Direct Methanol Fuel Cells [1 ed.] 9781611224474, 9781608768653

Citation preview

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

ENERGY SCIENCE, ENGINEERING AND TECHNOLOGY SERIES

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

DIRECT METHANOL FUEL CELLS

No part of this digital document may be reproduced, stored in a retrieval system or transmitted in any form or by any means. The publisher has taken reasonable care in the preparation of this digital document, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained herein. This digital document is sold with the clear understanding that the publisher is not engaged in rendering legal, medical or any other professional services.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

ENERGY SCIENCE, ENGINEERING AND TECHNOLOGY SERIES Oil Shale Developments Ike S. Bussell (Editor) 2009. ISBN: 978-1-60741-475-9 Power Systems Applications of Graph Theory Jizhong Zhu 2009. ISBN: 978-1-60741-364-6 Bioethanol: Production, Benefits and Economics Jason B. Erbaum (Editor) 2009. ISBN: 978-1-60741-697-5

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Bioethanol: Production, Benefits and Economics Jason B. Erbaum (Editor) 2009. ISBN: 978-1-61668-000-8 (Online Book) Introduction to Power Generation Technologies Andreas Poullikkas 2009. ISBN: 978-1-60876-472-3 Handbook of Exergy, Hydrogen Energy and Hydropower Research Gaston Pélissier and Arthur Calvet (Editors) 2009. ISBN: 978-1-60741-715-6 Energy Costs, International Developments and New Directions Leszek Kowalczyk and Jakub Piotrowski (Editors) 2009. ISBN: 978-1-60741-700-2

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Radial-Bias-Combustion and Central-Fuel-Rich Swirl Pulverized Coal Burners for Wall-Fired Boilers Zhengqi Li 2009. ISBN: 978-1-60876-455-6 Syngas Generation from Hydrocarbons and Oxygenates with Structured Catalysts Vladislav Sadykov, L. Bobrova, S. Pavlova, V. Simagina, L. Makarshin, V. Julian, R. H. Ross, and Claude Mirodatos 2009. ISBN: 978-1-60876-323-8

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Corn Straw and Biomass Blends: Combustion Characteristics and NO Formation Zhengqi Li 2009. ISBN: 978-1-60876-578-2 Computational Techniques: The Multiphase CFD Approach to Fluidization and Green Energy Technologies (includes CD-ROM) Dimitri Gidaspow and Veeraya Jiradilok 2009. ISBN: 978-1-60876-024-4 Air Conditioning Systems: Performance, Environment and Energy Factors Tobias Hästesko and Otto Kiljunen (Editors) 2009. ISBN: 978-1-60741-555-8 Cool Power: Natural Ventilation Systems in Historic Buildings Carla Balocco and Giuseppe Grazzini 2010. ISBN: 978-1-60876-129-6

A Sociological Look at Biofuels: Understanding the Past/Prospects for the Future Michael S. Carolan 2010: ISBN: 978-1-60876-708-3

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Ethanol Biofuel Production Bratt P. Haas 2010: ISBN: 978-1-60876-086-2 Thin-Film Solar Cells Abban Sahin and Hakim Kaya (Editors) 2010. ISBN: 978-1-60741-818-4

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

CFD Modeling and Analysis of Different Novel Designs of Air-Breathing Pem Fuel Cells Maher A.R. Sadiq Al-Baghdadi 2010. ISBN: 978-1-60876-489-1 Bio Resource-Based Energy for Sustainable Societies K. A. Vogt, D. J. Vogt, M. Shelton, R. Cawston, L. Nackley, J. Scullion, M. Marchand, M. Tulee, T. Colonnese, L. Stephan, S. Doty, R. K. Hagmann, T. C. Geary, T. A. House, I. Nwaneshiudu, P. A. Roads, S. Candelaria, L. L. James, E. Kahn, L.X. Lai, A. M. Lee, S. J. Rigdon, and R.M. Theobald 2010. ISBN: 978-1-60876-803-5 Jatropha Curcas as a Premier Biofuel: Cost, Growing and Management Claude Ponterio and Costanza Ferra (Editors) 2010. ISBN: 978-1-60876-003-9 Direct Methanol Fuel Cells A. S. Arico, V. Baglio and V. Antonucci 2010. ISBN: 978-1-60876-865-3

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

ENERGY SCIENCE, ENGINEERING AND TECHNOLOGY SERIES

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

DIRECT METHANOL FUEL CELLS

ANTONINO S. ARICÒ VINCENZO BAGLIO VINCENZO ANTONUCCI

Nova Science Publishers, Inc. New York

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010 by Nova Science Publishers, Inc. All rights reserved. No part of this book may be reproduced, stored in a retrieval system or transmitted in any form or by any means: electronic, electrostatic, magnetic, tape, mechanical photocopying, recording or otherwise without the written permission of the Publisher. For permission to use material from this book please contact us: Telephone 631-231-7269; Fax 631-231-8175 Web Site: http://www.novapublishers.com

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

NOTICE TO THE READER The Publisher has taken reasonable care in the preparation of this book, but makes no expressed or implied warranty of any kind and assumes no responsibility for any errors or omissions. No liability is assumed for incidental or consequential damages in connection with or arising out of information contained in this book. The Publisher shall not be liable for any special, consequential, or exemplary damages resulting, in whole or in part, from the readers‘ use of, or reliance upon, this material. Independent verification should be sought for any data, advice or recommendations contained in this book. In addition, no responsibility is assumed by the publisher for any injury and/or damage to persons or property arising from any methods, products, instructions, ideas or otherwise contained in this publication. This publication is designed to provide accurate and authoritative information with regard to the subject matter covered herein. It is sold with the clear understanding that the Publisher is not engaged in rendering legal or any other professional services. If legal or any other expert assistance is required, the services of a competent person should be sought. FROM A DECLARATION OF PARTICIPANTS JOINTLY ADOPTED BY A COMMITTEE OF THE AMERICAN BAR ASSOCIATION AND A COMMITTEE OF PUBLISHERS. LIBRARY OF CONGRESS CATALOGING-IN-PUBLICATION DATA Aricr, A. S. (Antonio Salvatore) Direct methanol fuel cells / A.S. Arico, V. Baglio, and V. Antonucci. p. cm. Includes bibliographical references and index.

ISBN:  (eBook)

1. Proton exchange membrane fuel cells. 2. Direct energy conversion. 3. Methanol as fuel. I. Baglio, V. (Vincenzo) II. Antonucci, V. (Vincenzo) III. Title. TK2933.P76A75 2009 621.31'2429--dc22 2009050556

Published by Nova Science Publishers, Inc.  New York Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

CONTENTS

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Preface

ix

Chapter 1

Introduction

1

Chapter 2

Basic Aspects of Direct Methanol Fuel Cells

3

Chapter 3

Development of Direct Methanol Fuel Cells Components and Devices

19

Chapter 4

DMFC Applications

39

Chapter 5

Fundamental Aspects: Status of Knowledge

73

Chapter 6

Technology Development

129

Chapter 7

Techno-Economical Challenges

145

Chapter 8

Conclusion

149

Acknowledgments

153

References

155

Index

169

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

PREFACE

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

This book deals with an analysis of materials issues, status of technologies and potential applications of direct methanol fuel cells. The principle of operation of direct methanol fuel cells and the status of knowledge in the basic research areas are presented. The technology of direct methanol fuel cells is discussed in this book with particular regard to fabrication methodologies for the manufacturing of catalysts, electrolytes membrane-electrode assemblies, stack hardware and system design.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

ABSTRACT The candidate fuel for fuel cells is usually considered hydrogen. However, at present, no suitable large-scale infrastructure exists for hydrogen production, storage and distribution. Significant efforts have been addressed in the last decades to the direct electrochemical oxidation of alcohol and hydrocarbon fuels. Organic liquid fuels are characterized by high energy density whereas the electromotive force associated to their electrochemical combustion to CO2 is comparable to that of hydrogen combustion to water. Among the liquid organic fuels, methanol has promising characteristics in terms of reactivity at low temperatures, storage and handling. Accordingly, a methanol-feed proton exchange membrane fuel cell would help to alleviate some of the issues surrounding fuel storage and processing for fuel cells. Technological improvements in direct methanol fuel cells (DMFCs) are thus fuelled by their perspectives of applications in portable, transportation and stationary systems especially with regard to the remote and distributed generation of electrical energy. This book deals with an analysis of materials issues, status of technologies and potential applications of direct methanol fuel cells. The principle of operation of direct methanol fuel cells and the status of knowledge in the basic research areas are presented. The technology of direct methanol fuel cells is discussed with particular regard to fabrication methodologies for the manufacturing of catalysts, electrolytes membraneelectrode assemblies, stack hardware and system design.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Chapter 1

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

1. INTRODUCTION Fuel cells have reached a mature level of technology. These systems appear now ready for electro-traction, portable power sources, distributed power generation and stationary applications [1]. The high thermodynamic efficiency and the near zero emission levels make them an attractive alternative to internal combustion engines, batteries and thermal combustion power plants. Like storage batteries, fuel cells deliver energy by consuming electroactive chemicals, but differ significantly in that these chemicals are delivered on-demand to the cell. As a result, a fuel cell can generate energy continuously and for as long as the electroactive chemicals are provided to the cell. The candidate fuel for fuel cells is usually considered hydrogen. However, at present, no suitable large-scale infrastructure exists for hydrogen production, storage and distribution. Significant efforts have been addressed in the last decades to the direct electrochemical oxidation of alcohol and hydrocarbon fuels. Organic liquid fuels are characterized by high energy density, whereas, the electromotive force associated to their electrochemical combustion to CO2 is comparable to that of hydrogen combustion to water [1-3]. Among the liquid organic fuels, methanol has promising characteristics in terms of reactivity at low temperatures, storage and handling. Accordingly, a methanol-feed proton exchange membrane fuel cell would help to alleviate some of the issues surrounding fuel storage and processing for fuel cells. Technological improvements in direct methanol fuel cells (DMFCs) are thus fuelled by their perspectives of applications in portable, transportation and stationary systems especially with regard to the remote and distributed generation of electrical energy [4-5]. Methanol is cheap and it can be distributed by using the present infrastructure for liquid fuels. It can be obtained from fossil fuels, such as natural gas or coal, as well as from sustainable sources through fermentation of agricultural products and from biomasses. With respect to ethanol, methanol has

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

2

A. S. Aricò, V. Baglio and V. Antonucci

the significant advantage of high selectivity to CO2 formation in the electrochemical oxidation process [1-3]. Thus, the direct methanol fuel cell is a technology receiving a great deal of attention because of specific advantages it has over hydrogen-based fuel cell systems. As above mentioned, hydrogen-based systems for mobile applications require new technologies and new infrastructure for fuel storage, delivery, and handling safety. Liquid fuels can greatly simplify handling, furthermore, they are characterized by energy density larger than hydrogen stored as a compressed gas or metal hydride. Alternatively, hydrogen can be supplied by converting hydrocarbon and alcohol fuels, but a DMFC still presents a much simpler system for mobile applications by eliminating the need for a fuel processor. However, despite these practical system benefits, DMFCs are characterized by a significantly lower power density and lower efficiency than a polymer electrolyte fuel cell (PEMFC) operating with hydrogen because of the slow oxidation kinetics of methanol and methanol cross-over from the anode to the cathode [1-3]. This book deals with an analysis of the history, current status of technology, potential applications and techno-economic challenges of direct methanol fuel cells. The basic aspects related to DMFC operation are presented with particular regard to thermodynamics, performance, efficiency and energy density characteristics. The historical development of DMFC devices and components is analyzed with special regard to the study of catalysts and electrolytes. The status of knowledge in the basic research areas is presented and particular emphasis is given to required breakthroughs. The section on fundamentals is focused on the electrocatalysis of the methanol oxidation reaction and oxygen electro-reduction. To this regard, particular relevance is given to the interpretation of the promoting effects for methanol oxidation and on the features that govern methanol tolerance for oxygen reduction catalysts. The technology section deals with the fabrication methodologies for the manufacturing of membrane-electrode assemblies, stack hardware and system design. The recent efforts in developing DMFC stacks for both portable and electro-traction applications are reported.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Chapter 2

2. BASIC ASPECTS OF DIRECT METHANOL FUEL CELLS

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

2.1. FUEL CELL PROCESS The core of the present direct methanol fuel cells is a polymer electrolyte ion exchange membrane. The electrodes (anode and cathode) are in intimate contact with the membrane faces. A scanning electron micrograph of a DMFC MEA is shown in Figure 1. The electrodes usually consist of three-layers: catalytic layer, diffusion layer and backing layer, but there are also several different configurations. The catalytic layer is composed by a mixture of catalyst and ionomer and it is characterized by a mixed electronic-ionic conductivity. The catalysts are often based on carbon supported or unsupported PtRu and Pt materials at the anode and cathode, respectively. The membrane as well as the ionomer consists, in most cases, of a perfluorosulfonic acid polymer. The diffusion layer is usually a mixture of carbon and Polytetrafluoroethylene (Teflon). The hydrophobic properties of this layer are determined by the need to transport oxygen molecules to the catalytic sites at the cathode or to favor the escape of CO2 from the anode. The package formed by electrodes and membrane is called ―membrane and electrode assembly‖ (MEA). The overall thickness of this package is generally smaller than one millimeter. Each MEA forms a cell. Several cells are usually connected in series to form a fuel cell stack that is integrated in a system which contains the auxiliaries allowing stack operation and delivering of the electrical power to the external load.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

4

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Figure 1. SEM micrograph of a DMFC membrane and electrode assembly equipped with Nafion 112 membrane.

In the literature it is usually made a distinction between ―active‖ and ―passive‖ operation mode [5]. In the active mode, the auxiliaries such as pumps, blowers, sensors etc. are used to supply reactants and to control the stack operation in order to optimize working conditions. This allows to achieve the most appropriate electrical characteristics. In the passive mode, there are no energy consuming auxiliaries (excluding step-up DC/DC converters) and the reactants reach the catalytic sites by natural convection or by effect of the capillary forces or due to the concentration/partial pressure gradient. The system is more simple than in the active mode; no significant amount of power from the stack is thus dissipated on auxiliaries, but, the operating conditions may not be optimal to achieve the best efficiency and performance. DMFCs usually operate at temperatures below 100°C. This leads to significantly lower power density and lower efficiency than a PEMFC operating with hydrogen because of the slow oxidation kinetics of methanol, in this temperature range, and methanol crossover from the anode to the cathode. [2-4] Typical perfluorosulfonic acid membranes used in the present DMFCs, such as Nafion (DuPont), are permeable to methanol transport, thereby reducing significantly the fuel utilization efficiency of the device. An increase of the power density can be achieved by raising the fuel cell operating temperature to increase

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Basic Aspects of Direct Methanol Fuel Cells

5

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

the kinetics of methanol oxidation and to reduce the coverage of its oxidation intermediates on the anode catalyst. Unfortunately, typical membranes experience significant membrane dehydration at higher temperatures, leading to high ohmic drop and poor cell performance [2]. Furthermore, for some applications such as portable systems, low operating temperatures are preferable. There are alternative membranes that can extend the operating temperature above 100 °C such as phosphoric acid doped polybenzoimidazole or composite membranes [2]. Thus, it can be affirmed that polymer electrolyte membrane DMFCs can operate from ambient temperature up to 200 °C, depending on the membrane. The membrane determines the optimum performance in a specific temperature range. However, what is required for the DMFC is an optimal operation in a wide temperature range. In the presence of a protonic membrane, the DMFCs are directly fed with methanol/water mixture at the anode. Methanol is directly oxidized to carbon dioxide although it is not excluded the possible formation of compounds such as formaldehyde, formic acid and other organic molecules. The formation of such organic molecules decreases the fuel utilization in the process. A scheme of the overall reaction process occurring in a DMFC equipped with a proton conducting electrolyte is outlined below: CH3OH + H2O  CO2 + 6 H+ + 6e- (anode)

(1)

3/2 O2 + 6 H+ + 6e-  3H2O (cathode)

(2)

CH3OH + 3/2 O2  CO2 + 2H2O (overall)

(3)

In the presence of an alkaline electrolyte, this process can be written as follows: CH3OH + 6 OH-  CO2 + 5 H2O + 6e- (anode)

(4)

3/2O2 + 3H2O + 6e-  6OH- (cathode)

(5)

CH3OH + 3/2O2  CO2 + 2H2O (overall)

(3‘)

The thermodynamic efficiency of the process is given by the ratio between the Gibbs free energy, i.e. the maximum value of electrical work (G°) that can be obtained and the total available energy for the process, i.e. the enthalpy (H°). Under standard conditions:

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

6

A. S. Aricò, V. Baglio and V. Antonucci rev= G°/ H° ; reversible energy efficiency

(6)

with G°= H° – (TS°);

(7)

G°=-nFErev

(8)

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

and

Erev is the electromotive force. At 25°C, 1 atm and with pure oxygen feed the reversible potential for methanol oxidation is 1.18 V [3]. It does not vary significantly in the operating range 20°-130 °C and 1-3 bar abs. pressure. Usually, the open circuit voltage of a polymer electrolyte direct methanol fuel cell is significantly lower than the thermodynamic or reversible potential for the overall process. This is mainly due to methanol cross-over that causes a mixed potential at the cathode and to the irreversible adsorption of intermediate species at electrode potentials close to the reversible potential. The coverage of methanolic species is larger at high cell potentials i.e. at low anode potentials. This determines a strong anode activation control that the overall polarization curve (Figure 2). This can be observed in a polarization plot (Figure 2) where the terminal voltage of the cell is deconvoluted into the anode and cathode polarizations according to the equation: Ecell= Ecathode- Eanode.

