinium-based systems, and in tetraalkylammonium dicyanamide, the onset of these rotations is not instantaneous but rather occurs over a broad temperature range, as shown in Figure 25.5 for [Me4N][N(CN)2]. In this case, NMR second moment analyses have shown that several intermediate situations develop, in which some species are undergoing restricted rotations and others isotropic rotation. With increasing temperature, the number of species undergoing isotropic rotations increases dramatically until finally this is the dominant process, coinciding with the peak of the thermal phase transition. Interestingly, in the case of [Et4N][N(CN)2], the transition into phase I is accompanied by a substantial increase in volume as well as a subsequent decrease in the rate of change of volume for phase I compared with phase II. Although we have not yet quantified the vacancy size and number in this dicyanamide system, other work on the pyrrolidinium tetrafluoroborate family of salts using positron annihilation lifetime spectroscopy has suggested that a significant step in the vacancy size occurs concomitantly with a drop in apparent vacancy number. This suggests that the coalescence of vacancies may occur, leading to larger scale defects such as dislocations. Evidence for this also has been found through SEM analysis [33]. Figure 25.12 presents an scanning electro microgram micrograph for [C3mpyr][BF4] showing evidence of slip

Figure 25.12. Scanning electron micrograph of [C3mpyr][BF4] showing evidence of slip planes.

c25.indd 364

1/20/2011 10:28:47 AM

365

Diffusion Coefficient / (m2 s-1)

TRANSPORT MECHANISMS IN PLASTIC CRYSTAL PHASES

1E-10

1E-11 19 19

F in [C3 mpyr][BF4 ]

F in 2wt% LiBF4 in [C3mpyr][BF4 ]

7

Li in 2wt% LiBF4 in [C3 mpyr][BF4]

1

H in [C3 mpyr][BF4 ]

1E-12

1

H in 2wt% LiBF4 in [C3 mpyr][BF4 ]

290

300

310

320

330

340

350

temperature / K Figure 25.13. The effect of lithium doping and temperature on the 7Li, 19F, and 1H diffusion in [C3mpyr][BF4]. Doping the material with 2% Li BF4 results in a decrease in cation diffusion. It is believed that an increase in salt concentration results in an increased ion interaction and thus a loss of mobility.

planes separated by about 50 nm. Such defects are known to enhance selfdiffusion by a process known as pipe-diffusion, whereby the defect acts as a pipe along which molecules can diffuse [2,94]. Diffusion coefficients for different species in organic ionic plastic crystals have been determined via pulsed field gradient NMR measurements and have shown that, despite its apparently large size, the pyrrolidinium cation diffuses rapidly in the higher temperature solid-state phases. For example, in [C2mpyr] [NTf2], diffusion coefficients of the order of 10−11 m2 s−1 have been measured [33]. Figure 25.13 shows the temperature-dependent diffusion coefficients for the cation and anion in the [C3mpyr][BF4] system. It is shown that although both the cation and the anion are diffusive in this system, the pyrrolidinium cation is more mobile, despite its larger size. The addition of LiBF4 to [C3mpyr] [BF4] leads to a slight decrease in the diffusivity of the matrix ions in this case (which is in contrast to the effect of Li NTf2 additions to [C2mpyr][NTf2] [32]). In this system, the lithium and pyrrolidinium cation diffusivities are comparable, and the anion is diffusing at a faster rate. In previous work where Li NTf2 was doped into [C2mpyr][NTf2], 7Li NMR line-width measurements showed that even in the lower temperature phases some fraction of the lithium species is mobile, until eventually in phase I, all lithium ions are diffusing, as shown by a single, narrow resonance [32]. Changes in the ionic interactions with temperature in the organic ionic plastic crystal [C1mpyr][SCN] also have been studied in detail [95]. This

c25.indd 365

1/20/2011 10:28:48 AM

366

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS

material has a solid–solid phase transition at 82 °C, melts at 122 °C, and has conductivity in phase I of around 10−3 S cm−1. There is a one order of magnitude step in the conductivity at each of the phase transitions. The activation energy decreases by a factor of two at each transition, which indicates a significant decrease in the energy barrier to ionic mobility. The vibrational modes of the thiocyanate anion are sensitive to its coordination environment, and Raman spectroscopy confirms that the cation, and probably the anion, are undergoing isotropic tumbling in phase I. 1H NMR analysis of the [C1mpyr][SCN] shows a significant decrease in the linewidths across the phase II to phase I transition, indicating an onset of mobility of the cation, which is consistent with the conductivity increase. The second moment analysis of the 1H NMR data indicates that the cation goes from being static in phase II to full isotropic tumbling in phase I. This is an unusually large increase in rotational mobility; the iodide and NTf2 salts of this cation [96], and the plastic crystal [Me4N][DCA] [80], all go through several phases with anisotropic tumbling of the cations before the state of isotropic tumbling is reached [77].

25.6

DEVELOPING AREAS OF RESEARCH

In recent years, the potential application of organic ionic plastic crystals as solid-state electrolytes, for example, in lithium batteries, has been increasingly recognized. This increasing interest has resulted in numerous new materials being reported in recent years [97–103], including 52 different salts using quaternary ammonium cations and a range of perfluoroalkyltrifluoroborate anions, several of which exhibit one or more solid–solid phase transitions [98]. A range of different fluorinated anions also can be used (Figure 25.14) [104,105]. The nonafluoro-1-butanesulfonate salts are particularly interesting materials as all six of the series are in the plastic phase I at ambient temperature, up to at least 90 °C, and they exhibit excellent electrochemical and thermal stability. A new class of proton-conducting OIPC also has been reported (Figure 25.14b) [99]. These use the dihydrogenphosphate (DHP) anion, which can generate protons as a carrier and thereby eliminates the need for the addition a dopant acid that may be incompatible with the host matrix. Choline DHP is in phase II between 23 °C and 119 °C, and the conductivity increases from 10−6 to 10−3 S cm−1 in phase I. This phase behavior makes it suitable for use in fuel cells, which is currently being investigated. The applicability of organic ionic plastic crystals in lithium batteries also has made significant progress in recent years, driven by their nonflammability, nonvolatility, plasticity, and high electrochemical and thermal stability [77]. For example, a range of doped N,N’-cyclized pyrazolium NTf2 plastic crystal electrolytes have been studied for use in lithium ion batteries (Figure 25.14c) [106–109]. However, the relatively poor cathodic stability of these materials

c25.indd 366

1/20/2011 10:28:48 AM

DEVELOPING AREAS OF RESEARCH

367

Figure 25.14. New organic ionic plastic crystal materials recently reported, (a) with new fluorinated anions [104], [105], (b) proton conducting OIPCs [99], and (c) cations used in NTf2-based plastic crystals for lithium battery applications.

c25.indd 367

1/20/2011 10:28:48 AM

368

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS

(∼1.4 V vs. Li+/Li) limits the electrolytes to application in lithium ion batteries rather than in lithium metal batteries. Quaternary ammonium salts using the perfluoroalkyltrifluoroborate anion can be electrochemically stable, and a 5 mol% LiCF3BF3 doped electrolyte displays efficient lithium deposition and dissolution at 25 °C [98], which is promising for their application in lithium battery systems. Thus far, there are no reports describing their application in cells with lithium intercalation electrodes. There are also a range of other lithium-compatible organic ionic plastic crystals that have the physical properties and electrochemical stability sufficient for lithium battery applications, such as the [Cnmpyr][BF4] salts (n = 1 − 4) [29], [Cnmpyr] nonafluoro-1-butane sulfonate (NfO) salts [104], and a range quaternary ammonium NTf2 salts, (Figure 25.14) [110–112] that support lithium deposition and dissolution. The [Cnmpyr][NTf2] plastic crystals also may be suitable for use in lithium metal batteries [82], [113]; the related ionic liquids have been investigated extensively and shown to support stable lithium electrochemistry [114,115]. Indeed, LiNTf2-doped [C2mpyr][NTf2] plastic crystal electrolytes can support substantial currents after initial cycling (approaching 0.5 mAcm−2 at 50 °C) [116]. There is a reduction in cell impedance with cycling (Figure 25.15), which seems to be related to a change in the microstructure of the plastic crystal as a result of the applied current. This behavior is believed to be related to the formation of improved interfacial contact between the electrode and the solid electrolyte (Figure 25.15b) [77].

(a)

(b) 1.5

Potential (V)

1 0.5 0 -0.5 -1 -1.5 0

10

20

30

40

50

Time (h)

Figure 25.15. (a) Galvanostatic cycling (0.01 mAcm−2) of a symmetrical lithium cell using 1 mol% LiNTf2 in [C2mpyr][NTf2] at 50 °C, and (b) back-scattered SEM image of the lithium/electrolyte interface taken from a “conditioned” cell. The dark region at the interface implies the formation of a conductive lithium-rich interphase that is approximately 10 μm thick [116]. Reproduced from Ref. 77 with permission from The Royal Society of Chemistry.

c25.indd 368

1/20/2011 10:28:48 AM

REFERENCES

25.7

369

CONCLUDING REMARKS

Ion-conducting plastic crystalline materials are promising solid-state electrolytes for a variety of electrochemical device applications. In their soft, plastic form, they posses close to the ideal material properties for a solid electrolyte. They are nonflammable, and as ionic materials, they have negligible vapor pressures, thereby eliminating problems associated with the loss of the electrolyte by evaporation. However, the ideal plastic crystalline electrolyte material is yet to be found. For an ambient temperature electrolyte, a material that exists in its plastic phase over a wide temperature range (i.e., between 0 °C and 100 °C) would be desirable. For Li and Li-ion battery use, a lithium doped material would be ideal, but it also would be preferable for the lithium ion to be the dominant charge carrier, whereas in the present systems, the matrix ions also contribute to the conductivity in the higher temperature phases. Nevertheless, the use of doped organic ionic plastic crystals in lithium ion batteries is starting to show real promise, which suggests that the wider use of organic ionic plastic crystals as electrolytes in a range of electrochemical devices soon could be a reality. Plastic crystals present many challenges in terms of elucidating the mechanisms of rotational motion, conduction, diffusion, mechanical deformation, and the interrelationship between these mechanisms. Furthermore, understanding the effect of doping or mixing these systems with additional components such as acids, inorganic salts, nanoparticles, or polymers on these transport mechanisms is still a challenge. The range of techniques used for the study of these unique materials is growing, with the application of PALS, X-ray crystallography, Raman spectroscopy, solid-state NMR, and so on, all providing valuable information about the conduction mechanisms as well as the physical changes that occur at the phase transitions. The development of new organic ionic plastic crystals is also paramount in enhancing our understanding of these materials and taking us closer to the target-specific design of plastic crystals with specific physical properties.

REFERENCES 1. J. Timmermans, J. Phys. Chem. Solid, 18, 1 (1961). 2. J. N. Sherwood, The Plastically Crystalline State (Orientionally Disordered Crystals), Wiley-Interscience, 1979. 3. G. J. Kabo, A. V. Blokhin, M. B. Charapennikau, A. G. Kabo, V. M. Sevruk, Thermochim. Acta, 345, 125 (2000). 4. D. R. MacFarlane, P. Meakin, J. Sun, N. Amini, M. Forsyth, J. Phys. Chem. B, 103, 4164 (1999). 5. D. R. MacFarlane, J. Huang, M. Forsyth, Nature, 402, 792 (1999). 6. H. Every, A. G. Bishop, M. Forsyth, D. R. MacFarlane, Electrochim. Acta., 45, 1279 (2000).

c25.indd 369

1/20/2011 10:28:49 AM

370

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS

7. M. Tansho, D. Nakamura, R. Ikeda, Ber. Bunsen-Ges. Phys. Chem., 95, 1643 (1991). 8. M. Tansho, D. Nakamura, R. Ikeda, J. Chem. Soc., Faraday Trans., 87, 3255 (1991). 9. M. Tansho, Y. Furukawa, D. Nakamura, R. Ikeda, Ber. Bunsen-Ges. Phys. Chem., 96, 550 (1992). 10. E. I. Cooper, C. A. Angell, Solid State Ionics, 18–19, 570 (1986). 11. R. Aronsson, B. Jansson, H. E. G. Knape, A. Lunden, L. Nilsson, C. A. Sjoeblom, L. M. Torell, J. Phys., Colloq., 35 (1980). 12. R. Aronsson, H. E. G. Knape, L. M. Torell, J. Chem. Phys., 77, 677 (1982). 13. R. Aronsson, L. Boerjesson, L. M. Torell, Phys. Lett. A, 98A, 205 (1983). 14. L. Nilsson, J. O. Thomas, B. C. Tofield, J. Phys. C, 13, 6441 (1980). 15. M. Witschas, H. Eckert, D. Wilmer, R. D. Banhatti, K. Funke, J. Fitter, R. E. Lechner, G. Korus, M. Jansen, Z. Phys. Chem., 214, 643 (2000). 16. M. Witschas, H. Eckert, J. Phys. Chem. A, 103, 10764 (1999). 17. S. Fukada, H. Yamamoto, R. Ikeda, D. Nakamura, J. Chem. Soc., Faraday Trans., 83, 3207 (1987). 18. H. Ishida, R. Ikeda, D. Nakamura, J. Phys. Chem., 86, 1003 (1982). 19. H. Ishida, T. Iwachido, R. Ikeda, Ber. Bunsen-Ges. Phys. Chem., 96, 1468 (1992). 20. H. Ishida, N. Hayama, R. Ikeda, Chem. Lett., 1333 (1992). 21. H. Suga, M. Sugisaki, S. Seki, Mol. Cryst., 1, 377 (1966). 22. J. M. Chezeau, J. H. Strange, Phys. Rep., 53, 1 (1979). 23. H. M. Hawthorne, J. N. Sherwood, Faraday Trans., 66, 1792 (1970). 24. P. Derollez, J. Lefebvre, M. Descamps, W. Press, H. Fontaine, J. Phys. Condens. Matter, 2, 6893 (1990). 25. A. Masui, S. Yoshioka, S. Kinoshita, Chem. Phys. Lett., 341, 299 (2001). 26. D. Chandra, J. H. Helms, A. Majumdar, J. Electrochem. Soc., 141, 1921 (1994). 27. D. R. MacFarlane, S. A. Forsyth, J. Golding, G. B. Deacon, Green Chem., 4, 444 (2002). 28. J. Golding, N. Hamid, D. R. MacFarlane, M. Forsyth, C. Forsyth, C. Collins, J. Huang, Chem. Mater., 13, 558 (2001). 29. S. Forsyth, J. Golding, D. R. MacFarlane, M. Forsyth, Electrochim. Acta, 46, 1753 (2001). 30. J. M. Pringle, J. Golding, C. M. Forsyth, G. B. Deacon, M. Forsyth, D. R. MacFarlane, J. Mater. Chem., 12, 3475 (2002). 31. H. Every, Ph.D. Dissertation, Monash University (2001). 32. M. Forsyth, J. Huang, D. R. MacFarlane, J. Mater. Chem., 10, 2259 (2000). 33. D. R. MacFarlane, M. Forsyth, Adv. Mater., 13, 957 (2001). 34. J. Adebahr, A. Seeber, D. R. MacFarlane, M. Forsyth, J. Phys. Chem. B, 109, 20087 (2005). 35. S. Long, D. R. Macfarlane, M. Forsyh, Solid State Ionics, 161, 105 (2003). 36. C. A. Angell, L. E. Busse, E. I. Cooper, R. K. Kadiyala, A. Dworkin, M. Ghelfenstein, H. Szwarc, A. Vassal, J. Chim. Phys. Phys.-Chim. Biol., 82, 267 (1985). 37. T. Asanuma, H. Nakayama, T. Eguchi, N. Nakamura, Mol. Cryst. Liq. Cryst. Sci. Tech., 326, 395 (1999).

c25.indd 370

1/20/2011 10:28:49 AM

REFERENCES

371

38. V. Balevicius, A. Marsalka, V. Sablinskas, L. Kimtys, J. Mol. Struct., 219, 123 (1990). 39. R. H. Baughman, D. Turnbull, J. Phys. Chem. Solid, 33, 121 (1972). 40. B. Beck, E. Roduner, H. Dilger, P. Czarnecki, D. G. Fleming, I. D. Reid, C. J. Rhodes, Physica B, 289&290, 607 (2000). 41. M. Bee, J. P. Amoureux, Mol. Phys., 48, 63 (1983). 42. M. Bee, J. Chim. Phys. Phys.-Chim. Biol., 82, 205 (1985). 43. P. Bladon, N. C. Lockhart, J. N. Sherwood, Mol. Phys., 22, 365 (1971). 44. K. J. Adamic, S. G. Greenbaum, K. M. Abraham, M. Alamgir, M. C. Wintersgill, J. J. Fontanella, Chem. Mater., 3, 534 (1991). 45. N. Boden, J. Cohen, P. P. Davis, Mol. Phys., 23, 819 (1972). 46. S. Braeuninger, S.-Q. Dou, H. Fuess, W. Schmahl, R. Strauss, A. Weiss, Ber. BunsenGes. Phys. Chem., 98, 1096 (1994). 47. A. R. Britcher, J. H. Strange, Mol. Phys., 37, 181 (1979). 48. H. M. Hawthorne, J. N. Sherwood, J. Trans. Faraday Soc., 66, 1783 (1970). 49. H. M. Hawthorne, J. N. Sherwood, J. Trans. Faraday Soc., 66, 1799 (1970). 50. H. Honda, S. I. Ishimaru, N. Onoda-Yamamuro, R. Ikeda, Z. Naturforsch., A, 50, 871 (1995). 51. H. Honda, M. Kenmotsu, H. Ohki, R. Ikeda, Y. Furukawa, Ber. Bunsen-Ges., 99, 1009 (1995). 52. R. M. Hooper, J. N. Sherwood, Surf. Defect Prop. Solids, 6, 308 (1977). 53. K. Horiuchi, H. Takayama, S. I. Ishimaru, R. Ikeda, Bull. Chem. Soc. Jpn., 73, 307 (2000). 54. H. Ishida, R. Ikeda, D. Nakamura, Bull. Chem. Soc. Jpn., 59, 915 (1986). 55. H. Ishida, R. Ikeda, D. Nakamura, Bull. Chem. Soc. Jpn., 60, 467 (1987). 56. H. Ishida, N. Matsuhashi, R. Ikeda, D. Nakamura, J. Chem. Soc., Faraday Trans., 85, 111 (1989). 57. H. Ishida, K. Takagi, R. Ikeda, Chem. Lett., 605 (1992). 58. H. Ishida, Y. Furukawa, S. Kashino, S. Sato, R. Ikeda, Ber. Bunsen-Ges. Phys. Chem., 100, 433 (1996). 59. S. Iwai, R. Ikeda, D. Nakamura, Can. J. Chem., 66, 1961 (1988). 60. S. Iwai, M. Hattori, D. Nakamura, R. Ikeda, J. Chem. Soc., Faraday Trans., 89, 827 (1993). 61. D. W. Larsen, B. A. Soltz, F. E. Stary, R. West, J. Chem. Soc., Chem. Commun., 1093 (1978). 62. K. J. McGrath, R. G. Weiss, Langmuir, 13, 4474 (1997). 63. P. McKay, J. N. Sherwood, J. Chem. Soc., Faraday Trans., 71, 2331 (1975). 64. H. Ono, R. Seki, R. Ikeda, H. Ishida, J. Mol. Struct., 345, 235 (1995). 65. H. Ono, R. Ikeda, H. Ishida, Ber. Bunsen-Ges. Phys. Chem., 100, 1833 (1996). 66. H. Ono, S. i. Ishimaru, R. Ikeda, H. Ishida, Bull. Chem. Soc. Jpn., 70, 2963 (1997). 67. H. Ono, S. Ishimaru, R. Ikeda, H. Ishida, Ber. Bunsen-Ges. Phys. Chem., 102, 650 (1998). 68. H. Ono, S. I. Ishimaru, R. Ikeda, H. Ishida, Bull. Chem. Soc. Jpn., 72, 2049 (1999).

c25.indd 371

1/20/2011 10:28:49 AM

372

ION CONDUCTION IN ORGANIC IONIC PLASTIC CRYSTALS

69. L. S. Prabhumirashi, R. Ikeda, D. Nakamura, Ber. Bunsen-Ges. Phys. Chem., 85, 1142 (1981). 70. S. Sato, R. Ikeda, D. Nakamura, Ber. Bunsen-Ges. Phys. Chem., 86, 936 (1982). 71. T. Shimizu, S. Tanaka, N. Onoda-Yamamuro, S. I. Ishimaru, R. Ikeda, J. Chem. Soc., Faraday Trans., 93, 321 (1997). 72. T. Tanabe, D. Nakamura, R. Ikeda, J. Chem. Soc., Faraday Trans., 87, 987 (1991). 73. S. Tanaka, N. Onoda-Yamamuro, S. I. Ishimaru, R. Ikeda, Bull. Chem. Soc. Jpn., 70, 2981 (1997). 74. H. Ishida, K. Takagi, T. Iwachido, M. Terashima, D. Nakamura, R. Ikeda, Z. Naturforsch. A., 45, 923 (1990). 75. S. J. Pas, J. Huang, M. Forsyth, D. R. MacFarlane, A. J. Hill, J. Chem. Phys., 122, 064704/1 (2005). 76. J. Huang, A. Hill, M. Forsyth, D. MacFarlane, A. Hollenkamp, Solid State Ionics, 177, 2569 (2006). 77. J. M. Pringle, P. C. Howlett, D. R. MacFarlane, M. Forsyth, J. Mater. Chem., 20, 2056 (2010). 78. W. A. Henderson, V. G. Young, Jr., S. Passerini, P. C. Trulove, H. C. De Long, Chem. Mater., 18, 934 (2006). 79. A. H. Wesley, M. Herstedt, G. Y. Victor, Jr., S. Passerini, C. D. L. Hugh, C. T. Paul, Inorg Chem, 45, 1412 (2006). 80. A. J. Seeber, M. Forsyth, C. M. Forsyth, S. A. Forsyth, G. Annat, D. R. MacFarlane, Phys. Chem. Chem. Phys., 5, 2692 (2003). 81. G. Annat, J. Adebahr, I. R. McKinnon, D. R. MacFarlane, M. Forsyth, Solid State Ionics, 178, 1065 (2007). 82. D. R. MacFarlane, P. Meakin, N. Amini, M. Forsyth, J. Phys. Condens. Matter, 13, 8257 (2001). 83. M. Hattori, S. Fukada, D. Nakamura, R. Ikeda, J. Chem. Soc., Faraday Trans., 86, 3777 (1990). 84. S. J. Pas, J. Efthimiadis, J. M. Pringle, M. Forsyth, D. R. MacFarlane, J. Mater. Chem., 14, 2603 (2004). 85. S. J. Pas, J. M. Pringle, M. Forsyth, D. R. MacFarlane, Phys. Chem. Chem. Phys., 6, 3721 (2004). 86. J. Efthimiadis, G. Annat, J. Efthimiadis, D. R. MacFarlane, M. Forsyth, Solid State Ionics, 177, 95 (2006). 87. J. Adebahr, N. Ciccosillo, Y. Shekibi, D. R. MacFarlane, A. J. Hill, M. Forsyth, Solid State Ionics, 177, 827 (2006). 88. Y. Shekibi, A. Gray-Weale, D. R. MacFarlane, A. J. Hill, M. Forsyth, J. Phys. Chem. C, 111, 11463 (2007). 89. Y. Shekibi, S. J. Pas, N. M. Rocher, B. R. Clare, A. J. Hill, D. R. MacFarlane, M. Forsyth, J. Mater. Chem., 19, 1635 (2009). 90. Y. Shekibi, J. M. Pringle, J. Sun, S. J. Pas, N. M. Rocher, B. R. Clare, A. J. Hill, D. R. MacFarlane, M. Forsyth, J. Mater. Chem., 20, 338 (2010). 91. J. Maier, Prog. Solid State Chem., 23, 171 (1995). 92. J. Maier, Phys. Chem. Chem. Phys., 11, 3011 (2009).

c25.indd 372

1/20/2011 10:28:49 AM

REFERENCES

373

93. J. M. Pringle, J. Adebahr, D. R. MacFarlane, M. Forsyth, Phys. Chem. Chem. Phys. 2010. In Press. 94. D. A. Porter, K. E. Easterling, Phase Transformation in Metals and Alloys, CRC Press, 1981. 95. J. Adebahr, M. Grimsley, N. M. Rocher, D. R. MacFarlane, M. Forsyth, Solid State Ionics, 178, 1798 (2008). 96. J. Adebahr, A. J. Seeber, D. R. MacFarlane, M. Forsyth, J. Appl. Phys., 97, 093904/1 (2005). 97. J. Sun, D. R. MacFarlane, M. Forsyth, Solid State Ionics, 148, 145 (2002). 98. Z.-B. Zhou, H. Matsumoto, K. Tatsumi, Chem. Eur. J., 12, 2196 (2006). 99. M. Yoshizawa-Fujita, K. Fujita, M. Forsyth, D. R. MacFarlane, Electrochem. Commun., 9, 1202 (2007). 100. J. Efthimiadis, G. J. Annat, J. Efthimiadis, M. Forsyth, D. R. MacFarlane, Phys. Chem. Chem. Phys., 5, 5558 (2003). 101. P. M. Dean, B. R. Clare, V. Armel, J. M. Pringle, C. M. Forsyth, M. Forsyth, D. R. MacFarlane, Aust. J. Chem., 62, 334 (2009). 102. S. A. Forsyth, D. R. MacFarlane, J. Mater. Chem., 13, 2451 (2003). 103. S. A. Forsyth, S. R. Batten, Q. Dai, D. R. MacFarlane, Aust. J. Chem., 57, 121 (2004). 104. S. A. Forsyth, K. J. Fraser, P. C. Howlett, D. R. MacFarlane, M. Forsyth, Green Chem., 8, 256 (2006). 105. K. M. Johansson, J. Adebahr, P. C. Howlett, M. Forsyth, D. R. MacFarlane, Aust. J. Chem., 60, 57 (2007). 106. Y. Abu-Lebdeh, P.-J. Alarco, M. Armand, Angew. Chem., Int. Ed., 42, 4499 (2003). 107. P.-J. Alarco, Y. Abu-Lebdeh, N. Ravet, M. Armand, Solid State Ionics, 172, 53 (2004). 108. Y. Abu-Lebdeh, A. Abouimrane, P.-J. Alarco, M. Armand, J. Power Source, 154, 255 (2006). 109. Y. Abu-Lebdeh, P. J. Alarco, A. Abouimrane, L. Ionescu-Vasii, A. Hammami, M. Armand, J. New Mater. Electrochem. Syst., 8, 197 (2005). 110. M. Yoshizawa-Fujita, D. R. MacFarlane, P. C. Howlett, M. Forsyth, Electrochem. Commun., 8, 445 (2006). 111. T. Ruther, J. Huang, A. F. Hollenkamp, Chem. Comm., 5226 (2007). 112. Y. Abu-Lebdeh, E. Austin, I. J. Davidson, Chem. Lett., 38, 782 (2009). 113. J. Huang, M. Forsyth, D. R. MacFarlane, Solid State Ionics, 136–137, 447 (2000). 114. P. C. Howlett, N. Brack, A. F. Hollenkamp, M. Forsyth, D. R. MacFarlane, J. Electrochem. Soc., 153, A595 (2006). 115. P. C. Howlett, D. R. MacFarlane, A. F. Hollenkamp, Electrochem. Solid State Lett., 7, A97 (2004). 116. P. C. Howlett, Y. Shekibi, D. R. MacFarlane, M. Forsyth, Adv. Eng. Mater., 11, 1044 (2009).

c25.indd 373

1/20/2011 10:28:49 AM

26 LIQUID CRYSTALLINE IONIC LIQUIDS Takashi Kato and Masafumi Yoshio

Liquid crystals are anisotropically ordered and mobile materials [1] and have been successful in the field of electrooptic devices and high-strength fibers [1]. In addition, they are expected to serve as anisotropic conductors because of their self-organized structures. Recently, transportation of charges [2–5] and ions [6–30] in liquid crystals has attracted much attention. For this purpose, the design and control of molecular interactions and microphase-segregated structures in liquid crystals is essential [6–10,31–34]. Moreover, the macroscopic orientation of self-organized monodomains plays a key role in the enhancement of properties because the boundary in randomly oriented polydomains disturbs high and anisotropic transportation of charges and ions. Percec [35–37], Nolte [38], Lehn [39], and Kimura [40] incorporated oligo(ethylene oxide)s (PEOs), crown ethers, and azacrowns, which function as ion-conducting moieties into mesogenic rodlike and fanlike molecules. These materials form layer and columnar structures. The ionic conductivities of some of these materials were measured. However, the ionic conductivities of these earlier examples were lower than expected because no macroscopic orientation had been achieved. Recently, Kato, Ohno, and coworkers have achieved macroscopic orientation of smectic liquid crystals containing PEO moieties with lithium salts and succeeded in anisotropic (two-dimensional) measurements of the ionic conductivities [11–17]. For the PEO-based columnar materials, one-dimensional ion conduction in a macroscopic monodomain has not yet been measured. These PEO-based anisotropic materials have potential applications as functional electrolytes in lithium ion batteries. Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

375

c26.indd 375

1/20/2011 10:28:55 AM

376

LIQUID CRYSTALLINE IONIC LIQUIDS

Recently, ionic liquid-based liquid crystalline materials have been developed as a new family of anisotropic (one- or two-dimensional) electrolytes [18–26] and isotropic electrolytes forming three-dimensionally interconnected ion channels [27]. Ionic liquids are isotropic organic liquids consisting entirely of ions [41–47]. They have advantages for applications as nonvolatile liquid electrolytes in a variety of electrochemical devices such as lithium ion batteries, capacitors, dye-sensitized solar cells, organic light-emitting diodes, and actuators. In this chapter, we review the development of liquid crystalline materials based on ionic liquids. Self-organization behavior and ion-conductive properties of these anisotropic materials are described.

