ELECTROCHEMICAL ASPECTS OF IONIC LIQUIDS ppt

Basic and Possible Characteristics of Organic Ionic Liquids Low melting point • Treated as liquid at ambient temperature • Wide usable temperature range Nonvolatility • Thermal sta

ELECTROCHEMICAL

ASPECTS OF IONIC

LIQUIDS

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 specifi cally disclaim any implied

warranties of merchantability or fi tness 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 profi t 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.

10 9 8 7 6 5 4 3 2 1

PREFACE TO THE SECOND EDITION ix

CONTRIBUTORS xv

1 Importance and Possibility of Ionic Liquids 1

Hiroyuki Ohno

2 Physical Chemistry of Ionic Liquids: Inorganic and Organic

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

Hiroyuki Ohno, Masahiro Yoshizawa-Fujita, and Tomonobu Mizumo

Hiroyuki Ohno and Kyoko Fujita

Toshio Fuchigami and Shinsuke Inagi

9 Electrodeposition of Metals in Ionic Liquids 129

Yasushi Katayama

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

Kyoko Fujita and Hiroyuki Ohno

18 Actuators 271

Kinji Asaka

Rika Hagiwara and Kazuhiko Matsumoto

Hiroyuki Ohno

Masahiro Yoshizawa-Fujita, Asako Narita, and Hiroyuki Ohno

Wataru Ogihara, Masahiro Yoshizawa-Fujita, and Hiroyuki Ohno

Tomonobu Mizumo and Hiroyuki Ohno

24 Electric Conductivity and Magnetic Ionic Liquids 337

Gunzi Saito

PART V IONIC LIQUIDS IN ORDERED

STRUCTURES 347

25 Ion Conduction in Organic Ionic Plastic Crystals 349

Maria Forsyth, Jennifer M Pringle, and Douglas R MacFarlane

Takashi Kato and Masafumi Yoshio

Kenji Hanabusa

Masahiro Yoshizawa-Fujita and Hiroyuki Ohno

Naomi Nishimura and Hiroyuki Ohno

viii CONTENTS

Hiroyuki Ohno and Masahiro Yoshizawa-Fujita

Hiroyuki Ohno, Masahiro Yoshizawa-Fujita, and Wataru Ogihara

Masahiro Yoshizawa-Fujita and Hiroyuki Ohno

Hiroyuki Ohno

INDEX 465

PREFACE TO THE SECOND EDITION

ix

The fi rst edition of this book was published in 2005 as the fi rst 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

indus-try, 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

fi nding 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 suffi cient

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 fi nd some interesting point regarding ionic liquid science

that aids in their research

HIROYUKI OHNO

PREFACE TO THE FIRST EDITION

xi

This book introduces some basic and advanced studies on ionic liquids in the

electrochemical fi eld 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 fi eld 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 fi eld

The applications of ionic liquids in this book have been narrowed to the

latest results of electrochemistry For this reason only the results on room

temperature ionic liquids are presented, and not on high - temperature melts

The reader of this book should have some basic knowledge of

electrochem-istry Those who are engaged in work or study of electrochemistry will get to

know the great advantages of using ionic liquids Some readers may fi nd the

functionally designed ionic liquids to be helpful in developing novel materials

not only in electrochemistry but also in other scientifi c fi elds This book covers

a wide range of subjects involving electrochemistry Subjects such as the

solu-bilization of biomolecules may not seem to be necessary for electrochemistry

concerning ionic liquids, but some readers will recognize the signifi cance of

solubility control of functional molecules in ionic liquids even in an

electro-chemical fi eld Many more examples and topics on ionic liquids as solvents