(9)

The anodic potential, the cathodic potential and the terminal cell voltage can be measured simultaneously by using a proper reference electrode. This can be a dynamic hydrogen electrode (DHE) formed by two small circular pieces of electrodes, in close contact with the membrane; hydrogen is forced to evolve on one of these small electrodes, which is used as a reference, by applying a small current. These electrodes are electrically insulated from the anode and cathode current collectors but are humidified by draining a small amount of liquid from the main compartments through a channel (Figure 3). Alternatively, the anode polarization can be measured in the driven mode, in an independent experiment carried out in the same conditions of the overall polarization curve. The cathode curve is mathematically calculated from equation 9. In the driven mode hydrogen is fed to the cathode that acts as both counter and reference electrode. This is also the usual mode to carry out in-situ cyclic voltammetry experiments for the anode.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Basic Aspects of Direct Methanol Fuel Cells

7

1 T = 60°C

Cell

Potential / V

0.8

Anode Cathode

0.6

0.4

0.2

0 0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Current density / A cm -2

Figure 2. Single cell and in-situ half-cell electrode polarizations for a DMFC operating at 60 °C, ambient pressure, with 1 M methanol at the anode and air feed at the cathode.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Air

H2

Ref: H+/H2

H+/H2 DHE

CH3OH/H2O

CH3OH/H2O

Figure 3. Reference electrode configurations for in-situ half-cell electrode polarization measurements.

Besides the strong activation control at the anode, the effect of the mixed potential on the cathode polarization curve is clearly observed in Figure 2. The onset potential for the oxygen reduction in the presence of methanol cross-over is below 0.9 V versus the reversible hydrogen electrode (RHE). This is much lower

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

8

A. S. Aricò, V. Baglio and V. Antonucci

than the reversible potential for the oxygen reduction in the absence of methanol, i.e. 1.23 V vs. RHE. As it was pointed out above, such a result is mainly due to the effect of the mixed potential. In addition, the cathode polarization curve in the presence of cross-over does not present a clear sigmoidal shape as in hydrogenfed PEMFCs since the methanol adsorption on the cathode mainly influences the region of activation control for oxygen reduction. In fact, at high cathode potentials, oxygen reduction is slow and oxidation of methanol permeated through the membrane is enhanced by the elevated potential. The two opposite reactions compete each other and no spontaneous current is registered above 0.9 V (Figure 2). At high currents (Figure 2), both anodic and cathodic polarization curves show the onset of mass transport constraints due to the removal of the CO2 from the anode and the effect of flooding at the cathode. In the methanol fuel cell, the flooding of the cathode is not only due to the water formed by the electrochemical process; but, it especially occurs as a consequence of the fact that a liquid or a vapour (and not a humidified gas) is fed to the anode and this water/methanol mixture permeates through the hydrophilic membrane to the cathode. The polarization curves of a DMFC device can be registered at different temperatures in order to study in detail the activation behavior (Figure 4). It is clearly envisaged the presence of a strong effect of the temperature on the activation process. Thus, the temperature represents together with the methanol concentration the most important variables determining the performance and efficiency. According to the above considerations, the polarization behavior of a DMFC is predominantly controlled by the cross-over of methanol from anode to cathode and the slow kinetics of its electro-oxidation at the anode, when compared to that of a H2/air PEMFC [6-18]. Most of the present DMFCs are based on Pt-Ru and Pt as anode and cathode electrocatalysts, respectively. In the following, the typical polarization curve of a conventional DMFC based on such catalysts and Nafion membrane, as reported in Figure 2, is analyzed. The observed open circuit potentials, at present, are about 150 to 200 mV lower than those observed in PEMFCs, with similar oxygen partial pressure and cell temperature conditions, even though the reversible potential for methanol oxidation differs by only 40 mV from that of the reversible hydrogen electrode (RHE) [19]. In the temperature range of 60 °C -130 °C, the characteristic polarization curves of a DMFC show an open circuit voltage ranging from 0.7 to 0.9 V, depending on the electrocatalyst, membrane and operating conditions (Figure 4).

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Basic Aspects of Direct Methanol Fuel Cells

9

Cell Potential / V

0.8 T T T T T

0.6

= 90°C = 100°C = 110°C = 120°C = 130°C

0.4

0.2

Power Density / W cm

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

-2

0

T T T T T

0.5 0.4 0.3

= 90°C = 100°C = 110°C = 120°C = 130°C

0.2 0.1 0 0

0.5

1

1.5

2

2.5

3

-2

Current Density / A cm

Figure 4. Galvanostatic polarisation data for the DMFC equipped with CNR-ITAE Pt-Ru (anode) and 90% Pt/C (cathode) catalysts; 2 M CH3OH, oxygen feed, interdigitated flow field. Reprinted from Ref. [184] with permission from Elsevier.

A potential drop of about 0.15-0.2 V, observed at very low current density, can be ascribed to the methanol oxidation reaction and methanol cross-over [10]. Thereafter, the current density increases markedly with decrease in cell potential. This behavior is related to a first methanol dehydrogenation step which occurs at low anodic potentials and gives rise to an almost complete monolayer of CO coverage on the Pt surface. Oxidation of chemisorbed residues occurs as the anode overpotential becomes sufficiently high for a fast water discharge reaction

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

10

A. S. Aricò, V. Baglio and V. Antonucci

on Ru sites producing OH species [10]. According to previous studies in the literature [10], the decrease in cell potential with increase in current density in the activation controlled region of the polarization curve suggests that water discharging occurs quite slowly at high cell potentials. On the other hand, at intermediate and low cell potentials, the observed variation of the fuel cell performance with methanol concentration indicates that the reaction rate is directly related to the coverage of methanolic residues. These evidences suggest that the rate of methanol oxidation depends on a suitable coverage of both OH and CO-like species on the anode surface. At the potential corresponding to the maximum power output, the rate of the water discharging process is extremely fast due to the high anode overpotential (Figure 2) . Hence, the surface reaction between Pt-bonded CO-like species and OH species adsorbed on Ru sites could become the rate determining step of the overall process. The chemical interaction between these adsorbed intermediates requires a good atomic mixing between Pt and Ru sites. In general, when a significant fraction of Ru atoms is alloyed with Pt, these sites actively contribute to accelerate the methanol oxidation reaction. As above observed, the oxygen reduction process is affected by methanol cross-over especially at high cell voltages; a mixed potential at the cathode determines low OCV values. At high current densities, due to the fast methanol consumption at the anode, the overpotential for oxygen reduction in a DMFC is less affected by methanol cross-over but it is still larger than in the case of a H2/air fuel cell due to the cathode flooding caused by water permeation through the membrane.

2.2. OPERATION OF DIRECT METHANOL FUEL CELLS It is useful to report in a polarization diagram, beside the terminal voltage and the power density also the variation of the ohmic resistance and the cross-over current (equivalent current density) as a function of the electrical current density. Usually, the internal resistance does not vary significantly in the current density range of a DMFC. Whereas, the equivalent current density is generally quite important for the methanol fuel cell, because it determines the fuel utilization and influences the overall performance. It represents the current corresponding to the methanol permeation rate. A direct comparison with the effective (measured) electrical current permits to evaluate the fuel lost due to the cross-over (Figure 5). The cross-over or permeation rate of methanol can be in-situ determined by the so-called CO2 sensor method. Alternatively, chromatographic analyses can be used. It is assumed that, in the presence of a Pt based catalyst, almost all the methanol that is permeated to the cathode is oxidized to CO2 at high

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Basic Aspects of Direct Methanol Fuel Cells

11

electrochemical potentials. From the CO2 flow rate, methanol oxidation reaction and Faraday law, the equivalent current is calculated according to the following equation I cross over= molMeOH cross-over ·6 · F

(10)

where molMeOH cross-over is the rate of methanol permeation to the cathode per unit of time and geometric electrode area (moles ·min-1 cm-2). The CO2-sensor technique appears the most appropriate in-situ method to determine the cross-over. This measurement is instantaneous and it is carried out simultaneously to the normal polarization experiment. It takes into proper account the effects of the electro-osmotic drag. However, a possible CO2 permeation from the anode compartment through the membrane may cause some interferences at high electrical current densities in the presence of thin membranes. Alternatively, the permeation can be measured in a separate experiment.

0.6 40 0.4

30

20 0.2

Equivalent Current density / mA cm

60

50

Cell Potential / V

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

0.8

-2

70

10

0

0 0

0.1

0.2

0.3

0.4

0.5

0.6

Current density / A cm -2

Figure 5. Cell potential and equivalent current density (due to methanol cross-over) as a function of electrical current density for a DMFC operating at 60°C.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

12

A. S. Aricò, V. Baglio and V. Antonucci

An inert gas is fed to the cathode compartment and the oxygen electrode is polarized anodically, while the methanol electrode is polarized cathodically. The measured anodic current is related to the methanol permeation rate through the membrane. This procedure discards two relevant phenomena usually occurring during fuel cell operation. In a DMFC equipped with a protonic membrane, the electro-osmotic drag causes methanol molecules which coordinate together with water the protons to migrate from the anode to the cathode. In the driven mode protons move to the methanol compartment. During practical DMFC operation, there is a significant methanol concentration gradient at the anode-electrolyte interface at high current densities. This determines the real methanol cross-over (Figure 5). Such an effect is not reproduced in the driven mode. By using the CO2 sensor method (Figure 5), it is observed that the equivalent current density usually decreases as a function of the electrical current density due to the methanol consumption at the anode/electrolyte interface that reduces the methanol concentration gradient between the anode and the cathode. Another important aspect is related to the oxygen stoichiometry at the cathode. Air is usually fed at stoichiometry of 2 in the normal mode. If the system operates at ambient pressure, the power consumption by the blower, in the case of a large air flow, is not as significant as in the case of a compressor (pressurized DMFC). An increase of air flow produces better performance (Figure 6a) and it is not unusual to see reported DMFC experiments with an air flow corresponding to a stoichiometry of 5 or even higher. A similar effect is produced by the cathode pressure. The negative effects of methanol poisoning at the cathode can be counteracted by the increase of oxygen partial pressure (Figure 6b). The Temkin adsorption isotherm is often used to model the adsorption of oxygen at the cathode in PEMFCs [20]. Accordingly, an increase of oxygen partial pressure influences significantly the coverage of adsorbed oxygen species. This process is in competition with the adsorption of methanol permeated through the membrane, on the cathode surface. It appears that the increase of air stoichiometry especially favors the physical removal from the cathode of the liquid mixture of water/methanol that permeates through the membrane or is formed by the reaction (water) avoiding the electrode flooding. The flooding of the cathode is more significant in a DMFC than in a PEMFC due to the supply of a plenty of liquid water together with methanol to the anode that permeates to the cathode through the hydrophilic membrane. This effect is less dramatic in a vapor-fed DMFC. For several applications, it is quite important to know how much heat is released during the operation of the fuel cell.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Basic Aspects of Direct Methanol Fuel Cells

13

Air Oxygen Air Oxygen

Cell Potential / V

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

130 °C

0.6

0.4 0.3

0.4

0.2

0.2

0.1

(b)

0

Power Density / W cm

0.8

-2

(a)

0 0

0.5

1

1.5

2

Current Density / A cm

2.5

3

-2

Figure 6. Polarization and power density curves recorded at (a) 60°C using different air flow rates at the cathode compartment and at (b) 130°C using different oxidant (air or pure oxygen).

Thus, another useful polarization diagram to qualify the behavior of the fuel cell should include the heat released in the process deconvoluted into the heat produced by the electrochemical process and that produced by the chemical reaction associated to the methanol cross-over (Figure 7). We are not aware of the

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

14

A. S. Aricò, V. Baglio and V. Antonucci

use of such a plot in the literature. Yet, since we are convinced of its importance, we have reported in Figure 7 the results concerning with the heat release calculation that we have carried out in the framework of an European Community project called Morepower and dealing with the development of a DMFC system for portable applications [21]. For sake of simplicity, the calculation here reported concerns with the single cell behavior. The heat produced per unit of time by the electrochemical process only is derived from the reaction enthalpy and the methanol consumption rate from the following formula: Q=molMeOH (Hr - nFEcell);

(11)

molMeOH are the number of moles of methanol which are consumed per unit of time. This term is calculated from the electrical current density and the Faraday law; Ecell is the cell voltage at the operating current density. The redox process occurring at the cathode, associated to the methanol crossover, can be assumed as a chemical oxidation to CO2 by effect of the oxygen molecules, mediated by the Pt catalyst, since there is no electrical work produced: (12) 0.6

0.8

V cell 0.6

0.4

W el

Atmospheric pressure

0.3

W therm 0.4

Total heat 0.2

Heat due to MeOH cross-over

0.2

0.1

0

0 0

0.1

0.2

0.3

0.4

0.5

0.6

Current density / A cm-2

Figure 7. Electrical and thermal characteristics of a DMFC operating at 60°C under atmospheric pressure.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Power density / W cm-2

0.5

Cell potential / V

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Q=molMeOH ·Hr;

Basic Aspects of Direct Methanol Fuel Cells

15

In Figure 7 the effect of heat release due to the electrochemical process is compared to that related to methanol cross-over for a device operating at 60 °C with 1 M methanol and air feed at the cathode. The cross-over decreases significantly as a function of electrical current density; accordingly, the amount of heat released diminishes. The heat due to the cross-over is comparable to the heat due to the electrochemical process at 0.1 A cm-2 , whereas, the effect of cross-over is much less significant above 250 mA cm-2. The chemical energy that is dissipated as heat represents a net loss of efficiency. The heat released increases exponentially as a function of current and it reaches a maximum at the short circuit. From the considerations made in the previous sub-section, during fuel cell operation at a defined current, the cell voltage is lower than the reversible potential for the overall process. At a defined current density, the voltage efficiency is thus defined as the ratio between the terminal cell voltage and the reversible potential for the process at the same temperature and pressure.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

v= V/ Erev

(13)

As a consequence of cross-over, the current delivered by the DMFC device is smaller than that calculated on the basis of overall methanol consumption. The ratio between the measured electrical current (I) and that calculated from the Faraday law on the basis of the total methanol consumption (Itotal) is defined as fuel efficiency: f= I/Itotal

(14)

The fuel efficiency should be calculated after a defined operation cycle. In the active mode, the methanol is recirculated through the anode; accordingly, it is fed to the anode compartment at a rate much larger than the stoichiometry value determined by the operating current density using the Faraday law. A determination of the fuel efficiency only based on the methanol cross-over may represent a source of error if there are other side effects such a loss of methanol by evaporation. This may occur for a methanol fuel cell operating at high temperature; but, also for the passive mode operation at room temperature if the tank or membrane borders are not perfectly sealed. For a passive DMFC, the overall efficiency can be thus expressed as: = rev  v  f

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

(15)

16

A. S. Aricò, V. Baglio and V. Antonucci

In the active mode, the amount of energy consumed by the auxiliaries (pumps, blowers etc.) as compared to that delivered by the stack must be taken into consideration when the system efficiency is calculated. Besides performance and efficiency of the DMFC device, the energy density of the fuel plays a significant role for several practical applications including transportation and portable power sources. It is also a relevant factor for stationary generation since it determines which infrastructure is appropriate for fuel distribution [22]. The energy density of a fuel is defined with respect to the weight (kWh/kg) or volume (kWh/l) as

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

We= (-G/3600·M) or Ws=(-G·/3600·M);

(16)

where M is the molecular weight (g/moles) and  the density (g/l). Table 1 summarizes the energy density for various fuels. The gravimetric energy density of pure methanol is about one order of magnitude larger than that of H2 stored in a pressurized tank (e.g. at 200 bar) and in a metal hydride system (4-5%). Similar considerations can be made with regard to the volume. The energy density of pure methanol is also quite larger than Li-Ion batteries but lower than conventional liquid fuels used in transportation such as Gasoline and Diesel (Figure 8). One of the most promising applications of DMFCs presently concerns with their use as portable power sources [2]. In this regard, increasing interest is devoted towards the miniaturization of these fuel cell devices in order to replace the current Li-ion batteries. Table 1. Volumetric and Gravimetric Energy density for various fuels of technical interest for low temperature fuel cells Fuels Diluted Hydrogen (1.5%) Hydrogen Methanol Ethanol Formic acid Dimethyl ether (DME) Ethylene glycol

Volumetric Energy density (kWh l-1) -

Gravimetric Energy density (kWh kg-1) 0.49

0.18 (@ 1000 psi, 25°C) 4.82 (100 wt.%) 6.28 (100 wt.%) 1.75 (88 wt.%) 5.61 (in liquid of 100 wt.%) 5.87 (100 wt.%)

6.1 8 8.4 5.3

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Basic Aspects of Direct Methanol Fuel Cells

17

Ni-Cd

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Figure 8. Gravimetric and volumetric energy density of various fuels/devices.