26.1 LIQUID CRYSTALLINE IONIC LIQUIDS BY THE CHEMICAL MODIFICATION OF IONIC LIQUIDS Ionic liquid crystals are liquid crystals containing ionic moieties [6,7,48]. Ionic liquid crystals with ammonium and pyridinium salts were reported to show stable thermotropic liquid crystalline behavior [49–54]. Bowlas et al. showed that some imidazolium and pyridinium salts exhibit low melting points and liquid crystalline behavior [55]. Ionic liquid crystals with lower temperature mesophase ranges are called liquid crystalline ionic liquids. A large variety of thermotropic liquid crystalline ionic liquids exhibiting anisotropic fluid states were prepared. In particular, the ionic liquid crystals based on imidazolium salts 1 [56–60], 2 [61], 3 [62], and 4 [63] as well as pyridinium salts 5 [56] containing weakly coordinating perfluorinated anions, such as BF4− and PF6− are representative liquid crystalline ionic liquids because of their thermal and electrochemical stabilities (Figure 26.1). Liquid crystalline ionic liquids with a benzimidazolium unit 6 also were prepared [64]. For these materials, liquid crystalline phases are induced by the microphase segregation of ionic moieties and long alkyl or perfluoroalkyl chains. The types of liquid crystalline phases exhibited are dependent on the molecular shape and the volume ratio of incompatible ionic and nonionic parts. Most ionic molecules containing single chains form bilayer smectic structures with the interdigitation of alkyl chains. The influence of the anion type and chain length on the liquid crystalline phases has been systematically studied for 1-alkyl-3-methylimidazolium salts 1(n/X−), where n is the alkyl chain length and X− indicates the anion species [56–59,65]. The salts with X− = Cl−, Br−, BF4−, PF6−, and CF3SO−3 (n = 12, 14, 16, 18) form smectic A phases. For example, 1(12 /BF4−) shows a smectic A phase from 26 °C to 39 °C, and 1(18 /BF4− ) exhibits a smectic A phase from 67 °C to 215 °C. The melting and clearing points increase with increasing chain length. The triflate salts 1(12 /CF3SO−3 ) and 1(14 /CF3SO−3 ) as well as the bis(trifluoromethanesulfonyl)imide (TFSI) salts 1(n/TFSI−) with alkyl chains of n = 12–18 are nonmesomorphic. The order of the stabilization effects on liquid crystalline structures is Cl − >Br − > BF4− > PF6− > CF3SO−3 > (CF3SO2 )2 N − .

c26.indd 376

1/20/2011 10:28:55 AM

LIQUID CRYSTALLINE IONIC LIQUIDS

377

Figure 26.1. Molecular structures of ionic liquid crystals.

The formation of layered assemblies can be induced by the addition of a small amount of water to isotropic ionic liquid 1(10/Br−) [66,67]. A lyotropic liquid crystalline gel consisting of 1(10/Br−) and water of 16 wt% has been prepared. The addition of water to 1(10/Br−) induces the formation of a lamellar structure with hydrogen bonding between the imidazolium ring (in particular, the proton on C2) and Br−, and Br−, and water, whereas a dried 1(10/Br−) including water of less than 1.6 wt% shows no liquid crystalline phases. Metal-based ionic liquid crystals containing a tetrahedral tetrahalometalate ion (MX 2− 4 ; X = Cl, Br, I; M = Co, Ni, Cu, Pd, etc.) 7 [55,68] and 8 [69–71] were studied (Figure 26.1). They provide various coordination geometries. For

c26.indd 377

1/20/2011 10:28:55 AM

378

LIQUID CRYSTALLINE IONIC LIQUIDS

example, N-alkylpyridinium salts 8, often denoted as [CnPy]2[CuCl4] (n = 12–18), show a variety of liquid crystalline phases from smectic to columnar or even cubic phases [71]. These metal-based materials can exhibit properties as metal complexes such as chromism, magnetism, polarizability, redox behavior, and catalysis. Bipyridinium cation 9 [72], phosphonium [73–76], guanidinium [77], and vinamidinium [78] cations were used to form thermotropic ionic liquid crystals. For example, the mixture of diheptyl and dioctyl viologens 9 (20 : 80 by wt%) with TFSI anion forms a smectic A phase between 22 °C and 132 °C. It was reported that the viologens show fluorescence in both nonpolar and polar organic solvents, which can be useful for the development of biological and chemical sensors [79]. The direct connection of imidazolium moieties to rod-shaped aromatic mesogens such as biphenyls can lead to the development of a new family of thermotropic smectic ionic liquid crystals 10 [80]. They were synthesized by a copper-catalyzed Ullman-type coupling of aryl halides and imidazoles followed by quarterization with alkyl halides and subsequent anion exchange reactions. For example, compound 10 containing BF4− shows a smectic A phase from 8 °C to 183 °C as well as a glassy phase below 8 °C by suppression of crystallization. The clearing points of 10 with different anions decrease in the order of I − > BF4− > ClO−4 > PF6−. Thermotropic columnar liquid crystalline phases are formed by selforganization of fan-shaped imidazolium molecules 11a(n/X−) and 11b(12/X−), where n is the alkyl chain length and X− indicates the anion species (Figure 26.2) [22–24]. The salts form hexagonal columnar liquid crystalline phases, except for 11a(8/X−) and 11a(12/X−) containing CF3SO3− and (CF3SO2)2N−, as well as 11b(12/(CF3SO2)2N−), which melt directly to isotropic ionic liquids. Room-temperature columnar liquid crystalline phases are observed only for 11a( n/BF4− ) and 11a( n/PF6− ) with octyl and dodecyl chains and for 11b(12 /BF4− ). The melting points tend to increase as the chain length increases.

Figure 26.2. Molecular structure of fan-shaped imidazolium salts 11a(n/X−) and 11b(12/ X−).

c26.indd 378

1/20/2011 10:28:56 AM

LIQUID CRYSTALLINE IONIC LIQUIDS

379

(a)

(b)

Figure 26.3. Polarizing optical microscopic images and schematic illustrations of the oriented and self-assembled structures of 11a(12 /BF4− ) in the hexagonal columnar state: (a) Before shearing; (b) after shearing the material along the direction perpendicular to the Au electrodes. Directions of A: analyzer; P: polarizer; S: shearing. (Reproduced with permission from J. Am. Chem. Soc., 126, 994–995 (2004). Copyright 2004 American Chemical Society).

The clearing points show a decreasing trend with an increase in the size of anions. The bulky anions may hinder their columnar packing. In these structures, the imidazolium part forms a one-dimensional ionic path inside the column. These columnar materials are macroscopically aligned by shearing on the glass substrate. Figure 26.3 shows polarizing microscope images of compound 11a(12 /BF4− ) before and after shearing. The monodomain of the columnar phase is formed between Au electrodes after the polydomain material is sheared in the sandwiched glasses without rubbing treatment. One-dimensional ionic conductivities were measured by the cell with comb-shaped gold electrodes, which is described in a succeeding section. Recently, thermotropic bicontinuous cubic liquid crystalline phases have been induced by self-organization of fan-shaped ammonium salts 12a,b (Figure 26.4 and 26.5) [27]. Compound 12a exhibits a bicontinuous cubic phase from 42 °C to 82 °C, whereas compound 12b shows a bicontinuous cubic phase from 32 °C to 49 °C as well as a hexagonal columnar phase from 49 °C to 126 °C. In

c26.indd 379

1/20/2011 10:28:56 AM

380

LIQUID CRYSTALLINE IONIC LIQUIDS

Figure 26.4. Molecular structure of fan-shaped ammonium salts 12a,b.

Figure 26.5. Bicontinuous cubic liquid crystalline structure forming three-dimensionally interconnected nano-ion channels.

these bicontinuous cubic structures, ammonium moiety forms threedimensionally interconnected nano-ionic channels (Figure 26.5). These materials have great potential as a new type of alignment-free low-dimensional ion conductors because of the formation of three-dimensional ion channel networks. Oriented structures of self-organized ionic liquids can be preserved in solid films [20]. Self-standing nanostructured two-dimensional polymer films were prepared by in situ photopolymerization of ionic liquid crystalline monomer 13 that forms homeotropic monodomains of the smectic A phase on a glass plate (Figure 26.6). The film of 14 has a macroscopically oriented layered nanostructure as presented in Figure 26.7.

26.2 SELF-ASSEMBLY OF IONIC LIQUIDS WITH LIQUID CRYSTAL MOLECULES Simple ionic liquids can be self-organized by mixing with liquid crystalline molecules that have functional groups interacting with ionic liquids [18,21,26]. Mesogenic compounds 15, 16, and 17 were partially compatible with ionic liquids. The mixtures of 1-ethyl-3-methylimidazolium tetrafluoroborate 18 and

c26.indd 380

1/20/2011 10:28:56 AM

SELF-ASSEMBLY OF IONIC LIQUIDS WITH LIQUID CRYSTAL MOLECULES

381

Figure 26.6. Molecular structures of ionic liquid crystalline monomer 13 and polymer 14.

Figure 26.7. SEM image and schematic illustration of a nanostructured ion-conductive flexible film obtained by in situ photopolymerization of ionic liquid crystalline monomer 13 in the smectic A phase.

c26.indd 381

1/20/2011 10:28:56 AM

382

LIQUID CRYSTALLINE IONIC LIQUIDS

Figure 26.8. Molecular structures of hydroxy-terminated mesogenic compounds 15–17, ionic liquid 18, and mesogenic compound 19 without functional groups.

compounds 15, 16, or 17 form stable layered assemblies (Figures 26.8 and 26.9). In this case, the intermolecular interactions between the hydroxy groups of mesogenic compounds and the ionic liquid enhance the miscibility and lead to the formation of microphase-segregated layered structures at the nanometer scale, as illustrated in Figure 26.9. The phase transition behavior for the mixture of 15 and 18 is shown as a function of the mole fraction of 18 in Figure 26.10 [18]. Compounds 15 and 18 are miscible up to the mole fraction of 0.7 for 18 in the mixture. These mixtures exhibit smectic phases, whereas compound 15 alone gives a columnar phase (Col) from 79 °C to 210 °C. For example, an equimolar mixture of 15 and 18 (15/18) shows a glass transition at 14 °C and subsequent smectic phases until 198 °C on heating. In addition, the mixtures of 16 and 18 form thermally stable layered structures. The layer spacing of the smectic A phase for the equimolar mixture of 16 and 18 (16/18) at 100 °C is 5.0 nm, whereas the spacing for 16

c26.indd 382

1/20/2011 10:28:56 AM

SELF-ASSEMBLY OF IONIC LIQUIDS WITH LIQUID CRYSTAL MOLECULES

383

Figure 26.9. Schematic illustration of the microphase-segregated layered structure for the mixtures of dihydroxy-terminated mesogenic compounds 15 and 17 with ionic liquid 18.

250 Iso

Iso

Temperature / °C

200 Col

150

SA SA

100 SB

SB

50 SX

Cr

Cr

SX

0 G –50

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Mole fraction of 18 in the mixture of 15 and 18

Figure 26.10. Phase transition behavior for a mixture of 15 and 18 on the heating runs. Col: columnar; Cr: crystalline; G: glassy; Iso: isotropic; SA: smectic A; SB: smectic B; SX: unidentified smectic.

c26.indd 383

1/20/2011 10:28:56 AM

384

LIQUID CRYSTALLINE IONIC LIQUIDS

Figure 26.11. Molecular structures of dihydroxy-functionalized mesogenic compound 20 and imidazolium salt 21.

alone is 4.6 nm at the same temperature. The incorporation of the ionic liquid layers between the mesogenic layers may extend the layer spacing for the mixture. In contrast, phase separation in the macroscopic scale is observed for the mixture of 18 and 19 without hydroxy groups because of the lack of intermolecular interactions. Columnar liquid crystalline assemblies are formed for the mixtures of a fan-shaped mesogenic molecule containing dihydroxy groups 20 and imidazolium salt 21 (Figure 26.11) [26]. The ionic liquid is incorporated into the inner part of columnar structures. Compound 20 exhibits a hexagonal columnar (Colh) phase from 7 °C to 27 °C on heating. The phase transition behavior for the mixture of 20 and 21 is shown as a function of the mole fraction of 21 in Figure 26.12. Compounds 20 and 21 are miscible up to the mole fraction of 0.5 for 21 in the mixture. The mixing of 20 and 21 leads to the thermal stabilization of the Colh phase compared with the behavior of 20 itself. The equimolar mixture of 20 and 21 shows a Colh phase from −4 °C to 63 °C. When the mole fraction of 21 exceeds 0.5, a macroscopic phase separation occurs to form two phases comprising the isotropic liquid and the columnar liquid crystalline phases.

26.3 ANISOTROPIC IONIC CONDUCTIVITIES OF IONIC LIQUID CRYSTALS Liquid crystalline ionic liquids can be used as low-dimensional ion conductors [18–27]. For this purpose, the formation of oriented monodomains in the macroscopic scale is important because the boundary in the randomly oriented polydomains disturbs high and anisotropic ion conduction. For the smectic liquid crystalline structures of 15 and 18 (Figure 26.8), alternating ionic liquid layers (conducting parts) and mesogenic layers

c26.indd 384

1/20/2011 10:28:57 AM

ANISOTROPIC IONIC CONDUCTIVITIES OF IONIC LIQUID CRYSTALS

385

150

Iso Temperature / °C

100

50 Colh

0 Cr –50

0

0.1

0.2

0.3

0.4

0.5

Mole fraction of 21 in the mixture of 20 and 21

Figure 26.12. Phase transition behavior for a mixture of 20 and 21 on the heating runs. Colh: hexagonal columnar; Cr: crystalline; Iso: isotropic.

(insulating parts) microphase-segregate at the nanometer scale. Successful control of the orientation of the layered nanostructure on substrates leads to the preparation of two-dimensionally ion-conductive materials. The layered assemblies spontaneously form a homeotropic alignment on glass and indium tin oxide (ITO) surfaces. The homeotropic orientation of the materials to the substrates is confirmed by using a polarizing microscope. Figure 26.13a shows orthoscopic and conoscopic images of the sample oriented homeotropically in the smectic A phase. No birefringence is observed under the crossed Nicols condition. However, focal conic fan texture is observed when the sample is randomly aligned in the smectic A phase (Figure 26.13b). Anisotropic ionic conductivities were measured by an alternating current impedance method [81,82] for the samples forming oriented monodomains in two types of cells shown in Figure 26.14. The comb-shaped gold electrodes with a thickness of 0.8 μm on a glass plate are used for the measurement of ionic conductivity along with the direction parallel (σ||) to the layer (Figure 26.14a). A pair of ITO electrodes fixed by a Teflon spacer of 30 μm can be useful for the measurement of ionic conductivity along with the direction perpendicular (σ⊥) to the layer (Figure 26.14b). Figure 26.15 shows anisotropic ionic conductivities of the equimolar assemblies (15/18, 16/18) of mesogenic compounds 15 and 16 with ionic liquid 18 as a function of temperature. The ionic conductivities parallel to the smectic layer (σ||) are higher than those perpendicular to the layer (σ⊥). For 15/18, the highest σ|| value of 4.1 × 10−3 S cm−1

c26.indd 385

1/20/2011 10:28:57 AM

386

LIQUID CRYSTALLINE IONIC LIQUIDS

(a)

(b)

Figure 26.13. Polarizing optical micrographs of the equimolar mixture of 16 and 18 in the smectic A liquid crystalline state: (a) homeotropically aligned monodomain is formed on a glass substrate when the sample is cooled slowly from the isotropic state at the rate of 3 °C min−1. Inset is a conoscopic image; (b) unaligned polydomains are formed on a glass substrate when the sample is cooled from the isotropic state at the rate of 20 °C min−1.

is observed in the smectic A phase at 192 °C. The magnitude of anisotropy (σ||/σ⊥) at 192 °C is 2.9 × 10. The magnitude of maximal anisotropy is about 3.1 × 103 at 73 °C in the smectic B phase. In the smectic B phase, hexagonally ordered packing of the rigid-rod mesogen effectively may disturb the ion conduction perpendicular to the layer. The liquid crystal assembly of 16/18 shows the highest σ|| value of 2.8 × 10−3 S cm−1, and the σ||/σ⊥ value of 8.2 × 10 is obtained in the smectic A phase at 132 °C. For the anisotropic self-assembled materials containing ionic liquids, the discontinuous changes of ionic conductivities are observed at the phase transition temperatures. In particular, abrupt changes are observed at the isotropic (disorder)-liquid crystalline (order) transitions. To obtain information on the mobility of ionic liquids in the layered nanostructures, the activation energies in the smectic A phases were estimated from the slope of the ionic conductivities (σ||) (Figure 26.15). The energy required for 15/18 is about twice greater than that for 16/18. The ionic conductivities

c26.indd 386

1/20/2011 10:28:57 AM

ANISOTROPIC IONIC CONDUCTIVITIES OF IONIC LIQUID CRYSTALS

387

Figure 26.14. Schematic illustration of the cells for the anisotropic ion-conductive measurements. The samples forming homeotropically aligned monodomains in the smectic liquid crystalline states are filled between electrodes.

along the direction parallel to the smectic layer of 16/18 are higher than those of 15/18, which suggests that the conductivities depend on the number of hydroxy groups of the mesogenic molecules in the mixtures. It was found that the increase of the intermolecular interactions between the hydroxy groups and ionic liquids decreases the mobility. No anisotropy of the ionic conductivity is observed when the assemblies form isotropic states. The formation of the smectic layered structures through the microphase segregation at the nanometer level between ionic liquid and mesogenic compounds results in the spontaneous formation of anisotropic ion-conductive pathways with longrange order. Anisotropic ionic conductivities also were observed for nanostructured two-dimensional films consisting of polymer 14 (Figure 26.6) [20]. The conductivities parallel to the layer are 3.2 × 10−2 S cm−1 in the smectic A phase at 209 °C and 5.9 × 10−6 S cm−1 in the solid state at 50 °C. One-dimensional ion conduction is achieved for columnar liquid crystalline ionic liquids 11a(n/X−) [22–24]. In the macroscopically ordered states of these columnar materials, ionic conductivities parallel to the columnar axis (σ||) are higher than those perpendicular to the axis (σ⊥). No anisotropy of conductivities is observed when the materials form isotropic liquid phases. For example,

c26.indd 387

1/20/2011 10:28:57 AM

388

LIQUID CRYSTALLINE IONIC LIQUIDS Temperature / °C 200

Ionic Conductivity σ / S cm–1

10–2

10

–3

150 Iso N

50

100 SA

Cr

Iso

10–4

10–5

SA

10–6

10–7 10

SB

SX

–8

1.8

2.0

2.2

2.4

2.6

1000/T / K

2.8

3.0

3.2

–1

Figure 26.15. Anisotropic ionic conductivities of the complexes comprising equimolar amounts of mesogenic compounds 15, 16, and ionic liquid 18 as a function of temperature: (•) parallel and (■) perpendicular to the smectic layer for the material based on 15; (○) parallel and (□) perpendicular to the smectic layer for the material based on 16.

compound 11a(8 /BF4− ) shows the conductivities of 3.1 × 10−5 S cm−1 (σ||), 7.5 × 10−7 S cm−1 (σ⊥), and anisotropy (σ||/σ⊥) of 41 at 100 °C. These materials function as self-organized electrolytes. They dissolve a variety of ionic species such as lithium salts. Compound 11a(12 /BF4− ) containing LiBF4 (molar ratio of LiBF4 to 11a(12 /BF4− ): 0.25) exhibits the conductivities of 7.5 × 10−5 S cm−1 (σ||) and 8.0 × 10−7 S cm−1 (σ⊥) at 100 °C. The value of σ||/σ⊥ at 100 °C is 94. It should be noted that by adding salts only conductivities of σ|| are enhanced. The ionic conductivities depend on the nature of anions. The σ|| values for 11a(16/X−) in the Colh phases at 110 °C increase in the order of PF6− > BF4− > CF3SO−3 . The anisotropy of ionic conductivities decreases as the size of the anions increases. For example, compounds 11a(16 /BF4−), 11a(16 /PF6− ), and 11a(16 /CF3SO−3 ) show the anisotropy of 110, 81, and 15 at 110 °C, respectively. For these columnar liquid crystalline ionic liquids, the cationic imidazolium moiety is considered to be less mobile than the anions in the ionic channels because the cationic moiety is covalently connected to the alkoxyphenyl groups forming columnar structures. The observed ionic conductivities mostly may be caused by the transportation of the anions inside the columns because the conductivities strongly depend on the anions.

c26.indd 388

1/20/2011 10:28:57 AM

REFERENCES

26.4

389

RELATED MATERIALS

Metallomesogens are materials related to liquid crystalline ionic liquids because some of them have ionic structures [83,84]. Gels based on ionic liquids are also related materials. The physical gelation of ionic liquids is accompanied by the formation of bilayer structures of glycolipids and a cationic amphiphile in ionic liquids [85,86]. Ionic liquids also are physically gelled by the low molecular weight organogelators based on cholesterol and amino acids to tune the fluidity of ionic liquids for application in electrochemical devices [87,88]. Moreover, gelation of ionic liquids is induced by physical cross-linking of single-walled carbon nanotubes, which is mediated by the local molecular ordering of ionic liquids around the nanotubes [89]. Chemical gels of ionic liquids also have been prepared [90–92]. 26.5

SUMMARY

In this chapter, we have described the mesomorphic behavior and ionic conductivities of ionic liquid-based liquid crystalline materials. These ion-active anisotropic materials have great potential for applications not only as electrolytes that anisotropically transport ions at the nanometer scale but also as ordered solvents for reactions. Ionic liquid crystals also have been studied for uses as diverse as nonliner optoelectronic materials [75,76], photoluminescent materials [93–95], structure-directing reagents for mesoporous materials [96,97], and ordered solvents for organic reactions [61,98]. Approaches to selforganization of ionic liquids may open a new avenue in the field of material science and supramolecular chemistry. REFERENCES 1. D. Demus, J. W. Goodby, G. W. Gray, H.-W. Spiess, V. Vill, eds., Handbook of Liquid Crystals, Wiley-VCH, 1998. 2. D. Adam, P. Schuhmacher, J. Simmerer, L. Häussking, K. Siemensmeyer, K. H. Etzbach, H. Ringsdorf, D. Haarer, Nature, 371, 141 (1994). 3. N. Boden, R. J. Bushby, J. Clements, B. Movaghar, J. Mater. Chem., 9, 2081 (1999). 4. R. J. Bushby, O. R. Lozman, Curr. Opin. Solid State Mater. Sci., 6, 569 (2002). 5. M. O’Neill, S. M. Kelly, Adv. Mater., 15, 1135 (2003). 6. T. Kato, Science, 295, 2414 (2002). 7. T. Kato, N. Mizoshita, K. Kishimoto, Angew. Chem. Int. Ed., 45, 38 (2006). 8. T. Kato, T. Yasuda, Y. Kamikawa, M. Yoshio, Chem. Commun., 729 (2009). 9. M. Funahashi, H. Shimura, M. Yoshio, T. Kato, Struct. Bond., 128, 151 (2008). 10. T. Kato, K. Tanabe, Chem. Lett., 38, 634 (2009). 11. T. Ohtake, K. Ito, N. Nishina, H. Kihara, H. Ohno, T. Kato, Polymer J., 31, 1155 (1999).

c26.indd 389

1/20/2011 10:28:57 AM

390

LIQUID CRYSTALLINE IONIC LIQUIDS

12. T. Ohtake, M. Ogasawara, K. Ito-Akita, N. Nishina, S. Ujiie, H. Ohno, T. Kato, Chem. Mater., 12, 782 (2000). 13. T. Ohtake, Y. Takamitsu, K. Ito-Akita, K. Kanie, M. Yoshizawa, T. Mukai, H. Ohno, T. Kato, Macromolecules, 33, 8109 (2000). 14. K. Hoshino, K. Kanie, T. Ohtake, T. Mukai, M. Yoshizawa, S. Ujiie, H. Ohno, T. Kato, Macromol. Chem. Phys., 203, 1547 (2002). 15. K. Kishimoto, M. Yoshio, T. Mukai, M. Yoshizawa, H. Ohno, T. Kato, J. Am. Chem. Soc., 125, 3196 (2003). 16. K. Kishimoto, T. Suzawa, T. Yokota, T. Mukai, H. Ohno, T. Kato, J. Am. Chem. Soc., 127, 15618 (2005). 17. Y. Iinuma, K. Kishimoto, Y. Sagara, M. Yoshio, T. Mukai, I. Kobayashi, H. Ohno, T. Kato, Macromolecules, 40, 4874 (2007). 18. M. Yoshio, T. Mukai, K. Kanie, M. Yoshizawa, H. Ohno, T. Kato, Adv. Mater., 14, 351 (2002). 19. M. Yoshio, T. Mukai, K. Kanie, M. Yoshizawa, H. Ohno, T. Kato, Chem. Lett., 320 (2002). 20. K. Hoshino, M. Yoshio, T. Mukai, K. Kishimoto, H. Ohno, T. Kato, J. Polym. Sci., Part A: Polym. Chem., 41, 3486 (2003). 21. M. Yoshio, T. Mukai, M. Yoshizawa, H. Ohno, T. Kato, Mol. Cryst. Liq. Cryst., 413, 2235 (2004). 22. M. Yoshio, T. Mukai, H. Ohno, T. Kato, J. Am. Chem. Soc., 126, 994 (2004). 23. M. Yoshio, T. Mukai, H. Ohno, T. Kato, ACS Symp. Ser., 975, 161 (2007). 24. M. Yoshio, T. Ichikawa, H. Shimura, T. Kagata, A. Hamasaki, T. Mukai, H. Ohno, T. Kato, Bull. Chem. Soc. Jpn., 80, 1836 (2007). 25. M. Yoshio, T. Kagata, K. Hoshino, T. Mukai, H. Ohno, T. Kato, J. Am. Chem. Soc., 128, 5570 (2006). 26. H. Shimura, M. Yoshio, K. Hoshino, T. Mukai, H. Ohno, T. Kato, J. Am. Chem. Soc., 130, 1759 (2008). 27. T. Ichikawa, M. Yoshio, A. Hamasaki, T. Mukai, H. Ohno, T. Kato, J. Am. Chem. Soc., 129, 10662 (2007). 28. F. B. Dias, S. V. Batty, A. Gupta, G. Ungar, J. P. Voss, P. V. Wright, Electrochim. Acta, 43, 1217 (1998). 29. H. V. St. A. Hubbard, S. A. Sills, G. R. Davies, J. E. MacIntyre, I. M. Ward, Electrochim. Acta, 43, 1239 (1998). 30. C. T. Imrie, M. D. Ingram, G. S. McHattie, Adv. Mater., 11, 832 (1999). 31. T. Kato, Struct. Bond., 96, 95 (2000). 32. T. Kato, N. Mizoshita, Curr. Opin. Solid State Mater. Sci., 6, 579 (2002). 33. C. Tschierske, J. Mater. Chem., 8, 1485 (1998). 34. C. Tschierske, J. Mater. Chem., 11, 2647 (2001). 35. V. Percec, G. Johansson, J. Heck, G. Ungar, S. V. Batty, J. Chem. Soc., Perkin Trans. 1, 1411 (1993). 36. V. Percec, J. A. Heck, D. Tomazos, G. Ungar, J. Chem. Soc., Perkin Trans. 2, 2381 (1993). 37. V. Percec, G. Johansson, R. Rodenhouse, Macromolecules, 25, 2563 (1992).

c26.indd 390

1/20/2011 10:28:57 AM

REFERENCES

38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. 61. 62. 63. 64. 65. 66. 67.

c26.indd 391

391

C. F. van Nostrum, R. J. M. Nolte, Chem. Commun., 2385 (1996). J.-M. Lehn, J. Malthête, A.-M. Levelut, J. Chem. Soc., Chem. Commun., 1794 (1985). H. Tokuhisa, M. Yokoyama, K. Kimura, J. Mater. Chem., 8, 889 (1998). T. Welton, Chem. Rev., 99, 2071 (1999). P. Wasserscheid, W. Keim, Angew. Chem. Int. Ed., 39, 3772 (2000). R. D. Rogers, K. R. Seddon, Science, 302, 792 (2003). P. Bonhôte, A.-P. Dias, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). H. Ohno, M. Yoshizawa, Solid State Ionics, 154–155, 303 (2002). H. Ohno, Bull. Chem. Soc. Jpn., 79, 1665 (2006). M. Armand, F. Endres, D. R. MacFarlane, H. Ohno, B. Scrosati, Nat. Mater., 8, 621 (2009). K. Binnemans, Chem. Rev., 105, 4148 (2005). C. G. Bazuin, D. Guillon, A. Skoulios, J.-F. Nicoud, Liq. Cryst., 1, 181 (1986). S. Ujiie, K. Iimura, Macromolecules, 25, 3174 (1992). V. Hessel, H. Ringsdorf, Makromol. Rapid Comm., 14, 707 (1993). Y. Kosaka, T. Kato, T. Uryu, Liq. Cryst., 18, 693 (1995). C. Chovino, Y. Frere, D. Guillon, P. Gramain, J. Polym. Sci., Part A: Polym. Chem., 35, 2569 (1997). Y. Haramoto, Y. Akiyama, R. Segawa, S. Ujiie, M. Nanasawa, J. Mater. Chem., 8, 275 (1998). C. J. Bowlas, D. W. Bruce, K. R. Seddon, Chem. Commun., 1625 (1996). C. M. Gordon, J. D. Holbrey, A. R. Kennedy, K. R. Seddon, J. Mater. Chem., 8, 2627 (1998). J. D. Holbrey, K. R. Seddon, J. Chem. Soc., Dalton Trans., 2133 (1999). A. E. Bradley, C. Hardacre, J. D. Holbrey, S. Johnston, S. E. J. McMath, M. Nieuwenhuyzen, Chem. Mater., 14, 629 (2002). C. Hardacre, J. D. Holbrey, S. E. J. McMath, M. Nieuwenhuyzen, ACS Symp. Ser., 818, 400 (2002). J. De Roche, C. M. Gordon, C. T. Imrie, M. D. Ingram, A. R. Kennedy, F. Lo Celso, A. Triolo, Chem. Mater., 15, 3089 (2003). C. K. Lee, H. W. Huang, I. J. B. Lin, Chem. Commun., 1911 (2000). T. L. Merrigan, E. D. Bates, S. C. Dorman, J. H. Davis Jr., Chem. Commun., 2051 (2000). K.-M. Lee, Y.-T. Lee, I. J. B. Lin, J. Mater. Chem., 13, 1079 (2003). K. M. Lee, C. K. Lee, I. J. B. Lin, Chem. Commun., 899 (1997). A. Downard, M. J. Earle, C. Hardacre, S. E. J. McMath, M. Nieuwenhuyzen, S. J. Teat, Chem. Mater., 16, 43 (2004). M. A. Firestone, J. A. Dzielawa, P. Zapol, L. A. Curtiss, S. Seifert, M. L. Dietz, Langmuir, 18, 7258 (2002). M. L. Dietz, J. A. Dzielawa, M. P. Jensen, M. A. Firestone, ACS Symp. Ser., 856, 526 (2003).