have been summarized and published elsewhere, and the interested reader will

benefi t from studying the references that are provided at the end of each

chapter

Hiroyuki Ohno

ACKNOWLEDGMENTS FOR THE

SECOND EDITION

xiii

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

H iroyuki O hno

Maria Forsyth, Department of Materials Engineering, Monash University

Toshio Fuchigami, Department of Electronic Chemistry, Tokyo Institute of

Agriculture and Technology

Institute and School of Materials, Arizona State University

University

Kenji Hanabusa, Graduate School of Science and Technology, Shinshu

University

Akitoshi Hayashi, Department of Applied Chemistry, Graduate School of

Engineering, Osaka Prefecture University

Technology

Noritaka Iwai, Department of Bioengineering, Tokyo Institute of Technology

Yasushi Katayama, Department of Applied Chemistry, Faculty of Science and

Technology, Keio University

Engineering, The University of Tokyo

School of Engineering, Kyushu University

Technology

University

Douglas R MacFarlane, School of Chemistry, Monash University

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

Engineering, Kyoto University

Agriculture and Technology

Akihiro Noda, Honda R & D Co., Ltd

Wataru Ogihara, Nissan Motor Co., Ltd

Agriculture and Technology

Chemistry, Monash University

Gunzi Saito, Research Institute, Meijo University

CONTRIBUTORS xvii

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

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

Yokohama National University

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

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

ordi-nary 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

prop-erties 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

dif-ferent components So there are unlimited possibilities open to the new fi eld

of ionic liquids The most compelling idea is that ionic liquids are “ designable ”

or “ fi ne - tunable ” Therefore, we can easily expect explosive developments in

fi elds 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.

2 IMPORTANCE AND POSSIBILITY OF IONIC LIQUIDS

1.2 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

proper-ties of molecular solvents and molten salts have already been introduced and

summarized in several books Thousands of papers have already been

pub-lished on their outstanding characteristics and effectiveness for a variety of

fi elds 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 fl ame retardancy The organic liquids

with both high ionic conductivity and fl ame retardancy are practical materials

for use in electrochemistry At the same time, the fl ame retardancy based on

nonvolatility inherent in ion conductive liquids opens new possibilities in other

fi elds 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

develop-ment 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

TABLE 1.1 Basic and Possible Characteristics of Organic Ionic Liquids

Low melting point • Treated as liquid at ambient temperature

• Wide usable temperature range

Nonvolatility • Thermal stability

• Flame retardancy

Composed by ions • High ion density

• High ion conductivity

Organic ions • Various kinds of salts

• Designable

• Unlimited combinations

ion conductive materials The aqueous salt solution is one of the best

electro-lyte 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

misunder-standing that the ionic liquids are never vaporized and are stable even when

on fi re 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 fi nishing

their role? It is not diffi cult to design novel ionic liquids that can be

decom-posed 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

electro-lyte solutions The astute reader will fi nd the future prospects for ionic liquids

between the lines in all chapters of this book

2

5

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) fi eld ( < 100 ° C ILs), the subject of the present volume, is still under

development The many interesting studies on transport and thermodynamic

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

prop-erties 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 fi rst 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.

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

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 effi ciently 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 fi rst industrial IL process

[1b] , because of the new fi nding that its members can have aqueous solution

like conductivities [17] and can serve as novel electrolytes for fuel cells [18] ,

and fi nally, because of the evidence that these liquids, in hydrated form, can

be used as tunable solvents for biomolecules, on which stability against

aggre-gation and hydrolysis may be provided under the right tuning [19] This class

is closely related to the fi rst but differs in that the cation has been formed by

transfer of a proton from a Br ø nsted acid to a Br ø nsted 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 fl uid If the gap is large, as in the case of ammonium nitrate, 87 kJ/mol (from

data for HNO 3 + NH 3 → NH 4 NO 3 ), 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 trifl ic acid, HSO 3 CF 3, 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

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 fl uid than the aprotic ionic liquids [17] , most likely

because of their generally reduced ionicity (see Section 2.6 )

CLASSES OF IONIC LIQUIDS 7

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

mis-match of large anions like tetrachloroaluminate or iodide with small cations

like Li + The eutectic in the system LiAlCl 4 – LiAlI 4 system, for instance, lies at

65 ° C, and the liquid is highly fl uid — more fl uid 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

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 fi rst “ ionic

liquids ” made were probably of this type, namely mixtures of ammonium salts

Figure 2.1 Correlation of the excess boiling point (determined at the 1 : 1 composition)

with the difference in aqueous solution pK a values for the component Br ø nsted acids

and bases of the respective ionic liquids Δ pK a The Δ T b value is determined as the

dif-ference between the measured boiling point and the value, at 1 : 1, of the linear

con-nection 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, α Pic =

α - picolinium = 2 - methylpyridinium = 2MPy, FA = formate, TFAc = trifl uoroacetate,