Theoretically, methanol has a superior specific energy density (6000 Wh/kg) in comparison with the best rechargeable battery, lithium polymer and lithium ion polymer (600 Wh/kg) systems. This means longer conversation times using mobile phones, longer operating times of laptop computers before battery recharging and more power available on these devices to support consumer demand. Another significant advantage of the DMFC over the rechargeable battery is its potential for instantaneous refueling. These significant advantages make DMFCs an exciting development in the portable electronic devices market [2,4]. However, to use all potential energy density associated to the methanol combustion, a tank with pure methanol should be used. When considering the range i.e. the driving distance of a fuel cell car compared to an internal combustion engine or the operating time of a methanol portable power source compared to a Li-battery, it should be considered how much is the fuel diluted with water and the overall efficiency of the processes. The efficiency of a methanol fuel cell should be in principle larger than that of a conventional ICE that uses a thermal cycle; concerning the comparison with batteries, the electrochemical processes occurring in the battery are quite reversible compared to methanol fuel cells. In the transportation application, the water management (water recovery) can be addressed at the system level thus allowing the use of pure methanol in the tank. Recent progress in the development of DMFCs systems

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

18

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

operating in the passive mode has allowed the use of an almost pure fuel in the cartridge of these miniaturized devices. In some approaches the water produced at the cathode diffuses to the anode through the hydrophilic channels in the membrane allowing the methanol oxidation reaction to occur. It is pointed out that a proper comparison of the actual energy density (range) for the different technologies should be made under practical operating conditions.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Chapter 3

3. DEVELOPMENT OF DIRECT METHANOL FUEL CELLS COMPONENTS AND DEVICES

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

3.1. DMFC CATALYSTS Although the concept and the practice of direct methanol-air fuel cells have been under scrutiny for decades, the overwhelming problem associated with anode catalyst poisoning particularly at or near ambient temperatures has limited the progress of these efforts. The present section aims to discuss the historical development of DMFC catalysts, whereas a more detailed discussion of the mechanism and an overview of the status of knowledge is provided in a successive section. The methanol electro-oxidation process was explored for the first time by E. Muller in 1922 [23]. However, the concept of methanol fuel cells started to be investigated in the early fifties by Kordesch and Marko [24] as well as by Pavela [25]. Accordingly, a wide research on the anode and cathode electrocatalysts for such application initiated in the same years [1, 26]. The investigation of specific catalysts was closely related to the selected electrolyte. Alkaline electrolytes were initially used for methanol fuel cells; the search for the active anode and cathode catalysts mainly regarded Nickel or Platinum for methanol oxidation reaction and silver for the oxygen reduction process [25, 26]. The investigation of methanol oxidation reaction was conducted in parallel in acidic electrolytes such as sulfuric acid in the same period [3, 23]. It was observed that the kinetics of methanol electro-oxidation was slower in acidic environment as compared to the alkaline electrolyte. Yet, this aspect, instead of depressing the search of active and stable (corrosion resistant) catalysts for operation in conjunction with acidic electrolytes, further stimulated the research in this field.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

20

A. S. Aricò, V. Baglio and V. Antonucci

This was essentially due to the fact that better perspectives were envisaged for the acid electrolyte based direct methanol fuel cells. The main issue of a liquid alkaline electrolyte, such as KOH was its chemical interaction with the reaction product of methanol oxidation, i.e. carbon dioxide, to give rise to the formation of carbonate. Among the pioneering studies carried out on catalysts for methanol oxidation in acidic media, it should be mentioned the work of Cathro [27] that investigated the Pt-Sn system and that of Jansen and Molhuysen [28] which considered all the possible combinations of Pt with other elements of the periodic table. The latter work represented the first attempt to make a screening of bimetallic catalysts by using a systematic approach. Pt-Sn and Pt-Ru were individuated as the most promising anode formulations [28]. In effect, PtSn resulted initially a better catalytic system than PtRu [28]; this was essentially due to the use of the adatoms approach for both the formulations. Successive studies by Watanabe and Motoo [29] in the sixties showed the large potentialities of the Pt-Ru system especially when Pt and Ru were combined together in a solid solution (face centered cubic alloy). In the first decades, the activity on DMFC was mainly addressed to the search of active anode formulations; it resulted from half cell studies that the methanol oxidation process was slower than the oxygen reduction; thus, the anode reaction attracted more interest being considered the rate determining step of the overall DMFC process [1,2]. The first decades of activity on anode catalysts were mainly addressed to the investigation of the mechanism and the search of new or improved catalyst formulations. One of the first attempts to rationalize the methanol oxidation process was due to Bagotzky and Vassiliev [30]. Their work was essentially carried out on pure platinum; they proposed some relevant kinetic equations for methanol electro-oxidation rate as a function of the coverage of methanolic residues and oxygen species adsorbed on the electrodes. These studies served as a basis for the successive formulation of the bifunctional theory [29] for bimetallic catalysts. It is also worthy mentioning the work of Shibata and Motoo on the effect of ad-atoms [31] that individuated the influence of steric effects. Of relevant interest were also the successive attempts to further rationalize the mechanism of methanol oxidation by electrochemical studies of McNicol [32], Parsons and VanderNoot [33], Aramata [34]. However, an in- depth analysis of the methanol oxidation process, initially for smooth electrode surfaces, was made possible by the use of spectro-electrochemical methods to investigate the methanol oxidation in-situ. This work was carried out by several groups including those of Lamy [35] and Bockris [36], Christensen [37]. These essentially regarded the study of adsorbed methanolic residues by infrared spectroscopy. Whereas, for

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Development of Direct Methanol Fuel Cells Components and Devices

21

the adsorbed oxygen species, ellipsometry resulted as one the most proper methods to individuate the characteristics of such species adsorbed on the surfaces [38]. Further completion of the knowledge of the methanol oxidation process was provided by the use of the in-situ mass spectrometry. This method allowed to detect the anode potentials at which CO2 was formed. In the late eighties and beginning of nineties, a relevant amount of the work was addressed to the amelioration of the catalyst formulations and to a further investigation of the structural, surface and electronic properties of the most promising formulations, essentially Pt-Ru. The work carried out by Goodenough, Hamnett and Shukla in the eighties and nineties [39-40] was of relevant interest in this regard. They focused their attention not only on the catalyst but also on the electrode structure (including diffusion and backing layers). Mc Breen and Mukerjee [41] Ross et al. [42] used advanced physico-chemical tools such as Extended X-ray Absorption Fine Structure (EXAFS), Low-energy ion scattering spectroscopy, Auger and Xray photoelectron spectroscopy (XPS) to characterize model and practical anode and cathode catalysts. The 1990s opened a new era for the direct methanol fuel cells; the investigation of catalysts formulations in polymer electrolyte single cells progressively replaced the half cell studies in liquid electrolytes. The number of anodic formulations investigated reduced consistently. More attention was addressed to the behavior of the catalyst inserted in a practical MEA in single cell. Increasing emphasis was given to the development of the complete device. The studies on the anode catalysts continued with the study of the catalytic layer i.e. a composite layer of catalyst and ionomer with mixed conducting properties; whereas, further spectro-electrochemical studies were carried out on practical (high active surface area) catalytic systems. It clearly appears that from the enormous amount of work on the anode catalysts essentially one formulation has been selected and used in practical systems equipped with protonic electrolytes i.e. PtRu. Yet, despite all the work carried out in the 1990s and in the present decade and addressed to the amelioration of this catalytic system, neither unsupported nor carbon supported PtRu catalysts have completely solved the issue of poor anode reaction kinetics. Furthermore, operation at high anode potentials often causes Ru dissolution and migration through the membrane to the cathode [43]. It seems that the selection of PtRu was restrictive compared to all the work carried out in the first decades. Although, it was recognized by many researchers as one of the most important routes to improve methanol oxidation kinetics, the approach of using a multifunctional catalysts has achieved little success.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

22

A. S. Aricò, V. Baglio and V. Antonucci

The extensive use of combinatorial catalyst discovery studies in the 1990s suggested that this direction could be of significant importance [44]. In the successive paragraphs, we have tried to resume some properties related to the anode catalyst development, especially considering those aspects that have been studied in the first decades. Possible interpretations of promoting effects are provided by using unifying concepts. The aim is to reconsider some of the efforts that have been carried out to investigate possible alternatives and/or oxidation routes as a basis for a further development of methanol oxidation catalysts. It is pointed out that being this analysis mainly based on the studies carried out using ad-atoms there is a strong effect of the chemical oxidation state. As example, it was cited above that in the studies of Janssen and Moolhuysen [28], the activity of PtSn was larger than PtRu. However, it is considered that the analysis of several formulations may overcome this difficulty and allows to derive useful indications on possible alternative routes for catalyst development. It was established in the seventies and eighties that the activity of a methanol oxidation catalyst depends on several factors including catalyst formulation, the catalyst support, the electrode structure and operating conditions. Most of the work was concentrated on examining the effect of changing the catalyst formulation as a means of enhancing catalytic activity; alloys of various compositions were used as electrode materials, although most of these alloys were based on Pt. In fact, they are the only ones, to date, showing appreciable activity for CH3OH oxidation at low potentials in an acidic environment. It was evident that the reaction rate was improved by electrocatalysts adsorbing water and/or oxygen species at potentials similar to the reversible potential of the CH3OH oxidation reaction and/or are able to minimize poisoning by the methanolic residue [45]. The explanations for the mode of action of bimetallic catalysts have also been advanced as hypotheses for the catalytic enhancement of electrodes modified by coverage of foreign metal adatoms. It was recognized that the presence of an alloying metal or ad-atom either: (i) modified the electronic nature of the surface; (ii) modified the physical structure; (iii) blocked the poison formation reactions; (iv) adsorbed oxygen/hydroxyl species which take part in the main oxidation reaction. Adatoms were seen as a means to improve the electrocatalytic behavior of electrodes either by minimizing the poisoning reaction or enhancing the main oxidation reaction, as above said. Besides the several proposed reaction paths, there were three possible mechanisms that appeared more likely to occur on the catalytic systems more in-depth investigated [33]. A mode of action suggested that the adatoms either altered the electronic properties of the substrate or acted as redox intermediates. This hypothesis, supported by experimental evidences, also

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Development of Direct Methanol Fuel Cells Components and Devices

23

led to consider the influence of a possible steric effect on the enhanced oxidation rate. A second way by which adatoms might work is to block the poison formation reaction. In the first decades, the poison reaction was assumed to occur on a certain number of sites of a certain type which was greater than the number required for the main reaction. The idea was thus to tailor the coverage by the adatoms so as to leave sufficient space for the main reaction to occur but not the poison reaction. The third hypothesis of adatom activity invoked the idea of a bifunctional catalyst. In this mechanism, the oxidation reaction of either the fuel or the poisoning intermediate was enhanced by the adsorption of oxygen or hydroxyl radicals on adatoms adjacent to the reacting species. Accordingly, there was a geometric requirement for the oxygen species transfer and the rates of adsorption and donation as well as the total amount of OH adsorbed affected the degree of enhancement. The potential at which this enhancement occurred was governed by the potentials where the adatoms were capable to adsorb oxygen species. These adsorption potentials were found to be lower than Pt [33]. Amongst the different combinations investigated by different researchers, the best ones reported in the first decades of study of methanol electro-oxidation were PtRu, Pt-Sn, Pt-Re, Pt-Ti, Pt-Os, Pt-Ru-Re, Pt-Ru-Sn, Pt-Ru-Au [28]. Although, at that time, it was not certain by which mechanism these elements enhanced the activity, enhancement factors up to 40 times were recorded with respect to bare Pt [46]. The following analysis tries to bring to a "rationale" main features of different bimetallic Pt-based electrocatalysts (Figures 9-11) The aim is to try to individuate possible correlations between the electrocatalytic activity of the systems that were investigated in the first decades and some physicochemical properties of the elements included in the formulation of the electrocatalysts. In a first instance, we have considered: a) the influence of the interaction energies of CO on transition metals; b) the influence of electronegativity; c) possible steric and electronic effects, reflected by the influence of atomic radius and ionic potential values. Such choices have been prevalently determined by the above reported interpretations, according to which, however, the nature of the catalyst surface plays in any case the determining role. As for the involvement of different poisons in the oxidation mechanism, spectroelectrochemical methods allowed to establish in the eighties that the two more probable adsorbed species were -CHO and CO, being evidences to support both possibilities [33]. On the whole, the evidence for CO as the poisoning species appeared to be, however, more conclusive, coming from various "in situ" spectroscopic techniques. Accordingly, both linear and bridged CO species have been detected on the electrode surface. Furthermore, the

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

24

A. S. Aricò, V. Baglio and V. Antonucci

L og . ( I / mA c m

-2

)

potentials where the poison was oxidized also corresponded very closely to those where adsorbed species from pure CO were oxidized in separate experiments [33]

2.4 2 1.6

Ru

1.2

Re

Os

Ti

0.8 0.4

Mo

Nb

V

0

W

-20

0

20

40

60

80

100

120

Zr

140

K c al/mole Figure 9. Methanol electrocatalytic activity vs. calculated interaction energy of CO adsorption on various Pt alloyed transition metals.

I / μAcm -2

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

40 Ag Hg Bi Te Se S

30 20 10 0 0

1

2 Electronegativity / eV

3

Figure 10. Methanol electro-oxidation activity vs. electronegativity of the adatoms used for Pt modification.

A plot of several electro-catalytic activity results [28, 46-48] as a function of the adsorption heats of CO on transition metals taken from Miyazaki [49] is reported in Figure 9. It appears that a higher electro-oxidation activity is found for metals having low interaction energies with the CO molecule. The proposed linear trend is furthermore strengthened by the experimental evidences [46], demonstrating in that period that Pt-Ru binary alloys resulted one of the two best

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Development of Direct Methanol Fuel Cells Components and Devices

25

performing catalysts in the CH3OH electro-oxidation, together with Pt-Sn. Such a correlation seemed to confirm also the possible role of the added metal component in the formulation of the binary catalyst, i.e. its capability to lower the extent of poison adsorption. This could happen if the adsorption of the poisoning specie was lower at the alloy surface as compared to the pure Pt surface. The experiences by Shibata et al. [31], related to the enhancement of CO oxidation on Pt by the electronegativity of adatoms showed that a strong interaction between the adatom site and either the hydrated hydrogen ion or the adsorbed CO molecule was required to obtain high enhancement effects. Yet, such evidences were not observed for the CH3OH oxidation reaction (see Figure 10); this is, in our opinion, explainable by the detrimental effects of non transition elements upon the methanol dehydrogenation which, as well known, constitutes the first step in the overall electro-oxidation reaction [50]. Accordingly, the positive influence of the more electronegative elements still holds for transition metals which are known to favor CH3OH dehydrogenation. According to the adatoms theory, two distinct effects were identified from a catalytic view-point in enhancing the electrooxidation of organic molecules; the first was related to the modification of the electronic environment of the adsorption site, the other was linked to the steric factor which also influenced both the extent and the strength of the adsorption process. As for the latter, a correlation between the overpotential at which a sustained current is obtained for CH3OH electro-oxidation and the atomic radius of the alloyed metal with Pt (Figure 11). The combination of labile adsorption intermediates with a large metal area utilization was recognized to be favored by the small dimensions of the alloyed element. For what concerned the influence of adatoms on the modification of the electronic environment of Pt, it was anticipated in the literature that the positive catalytic effects of Pt-Ru and Pt-Sn formulations were due to the adsorption, by these elements, of active oxygen on the catalyst surface at low potentials. These species were identified in the late eighties as adsorbed OH species by using spectro-electrochemical methods. Although a correlation between the electrooxidation activity and the thermodynamic formation at such potentials of Sn and Ru ionic species was not established clearly, cyclic voltammetry experiments showed that, both on smooth Pt as well as on carbon-supported Pt electrodes, methanol oxidation activity was observed at low potentials especially when labile oxygen species were detected at the electrode surface. Regarding the development of catalysts for the oxygen reduction process in methanol fuel cells during the past decades, it should be mentioned that silver was mainly considered at the beginning especially for an operation in conjunction with alkaline electrolytes [25, 26]. Silver is presently used in oxygen depolarized

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

26

A. S. Aricò, V. Baglio and V. Antonucci

cathodes for industrial cloro-alkali cells which utilize the same liquid electrolyte (KOH) [51]. Whereas, Pt and its alloys were employed for oxygen reduction in the presence of acidic electrolytes both in unsupported and carbon supported form.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Ru Os Ir Pt Pd Au

.