1/20/2011 10:28:57 AM

392

LIQUID CRYSTALLINE IONIC LIQUIDS

68. C. Hardacre, J. D. Holbrey, P. B. McCormac, S. E. J. McMath, M. Nieuwenhuyzen, K. R. Seddon, J. Mater. Chem., 11, 346 (2001). 69. F. Neve, A. Crispini, S. Armentano, O. Francescangeli, Chem. Mater., 10, 1904 (1998). 70. F. Neve, A. Crispini, O. Francescangeli, Inorg. Chem., 39, 1187 (2000). 71. F. Neve, O. Francescangeli, A. Crispini, J. Charmant, Chem. Mater., 13, 2032 (2001). 72. P. K. Bhowmik, H. Han, J. J. Cebe, R. A. Burchett, Liq. Cryst., 30, 1433 (2003). 73. D. J. Abdallah, A. Robertson, H.-F. Hsu, R. G. Weiss, J. Am. Chem. Soc., 122, 3053 (2000). 74. H. Chen, D. C. Kwait, Z. S. Gönen, B. T. Weslowski, D. J. Abdallah, R. G. Weiss, Chem. Mater., 14, 4063 (2002). 75. A. Kanazawa, T. Ikeda, J. Abe, Angew. Chem. Int. Ed., 39, 612 (2000). 76. A. Kanazawa, T. Ikeda, J. Abe, J. Am. Chem. Soc., 123, 1748 (2001). 77. F. Mathevet, P. Masson, J.-F. Nicoud, A. Skoulios, Chem. Eur. J., 8, 2248 (2002). 78. C. P. Roll, B. Donnio, D. Guillon, W. Weigand, J. Mater. Chem., 13, 1883 (2003). 79. D. Wang, J. Wang, D. Moses, G. C. Bazan, A. J. Heeger, Langmuir, 17, 1262 (2001). 80. P. H. J. Kouwer, T. M. Swager, J. Am. Chem. Soc., 129, 14042 (2007). 81. J. R. Macdonald, Impedance Spectroscopy Emphasizing Solid Materials and Systems, Wiley, 1987. 82. Y. Tominaga, H. Ohno, Electrochim. Acta, 45, 3081 (2000). 83. D. W. Bruce, Acc. Chem. Res., 33, 831 (2000). 84. R. Giménez, D. P. Lydon, J. L. Serrano, Curr. Opin. Solid State Mater. Sci., 6, 527 (2002). 85. N. Kimizuka, T. Nakashima, Langmuir, 17, 6759 (2001). 86. T. Nakashima, N. Kimizuka, Chem. Lett., 1018 (2002). 87. A. Ikeda, K. Sonoda, M. Ayabe, S. Tamaru, T. Nakashima, N. Kimizuka, S. Shinkai, Chem. Lett., 1154 (2001). 88. W. Kubo, T. Kitamura, K. Hanabusa, Y. Wada, S. Yanagida, Chem. Commun., 374 (2002). 89. T. Fukushima, A. Kosaka, Y. Ishimura, T. Yamamoto, T. Takigawa, N. Ishii, T. Aida, Science, 300, 2072 (2003). 90. R. T. Carlin, J. Fuller, Chem. Commun., 1345 (1997). 91. A. Noda, M. Watanabe, Electrochim. Acta, 45, 1265 (2000). 92. H. Ohno, M. Yoshizawa, W. Ogihara, Electrochim. Acta, 48, 2079 (2003). 93. D. Haristoy, D. Tsiourvas, Chem. Mater., 15, 2079 (2003). 94. K. Tanabe, T. Yasuda, T. Kato, Chem. Lett., 37, 1208 (2008). 95. A. J. Boydston, C. S. Pecinovsky, S. T. Chao, C. W. Bielawski, J. Am. Chem. Soc., 129, 14550 (2007). 96. Y. Zhou, M. Antonietti, Adv. Mater., 15, 1452 (2003). 97. Y. Zhou, M. Antonietti, Chem. Mater., 16, 544 (2004). 98. R. G. Weiss, Tetrahedron Lett., 44, 3413 (1988).

c26.indd 392

1/20/2011 10:28:57 AM

PART VI GEL-TYPE POLYMER ELECTROLYTES

c27.indd 393

1/20/2011 10:28:58 AM

27 IONIC LIQUID GELS Kenji Hanabusa

Recently, ionic liquids have received much attention for materials in greensustainable chemistry [1–3]. Ionic liquids comprising ions only have good properties such as extremely low volatility, high thermal stability, a wide temperature range for liquid phase, nonflammability, high chemical stability, high ionic conductivity, and a wide electrochemical window. Here, the preparation of ionic liquid gels and the properties of formed gels are described. 27.1 GELATION OF IONIC LIQUIDS BY LOW MOLECULAR WEIGHT GELATORS Low molecular weight compounds capable of hardening organic liquids, which are called “gelators,” are of special interest for not only academic objects but also for potential applications [4,5]. Gelators have unique characteristics of both good solubility upon heating and the inducement of smooth gelation of organic fluids at a low concentration. For example, when the hot solution containing gelators is cooled to room temperature, the gel is easily formed in the course of the cooling process. The formed gels always exhibit a thermally reversible sol-to-gel phase transition, because the driving forces of physical gelation by gelator are cooperating noncovalent interactions of gelator molecules, such as hydrogen bonding, hydrophobic interaction, π–π interaction, and electrostatic interaction. With a low melting point in mind, nine ionic liquids shown in Figure 27.1 were tested for gelation. The six kinds of imidazolium salts, the two kinds of ammonium salts, and a pyridinium salt are tested. The imidazolium and Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

395

c27.indd 395

1/20/2011 10:28:58 AM

396

IONIC LIQUID GELS BF4−

N + N

N + N

C2mim / BF4

N + N

C4mim / BF4

O O O O S S F3C N CH3 −

N + N

BF4−

C4mim / TFSI

C4mim / PF6

BF4−

PF6−

N +

N + N C6mim / BF4

BF4− N

O

O

BF4−

O

H O

HN

NH

O

N+ B

H HN

O

T

C4Py / BF4

O

C6Py / PF6

O O O O O S S N+ F3C N CF3 −

+

PF6−

O NH

O H

H 1

2

Figure 27.1. Structures of ionic liquids and gelators.

pyridinium salts were prepared by ordinary methods [6], and two ammonium salts were supplied by Nisshinbo Inc. (Japan). From a screening test of many gelators, it was found that cyclo(l-β-3,7-dimethyloctylasparaginyl-lphenylalanyl) (1) and cyclo(l-β-2-ethylhexylasparaginyl-L-phenylalanyl) (2), prepared from artificial sweetener “Aspartame,” were excellent gelators for the ionic liquids (Figure 27.1) [7]. Other gelators [8–12] could not gel up ionic liquids because of the low solubility. The results of the gelation test of ionic liquids by gelators 1 and 2 are summarized in Table 27.1, where values indicate the minimum gel concentrations (MGC) necessary to harden ionic liquids. The gelators gelled up all ionic liquids and formed translucent or transparent gels by adding less than 1 wt% (gelator/ionic liquid). For example, 3 g of gelator 1 can form 1 L of translucent gel of C2mim/BF4.

c27.indd 396

1/20/2011 10:28:58 AM

397

GELATION OF IONIC LIQUIDS

TABLE 27.1. Results of Gelation Test of Ionic Liquids by Gelators 1 and 2 at 25 °C Ionic Liquid

1

2

C2mim/BF4 C4mim/BF4 C4mim/PF6 C4mim/TFSI C6mim/BF4 C6mim/PF6 C4Py/BF4 T B

G-Tl (3)a G-Tl (5) G-Tl (5) G-Tl (8) G-Tp (8) G-Tl (8) G-Tp (3) G-Tl (9) G-Tl (5)

G-Tl (5) G-Tl (7) G-Tl G-Tl (12) G-Tp (12) G-Tl G-Tp (5) G-Tl (13) G-Tl (10)

a

G-Tp: transparent gel, G-Tl: translucent gel. Note: Values mean minimum gel concentrations, whose unit is g/L (gelator/ionic liquid).

3000

Gel strength (g/cm2)

2500

2000

E G

G E G G E E

G E

1500

E G

1000

G E

500

E G

G E

E G

0 0

20

40

60

80

100

120

Concentration of 1 (g/L)

Figure 27.2. Gel strength of ionic liquids gels formed by gelator 1: (□) C4mim/BF4 and (○) C4Py/BF4.

Gel strength is an important factor in the application of gels. Gel strength of C4mim/BF4 and C4Py/BF4 gels are plotted against the concentration of 1 (Figure 27.2). The gel strength is evaluated as the power necessary to sink a cylinder bar (10 mm in diameter) 4-mm deep in the gels. It is clear that gel strength increases with an increasing concentration of the added gelator. The

c27.indd 397

1/20/2011 10:28:59 AM

398

IONIC LIQUID GELS

Concentration of gelator (g/L)

80 70 60 50

Gel phase

40 30 20

Sol phase

10 0 40

60

80

100

120

Sol-to-gel phase transition temperature

140 (EC)

Figure 27.3. Sol-to-gel phase transition diagram: (◇) gelator 1 in C4mim/BF4 and (○) gelator 1 in C4Py/BF4.

addition of 60 g of gelator 1 to 1 L of C4Py/BF4 resulted in a transparent gel, whose gel strength was around 1500 g cm−2. The formed gels showed a thermally reversible sol-to-gel transition because of the noncovalent interactions. The sol-to-gel phase diagrams for C4mim/BF4 and C4Py/BF4 are shown in Figure 27.3, where the MGC are plotted against the temperature of ionic liquids. The area above the plots is the gel phase, and the area below is the sol phase. The transition temperature from gel to sol increases with the increasing concentration of gelator. For instance, when the gel is prepared at the concentration of 70 g L−1, the phase transition temperature of the gel-to-sol transition is more than 140 °C. This means that the thermally stable gels can be obtained using gelators. 27.2

CONDUCTIVITY OF IONIC LIQUID GELS

Solid electrolytes, which are electrically conductive solids with ionic carriers, have received special attention because of their potential use in solid-state batteries, fuel cells, energy storage, and chemical sensors [13,14]. The gelation of ionic liquids by gelator is a conventional method for making gel electrolytes; therefore, it is expected to play a role in the field of solid electrolytes. Figure 27.4 shows the ionic conductivity of the various ionic liquids gels formed by 1. In general, trifluoromethanesulfonimide and tetrafluoroborate of 1-butyl-3-methylimidazolium (C4mim) show a higher ionic conductivity. The plots at the left end give the ionic conductivities of the neat ionic liquids without a gelator. The ionic conductivities of gels are nearly similar to those of the neat ionic liquids. Namely, the ionic conductivity decreased slightly with

c27.indd 398

1/20/2011 10:28:59 AM

399

CONDUCTIVITY OF IONIC LIQUID GELS

4H B

H

Ionic conductivity (mS/cm)

B

H

H

H B

H B

B B

3

H

H

B

B

2 JC E F 1I G

JC E F I

JC

JC

JC

JC

E F I

E F I

E F I

E F I

JC E F I

JC E F I

G

G

G

G

G

G

G

0 0

20

40

60

80

Concentration of gelator (g/L)

Figure 27.4. Dependence of ionic liquid gel conductivity on concentration of gelator 1 at room temperature: (▲) C4mim/TFSI, (■) C4mim/BF4, (•) C4mim/PF6, (△) T, (○) C4Py/BF4, (◆) C6mim/BF4, (×) B, and (□) C6Py/PF6.

the increasing concentration of gelator. The gels of ionic liquids were characterized by high ionic conductivity. For instance, the ionic conductivity of the C4mim/BF4 gel formed by 1 (20 g L−1) was 3.20 mS cm−1. This result indicates that the gelator molecules hardly interfere with the mobility of the anion and cation. When, in practice, ionic liquids are used for electrolytes in a device, the addition of polar solvents is recommended because the addition of polar solvent reduces the viscosity and promotes the dissociation of ionic liquids. The addition of a polar solvent consequently improves conductivity. The gelation of a mixture containing C4mim/BF4 and propylene carbonate (PC) by gelator 1 was studied. In this case, the ionic liquids in PC were thought to be supporting electrolytes. The MGCs of gelator 1 were plotted against the PC ratio, as shown in Figure 27.5. In the figure, the 0% of the PC ratio means 100% of the ionic liquid, and 100% of the PC ratio means pure PC. The gelator was found to gel up mixtures regardless of the PC ratio, and the MGCs were almost proportional to the ratio of PC in the mixture. The ionic conductivity of gels comprising C4mim/BF4, PC, and 1 is shown in Figure 27.6. The plots at the left end show the ionic conductivities of the neat mixture of C4mim/BF4 and PC without the gelator. It is apparent that the ionic conductivities of the mixture increase with the amount of added PC. It

c27.indd 399

1/20/2011 10:28:59 AM

400

IONIC LIQUID GELS

25

F

MGC(g/L)

20

15

F F

10

F

F

5F

0 0

20

40

60

80

100

PC ratio in mixture (vol-%)

Figure 27.5. MGC necessary for gelation of mixtures of C4mim/BF4 and PC. 8

Ionic conductivity (mS/cm)

B

6

J

H F 4

E

B

B

B

B

J

J

J

J

H F

H F

H F

H F

E

E

E

E

E

E

E

2

0 0

20

40

60

80

Concentration of gelator 1 (g/L)

Figure 27.6. Ionic conductivity of C4mim/BF4 gel containing PC: (■) C4mim/BF4: PC = 1 : 1; (•) C4mim/BF4: PC = 2 : 1; (▲) C4mim/BF4: PC = 5 : 1; (◆) C4mim/BF4: PC = 10 : 1; (○) C4mim/BF4 only.

c27.indd 400

1/20/2011 10:28:59 AM

REFERENCES

401

is noteworthy that the ionic conductivities decrease slightly with the increasing concentration of gelator. The slight decrease of ionic conductivity suggests that the gelators can be useful in making ionic liquid gels. 27.3

POLYMER GELS OF IONIC LIQUIDS

Zwitterionic molten salts comprising imidazolium cations containing covalently bound anionic sites were synthesized. The radical polymerization of the zwitterionic imidazolium salts containing the vinyl group resulted in rubberlike polymers [15]. The ionic conductivity of pure zwitterionic-type imidazolium salt polymers was less than 10−9 S cm−2, but when mixed with 100 mol% LiTFSI, these polymers showed good ionic conductivities of around 10−5 S cm−2 at 50 °C, despite their rubberlike properties. However, radical polymerization of common vinyl monomers in situ in ionic liquids [16] produced new polymer-electrolytes called “ion gels.” These ion gels have completely compatible combinations of ionic liquids and the resulting network polymers. The ion gels can softly bend, and they showed a fast ionic conductivity of 10−2 S cm−2 at room temperature. REFERENCES 1. T. Welton, Chem. Rev., 99, 2071 (1999). 2. P. Wasserscheid, T. Welton, Ionic Liquids in Synthesis, Wiley-VCH, 2002. 3. R. D. Rogers, K. R. Seddon, Ionic Liquids—Industrial Applications for Green Chemistry, ASC Symp. Ser., 818 ACS, 2002. 4. P. Terech, R. G. Weiss, Chem. Rev., 97, 3133 (1997). 5. J. H. Esch, B. L. Feringa, Angew. Chem. Int. Ed., 39, 2263 (2000). 6. P. Bonhôte, A. Dias, N. Papagerorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). 7. K. Hanabusa, M. Matsumoto, M. Kimura, A. Kakehi, H. Shirai, J. Colloid Interface Sci., 224, 231 (2002). 8. K. Hanabusa, K. Hiratsuka, M. Kimura, H. Shirai, Chem. Mater., 11, 649 (1999). 9. K. Hanabusa, M. Yamada, M. Kimura, H. Shirai, Angew. Chem. Int. Ed., 35, 1949 (1996). 10. K. Hanabusa, J. Tange, Y. Taguchi, T. Koyama, H. Shirai, Chem. Comm., 1993, 390 (1993). 11. K. Hanabusa, H. Nakayama, M. Kimura, H. Shirai, Chem. Lett., 2000, 1070 (2000). 12. K. Hanabusa, K. Shimura, K. Hirose, M. Kimura, H. Shirai, Chem. Lett., 1996, 885 (1996). 13. P. G. Bruce, Solid State Elecrochemistry, Cambridge University Press, 1995. 14. B. V. Ratnakumar, S. R. Narayanan, Handbook Solid State Batteries Capacitors, World Scientific, 1995. 15. M. Yoshizawa, M. Hirao, K. Ito-Akita, H. Ohno, J. Mater. Chem., 11, 1057 (2001). 16. A. Noda, M. Watanabe, Electochim. Acta, 45, 1265 (2002).

c27.indd 401

1/20/2011 10:28:59 AM

28 ZWITTERIONIC LIQUID/POLYMER GELS Masahiro Yoshizawa-Fujita and Hiroyuki Ohno

Typical polymer gel electrolytes are obtained by mixing a polar solvent, salt, and polymer to form a matrix. Because the polar solvent transports ions, these polymer gel electrolytes show relatively high ionic conductivity despite the film property. In addition, the desirable ion conductive behavior and solid property can be attained easily by appropriately choosing the kind of polymer used as a matrix and the ratio of solvent to polymer. These polymer gel electrolytes have been used as flexible electrolyte films for dry ionics devices. However, the temperature range of this kind of polymer gel electrolyte is not wide enough because the thermal stability of polymer gel electrolytes is dependent on the solvent’s properties. Although there are excellent organic solvents such as propylene carbonate and ethylene carbonate, the upper limit of the available temperature for all these remains at about 200 °C. It is difficult to satisfy each parameter simultaneously, such as low viscosity, high polarity, low melting point, and high boiling point. This has prevented the application of polymer gel electrolytes in many fields. In particular, the volatility of the polar organic solvent is a serious problem (Table 28.1, type A). Ionic liquids (ILs) are being investigated as a potential solution to this problem. ILs have attractive properties such as nonvolatility, a liquid state in a wide temperature range, and capacity to dissolve many compounds [1–3]. Furthermore, ILs with relatively low viscosity show a high ionic conductivity of about 10−2 S cm−1 at room temperature [4,5]. Although the temperature property of most ILs is not satisfactory, studies have been done to overcome

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

403

c28.indd 403

1/20/2011 10:29:00 AM

404

ZWITTERIONIC LIQUID/POLYMER GELS

TABLE 28.1. Comparison of Gel Type Polymer Electrolytes Matrix Component

Type A

Type B

Polymer, Solvent + Salt

Polymer Ionic Liquid

Target ion transport Thermal stability

×

○

×

CH2CF2

CF2CF m

CF3 m : n = 85 : 15 P(VdF-co-HFP)

n

○

N

N

SO3 x

x = 3 ; EIm3S x = 4 ; EIm4S

Figure 28.1. Structures of P(VdF-co-HFP) and ZILs.

the upper limit by choosing a suitable combination of cation and anion. As reported in Chapter 27, many researchers already have succeeded in making polymer gel electrolytes using ILs [6–20]. Taking the excellent properties of ILs into account, it is easy to reach a conclusion that all these problems can be solved by using ILs as the electrolyte solutions. However, the high ionic conductivity of ILs is based on the IL itself. The low transference number of target ions such as the lithium cation and proton remains a drawback, as shown in Table 28.1, type B. To overcome this drawback, zwitterionic liquids (ZILs), which have both a cation and an anion in the same molecule, have started to be investigated (refer to Chapter 21 for details) [21]. Because ZILs do not migrate even under a potential gradient, they can provide the ion conductive pathway to transport only target ions. In other words, ZILs should be suitable as the nonvolatile solvent for polymer gel electrolytes. It is not so difficult to design a suitable structure of the IL for a particular purpose and to prepare the IL. The ease with which this can be done is a big advantage for organic ILs. Therefore, the favored IL is often called the “designer solvent” [22] or even “my solvent” [23] because the matrix prepared by the polymerized IL is designed to overcome the aforementioned problem [24,25]. Such a matrix also is designed to maximize the merits of organic ILs. The structure of a host polymer with ZILs is shown in Figure 28.1 [26]. The copolymers of vinylidene fluoride and hexafluoropropylene (P(VdF-co-HFP)) are used as the host polymer. This copolymer is used to prepare polymer gel electrolytes containing simple ILs because of their good miscibility with this component [17]. A fluorine-containing polymer generally has excellent compatibility with hydrophobic ILs. The following ZILs were used as a solvent: 1-(N-ethylimidazolio)propane-3-sulfonate (EIm3S) and 1-(Nethylimidazolio)butane-4-sulfonate (EIm4S). The thermal stability of P(VdF-

c28.indd 404

1/20/2011 10:29:00 AM

405

ZWITTERIONIC LIQUID/POLYMER GELS 0

100

200

300

400

0

100

200

300

400

P(VdF-co-HFP)

Pure ZIL

ZIL + LiTFSI

Polymer Gel Electrolyte

Temperature / °C

Figure 28.2. Thermal stabilities of P(VdF-co-HFP), EIm3S, and their mixture.

co-HFP), ZILs, and their mixture is summarized in Figure 28.2. The decomposition temperature of P(VdF-co-HFP) is about 400 °C, whereas that of EIm3S and EIm4S is 330 °C. No weight loss was confirmed by means of thermogravimetric differential thermal analysis (TG/DTA) up to each decomposition temperature. Interestingly, the mixture of ZILs and an equimolar amount of LiTFSI remained stable up to 390 °C, which is about 60 °C higher than that of pure ZILs. In simple ILs, their thermal stability deeply depends on the combination of imidazolium cation and counter anion. For instance, when the imidazolium cation is combined with a TFSI anion, the decomposition temperature generally is improved at above 400 °C [27,28]. The thermal stability of ZILs is also a function of the added imidazolium cation of ZILs and the TFSI anion. The thermal stability of the polymer gel electrolyte containing the ZIL/LiTFSI mixture is about 390 °C, which is consistent with that of the ZIL/LiTFSI mixture. The thermal stability of common polymer gel electrolytes is much less. Systems containing the ZIL/LiTFSI mixture are expected to be extremely stable gel electrolytes. Figure 28.3 shows the temperature dependence of the ionic conductivity of these novel polymer gel electrolytes. P(VdF-co-HFP) was mixed with EIm3S/ LiTFSI at a mixing ratio of 1 : 2 by weight. The ionic conductivity of this system is about 10−4 S cm−1 at 200 °C. When the EIm4S/LiTFSI mixture was used, a similar tendency was observed for the Arrhenius plot of the polymer gel electrolyte. Although the ionic conductivity could not be measured in a higher temperature range because of an equipment limitation, the ionic conductivity of these polymer gel electrolytes is guaranteed by the results of the TG/DTA analysis. The thermal stability of this film still needs to be improved in the present system. As shown in Figure 28.3, there is an inflection point in the ionic conductivity at 150 °C. This point agrees with the melting point of

c28.indd 405

1/20/2011 10:29:00 AM

406

ZWITTERIONIC LIQUID/POLYMER GELS

Temperature / °C

Ionic conductivity log(σi / Scm–1)

240

200

160

120

80

40

– 4.0

– 4.5

– 5.0

– 5.5

2.0

2.4

2.8

3.2

1000 T –1 / K –1 Figure 28.3. Temperature dependence of ionic conductivity for the P(VdF-co-HFP)/ equimolar mixture of EIm3S and LiTFSI. The ratio of polymer to ZIL is fixed at 1 : 2 by weight.

P(VdF-co-HFP). It means that, in this case, the host polymer could not keep a solid state above 150 °C. This drawback was overcome by using other network polymers [29]. Several polymer gel electrolytes were prepared by changing the mixing ratio of ZILs to polymer. The polymer gel electrolytes were obtained as translucent films up to 80 wt% of the ZIL/LiTFSI content. When the content of the ZIL/LiTFSI mixture was less than 40 wt%, they were obtained as flexible but white films. Figure 28.4 shows the effect of a ZIL/LiTFSI mixture content on the ionic conductivity for polymer gel electrolytes. The ionic conductivity of these polymer gel electrolytes increased with the increasing ZIL/LiTFSI mixture content. When the ZIL/LiTFSI mixture content was 66 wt%, the ionic conductivity exceeded 10−6 S cm−1 at 50 °C, which is almost equivalent to that of ZIL/LiTFSI simple mixture. We prepared a new polymer gel electrolyte of superior thermal stability by using ZILs. Several properties of the polymer gel electrolytes also will be improved through the synthesis of novel liquid ZILs at room temperature. Unfortunately, most ZILs are solid at room temperature without LiTFSI, so we found it necessary to design new ZILs to lower the melting point of the ZILs. We are expanding the possibility of using liquid ZILs at room temperature. Our research with liquid ZILs will be published elsewhere soon.

c28.indd 406

1/20/2011 10:29:00 AM

407

REFERENCES

Film

Ionic conductivity log(σi / Scm–1)

–5

–6

50 °C

–7 30 °C –8 40 60 80 Zwitterionic liquid / wt%

100

Figure 28.4. Effect of ZIL/LiTFSI content on the ionic conductivity for the polymer gel electrolyte.

REFERENCES 1. P. Wasserscheid, T. Welton, eds., Ionic Liquids in Synthesis, Wiley-VCH, 2003. 2. H. Ohno, ed., Ionic Liquids-The Front and Future of Material Development, in Japanese, CMC, 2003. 3. T. Welton, Chem. Rev., 99, 2071 (1999). 4. P. Bonhôte, A.-P. Dias, M. Armand, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). 5. R. Hagiwara, Y. Ito, J. Fluorine Chem., 105, 221 (2000). 6. (a) A. Noda, M. Watanabe, Electrochim. Acta, 45, 1265 (2000). (b) T. Tsuda, T. Nohira, Y. Nakamori, K. Matsumoto, R. Hagiwara, Y. Ito, Solid State Ionics, 149, 295 (2002). 7. H. Nakagawa, S. Izuchi, K. Kuwana, Y. Aihara, J. Electrochem. Soc., 150, A695 (2003). 8. (a) J. Sun, D. R. MacFarlane, M. Forsyth, Solid State Ionics, 147, 333 (2002). (b) M. Forsyth, J. Sun, F. Zhou, D. R. MacFarlane, Electrochim. Acta, 48, 2129 (2003). 9. M. Doyle, S. K. Choi, G. Proulx, J. Electrochem. Soc., 147, 34 (2000). 10. (a) J. Sun, L. R. Jordadn, M. Forsyth, D. R. MacFarlane, Electrochim. Acta, 46, 1703 (2001). (b) J. Sun, D. R. MacFarlane, M. Forsyth, Electrochim. Acta, 48, 1971 (2003). 11. W. Kubo, Y. Makimoto, T. Kitamura, Y. Wada, S. Yanagida, Chem. Lett., 31, 948 (2002). 12. P. Wang, S. M. Zakeeruddin, I. Exnar, M. Grätzel, Chem. Comm., 2972 (2002). 13. A. Lewandowski, A. Swiderska, Solid State Ionics, 161, 243 (2003).

c28.indd 407

1/20/2011 10:29:00 AM

408

ZWITTERIONIC LIQUID/POLYMER GELS

14. M. Ue, M. Takeda, A. Toriumi, A. Kominato, R. Hagiwara, Y. Ito, J. Electrochem. Soc., 150, A499 (2003). 15. J. D. Stenger-Smith, C. K. Webber, N. Anderson, A. P. Chafin, K. Zong, R. Reynolds, J. Electrochem. Soc., 149, A973 (2002). 16. (a) W. Lu, A. G. Fadeev, B. Qi, E. Smela, B. R. Mattes, J. Ding, G. M. Spinks, J. Mazurkiewicz, D. Zhou, G. G. Wallace, D. R. MacFarlane, S. A. Forsyth, M. Forsyth, Science, 297, 983 (2002). (b) D. Zhou, G. M. Spinks, G. G. Wallace, C. Tiyaboonchaiya, D. R. MacFarlane, M. Forsyth, J. Sun, Electrochim. Acta, 48, 2355 (2003). 17. (a) J. Fuller, A. C. Breda, R. T. Carlin, J. Electrochem. Soc., 144, L67 (1997). (b) J. Fuller, A. C. Breda, R. T. Carlin, J. Electroanal. Chem., 459, 29 (1998). 18. C. Tiyapiboonchaiya, D. R. MacFarlane, J. Sun, M. Forsyth, Macromol. Chem. Phys., 203, 1906 (2002). 19. T. Fukushima, A. Kosaka, Y. Ishimura, T. Yamamoto, T. Takigawa, N. Ishii, T. Aida, Science, 300, 2072 (2003). 20. N. Nishimura, H. Ohno, J. Mater. Chem., 12, 2299 (2002). 21. M. Yoshizawa, M. Hirao, K. Ito-Akita, H. Ohno, J. Mater. Chem., 11, 1057 (2001). 22. M. Freemantle, Chem. Eng. News, March 30, 32 (1998). 23. H. Ohno, Petrotech, 23, 556 (2000). in Japanese. 24. (a) H. Ohno, K. Ito, Chem. Lett., 27, 751 (1998). (b) M. Hirao, K. Ito, H. Ohno, Electrochim. Acta, 45, 1291 (2000). (c) H. Ohno, Electrochim. Acta, 46, 1407 (2001). 25. M. Yoshizawa, W. Ogihara, H. Ohno, Polymer Adv. Technol., 13, 589 (2002). 26. H. Ohno, M. Yoshizawa, W. Ogihara, Electrochim. Acta, 48, 2079 (2003). 27. H. L. Ngo, K. LeCompte, L. Hargens, A. B. McEwen, Thermochim. Acta, 357–358, 97 (2000). 28. A. B. McEwen, H. L. Ngo, K. LeCompte, J. L. Goldman, J. Electrochem. Soc., 146, 1687 (1999). 29. (a) H. Ohno, M. Yoshizawa, Polymer Preprints, Japan, 51, 3069 (2002). (b) M. Yoshizawa, M. Tamada, H. Ohno, Polymer Preprints, Japan, 52, 2615 (2003).

c28.indd 408

1/20/2011 10:29:00 AM

29 IONIC LIQUIDIZED DNA Naomi Nishimura and Hiroyuki Ohno

The typical cation structures of ionic liquids are heteroaromatic rings such as imidazolium cations or pyridinium cations. Molecular orbital calculations suggested that imidazole is a good starting material to prepare onium cations for excellent ionic liquids. Many other heteroaromatic rings also have been examined for this purpose. DNA consists of four nucleic acid bases, and these bases are heteroaromatic rings. Nucleic acid bases have been examined as starting materials to prepare ionic liquids by neutralization. Based on the ionic liquids comprising these nucleic acid bases, DNA was ionic liquidized to prepare new ion conductive materials. Reversible redox reaction of dye molecules fixed in the DNA also was mentioned as one application of ionic liquidized DNA [1].

29.1

DNA

DNA is generally recognized as a typical biopolymer that has genetic information. It is not so well known that the DNA is obtained easily with an inexhaustible supply in nature if proteins are separated from the gene. Because DNA has superior characteristics compared with the synthetic polymers, DNA is an attractive material, as shown in Figure 29.1. DNA is also readily available; salmon roe is frequently used as a Japanese dish, but considerable amounts of salmon milt, which is rich in DNA, are discarded. DNA is a kind of biomass. DNA has four kinds of nucleic acid bases, and the information is kept by hydrogen bonding between complementary base pairs. DNA has nucleic acid bases with high density. Therefore, much interest has been shown using these

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

409

c29.indd 409

1/20/2011 10:29:01 AM

410

IONIC LIQUIDIZED DNA

Characteristics; Rigid rod-like polymer, having a hetero aromatic rings, π stacked bases, inexhaustible supply in nature, and biodegradability

Figure 29.1. Characteristics of DNA.

characteristics of DNA in creating functional materials for electronics [2], optics, and the like [3]. In particular, there have been some attempts to prepare electroconductive materials using the π–π stacking of base pairs. However, except for this research, the application of ion conductive materials using DNA has not been undertaken. We have used a composite of DNA and poly(ethylene oxide) (PEO), and we found this composite to have excellent ionic conductivity [4].