Tf = trifl ate = trifl uoromethanesulfonate) (from Yoshizawa et al [21] ) Data for the

three protic nitrates of [17] (ethylammonium nitrate, dimethylammonium nitrate, and

methylammonium nitrate) fi t precisely on this diagram

0 50 100 150 200 250 300 350 400 450 500

PA-TFAc

αPic-Tf

PA-Tf

EA-FA

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

the latter is on record, but its melting point is unknown We have obtained a

components [26] , but its conductivity is too low to class it as an ionic liquid

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 fi eld cations

and have no solvent function Such “ molten solvates ” have low vapor pressures

instance, LiZnBr 4 · 3H 2 O, has T b = 190 ° C, whereas T g = − 120 ° C [34]

To provide a better perspective on ionic liquids, we fi rst make some

obser-vations on inorganic salts and the factors that make it possible to observe them

as ionic liquids below 100 ° C

Figure 2.2 Phase diagram of the system LiAlCl 4 + LiAlI 4 (from Lucas et al [22] )

showing ionic liquid domain (T l < 100 ° C) in the middiagram

LiAlCl

0 0 50 100 150 200 250

65

20 40 60 80 100

Mol% LiAlI4

LOW-TEMPERATURE LIQUID BEHAVIOR OF IONIC MELTS 9

2.2 LOW - TEMPERATURE LIQUID BEHAVIOR OF IONIC MELTS

Most inorganic salts, when they melt, are found to fl ow 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

occupy much more volume in the liquid state than KI, but they melt at much

viscos-ity temperature dependence according to any but the most precise

measure-ments It is only with deep eutectics like those for the ternary systems

LiNO 3 – NaNO 3 – KNO 3 ( T E = 143 ° C) and LiNO 3 – NaNO 2 – KNO 3 ( T E = 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

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

tem-perature range at T < 2 T 0 [6] , where T 0 is the theoretical low - temperature limit

extrapola-tions 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 fl uid fl ow

would diverge [7,35] The practical low - temperature limit to the liquid state

T 0 , by an amount that depends on the liquid fragility Typically T g / T 0 = 1.2 – 1.3,

unless the liquid is “ strong ” [35b] So the low - temperature domain is entered

at T g / T ≈ 0.53 – 0.66 For Arrhenius behavior, which represents the “ strong ”

liquid limit of behavior is rarely observed, T 0 = OK and T g / T 0 = ∞ The upper

end of this range, T g / T = 0.66, is the number usually associated with the ratio

of T g / T m 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

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 T g / T = 0.625

The deviations from Arrhenius behavior observed in the low - temperature

parameter Vogel – Fulcher – Tammann (VFT) equation [37] in the modifi ed

form:

where η 0 , D , and T 0 are constants, and T 0 , the vanishing mobility temperature,

of magnitude) are included in the data fi tting The different curvatures in

Figure 2.3 then are reproduced by variations in the parameter D of equation

(2.1) [6b]

Figure 2.3 T g - scaled Arrhenius plot showing data for molten salts ZnCl 2 and calcium

potassium nitrate (CKN), with data for the calcium nitrate hydrate (CaNO 3 · 8H 2 O)

and the tetrafl uoroborates of quaternary ammonium (MOMNM 2 E, 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] )

1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0 -5 -3 -1 1 3 5 7 9 11 13

MOMNMe2EBF4 BMIBF4 BMIBOB

STRONG

LOW-TEMPERATURE LIQUID BEHAVIOR OF IONIC MELTS 11

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 (defi ned by easily recognized

“ super - Arrhenius ” transport behavior) — that is, they melt within the same

range that is generally diffi cult 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 superfi cially 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

rapidly to avoid crystallization What is interesting is that the conductivity at

IL measured at its glass temperature (which requires special equipment) The

explanation for this dramatic difference lies in the ability of the small cation,

Figure 2.4 Arrhenius plots for conductivity of lithium haloaluminate or

pseudohalo-aluminate melts, showing super - Arrhenius conductivity that remains high at the glass

temperature because of decoupling of Li + motion from its surroundings, in the “ low

temperature 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 T g and the structure becomes fi xed (from Lucas et al [22] )