.

.

Figure 11. Methanol electro-oxidation activity vs. atomic radius of various metals utilised to form Pt-alloys.

The development of cathode catalysts for proton conducting electrolyte based DMFCs was initially influenced by the similar studies carried out on phosphoric acid fuel cells (PAFC). The catalytic layer was essentially a mixture of Pt/C catalysts and PTFE binder sintered at around 350 °C. These hydrophobic electrodes were useful in reducing the flooding caused by liquid electrolytes. However, it was rapidly envisaged the need of using a methanol tolerant cathode catalyst. Accordingly, the research was also directed towards alternative catalytic systems also including non noble metals. In the late eighties, the development of methanol tolerant oxygen reduction catalysts became of practical interest with the development of metal

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Development of Direct Methanol Fuel Cells Components and Devices

27

chalcogenides [52], phtalocyanines and phorphirins [53] –based cathodes with catalytic activities approaching those of Pt in the presence of methanol poisoning. As for the development of non noble metal catalysts for the oxygen reduction reaction, essentially three classes of materials were investigated in the first decades of DMFC development. These included oxides, chalcogenides different than oxides and organometallic compounds. Metal oxides were initially considered being these materials the most obvious candidates as Pt substitutes in electrocatalysis. Oxides are present on the surfaces of all non-noble metals at potentials useful for oxygen reduction. It was assumed that oxygen reduction occurred by exchange of oxygen atoms between molecular oxygen, surface oxide, and water (referred to as a regenerative mechanism). Several non noble metal oxides with metallic conductivity are available; a few are stable in acid electrolytes. These includes mainly tungsten oxides as a possible candidate. Only recently, Co-oxides with perovskite structure similar to those used in intermediate temperature solid oxide fuel cells (IT-SOFCs) have been considered [54]. These may represent a promising candidate material for oxygen reduction especially in alkaline electrolytes. The sulfides, selenides, and tellurides promise higher conductivity than oxides and better resistance to catalyst poisoning by effect of cross-over than metals or oxides. A comparison of oxides with other chalcogenides allows to draw some conclusions about the electronic structure required for effective catalysis. The activity of hemoglobins for oxygen transport suggested that materials similar to the central heme group could prove active in oxygen electrocatalysis. The large stabilization energy by resonance of the aromatic macrocyclic porphyrins, phthalocyanines, and related compounds gives them a much higher stability at oxygen reduction potentials than would be found for the smaller organometallic molecules used as catalysts in nonaqueous media. The planar aromatic structure reminiscent of graphite gave hope that high dispersions of these compounds could be achieved on carbon supports. It was, however, considered that the relatively large size of the molecules, however, could limit the surface density of active sites. To conclude this survey of the activities carried out on the DMFC catalysts, it appears from the large number of catalytic formulations, studied especially in the sixties, seventies and eighties, that only a few catalytic systems have been considered in the successive DMFC device development activity. The activity in the nineties was essentially addressed to the amelioration of these formulations mainly PtRu at the anode, Pt for conventional cathodes, Ru selenides for methanol tolerant cathodes. Pt-Sn is now mainly used in direct ethanol fuel cells despite the suitable perspectives that were envisaged for this catalyst at the beginning of the

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

28

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

research on methanol oxidation catalysts. Figure 12 shows the decrease in overpotential for methanol oxidation (2 M) at various bimetallic catalysts in sulfuric acid.

Figure 12. IR-free polarization curves for the oxidation of methanol (2 M) at various bimetallic catalysts in 0.5 M H2SO4 at 60 °C. Pt loading 0.8 mg cm-2. Reprinted from Ref. [2] with permission from Wiley-VCH.

On the whole, it appears however that the complexity of the CH3OH electrooxidation mechanism has strongly hindered the disclosure of a practical catalyst, and much remains to be done; amongst other, the evaluation of new materials as potential catalysts is still one of the primary items. In any case, it is a general feeling the need of major breakthroughs in catalysts and electrolytes to make competitive DMFC systems. A discussion of the present status of development of the catalysts is made in a next section.

3.2. ELECTROLYTES FOR DMFCS It is widely recognized that the electrolyte is a key component in DMFCs. The main drawbacks which affect DMFCs concern with methanol cross-over,

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Development of Direct Methanol Fuel Cells Components and Devices

29

slow methanol oxidation reaction rate and high cost of active components. A proper choice of the electrolyte can reduce these constraints. The electrolyte determines the fuel permeation rate, the choice of the catalysts and it influences the reaction rate. It is well known that the use of non-noble metal catalysts is possible in the presence of alkaline electrolytes. Proton conducting electrolytes have been preferred to alkaline electrolytes for several decades for practical reasons, e.g. to avoid carbonation. The standard electrolyte membrane for DMFCs is usually a perfluorosulfonic acid membrane such as Nafion that is also widely used in PEMFCs. Most of the electrolytes alternative to Nafion both proton conducting and alkaline-type are cheaper than the classical perfluorosulfonic membranes used in PEMFCs; in some cases, they are also characterized by smaller methanol cross-over; however, life-time characteristics similar to those shown by Nafion-type membranes in fuel cells (60000 hrs of operation) have not yet been demonstrated for the alternative membranes [55, 56]. Concerning with the conductivity, only recently, membranes alternative to Nafion-type have shown similar levels of performance. One critical aspect is related to the fact that the presence of water is a requirement of low-temperature DMFCs for the occurrence of the electrochemical reactions and to promote ion conductivity. Being methanol highly soluble in water, the transport of water through the membrane is commonly associated to methanol permeation. This effect is more critical with protonic membranes because besides methanol transport due to the concentration gradient (diffusion) there is an effect due to the electro-osmotic drag. High ionic conductivities are often associated to the presence of large water uptake by the membrane; whereas, what is required is a low water uptake. These aspects are mainly related to polymer electrolyte membrane direct methanol fuel cells. No drawbacks in terms of methanol crossover with consequent cathode poisoning and poor anode reaction kinetics are envisaged in intermediate temperature solid oxide fuel cells (IT-SOFCs) which employ dense ceramic anionic electrolytes and operate at 500°-750°C. However, such devices are less suitable for most of the applications of methanol fuel cells including portable and assisted power units (APU) uses. Thus, the research efforts on low temperature methanol fuel cell membranes have been addressed to improve the conductivity, reduce cross-over of methanol and reduce catalyst degradation. The latter regards for example dissolution of Ru and Ru ion migration from the anode to the cathode [43], Co or Fe dissolution from the cathode into the membrane (e.g. in the case of a PtCo or PtFe alloy cathode) etc. Other important aspects are concerning with the increase of chemical and electrochemical stability, reduction of water uptake and swelling, extension of the operating temperature range and finally cost reduction for market application. In

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

30

A. S. Aricò, V. Baglio and V. Antonucci

this section a brief analysis of the historical development of DMFC electrolytes is presented. A discussion of the present status of development of DMFC membranes is discussed in a successive section. The electrolytes that were first used in DMFC devices in the fifties consisted of a concentrated alkaline solution e.g. KOH which contained dissolved methanol [3, 25, 26, 57-58]. Although some attempts were also carried out with anion exchange membranes onto which the electrodes were pressed [59], thus, using an approach similar to the modern polymer electrolyte membrane fuel cells, the best results in terms of output power were achieved with 5 M KOH [58]. The reasons why alkaline instead of protonic electrolytes were selected in the first stage of methanol fuel cell development were due to the fact that methanol oxidation in alkaline media was more rapid than in the presence of acidic electrolytes [59]. This made possible the use of low cost catalysts e.g. Ni at the anode and silver at the cathode [58]. Unfortunately, there were several practical constraints that induced most of the DMFC developers to abandon the alkaline electrolyte approach [2,3,60, 61] focusing the attention on protonic electrolytes. This approach was under taken despite the fact that the number of possible catalyst formulations was restricted to those stable in acidic environments [60-61]. The main problem of alkaline electrolytes consisted in the acid-base reaction between CO2 and the alkaline solution with carbonate precipitation in the catalyst pores. The lack of adequate alkaline polymer electrolyte membranes with conductivity comparable to Nafion has retarded the development of the newgeneration of anionic membrane direct methanol fuel cells (AMDMFC) [59]. Most of the anion-exchange membrane, used in the past, required KOH recirculation at the anode [59, 60]. Regeneration of this electrolyte due to carbonation and dilution was considered a significant constraint for practical applications. The use of carbonate/bicarbonate media instead of KOH reduced part of the kinetic advantage of alkaline media versus protonic electrolyte for methanol electro-oxidation [59]. Proton conducting solid polymer electrolyte membranes for hydrogen-fed fuel cells were initially developed at General Electric for the Gemini Earth-orbiting program in the early 1960s [62]. These devices were based on a polystyrene sulfonic acid membrane that exhibited poor oxidative stability and thus no suitable long-term performance [62]. The poor stability was due to the oxidation of the C-H bonds occurring at high potentials, at the cathode, especially in the presence of hydrogen peroxide-type radicals. A large improvement in terms of stability was achieved when Nafion replaced the sulfonated polystyrene-divinylbenzene membrane in the late sixties [55, 56]. Nafion was originally developed for cloro-alkali electrolyzers; its peculiar

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Development of Direct Methanol Fuel Cells Components and Devices

31

characteristics rely on the excellent chemical, electrochemical and high proton conductivity that derives from its unique chemical structure [62]. The Nafion membrane has a chemical structure similar to Teflon, except that it has side chains with ether-type bonds followed by –CF2 groups bonded to the terminal sulfonic acid [62]. The fluorine exerts a high electronic affinity enhancing the polarization of the SO3 – _H+ group. Furthermore, the chemical stability of the C-F bond is quite larger than the C-H bond. The excellent characteristics of Nafion promoted the development of membranes with similar structure such us the Dow and Asahi glass membranes, the Flemion from Aciplex, the Hyflon from Solvay Solexis. One of the characteristics of some of these membranes, especially Hyflon, is the presence of short side chains and an equivalent weight smaller than Nafion [63, 64]. Hyflon was employed with success in DMFCs especially at high temperature [64]. The peculiar characteristic of Hyflon is a glass transition temperature larger than Nafion that makes this polymer more stable at high temperature [64]. Yet, a pressurized system is necessary to favor water retention above 100 °C and the problem of the methanolcross over is still quite significant. Despite the large methanol cross-over shown by the Nafion membranes, these membranes were the most used electrolyte in DMFCs in the 1990s [55]. Still now Nafion 117 is considered as a standard electrolyte to compare the performance, conductivity, methanol cross-over of alternative or newly developed membranes for DMFCs [1, 2]. Prior of the advent of Nafion in DMFCs that essentially occurred in the late 1980s, various acid electrolytes were used for direct methanol fuel cells such as sulfuric acid and phosphoric acid. In these devices, the anode and cathode were separated by a ceramic matrix, e.g. a porous silicon carbide separator impregnated with the acidic electrolyte. The electrodes were impregnated with the same acid. Yet, due to the extremely high methanol crossover, this concept was abandoned in favor of Nafion. However, in the late 1990s, a similar approach was developed by Peled et al. at the Tel Aviv University by using a nanoporous membrane based on PEO filled with silica nanoparticles and impregnated with sulfuric acid or TFMFSA [65]. The projected cost of this nanoporous membrane filled with sulfuric acid was considered quite lower compared to Nafion. However, the best performance was achieved with TMFSA due to the lower adsorption of anionic species from the electrolyte on the electrode [66]. The latter system showed high performance and less constraints due to the water management under high temperature operation [66]. In general, in the 1990s, the development of DMFCs, strictly followed the concepts used in PEMFCs. However, some differences were adopted to reduce the main constraints associated to methanol fuel cells i.e. poor anode reaction kinetics

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

32

A. S. Aricò, V. Baglio and V. Antonucci

and methanol cross-over. In the DMFCs, a PtRu anode was used instead of Pt. Since the membrane thickness was recognized to be effective in counteracting the methanol cross-over, thick membranes like Nafion 117 (170 microns) were mainly used especially at low temperatures. The optimum membrane thickness was determined by both cell resistance and cross-over [63]. One of the parameter used to evaluate the membrane performance was determined by the ratio between conductivity and cross-over normalized by the thickness of the membrane. Methanol cross-over in several protonic membranes occurs through the ion-cluster channels [1, 3]. These channels also govern the proton conduction through the socalled vehicle mechanism assisted by the water molecules [67]. This mechanism is present in most of the low-temperature membranes. Thus, it is quite difficult to deconvolute proton conduction from methanol cross-over. The strong activation control of the methanol oxidation reaction indicated in the high temperature operation the most useful strategy to increase the performance. High temperature operation allowed to achieve high current densities with consequent fast methanol consumption at the anode/electrolyte interface. This effect reduces the concentration gradient allowing to decrease the methanol cross-over. In this regard, the use of thin membranes like Nafion 112 was sometime adopted for high temperature operation [9, 68]. The high temperature operation was recognized as the most appropriate approach to increase the DMFC performance with regard to automotive, APU and stationary applications. In this regard, several approaches were investigated to increase the operating temperature of methanol fuel cells. The most promising strategies were concerning with the use of phosphoric acid doped polybenzo-imidazole membranes operating at about 180-200 °C [69] and composite perfluorosulfonic acid membranes operating up to 145°C including inorganic fillers such as silica, zirconium phosphate, heteropolyacid doped silica, titanium oxide [70-71]. In several attempts, the filler was in-situ formed e.g. silica was synthesised inside the membrane by using a sol-gel type procedure by using tetraethyl orthosilicate (TEOS) as precursor [72]. Although, both approaches were demonstrated appropriate to extend the operating temperature range, the main constraint of phosphoric acid doped polybenzoimidazole was represented by the leaching of acid molecules from the membrane in the presence of hot methanol; whereas, composite membrane operated properly at 145 °C in the presence of 3 bar, abs. pressure [71]. Successively, it was shown that the water retention properties in composite membranes were promoted by the presence of acidic functionalities on the filler surface. Operation at high temperature (145 °C) and reasonable pressure (1.5 bar, abs) with acceptable level of performance was made possible by ameliorating of membrane properties [71, 73]. The composite membrane

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Development of Direct Methanol Fuel Cells Components and Devices

33

approach was also extended to membranes alternative to Nafion such as s-PEEK and polysulfone-type [74]. The mechanism for reducing the cross-over in phosphoric acid doped polybenzoimidazole and composite membranes is essentially different. In the composite membrane a decrease of cross-over is caused by an increase of the tortuosity factor [70]; whereas, in the phosphoric acid doped polybenzoimidazole membrane the proton conduction mainly occurs through a Grotthus-type mechanism involving proton hopping. Thus, the methanol permeation assisted by water transport through the ion clusters channels is less significant. However, it appears, at the present, that there is no valid membrane that covers properly the entire temperature range including ambient temperature conditions. The Grotthustype mechanism is essentially activated by the high temperature. Concerning with the high temperature operation, new commercial membranes with promising properties and the aim to better cover a wide range of temperature have been recently developed by Fumatech and BASF [63]. The potentialities of these new systems for DMFCs appear to be yet not fully explored. For what concerns membrane stability, some aspects related to the operating conditions appear less critical in a DMFC than in a hydrogen-fed fuel cell. As example, the cathode never experiences in a methanol fuel cell electrochemical potentials above 1 V. Furthermore, the formation of hydrogen-peroxide radicals, which can cause significant membrane degradation, mainly occurs in a PEMFC by effect of hydrogen cross-over to the cathode. On the other hand, it is observed that, the presence of hot concentrated methanol in DMFCs may increase membrane swelling. However, due to the lower electrochemical stability requirements, the range of membranes explored for DMFCs appears larger than in PEMFCs [2, 63]. Partially fluorinated, non-fluorinated aromatic polymers, radiation grafted ethylene tetrafluoroethylene (ETFE) based membranes, acidbase blends etc. [2, 63], have been explored as alternative to Nafion. This is not an exhaustive list of alternative membranes. In effect a large variety of electrolytes has been investigated. Several excellent reviews have been already published on this topic [61, 63]. Most of these alternative electrolytes have the characteristics of lower methanol cross-over but also smaller conductivity than Nafion (as mentioned above these aspects are often inter-related) and especially the projected costs appear quite promising compared to the classical perfluorosulfonic membranes. Among the various proposed membranes, s-PEEK [75], despite the promising properties in terms of conductivity, fuel permeation and costs, seem still affected by a large swelling; the properties of sulfonated poly (aryl-ether) type membranes such as polysulfone or polyimide ionomer membranes as well as acid-base blends appears more promising [62, 74, 76].