29.2 29.2.1

IONIC LIQUIDIZED DNA Inner Column

Although the details are omitted, four kinds of bases as model compounds were investigated [5]. Because adenine and cytosine formed ionic liquids after neutralization, these subunits in DNA also should form ionic liquids after acid treatment. DNA has many bases aligned on the chain. If all these bases are converted into ionic liquids, then successive ionic liquid domains should be created along the chain. According to the results in [5], HBF4, HTFSI, and CF3SO3H are strongly indicated to be effective acids for this purpose. DNA was treated with these acids. The four kinds of bases were assumed to be contained equally in the DNA. Acids of 50 mol% to the total bases were mixed with DNA to neutralize all adenines and cytosines. After neutralization with the stated acids individually in pure water, all products were obtained as precipitates. It is known that purine rings are dissociated from DNA by strong acid treatment. There was some fear of the dissociation of the purine rings occurring in the present treatments. Nevertheless, we proceeded to prepare the ionic liquid domain of adenine salts in the DNA. The resulting samples were washed with water to remove unreacted acid, dissociated purine (salts),

c29.indd 410

1/20/2011 10:29:01 AM

411

IONIC LIQUIDIZED DNA

and so forth. This process gave us an acid-treated DNA without the dissociated purines. The hydrogen bonding between complementary base pairs was broken after neutralization, and the hydrophobic bases turned out from the helix. Circular dichroism (CD) and infrared (IR) spectra gave strong indications that these DNAs lose their double-strand helical structure. The ionic conductivity of the neutralized DNA was approximately 1 × 10−9 S cm−1 at room temperature. This can be explained by the insufficient ionic liquid fraction. The weight fraction of all bases in the solid DNA was approximately 40 wt%. However, if all adenines and cytosines had yielded ionic liquids, the corresponding domain fraction would have been only approximately 20 wt%. A closely packed ion conduction path cannot be formed with this low fraction. C · TFSI, which had the lowest Tm of the neutralized bases, therefore, was added to the neutralized DNA to assist the construction of a continuous ion conduction path. The neutralized DNA and C · TFSI were mixed in a solvent, and then they were cast to prepare films. The problem that we encountered in the processing was that DNA · TFSI is excellent matrix from the viewpoint of ionic conductivity, but it cannot be dissolved in any solvent. Because DNA · BF4 can be dissolved in excess of water, we used the DNA · BF4 in other experiments. When the C · TFSI content in DNA · BF4 was less than 50 wt%, the ionic conductivity of the mixture was low, at around 1 × 10−8 S cm−1 at 50 °C (Figure 29.2; ▲). It abruptly increased when more than 70 wt% C · TFSI was added, reaching 4.76 × 10−5 S cm−1 at 50 °C when 80 wt% of C · TFSI was added to the DNA · BF4. The mixture, however, was obtained as a flexible film even with 80 wt% C · TFSI. Because the ionic conductivity of C · TFSI in the bulk was only 6.85 × 10−5 S cm−1, the DNA · BF4 and C · TFSI mixed film showed

logσi (S/cm)

-2

-4

-6

-8 0

20

40

60

80

100

Added Ionic Liquid Concentration (wt%)

Figure 29.2. Effect of added ionic liquid concentration on the ionic conductivity of DNA · BF4/C · TFSI (▲), DNA · BF4/[emim][BF4] (•) at 50 °C. Those of pure C · TFSI (△) and pure [emim][BF4] (○) also were depicted as references.

c29.indd 411

1/20/2011 10:29:01 AM

412

IONIC LIQUIDIZED DNA

reasonable conductivities, indicating the formation of a successive ionic liquid phase. To improve the ionic conductivity of the DNA · BF4 film, we had to mix the DNA · BF4 with a more conductive ionic liquid. For this purpose, ethylimidazolium tetrafluoroborate ([emim][BF4]) was added to the DNA · BF4 (Figure 29.2; •). [emim][BF4] shows high ionic conductivity of around 10−2 S cm−1 at 50 °C. Because both DNA · BF4 and [emim][BF4] have the same BF4− anion and are soluble in pure water, the DNA · BF4/[emim][BF4] mixture was conveniently prepared as a homogeneous film by casting. When the amount of the added [emim][BF4] was up to 10 wt%, the mixture’s ionic conductivity was the same as that of pure DNA · BF4. However, the ionic conductivity of the film containing 15 wt% [emim][BF4] was approximately 4.62 × 10−7 S cm−1 and that of film containing 23.7 wt% [emim][BF4] was 1.74 × 10−4 S cm−1 at 50 °C. This excellent ionic conductivity was maintained up to 85 wt%. The highest ionic conductivity, 5.05 × 10−3 S cm−1 at 50 °C, was observed when the film was prepared with 93 wt% [emim][BF4]. Continued addition of [emim][BF4] maintained high ionic conductivity, but no film was obtained. We found, surprisingly, that DNA with only 7 wt% ionic liquid forms a film and solidity with 93 wt% ionic liquid because of the high molecular weight of DNA and the strong affinity of [emim][BF4] for DNA · BF4. We have already reported on the preparation of excellent ion conductive films from a mixture of native DNA and [emim][BF4]. There is no difference in the ionic conductivity between the DNA · BF4/[emim][BF4] film and the DNA/[emim][BF4] film at high ionic liquid content (>70 wt%), as shown in Figure 29.3 [6]. However, the DNA · BF4/ [emim][BF4] film displayed much higher ionic conductivity at a low [emim] [BF4] content. This can be explained by the formation of an effective ionic liquid pathway in the DNA matrix as a result of neutralization of the bases. Furthermore, the DNA · BF4/[emim][BF4] film showed high ionic conductivity

logσi (S/cm)

-2.0

-4.0

-6.0

-8.0

0

20

40

60

80

100

Added Ionic Liquid Concentration (wt%)

Figure 29.3. The relationship of ionic conductivity at 50 °C and added [emim][BF4] concentration in native DNA.

c29.indd 412

1/20/2011 10:29:01 AM

IONIC LIQUIDIZED DNA

413

and excellent flexibility over a wide ionic liquid content. Significant improvement in both the ionic conductivity and the flexibility of the DNA films is observed compared with the DNA/PEO1000/NaClO4 film [4] or the simple mixture of DNA and ionic liquid [5]. DNA · BF4 was mixed with 40 wt% [emim][BF4], and the ionic conductivity of the film was 1.32 × 10−3 S cm−1 at room temperature. In other words, flexible film with a high ionic conductivity can be obtained when DNA · BF4 is mixed with only a small amount of [emim] [BF4]. The experiments were carried out in a dry nitrogen atmosphere at room temperature. DNA · BF4/[emim][BF4] film showed excellent stability, and no leakage of ionic liquid was detected. This stability was observed continually at room temperature over several months. 29.2.2

Outer Column

Low-molecular-weight nucleic acid bases were used to prepare ionic liquids. In the previous section, bases located inside the helix structure were ionic liquidized. Because these bases are located inside the double-stranded DNA, ionic liquidization of these bases disturbed the double-stranded structure. In this section, ionic liquidization of phosphate anions on the outer column of the DNA has been carried out by keeping the double-stranded DNA structure [7]. We preliminarily confirmed that salts, comprising phosphoric acid di-nbutyl ester (PDE) and alkylmethylimidazolium cation (Cnmim), showed low Tg. The counter cations of phosphate groups of DNA were changed into alkylmethylimidazolium cations to form Cnmim-DNA (DNA robed with ionic liquids). All these Cnmim-DNA are soluble in methanol and ethanol. These Cnmim-DNA methanol solutions were cast on the glass plates to prepare flexible films after the drying process (Figure 29.4). The double-stranded structure of thus prepared Cnmim-DNA was confirmed by CD and ultraviolet (UV) spectroscopy. Raman spectroscopy also showed a double-stranded structure of Cnmim-DNA (for details, see Reference [7]). All these results strongly suggested that ionic liquidization of the outer column should be effective to keep a double stranded helix structure. Because ionic liquids have low Tg, the obtained DNA robed with ionic liquids were flexible films with high ionic conductivity. Figure 29.5 shows the effect of alkyl chain length on the imidazolium cation (n) on the ionic conductivity at 100 °C. Among four samples, c2mim-DNA showed the highest ionic conductivity of 1.9 × 10−6 S cm−1. Cnmim-DNA with a shorter alkyl chin length on the imidazolium cation showed higher ionic conductivity. This was the same tendency as that in the case of low-molecular-weight model compounds [7]. This low bulk ionic conductivity of Cnmim-DNA was attributed to a longer distance between cations in which no effective ion hopping was allowed. To improve the ionic conductivity at room temperature, 11 mol% [emim][BF4] was mixed with Cnmim-DNA. All mixed systems were obtained as flexible and uniform films. When four kinds of Cnmim-DNA were mixed with 11 mol% [emim][BF4] to the phosphate groups, ionic conductivity was

c29.indd 413

1/20/2011 10:29:01 AM

414

IONIC LIQUIDIZED DNA

Figure 29.4. Photograph of C12mim-DNA film.

log (σi S/cm) at 100 °C

-5.5

-6.0

-6.5

-7.0

-7.5

0

2

4

6

8

10

12

14

Alkyl Chain length Figure 29.5. Effect of alkyl chain length of imidazolium cation on the ionic conductivity of Cnmim-DNA at 100 °C.

improved considerably. The 11 mol% [emim][BF4] was equivalent to the ratio of one [emim] cation to nine phosphate anion groups. The highest ionic conductivity of 5.4 × 10−6 S cm−1 at 25 °C was found in [emim][BF4]/C2mim-DNA mixture. Continued addition of [emim][BF4] did not contribute to the increase in the ionic conductivity but induced the phase separation.

c29.indd 414

1/20/2011 10:29:01 AM

REDOX REACTION OF DYE MOLECULES IN THE IONIC LIQUIDIZED DNA

415

29.3 REDOX REACTION OF DYE MOLECULES IN THE IONIC LIQUIDIZED DNA In this section, ionic liquidized DNA was used as the host for some dye molecules to realize stable redox reaction without dimerization of the dyes. One serious problem for an organic dye-based electrochromic system is irreversible dimerization of the dye molecules. To inhibit the dimerization, we applied specific interaction between dyes and DNAs. Viologens were groove-bound into the C12mim-DNA complex film. There was no peak for dimerized viologen molecules observed in the UV spectroscopy. The polymerized ionic liquids were used as polymer electrolyte for this system [8]. It is important to design an ionic liquid copolymer (CoIL) with a few different roles such as flexibility, suitable ionic (bromide anion) conductivity, low glass transition temperature, and suppressed solubility of dye molecules. Heptyl viologen (HV) was groovebound in C12mim-DNA in 20 mol% to the DNA base pair. Figure 29.6 shows the scheme of our electrochromic cell. Addition of N,N,N′,N′-tetramethyl-pphenylenediamine (TMPD) as an active material for a counter electrode certainly suppressed the charge-up and allowed easy operation under a lower voltage. Furthermore, to increase the electrode surface area, ITO nanoparticles (100∼140 nm) were added to the film. The cell containing 10 wt% ITO nanoparticles showed improved efficiency. A greater amount of ITO nanoparticle made the film turbid. Reversible color changes were smoothly carried out by giving the voltage of more than ±0.7 V [9]. By adding more than two different dyes (should have different absorption profile and different redox

CH3

H3C N

N

H3C

CH3

TMPD

n

m

Br-

PO3-

N

N

N

NH

CoIL

ITO nano-particle C12 mim-DNA

Figure 29.6. Scheme of DNA-based electrochromic cell.

c29.indd 415

1/20/2011 10:29:02 AM

416

IONIC LIQUIDIZED DNA

potential), multistep color change was realized by continuous potential sweep. p-Cyanophenyl viologen (CV) was selected as a good partner of HV. About 13% HV and 7% CV were fixed stepwise into c12mim-DNA. The cell showed three different colors depending on the given potential [10]. This complexation of DNA and dyes is effective for many different dyes to inhibit dimerization as long as they form complexes with DNA.

REFERENCES 1. T. Kakibe, H. Ohno, Chem. Comm., 377 (2008). 2. J. Berashevich, T. Chakraborty, J. Phys. Chem. B, 112, 14083 (2008). 3. C. K. Lee, S. R. Shin, J. Y. Mun, S. Han, I. So, J. Jeon, T. M. Kang, S. I. Kim, P. G. Whitten, G. G. Wallace, G. M. Spinks, S. J. Kim, Angew. Chem. Int. Ed., 48, 5116 (2009). 4. H. Ohno, N. Takizawa, Chem. Lett., 642 (2000). 5. N. Nishimura, H. Ohno, J. Mater. Chem., 12, 2299 (2002). 6. H. Ohno, N. Nishimura, J. Electrochem. Soc., 148, E168 (2001). 7. N. Nishimura, Y. Nomura, N. Nakamura, H. Ohno, Biomaterials, 26, 5558 (2005). 8. M. Yoshizawa, W. Ogihara, H. Ohno, Polymer Adv. Technol., 13, 589 (2002). 9. T. Kakibe, H. Ohno, J. Mater. Chem., 19, 4960 (2009). 10. T. Kakibe, R. Kurihara, H. Ohno, Chem. Lett., 38, 494 (2009).

c29.indd 416

1/20/2011 10:29:02 AM

PART VII POLYMERIZED IONIC LIQUIDS

c30.indd 417

1/20/2011 10:29:04 AM

30 ION CONDUCTIVE POLYMERS Hiroyuki Ohno and Masahiro Yoshizawa-Fujita

Because ionic liquids (ILs) consist only of ions, they offer two brilliant features: high concentration of ions [1] and high mobility of component ions at room temperature. Because many ILs show the ionic conductivity of more than 10−2 S cm−1 at room temperature [2,3], there are plenty of possible applications as electrolyte materials (for rechargeable lithium ion batteries [4–14], fuel cells [15–21], solar cells [22–34], and capacitors [35–43]). However, so far, the ILs studied as electrolyte materials have shown a fatal drawback in that their component ions migrate along with the potential gradient. As a result, there is a concern about the leakage of liquid ILs in battery technology as well as in other typical organic electrolyte applications. Certain kinds of polymer gel electrolytes containing ILs have been investigated because of their high ionic conductivity and good mechanical properties. For example, findings for poly(vinylidene fluoride-co-hexafluoropropylene) (P(VdF-co-HFP)) [19,20,25,44–47], Nafion [15,21], DNA [48], and some polymer electrolytes [49] containing ILs have been reported. Polymer-in-IL electrolytes also were prepared by in situ polymerization of vinyl monomers in ILs [50]. Because we first reported polymerized ILs and their ionic conductivities in 1998 [51], many systems have been investigated as new electrolyte materials for electrochemical applications. In this chapter, the ion conductive behavior of new ion conductive polymers that polymerize the IL itself will be reported.

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

419

c30.indd 419

1/20/2011 10:29:05 AM

420

ION CONDUCTIVE POLYMERS

30.1

POLYCATIONS

The IL monomers were prepared from vinylimidazole and polymerized to investigate the relationship between the IL polymer structure and their ionic conductivity. Figure 30.1 shows the structures of several IL monomers with different polymerizable cations. Monomer I was prepared by the neutralization of N-vinylimidazole and HBF4 [52]. It already has been reported that ILs can be obtained by the neutralization of the tertiary amines and organic acids (see Chapter 20) [53–55]. This procedure is simple, and only equimolar mixings of tertiary amines and acids provide salts without generation of by-products. Therefore, this is one of the most favored procedures to prepare pure ILs and their models. Monomer I showed its melting point (Tm) at 55 °C, whereas it showed its glass transition temperature (Tg) at −5 °C after polymerization. However, although monomer I showed the ionic conductivity of about 10−3 S cm−1 at 50 °C, the ionic conductivity of the corresponding polymer was about 10−8 S cm−1, which is 10,000 times lower than that of the monomer, as shown in Figure 30.2. This decrease should be based on the decreases of carrier ion number and their mobility. There are two rough strategies to overcome this drop of ionic conductivity of polymer I. One is the design of a polymer structure to lower the Tg. Recently, Chen et al. reported interesting results [56]. New polymerized ILs comprising hexyl methacrylate (nonionic) and an imidazolium-based methacrylate (ionic) were synthesized as shown in Figure 30.3. The ionic conductivity of the ionic polymer was 5 × 10−3 S cm−1 at 100 °C. When the hexyl methacrylate content was 61.7 mol%, the copolymer showed the ionic conductivity of 2 × 10−2 S cm−1 at 100 °C. The ionic conductivity of nonionic-ionic copolymers increased by almost an order of magnitude with increasing hexyl methacrylate composition despite a decrease in carrier ion

N + NH

TFSI−

BF4−

N + N

I

II SO3−

N + N III TFSI− N + N

IV

Figure 30.1. Structures of IL monomers with different polymerizable cations.

c30.indd 420

1/20/2011 10:29:05 AM

421

POLYCATIONS

Ionic conductivity log(σi / S cm-1)

-2

monomer I -4

-6 + LiBF4 -8 polymer I -10 3.0

3.1

3.2

3.3 -1

3.4

3.5

3.6

-1

1000 T /K

Figure 30.2. Arrhenius plots of the ionic conductivity for monomer I, its polymer, and the polymer containing LiBF4 (polymer: LiBF4 = 4 : 1, mole fraction).

y

x

O

O

O

O

N

BF4 N

Figure 30.3. Structure of the nonionic–ionic copolymers.

numbers. Such behavior should be explained by the lowering of Tg. The Tg of the ionic polymer was 71.5 °C, whereas the Tg of copolymers was below 10 °C when the hexyl methacrylate composition was more than 50 mol%. The introduction of nonionic units significantly lowered the Tg. The effect of lowering of Tg on the ionic conductivity also will be mentioned in Chapter 32. The other strategy is to add more carrier ions. In the present chapter, we mainly consider

c30.indd 421

1/20/2011 10:29:05 AM

422

ION CONDUCTIVE POLYMERS

the case in which lithium salts were added to the polymer matrix. The ionic conductivity of polymer I containing LiBF4 25 mol% to imidazolium cation was about 10−4 S cm−1 at 50 °C, which is almost equivalent to that of monomer I. The carrier ions increased with the added LiBF4, but the number of ions was not the only factor to govern the ionic conductivity. That is, when LiCl was added to the polymer, no improvement of ionic conductivity was found. The dissociation degree of the added salt as well as changes in viscosity (in terms of segmental motion) would turn out to be important factors in improving the ionic conductivity of the IL polymer. Monomer II is also a polymerizable IL comprising quaternized imidazolium salt [51], as shown in Figure 30.1. This monomer is liquid at room temperature and shows a Tg only at −70 °C. Its high ionic conductivity of about 10−2 S cm−1 at room temperature reflects a low Tg. Although the ionic conductivity of this monomer decreased after polymerization, as in the case of monomer I, it was considerably improved by the addition of a small amount of LiTFSI. Figure 30.4 shows the effect of LiTFSI concentration on the ionic conductivity and lithium transference number ( tLi + ) for polymer II. The bulk ionic conductivity of polymer II was 10−5 S cm−1 at 50 °C. When LiTFSI was added to polymer II, the ionic conductivity increased up to 10−3 S cm−1. After that, the ionic conductivity of polymer II decreased gradually with the increasing LiTFSI concentration. However, when the LiTFSI concentration was 100 mol%, the tLi + of this system exceeded 0.5. Because of the fixed imidazolium cations on the polymer chain, more mobile anion species than cation species exist in the polymer matrix at this concentration. Because the TFSI anions form the IL domain

-2

o

log(σi / S cm-1) at 50 C

-3 0.50

-4

0.25

-5

-6 0

2

4

6

8

10

100

200

300

400

Li transference number

0.75

0.00 500

LiTFSI concentration / mol% to monomer unit

Figure 30.4. Effects of LiTFSI concentration on the ionic conductivity and lithium transference number for polymer II.

c30.indd 422

1/20/2011 10:29:05 AM

423

POLYCATIONS

TFSI N

N

TFSI

OMe N

B

N

n

B n

Figure 30.5. Structure of the organoboron polymer electrolytes.

with the imidazolium cations, the anions can supply a successive ion conduction path for the lithium cation. Such behavior is not observed in monomeric IL systems and is understood to be a result of the concentrated charge domains created by the polymerization. In recent years, novel polymerized ILs with anion trapping sites were synthesized to improve tLi + strongly. Organoboron polymer electrolytes were obtained by hydroboration polymerization of 1,3-diallylimidazolium bromide and subsequent anion exchange reaction (Figure 30.5) [57]. After the addition of an equimolar amount of LiTFSI to the organoboron unit, the mixtures showed the ionic conductivity of about 10−5 S cm−1 at 50 °C. The tLi + for these systems was in the range of 0.45 and 0.87 at 30 °C, which was much higher than those of polyether systems. This strategy was applied to the preparation of novel ion conductive composite materials [58]. Inorganic–organic composites were prepared with polymerized ILs and a borosilicate network. The tLi + for the composites increased with an increasing boron content. These results demonstrate that anion trapping of the boron unit was effective in IL-based matrices. Polymers I and II showed inherently low ionic conductivity and had poor mechanical properties. IL monomers with benzene rings were prepared to improve the properties of these polymer because it is known that most imidazolium salts that have benzene rings form ILs at room temperature and show relatively high ionic conductivity [53(c)]. In addition, the polymers with benzene rings are generally superior in terms of thermal stability and film preparation. Two kinds of IL monomers, which have a benzene ring on either the cation or the anion, were synthesized [59], as shown in Figure 30.1. Monomer III was obtained by the reaction of vinylimidazole and by an equimolar amount of p-toluenesulfonic acid ethyl ester (p-TSE). This is also a simple reaction similar to the neutralized amines. Because p-TSE is cheap, various ILs were prepared by this reaction [60]. The monomer IV was prepared by quaternization of N-ethylimidazole with 4-chlorostyrene. Monomer III showed a Tm at 83 °C, whereas monomer IV showed a Tg at −59 °C. The ionic conductivity of monomer IV (3.07 × 10−3 S cm−1 at 50 °C) was four orders higher than that of monomer III (1.64 × 10−7 S cm−1 at 50 °C), reflecting its improved thermal property. Monomer IV is a kind of IL, and it shows high ionic conductivity. Figure 30.6 gives the ionic conductivity values of III and IV after polymerization. The effects of the LiTFSI addition on the

c30.indd 423

1/20/2011 10:29:05 AM

424

ION CONDUCTIVE POLYMERS

Ionic conductivity log(σi / S cm-1)

-3

-4

-5

-6

-7 + LiTFSI -8 3.0

3.1

3.2

3.3 -1

3.4

3.5

3.6

-1

1000 T /K

Figure 30.6. Effects of LiTFSI addition on the ionic conductivities of polymer III and IV. △: polymer IV; ▲: polymer IV + LiTFSI; •: polymer III + LiTFSI.

ionic conductivities of polymers III and IV are shown in this figure. Polymer III acted mostly as an insulator, so its conductivity was too low to appear on this scale. Polymer IV’s ionic conductivity was about 10−6 S cm−1 at room temperature. When an equimolar amount of LiTFSI was added to the imidazolium unit, the ionic conductivity values of these polymers increased considerably. Polymer III especially showed an ionic conductivity of more than 10−4 S cm−1 at room temperature, which is higher than that of the corresponding monomer. Figure 30.7 shows the effects of the LiTFSI concentration on the ionic conductivities at 30 °C for polymers III and IV. Although the ionic conductivity of polymer III was poor in bulk, it improved considerably with the increasing LiTFSI concentration. When the LiTFSI concentration was 100 mol% to the imidazolium unit, the ionic conductivity of polymer III reached its maximum value. After more additions, there was no obvious increase of ionic conductivity. The Tg of polymer III became lower with increasing LiTFSI concentration. The Tg fell to −80 °C when the LiTFSI concentration was 100 mol%. The fall of Tg was brought on by the increase in the number of imidazolium cations and TFSI anion pairs. However, although the addition of TFSI salt generally can be expected to produce a plasticizing effect, this effect cannot be used to fully explain the low Tg at the 100 mol% addition. However, the ionic conductivity of polymer IV also increased with the addition of LiTFSI. Although the

c30.indd 424

1/20/2011 10:29:05 AM

425

POLYCATIONS

log(σi / S cm-1) at 30 °C

-2

polymer III

-4

-6 polymer IV -8

-10 0 100 200 300 400 500 LiTFSI concentration / mol% to monomer unit

Figure 30.7. Effects of LiTFSI concentration on the ionic conductivities of polymers III and IV at 30 °C.

increase was small, it had to do with the relatively high ionic conductivity of polymer IV in bulk. Two Tg values of polymer IV were observed at the high salt concentration, indicating that polymer IV had a microphase separation with LiTFSI. For example, two Tg values were observed at −4 °C and −49 °C when the salt concentration was 100 mol%. This phase separation would be caused by the low solubility of salts in the polymer matrix that have a benzene ring. Polycation-type ILs can provide an ion conductive path for the added lithium salts. High rate performance of a lithium polymer battery fabricated with a composite comprising a lithium salt dissolved in an IL and a polymerized IL has been reported by Sato et al. [61]. A lithium ion cell retained 83% of its discharge capacity at 3 °C and showed relatively good cycle performance. In addition, polymerized ILs were designed as electrolyte materials for dye-sensitized solar cells [62,63]. Figure 30.8 shows structures of IL monomers for iodide ion conductors. Gel polymer electrolytes based on these monomers showed the ionic conductivity values of 10−3–10−4 S cm−1 at room temperature. Polycation-type ILs will be suitable as highly anion-conducting solid electrolytes because iodide ions can migrate and onium cations are immobilized. Polymerized ILs usually have been synthesized via polymerization of vinyl monomers as mentioned previously. Recently, Vygodskii et al. reported novel polymer electrolytes derived from poly(norbornene)s with pendant imidazolium moieties and three different counter anions, as shown in Figure 30.9 [64].

c30.indd 425

1/20/2011 10:29:05 AM

426

ION CONDUCTIVE POLYMERS

O

MeO

O

MeO N

Si

I

OMe

I

N

N

N

Figure 30.8. Structures of IL monomers for iodide ion conductors.

n O O

X

N N

X = BF4, PF6, TFSI

Figure 30.9. Structure of poly(norbornene)s with pendant imidazolium moieties.

The weight-average molecular weights of the polymers were determined via static light scattering to be in the range of 8100 and 44,000. It seems that even larger values might be suitable for free-standing films. The polymer with a TFSI anion showed the highest ionic conductivity value of 1.44 × 10−4 S cm−1 at 50 °C. The ionic conductivities increased in the order TFSI > PF6 > BF4, whereas the Tg values decreased in the same order. The ionic conductivities and the Tg values showed a good correlation in the same way as the vinyl systems did. These results suggest that the properties of polycation-type ILs mainly depend on the nature of the counteranions.

30.2

POLYANIONS

The IL polymers were prepared from polymerizable anions and free imidazolium cations. To investigate the effect of the anion structure on the properties, we synthesized four kinds of IL monomers by neutralizing Nethylimidazole with acrylic acid, p-styrenesulfonic acid, vinylsulfonic acid, and

c30.indd 426

1/20/2011 10:29:05 AM

427

POLYANIONS SO3− COO− HN + N

HN + N

ElmAC

ElmSS

O HN + N

SO3−

OH O− P O

ElmVS

HN + N ElmVP

Figure 30.10. Structures of IL monomers with different polymerizable anions.

TABLE 30.1. Ionic Conductivity and Tg for Ionic Liquid Monomers and Their Polymers σi/S cm−1 (30 °C)

Tg/°C Monomer EImAC EImSS EImVS EImVP

−61 −77 −95 −78

Polymer −32 — −63 −56

Monomer −4

1.4 × 10 8.7 × 10−5 9.0 × 10−3 1.5 × 10−4

Polymer 1.2 1.1 1.1 2.9

× × × ×

10−6 10−8 10−4 10−5

Note: — not detected.

vinylphosphonic acid, respectively, as shown in Figure 30.10 [65]. Table 30.1 summarizes the properties of IL monomers comprising the polymerizable anion and an ethylimidazolium cation. They are liquid at room temperature and show low Tgs, in the range of −100 °C to −60 °C. EImVS showed the lowest Tg of −95 °C among these four IL monomers. In addition, it showed the highest ionic conductivity of about 10−2 S cm−1 at 30 °C, which is comparable with those of the best ILs such as the 1-ethyl-3-methylimidazolium salts. The IL monomers comprising polymerizable acid with a lower pKa show higher ionic conductivity. Table 30.1 also summarizes the properties of the polymerized IL monomers. The Tg was elevated about 30 °C after polymerization. P(EImVS) (P implies polymer) showed the lowest Tg of −63 °C and the highest ionic conductivity of 1.1 × 10−4 S cm−1 at 30 °C among these four polymers. The IL with the VS anion kept its matrix flexibility high even after polymerization. However, the ionic conductivity of P(EImSS) was extremely low. Although the Tg of P(EImSS) was not detected by the differential scanning calorimetry measurement, the Tg of P(EImSS) might have increased considerably after polymerization. As mentioned previously, there is a linear relationship between Tg and the ionic conductivity [66], and accordingly,

c30.indd 427

1/20/2011 10:29:05 AM

428

ION CONDUCTIVE POLYMERS

this can be used to estimate the Tg from its ionic conductivity. When the ion concentration is comparable, the ionic conductivity can roughly be estimated from this relation. As a result, the Tg of P(EImSS) can be estimated to be around 5 °C based on the low ionic conductivity of about 10−8 S cm−1 at 30 °C.