20 –7 –6 –5 –4 –3

–1 ) –2 –1

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

some-what 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 ( G liq < G crys) at low temperatures, relative to their cohesive

energies

The simplest explanation that can be given for this circumstance focuses on

the diffi culty that the more complex and size - mismatched ions characteristic

of IL and inorganic salt hydrates fi nd in packing closely This causes the

cohe-sive energy (zero in the gas phase) to be fi xed at a value less negative than

would be obtained if the centers of attraction could be packed together more

closely The lattice energies refl ecting 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 fi ctitious substance, along with that of the liquid

phase, over a range of temperatures The crystals are supposed to be

noncon-vertible except to the liquid and are supposed to differ only in their lattice

energies (lattice energy, E L = 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 verifi ed using data for isomers

of the same substance (e.g., xylene in which all three isomers o - , m - , and p -

have 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

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,

< 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] )

ELECTRICAL CONDUCTIVITY AND LOW VAPOR PRESSURE 13

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

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 effi cient

packing in three - dimensional (3D) order This improves the competitive status

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 effi cient in minimizing the

Figure 2.5 Gibbs free energies for a system in which there are multiple

nonintercon-vertable crystalline phases, each with a different lattice energy ( E L = 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

glass

liquid

Tmelt

XtI1XtI2XtI3XtI4XtI5

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 diffi cult

to estimate quantitatively, depending on details of crystal symmetry [44] , and

direct calculations for isotropic phases are not available to the authors ’

knowl-edge 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 refl ection 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

fl uctuation h suffi cient to permit an ion pair to escape from the bulk liquid

into the vapor is proportional to the Boltzmann factor:

molecule (ion pair), which is dominated by the Madelung energy

In an ideal quasilattice, no ion - pairs can be distinguished over statistically

signifi cant time intervals because all sites are equivalent This is also the

condi-tion that leads to the maximum ionic conductivity for a given fl uidity (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 fl uid fl ow without an ionic current fl ow

These relations are summed up in some classical equations of

electrochem-istry, which were derived by considering dilute aqueous solutions in which

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,

outline in the following list

The classical equations to which we refer are as follows:

1 The Nernst – Einstein equation connecting diffusion and partial

equiva-lent ionic conductivity λ i ,

where F is the Faraday (charge per equivalent)

ELECTRICAL CONDUCTIVITY AND LOW VAPOR PRESSURE 15

2 The Stokes – Einstein equation connecting diffusivity D i of ionic species

i of charge z i and radius r i , with the viscosity η of the medium in which the diffusion is occurring,

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

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

fl uidity are chosen to be reciprocal poise and when those for equivalent

in Figure 2.6

To fi x 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

dissoci-ated and to have ions of equal mobility [46] We have included some data for

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]

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

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 classifi cation 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 fl uidities are high, but

the conductivity is much lower than it would be if all moving particles were

cations or anions

Among the subionic liquids are the cases of proton transfer salts in which

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

elec-trolytes 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

Figure 2.6 Walden plot for tetrafl uoroborate salts of various cations, showing subionic

(or “ supercoupled ” ) behavior of the salt of the methoxymethyldimethylethyl

ammo-nium cation MOMNM 2 E + By contrast, the salt of MOENM 2 E + may become decoupled

at low temperature The dotted lines passing through the LiAlCl 4 points and the

AgCl – AgI – CsCl points are the fi ts 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 classifi cation 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)

-5 -4 -3 -2 -1 0 1 2 3

Poor ionic liquids

Non-ionic liquids

Superionic liquids Superionic glasses

High vapor pressures

MOENM2E-BF4 MOMNM2E-BF4 BPy-BF4 BMI-BF4 LiAlCl4 35AgCl-45AgI-20CsCl

Good ionic liquids

ELECTRICAL CONDUCTIVITY AND LOW VAPOR PRESSURE 17

in Figure 2.4 would provide a good example of this behavior, but unfortunately,

fl uidity data are not available to complete the plot Instead, we show classic

halide ionic liquid for which fl uidity and conductivity data were reported by

McLin and Angell [49] In this case, the silver cation can slip through the

chan-nels 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

been defi ned from the ratio of the conductivity and fl uidity relaxation times

Walden rule,

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

fl uidity 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 confi gurational excitation,

is distributed in ionic liquids This effect, which leads all except infi nitely dilute