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

34

A. S. Aricò, V. Baglio and V. Antonucci

A critical aspect related to the use of membranes alternative to perfluorinated sulfonic acid electrolytes (e.g. Nafion) is often the absence of a suitable ionomer solution to form composite catalytic layers with mixed ionic-electronic conductivity and suitable membrane adhesion characteristics. An important issue is concerning the recent discover of Ru cross-over through the membrane [43], especially when poorly alloyed PtRu anodic catalysts are used. This indicates the need to utilize membranes which are resilient to Rucrossover as well as an anode catalyst with proper chemical properties. The aspects related to the chemistry of Pt-Ru catalysts are discussed in a next session dealing with the discussion of the present status of development of DMFC components. More recently, several attempts have been carried out to develop a new generation of alkaline anion exchange membranes (AAEMs). The availability of new anionic polymers with conductivity approaching values which are half of the conductivity of Nafion [77, 78], but, characterized by much lower methanol crossover has given new emphasis to the development of alkaline methanol fuel cells. The new membranes significantly reduced the drawbacks associated to conventional aqueous KOH electrolyte fuel cells i.e. carbonate formation and need to frequently regenerate the electrolyte. OH- ions, necessary for ion conduction, are formed at the cathode by the water employed to humidificate the oxidant. These ions migrate to the anode reducing methanol cross-over by the electro-osmotic drag. The alkaline environment allows the use of non-noble metals catalysts (usually unstable in the acidic environment), the catalyst corrosion and membrane degradation problems are significantly mitigated due to the high pH [77]. Accordingly, cheap catalysts and hydrocarbon-only membranes have been explored [59, 60, 61, 77]. There are however some drawbacks which concern with the formation of a pH gradient between anode and cathode [59], the need of cathode humidification (protonic membranes based DMFCs are usually fed by dry air) and the need to increase the operating temperature to enhance conductivity that may be not useful for portable applications. The liquid electrolytes that have been used in half cell studies to simulate the behavior of solid state electrolytes used in single cells and stacks include several categories: Inorganic Acids such as H2SO4, HClO4 and H3PO4 Superacids such as trifluoromethansulfonic acid (TFMSA). Buffers such as CO32- /HCO3-. Alkaline electrolytes such as KOH.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Development of Direct Methanol Fuel Cells Components and Devices

35

TFMSA appears as the most appropriate liquid electrolyte medium to simulate the behavior of Nafion-type membranes Yet, for practical reasons H2SO4 or HClO4 have been mostly used. The latter is less affected by anion adsorption on the catalyst. Phosphoric acid is usually selected for half cell studies of catalytic systems that are used in conjunction with PBI membranes. As conclusion, we can say that significant progress on the electrolyte development for DMFCs has been made but it is difficult at the present to individuate the best membrane. The choice of the proper membrane results from several considerations that include device application, range of operating conditions, costs etc. One of the commercial membranes that shows a suitable compromise among the various requested properties is produced by Polyfuel [79]. The Polyfuel membrane is especially suited for passive DMFCs. The performance are comparable and even superior to Nafion based devices, whereas life-time exceeds 5000 h. Although, its stability is not comparable to that of Nafion (60.000 h in PEMFCs), it appears already acceptable for portable application.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

3.3. DMFC STACKS DEVELOPMENT Even though from the 1960s to the 1980s the development of DMFCs was a ―Fuel Cell Researcher‘s Dream‖ [55], only in the last two decades, the attractive features of DMFC power sources (portable liquid fuel with energy density about half of that of gasoline, environmentally friendly technology, a ten-fold increase in power density with a proton exchange membrane electrolyte) have clearly indicated perspectives for their applications in transportation, portable power and power generation/cogeneration applications. There were only a few attempts to develop DMFC stacks/systems in the decades preceding the 1990s. The first attempts to develop methanol fuel cells [23, 24] were carried out by Kordesch and Marko in 1951 on the basis of the earlier studies by E. Muller. The DMFC devices initially developed were based on alkaline electrolytes, Ni-based or Pt-Pd-based anodes and silver cathodes [25, 26]. One of the first DMFC stack of reasonable power, based on alkaline electrolytes, was developed in the sixties by Murray and Grimes at Allis-Chalmers in 1963 [58]. It was operating at 50 °C and consisted of an aqueous alkaline electrolyte (5 M KOH) Pt-Pd anode and Ag cathode catalysts. Porous Ni sheet was used as backing layer for the electrode. The stack composed of 40 cells provided a maximum electrical power of 750 W at 9 V with an average cell power density of about 40 mW/cm2. The approach of using concentrated KOH as electrolyte was similar to that of hydrogen-fed alkaline fuel cells developed in the same period

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

36

A. S. Aricò, V. Baglio and V. Antonucci

mainly for space applications [55, 56, 62]. However, it was rapidly recognized the problem of an acid-base reaction between the electrolyte and the reaction product at the anode, i.e. CO2 with formation of potassium carbonate. This caused carbonate precipitation inside catalyst pores with occlusion and consequent increase of mass transport constraints. The increase of resistance with time and the need to regenerate the cell (excess of carbonate removal) induced most of the DMFC developers to address their efforts towards the development of DMFCs based on proton conducting electrolytes. Some attempts were addressed to use carbonate electrolytes working at high temperature or anion exchange membranes [59]. Unfortunately, the performance achieved by using the latter approach was not satisfying in the past [59]. In recent years, the approach of anion exchange membranes for DMFCs has been reconsidered. The new anionic membranes show proper conductivity values even in the absence of KOH recirculation [77]. The development of DMFC devices based on acidic electrolytes was initially conducted in the mid of sixties by some leading laboratories such as Shell, Exxon and Hitachi [2]. In all these cases, 1-2 M sulfuric acid was used as the electrolyte and unsupported platinum black was initially used as electrocatalyst. However, studies conducted by researchers at Shell in 1968 individuated Pt-Ru as one of the most effective anode electro-catalysts and developed a 300 W prototype [23]. Esso developed a 100 W stack for communication applications [23]. In terms of stack development, another highlighting result in terms of performance was the development of a 50 W DMFC stack at Hitachi [2].The stimulation for developing DMFCs was enhanced in the early 1990s when the sulfuric acid electrolyte was replaced by a solid state proton conductor (Nafion). There were two significant effects, a) an increase in electro-catalytic activity of the electrodes, and b) improved open circuit potential of the cell due to reduced methanol cross-over. In addition, an enhanced oxygen electrode performance was observed because of the replacement of the liquid electrolyte with the perfluorosulfonic acid solid polymer. The initial interest in stack development was for the transportation application. Recently, due to the lower efficiencies and power densities of DMFCs than PEMFCs and higher projected cost of the DMFC power source (mainly because of the significantly higher noble metal loading) the near-term projected applications were directed towards portable applications [4]. For what concerns the active components used in the stacks, the focus of DMFC researchers was directed to the anode catalyst development for operation in conjunction with proton conducting electrolytes, especially after it was recognized the limited applicability of the liquid alkaline electrolytes. In the 1990s, most of the efforts were concentrated on the development of membranes alternative to

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Development of Direct Methanol Fuel Cells Components and Devices

37

Nafion. In parallel, the demonstration of DMFC devices based on polymer electrolytes was progressively intensified primarily for automotive applications and later for portable uses. Concerning with the catalysts, the PtRu formulation selected by the studies conducted in the decades before the 1990s was mainly used at the anode in DMFC devices, whereas, Pt was used at the cathode; alternatives such as methanol tolerant cathode catalysts were used in the same period. Due to the poor anode reaction kinetics and cathode poisoning by methanol cross-over, high noble metal loadings were used in both electrodes. Initially, the amount was about 10 mg cm-2 (unsupported catalysts); this progressively decreased up to reach 2 mg cm-2 and even lower. For portable application, it is still used a high loading of 4 mg cm-2 in conjunction with unsupported PtRu and Pt catalysts formed by nanosized primary particles (2-3 nm); although for portable uses the catalyst loading is less critical than for large-size stacks, further efforts are being addressed to a further decrease of the noble metal content. Yet, due to the kinetic drawbacks, it is quite difficult to reach the ultra-low loading achieved for PEMFCs. However, one approach towards this direction regards the decorated catalyst approach; this approach although demonstrated in single cell [80] has not yet been used for DMFC stacks. Successively to the use of unsupported catalyst, high concentration carbon supported catalysts (e.g. 85% PtRu and 60 %Pt) have been used in practical stacks [81] especially in high temperature applications. As above said, the carbon supported catalysts have not yet completely replaced the unsupported catalysts in the DMFCs for portable devices. To achieve high reaction rates and fast transport of reactants/reaction products in the passive mode at ambient temperature, it is appropriate to use very thin catalytic layers, with high catalyst content, in close contact with the polymer electrolyte membrane. Unsupported catalysts match several of the above requirements. Similarly, due to the methanol cross-over constraints is still preferable a thick membrane for a practical device. The large catalyst content and the expensive ionomer membranes per unit of active cell area is less critical for portable applications as compared to automotive and stationary uses. Yet, up to now, it appears that the choice of the active components for practical systems has been mainly determined by the reliability of the materials than their real perspectives for wide scale application.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Chapter 4

4. DMFC APPLICATIONS

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

4.1. PORTABLE POWER SOURCES The potential market for portable fuel cell systems mainly deals with the energy supply for electronic devices but it also includes remote and microdistributed electrical energy generation. Accordingly, DMFC power sources can be used in mobile phones, lap-top computers, as well as energy supply systems for weather stations, medical devices, auxiliary power units (APU) etc.. Direct Methanol Fuel Cells (DMFCs) are promising candidates for these applications because of their high energy density, lightweight, compactness, simplicity as well as easy and fast recharging [24, 82-84]. Theoretically, methanol has a superior specific energy density (6000 Wh/kg) in comparison with the best rechargeable battery, lithium polymer and lithium ion polymer (theoretical, 600 Wh/kg) systems. This performance advantage translates into longer conversation times using cell phones, longer times for use of laptop computers between replacement of fuel cartridges and more power available on these devices to support consumer demand. In relation to consumer convenience, another significant advantage of the DMFC over the rechargeable battery is its potential for instantaneous refueling. Unlike rechargeable batteries that require hours for charging a depleted power pack, a DMFC can have its fuel replaced in minutes. These significant advantages make DMFCs an exciting development in the portable electronic devices market. Several organizations (Table 2) are actively engaged in the development of low power DMFCs for cellular phone, laptop computer, portable camera and electronic game applications [68, 82-84, 85]. The initial goal of this research is to develop proof of concept DMFCs capable of replacing high performance rechargeable batteries in the US$ 6-billion portable electronic devices market.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Table 2. DMFC power sources for portable applications Developer

Number/area of cells

Power density

T(°C)

Oxidant

Methanol concentration (M)

Anode catalyst and loading

Electrolyte

Motorola Labs

4 cells (planar stack)/ 13-15 cm2 Planar stack

12-27 mW cm-2

21

Ambient air*

1

PtRu alloy, 6-10 mg cm-2

Nafion 117

Cathode catalyst and loading 6-10 mg cm-2

3-5 mW cm-2

25

Ambient air*

1

PtRu alloy

Nafion

Pt

6 cells (flat pack)/6-8 cm2 5 cells/45 cm2

6-10 mW cm-2

2025 60

Ambient air*

1

Air (3-5 times stoichiometry)

0.5

PtRu alloy, 4-6 mg cm-2 PtRu alloy, 0.816.6 mg cm-2

Nafion 117 Nafion

40 cells/ 100 cm2

45-55 mW cm-2

5070

O2 (3 atm)

1

PtRu, 2 mg cm-2

Nafion 115

Pt, 4-6 mg cm-2 Pt, 0.816.6 mg cm-2 Pt, 2 mg cm-2

12 cells (monopolar)/ 2 cm2

23 mW cm-2

25

Ambient air*

5 Passive mode

PtRu, 3-8 mg cm-2

Hybrid membrane

Pt, 3-8 mg cm-2

6 cells (bipolar)/ 52 cm2 6 cells (monopolar)/ 6 cm2 20 cm2

121-207 mW cm-2

2550

O2 (300 ml min-1), ambient pressure

2.5 Active mode

PtRu/C

Nafion 115 & 117

Pt black

40 mW cm-2

25

Ambient air*

4 Passive mode

PtRu

Nafion 115

Pt

60-100 mW cm-2

25

Ambient air*

30-5%

PtRu

Liquid electrolyte

Pt

Energy Related Devices Jet Propulsion Lab Los Alamos National Labs Forschungszentr um Julich GmbH Samsung advanced Institute of Technology Korea Institute of Energy Research Korea Institute of Science & Technology More Energy Ltd.

300 W/l

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Developer

Number/area of cells

Power density

T(°C)

Oxidant

Methanol concentration (M)

Anode catalyst and loading

Electrolyte

Institute for Fuel Cell Innovation, Canada University of Connecticut, USA Honk Kong University

3 cells (monopolar)

8.6 mW cm-2

25

Ambient air*

2 Passive mode

80%PtRu, 4 mg cm-2

Nafion 117

Cathode catalyst and loading Pt black, 4 mg cm2

4 cells/ 18-36 cm2

30 mW cm

-2

25

Ambient air*

2-5 Passive mode

PtRu alloy, 7 mg cm-2

Nafion 117

Pt, 6.5 mg cm-2

Single cell/4 cm2

28 mW cm-2

22

Ambient air*

4 Passive mode

PtRu, 4 mg cm-2

Nafion 115

The Pennsylvania State University, USA Harbin Institute of Technology

Single cell/5 cm2

93 mW cm-2

85

Air (700 ml min-1 and 15 psig)

2 Active mode

PtRu, 4 mg cm-2

Nafion 112

40% Pt/C, 2 mg cm-2 40% Pt/C, 1.3 mg cm-2

Single cell

9 mW cm-2

30

Ambient air*

2 Passive mode

40% PtRu/C, 2 mg cm-2

Nafion 117

Tel-Aviv University, Israel Tekion Inc., USA University of California, USA

Flat fuel cell/6 cm2

12.5 mW cm-2

25

Ambient air*

1-6 in H2SO4 Passive mode

PtRu, 5-7 mg cm-2

NP-PCM

Single cell/5 cm2 µ-Single cell/1.625 cm2

65 mW cm-2

60

Ambient air*

PtRu

Nafion

Pt

16-50 mW cm-2

2560

Air (88 ml min-1)

2 Active mode 2 Active mode

PtRu, 4-6 mg cm-2

Nafion 112

Waseda University, Japan

µ-Single cell/0.018 cm2

0.8 mW cm-2

25

O2 (10 µl min-1) sat. in H2SO4

2

PtRu, 2.85 mg cm-2

Nafion 112

40% Pt/C, 1.3 mg cm-2 Pt, 2.4 mg cm-2

40% Pt/C, 2 mg cm-2 Pt, 4-7 mg cm-2

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Table 2. (Continued)

*

Developer

Number/area of cells

Power density

T(°C)

Oxidant

Methanol concentration (M)

Anode catalyst and loading

Electrolyte

Institute of Microelectronic of BarcelonaCNM, Spain Yonsei University, Korea CNR-ITAE, Italy

µ-Single cell

11 mW cm-2

25

Ambient air*

4-5 Passive mode

PtRu, 4 mg cm-2

Nafion 117

Multi-cell structure (monopolar) 3 cells (monopolar)/4 cm2

33 mW cm-2

80

O2 (30 ml min-1)

2

60% PtRu/C, 4 mg cm-2

Nafion 117

20 mW cm-2

21

Ambient air*

5 Passive mode

PtRu, 4 mg cm-2

Nafion 117

Ambient air usually refers to the air breathing mode.