REFERENCES 1. P. Wasserscheid, T. Welton, eds., Ionic Liquids in Synthesis, Wiley-VCH, 2003. 2. R. Hagiwara, Y. Ito, J. Fluorine Chem., 105, 221 (2000). 3. P. Bonhôte, A.-P. Dias, M. Armand, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). 4. (a) H. Sakaebe, H. Matsumoto, Electrochem. Commun., 5, 594 (2003). (b) H. Sakaebe, H. Matsumoto, K. Tatsumi, J. Power Sourc, 146, 693 (2005). (c) H. Matsumoto, H. Sakaebe, K. Tatsumi, J. Power Sourc, 146, 45 (2005). (d) H. Sakaebe, H. Matsumoto, K. Tatsumi, Electrochim. Acta, 53, 1048 (2007). 5. H. Nakagawa, S. Izuchi, K. Kuwana, T. Nukuda, Y. Aihara, J. Electrochem. Soc., 150, A695 (2003). 6. (a) P. C. Howlett, D. R. MacFarlane, A. F. Hollenkamp, Electrochem. Solid State Lett., 7, A97 (2004). (b) P. C. Howlett, N. Brack, A. F. Hollenkamp, D. R. MacFarlane, M. Forsyth, J. Electrochem. Soc., 153, A595 (2006). 7. B. Garcia, S. Lavallee, G. Perron, C. Michot, M. Armand, Electrochim. Acta, 49, 4583 (2004). 8. S. Y. Lee, H. H. Yong, Y. J. Lee, S. K. Kim, S. Ahn, J. Phys. Chem. B, 109, 13663 (2005). 9. (a) J. H. Shin, W. A. Henderson, S. Passerini, J. Electrochem. Soc., 152, A978 (2005). (b) J. H. Shin, W. A. Henderson, S. Passerini, Electrochem. Commun., 5, 1016 (2003). 10. (a) S. Seki, Y. Kobayashi, H. Miyashiro, Y. Ohno, Y. Mita, A. Usami, N. Terada, M. Watanabe, Electrochem. Solid State Lett., 8, A577 (2005). (b) S. Seki, Y. Kobayashi, H. Miyashiro, Y. Ohno, A. Usami, Y. Mita, N. Kihira, M. Watanabe, N. Terada, J Phys. Chem. B, 110, 10228 (2006). (c) S. Seki, Y. Kobayashi, H. Miyashiro, Y. Ohno, A. Usami, Y. Mita, M. Watanabe, N. Terada, Chem. Comm., 544 (2006). 11. M. Yoshizawa-Fujita, D. R. MacFarlane, P. C. Howlett, M. Forsyth, Electrochem. Commun., 8, 445 (2006). 12. A. I. Bhatt, A. S. Best, J. Huang, A. F. Hollenkamp, J. Electrochem. Soc., 157, A66 (2010). 13. G. B. Appetecchi, G.-T. Kim, M. Montanino, M. Carewska, R. Marcilla, D. Mecerreyes, I. De Meatza, J. Power Sourc, 195, 3668 (2010). 14. (a) K. Tsunashima, F. Yonekawa, M. Sugiya, Electrochem. Solid State Lett., 12, A54 (2009). (b) K. Tsunashima, F. Yonekawa, M. Sugiya, Chem. Lett., 37, 314 (2008). (c) K. Tsunashima, M. Sugiya, Electrochem. Solid State Lett., 11, A17 (2008). 15. M. Doyle, S. K. Choi, G. Proulx, J. Electrochem. Soc., 147, 34 (2000). 16. R. F. de Souza, J. C. Padilha, R. S. Goncalves, J. Dupont, Electrochem. Commun., 5, 728 (2003). 17. M. A. Navara, S. Penero, B. Scrosati, Electrochem. Solid State Lett., 8, A324 (2005).

c30.indd 428

1/20/2011 10:29:06 AM

REFERENCES

429

18. (a) K. Kudo, S. Mitsushima, N. Kamiya, K. Ota, Electrochemistry, 73, 668 (2005). (b) K. Kudo, S. Mitsushima, N. Kamiya, K. Ota, Electrochemistry, 73, 272 (2005). 19. S. S. Sekhon, P. Krishnan, B. Singh, K. Yamada, C. S. Kim, Electrochim. Acta, 52, 1639 (2006). 20. (a) A. Fernicola, M. A. Navarra, S. Panero, J. Appl. Electrochem., 38, 993 (2008). (b) A. Fernicola, S. Panero, B. Scrosati, J. Power Sourc. 178, 591 (2008). 21. (a) E. Cho, J.-S. Park, S. S. Sekhon, G.-G. Park, T.-H. Yang, W.-Y. Lee, C.-S. Kim, S.-B. Park, J. Electrochem. Soc., 156, B197 (2009). (b) S. S. Sekhon, J.-S. Park, E. Cho, Y.-G. Yoon, C.-S. Kim, W.-Y. Lee, Macromolecules, 42, 2054 (2009). (c) S. S. Sekhon, J.-S. Park, J.-S. Baek, S.-D. Yim, T.-H. Yang, C.-S. Kim, Chem. Mater., 22, 803 (2010). 22. N. Papageogiou, Y. Athanassov, M. Armand, P. Bonhôte, H. Pettersson, A. Azam, M. Grätzel, J. Electrochem. Soc., 143, 3099 (1996). 23. R. Kawano, M. Watanabe, Chem. Comm., 330 (2003). (b) R. Kawano, H. Matsui, C. Matsuyama, A. Sato, M. A. B. H. Susan, N. Tanabe, M. Watanabe, J. Photochem. Photobio. Chem. 164, 87 (2004). (c) R. Kawano, M. Watanabe, Chem. Comm., 2107 (2005). 24. W. Kubo, T. Kitamura, K. Hanabusa, Y. Wada, S. Yanagida, Chem. Comm., 374 (2002). 25. P. Wang, S. M. Zakeeruddin, I. Exnar, M. Grätzel, Chem. Comm., 2972 (2002). 26. H. Matsumoto, T. Matsuda, T. Tsuda, R. Hagiwara, Y. Ito, Y. Miyazaki, Chem. Lett., 26 (2001). 27. (a) S. Murai, S. Mikoshiba, H. Sumino, T. Kato, S. Hayase, Chem. Comm., 1534 (2003). (b) Y. Shibata, T. Kato, T. Kado, R. Shiratuchi, W. Takashima, K. Kaneto, S. Hayase, Chem. Comm., 2730 (2003). 28. P. Wang, S. M. Zakeeruddin, J.-E. Moser, M. Grätzel, J. Phys. Chem. B, 107, 13280 (2003). (b) P. Wang, S. M. Zakeeruddin, R. Humphry-Baker, M. Grätzel, Chem. Mater., 16, 2694 (2004). 29. P. Wang, B. Wenger, R. Humphry-Baker, J.-E. Moser, J. Teuscher, W. Kantlehner, J. Mezger, E. V. Stoyanov, S. M. Zakeeruddin, M. Grätzel, J. Am. Chem. Soc., 127, 6850 (2005). 30. J. Chen, D. L. Officer, J. M. Pringle, D. R. MacFarlane, C. O. Too, G. G. Wallace, Electrochem. Solid State Lett., 8, A528 (2005). 31. N. Mohmeyer, D. Kuang, P. Wang, H.-W. Schmidt, S. M. Zakeeruddin, M. Grätzel, J. Mater. Chem., 16, 2978 (2006). 32. D. Kuang, S. Uchida, R. Humphry-Baker, S. M. Zakeeruddin, M. Grätzel, Angew. Chem. Int. Ed., 47, 1923 (2008). 33. Y. Bai, Y. Cao, J. Zhang, M. Wang, R. Li, P. Wang, S. M. Zakeeruddin, M. Grätzel, Nat. Mater., 7, 626 (2008). 34. P. Chen, J. Brillet, H. Bala, P. Wang, S. M. Zakeeruddin, M. Grätzel, J. Mater. Chem., 19, 5325 (2009). 35. A. Lewandowski, A. Swiderska, Solid State Ionics, 161, 243 (2003). 36. A. Lewandowski, M. Galinski, J. Phys. Chem. Solid, 65, 281 (2004). 37. M. Ue, M. Takeda, A. Toriumi, A. Kominato, R. Hagiwara, Y. Ito, J. Electrochem. Soc., 150, A499 (2003).

c30.indd 429

1/20/2011 10:29:06 AM

430

ION CONDUCTIVE POLYMERS

38. J. D. Stenger-Smith, C. K. Webber, N. Anderson, A. P. Chafin, K. Zong, R. Reynolds, J. Electrochem. Soc., 149, A973 (2002). 39. A. B. McEwen, H. L. Ngo, K. LeCompte, J. L. Goldman, J. Electrochem. Soc., 146, 1687 (1999). 40. Y.-J. Kim, Y. Matsuzawa, S. Ozaki, K. C. Park, C. Kim, M. Endo, H. Yoshida, G. Masuda, T. Sato, M. S. Dresselhaus, J. Electrochem. Soc., 152, A710 (2005). 41. T. Sato, G. Masuda, K. Takagi, Electrochim. Acta, 49, 3603 (2004). 42. A. Balducci, U. Bardi, S. Caporali, M. Mastragostino, F. Soavi, Electrochem. Commun., 6, 566 (2004). 43. T. Katakabe, T. Kaneko, M. Watanabe, T. Fukushima, T. Aida, J. Electrochem. Soc., 152, A1913 (2005). 44. (a) J. Fuller, A. C. Breda, R. T. Carlin, J. Electrochem. Soc., 144, L67 (1997). (b) J. Fuller, A. C. Breda, R. T. Carlin, J. Electroanal. Chem., 459, 29 (1998). 45. H. Ohno, M. Yoshizawa, W. Ogihara, Electrochim. Acta, 48, 2079 (2003). 46. (a) K.-S. Kim, S.-Y. Park, S. Choi, H. Lee, J. Power Sourc, 155, 385 (2006). (b) K.-S. Kim, S.-Y. Park, S.-H. Yeon, H. Lee, Electrochim. Acta, 50, 5673 (2005). (c) S.-H. Yeon, K.-S. Kim, S. Choi, J.-H. Cha, H. Lee, J. Phys. Chem. B, 109, 17928 (2005). 47. (a) B. Singh, M. S. Hundal, G.-G. Park, J.-S. Park, W.-Y. Lee, C.-S. Kim, K. Yamada, S. S. Sekhon, Solid State Ionics, 178, 1404 (2007). (b) B. Singh, S. S. Sekhon, Chem. Phys. Lett., 414, 34 (2005). (c) B. Singh, S. S. Sekhon, J. Phys. Chem. B, 109, 16539 (2005). 48. (a) H. Ohno, N. Nishimura, J. Electrochem. Soc., 148, E168 (2001). (b) N. Nishimura, H. Ohno, J. Mater. Chem., 12, 2299 (2002). 49. (a) M. Forsyth, J. Sun, F. Zhou, D. R. MacFarlane, Electrochim. Acta, 48, 2129 (2003). (b) C. Tiyapiboonchaiya, J. M. Pringle, D. R. MacFarlane, M. Forsyth, J. Sun, Macromol. Chem. Phys., 204, 2147 (2003). 50. (a) A. Noda, M.Watanabe, Electrochim.Acta, 45, 1265 (2000). (b) C.Tiyapiboonchaiya, D. R. MacFarlane, J. Sun, M. Forsyth, Macromol. Chem. Phys., 203, 1906 (2002). (c) T. Tsuda, T. Nohira, Y. Nakamori, K. Matsumoto, R. Hagiwara, Y. Ito, Solid State Ionics, 149, 295 (2002). 51. H. Ohno, K. Ito, Chem. Lett., 751 (1998). 52. M. Hirao, K. Ito, H. Ohno, Electrochim. Acta, 45, 1291 (2000). 53. (a) M. Hirao, H. Sugimoto, H. Ohno, J. Electrochem. Soc., 147, 4168 (2000). (b) M. Yoshizawa, W. Ogihara, H. Ohno, Electrochem. Solid State Lett., 4, E25 (2001). (c) H. Ohno, M. Yoshizawa, Solid State Ionics, 154–155, 303 (2002). 54. M. Yoshizawa, W. Xu, C. A. Angell, J. Am. Chem. Soc., 125, 15411 (2003). 55. T. L. Greaves, C. J. Drummond, Chem. Rev., 108, 206 (2008). 56. H. Chen, J.-H. Choi, D. Salas-de la Cruz, K. I. Winey, Y. A. Elabd, Macromolecules, 42, 4809 (2009). 57. N. Matsumi, K. Sugai, M. Miyake, H. Ohno, Macromolecules, 39, 6924 (2006). 58. T. Mizumo, T. Watanabe, N. Matsumi, H. Ohno, Polymer Adv. Technol., 19, 1445 (2008). 59. M. Hirao, K. Ito-Akita, H. Ohno, Polymer Adv. Technol., 11, 534 (2000). 60. J. Golding, S. Forsyth, D. R. MacFarlane, M. Forsyth, G. B. Deacon, Green Chem., 4, 223 (2002).

c30.indd 430

1/20/2011 10:29:06 AM

REFERENCES

431

61. T. Sato, S. Marukane, T. Narutomi, T. Akao, J Power Sourc, 164, 390 (2007). 62. B. Yo, F. Zhou, C. Wang, W. Liu, Eur. Polymer J., 43, 2699 (2007). 63. E. Stathatos, V. Jovanovski, B. Orel, I. Jerman, P. Lianos, J. Phys. Chem. C, 111, 6528 (2007). 64. Y. S. Vygodskii, A. S. Shaplov, E. I. Lozinskaya, K. A. Lyssenko, D. G. Golovanov, I. A. Malyshkina, N. D. Gavrilova, M. R. Buchmeiser, Macromol. Chem. Phys., 209, 40 (2008). 65. H. Ohno, M. Yoshizawa, W. Ogihara, Electrochim. Acta, 50, 255 (2004). 66. (a) Y. Nakai, K. Ito, H. Ohno, Solid State Ionics, 113–115, 199 (1998). (b) Y. Tominaga, N. Takizawa, H. Ohno, Electrochim. Acta, 45, 1285 (2000). (c) M. Yoshizawa, K. ItoAkita, H. Ohno, Electrochim. Acta, 45, 1617 (2000).

c30.indd 431

1/20/2011 10:29:06 AM

31 AMPHOTERIC POLYMERS Hiroyuki Ohno, Masahiro Yoshizawa-Fujita, and Wataru Ogihara

The polymerized ionic liquid (IL) shows great promise for diverse applications. Some polymerization methods have already been oriented toward specific applications. Polymerized ILs are useful in polar environments or where there are ion species for transport in the matrix. Amphoteric polymers that contain no carrier ions are being considered for several purposes in polymer electrolytes. Zwitterionic liquids (ZILs) were introduced in Chapter 21 as ILs in which component ions cannot move with the potential gradient. ZILs can provide ion-conductive paths upon the addition of salt to the matrix. It is therefore possible to realize selective ion transport in an IL matrix. If the resulting matrix can form a solid film over a wide temperature range, many useful ionic devices can be realized. This chapter focuses on the preparation and characteristics of amphoteric IL polymers. Amphoteric IL polymers contain no carrier ions, because the migration of component ions is completely excluded by fixing them on the polymer chain. They can therefore transport only added target ions in the matrix. Research into solid polymer electrolytes has so far mainly involved the use of polyether matrices, because these polyethers can dissolve salts and transport the dissociated ions along with intramolecular and intermolecular motion of these polyether chains [1,2]. Polymer matrices having both a high polarity and a low glass transition temperature are rare. However, the ionic conductivity of certain polyether systems is approximately 10−4 S cm−1 at room temperature [3], and it is difficult to improve cationic conductivity. This is a consequence of the ion– dipole interaction of polyethers to target ions, known as strong solvation of

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

433

c31.indd 433

1/20/2011 10:29:07 AM

434

AMPHOTERIC POLYMERS

polyethers to cations. Alternative matrices have therefore been sought to overcome the problem. Some polyampholytes, in which both cation and anion are fixed in the polymer main chain and have a low glass transition temperature, have been studied as a possible polyether-alternative ion-conductive matrix. These amphoteric polymers can provide ion-conductive paths for target ions without migration of the IL itself. There are two main categories of polyampholytes: copolymers and poly(zwitterion)s.

31.1

COPOLYMERS

Copolymers can be obtained by the polymerization of IL monomers synthesized by the neutralization of acid and base, which both have vinyl groups [4]. In the case of copolymer systems, the ionic conductivity of IL copolymers should be low because there are no carrier ions in the matrix. However, IL copolymers are expected to provide ion-conductive paths for target ions upon adding appropriate salts. Figure 31.1 shows the structure of IL monomers where both cation and anion have vinyl groups. To compare the effect of the alkyl spacer on the properties of IL copolymers, monomer 2 was also synthesized and polymerized, having the alkyl spacer between the vinyl group and the sulfonate group. Table 31.1 summarizes the effect of an alkyl spacer on the properties of IL copolymers. Both monomers had only a glass transition temperature (Tg), and

HN + N

SO3 1

O

HN + N

SO3

O

2

Figure 31.1. Structures of IL monomers composed of cation and anion both having a vinyl group.

TABLE 31.1. Thermal Property and Ionic Conductivity of IL Monomers and Their Polymers σi at 30 °C/S cm−1

Tg/°C Monomer 1 2

c31.indd 434

−83 −73

Polymer — −31

Monomer −3

3.5 × 10 6.5 × 10−4

Polymer <10−9 <10−9

1/20/2011 10:29:07 AM

435

COPOLYMERS

were colorless transparent liquids at room temperature. The ionic conductivity of the monomers was 10−4–10−3 S cm−1 at room temperature. This value reflects the low Tg values of −83 °C and −73 °C, respectively. However, 1 did not show Tg after polymerization, whereas polymerized 2 had Tg = −31 °C, implying that 2 maintains a high mobility of charged units. Both copolymers displayed low ionic conductivity, however. It was confirmed that neither contains any carrier ions that can support a long-distance migration. Li salts were added to P1 (P1 implies polymerized 1) so as to generate carrier ions, and their ionic conductivity was measured. Three distinct Li salts (LiTFSI, LiBF4, and LiCF3SO3) were used as additives. It is well known that these anions can form ILs at room temperature after mixing with imidazolium cations. Figure 31.2 shows the temperature dependence of the ionic conductivity for P1 containing an amount of Li salt equimolar to the imidazolium unit. When LiTFSI was added to P1, it had ionic conductivity of 7.2 × 10−7 S cm−1 at 50 °C. In contrast, the ionic conductivity of P1 containing the other Li salts was approximately 10−9 S cm−1, which is almost insulating. The ionic conductivity of P1 improved significantly upon adding LiTFSI even though other anion species have the ability to form ILs. This difference will be based on the plasticizing effect of the TFSI anion [5]. A similar effect has been observed in ZILs [6]. Unfortunately, their Tg has not previously been settled, although the DSC

60

50

Temperature / °C 40 30

20

10

Ionic conductivity log(σi / S cm-1)

-5

-6

-7

-8

-9

-10 3.0

3.1

3.2

3.3 -1

3.4

3.5

3.6

-1

1000 T /K

Figure 31.2. Temperature dependence of the ionic conductivity for P1 containing an equimolar amount of LiX. ○: X = TFSI; ▲: BF4; □: CF3SO3.

c31.indd 435

1/20/2011 10:29:07 AM

436

AMPHOTERIC POLYMERS Temperature / °C 60

50

40

30

20

10

Ionic conductivity log(σi / S cm-1)

-4

-5

-6

-7

-8

-9 3.0

3.1

3.2

3.3 -1

3.4

3.5

3.6

-1

1000 T /K

Figure 31.3. Temperature dependences of the ionic conductivities for P1 (○) and P2 (□) containing an equimolar amount of LiTFSI.

measurement has been carried out. The addition of the TFSI anion is effective in improving the ionic conductivity of IL copolymers. An amount of LiTFSI equimolar to the imidazolium cation unit was added to P1 and P2 in order to study the effect of an alkyl spacer on the ionic conductivity. Figure 31.3 shows Arrhenius plots of the ionic conductivity for the copolymers after addition of salt. Within the measured temperature regime, P2 with an alkyl spacer had ionic conductivity one order higher than P1 without a spacer. There are two possible explanations. An increase in the length of the alkyl spacer causes an increase of free volume and maintains the high mobility of the IL domain. Although the spacer unit is effective in improving the ionic conductivity of copolymers, a spacer of appropriate length should be used because it is difficult to form successive domains for ion conduction with a longer spacer. Proton-conducting copolymers of vinylphosphonic acid and 4-vinylimidazole have been reported recently by Bozkurt et al. [7]. Because the imidazole ring can act as a proton-hopping site in the polymer matrix [8–10], these copolymers had an ionic conductivity of approximately 10−6 S cm−1 at 60 °C without solvent or salt. To realize fast proton transport in copolymer systems, it is essential to design ion-conductive paths that use the IL domain.

c31.indd 436

1/20/2011 10:29:07 AM

437

POLY(ZWITTERIONIC LIQUID)S

31.2

POLY(ZWITTERIONIC LIQUID)S

ZIL monomers, having both cation and anion in the same monomer unit, were prepared to give different types of ampholyte polymers. Figure 31.4 shows the structure of the monomers used here [6]. Four different monomers were synthesized to study the relation between structure and such properties as ionic conductivity and Tg. Polymers having a ZIL unit should provide an excellent ion-conductive matrix for added target ions alone, as stated in Chapter 21. Accordingly, poly(ZIL)s were synthesized, and their ionic conductivity and thermal properties were studied. P6 showed the lowest Tg, both before and after addition of salt. This matrix proved to have higher ionic conductivity than the other polymers. Figure 31.5 shows the temperature dependence of the ionic conductivity for poly(ZIL)s containing equimolar amounts of LiTFSI. Two groups of poly(ZIL)s were synthesized, with an imidazolium cation and with a sulfonamide group on the main chain, to study the effect of freedom of the imidazolium cation. The ionic conductivity of polymers having an imidazolium cation on the main chain, P3 and P4, containing 100 mol% LiTFSI to imidazolium salt unit, was approximately 10−8−10−9 S cm−1. Poly(ZIL)s having an immobilized counter anion on the main chain (P5 and P6 containing LiTFSI) had relatively high ionic conductivity of approximately 10−5 S cm−1 at 50 °C despite their rubberlike properties. P6, with a long alkyl spacer (m = 3), showed slightly greater ionic conductivity. The difference in ionic conductivity can be attributed to the difference in freedom of the imidazolium cation. This effect has already been confirmed empirically in our laboratory [11,12]. For a simple system such as 1-ethyl-3vinylimidazolium TFSI, the ionic conductivity decreased by approximately

N + N

SO3 3 O

N + N

S

N

CF3

O 4 O N + N

O

S mO

N

5; m = 2 6; m = 3

Figure 31.4. Structures of zwitterionic monomers.

c31.indd 437

1/20/2011 10:29:07 AM

438

AMPHOTERIC POLYMERS

60

50

Temperature / °C 40 30

20

10

Ionic conductivity log(σi / S cm-1)

-4

-5

-6

-7

-8

-9

-10 3.0

3.1

3.2

3.3 -1

3.4

3.5

3.6

-1

1000 T /K

Figure 31.5. Arrhenius plots of the ionic conductivity for poly(ZIL)s containing LiTFSI. □: P3; ▼: P4; ○: P5; △: P6.

four orders upon polymerization. The ionic conductivity of IL polymer brushes having PEO or hydrocarbon chains as spacers between the vinyl group and imidazolium salt are almost the same before and after polymerization. Despite their rubberlike physical properties, IL polymer brushes had excellent ionic conductivity, around 10−4 S cm−1, at room temperature. The details of the IL polymer brushes are given in Chapter 32. The ionic conductivity of P3 and P4, in which the imidazolium cation is fixed to the main chain, was very low even after salts were added to the matrix. P5 and P6 having the counter anion on the main chain displayed an ionic conductivity four orders higher than P3 or P4 (about 10−5 S cm−1 at 50 °C). The distance between the vinyl polymer and the imidazolium cation is important for the high ionic conductivity in IL polymers. Poly(ZIL)s composed of aliphatic ammonium cation have also been reported by Galin et al. [13,14]. These authors studied the ionic conductivity and dipole moment of poly(ZIL)s. The dipole moment increased with increasing distance between the cation and the anion fixed on the polymer chain. This observation suggests that these poly(ZIL)s have different solubility from the added salts, which is reflected in their polarity. However, the ionic conductivities of P4 and P5 were very different, although the spacer distance between the cation and the anion was nearly identical. This result suggests that the

c31.indd 438

1/20/2011 10:29:07 AM

REFERENCES

439

distance between the vinyl polymer and the imidazolium cation is important in maintaining the high ionic conductivity of IL polymers. Although other factors muddy the comparison, the ionic conductivity of P6 is nevertheless 100 times higher than that of the ammonium systems prepared by Galin et al. We attribute this difference not only to the fixed position of the cation but also to the use of a suitable anion to form the ILs. Ampholyte-type polymers should become attractive materials once successive and effective ion-conductive paths are successfully designed in the matrix.

REFERENCES 1. J. R. MacCallum, C. A. Vincent, eds., in Polymer Electrolyte Reviews 1 and 2, Elsevier Applied Science, 1987 and 1989. 2. P. G. Bruce, ed., in Solid State Electrochemistry, Cambridge University Press, 1995. 3. R. Hooper, L. J. Lyons, M. K. Mapes, D. Schumacher, D. A. Moline, R. West, Macromolecules, 34, 931 (2001). 4. M. Yoshizawa, W. Ogihara, H. Ohno, Polymer Adv. Technol., 13, 589 (2002). 5. S. Besner, A. Vallée, G. Bouchard, J. Prud’homme, Macromolecules, 25, 6480 (1992). 6. M. Yoshizawa, M. Hirao, K. Ito-Akita, H. Ohno, J. Mater. Chem., 11, 1057 (2001). 7. A. Bozkurt, W. H. Meyer, J. Gutmann, G. Wegner, Solid State Ionics, 164, 169 (2003). 8. (a) M. A. B. H. Susan, A. Noda, S. Mitsushima, M. Watanabe, Chem. Comm., 938 (2003). (b) A. Noda, M. A. B. H. Susan, K. Kudo, S. Mitsushima, K. Hayamizu, M. Watanabe, J. Phys. Chem. B, 107, 4024 (2003). 9. M. Yoshizawa, W. Xu, C. A. Angell, J. Am. Chem. Soc., 125, 15411 (2003). 10. (a) H. G. Herz, K. D. Kreuer, J. Maier, G. Scharfenberger, M. F. H. Schuster, W. H. Meyer, Electrochim. Acta, 48, 2165 (2003). (b) K. D. Kreuer, A. Fuchs, M. Ise, M. Spaeth, J. Maier, Electrochim. Acta, 43, 1281 (1998). (c) M. Schuster, W. H. Meyer, G. Wegner, H. G. Herz, M. Ise, M. Schuster, K. D. Kreuer, J. Maier, Solid State Ionics, 145, 85 (2001). 11. H. Ohno, K. Ito, Chem. Lett., 27, 751 (1998). 12. M. Yoshizawa, H. Ohno, Electrochim. Acta, 46, 1723 (2001). 13. M. Galin, A. Mathis, J.-C. Galin, Polymer Adv. Technol., 12, 574 (2001). 14. M. Galin, E. Marchal, A. Mathis, J.-C. Galin, Polymer Adv. Technol., 8, 75 (1997).

c31.indd 439

1/20/2011 10:29:07 AM

32 POLYMER BRUSHES Masahiro Yoshizawa-Fujita and Hiroyuki Ohno

When ionic liquids (ILs) are applied as electrolyte materials, the major mobile ions are the component ions, and therefore, any specific target ion transport is impaired. Target ion transport in ILs has already been discussed in Chapters 21 and 31 [1–8]. Polymerized derivatives of ILs are required for applications in ionics devices [9]. When an IL unit is polymerized, ion transport becomes selective [10]. In the polymer matrix, since the fixed cation or anion cannot move along with the potential gradient, only the counter ion can move. Therefore, IL polymers are expected to be novel single ion conductive matrices. Still, as mentioned in Chapter 31, the ionic conductivity values of ILs composed of vinylimidazolium cation decrease considerably after polymerization [10]. It is, therefore, necessary to design new IL polymers to maintain a low Tg. Certain IL-type polymer brushes having a flexible spacer between polymerizable group and the IL moiety have been developed to minimize this drawback [11–16]. Figure 32.1 shows the structure of IL monomers having a flexible spacer. IL monomers having an ethylene oxide (EO) or a hydrocarbon (HC) spacer between the vinyl group and the imidazolium cation were synthesized, and the effects of the structure and the length of the spacer groups on both the ionic conductivity and the thermal properties were investigated. First, the effect of the EO spacer on the physical properties for IL-type polymer brushes is discussed. Both IL monomers, a8Cl and a8TFSI, are liquids at room temperature and have very low Tg’s at −74 °C and −72 °C, respectively. These monomers showed very high ionic conductivity at room temperature,

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

441

c32.indd 441

1/20/2011 10:29:08 AM

442

POLYMER BRUSHES (a)

A = Cl or TFSI

O

N

O

N

m-1

O amA; m = 2 or 8 A = Br or TFSI

O

N

N

R1

n

O

R2

bnA; R1 = Me, R2 = H, n = 6 cnA; R1 = Et, R2 = H, n = 2, 3, 6, 9, or 12 dnA; R1 = Bu, R2 = H, n = 6 enA; R1 = Et, R2 = Me, n = 6 TFSI

O

TFSI

N

O

N

6

6

O

O f6TFSI

g6TFSI TFSI

O

R3

N 6

O h6TFSI; R3 = Me i6TFSI; R3 = Et (b) Conductive path Flexible spacer Vinyl polymer

Figure 32.1. (a) Structures of IL monomers having a flexible spacer between vinyl group and imidazolium cation. (b) Model structure of IL polymer brushes.

c32.indd 442

1/20/2011 10:29:09 AM

443

POLYMER BRUSHES Temperature / °C 60

50

40

30

20

10

Ionic c onductivit y log(σi / S cm-1)

-2

-3

-4

-5

-6

-7 3.0

3.1

3.2 1000 T

3.3 -1

/K

3.4

3.5

3.6

-1

Figure 32.2. Arrhenius plots of the ionic conductivity for IL monomers having EO spacer a8A and their polymers. Open plots: monomers; closed plots: polymers. ○,•: TFSI−; △,▲: Cl−.

as shown in Figure 32.2. The ionic conductivity of a8TFSI was 5.51 × 10−4 S cm−1 at 30 °C, which was 20 times higher than that with Cl− as counter anion. This improvement can be attributed to the IL formation by the combination of the imidazolium cation and TFSI anion and to a much weaker interaction force of ether oxygens toward TFSI− than that to Cl−. Since the ion radius of TFSI− (325 pm) [17] is larger than that of Cl− (181 pm), the delocalized surface charge density of TFSI− was much smaller, permitting the weaker interaction. Then the ionic conductivity for these samples was measured after polymerization. The IL polymer brushes having EO spacers are rubber-like solids regardless of the anion species. The Tg’s of these polymers P(a8Cl) and P(a8TFSI) (P implies polymer) are −62 °C and −64 °C, respectively. From comparing the Tg’s for both monomers and polymers, we confirmed that the Tg remained low even after polymerization. Despite their rubber-like properties, these IL polymer brushes showed high ionic conductivity, reflecting their low Tg. That is, the flexibility of the side-chain containing the imidazolium salts was effective in maintaining the high ionic conductivity. In particular, when the counter anion species of the polymer was TFSI−, excellent ionic conductivity (1.20 × 10−4 S cm−1 at 30 °C) was observed. The EO-tethering of the IL and the

c32.indd 443

1/20/2011 10:29:09 AM

444

POLYMER BRUSHES -2.0

log(σi / S cm-1) at 30 °C

-3.0 -40 -3.5 -4.0

-60

-4.5 -80

-5.0 -5.5

2

4

6

8

10

Glass transition temperature / °C

-20

-2.5

12

Carbon number of spacer / n

Figure 32.3. Effects of HC spacer length (n) on the ionic conductivities and Tg’s for IL monomers cnTFSI.

polymer matrix was effective in suppressing the decrease of ionic conductivity after polymerization. Figure 32.3 shows the effect of HC spacer length (n) on the ionic conductivity and Tg for cnTFSI. Note that all monomers show almost the same Tg between −80 °C and −70 °C. There is a good relationship between the ionic conductivity and the Tg’s of ordinary ILs [18]. In other words, ILs having lower Tg’s show higher ionic conductivity. The ionic conductivities of these IL monomers decreased monotonically with the increasing spacer chain length. No increase in the ionic conductivity could be detected with the drop in Tg. This may be caused by the decrease of ion density and the increase of viscosity with the increasing HC length. Figure 32.4 shows the relationship between the carbon number (n) of the HC spacer and the ionic conductivity or Tg for P(cnTFSI). The ionic conductivity of poly(1-ethyl-3-vinylimidazolium TFSI) (P(EVImTFSI)) is also depicted as a reference, with n = 0 for comparison [10a]. P(EVImTFSI) was obtained as a glassy solid at room temperature. The ionic conductivity decreased more than four orders, compared with the system before polymerization. However, the P(cnTFSI)s were sticky and rubbery, and their ionic conductivity was almost the same before and after polymerization. The decrease of ionic conductivity caused by polymerization was minimized in these cases to within about one order. The Tg of P(cnTFSI)s was almost the same value, around −60 °C except for n = 2 (−47.8 °C). Nearly equal ionic conductivity values of a series of P(cnTFSI) can be explained by the excellent

c32.indd 444

1/20/2011 10:29:09 AM

445

POLYMER BRUSHES

-20

-4 -40 -5

-60

-6

-7

Glass transition temperature / °C

log(σi / S cm-1) at 30 °C

-3

-80 0

2

4

6

8

10

12

Carbon number of spacer / n

Figure 32.4. Effects of HC spacer length (n) on both the ionic conductivities (▲: the data of P(EVImTFSI) was also shown as reference) and the Tg’s for IL polymer brushes P(cnTFSI).