solutions to have limiting high - temperature conductivities that are smaller

than expected from their fl uidities, is well illustrated in Figure 2.7 [17] , so that

comparison can be made with the high temperature limit in Figure 2.7 b

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 2.7 a, 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

ter-minates at the universal high T limit for fl uidity implied by Figure 2.3 , namely

at 10 4.5 poise

Figure 2.7 b shows the behavior, in scaled Arrhenius form, of the

conductiv-ity for these systems, to be discussed later on In principle, the conductivconductiv-ity of

the infi nitely dilute aqueous solution, in which the ions move without any

infl uence on each other, should have the same value at 1/ T = 0 as does the

aqueous KCl solutions at infi nite 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 T g / T < 0.5 (see Fig 2.3 ), we see that this

correspondence is almost realized Certainly, the infi nitely dilute solution (in

which cation – anion friction is absent) has a much higher equivalent

conductiv-ity at the high - temperature limit than do any of the other systems

–1 0 1 2 3 4

log[ η –1 (Poise –1 )]

35AgCl-45AgI-20CsCl LiCl-6H2O

7:3(mol)I-II [EtNH3][NO3]

[BMIM][BF4]

[EtNH3][HCO2]

Ideal line

superionic liquids

subionic liquids (high vapor pressures)

–1 0 1 2 3 4

Tg/T

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)

2 mol –1 ) (a)

(b)

ELECTRICAL CONDUCTIVITY AND LOW VAPOR PRESSURE 19

Counted among these superionic electrolytes are the solutions of strong

acids in water and other hydrogen - bonded solvents (e.g., glycerol, and

hydro-gen 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 bifl uoride,

which equals 7.7 M lithium chloride (LiCl · 6H 2 O) [53] , suggests that some

salts with protons in both cation and anions may have superprotonic

Even without superprotonic con duction, it seems that protic ionic liquids,

which are mostly neutral and noncorrosive, can serve as fuel cell

elec-trolytes with unusual performance properties, as we describe in Section 2.6

Figure 2.7 (a) Relation of equivalent conductivity to fl uidity 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 fl uidity because the only force

imped-ing the motion of an ion under fi xed potential gradient is the viscous friction The

position of the ideal line is fi xed by data for 1 M aqueous KCl solution at ambient

temperature The data for LiCl · 6H 2 O 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 fi gure A “ fractional Walden rule ” Λ η α = 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) T g - scaled Arrhenius plot to display the temperature dependence of the

equiv-alent conductivity in relation to the temperature at which the fl uidity 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 interaction

free) behavior for conductivity is shown by the dashed line plot of infi nite dilution

conductivities for KCl in H 2 O (data from Robinson and Stokes [46] ) To include these

data, we assign T g as 138 K because high - temperature viscosity fi tting suggests it The

real value for water is controversial Notation: I - [MeNH 3 ][NO 3 ], II - [Me 2 NH 2 ][NO 3 ]

(From Xu and Angell [17] by permission of the AAAS)

after considering the cohesion factors that can be important in the

determina-tion 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 - fl uidity and high - conductivity media at low temperatures It would seem

that increasing the size of the ions would be the way to achieve high fl uidities

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

Figure 2.8 Specifi c conductivities of protic versus aprotic ionic liquids, showing

match-ing 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)

–7 –6 –5 –4 –3 –2 –1 0

4:6 (mol)V-II [EtNH3][NO3]

[N1-O-2,211][NO3]

COHESION AND FLUIDITY: THE TRADE-OFF 21

Figure 2.9 Dependence of the cohesion of salts of weakly polarizable cations and

anions, assessed by T g value, on the ambient temperature molar volume V m , hence on

interionic spacing d A broad minimum in the ionic liquid cohesive energy is observed

at a molar volume of 250 cm 3 mol − 1 , which corresponds to an interionic separation of

about 0.6 nm, assuming face - centered cubic packing of anions about cations The lowest

T g in the plot probably should be excluded from consideration because of the nonideal

Walden behavior for this IL ( MOMNM E BF2 + 4−) [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.)