Cathode catalyst and loading Pt, 4 mg cm-2

60% Pt/C, 4 mg cm-2 Pt, 4 mg cm-2

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

DMFC Applications

43

Motorola Labs—Solid State Research Center, USA, [4] in collaboration with Los Alamos National Laboratory (LANL), USA, is actively engaged in the development of low power DMFCs (greater than 300 mW) for cellular phone applications [86]. Motorola has recently demonstrated a prototype of a miniature DMFC based on a membrane electrode assembly (MEA) set between ceramic fuel delivery substrates [4]. Motorola utilized their proprietary low temperature cofired ceramic (LTCC) technology to create a ceramic structure with embedded microchannels for methanol/water mixing and delivery to the MEA and, exhausting by-product CO2. In addition, processing of the ceramic material into a grid screen design facilitated the delivery of ambient air to the MEA. Substrates were processed in multiple layers after aligning, tacking and laminating at approximately 3.45×106 Pa. The final monolithic integrated ceramic substrate was formed after sintering at 850 °C. In this design, the MEA was mounted between two porous ceramic plates. Thin films of electrocatalysts were applied in a proprietary process using carbon cloth gas diffusion layers. For the anode, an unsupported Pt/Ru (1:1) alloy at a high loading of 6–10 mg cm-2, and for the cathode Pt black were used as electrocatalysts. Nafion 117 membranes were used as the electrolyte and were hydrated by running deionized (DI) water through the cell for 18 h. The active electrode area for a single cell was approximately 3.5–3.6 cm2. In the stack assembly, four cells were connected in series in a planar configuration with a MEA area of 13–14 cm2, the cells exhibited average power densities between 15–22 mW cm-2. Four cells (each cell operating at 0.3 V) were required for portable power applications because DC–DC converters typically require 1V to efficiently step up to the operating voltage for electronic devices. The fuel cell consumed oxygen from ambient air (21 °C and 30% RH) and the fuel from 1.0 M methanol pumped at a rate of 0.45 ml min-1 using a peristaltic pump. Variations in time of operation, temperature, fuel mixing, flow rate and humidity gradually led to improved performance characteristics of the system. In addition, improved assembly and fabrication methods have led to peak power densities greater than 27 mW cm-2. Motorola is currently improving their ceramic substrate design to include micro-pumps, methanol concentration sensors and supporting circuitry for second generation systems. Energy Related Devices Inc. (ERD), USA, is working in alliance with Manhattan Scientific Inc., US) to develop miniature fuel cells for portable electronic applications [82,87]. A relatively low-cost sputtering method, similar to the one used by the semiconductor industry for production of microchips, was used for deposition of electrodes (anode and cathode) on either side of a microporous plastic substrate; the micropores (15 nm to 20 µm) are etched into the substrate using nuclear particle bombardment. Micro-fuel arrays, with external

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

44

A. S. Aricò, V. Baglio and V. Antonucci

connections in series, were precisely fabricated and had a thickness of about a millimeter. The principal advantages of the cell include the high utilization of catalyst, controlled pore geometry, low-cost materials and minimum cell thickness and weight. A MicroFuel CellTM was reported to have achieved a specific energy density of 300 Wh/kg using methanol and water and air as the anodic and cathodic reactants, respectively [4]. The anode design that was developed by MicroFuel CellTM has represented a critical new advance in the development of a cost-effective pore-free electrode that is only permeable to hydrogen ions [4]. This increases the efficiency of a methanol fuel cell because it blocks the deleterious effect of methanol crossover across the membrane. The first layer of the anode electrode formed a plug in the pore of the porous membrane; an example is a 20 nm thick palladium metal film on a Nuclepore filter membrane with 15 nm diameter pores. The second layer (platinum) was deposited to mitigate the hydration induced cracking that occurs in many of these films. The third layer was deposited over the structural metal film and was the most significant layer because it needed to be catalytically active to methanol and capable of accepting hydrogen ions. An alternate method of forming the electrode was to include on the surface of the metal films powder catalyst particles (Pt/Ru on activated carbon) to enhance the catalytic properties of the electrode. Between the anode electrode and the cathode electrode was the electrolyte filled pore, the cell interconnect and the cell break. In the pores of the membrane the electrolyte (Nafion) was immobilized and ERD claims this collimated structure results in improved protonic conductivity. Each of the cells was electrically separated from the adjacent cells by cell breaks, useless space occupying the central thickness of the etched nuclear particle track plastic membrane. The cathode was formed by first sputter depositing a conductive gold film onto the porous substrate followed by a platinum catalyst film. The electrode was subsequently coated with a Nafion film. Alternatively, platinum powder catalyst particles were added to the surface of the electrode via an ink slurry of 5% Nafion solution. A hydrophobic coating was then deposited onto this Nafion layer in order to prevent liquid product water from condensing on the surface of the air electrodes. ERD developed a novel configuration to utilize their fuel cell as a simple charger in powering a cellular phone. The fuel cell was configured into a plastic case that was in close proximity to a rechargeable battery. Methanol is delivered to the fuel cell via fuel needle and fuel ports, which allow methanol to wick or evaporate out into the fuel manifold, and was delivered to the fuel electrodes. The Jet Propulsion Laboratory (JPL), USA, has been actively engaged in the development of ―miniature‖ DMFCs for cellular phone applications over the last 2

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

DMFC Applications

45

years [84, 88]. According to their analysis, the power requirement of cellular phones during the standby mode is small and steady at 100–150 mW. However, under operating conditions the power requirements fluctuates between 800–1800 mW. In the JPL DMFC, the anode was formed from platinum-ruthenium alloy particles, either as fine metal powders (unsupported) or dispersed on high surface area carbon. Alternatively, a bimetallic powder made up of submicron platinum and ruthenium particles was reported to give better results than the platinumruthenium alloy. Another method describes the sputter-deposition of platinumruthenium catalyst onto the carbon substrate. The preferred electrolyte was Nafion 117; however, other materials may be used to form proton-conducting membranes. Air was delivered to the cathode by natural convection and the cathode is prepared by applying a platinum ink to a carbon substrate. Another component of the cathode was the hydrophobic Teflon polymer utilized to create a three-phase boundary and to achieve efficient removal of water produced by electro-reduction of oxygen. Sputtering techniques can also be used to apply the platinum catalyst to the carbon support. The noble metal loading in both electrodes was 4–6 mg cm-2. The MEA was prepared by pressing the anode, electrolyte and cathode at 8.62 × 106 Pa and 146 °C. JPL opted for a ―flat-pack‖ instead of the conventional bipolar plate design, but this resulted in higher ohmic resistances and non-uniform current distribution. In this design the cells were externally connected in series on the same membrane, with through membrane interconnect and air electrodes on the stack exterior. Two ―flat packs‖ were deployed in a back to back configuration with a common methanol feed to form a ―twin-pack‖ [4]. Three ―twin-packs‖ in series were needed to power a cellular phone. In the stack assembly, six cells were connected in series in a planar configuration, which exhibited average power densities between 6–10 mW cm-2. The fuel cell was typically run at ambient air, 20–25 °C with 1 M methanol. Improvements of configuration and interconnect design have resulted in improved performance characteristics of the six cell ―flat-pack‖ DMFC. Based on the results of the current technology, the JPL researchers predict that a 1 W DMFC power source, with the desired specifications for weight and volume and having an efficiency of 20% for fuel consumption, can be developed for a 10 h operating time, prior to replacement of methanol cartridges. As stated earlier Los Alamos National Laboratory (LANL) has been in collaboration with Motorola Labs—Solid State Research Center to produce a ceramic based DMFC, which provides better than 10 mW cm-2 power density. LANL researchers have also been engaged in a project to develop a portable DMFC power source, capable of replacing the ―BA 5590‖ primary lithium battery, used by the US Army in communication systems [89]. A 30-cell DMFC

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

46

A. S. Aricò, V. Baglio and V. Antonucci

stack, with electrodes having an active area of 45 cm2, was constructed, an important feature of this stack being a narrow width (i.e., 2 mm) of each cell. MEAs were made by the decal method; that is, thin film catalysts bonded to the membrane resulting in superior catalyst utilization and overall cell performance. Anode catalyst loading of Pt between 0.8–16.6 mg cm-2 in unsupported PtRu and carbon supported PtRu were used. A highly effective flow field for air made it possible to use a dry air blower for operation of the cathode at three to five times stoichiometry. The stack temperature was limited to 60 ◦C and the air pressure was 0.76 atm, which is the atmospheric pressure at Los Alamos (altitude of 2500 m). To reduce the cross-over rate, methanol was fed into the anode chamber at a concentration of 0.5 M. Since water management becomes more difficult at such low methanol concentrations, a proposed solution was to return water from the cathode exhaust to the anode inlet, while using a pure methanol source and a methanol concentration sensor to maintain the low methanol concentration feed to the anode. The peak power attained in the stack near ambient conditions was 80 W at a stack potential of 14 V and approximately 200 W near 90 °C. From this result, it was predicted that this tight packed stack could have a power density of 300 W/l. An estimate of an energy density of 200 Wh/kg was made for a 10 h operation, assuming that the weight of the auxiliaries is twice the weight of the stack. Forschungszentrum Julich GmbH (FJG), Germany, has developed and successfully tested a 40-cell 50W DMFC stack [90]. The FJG system consisted of the cell stack, a water/methanol tank, a pump and ventilators as auxiliaries. The stack was designed in the traditional bipolar plate configuration, which results in lower ohmic resistances but heavier material requirements. To circumvent the weight limitations current collectors were manufactured from stainless steel (MEAs were mounted between current collectors) and were inserted into plastic frames to reduce stack weight. The 6 mm distance between MEAs (cell pitch) revealed very tight packaging of the stack design. Each frame carried two DMFC single cells that were connected in series by external wiring [4]. MEAs were fabricated in house with anode loading of 2 mg cm-2 PtRu black, catalyst loading of 2 mg cm-2 Pt black and cell area of 100 cm2 for each of the 40 cells. At the anode a novel construction allowed the removal of CO2 by convection forces at individual cell anodes. The conditions for running the stack were 1M methanol, 60 °C and 3 bar O2 which led to peak energy densities of 45–55 mW cm-2. The cathode used air at ambient or elevated pressures; when the stack operated at temperatures above 60°C the air was fed into the cathode by convection forces. Further evaluation of the system revealed that current collectors made of stainless steel showed an inhomogeneous distribution of contact resistance and as a result

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

DMFC Applications

47

single cells displayed fluctuating power densities. It was postulated that the pressure of the current collectors on the MEAs was not high enough to prevent delamination of the electrocatalyst layer. Recent developments include a three-cell short stack design which has reduced the cell pitch to only 2 mm. The individual cell area of this design is larger, 145 cm2, than the previous prototype and although it is not air-breathing it works with low air stoichiometric rates (more efficient cathodic flow distribution structure). The short stack was tested under ambient pressure (a low power-consuming compressor provided air to the cathode) and operated at 45 °C. Samsung Advanced Institute of Technology (SAIT), South Korea, has developed a small monopolar DMFC cell pack (2 cm2, 12 cells, CO2 removal path, 5–10 M methanol, air breathing and room temperature) of 600 mW for mobile phone applications [91, 92]. Unsupported PtRu and Pt catalysts were coated onto the diffusion electrode of porous carbon substrate of anode and cathode, respectively. In order to allow methanol wicking and air breathing, short and capillary paths were designed as the diffusion layer. Catalyst loading was around 3–8 mg cm-2. Ternary alloys with low binding energy for CO adsorption were investigated with the aid of quantum chemical methods. Inorganic phase dispersed hybrid membranes based on Nafion or Co-PTFS were prepared and applied to the MEA for attaining a high fuel efficiency and preventing the voltage loss on the cathode. A gas chromatography (GC) method was utilized in situ during the electrochemical polarization. In this way, the cathode output stream gas was analyzed and it was calculated the amount of carbon dioxide produced by the permeated methanol, which is consumed at the cathode. A monopolar structure was investigated; 12 cells of 2 cm2 were connected in series within a flat cell pack. Fuel storage was attached to the cell pack and power characteristics were measured in the free-standing basis without any of fuel and air supply systems. A power density of 50 mW cm-2 at 0.3 V was achieved in the normal diffusion electrode design. For application in portable electronic devices, methanol wicking and air breathing electrodes were required; the MEA having this novel diffusion electrode showed 10 mW cm-2 at 0.3 V of power density without the aid of any external fueling system. In this MEA, the anode contained a microlayer for the methanol flow field with capillary wicking structure and the cathode-contained a microlayer for the air flow field with breathing structure. A hybrid membrane with inorganic phase dispersions was utilized. This was operated as methanol blocking medium in the hydrophilic channel of the ionomer assisting to reduce the amount of methanol cross-over. As measured by GC, the hybrid membrane allowed a 20–40% reduction of methanol permeation, at the nominal potential of 0.3 V, within the various range of methanol concentrations from 1 to 5 M. If a

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

48

A. S. Aricò, V. Baglio and V. Antonucci

conductivity approaching that of plain Nafion i.e. near 10-1 S/cm, could be achieved with this system, such a process offers the possibility of the development of functional membranes for DMFCs. A monopolar design consisting of 12 cells flat pack was assembled and tested in the severest condition that is methanol wicking and air breathing at room temperature. Each cell had the active area of 2 cm2 and the pack was equipped with a path of CO2 removal at the anode. The maximum power output was 560 mW at 2.8 V, close to that required by the cellular phone. For this cell pack condition with small active area, the unit cell power density was 23 mW cm-2, which is rather higher than that achieved in the single MEA test (10 mW cm-2). This result could be attributed to the uniform fuel distribution and efficient current collecting design of smaller single cells. The Korea Institute of Energy Research (KIER, South Korea) has developed a 10 W DMFC stack (bipolar plate, graphite construction) fabricated with six single cells of 52 cm2 electrode area [93]. The stack was tested at 25–50 °C using 2.5 M methanol, supplied without a pumping system, and O2 at ambient pressure, at a flow rate of 300 cc min-1. The maximum power densities obtained in this system were 6.3 W (121 mW cm-2) at 87 mA cm-2 at 25 °C and 10.8 W (207 mW cm-2) at 99 mA cm-2 at 50 °C. MEAs using Nafion 115 and 117 were formed by hot pressing and the electrodes were produced from carbon supported Pt-Ru metal powders and Pt-black for anode and cathode electrodes, respectively. More Energy Ltd. (MEL), ISRAEL, a subsidiary of Medis Technologies Ltd. (MDTL, USA), is developing a direct liquid methanol (DLM) fuel cells (a hybrid PEM/DMFC system) for portable electronic devices [94]. The key features of the DLM fuel cell are as follows: (i) the anode catalyst extracts hydrogen from methanol directly, (ii) the DLM fuel cell uses a proprietary liquid electrolyte that acts as the membrane in place of a solid polymer electrolyte (Nafion) and (iii) novel polymer and electrocatalyst enable the fabrication of more effective electrodes. The company‘s fuel cell module delivers approximately 0.9 V and 0.24 W at 60% of its nominal capacity for eight hours. This translates into energy densities of approximately 60 mW cm-2 with efforts under way to improve that result to 100 mW cm-2. The high power capacity of the cell is attributed to the proprietary electrode ability to efficiently oxidize methanol. In addition Medis claims the use of high concentrations of methanol (30%) in its fuel stream with plans for increasing that concentration to 45% methanol. The increased concentration of methanol in the feed stock results in concentration gradients that should lead to higher methanol crossover rates. However, this technical concern is not mentioned in the company‘s literature. At the Institute for Fuel Cell Innovation in Vancouver, Canada, a passive (air breathing) planar three-cell DMFC stack was designed, fabricated and tested [95].