Tg values. It was also confirmed that the HC spacer is as effective in maintaining high ionic conductivity, even after polymerization, as the effect of the EO spacer between the vinyl group and the imidazolium cation [13]. The IL monomers tethering different cation structures on the spacer end were synthesized to study the effect of the cation structure on the ionic conductivity. Figure 32.5 shows the Arrhenius plots of the ionic conductivity for P(x6TFSI). P(c6TFSI) having the ethylimidazolium cation showed the highest ionic conductivity among these polymer brushes. Its ionic conductivity value was 1.4 × 10−4 S cm−1 at 30 °C, which is one order higher than those of P(e6TFSI) having the 1-ethyl-2-methylimidazolium cation and P(h6TFSI) having the 1-methylpiperidinium cation. It appears that the ethylimidazolium cation is suitable for higher ionic conductivity for the IL polymer brushes as well as for simple ILs [19]. The Tg’s of P(x6TFSI) are −53 °C (x = b), −59 °C (x = c), −51 °C (x = d), −42 °C (x = e), −43 °C (x = f), −40 °C (x = g), −38 °C (x = h), and −55 °C (x = i). The system with lower Tg’s showed the higher ionic conductivity. In addition a thermogravimetric analysis of these polymers was carried out. Their thermal stability was evaluated in the following sequence of the onset decomposition temperature (Td): P(e6TFSI) (T = 389 °C) > P(d6TFSI) (T = 382 °C) > P(c6TFSI) (T = 381 °C) > P(g6TFSI) (T = 372 °C) > P(b6TFSI) (T = 371 °C) = P(f6TFSI) (T = 371 °C) > P(i6TFSI) (T = 364 °C) > P(h6TFSI) (T = 362 °C). From these results, the IL-type polymer brushes were

c32.indd 445

1/20/2011 10:29:09 AM

446

POLYMER BRUSHES -3.0

-3.5

log(σi / S cm-1)

-4.0

-4.5

-5.0

-5.5

-6.0

-6.5 3.0

3.1

3.2

3.3

3.4

3.5

3.6

Temperature / °C

Figure 32.5. Arrhenius plots of the ionic conductivity for various polycation-type ILs P(x6TFSI). ○: x = c; △: x = d; □: x = b;◇: x = e.

determined to be thermally stable up to around 400 °C. This is an excellent property for application to electrochemical devices. Although P(c6TFSI), which showed the highest ionic conductivity among all systems of P(x6TFSI), had an ionic conductivity of more than 10−4 S cm−1 at 30 °C, the sticky characteristics of this polymer made it difficult to handle. To surmount this problem, we polymerized c6TFSI with various cross-linkers, as shown in Figure 32.6. First, the CLa3 was used as a cross-linker so that the effects of various amounts of the cross-linker on their properties could be examined. The sticky rubbery property of IL polymers was improved by the addition of the cross-linkers. Although they were sticky rubbers without crosslinkers, they formed flexible films after the addition of the cross-linker. In addition they could be obtained as films by adding just a small amount of the cross-linker. Figure 32.7 shows a typical photograph of a flexible film prepared by the polymerization of c6TFSI with 3 mol% CLa3 as the cross-linker. It was confirmed a small amount of cross-linker of approximately 1 mol% was enough to obtain the IL polymers as flexible films. Figure 32.8 shows the effect of the cross-linker (CLa3) concentration on the ionic conductivity and the Tg for the IL-type network polymer P(c6TFSI– CLa3). After cross-linking, the Tg of the network was increased approximately 5 °C, compared with that of the linear polymer, and the ionic conductivity decreased by approximately one third. Within the range of the amount of

c32.indd 446

1/20/2011 10:29:09 AM

447

POLYMER BRUSHES

O

O

n

CLa n; n = 2 or 3 O O

O

n

O CLbn; n = 4 or 13-14 O O

O

n

O CLcn; n = 2, 4, or 9 TFSI

O O

N

TFSI

N

O N

N

O

n

CLdn; n = 4 or 6

TFSI

N

TFSI

N

N

N

n

CLen; n = 4 or 6

Figure 32.6. Structures of various cross-linkers used for obtaining film electrolytes.

cross-linker from 0.5 to 5 mol%, the ionic conductivity was almost the same. Generally, cross-linking restricts the motion of the polymer backbone. Therefore, higher Tg should be found with increasing degrees of cross-linking. The ionic conductivities of network polymer electrolytes are usually much lower than those of the corresponding linear polymers [20]. There is a report that ionic conductivity remained almost the same as that for the linear polymers when monomers containing the EO spacer were polymerized with the cross-linkers (up to 5 mol%) [21]. This is caused by the successive ion conduction in the domain even after cross-linking despite a remarkable increase in viscosity. Because the IL structure is tethered to the end of the flexible spacer,

c32.indd 447

1/20/2011 10:29:09 AM

448

POLYMER BRUSHES

Figure 32.7. Photograph of film electrolyte composed of polymerized IL (cross-linker CLa3 3 mol% to c6TFSI).

-30 1.4

-40

1.0 0.8 0.6

-50 0.4 0.2 0.0

0

2 4 6 8 10 Cross-linker CLf3 / mol% to c6TFSI unit

Glass transition temperature / °C

σi x 104 / S cm-1 at 30 °C

1.2

-60 12

Figure 32.8. Effects of cross-linker concentration CLa3 on the ionic conductivity and Tg of the IL-type network polymer P(c6TFSI-CLa3).

c32.indd 448

1/20/2011 10:29:09 AM

POLYMER BRUSHES

449

almost no change in Tg is found after polymerization in the presence of small amount of cross-linker. This seems to be the mechanism that maintains the high ionic conductivity. Using various EO cross-linkers (CLan, CLbn, and CLcn), we next synthesized network polymers to investigate the effect of the spacer length and the polymerizable group of EO cross-linkers on ionic conductivity. The added amount was fixed at 1.3 mol% because the addition of approximately 1 mol% CLa3 has been confirmed to be enough to obtain these polymers as films. Although all the obtained network systems showed almost the same Tg’s around −55 °C to −60 °C, there appeared some differences in the ionic conductivity, as shown in Figure 32.9. In the CLa3 system, the network exhibited higher ionic conductivity than the network of CLa2, and in the CLbn system, the network of CLb4 exhibited higher ionic conductivity than the network of CLb13–14. There was a suitable length in the spacer of the cross-linker to maintain the high ionic conductivity despite the small amount of cross-linkers. Besides the CLcn system, the temperature dependence of the ionic conductivity was the same regardless of the spacer’s length. These results indicate that an appropriate spacer length must be found for every cross-linker. When CLa3 was used as the cross-linker, a minimum amount of 1 mol% was necessary to obtain a film. However, the minimum amount depended on the structure of the cross-linker. Among these, CLbn, CLc4, and CLc9 were effective in forming films after the addition of only 0.5 mol%. In particular, CLb4 was the best and induced the highest ionic conductivity. The ionic conductivity of the transparent and flexible film obtained by adding 0.5 mol% of CLb4 was 1.1 × 10−4 S cm−1 at 30 °C, which is almost equivalent to that of the monomer. Using various EO cross-linkers is a useful methodology to achieve film forming ability for IL-type polymer brushes. However, those films are brittle. In order to further improve both thermal stability and mechanical strength, researchers synthesized new cross-linking monomers containing an IL moiety [22]. Monomer c2TFSI was polymerized in the presence of the cross-linking monomer, CLdn or CLen. The Td of the cross-linked polymers was above 400 °C, which was higher than those of the IL-type network polymers with EO cross-linkers. This was attributable to the IL moiety with TFSI anion. There is no fear of thermal degradation starting from the cross-linkers. Copolymerization of c2TFSI and a cross-linker (CLd4, 1 mol%) led to a solid polymer electrolyte showing a high ionic conductivity: 1.36 × 10−4 S cm−1 at 50 °C. The ionic conductivity decreased with increasing CLd4 content, as was true for the EO crosslinkers. The mechanical strength of the films with IL moiety was slightly improved as compared with those of the films with EO cross-linkers. To investigate the effect of the flexible spacer on the physical properties of the polyanion-type ILs, we synthesized polyanion-type ILs having a hydrocarbon or an EO spacer. Figure 32.10 shows the structure of the IL monomers having a flexible spacer between a polymerizable group and an anion. The effect of the HC spacer on the ionic conductivity of the polyanion-type ILs was analyzed. Figure 32.11 shows the Arrhenius plots of the ionic conductivity

c32.indd 449

1/20/2011 10:29:09 AM

(a)

-1

Ionic conductivity log(σi / S cm )

-3.0

-3.5

-4.0

-4.5

-5.0

-5.5 3.0

3.1

3.2

3.3 -1

3.4

3.5

3.6

3.4

3.5

3.6

3.4

3.5

3.6

-1

1000 T / K

(b)

Ionic conductivity log(σi / S cm-1)

-3.0

-3.5

-4.0

-4.5

-5.0

-5.5 3.0

3.1

3.2

3.3 -1

-1

1000 T / K

(c)

-1

Ionic conductivity log(σi / S cm )

-3.0

-3.5

-4.0

-4.5

-5.0

-5.5 3.0

3.1

3.2

3.3 -1

-1

1000 T / K

Figure 32.9. Arrhenius plots of the ionic conductivity for IL-type network polymers. (a) P(c6TFSI-CLan); (b): P(c6TFSI-CLbn); (c) P(c6TFSI-CLcn).

c32.indd 450

1/20/2011 10:29:09 AM

451

POLYMER BRUSHES

O

SO3

HN

N

O EImC3S

R2 O

O

O HN

n-1

O

R2

O

N

SO3

R1

EImEOnBS; R1 = Et, R2 = H, n = 2 or 8 MImEOnBS; R1 =Me, R2 = H, n = 6 EImPOnBS; R1 = Et, R2 = Me, n = 9 N

O O

N H

O

N H

O

O

N

2

O

N CF3

S

O

EMImUEO3SA

Figure 32.10. Structures of IL monomers having a flexible spacer between a polymerizable group and an anion structure.

Ionic conductivity log(σ i / S cm-1)

-1

-2

-3

-4

-5

3.0

3.1

3.2

3.3 -1

1000 T / K

3.4

3.5

3.6

-1

Figure 32.11. Temperature dependences of the ionic conductivities for EImVS, EImC3S, and their polymers. ○: EImVS; •: P(EImVS); △: EImC3S; ▲: P(EImC3S).

c32.indd 451

1/20/2011 10:29:09 AM

452

POLYMER BRUSHES

for N-ethylimidazolium vinyl sulfonate (EImVS), EImC3S, and their polymers. The ionic conductivity of EImC3S was approximately 10−3 S cm−1 at room temperature, which is one order lower than that of EImVS. This difference could be a result of the large-size anion of EImC3S. After polymerization, both systems showed an ionic conductivity of approximately 10−4 S cm−1 at room temperature. Although the alkyl spacer group was introduced between the vinyl group and the sulfonate, no suppression of the drop in ionic conductivity was observed. The flexibility of the imidazolium cation is inherently high in the polyanion-type IL containing sulfonate. Higher ionic conductivity of the polymerized ILs was expected when the motion of imidazolium cation was coupled with that of anion containing the tethered vinyl group. For this reason, some IL monomers were prepared, as shown in Figure 32.10. Figure 32.12 shows the effect of the spacer length on ionic conductivity and the Tg’s for EImEO2BS, EImEO8BS, EImPO9BS, and their polymers. N-ethylimidazolium p-styrenesulfonate (EImSS) was used as the reference with a spacer length of 0. The ionic conductivity of EImSS was approximately 10−4 S cm−1 at room temperature. However, its polymer showed a low ionic conductivity of approximately 10−8 S cm−1. As mentioned, the lowering of ionic conductivity after polymerization seems to reflect the increase of Tg. On the other hand, polyanion-type ILs having an EO or a PO spacer showed ionic conductivity of more than 10−6 S cm−1 at room temperature. When the repeating unit number of the spacer was 2, suppression of the ionic conductivity was diminished. In particular, the ionic conductivity of the PO system was approximately 10−5 S cm−1, which is nearly equal to that of the corresponding monomer. This effect should be caused by the interaction between the imidazolium cation and the PO units being weaker than that with the EO units. The ionic conductivity of polyanion-type ILs containing benzene sulfonate was improved by the introduction of the polyether spacer, whereas their Tg’s increased with the increasing spacer unit number. Generally, the increase of Tg causes a considerable drop in ionic conductivity [9]. Although the Tg’s of the polyanion-type ILs having the polyether spacer increased with the increasing spacer unit number, they maintained ionic conductivities of more than 10−6 S cm−1 at room temperature. These Tg values are based on the polyether segment and not on the ionic liquid part of the spacer end. We suggest that the polyether spacer does not work as an ion conductive path; that is, the transport of imidazolium cation does not depend on the mobility of polyether segment but on the flexibility of the charged site. So the PO unit is a favorable spacer for keeping the ionic conductivity of the polymerized ILs high. Interpenetrating polymer networks (IPNs) based on polybutadiene (PB) and a polyanion-type IL, MImEO6BS, were prepared by Vidal et al. [23]. The polyanion-type IL network was obtained from the copolymerization of MImEO6BS and an EO cross-linker, whereas a PB network was formed by an isocyanate-alcohol addition between the hydroxyl end groups of PB and an isocyanate cross-linker. All polymerizable precursors were first dissolved in a DMSO/toluene mixture, and then the polymerization reactions occurred

c32.indd 452

1/20/2011 10:29:10 AM

453

POLYMER BRUSHES

-4

-1

o

log(σi / S cm ) at 30 C

-5

-6

-7

-8

o

Glass transition temperature / C

-30

-40

-50

-60

-70

-80 0

2

4

6

8

10

EO or PO unit number

Figure 32.12. Effects of spacer length and structure on the ionic conductivities and Tg’s for IL monomers and polymers containing polyether spacers. Structures are shown in Figure 32.10. ○, •: EO spacer; □, ■: PO spacer. Open plots: monomers; closed plots: polymers.

leading to an IPN. The tan δ of the IPN containing 72 wt% IL network showed a single relaxation at 11 °C in a DMA measurement. This will be an interesting methodology for improving mechanical properties of IL-type polymer brushes. A new polyanion-type IL composed of trifluoromethylsulfonylamide anion was synthesized as shown in Figure 32.10 [24]. The Tg value of EMImUEO3SA was −67 °C, which was lower than that of EImEO8BS.The Tg of P(EMImUEO3SA) remained at a low level of −46 °C because of the presence of trifluoromethylsulfonyl (TFS) groups on the pendant chain end and a flexible EO spacer. The

c32.indd 453

1/20/2011 10:29:10 AM

454

POLYMER BRUSHES

TFS groups should work as a plasticizer [25] and should induce the charge delocalization by the electron-withdrawing effect [26]. Ionic conductivity measurements were performed for EMImUEO3SA and its polymer. EMImUEO3SA showed the ionic conductivity value of 2.1 × 10−3 S cm−1 at 20 °C, whereas the ionic conductivity of P(EMImUEO3SA) was 6.5 × 10−4 S cm−1 at 20 °C, clearly showing that the IL-type polymer retained a high ionic conductivity after the polymerization. The introduction of highly dissociable ion structures will be effective to develop highly conductive matrices. Recently, novel IL-type polymer brushes have been reported that are derived from conducting polymers such as poly(fluorene) (1) [27], poly(thiophene) (2) [28], poly(p-phenylene) (3) [29], poly(3,4ethylenedioxythiophene) (4) [30], and poly(pyrrole) (5) [31] as shown in Figure 32.13. Polymer 3 with bromide anions was soluble in water and showed a strong UV absorption at 332 nm in a dilute aqueous solution. The maximum PL emission of 3 was observed at 404 nm when excited at 370 nm, indicating that the polymer will be an efficient blue fluorescent material. Polymers 4 and 5 were synthesized by electropolymerization on electrodes. The electropolymerization of 4 with various anions was carried out by cyclic voltammetry leading to a blue-colored film on a working electrode. The polymer films on an electrode exhibited the corresponding color change from dark blue to pale blue between −0.2 and 1 V. Interestingly, the hydrophobicity of these films could be modified depending on the anion species.

X N

N X

n N 8

N

O

1 S

N N X

n

n O

2

N

N O

O

S

4

N

X

3

X N n

N

n N m

n

5

X

PF6

O

N

O 3

O

N

N

6

Figure 32.13. Structures of IL-type polymers derived from conducting polymers and poly(norbornene).

c32.indd 454

1/20/2011 10:29:10 AM

REFERENCES

455

Zheng et al. synthesized poly(norbornene)s with imidazolium salts— functionalized EO pendant groups via ruthenium-catalyzed ring-opening metathesis polymerization [32]. Polymer 6 showed a relatively low Tg value of −30 °C. The flexible EO segment gave a low Tg for IL-type polymer brushes regardless of their main chain structures. In 2003, a new composite material based on the IL-type polymer brush and the single-walled carbon nanotube (SWNT) was reported by Aida et al. [33]. The film of the IL polymer brush including SWNT (3.8 wt%) showed a nearly 400% increase in dynamic hardness compared with that of the pure IL polymer film. It is thought that the good dynamic hardness was obtained because of a strong SWNT–imidazolium cation interaction. In addition, the film showed a conductivity of 0.56 S cm−1 at room temperature, which is equal to that of typical conductive polymers whose films are being engineered as coating materials, antistatic materials, and novel electronic devices [34–37].

REFERENCES 1. (a) M. Yoshizawa, M. Hirao, K. Ito-Akita, H. Ohno, J. Mater. Chem., 11, 1057 (2001). (b) M. Yoshizawa, A. Narita, H. Ohno, Aust. J. Chem., 57, 139 (2004). 2. (a) M. Yoshizawa, H. Ohno, Chem. Comm., 1828 (2004). (b) T. Mizumo, T. Watanabe, H. Ohno, Polymer J., 40, 1099 (2008). 3. A. Narita, W. Shibayama, H. Ohno, J. Mater. Chem., 16, 1475 (2006). 4. A. Narita, W. Shibayama, K. Sakamoto, T. Mizumo, N. Matsumi, H. Ohno, Chem. Comm., 1926 (2006). 5. M. Yoshizawa, W. Ogihara, H. Ohno, Polymer Adv. Technol., 13, 589 (2002). 6. W. Ogihara, N. Suzuki, N. Nakamura, H. Ohno, Polymer J., 38, 117 (2006). 7. A. Narita, W. Shibayama, M. Tamada, H. Ohno, Polymer Bull., 57, 115 (2006). 8. A. Narita, W. Shibayama, N. Matsumi, H. Ohno, Polymer. Bull., 57, 109 (2006). 9. J. R. MacCallum, C. A. Vincent, eds., Polymer Electrolyte Reviews, Elsevier Applied Science., vol. 1, 1987, and vol. 2, 1989. 10. (a) H. Ohno, K. Ito, Chem. Lett., 27, 751 (1998). (b) M. Hirao, K. Ito, H. Ohno, Electrochim. Acta, 45, 1291 (2000). (c) M. Hirao, K. Ito-Akita, H. Ohno, Polymer Adv. Technol., 11, 534 (2000). 11. M. Yoshizawa, H. Ohno, Chem. Lett., 28, 889 (1999). 12. H. Ohno, Electrochim. Acta, 46, 1407 (2001). 13. M. Yoshizawa, H. Ohno, Electrochim. Acta, 46, 1723 (2001). 14. S. Washiro, M. Yoshizawa, H. Nakajima, H. Ohno, Polymer, 45, 1577 (2004). 15. H. Ohno, M. Yoshizawa, W. Ogihara, Electrochim. Acta, 50, 255 (2004). 16. W. Ogihara, S. Washiro, H. Nakajima, H. Ohno, Electrochim. Acta, 51, 2614 (2006). 17. M. Ue, J. Electrochem. Soc., 141, 3336 (1994). 18. M. Hirao, H. Sugimoto, H. Ohno, J. Electrochem. Soc., 147, 4168 (2000). 19. H. Ohno, M. Yoshizawa, Solid State Ionics, 154–155, 303 (2002).

c32.indd 455

1/20/2011 10:29:10 AM

456

POLYMER BRUSHES

20. (a) W. H. Meyer, Adv. Mater., 10, 439 (1998). (b) M. Watanabe, T. Hirakimoto, S. Mutoh, A. Nishimoto, Solid State Ionics, 148, 399 (2002). 21. J. M. G. Cowie, K. Sadaghianizadeh, Polymer, 30, 509 (1989). 22. H. Nakajima, H. Ohno, Polymer, 46, 11499 (2005). 23. F. Vidal, J. Juger, C. Chevrot, D. Teyssié, Polymer Bull., 57, 473 (2006). 24. J. Juger, F. Meyer, F. Vidal, C. Chevrot, D. Teyssié, Tetrahedron Lett., 50, 128 (2009). 25. S. Besner, A. Vallée, G. Bouchard, J. Prud’homme, Macromolecules, 25, 6480 (1992). 26. P. Bonhôte, A.-P. Dias, M. Armand, N. Papageorgiou, K. Kalyanasundaram, M. Grätzel, Inorg. Chem., 35, 1168 (1996). 27. D. Zhang, L. Chen, J. Chen, Y. Liang, H. Zhou, Synth. Commun., 35, 2609 (2005). 28. C. T. Burns, S. Lee, S. Seifert, M. A. Firestone, Polymer Adv. Technol., 19, 1369 (2008). 29. H. Wan, D. Zhang, L. Chen, J. Chen, H. Zhou, J. Disper. Sci. Technol., 29, 1158 (2008). 30. M. Döbbelin, C. Pozo-Gonzalo, R. Marcilla, R. Blanco, J. L. Segura, J. A. Pomposo, D. Mecerreyes, J. Polymer Sci. Polymer Chem., 47, 3010 (2009). 31. W. Zhang, Y. Li, C. Lin, Q. An, C. Tao, Y. Gao, G. Li, J. Polymer Sci. Polymer Chem., 46, 4151 (2008). 32. L. Zheng, F. Chen, M. Xie, H. Han, Q. Dai, Y. Zhang, C. Song, Reactive Func. Polym., 67, 19 (2007). 33. T. Fukushima, A. Kosaka, Y. Ishimura, T. Yamamoto, T. Takigawa, N. Ishii, T. Aida, Science, 300, 2072 (2003). 34. X. Pei, Y. Xia, W. Liu, B. Yu, J. Hao, J. Polymer Sci. Polymer Chem., 46, 7225 (2008). 35. B. Wu, D. Hu, Y. Kuang, B. Liu, X. Zhang, J. Chen, Angew. Chem. Int. Ed., 48, 4751 (2009). 36. X. Chu, B. Wu, C. Xiao, X. Zhang, J. Chen, Electrochim. Acta, 55, 2848 (2010). 37. C. Xiao, X. Chu, B. Wu, H. Pang, X. Zhang, J. Chen, Talanta, 80, 1719 (2010).

c32.indd 456

1/20/2011 10:29:10 AM

PART VIII CONCLUSION

c33.indd 457

1/20/2011 10:29:11 AM

33 FUTURE PROSPECTS Hiroyuki Ohno

Ionic liquids are a repository of a wealth of creative possibility. In the second edition of this book, we did our best to introduce the latest electrochemical trends using ionic liquids. We expect the future of ionic liquids to exceed our anticipated applications. The variety of ions is endless. One gets unlimited species of ions when organic compounds are ionized because there are millions of organic molecules. More species of ions can be prepared by every molecule, and therefore in principle, combinations of cation and anion can provide more than trillions of salts. There should be many ionic liquids in the group. Furthermore, there are endless combinations of mixed salts. Their combinations can be designed on demand. We see at least two paths of development in the future. Already novel ionic liquids are being vigorously developed by scientists, but the next step should be the fine-tuning of their functionalities, by mixing two or more ionic liquids just like wines, colors, and/or other cultural materials. Because blending of ILs can generate unexpected functions and abilities, an evolution of ionic liquids can be envisioned. Some functional groups could take the form, for example, of advanced deodorant liquids that absorb smells and decompose them into odorless molecules will be synthesized. Liquidized metal complexes should be effective for this kind of task. Surface modifications with some ionic liquids may provide maintenance-free surfaces. Some equipments will always be clean and smooth because surface-modified ionic liquids are harnessed to decompose hazardous molecules. Such a system may be even applicable to domestic life. Because catalytic activities are supported electrochemically, the

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

459

c33.indd 459

1/20/2011 10:29:11 AM

460

FUTURE PROSPECTS

TABLE 33.1. Some Expected Functions and Applications of Present and Future Ionic Liquids Functions Ion conduction, Electro-conduction, Thermal conduction, Magnetism, Emission, Luminosity, Refractive-index control, Polarity, Solubility, Chirality, Lubrication, Degradability, Degradation switching, Edibility, Sterilization, Surface protection Features Liquid, Liquid crystal, Liquid (micro or nano) particle, Liquid sheet, Gel, Supramolecular assembly, Gel, Solid, Metal complex, Polymer, Film, Nano-sheet, Membrane, Rubber, Composite Devices Lithium battery, Lithium ion battery, Fuel cell, Solar cell, Capacitor, Super capacitor, Actuator, Sensory system , Nano-switch, Biochip, Catalyst, Optical switch, Liquid switch, Liquid brake, Special Solvents (for separation, extraction, isolation, etc.), Biocompatible surface, Cell cultivation sheet, Organ storage solution, Preserver, Drug carrier, Column, Biodegradable materials.

electrochemical characteristics of ionic liquids will continue to find meaningful applications. Solvents that possess catalytic activity are intelligent solvents. They show physicochemical properties such as polarity, basicity, acidity, solution temperature, and even reaction step by a color change. A liquid brake in which the viscosity can be controlled by a given potential may be prepared soon. A conductive liquid can transport not only ions but also electrons to supply organic solder, information display liquid, microchips, and so on. A liquid battery can be prepared by coating three layers of electrodes and electrolyte. Unlimited applications can be imagined with functionalized ionic liquids for human consumption such as finely blended alcoholic drinks. Some expected features of future ionic liquids are summarized by keywords in Table 33.1. Not only liquid systems but also a system with polymerized features will become attractive and be developed as useful sheets. The combination of polymers and ionic liquids could become a key system for the expanding technologies. Some applications could require the electrochemical regulation of their functions by other ionic liquids. One remarkable progress in these few years is the application of ionic liquids for bioscience and biotechnology fields. Some examples for the bioelectrochemical application of ionic liquids have been introduced in this book, (see Part II). In the next decade, application of ionic liquids on the bioscience fields should increase considerably. Bio-related electrochemistry should also be improved by the contribution of ionic liquids. Together with the accumulated knowledge and data on basic properties of ionic liquids such as toxicity and biodegradability, there will be more contribution of ionic liquids in the bio-related electrochemistry.

c33.indd 460

1/20/2011 10:29:11 AM

FUTURE PROSPECTS

461

With advances in the science of ionic liquids, there are bound to be some excellent treatments for ionic liquid decomposition. Ionic liquids could be developed to have electrolytes that carry a decomposition switch. Both developments would make ionic liquids more “green” and useful electrolyte materials. The recovering process, recycling process, and other surrounding technologies should be developed along with the progress made in the functional design of ionic liquids. As discussed in Chapter 12, cellulose is now soluble in certain ionic liquids, and they should be treated as energy source for some different fields. These are only a handful of suggestions on the prospects for ionic liquids discussed in this book. Readers may be inspired to come up with their own ideas for the development of ionic liquids based on their unique backgrounds. Finally I would like to add very unique ionic liquid as the key material for the development of this field; [B+][NO−]. This should support our researches on ionic liquids. The [B+][NO−] has been explained at the closing remarks of the Second International Symposium on Ionic Liquids in 2007. You may understand what this is. If not, please contact me.

c33.indd 461

1/20/2011 10:29:11 AM

APPENDIX: STRUCTURES OF ZWITTERIONS

N

N

N

N

CN

CN

O

17

1

O

O

SO3

SO3

N

O

CN

Br

N

N

SO3

SO3

O

O

N

CN

O

18

O N

O

CN 44

31

CN N

N

N

SO3 3

CN

O

CN

2

N

CN O

43

30

N

O

O

CN

CN

N

O

O

CN

O

SO3

19

O

O 45

32

O N

N

N

N

SO3 O

O

O

O

N

20

N

CN

O

O

CN

33

4

O

CN

O

CN 46

O N

N N

SO3

O

N

O

CN O

N

N

N

22

CN O

47

CN

O O

O

O

34

SO3 9

O

21

8

N

CN

CN

O N

O CN

O CN 35

N

O

O

O

CN O

48

(Continued) Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

463

bapp.indd 463

1/20/2011 10:26:33 AM

464

APPENDIX: STRUCTURES OF ZWITTERIONS O N

N N

CN

O

SO3

N

23

O

36

SO3

N

N

CN O

10

N

N

O

O

N

O

24

11

CN

O CN 37

N

N

O

SO3

O

N

N

N

O

12 25

N

N

CN N

O

O

O

N

SO3 14

N

N

N

CN

O

N

O

O

CN

N

N

SO3

16

bapp.indd 464

CN

O CN 29

CN

41

O N

CN

O

O 28

N

CN

40

CN N

15

CN

O

8 CN

27

SO3

O

O

N

N

CN O

39

26

N

CN

38

N

N

SO3 13

CN

O

O

O O

N 42

CN

O CN

1/20/2011 10:26:33 AM

INDEX

A/ANMR ratio, 82–83 Ab initio molecular orbital theory, 249 Absorption, 96, 165–166, 172, 193–194, 196 AC/IL systems, 260 Acceptors, 6, 27, 172, 198, 337–339 Acetates, 7, 105–107, 113, 186 Acetoamidation, 105 Acetonitrile, 125 ACF/IL/ACF system, 260 Acidity, 24–27 Acids, 24–25, 420, 426. See also specific types of acids Acrylic acid, 426 Activation energy, 341, 366 Actuators, 115, 271, 376, 460. See also specific types of actuators ADA Technologies, 260 Additives, 113, 133, 212, 218, 359, 361 Adenines, 410–411 Aggregation, 6, 81 Agrochemicals, 105 Alcaligenes bronchisepticus, 160 Alcohols electrolytic reactions, 124 ethyl, 294 optically active, 159–160 secondary, 160

Alimates, 331 Aliphatics, 60, 212, 259 Alkali halides, 7, 9 Alkali metal, 37 Alkali metal ionic liquids (AMILs) characteristics of, 317–320 conductivity, 321–323 gelation of, 320 triple ion-type imidazolium salt, 320–323 Alkaline, electrodeposition, 131, 133 Alkenes, 287 Alkylamine, 36 Alkylation, 299 Alkylcarbazole, 115 Alkyl groups, 6, 22, 296, 321–322, 452 Alkylimidazole, 36 Alkylimidazolium ionic liquids, 284–285, 341–342 Alkylpyridinium, 282 Alkylpyrolidine, 36 Alkylsultone, 302 Alkyl-3-methylimidazolium, 341–342 Allyl-1-methylmorphlinium (AMMor(FH)2,3F), 284 Allyl-1-methylpiperidinium (AMPip(FH)2,3F), 284

Electrochemical Aspects of Ionic Liquids, Second Edition. Edited by Hiroyuki Ohno. © 2011 John Wiley & Sons, Inc. Published 2011 by John Wiley & Sons, Inc.