-140 -120 -100 -80 -60 -40 -20 0 20

I II

III IV

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 perfl uorinated species among the anions of

ionic liquids used in practice However, perfl uorinated anions are expensive

to manufacture, and even perfl uorinated species begin to cause fl uidity

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 fl uorinated 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

Because anions less polarizable than the perfl uorinated species represented

in Figure 2.9 cannot be made, it would seem that if fl uorinated species are to

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 fl uidities 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 confi guration), then the

The lowest T g s are obtained with the four - bead anion BF4−, which is perfl

and two - bead anions, NO3−, formate HCO2−, 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

systems opens up, and the maximum ambient temperature ionic conductivity

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 - trifl uoromethanesulfonyl 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 fl owing, 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 fl owing was subject to occasional

PROTON TRANSFER IONIC LIQUIDS AS NOVEL FUEL CELL ELECTROLYTES 23

large fl uctuations, 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 Tefl on sandwich cell for

differ-ent 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

eth-ylammonium 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 - fl uoropyridine and trifl ic acid [59]

As shown in Figure 2.12 , this performance is exceeded only when 6 M trifl ic

acid is used as electrolyte (in the same cell) Unfortunately, the performance

of the cell with the 2 - fl uoropyridium trifl ate 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

cor-rosive 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

Figure 2.10 Potential of the simple bubbler type fuel cell, under load, using

ethylam-monium nitrate as electrolyte A comparison is made with the behavior of the identical

cell with phosphoric acid substituted for the ethylammonium nitrate electrolyte

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 1.3

H3PO4 98% - 100 °C

Current Density (mA/cm2)

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

[EtNH3][NO3] - 100 °C

2.7 IONICITY AND ACIDITY OF PIL S : 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

Figure 2.11 Tafel plot for a gas diffusion Tefl on sandwich fuel cell

1.3 1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

1.3 1.2 1.1 1.0 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 0.0

IONICITY AND ACIDITY OF PILS 25

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 confi rmed for

several cases by direct electrochemical interrogation under strictly anhydrous

conditions [61]

assem-bled 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 suffi cient ionicity that the Walden rule is a good

approximation to the behavior Accordingly, it is shown that “ good ” ionic

combinations It also is shown that, with an appropriate choice of couples, PILs

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

pro-cesses, do not extract a proton from water to give a basic PIL as might have

Figure 2.12 Tafel plots for Tefl on sandwich type H 2 /O 2 gas fuel cell using E - tech

elec-trodes and liquid electrolytes [59] The 2 - fl uoropyridine - HTf average pK a places it in

the acid electrolyte domain of Fig 2.13 , above the level for phosphoric acid

been expected However, a superacidic PIL, based on proton transfer to

pen-tafl uoropyridine, (pK a = – 13) from HTFSI, has been identifi ed positively and

is being described elsewhere [33] Separate experiments have yielded similar

bubbling over anhydrous AlCl 3 [26]

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 pK a of the couple between which the

proton transfer occurs, the more acidic is the PIL The parentheses in “ (HAlCl 4 ) ” in

this diagram indicate that the free acid does not exist

Occupied

Acid Electrolytes

Neutral Electrolytes

Basic Electrolytes Super-Acids

Super-Bases

Vacant HSbF 6

(HAlCl 4 ) HTFSI

SbF 6 TFSI- HSO 3 F

-pFPyH +

HTf HClO 4

H 2 SO 4

HPO 2 F 2

HNO 3

CH 3 SO 3 H 2-Fluoropyridine H +

CF 3 COOH

SO 3

F-–10 0.59 –9 0.53 –1.3 0.08

HSO 4

-PO 2 F 2

-NO 3

-CH 3 SO 3- (MS) 2-Fluoropyridine

-COMPARISON OF HIGHEST-CONDUCTING ORGANIC 27

A case of special interest to us is that of the basic molecule guanidine and

its protonated conjugate — the simple, symmetrical, and resonance - stabilized

guanidin-ium salts [62] and have studied some of their binary solutions Their high

conductivities provide the material for our fi nal 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

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

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 vitrifi ed 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

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 ( T g ≈ − 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

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

explo-sive liquids and hence are not destined for practical applications These high

conductivity results are all correlated with high fl uidities A Grotthus

mecha-nism might make a small contribution to conductivity in the case of

guani-dinium cations

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, classifi ed in Figure 2.13 for the case of protic ionic

liquids Because of the preoccupation of the fi eld 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]

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 fi gure