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

DMFC Applications

49

In order, to maintain design flexibility, polycarbonate was chosen for the plate material; whereas, 304 stainless steel mesh current collectors were used. In order to test the DMFC in different electrical cell configurations (single cell, multiple cells connected in series or in parallel), a stainless threaded rod was attached to each mesh current collector on the anode and cathode side to allow for an external electrical connection. Commercial electrodes from E-TEK were used. The catalyst loading was 4 mg cm-2 and consisted of an 80% Pt:Ru alloy on optimized carbon. Unsupported Pt black with a 4 mg cm-2 loading was used for the cathode. A Nafion 117 membrane was utilized as electrolyte. A power density of 8.6 mW cm2 was achieved at ambient temperature and passive operation. Stacks with a parallel connection of the single cells showed a significantly lower performance than in a series configuration. It was also identified that high electrical resistance was the dominant factor in the low performance as a result of the stainless steel hardware and poor contact between the electrodes and current collectors. At University of Connecticut, USA, the group of Z. Guo and A. Faghri developed a design for planar air breathing DMFC stacks [96]. This design incorporated a window-frame structure that provided a large open area for more efficient mass transfer with modular characteristics, making possible to fabricate components separately. The current collectors had a niobium expanded metal mesh core with a platinum coating. Two four-cell stacks, one with a total active area of 18 cm2 and the other with 36 cm2, were fabricated by inter-connecting four identical cells in series. These stacks were suitable for portable passive power source application. Peak power outputs of 519 and 870 mW were achieved in the stacks with active areas of 18 and 36 cm2, respectively. A study of the effects of methanol concentration and fuel cell self-heating on fuel cell performance was carried out. The power density reached its highest value in this investigation when 2 and 3 M methanol solutions were used. At the Honk Kong University of Science and Technology, China, the group of R. Chen and T.S. Zhao [97-100] studied the effect of methanol concentration on the performance of a passive DMFC single cell. They found that the cell performance was improved substantially with an increase in methanol concentration; a maximum of power density of 20 mW cm-2 was achieved with 5.0 M methanol solution. The measurements indicated that the better performance with higher methanol concentrations was mainly attributed to the increase in the cell operating temperature caused by the exothermic reaction between permeated methanol and oxygen on the cathode. This finding was subsequently confirmed by the fact that the cell performance decreased, when the cell running with higher methanol concentrations was cooled down to room temperature. Moreover, they proposed a new membrane-electrode assembly (MEA), in which the conventional

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

50

A. S. Aricò, V. Baglio and V. Antonucci

cathode gas diffusion layer (GDL) is eliminated while utilizing a porous metal structure, made of a metal foam, for transporting oxygen and collecting current. They showed theoretically that the new MEA [99] and the porous current collector enabled a higher mass transfer rate of oxygen and thus better performance. The measured polarization and constant-current discharging behavior showed that the passive DMFC with the new MEA and new current collector yielded better and much more stable performance than did the cell having the conventional MEA and the conventional perforated-plate current collector, in particular with high methanol concentration. The EIS spectrum analysis further demonstrated that the improved performance with the new MEA was attributed to the enhanced transport of oxygen as a result of the reduced mass transfer resistance in the fuel cell system; whereas, the improved performance for the porous current collector was attributed to the increased operating temperature as a result of the lower effective thermal conductivity of the porous structure and its fast water removal as a result of the capillary action [100]. Another group at the Honk Kong University, H.F. Zhang et al. [101], reported on a flexible graphite-based integrated anode plate for DMFCs operating at high methanol feed concentration under active mode. This anode structure which was made of flexible graphite materials not only provided a dual role for the liquid diffusion layer and flow field plate, but also served as a methanol blocker by decreasing methanol flux at the interface of catalyst and membrane electrolyte. DMFCs incorporating this new anode structure exhibited a much higher open circuit voltage (OCV) (0.51 V) than that (0.42 V) of a conventional DMFC at 10 M methanol feed. Cell polarization data showed that this new anode structure significantly improved the cell performance at high methanol concentrations (e.g. 12 M or above). M.A. Abdelkareem and N. Nakagawa from Gunma University, Japan, [102] studied the effect of oxygen and methanol supply modes (passive and active supplies of methanol, and air-breathing and flowing supplies of oxygen) on the performance of a DMFC. The experiments were carried out with and without a porous carbon plate (PCP) under ambient conditions using methanol concentrations of 2 M for the MEA without PCP and 16 M for that with PCP. For the conventional MEA, flowing oxygen and methanol were essential to stabilize the cell performance, avoiding flooding at the cathode and depletion of methanol at the anode. As a result of flowing oxygen, methanol and water fluxes, the conventional MEA performance increased by more than twice as compared to that obtained. From the air-breathing cell. For the MEA with a porous plate, MEA/PCP, the flow of oxygen and methanol had no significant effect on the cell performance, where the porous carbon plate, PCP, prevented the cathode from

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

DMFC Applications

51

flooding by reducing the mass transport through the MEA. Methanol and water fluxes through the MEA/PCP were not affected by flowing oxygen at 0.1 l min−1. However, the increase in oxygen flow rate from 0.1 l min−1 to 1 l min−1 had a negative effect on the cell performance either for the conventional MEA or for the MEA/PCP. This was probably due to the cooling effect for conventional MEA and the drying effect for the MEA/PCP. A moderate supply of oxygen to the cathode, like air-breathing, was appropriate for the DMFC with a PCP. The effect of operating conditions on energy efficiency for a small passive DMFC was analyzed by D. Chu and R. Jiang from U.S. Army Research Laboratory, Adelphi, USA [103]. Both Faradic and energy conversion efficiencies decreased significantly with increasing methanol concentration and environmental temperature. The Faradic conversion efficiency was as high as 94.8%, and the energy conversion efficiency was 23.9% in the presence of an environmental temperature low enough (10°C) under constant voltage discharge at 0.6 V with 3 M methanol for a DMFC bi-cell using Nafion 117 as electrolyte. Although higher temperature and higher methanol concentration allowed to achieve higher discharge power, they resulted in considerable losses of Faradic and energy conversion efficiencies by using Nafion electrolyte membrane. Their conclusion was that the development of alternative highly conductive membranes with a significantly lower methanol crossover is necessary to avoid loss of Faradic conversion efficiency with temperature and with fuel concentration. Various research groups have focused their attention on the critical aspects which need to be addressed for the design a high-performance DMFC. These are CO2 bubble flow at the anode [104] and water flooding at the cathode [105]. Lu and Wang from the Pennsylvania State University, USA, [106] developed a 5 cm2 transparent cell to visualize these phenomena in situ. Two types of membraneelectrode assembly (MEA) based on Nafion® 112 were used to investigate the effects of backing pore structure and wettability on cell polarization characteristics and two-phase flow dynamics. One employed carbon paper backing material and the other carbon cloth. Experiments were performed with various methanol feed concentrations. The transparent fuel cell reached a peak power of 93 mW cm-2 at 0.3 V, using Toray carbon-paper based MEA under 2 M methanol solution preheated at 85°C. For the hydrophobic carbon paper backing, it was observed that CO2 bubbles nucleated at certain locations and formed large and discrete bubble slugs in the channels. For the hydrophilic carbon cloth backing, the bubbles were produced more uniformly and of smaller size. It was thus shown that the anode backing layer of uniform pore size and more hydrophilicity was preferable for gas management in the anode. Flow visualization of water flooding on the cathode side of DMFC was also carried out.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

52

A. S. Aricò, V. Baglio and V. Antonucci

It was shown that the liquid droplets appeared more easily on the surface of carbon paper due to its reduced hydrophobicity at elevated temperature. For the single-side ELAT carbon cloth, liquid droplets tended to form in the corner between the current collecting rib and GDL since ELAT is highly hydrophobic and the rib (stainless steel) surface is hydrophilic. Even if this study was performed at relatively high temperature (85°C), such a basic understanding is indispensable for portable DMFC design and optimization. Lai et al. [107] investigated the long-term discharge performance of passive DMFC at different currents with different cell orientations. Water produced in the cathode was observed from the photographs taken by a digital camera. The results revealed that the passive DMFCs with anode facing upward showed the best longterm discharge performance at high current. A few independent water droplets accumulated in cathode when the anode faced upward. Instead, in the passive DMFC with vertical orientation, a large amount of produced water flowed down along the surface of current collector. The passive DMFC with vertical orientation showed relatively good performance at low current. It was concluded that the cathode produced less water in a certain period of time at smaller current. In addition, the rate of methanol crossover in the passive DMFC with anode facing upward was relatively high, which leaded to a more rapid decrease of the methanol concentration in anode. The passive DMFC with anode facing downward resulted in the worst performance because it was very difficult to remove CO2 bubbles produced in the anode. Water loss and water recycling in direct-methanol fuel cells (DMFCs) are significant issues that affect the complexity, volume and weight of the system and become of greater concern as the size of the DMFC decreases. A research group at Tel-Aviv University, Israel, [108] realized a flat micro DMFC in a plastic housing with a water-management system that controlled the flux of liquid-water through the membrane and the loss of water during operation. These cells contained a nanoporous proton-conducting membrane (NP-PCM). Methanol consumption and water loss were measured during operation in static air at room temperature for up to 900 h. Water flux through the membrane varied from negative, through zero, to positive values as a function of the thickness and the properties of the water-management system. The loss of water molecules (to the air) per molecule of methanol consumed in the cell reaction (defined as the w factor) varied from 0.5 to 7. When w was equal to 2 (water flux through the membrane was equal to zero) there was no need to add water to the DMFC and the cell was operating under water-neutral conditions. On the other hand, when W resulted smaller than 2, it was necessary to remove water from the cell and when

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

DMFC Applications

53

it was larger than 2, water was added. The cell showed stable operation up to 900 h and its maximum power was 12.5 mW cm-2. At the Korea Institute of Science and Technology (KIST), Kim et al. [109] developed passive micro-DMFCs with capacities under 5 W to be used as portable power sources. Research activities were focused on the development of membrane–electrode assemblies (MEAs) and design of monopolar stacks operating under passive and air-breathing conditions. The passive cells showed many unique features, much different from the active ones. Single cells with active area of 6 cm2 showed a maximum power density of 40 mW cm-2 at 4 M of methanol concentration at room temperature. A six-cell stack having a total active area of 27 cm2 was constructed in a monopolar configuration and it produced a power output of 1000 mW (37 mW cm-2). Effects of experimental parameters on the performance were also examined to investigate the operation characteristics of single cells and monopolar stacks. Application of micro-DMFCs as portable power sources were demonstrated using small toys and display panels powered by the passive monopolar stacks. Tekion Inc., Champaign, USA, [110] has developed an advanced air breathing direct methanol fuel cell for portable applications. A novel MEA was fabricated to improve the performance of air-breathing direct methanol fuel cells. A diffusion barrier on the anode side was designed to control methanol transport to the anode catalyst layer and thus suppressing the methanol crossover. A catalyst coated membrane with a hydrophobic gas diffusion layer on the cathode side was employed to improve the oxygen mass transport. The advanced DMFC achieved a maximum power density of 65 mWcm−2 at 60 °C with 2 M methanol solution. The value was nearly two times more than that of a commercial MEA. At 40°C, the power densities operating with 1 and 2 M methanol solutions were over 20 mW cm−2 with a cell potential at 0.3 V. Pennsylvania State University together with University of California at Los Angeles, USA, [111] developed a silicon-based micro DMFC for portable applications. Anode and cathode flow-fields with channel and rib width of 750 µm and channel depth of 400 µm were fabricated on Si wafers using the microelectromechanical system (MEMS) technology. A membrane-electrode assembly (MEA) was specially fabricated to mitigate methanol crossover. This MEA features a modified anode backing structure in which a compact microporous layer is added to create an additional barrier to methanol transport thereby reducing the rate of methanol crossing over the polymer membrane. The cell with the active area of 1.625 cm2 was assembled by sandwiching the MEA between two micro-fabricated Si wafers. Extensive cell polarization testing demonstrated a maximum power density of 50 mW cm-2 using 2 M methanol feed

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

54

A. S. Aricò, V. Baglio and V. Antonucci

at 60 °C. When the cell was operated at room temperature, the maximum power density was shown to be about 16 mW cm-2 with both 2 and 4 M methanol feed. It was further observed that the present µDMFC still produced reasonable performance under 8 M methanol solution at room temperature. The Waseda University, Japan, proposed a new concept for µDMFC (0.018 cm2 active area) based on MEMS technology [112]. The µDMFC was prepared using a series of fabrication steps from micro-machined silicon wafer including photolithography, deep reactive ion etching, and electron beam deposition. The novelty of this structure is that anodic and cathodic micro-channels arranged in plane were fabricated, dissimilar to the conventional bipolar structure. The first objective of the experimental trials was to verify the feasibility of this novel structure on basis of MEMS technology. The methanol anode and oxidant cathode were prepared by electroplating either Pt–Ru or Pt and Pt, respectively, onto the Ti/Au electrodes. The electroplating solution for Pt was 20 mM H2PtCl6 · 6H2O and 0.5 mM (CH3COO)Pb · 3H2O. The deposition was carried by applying a current density of 30 mA cm-2 during 10 min. The mass loading of Pt was 2.4 mg cm-2. The Pt–Ru for methanol oxidation was obtained from a solution containing 20 mM H2PtCl6 xH2O + 20 mM RuCl3 xH2O. The deposition was performed at 0.15 V vs. Ag/AgCl for 5 min. The mass loading of Pt–Ru was 2.85 mg cm-2. The electroplating process was carried out at 25 °C for both electrodes. Energy dispersive X-ray (EDX) analysis showed a platinum/ruthenium of 90/10 atomic ratio. A Nafion 112 membrane was used as electrolyte. The performance of the µDMFC was assessed at ambient temperature using 2 M CH3OH/0.5 M H2SO4/H2O as the fuel and O2-sat./0.5 M H2SO4/H2O as the oxidant. The O2 saturated solution was prepared by using oxygen bubbling into 0.5 M H2SO4/H2O solution. The supply of fuel was made by means of a microsyringe pump connected to the fabricated µDMFC unit. The OCV for the Pt cell was 300 mV while it was 400 mV for Pt–Ru cell. The maximum power density was 0.44 mW cm-2 at 3 mA cm-2 at Pt electrode. While, the maximum power density reached 0.78 mW cm-2 at 3.6 mA cm-2 for cell with Pt–Ru anode. The reason of this low performance could be due to the not optimal composition of Pt-Ru anode catalyst. The Institute of Microelectronic of Barcelona-CNM (CSIC), Spain, presented a passive and silicon-based micro DMFC [113]. The device was based on a hybrid approach composed of a commercial Membrane Electrode Assembly (MEA) consisting of a Nafion® 117 membrane with 4.0 mg cm-2 Pt-Ru catalyst loading on the anode side and 4.0 mg cm-2 Pt on the cathode (E-TEK ELAT) sandwiched between two microfabricated silicon current collectors. The silicon plates were provided with an array of vertical squared channels of 300 micrometers depth that covered an area of 5.0 x 5.0 mm. The fabrication process of the silicon plates

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

DMFC Applications

55

started with a double-side polished Si wafer 500 mm thick. A first photolithography was done on the front side to define an array of squared windows with 80 mm size. Subsequently, a second photolithography was performed on the back side to define the cavity for the fuel container. Then, a deep reactive ion etching (DRIE) was realized first on the front side to obtain 200 mm-deep channels, and continued at the back until the wafer was completely perforated. These channels allowed fuel transport to the electrode surface and their dimensions were set to 80 x 80 micrometers in order to ensure the prevalence of the capillary force versus gravity in the anode side regardless of device orientation. In order to provide the current collectors with an appropriate electrical conductivity, a 150 nm Ti/Ni sputtered layer was deposited covering the front side of the wafer. This conductive layer was used as a seed layer for the 4 mm thick Ni layer that was electrodeposited afterwards. This layer enhanced the electrical conductivity of the current collector; it was then covered by a thin Au layer to prevent oxidation. Finally, the wafer was cut into 10 x 14 mm chips. In order to guarantee uniform pressure over the active area of the cell, two micromilled methacrylate pieces tightened with four bolts were used as external casing. In addition to provide a mechanical support while testing, the cell was equipped with a 100 ml methanol reservoir. The cell was tested at ambient temperature and different methanol concentrations. It was found that methanol concentration had low impact on the fuel cell maximum power density, which reached a value around 11 mW·cm-2 and was comparable to values reported in the literature for larger passive and stainless-steel fuel cells. Temperature measurements were performed; the fuel cell temperature did not change significantly and was independent from the methanol crossover rate. A research group of Yonsei University, Korea, realized a DMFC on printed circuit board (PCB) substrates by means of a photolithography process [114]. The effects of channel pattern, channel width and methanol flow rate on the performance of the fabricated DMFC were evaluated over a range of flow-channel widths from 200 to 400 µm and flow rates of methanol from 2 to 80 ml min-1. A µDMFC with a cross-stripe channel pattern gave superior performance compared with zig-zag and serpentine type of pattern. A single cell with a 200 µm wide channel delivered a maximum power density of 33 mW cm-2 when using 2 M methanol feed at 80°C. Our group (CNR-ITAE, Messina, Italy) developed passive DMFC ministacks for portable applications [115] based on simple designs. Essentially, two designs of flow-fields/current collectors for a passive DMFC monopolar three-cell stack were investigated (see Figure 13).