465

bindex.indd 465

1/20/2011 10:26:34 AM

466 Allyl-1-methylpyrrolidinium (AMPyr(FH)2,3F), 284 Allylpyridinium (APy(FH)2,3F), 284 Allyl-3-methylimidazlium (AMIm(FH)2,3F), 284 Alternating current (ac), 68, 311, 344 Aluminum, 120, 122, 131, 147, 244 Amidation, 105 Amine(s) bonds, 175 functions of, 36, 121, 237, 420 neutralized, see Neutralized amines nitrogen, 6–7 Amino groups, 189 Ammonium, 16 Ammonium systems bicyclic, 367 bifluoride, 19 functions of, 60, 211–212, 218, 439 groups, 177 nitrates, 6, 19 nonaromatic heterocyclic, 284 salts, 379, 395–396 Amphiphiles, 170–171, 176 Amphoteric polymers characteristics of, 433–434 copolymers, 434–436 poly(zwitterionic liquid)s, 437–439 Amplitude, implications of, 69–70 Amylose, 173 Anhydride, 123 Anion-dipole interactions, 333 Anions, functions of, 15, 19, 21–22, 37, 44, 46, 60, 73–77, 131, 176, 422. See also specific types of anions Anisotropy, 92–93, 366 Antimony, 136–137, 143, 150 Antistatic materials, 455 Aprotic ionic liquids, 19–20 Aqueous solutions dilute, 14, 17 electrochemical processes, 41 Grotthus mechanism, 12 reference electrodes, 38–39 Arenes, fused, 104 Argon, 49, 89 Aromatic systems, 21, 60, 124, 171, 378 Arrhenium behavior, 67 Arrhenium law, 9, 17

bindex.indd 466

INDEX

Arrhenius plot(s) copolymers, 436 IL monomers, 443, 445 indications of, 10–11, 288, 297–298 ionic conductivity, 90–92, 421 molar conductivity, 80 polycation-type ILs, 446 polymer brushes, 449–450 poly(ZILs), 438 self-diffusion coefficients, 73, 75, 77 Arrhenius equation, 83 Arsenic, electrodeposition, 150 Arylethanols, 159 Aryl groups, 123, 378 Aspartame, 396 Attenuation, 69–70, 72–73 Avogadro’s number, 81 Azacrowns, 375 Backscattering, 368 Base metals, 129, 146 Base pairs, 409, 411–413, 415 Batteries components of, 102 ionic-liquid, 218 lithium, see Lithium batteries; Lithium ion batteries rechargeable, 244–245, 255 solid-state, 398 Beaker-type cell, 102 Benzenes, 105, 108, 121, 124–125, 423, 425, 452 Benzene sulfonate (BS), 452 Benzofuran, 105 Benzothioate, 113 Benzoylformate, 160 Benzoylformic acid, 121 Benzyl chloride, 123 Benzyltrimethylammonium (BTMA), 130 Benzyl-2-methylimidazole, 90 Bifluoride ions, 19, 283 Bilayer membranes charged, 176–177 glycolipid, 172–175 Biocatalysts, 165 Biochips, 187, 460 Biocompatible surfaces, 460 Biodegradable materials, 460 Bioelectroactivity, 199

1/20/2011 10:26:34 AM

INDEX

Bioelectrochemistry enzymatic reactions, 159–167 molecular self-assembly in ionic liquids, 169–181 redox reactions of proteins, 193–200 solubilization of biomaterials into ionic liquids, 183–191 Biomaterials, solubilization, 183–191 Biofuel cells, 183, 185, 187, 190–191, 197, 200 Biomasses, 184, 190, 197 Biomembranes, 170 Biomolecular energy conversion systems, 183 Biomolecules, functions of, 6, 188 Bioreactors, 187 Bioscience, future directions for, 460 Biosensors, 187 Biotechnology, future directions for, 460 Biphenyls, 378 Bipyridinium, 378 Birefringence, 385 6-Bis(Ethylenedithio)([BEDT]-TTF2Cu(NCS)2), 337 Bis(fluorosulfonyl)amide, 263–264 Bis(trifluoromethanesulfonyl)amide (NTf2), 260, 356, 359–361, 365, 367–368 Bis(trifluoromethanesulfonyl)imide HTFSI, 6, 8, 22, 90–91, 410 TFSI, 293–294 Bis(trifluoromethylsulfonyl)amide (N(CF3SO2)-2) ionic liquids, electrodeposition, 146–149, 173, 177 Bis(trifluoromethanesulfonyl)imide (TFSI) butylpyridinium (BPTFSI), 70–72, 74–82 characteristics of, 293–294, 309–312,376, 401, 405, 411, 443–446 ethyl-3-methylimidazolium (EMITFSI), 71–82, 189, 275–276, 318 imidazole (HTFSI), 237–238 lithium (LiTFSI), 422–424, 426, 435–436 P(EVImTFSI), 444 trimethylpropylammonium(TMPA-TFSI), 49–50, 60

bindex.indd 467

467 Bismuth, 134, 137 Boiling point, 6–7, 12 Boltzmann’s constant, 14, 78 Borabicyclo[3,3.1]nonane (9-BBN), 333 Boron, 331–333, 423, 461 Bromide, 169, 173, 353 Bromoaluminate systems, 52 Bromobenzene, 108 Brønsted acid-base, 236 Brownian motion, 65 Bucky gel actuators, 274–276 Bulky ions, 333 Butanol ionic liquids, 162 Butylisoquinolinium (TCNQ)2, 338–339 Butylmethylpyrrolidinium (PY13/PY14), 260, 266 Butylpyridinium (BP) characteristics of, 130, 224–225 fluorohydrogenate (BPy(FH)2,3F), 284 salts, 289–290 Butylpyridinium bis(trifluoromethylsulfonyl)imide (BPTFSI) molar conductivity, 80–82 self-diffusion, 71–72, 74–79 Butylpyridinium tetrafluorobrate (BPBF4) characteristics of, 120 molar conductivity, 80–82 self-diffusion, 71–72, 74–80 Butyl-3-methylimidazolium (BMI)/ (BMIM) characteristics of, 130, 186, 191, 343–345, 397–400 fluorohydrogenate (BMIm(FH)2,3F), 284–285 hexafluorophosphate ([BMIM][PF6]), 115, 121–125, 166, 260 tetrafluoroborate and ([BMIM)[BF4]), 115, 121–125, 166, 274–275 -3-(n-butanesulfonate), 303 Cadmium, 141, 144, 146 Calcium potassium nitrate (CKN), 10 Calorimetric studies accelerating rate (ARC), 217 differential scanning (DSC), 71, 92, 175–176, 307–308, 311, 351, 353–354, 435–436

1/20/2011 10:26:34 AM

468 Capacitance, implications of, 90, 257 Capacitance-voltage (CV), 104 Capacitors, 43, 89–90, 118, 376, 460. See also specific types of capacitors Capacity use, 214 Carbazoles, 295 Carbenes, 124 Carbohydrates, 169–170, 173, 175 Carbon, 37, 59, 70, 140, 165, 207, 260. See also Carbon dioxide (CO2); Carbon nanotube (CNT) Carbonates, 109, 403 Carbon dioxide (CO2), 119–121, 259, 361 Carbon-halogen bonds, 122 Carbon nanotube (CNT) actuators, 271–274, 277 CNT/IL/CNT system, 260 Carbonyl compounds, 121–122 Carboxylated-dialkylimidazolium zwitterion, 282 Carboxylates, 184, 303, 306–307, 313, 327–328 Carrier, generally ions, 421–422 transport, 223 Catalysis, 378 Catalyst, future research directions for, 460 Cathodes, types of, 119–122 Cations, functions of, 6, 40, 44, 46, 48, 52, 73–77, 420–426, 422, 424, 435, 437 Cation-fluorinated anion combination alkylimidazolium 208–210 nonimidazolium, 210–216 Cell constant, 88 Cell cultivation sheet, 460 Cellobiose, 185, 197–200 Cellulose, 184–185, 190–191, 461 Ceramic capacitors, 244 Cesium, 131, 147, 318 Cetyltrimethylammonium bromide (CTAB), 170–171 CF3BF3, 259, 261–262 CF3CO2, 260, 378 (CF3SO2)2N, 261–262, 266 CF3SO3H, 410 (CF3SO3)2N, 253–254, 259 C2F3BF3, 261–262 (C2F3SO2)2N, 252–254, 259, 261–263

bindex.indd 468

INDEX

C2F5BF3, 259, 266 Charge density, 321 Charge-discharge process efficiency, 252 double-layer capacitance, 258 in lithium batteries, 206–208, 210–211, 213–214, 216 Charge transfer, 223, 225, 327, 337–339 Chemoenzymatics, 160 Chirality, 460 Chlorides, 37, 130, 173, 186, 225, 443 Chloride salt, 353 Chloroaluminate anions, 207–208 Chloroaluminate ionic systems characteristics of, 37, 40–41, 49, 224–225 electrochemical polymerization, 115–119 electrochemical window of, 52–56, 59, 61 electrodeposition of metals, 129–142 moisture-sensitive, 115 transport behavior, 78 Chlorobenzene, 108 Chloroborate systems, 52 Chloroethylbenzene, 121 Chloroform systems, 118, 171, 176 Chlorogallate systems, 52 Chlorostyrene, 423 Chlorosulfonates, 22 Chlorozincate ionic systems, 40, 144–146 Cholines dihydrogen phosphate, 367 functions of, 186–187, 366 hydrogen phosphate, 197–198 Chromism, 378 Chromium, 138 Chronoamperometry, 37 Chronoprotentiometry, 37 Coating materials, 455 Cobalt, 134, 138–139, 146, 148, 344 Cohesion, 12, 20–22 Cole-Cole plot, 89 Column, future research directions, 460 Complex-impedance measurement method, 88 Condensation, 332 Condensers, types of, 338

1/20/2011 10:26:34 AM

INDEX

Conducting polymer (CP) actuators, 271–272, 277 Conductivity anisotropic, 287 diffusion and, 69, 80–83 direct connect (DC), 339 electrical, 7, 13–20 electrolytic, 247–248 fluidity and, 16–19 high electrolytic, 244 high ion, 1–3 high-temperature, 17 ionic, see Ionic conductivity low, impact of, 6 low electrolytic, 246 low ion, 3 neutralization and, 297–299 partial equivalent ionic, 14 proton, 238 reciprocal molar, 284 significance of, 205 solid-state ionic, 12 super-Arrhenius, 11 Constant current (CC), 215, 253 Constant current-constant voltage (CCCV), 215 Conversion efficiency, 225, 228, 230 Cooling process, ionic conductivity, 92 Copolymerization, 452 Copolymers, 180, 330, 415, 420–421, 434–436 Copper, 134–135, 139, 143, 146, 148–150, 229, 378 Coulombic efficiency, 207–208, 214, 313 Coulombic forces/interactions, 73, 78, 83, 227, 344 Counteranions, 426 Counter electrode, 37 Couplings, impact of, 15. See also specific types of couplings Covalent bonds, 3, 274 C4Py, 397–399 Crown ethers, 375 Crystal(s) liquid, 171–172 organic, see Organic ionic plastic crystals plastic, 353–354 Crystalline phase, 176–177, 327

bindex.indd 469

469 Crystalline state, 325 Crystallization, 11, 71–72, 92–93, 354, 378 C7TET-TTF, 339–340 Current density, 23, 44–45, 51, 103, 109 Cut-off current, 45 Cut-off density, 51 Cyano groups, 307, 309 Cyclic voltammetry (CV), 37, 44, 50, 108, 149, 193, 227 Cyclic voltammogram, 214 Cycling test, 255 Cyclohexane, 350 Cyt c, 186–187, 189–190, 194–196, 198 Cytosines, 410–411 Dai-ichi Kogyo Seiyaku Co., Ltd., 263–265 Dark-field optical microsocpy, 174–176, 178 dc four-probe measurement method, 88 Decoupling implications of, 12, 14–15 index, 17, 19 Degradability, 460 Dehalogenation, 123 Dehydration, 49 Dehydrocoupling, 332 [DEME][BF4], 166–167, 260, 263 DEME:N, 263 DEMENTf2, 260 [demo][TfO], 237, 239–240 Density capacitance, 257–258 charge, 321 conductivity measurement, 89 current, see Current density cut-off, 51 energy, 257–259 functional theory, 249 implications of, 284 power, 240 lithium batteries, 213 temperature and, 72 Depolymerization, 191 Deprotection, 287 Deprotonation, 113 Deracemization, 160, 162–164 Dialkylation, 299 Dialkylfluorene, 115

1/20/2011 10:26:34 AM

470 Dialkyl groups, 367 Dialkylimidazolium, 282, 343 Dialkylmorpholinium, 282 Dialkylpiperidium, 282 Dialkylpyrrlidinium, 282 Diastereoselectivity, 121 Dibutyl phosphate (dbp), 186 Dichloroindophenol sodium (DCIP), 197–198 Dichloromethane, 37 Dicyanamides (dcas), 22, 149–150, 186, 353, 355–358, 364, 366 Dicyanoamide (N(CN2), 149–150, 353, 355–356, 359–360, 362–364 Dicyanoethenolates, 303 Dielectric material, 243 Diffusion, see Diffusivity alkali metal ionic liquids (AMILs), 319 coefficient, molar conductivity and, 80–83 conductivity and, 365 fundamentals of, 65–66 impact of, 14–15, 23–24, 46 in-volatile solvents, 226–227, 229 ionic transport behavior studies, 70–72 organic ionic plastic crystals, 352 photoelectrochemical cells, 232 processes in liquids, 66–68 self-diffusion, 67–70, 72–80 Diffusivity, fundamentals of, 65–66 Difluorination, 107 Difluorodesulfurization, 113–114 Dihalides, 123 Dihydrogen phosphate (DHP), 19, 366 Dihydroxy groups, 384 Dilution conductivities, 17, 19 Dimethoxybenzene, 124 Dimethylamino-N-butylpyridinium (DMABP) system, 52–53, 60 Dimethylcyclo-hexanmethylamine, 295 Dimethylcyclohexylamine, 295 Dimethylformamide (DMF), 122 Dimethylimidazolium fluorohydrogenate (DMIm/DMIm(FH)2,3F), 283–285 Dimethylindole, 295 Dimethylpyrrolidium (Cnmpyr), 351–353, 355–356, 358–361, 363–366, 368

bindex.indd 470

INDEX

Dimethyl-3-propylimidazolium (DMPI), 130, 133 Dinitrobenzenes, 125 Dioxolane, 109 Dipolar liquids, 13–14 Direct current (DC), 344 Dislocations, 352, 364 Displacement, molecular, 67, 70 Dissociation constant, 328, 330 impact of, 15, 422 Distillable ionic liquids, 294 DNA characteristics of, 409–410 liquidized, see Ionic liquidized DNA Dodecyl-3-methylimidazolium fluorohydrogenate (C12MIm(FH)2F), 288 Donors, 27, 337–339 Dopant ions, 350, 359 Doping process, 221, 355, 359–361, 365, 368 Double-layer capacitors electrolyte material requirements, 245–246 fabrication of, 253–254 functions of, generally, 232 high-temperature characteristics of, 265 outline of, 243–245 performance of ionic liquids in, 252–258 recent research, 258–259, 263–266 Dropping mercury electrode (DME), 250–252 Drug carrier, 460 Drug delivery, 170 Dry heat test, 231 Dye molecules, ionic liquidized DNA, 415–416 Dye sensitized solar cells (DSSCs), 223, 225–229, 231–232 Earth metals, alkaline, 131, 133 Edibility, 460 Einstein’s equation, 67 Einstein-Stokes equation, 226 Electric energy, 185

1/20/2011 10:26:34 AM

INDEX

Electroactive polymer (EAP) actuator, 271, 276–277 Electrochemical capacitors, 244 Electrochemical methods equipment, 35 preparation of electrolytes, 36–37 reference electrodes, 37–41 working electrodes, 37 Electrochemical quartz-crystal microbalance (EQCM), 149 Electrochemical stability, 205, 368 Electrochemical windows 43–44, 51–60, 101 Electrochemistry applications of, generally, 2–3, 14 bio-related, 460 diffusion in ionic liquids and correlation with ionic transport behavior, 65–83 general techniques, 35–41 ionic conductivity, 87–93 organic, 125 room-temperature ionic liquids (RTILs), 36, 43–61 spectroelectrochemistry, 95, 193 Electrochromic cells, 415 Electrochromic windows, 115 Electroconductive materials, 2, 410 Electrocrystallization, 287 Electrodeposition, of metals chloroaluminate ionic liquids, 130–142 nonchloroaluminate ionic liquids, 142–150 overview of, 129–131 Electrodes, see Reference electrodes; Working electrodes binder-free, 207–208 bucky gel, 275 carbon, 215, 222, 245, 285 gold, 147 graphite, 207–208 ionic conductivity measurement and, 87–92 mercury, 243, 250 opaque, 193–194 polarizable, 243 saturated calomel (SCE), 104–105 silver, 147–148 tungsten, 146

bindex.indd 471

471 Electrolysis anhydrous, 103 constant current, 104 constant potential, 104 indirect, using mediators, 104 inorganic, 101 pulse, 105 Electrolytes, see Electrolytic reactions decomposition potential, 45 low-melting, 17 organic, 48 polyanionic, 330, 332 polymer, 265–266, 313 preparation of, 36–37 superionic, 19 Electrolytic reactions anodic oxidation, of alcohols and aromatic compounds, 124 cell design, 102–103 electrochemical fixation, of CO2, 119–121 electrochemical polymerization, 115–119 electroreduction of carbonyl compounds, 121–122 electroreductive coupling, using metal complex catalysts, 122–124 methodologies, 104 overview of, 101–102 selective anodic fluorination, 105–115 Electron paramagnetic resonance (EPR), 339 Electron transfer reaction, 196–197 Electronic devices, future directions for, 455 Electroorganic synthesis, 109–110, 125 Electrophoresis, 289 Electroplating processes, 102 Electropolymerization, 104, 115–116, 454 Electroreduction of carbonyl compounds, 121–122 coupling, using metal complex catalysts, 122–124 Electrospray mass spectrometry, 353 Electrostatic energy, 14 Electrostatic interactions, 303, 317, 395 Electrosynthesis, 101–102, 104, 119 EMIF ≅ 2.3HF, 252, 255–257, 259

1/20/2011 10:26:34 AM

472 EMIN (CN)2, 260 (FSO2)2, 260 Tf2, 260 EMIOTf, 260 Emission, 460 EMITaF6, 253–254, 256 Enantiomers, 160 Enantioselectivity, 162 Energy storage, 398 Enthalpy, 14 Entropy, 9, 350 Enzymatic reactions biocatalytic oxidation, 160 deracemization, 160, 162–164 fluorination reaction, 165–167 Geotrichium candidum-catalyzed synthesis, 159–160 Epoxides, 120–121, 287 Equilibrium neutralization and, 294 potentials, 249 self-diffusion and, 78 thermodynamic, 67 Esters, 159, 293, 423 Ethanols, 159–160, 162–163, 413 Ethers, 109, 294, 375 (E)-3-octen-2-ol, 162 Ethyl-2-methylimidazolium (EMI), 445 Ethylammonium characteristics of, 7, 23 nitrate (EAN), 171–172 Ethyl benzothioate, 113 Ethylcarbazole, 295 Ethylenedioxythiophene, 454 Ethylene oxide (EO), 441, 443, 445, 447, 449, 452–453 Ethylimidalizium salts, 318 Ethylimidazole (EIM) with acrylic acid (AC), 426 characteristics of, 297, 423 with styrenesulfonic acid (SS), 426, 428 with vinylsulfonic acid (VS), 427, 451–452 (Ethylimidazolio) -butane-4-sulfonate (EIm4S), 404 -propane-3-sulfonate (EIm3S), 404–405

bindex.indd 472

INDEX

Ethyl-1-methylpiperidium (EMPip(FH)2,3F), 284, 286 Ethyl-1-methylpyrrolidinium (EMPyr(FH)2,3F), 284, 286–287 Ethyl-2-phenylindole, 295 Ethyl-3-methylimidazolium (EMI)/ (EMIM) bis(trifluoromethylsulfonyl)imide (EMITFSI), see Ethyl-3methylimidazolium bis(trifluoromethylsulfonyl)imide (EMITFSI) characteristics of, 52–53, 60, 130, 133, 227, 246–247, 287, 341–342 CF3BF3, 259 CF3CO2, 260 CF3SO3, 252–254 (CF3SO3)2N, 253–254, 259 (C2F3SO2)2N, 252–254, 259, 263 C2F5BF3, 260, 264 F-HF, 283 fluoride [2,3HF], 111 fluorohydrogenate (FH)2,3F), 281, 283–286 (FSO2)2N, 263–265 (methylcarbonate) (OCO2CH3), 253–254 methylphosphonate, 184–185 PO4, 260 tetrafluoroborate (BF4,) see Ethyl-3-methylimidazolium tetrafluoroborate (EMI-BF4) trifluoromethanesulfonate [OTf], 111, 115–118 Ethyl-3-methylimidazolium bis(trifluoromethylsulfonyl)imide (EMITFSI) actuators, 275–276 characteristics of, 318 molar conductivity, 80–82 protein solubilization, 189 self-diffusion, 71–79 Ethyl-3-methylimidazolium (EMIM) salts, 288–289 Ethyl-3-methylimidazolium tetrafluoroborate (EMI-BF4) actuators, 275 characteristics of, 299, 318, 380, 412–416

1/20/2011 10:26:34 AM

473

INDEX

chloroaluminate ionic systems, 46–48 decomposition voltage, 249–250 double-layer capacitance, 251–256, 259–260, 263 fluoroanion salts and, 285 molar conductivity, 80–82 self-diffusion, 70–79 Ethylpiperidine, 295 Et3MeNBF4/PC, 252, 254–257, 259, 264 Et4NBF4/PC, 263 Europium, 149 Eutectic, generally low-melting, 11 melting, 359 mixtures, 9 Evaporation, 41 Excess boiling point, 6–7 Faraday equation, 14 Faraday’s constant, 81 Fast ion conductors, 350 Fast proton transport, 436 Ferricenium, 38–39, 41, 228 Ferrocene, 38–39, 41, 47–48, 52, 56, 59, 224–225, 228, 286 Fick’s laws, first/second, 66 Film(s), generally capacitors, 244 composite, 118 ion conductive, 412 palladium-polypyrrole composite, 118 passivation, 133 polymer, 114, 118–119, 238, 380, 454 polypyrrole, 117–118 polythiophene, 118 solid-electrolyte interphase, 146 thin, 95 two-dimensional, 387 Filtration, 38, 41 Fischer esterification, 313 Flame resistance, 205 Flame retardancy, 2 Flavin (FAD), 197–199 Fluidity cohesion and, 20–22 conductivity and, 16–19 influential factors, 14, 20 Fluorene, 454

bindex.indd 473

Fluoride ionic liquids, 19, 107–109, 111, 165–166, 282–283, 289 Fluorinase, 165–167 Fluorination, 105–115, 165–167 Fluorine, 81, 73, 165, 282, 365 Fluoroanions, 49, 249, 256, 284 Fluorodesulfurization, 115 Fluorohydrogenate ionic liquids, 281, 283–289 (Fluorophenyl)thiophene, 119 Fluoropyridine, 23 Fluorosulfuric acid (FSO2) (CF3SO2)N, 261–262, 266 N, 261–262, 266 Fluorosulfurization, 111 Formates, 7, 22 Fractional Walden rule, 17, 19 Freezing point, 9 Frequency range, significance of, 89 Fuel cells all-solid, 19 bubbler type, 22–23 characteristics of, 43, 102, 301, 398 electrolytes, 7 fluorohydrogenate ionic liquids and, 286–287 future research directions for, 460 hydrogen and, 350 hydrogen-oxygen, 22 nonhumidified, 240 phosphoric acid, 22, 238–239 proton membrane, 235–236 proton transfer ionic liquids, 6, 22–24 Functional design alkali metal ionic liquids, 317–323 amines, neutralized, 293–300 electric conductivity and magnetic ionic liquids, 337–345 novel fluoroanion salts, 281–290 polyether/salt hybrids, 325–333 zwitterionic liquids, 301–313 Functional groups, 459 Functional materials, 105 Furan, 105 (Furyl)ethanol, 162 Fusion, 14, 350 Gallium, 134–135, 147, 150, 344 Galvanostatic cycling, 368

1/20/2011 10:26:34 AM

474 Gel polymer electrolytes conductivity, 425 ionic liquid gels, 395–401 ionic liquidized DNA, 409–416 zwitterionic liquid/polymer gels, 403–407 Gelatinization, 173 Gelation, 170, 320, 389, 395–398 Gelators, 229, 231, 395–398 Gels bucky, 274 ionic liquid, 395–401 polymer electrolyte, 272 zwitterionic liquid/polymer, 403–407 Geotrichium candidum, 159–162 Germanium, electrodeposition, 142–143, 147 Gibbs free energies, 12–13 Glass frit, 38, 46 porous, 38 temperature, 9–11, 21–22 transition, 354 transition temperature, 299, 305, 327, 434 Vycor, 104 Glassy carbon (GC), 247, 252, 285 Globular proteins, 194 Glucose, 185, 191 Glucose oxidase (GOD), 173 Glycerol, 19 Glycolipids, 170, 172–175, 389 Glycosylation, 170 Gold electrodeposition, 135, 139, 141, 144 photoreducing, 180–181 Grätzell cell, in-volatile solvents containing a low viscosity ionic liquid, 227–228 defined, 223 with ionic-liquid-based on iodide, 225–226 Green chemistry, 27 Greenhouse gases, 235 Grotthus mechanism, 12, 19, 223, 238 Guanidine, 26–27 Gyromagnetic ratio, 70

bindex.indd 474

INDEX

Hahn spin-echo pulse sequence, 69 Halides alkyl, 378 aryl, 123, 378 characteristics of, generally, 36–37, 104 germanium, 142 metal, 150 metathesis, 281 organic, 122, 150 Halofluorination, 287 Halometalates, 343 Haven ratio, 69 Heating processes, 92, 304–305, 307–308 Heme proteins, 189, 196, 198–199 Henry reaction, 124 Hexafluorophosphate (PF-6), electrodeposition, 142 Hexylmin, 166–167 Hexyl-3-methylimidazolium iodide (HMI-I), 225–226, 230 tetrafluoroborate (HMIBF4), 275 High double-layer capacitance, 244 High-performance liquid chromatography (HPLC), 165 High-temperature ionic liquids, 5 Hofmeister series, 187 Homo-coupling, 122–123 Hopping fluorohydrogenate ionic liquids, 284 Grotthuss mechanism, 238 mechanisms, 223, 229, 238 proton, 436 H-type cell, 102–103 Humidifcation, 240 Hybrids, polyether/salts anion conductive, 333 chemical structures, representative, 326 design of, 326–327 ionic conductivity improvements, 327–330 overview of, 325–326 polyether/alminate, 331–333 polyether/borate, 331–333 polymerized PEO/salt, 329–331 schematic representation of, 328 Hydrated ionic liquids, 185, 197–200 Hydrates, types of, 9–10, 12

1/20/2011 10:26:34 AM

INDEX

Hydration, 25 Hydrazine, 171 Hydrazinium nitrate, 8 Hydrocarbon (HC) implications of, 171 spacers, 441, 444–445, 447 Hydrodynamics, 78 Hydrogels, 173, 179 Hydrogen, generally bonds/bonding, 83, 125, 171, 173, 176, 178, 184, 283, 395 characteristics of, 22, 73, 317–318 chloride (HCl), 130, 281 conductivity, 365 fluoride (HF), 282–283 -oxygen fuel cells, 239 peroxide, 19 phosphate (dhp), 186–187 sulfoxide (H2SO4/H2O), 252, 254, 259 tetrafluoroborate (HBF4), 253, 295–297, 410 Hydrolysis, 6, 115, 185, 289, 333 Hydrophilic ionic liquids, 287 Hydrophilicity, 236 Hydrophobic ionic liquids, 37, 49–50, 167, 395 Hydrophobicity, 236, 454 Hydroxide groups, 186 Hydroxyesters, 159 Hydroxy groups, 382, 384 Hydroxyl groups, 173, 452 Hysteresis, 92 Imidazole ionic liquids, 22, 49, 90, 238–239 Imidazole-trifluoromethanesulfonamide (HTFSI), 238 Imidazolium alkylated, 131 cations, 22, 414 ionic conductivity, 413 ionic liquids, 68, 111, 115, 121–125, 130, 133, 166, 186, 191, 225–226, 230, 284–285, 288, 299, 303, 317, 341–345, 376–377, 397–400 plastic crystals and, 367 salts, 378, 395–396 Imine groups, 25

bindex.indd 475

475 Impedance, 80, 88–89, 311, 338, 368 Incident light, 97 Independently moving ions, 14 Indium, 135, 147, 150 Indium tin oxide (ITO), 385 Indole, 295–296 Inelastic scattering, 14 Inert gases, 89, 102 Inorganic ionic liquids, distinguished from organic, 27 Interfacial materials chemistry, 180–181 Internal reflection, 96 International Union of Pore and Applied Chemistry (IUPAC), 46, 48 Interpenetrating polymer networks (IPNs), 452–453 Iodide, 225–227, 353, 425–426 Iodine, 340 Iodoarene, 113 Ion-dipole interaction, 188, 330, 333, 433 Ionic conductivity, see specific chemicals conductive materials, 2–3 future research directions, 460 measurement of, 87–93 significance of, 87 Ionic devices actuators, 271–277 double-layer capacitors, 243–266 dry, 403 fuel cells, 235–241 Li batteries, 205–218 neutralized amines, 293–300 novel fluoroanion salts, 281–290 photoelectrochemical cells, 221–232 zwitteronic liquids, 301–313 Ionicity, 6, 8, 24–27, 69 Ionic-liquid batteries, 218 Ionic liquid gels conductivity of, 398–401 gelation, 395–398 polymer, 401 strength of, 397–398 Ionic liquidized DNA dye molecules, redox reactions of, 415–416 inner column, 410–413 outer column, 413–414

1/20/2011 10:26:34 AM

476 Ionic liquids, generally classes of, 5–8 designable, 1, 3 fine-tunable, 1 fundamental properties of, 246–252 future directions for, 459–461 gels, see Ionic liquid gels importance of, 2 organic, see Organic ionic liquids overview of, 1 post-treatment technologies, 3 potential of, 2–3 protic, see Protic ionic liquids (PILS) proton transfer, 22–24 thermal stability, 3, 5–6 vaporization, 13–20 Ionic melts, low-temperature liquid behavior of, 2, 9–12 Ionic plastic crystals, 287. See also Organic ionic plastic crystals (OIPCs) Ionic polymer-metal composites (IPMCs) actuators, 271–272, 277 Ionic Processes in Solutions (Gurney), 24 Ionic transport, 68 Ionogels, 172–175, 178–179 Ionophilic-ionophobic interactions, 169, 177–180 IR drop, 44, 46, 254 Iridium, 344 Iron, 134, 138, 146, 148, 343–344 Isobaric systems, 66 Isomers, 12 Isothermal systems, 66 Isotopes, 289 Karl Fischer method, 172 Kauzmann extrapolation, 9 Ketones, 162–163 Kinetics, electrochemical, 216–217 Kosmotropicity, 187 Lactate, 186 Lanthanum, 141, 149 Lattice(s) crystal, 9 crystalline, 12–14, 349 defects, 358

bindex.indd 476

INDEX

energies, 12–13 quasi-, 13–14 Law of matter conservation, 66 Lead, 134, 136 Lewis acid, 130, 142, 144, 146 Light-emitting diodes, 376 Li-ion battery application examples, 206–216 characteristics of, generally, 376 future research directions, 217 performance improvements, 216–217 plastic crystals for, 369 rechargeable, 419 safety aspects, 205–206 thermal stability, 217–218 Linear sweep voltammetry (LSV), 44, 49–50, 251 Lipases, 183 Lipids, 183 Liquid brake, 460 Liquid crystal, 180 Liquid crystalline ionic liquids anisotropic ionic conductivity, 384–388 chemical modification of, 376–380 development of, 376 overview of, 375–376 self-assembly, 380–384 Liquid organic metals, 338 Lithium alkali metal ionic metals and, 317–318, 321 batteries, see Li-ion batteries; Lithium batteries bis(trifluoromethylsulfonyl)imide (LiTFSI), 422–424, 435–436 chloride, 19–20 conductivity, 359–361, 365 electrodeposition of, 131, 146 electrolyte solutions, 16 gelation process, 320 halides, 37 haloaluminate/pseudohaloaluminate, 11 metal, 89 organic ionic plastic crystals, 349, 359 polyether/salt hybrids, 327 salts, 355, 375 trifluoroborate (LiBF4), 260, 421 zwitterionic liquids and, 310–313

1/20/2011 10:26:34 AM

INDEX

Lithium batteries, see Li-ion battery characteristics of, 43, 301 conductivity, 213 design of, 350 future research directions for, 460 plastic crystals for, 367–368 polymer, 425 Low-temperature ionic liquids, 5–6 Lubrication, 460 Luggin capillary, 39, 103 Luminosity, 460 Lutidine, 295 Madelung energy, 14, 338–339 Magnesium, 120, 122, 133–134, 147 Magnetic ionic liquids, electric conductivity cation radical salts with low melting point, 340–341 EMI salts containing complexes with paramagnetic metals, 343–345 influential factors, 337–340 TCNQ anion radical salts with low melting point, 341–343 Magnetism, 378, 460 Magnetization, 69 Mandelic acid, 121, 159 Manganese, 148–150, 344 Mass transfer, 121 Mass transport, 45–46, 106, 130 Material synthesis, bilayer-templated, 170 Matrix transport, 238 Matsumoto, 259, 263 Media, generally aqueous, 170–171 in-volatile solvents, 228–229 low-melting, 27 nonaqueous, 41 organic, 170–171 Mediators, indirect electrolysis, 104 Melting points Arrhenius law, 9 electrochemistry and, 299 electrodisposition and, 146 of fluorohydrogenate ionic liquids, 289 gelation and, 395 implications of, 5–6, 252 lattice energy and, 12–13 lithium batteries, 213 low, 340–341

bindex.indd 477

477 of magnetic ionic liquids, 338 neutralization and, 296–297, 299 normal, 11 Mercury, 134–135, 141, 252 Mesogenic compounds, 381–385, 387 Metal, see specific metals alkaline, 131–132 catalysts, 23 complexes, 3, 459 ionic conductivity of, 89 ions, multiple valence, 104 oxides, 3 Metallic bonds, 337 Metallomesogens, 389 Metathesis, 281, 288, 353, 455 Methacrylate, 420 Methanesulfonate (MeSO4), 22, 186 Methanol, 413 Methionine, 165–167 Methoxy groups, 329 Methoxymethyldimethylethyl ammonium, 16 Methoxymethyl-1-methylpyrrolidium fluorohydrogenate (MOMMPyr(FH)2,3F), 284 Methylbenzimidazole, 295 Methylene blue, 96 Methyl-5-hepten-2-ol, 163 Methyl groups, 296, 305, 321, 357 Methylimidazole, 295, 297 Methylindole, 295 Methyl-isoquinolinium, 341 Methyl(methoxyethyl)pyrrlidinium, 266 Methyl(methoxymethyl)pyrrolidinium, 263 Methyl-N-butyl pyrrolidinium (C4mpyr), 186, 354 Methyl-1-pyrroline, 295 Methylphthalimide, 122 Methylpiperidinium, 284, 286, 445 Methylpropylpyrrolidinium, 263 Methylpyrazole, 295 Methylpyridium (MP), 7, 130 Methylpyrrole, 295 Methylpyrrolidine, 295 Methylpyrrolidium characteristics of, 284, 286–287, 351–353, 355–356, 358–361, 366, 368 salts, 318

1/20/2011 10:26:34 AM

478 Micelles, 170–172 Microcrystalline, 173, 176, 231 Microelectrodes, 37, 45–46, 229 Microscopic studies polarized optical microscopy, 287, 379, 385–386 scanning electron microscopy (SEM), 117–118, 149, 364, 368, 381 scanning transition microscopy (STM), 149 Minimum gel concentration (MGC), 396–400 Mitsubishi Chemical Corporation, 259–260 Mobility temperature, 10 Molar conductivity, 69, 80–83 Molar volume, 21 Molecular dynamics studies, 283 Molecular self-assembly in aqueous media, 170–171 characteristics of, 169–170 charged bilayer membranes, 176–177 future research directions, 180–181 glycolipid bilayer membrane formation, 170, 172–175 ionogel formation, 172–175, 179 ionophilic-ionophobic interactions, 177–180 liquid crystal formation, 171–172 micelles formation, 170–172 in organic media, 170–171 Molecular weight, significance of, 8, 72, 82, 189, 327, 389 Mole ratio, 52 Monofluorinated acetate, 105 Monomers cross-linking, 449 imidazolium (IL), 441–443, 445, 453 Monte Carlo simulation, 275 Multiwalled carbon nanotubes, 276 Mutual diffusion, 67–68 Nafion, 419 Nanocrystals, 181 Nanotechnology, 92–93 Naphthalene, 104–105 Napthylethanol, 162 Native proteins dissolution of, 185–187 in hydrated ionic liquids, 197–200

bindex.indd 478

INDEX

Neat ionic liquids, 216–217, 398 Nernst-Einstein equation, 14, 81 Network polymers, 447–450 Neutralization of amines, see Neutralized amines implications of, 294–296 ionic liquidized DNA, 410–412 polyanions, 426 Neutralized amines easy preparation requirements, 293–294 ionic conductivity, 297, 299 neutralization process, 294–296, 420 ordinary quaternized onium salt models, 299 properties of, ionic species effects on, 296–297 Nickel, 121–122, 134, 139, 148, 150, 344 Niobium, 140, 150 Nisshinbo Industries, Inc., 260, 263 Nitrates, 6, 8–10, 19, 22, 171–172 Nitric acid, 20 Nitrites, 22 Nitrogen, 49, 70, 89, 102, 122, 293, 352 Nitrosonium tetrafluoroborate (NOBF4), 340–341 Nitroxyl radicals, 104 Nonaqueous solutions, 38 Nonchloroaluminate ionic systems characteristics of, 40–41, 47 electrochemical window of, 56–61 electrodeposition of metals, 129, 142–150 lithium batteries and, 209–210 transport behavior 78 Noncovalent interactions, 180 Nonvolatility, 1–3 Norbornene, 454–455 Nuclear magnetic resonance (NMR) studies, 69–70, 81, 83, 282, 318–319, 351, 353, 355, 362, 364–365, 369 Nucleic acids, 409, 413 Nucleophilicity, 113 Nyquist plot, ionic conductivity, 89–91 Octane, 367 Octanol, 162–163 Octylmin, 166–167 Ohmic drop, 45 Olefins, 105, 123

1/20/2011 10:26:34 AM

INDEX

Oligo(ethylene oxide), 375 Oligomers, 115, 188, 325, 327, 331, 333 Oligosaccharides, 185, 197 OMIBF4, 275 Onium, 299–300, 307, 325, 425 Open circuit voltage (OCV), 214, 223 Ordered structures, ionic liquids in ion conduction in organic ionic plastic crystals, 349–369 liquid crystalline ionic liquids, 375–389 Orders of magnitude, 45, 284, 342, 359, 366, 420 Organic ionic liquids aprotic cation class, 6 characteristics of, 2 degradation, 3 highest-conducting, compared with inorganic, 27 Organic ionic plastic crystals (OIPCs) characteristics of, 349–350, 369 conductivity in, 355–362 electrolytes, synthesis and thermal properties of, 352–355 plastic crystal phases, 350–352, 362–366 research directions for, 366–368 thermal behavior analysis, 352–355 Organic semiconductive condenser (OSCON), 338 Organoborates, 304, 313 Organoboron polymers, 423 Organogelators, 389 Organ storage solution, 460 Osmotic stress, 275–276 Oxidation anodic, of alcohols and aromatic compounds, 124 biocatalytic, 160, 197 impact of, 43, 198, 249 potential, in RTILs, 44–46 RTILs, 50–51 Oxidation-reduction potential, 196. See also Redox reactions Oxides, liquid, 5 Oxofluorides, 289 Oxyanions, 21 Oxygen, 21, 23, 35, 49 Palladium, 118, 135, 140, 144, 149 Paramagnetic bifunctional ionic liquids, 344

bindex.indd 479

479 Paramagnetic salts, 343–345 Passivation, 104–105 Pentafluorides, 289 Pentafluoropyridine, 26 Pentaglycerine, 350 Perfluorinated species, 21–22 Perfluoroalkyltrifluoroborate, 259 Perfluoroanion, 216 P(EVImTFSI), 444 Pharmaceuticals, 105 Phenanthroline, 105 Phenylene, 454 Phenylethanol, 160, 163 Phenyl groups, 388 Phenylpropanol, 162 Phenylthioacetate, 106–107 3-Phenylthiophthalide, anodic fluorination, 112 Phenyl-2-butanol, 162 Phosgene, 37 Phosphate groups, 184, 186, 197, 413–414 Phosphoric acid, 22–23, 238–239 Photoconversion efficiency, 231 Photocorrosion, 225 Photocurrents, short-circuit, 223 Photoelectrochemical cells (PEC) characteristics of, 221, 223 performance evaluation of, 223–224 schematic illustration of, 222–223 Photopolymerization, 381 Photovoltaic cells, 102, 230 Phthalides, 110 Physical chemistry classes of ionic liquids, 5–8 cohesion, 20–22 electrical conductivity related to low vapor pressure, 13–20 fluidity, 20–22 highest conductivity, 27 low-temperature liquid behavior, 9–12 melting points, 12–13 proton (PILS), ionicity and acidity, 24–26 proton transfer, 22–23 Picolium, 7 B-B interactions, 395 stacking, 410 pK values, 6–7, 25, 236

1/20/2011 10:26:34 AM

480 Plastic crystal, 287, 353–354, 369. See also Organic ionic plastic crystals; Plastic crystal phases Plastic crystal phases, see Organic ionic plastic crystals background of, 350–352 transport mechanisms, 362–366 Platinum, 37, 59, 88 Polar ionic liquids, 184–185 Polarity, 460 Polarizability, 378 Polarization, 239 Polonium ions, 349 Polyacrylonitrile (PAN), 260, 266 Polyaniline, 115 Polyanions, 426–428 Polyarene, 115 Polybutadiene (PB), 452–453 Polycations, 420–426 Poly(3,4-ethylenedioxythiophene) (PEDOT), 119 Poly(ethylene glycol) (PEG), 113, 189 Poly(ethylene oxide) (PEO) double-layer capacitors, 266 liquid crystalline ionic liquids, 375 lithium salt hybrids, 329 polymerized/salt hybrids, 329–331, 333 solubilization process, 188–190, 193–196 Polyelectrolytes, 332 Polyether systems, 187–190, 423, 433–434, 452–453 Polyfluorenes, 114 Polymer brushes cross-linkers, 446–449 imidazolium, 441–455 thermal stability, 445–446 Polymerization anodic, 119 copolymers, 435 electrochemical, 115–119 electrooxidative, 115 hydroboration, 423 implications of, 104 in situ, 419 polycations, 422 polymer brushes, 443–444

bindex.indd 480

INDEX

Polymerized ionic liquids amphoteric polymers, 433–439 ion conductive polymers, 419–428 polymer brushes, 441–455 Polymers amphoteric, 433–439 brushes, see Polymer brushes conducting (CPs), 271–272, 277 electroactive (EAP), 271 electroconducting, 115 electrolyte, 3 gels, thermal stability, 404–405 glucose, 191 ion conductive, 89–91, 419–428, 454 synthetic, 409 vinyl, 439 water-absorbing, 165–166 Poly(norbornene)s, 425–426 Polyphenylene, 119 Poly(propylene oxide) (PPO)/salt hybrids, 329–331 Polypyrroles, 115, 117–118 Polysaccharides, 184 Polysiloxanes, 330 Polythiophene, 115, 118 Poly(vinylidene fluoride-cohexafluoropropylene) (P(VdF-coHPF)), 404–406, 419 Poly(vinylidene fluoridehexafluoropropylene) (PvdF(HFP)), 260, 274, 277, 320 Poly(zwitterionic liquid)s, 437–439 Poor ionic liquids, 15 Potassium characteristics of, 318 chloride (KCl), 88, 194–196 cyanide, 309 electrodeposition of, 131, 146 fluoride (KF), 165–166 Potential scanning polymerization, 104 Potentiostat, 104 PO2F2, 261 Power density, 240 PP13, 60 Preserver, 460 Pressure, in electrochemical carboxylation, 121 Propanol, 162 Propylammonium, 7

1/20/2011 10:26:34 AM

481

INDEX

Propylene, 253, 403 Propylene carbonate (PC), 245, 251, 257, 399 Propylisoquinolinium, 341 Proteins electron transfer reaction, 196–197 glycosylation, 170 native, 185–187 redox reaction, 193–200 solubilization methods, 187–190 Protic ionic liquids (PILS) ambient temperature, 7 characteristics of, 5–6, 295 conductivities of, 19–20 ionicity and acidity, in proton free energy level diagram, 24–27 research goals, 12 vaporization of, 13 Proton(s) free energy level diagram, 24–27 transfer process, 24–25, 238 Pseudoreference electrode, 39 Pulsed field gradient nuclear magnetic resonance (PFG-NMR), 318–320, 365 Pulsed-gradient spin-echo NMR (PGSENMR), 69–70, 81, 83, 284 Purification, electrolyte preparation for, 36–37 Purines, 410–411 Pyrazolium system, 211, 366–367 Pyridines, 23, 26, 36 Pyridinium characteristics of, 68, 317 salts, 396 Pyridylethanols, 159 (Pyrimidylthio)acetate, 113 Pyrridinium, 131, 376 Pyrroles 295, 454 Pyrrolidinium, 22, 186, 263, 284, 286–287, 350–351, 354–362, 364, 367 Quasilattice, 13–14 Quasi solid-state DSSCs, 229, 231 Quasireference electrode (QRE), 39, 48 Quaternary ammonium, 10, 22 Quaternary ammonium cation, 356 salts, 119, 259, 368 systems, 60, 211–212, 218, 227, 351, 366

bindex.indd 481

Quaternization, 294, 423 Quinolines, 108 Radical-producing processes, 95 Radical salts anion, 341–342 cation, 340–341 Radio frequency (RF) pulse, 69–70 Random walk, 67 Recovering process, 461 Recycling process, 461 Redox behavior, 378 Redox coupling, 133, 135, 224 Redox hopping, 223 Redox potential, 38–39 Redox reactions conductivity measurement, 89 optical waveguide spectroscopic studies, 96–98 ionic liquidized DNA, 415–416 Redox system, RTILs, 46–48 Reduction potential, in RTILs, 44–46 Reference electrodes characteristics of, 37–40, 247, 249 chloroaluminate systems, 40, 52–53 chlorozincate systems, 40 nonchloroaluminate systems, 40–41, 47 redox reactions, 194 RTILs, 44, 46–48 Refractive index, 97, 460 Relative permittivity, 243 Relaxation, structural, 12 Repulsion, electrostatic, 177, 179 R4N, 356 RMIm, 284–285, 287 Room-temperature ionic liquids (RTILs) electrochemical energy devices, 223 electrochemical windows, mutual comparisons of, 51–60 electrodeposition of metals, 129 overview of, 36, 43–44, 60–61 purity of, 44, 61 transport behavior, 71–72 viscosity, 228–229, 232 voltammetry, measurement conditions, 44–51

1/20/2011 10:26:34 AM

482 Room-temperature molten salt (RTMS), 9 Rotational motion, measurement of, 353–354 Rotational order, 65–66 Rotator(s) motions, 351 phases, 363–364 Rubidium, 131, 318 Ruthenium, 150, 344 Saccharinate, 186 S-adenosyl-L-methionine (SAM), 165–167 Safety, 245, 247 Salt(s), types of ammonium, 7, 37, 119, 259 aprotic, 6 butylpyridium, 289–290 characteristics of, 5 chloride, 184 cohesion of, 21 composition of, 2 EMI, 247–248, 343–345 fluoroanion, 281–290 hexafluorometallate, 288–289 hydrophobic, 293 imidazolium, 321–323 inorganic, 3, 8–9, 317 liquid fluoride, 107–109 lithium, 37, 310–311, 313, 422, 425 molten, 2, 6, 9, 11, 14, 101, 205, 285, 301, 317, 323, 326, 331, 401 organic, 3, 37, 309, 327 protic, 8, 12 proton transfer, 16 radical, 340–343 room-temperature, 101 silver, 41 solid-state behavior of, 354 triflate, 297 valence, 340 Walden plots, 16 Samarium, electrodeposition, 149 Scan rate, 49–50, 59, 92 Selenium, 147 Self-assembly, of ionic liquids with liquid crystal molecules, 380–384

bindex.indd 482

INDEX

Self-diffusion coefficient measurement using pulsedgradient spin-echo (PGSE) NMR, 69–70, 73 defined, 67 ionic transport in ionic liquids, 68–69 viscosity and, 72–80 Self-organization, liquid crystalline ionic liquids, 379, 388–389 Semiconductors, 221–222, 224, 337, 339 Sensor systems, 170, 350, 378, 398, 460 Shanxi Institute, 260 Shearing, 379 Signal-to-noise ratio, 193 Silicates, 5 Silicon, 147, 229, 361 Silver alkali halides, 17 characteristics of, 37 electrodeposition, 134, 140, 143, 149 EMI salts and, 343 halides, 7 redox reactions, 41 reference electrodes and, 48 Single-walled carbon nanotube (SWNT) actuator, 273–274, 276 gelation and, 389 polymer brushes, 455 Sodium borohydride (NaBH4), 163 characteristics of, 131, 318, 349 Solar cells, 43, 221, 223, 301, 376, 460 Sol-gel interactions, 180 transition, 398 Solid electrolyte interface (SEI), 207, 217 Solid polymer electrolyte (SPE), 101 Solid state electrolytes, 206, 369 implications of, 352 NMR, 369 quasi-, 229 Solubility, 460 Solubilization process, 180, 183–191 Solvation, time-averaged, 81 Solvent effect, 247 Solvent-free liquids, 14, 20 Solvents aprotic, 36, 38, 101, 105

1/20/2011 10:26:34 AM

483

INDEX

aqueous, 183 future research directions, 460 hydrogen-bonded, 19 in-volatile, 224–229 organic, 101, 105–106, 113, 159, 169, 171, 183, 206, 217, 221, 225, 229, 232, 258, 403 polar, 171 volatile, 205, 221 Solvodynamic models, 67 Solvophilic-solvophobic amphiphilicity, 169, 171, 176 Sonovoltamammogram, 106 Spacers, polymer brushes, 441–445, 447, 452–453 Spectral analysis, 193 Spectroelectrochemistry, 95, 193 Spectroscopic studies impedance, 68 infrared (IR), 282 optical waveguide (OWG), 95–98, 189, 193–195 positron annihilation lifetime (PALS), 352, 364, 369 Raman, 366, 369, 413 resonance Raman, 186–187 transport studies, 68 ultraviolet (UV), 413, 415 X-ray photoelectron (XPS), 147, 150 Stacking, 171, 410 Stainless steel, 88 Sterilization, 460 Stoichiometry, 283 Stokes-Einstein, see Einstein-Stokes equation equation, 15–16, 78 model, 67 Stokes radius, 78 Strain-induced defects, 361 Streptococcus faecalis, 160 Streptomyces cattleya, 165 Structure-property relationship, 296 Styrenesulfonic acid, 426 Subionic liquids, 15–16, 19 Sublattices, 17 Succinonitrile, 350 Sugar-philic ionic liquids, 172–175 Sulfates, 318 Sulfonate groups, 306, 328, 434

bindex.indd 483

Sulfonylamides, 304, 309–310 Superacidic ionic liquids, 6–8, 26 Super-Arrhenius transport, 6, 9, 11 Superbases, 6, 25–26 Super-capacitors, 244, 460 Superconductors, 337 Supercooling, 299, 342 Super-growth single-walled carbon nanotubes (SG-SWNTs), 276 Superionic, generally liquids, 15–16 inorganic, 19 Superprotonic conduction, 19 Surface protection, 460 Surfactants, 170, 172, 176, 180–181 Switches/switching degradation switching, 460 future directions for, 460 liquid, 460 nano-switch, 460 optical, 460 Tafel plot, 24 Tantalum characteristics of, 149 electrolytic capacitors, 244 TeEA-TSAC, 60 Teflon sandwich cell, 23–24 Tellurium, electrodeposition, 135, 137–138, 146 Temperature absolute, 78–79 ambient, 2–3, 7–8, 14, 19, 330, 366, 369 electrical conductivity and, 263 electrochemical carboxylation, 121 glass transition, 299, 305, 327, 434 ionic conductivity and, 90–91 mobility, 10 molar conductivity and, 81 transport behavior and, 72 viscosity and, 77 wide operational range, 245–247 Tetraalkylammonium, 263 Tetraalkylphosphonium, 282, 284 Tetrachloroaluminate, 7 Tetracyanoborates, 22 Tetracyanoquinodimethane (TCNQ), 337, 341–343 Tetraethylammonium (TEA), 130

1/20/2011 10:26:34 AM

484 Tetrafluoroborate (BF4) ionic liquids, 142, 186, 253, 261–262, 274–275, 295–297, 340–341, 398, 410 salts, 16 Tetrahydrofuran, 109 Tetramethylammonium (TMA), 186, 357–358 Tetramethyl-p-phenylenediamine (TMPD), 415 Tetramethyltetraselenafulvalene (TMTSF), 337 Tetrathiafulvalene (TTF), 337–340 Thallium, 136, 149 Thermal analysis, 351–352 Thermal conduction, 460 Thermal degradation, 449 Thermal energy, 65–66 Thermal stability, 3, 217–218, 240, 247, 338, 382, 384 Thermodynamics, 5, 12, 358 Thermogravimetric analysis, 445 Thermogravimetric differential thermal analysis (TG/DTA), 405 Thiocyanates, 22, 353, 366 Thiolates, 328 Thiophene, 115–116, 118–119, 454 (Thiophen-2-yl)ethanol, 162 Thiosaccharinato (TSAC) anion, 60 Thorium, 149 Time-averaged ions, 81 Tin, 134–136, 147, 150 Titanium, 135, 138, 148 Titanium oxide, 223 Toluene, 108 Toluenesulfonic acid ethyl ester (TSE), 423 Toxicity, 460 Transition metals, 104, 138–148 Translational diffusion, 66–67 Transmission spectrochemistry, 95 Transport influential factors on, 5–6, 9, 11, 16 mechanisms, 362–366 processes, 68–69 Trialkylalkoxyammonium, 259 Triarylamines, 104 Triazolium, 52 Trichlorofluoroethane, 122 Triflate (Tf), 7

bindex.indd 484

INDEX

Triflic acid, 6 Trifluoroacetate (TFAc), 7 Trifluorobenzenes, 108 Trifluoromethanesulfonate, 7 Trifluoromethanesulfonimide, 398 Trifluoromethylsulfonyl (TFS), 453–455 Triiodides, 229, 231 Trimethylhexylammonium (TMHA), 130 Trimethyl(methoxymethyl)ammonium, 263 Trimethylpropylammonium (TMPA), 263 Trimethylpropylammoniumbis(trifluoromethyl)imide (TMPATFSI), 49–50, 60 Tri-n-butyl-methylphosphonium fluorohydrogenate (P4441(FH)2,3F), 284 TRIS-HCL buffer, 165–167 TTC7-TFF, 340–341 Tungsten, 37, 59, 228, 289 Turbidity, influential factors, 200, 415 Ultracapacitors, 244 Ultrasonication, 106–107, 122 Ultraviolet (UV)-visible absorption, 96 Undivided cells, 103 Uni-univalent inorganic molten salts, 11 University research studies Kansai University, 260, 264 Poznan´ University of Technology, 260, 265–266 Shinshu University, 260 University of Bologna, 260, 266 University of Paul Sabatier, 260 Yokohama National University, 260 Vacancy defects, 356 migration process, 352 Vacuum drying, 35–37, 49 Vanadium, 344 van der Waals forces, 177 interactions, 20, 337 Vapor pressure at ambient temperature, 3 fuel cells, 236

1/20/2011 10:26:35 AM

485

INDEX

high, 6, 16 low, 8, 13–20, 68 Vinyl groups, 294, 434, 452 Vinylimidazole, 423, 436 Vinylphosphonic acids, 427, 436 Vinylsulfonic acid, 426 Viologens, 104, 340, 415–416 Viscosity alkali metal ionic liquids, 317 cohesion and, 20 conductivity, 14–15 electrodeposition and, 146 electrolytic conductivity and, 248 electroreductive coupling, 122 fluoroanion salts, 281, 283–284 fluorohydrogenate ionic liquids, 284 ionic conductivity and, 92, 327–328 ionic liquid gels, 399 lattice energy and, 12–13 lithium batteries, 213 low, 231 polar ionic liquids, 184 polar organic solvents, 403 polymer brushes, 447 RTILs, 45–46 self-diffusion coefficient and, 72–80 significance of, 205, 247 super-Arrhenius transport and, 9 zwitterions, 309 Vogel-Fulcher-Tammann (VFT) equation, 10, 67, 75–78, 80–81, 83, 92 Volatility, influential factors, 287, 403 Voltage cell design and, 102 constant DC, 339 decomposition and, 244–245, 249–250 double capacitance, 258 fluoroanion salts, 285–286 lithium batteries, 214 Voltammetry/voltammogram RTILS, 44–46, 48 scan rate effects, 50–51, 59 water effects on, 49–50

bindex.indd 485

Walden plot, 284–285 Walden’s rule, 15, 17, 19, 284 Warburg impedance, 89 Water as electrolytic solvent, 101 contamination, 38, 49 effect on RTILs, 44, 49–50 -stable ionic liquids, 37 Waves, evanescent, 96 Wire(s), types of aluminum, 47, 52 platinum, 41, 98 silver, 98 Working electrodes, 37, 39, 45–46 Wunderlich’s “bead” notation, 22 XGG/IL/XGC system, 260 X-ray crystallography, 369 X-ray diffraction (XRD), 147–150, 283, 287, 344, 362–363 Zero-strain insertion material, 210 Zinc characteristics of, 180–181 chloride, 144–145 electrodeposition, 134, 140, 144–145 Zinc-Nickel (Zn-Ni) alloys, 146 Zwitterionic liquids, see Zwitterions alkali metal ionic liquids (AMILs), 320–321 characteristics of, 301–302, 313 heating process, 304–305, 307–308 ionic (ZILs), 321, 323, 404, 433 poly, 437–439 thermal stability, 404, 406 Zwitterions characteristics of, 282, 302 ammonium, 303 ionic conductivity, 311–312 liquid/polymer gels, 403–407 melting points of, 305, 308–310 molten salts, 401 synthesis of, 302–304 thermal behavior of, 310

1/20/2011 10:26:35 AM