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

56

A. S. Aricò, V. Baglio and V. Antonucci Design A

Design B

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Figure 13. Pictures of two different monopolar plates for application in a DMFC three-cell stack operating under passive mode.

The first design (A) consisted of two plastic plates covered by thin gold film current collectors with a distribution of holes through which methanol (from a reservoir) and air (from ambient) could diffuse into the electrodes. The second design (B) consisted of thin gold film deposited on the external borders of the fuel and oxidant apertures in the PCBs where the electrodes were placed in contact. A big central hole allowed a direct exposure of electrodes to ambient air (cathodes) and methanol solution (anodes). A methanol reservoir (containing in total 21 ml of methanol solution and divided in three compartments), with 3 small holes in the upper part to fill the containers and to release the produced CO2, was attached to the anode side (Figure 14). The electrodes were composed of a commercial gasdiffusion layer-coated carbon cloth HT-ELAT and LT-ELAT (E-TEK) at the anode and cathode, respectively. Unsupported Pt-Ru (Johnson-Matthey) and Pt (Johnson-Matthey) catalysts were mixed with 15 wt.% Nafion ionomer (Ion Power, 5 wt.% solution) and deposited onto the backing layer for the anode and cathode, respectively. Nafion 117 (Ion Power) was used as electrolyte. The MEAs for the two stack designs (3 cells) were manufactured by assembling simultaneously three sets of anode and cathode pairs onto the membrane (Figures 14b), afterwards they were sandwiched between two PCBs.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

DMFC Applications

57

(a)

(b)

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Figure 14. Pictures of the DMFC design B used for a three-cell stack (a) and MEA formed by a single membrane and three couples of electrodes (b).

The geometrical area of each electrode was 4 cm2 and the total area of the stack was 12 cm2. The cells were connected in series externally through the electric circuit. The electrochemical characterization was carried out varying the catalyst loading and methanol concentration. A loading of 4 mg cm-2 Pt loading provided the best electrochemical results in the presence of unsupported catalysts. This appeared to be the best compromise between electrode thickness and amount of catalytic sites. Similar performances in terms of maximum power were recorded for the two designs; whereas, better mass transport characteristics were obtained with the design B. On the contrary, OCV and stack voltage at low current were higher for the design A as a consequence of lower methanol crossover. A maximum power of 220-240 mW was obtained at ambient temperature for the three-cell stack with 4 mg cm-2 Pt loading on each electrode using both 2 M and 5 M methanol concentration at the anode, corresponding to a power density of about 20 mW cm-2. The use of highly concentrated methanol solutions caused a significant decrease of OCV that reflected on the overall polarization curve; however, the activation losses were similar to diluted methanol solutions. An investigation of the discharge behaviour of the two designs was carried out (Figure 15).

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

58

A. S. Aricò, V. Baglio and V. Antonucci

(a)

2.1

(b)

Stack Voltagel / V

1.8

5M MeOH

1.5

Design A 1.2

Design B

0.9 0.6 0.3

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

Current / A

1.5 Design A

I = 250mA

1.2 Stack Voltagel / V

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

0.0

Theoretical

Design B

0.9

0.6

(c)

0.3

0.0 0

150

300

450

600

750

900

1050

1200

1350

Time / min

Figure 15. Polarization curves for the design A stack with a Pt loading of 4 mg cm −2 on each electrode at different methanol concentrations (a), comparison between the polarization curves obtained with the two different designs with a Pt loading of 4 mg cm −2 on each electrode and 5 M methanol solution (b), and chrono-potentiometric results at 250 mA obtained with the two designs using a Pt loading of 4 mg cm-2 and 5 M methanol solution (c).

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

DMFC Applications

59

A longer discharge time (17 h) with an unique MeOH charge was recorded with design B at 250 mA compared to the design A (5 h). This was attributed to an easier CO2 removal from the anode and better mass transport properties. In fact, in the design A, CO2 did not escape easily from the anode hindering the methanol diffusion to the catalytic sites by natural convection. When the small stack based on the A design was mechanically agitated, the effect of this forced convection increased the discharge time. As mentioned above, the potential market for portable fuel cell devices mainly concerns with small electronic devices, mobile phones and lap-top computers but it also includes weather stations, medical devices, signal units, auxiliary power units (APU), gas sensors units, etc. In this regard, a recent European project called MOREPOWER was addressing the development of a low cost, low temperature (30–60 C) portable direct methanol fuel cell device of compact construction and modular design in the range of hundred Watts power. The project was coordinated by GKSS (Germany) and included as partners Solvay, Johnson Matthey, CNR-ITAE, CRF, POLITO, IMM and NedStack. The electrical characteristics of the device were 40 A, 12.5 V (total power 500 W). The single cell performance was approaching 0.2 A·cm-2 at about 0.5 V/cell at 60 °C and atmospheric pressure [21]. Several new membranes were investigated in this project. One of the most promising was a low-cost proton exchange membrane produced by SOLVAY by using a radiochemical grafting technology (Morgane® CRA type membrane) which showed a suitable compromise in terms of reduced methanol cross-over and suitable ionic conductivity [75]. Inorganic fillers-modified SPEEK membranes were also developed in the same project by GKSS (Germany) to reduce the permeability to alcohols while keeping high proton conductivity [75].

4.2. TRANSPORTATION Though the application of fuel cells in transportation has drawn great enthusiasm and stimulation since the late 1970‘s, it is still considered a formidable venture for fuel cell powered vehicles to compete with the conventional internal combustion and diesel engine powered vehicles. This is not surprising since fuel cell development is still at an infant stage, compared to the highly advanced IC or diesel engine technology which has taken over 100 years to reach the high levels of performance with respect to operating characteristics (start-up time, acceleration, lifetime, considerable reduction in level of environmental pollutants, etc.). The impetus for developing battery and fuel cell-powered vehicles derived

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

60

A. S. Aricò, V. Baglio and V. Antonucci

from the energy crisis in 1973; in the late 1980s and 1990s the environmental legislation in United States, to reduce greenhouse gas emissions gave a further stimulation. The United States Partnership for new Generation of Vehicles Program was implemented to make ―Quantum Jumps‖ in the performance of automobiles, such as (i) tripling the efficiency for fuel consumption, (ii) reach a range of 500 km between refueling, and (iii) ultra low or zero emission of pollutants, while being cost competitive with the current automobile technology [2, 4]. Similar objectives have been addressed in European Research Programs FP5 and FP6. The only types of vehicles that have the potential of reaching these goals are IC or diesel engine/battery and fuel cell/battery hybrid vehicles. The former type of power plants are more advanced than the latter and, in fact, Toyota and Honda have commercialized IC engine/battery hybrid vehicles in the last years. Other companies have also started commercialization of diesel engine/battery hybrids [2, 4]. Nowadays, however, due to the considerable progress made in this field, DMFCs appear much more ready than in the past for the application in electrotraction systems. With the development of highly active catalysts and appropriate ionomeric membranes, these systems have been successfully operated at temperatures close to or above 100°C allowing the achievement of interesting performances [68, 89]. In particular, it was shown that the overall efficiency of recent DMFCs devices is comparable or superior to the combination of reformer-H2:air fuel cells [90]. These aspects, together with the intrinsic advantages of methanol fuel cells with respect to the hydrogenconsuming devices, mainly due to the liquid fuel feed and the absence of a cumbersome reformer, would claim for a close demonstration of DMFCs in electric vehicles. However, DMFC devices could be employed in a fuel cell vehicle if they fulfil some specific requirements in terms of power density, durability, cost and system efficiency. Accordingly, more active catalysts need to be developed together with high temperature and cross-over resilient membranes. Besides, great deal should be devoted to bipolar plates and flow-fields both in terms of design and materials. Practically all the worldwide activities on fuel cell/battery hybrid vehicles (Daimler/Chrysler/Ballard, Ford, Toyota, General Motors/Opel, Honda, Volkswagen, Fiat) are essentially on PEMFC or PEMFC/battery hybrid vehicles [2, 4]. In several demonstration vehicles, hydrogen was the fuel carried on board, mostly as a compressed gas or as a metal hydride. However, in order to meet the technical targets of the vehicle and to minimize problems caused by changes needed in the infrastructure and fuel distribution network, the emphasis has been on carrying the conventional gasoline fuel or methanol on board and processing it to hydrogen. However, due to (i) the efficiency losses in fuel processing, (ii) the

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

DMFC Applications

61

significantly high weight of the fuel processing system and (iii) the progress made in DMFC technology with respect to efficiency, specific power and power density, there is an increasing interest in developing DMFCs in the last years, a 3 kW DMFC in a one-passenger vehicle prototype was demonstrated by DaimlerChrysler/Ballard [2, 4]. Though the DMFC technology is quite promising, major breakthroughs are still needed for it to compete with the PEMFC technology, even though the latter has the burden of carrying a heavy fuel processor to produce hydrogen from gasoline or methanol or compressed hydrogen fuel [2, 4]. DMFC technology offers a solution for transportation applications in the transition towards a zero emission future. Using methanol as a fuel circumvents one of the major hurdles plaguing PEMFC technology, that is the development of an inexpensive and safe hydrogen infrastructure to replace the gasoline/diesel fuel distribution network. It has been well established that the infrastructure for methanol distribution and storage can be easily modified from the current gasoline intensive infrastructure. Another drawback in using PEMFC technology is the need to store hydrogen (at very high pressures) or carry a bulky fuel processor to convert the liquid fuel into hydrogen on board the vehicle. Methanol is an attractive fuel because it is a liquid under atmospheric conditions and its energy density is about half of that of gasoline. Despite the compelling advantages of using DMFCs in transportation applications, major obstacles to their introduction remain. These barriers include the high costs of materials used in fabricating DMFCs (especially the high cost of platinum electrocatalysts), the crossover of methanol through the electrolyte membrane from the anode to the cathode and, the lower efficiency and power density performance of DMFCs in comparison to PEMFCs. Despite these obstacles a number of institutions (particularly in the last ten years) have become actively engaged in the development of DMFCs for transport applications. The most remarkable results achieved in this field are summarized in Table 3. These institutions have directed their resources toward improving every facet of the DMFC in the quest for competitive balance with PEMFCs, as stated below. Ballard Power Systems Inc. (BPSI, Canada) in collaboration with Daimler–Chrysler (Germany) recently reported the development of a 3 kW DMFC system that is at a very preliminary stage in comparison to Ballard‘s PEMFC products [116]. Daimler–Chrysler (Germany) demonstrated this system for the transportation application in a small one-person vehicle at its Stuttgart Innovation Symposium in November 2000. The DMFC go-cart weighed approximately 100 kg, required an 18 V/1 Ah battery system for starting the electric motor on its rear wheels, and had a range of 15 km and a top speed of 35 km/h. The stack used 0.5 l methanol (the concentration of methanol was unclear)

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

62

A. S. Aricò, V. Baglio and V. Antonucci

as fuel and operated at approximately100 ◦C. In January 2001 Ballard revealed that they have built and operated a 6 kW stack (60 V) based on the same stack design as the prototype shown in Stuttgart. No details are available with respect to the stack design and performance of the DMFC power source. However, the patent literature indicates fabrication techniques for producing DMFC electrodes [117]. The anode was prepared by first oxidizing the carbon substrate (carbon fiber paper or carbon fiber non-woven) via electrochemical methods in acidic aqueous solution (0.5 M sulfuric acid) prior to incorporation of the protonconducting ionomer. Oxidation results in the formation of various acidic surface groups on the carbonaceous substrate and can be achieved by constructing a simple electrochemical cell comprising the carbonaceous electrode substrate as the working electrode. During the treatment of the carbon substrate a voltage of greater than 1.2 V and more than 20 coulombs/cm2 was used in the process. The second step involves the impregnation of a proton-conducting ionomer such as a poly(perfluorosulfonic acid) into the carbon substrate and then drying off the carrier solvent; the amount impregnated into the substrate was usually greater than 0.2 mg/cm2. The anode preparation is completed by applying aqueous electrocatalyst ink to the carbon substrate without extensive penetration in the substrate. This method ensures that less electrocatalyst is used and, the catalyst is applied to the periphery of the electrode where it will be utilized more efficiently. The performance enhancements associated with the treatment of the carbonaceous substrate may be related to the increase in the wettability of the carbonaceous substrate. This may result in the more intimate contact of an ionomer coating with the electrocatalyst thereby improving proton access to the catalyst. Another theory concludes that the presence of the acidic groups on the carbon substrate itself may improve proton conductivity or, the surface active acidic groups may affect the reaction kinetics at the electrocatalyst sites. The assembly of the MEA and single cell occurred via conventional methods, that is, hot pressing the anode and cathode to a solid polymer membrane electrolyte. Oxygen and methanol flow fields are subsequently pressed against cathode and anode substrates, respectively but details of this assembly have not been forthcoming. IRD Fuel Cell A/S (Denmark) has developed DMFCs primarily for transportation applications (0.7 kW) [118]. The stack was constructed with separate water and fuel circuits and the bipolar flow plates are made of a special graphite/carbon polymer material for corrosion reasons.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Table 3. DMFC prototypes for stationary, APU and automotive applications Single Cell/ Stack Developer

Power/ Cell Power density

T(°C)

Oxidant

Methanol Concentration (M)

Anode Catalyst

Membrane Electrolyte

Cathode catalyst

Number of cells/Surface area (cm2)

Ballard Power Systems, Inc.

3 kW

100

Air

1

Pt/Ru

Nafion

Pt

-

IRD Fuel Cell A/s

100 mW cm-2

90-100

1.5 atm air

-

Pt/Ru

Nafion

Pt

4 / 154cm2 bipolar

Thales, CNRITAE, Nuvera FCs

140 mW cm-2

110

3 atm air

1

Pt/Ru

Nafion

Pt

5 / 225cm2 bipolar

Siemens Ag

250 mW cm-2 /90 mW cm-2

110 /80

3 atm O2 (1.5 atm air)

0.5 (0.5)

Pt/Ru

Nafion 117

Pt-black

3 cm2 per cell

Los Alomos National Labs

1 kW/l

100

3 atm air

0.75

Pt/Ru

Nafion 117

Pt

30 / 45 cm2 bipolar

Thales, CRF-Fiat, CNR-ITAE, Solvay (DREaMCAR Project)

5 kW/ 160 mW cm-2

130

3 atm air

1-2

85%PtRu/C

Hyflon

60%Pt/C

100 / 300 cm2 bipolar

64

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

The MEAs had an active cell area of 154 cm2. The air pressure was 1.5 bar at the cathode. A nominal cell voltage of 0.5 V was observed for IRDs stack at a current density at 0.2 A/cm2 and electric power was generated at 15 W per cell. A consortium composed of Thales-Thompson (France), Nuvera Fuel Cells (Italy), LCR (France) and Institute CNR-ITAE (Italy) has developed a five-cell 150W stainless steel based air fed DMFC stack in the framework of the Nemecel project with financial support of the European Union Joule Program [119]. Bipolar plates were utilized in the stack design and MEAs were fabricated using Nafion as the solid polymer electrolyte and high surface area carbon supported Pt-Ru and Pt electrocatalyst for methanol oxidation and oxygen reduction, respectively. The electrode area was 225 cm2 and stack was designed to operate at 110 ◦C, using 1 M methanol and 3 atm air achieving an average power density of 140 mW/cm2. Figure 16 shows the overall stack performance. A comparison of the polarization curves for single cells in the stack and a prototypal cell is shown in Figure 17. The different diffusion characteristics of the cells in the stack indicate that the stack fluidodynamics should be enhanced in terms of homogeneity of distribution of reactant over the electrodes. Siemens Ag (Germany) optimized its DMFC system (high oxygen pressure operation) for a niche market, and examined DMFCs in the low temperature, low pressure air operation for more general purposes [120].

Figure 16. Galvanostatic polarisation and power density for a 5-cell air-feed DMFC stack at 110°C. Electrolyte Nafion 117. Catalysts: 85 % Pt-Ru/C and 85 % Pt/C; 2 mg Pt cm-2; methanol 1 M, electrode surface 225 cm2. Reprinted from Ref. [2] with permission from Wiley-VCH.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

65

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

mW cm-2

DMFC Applications

Figure 17. Galvanostatic polarisation data and power densities at 110°C for two 225 cm2 cells along a 150 W air–fed DMFC stack section (from the reactant inlet) and comparison with a 5 cm2 graphite single cell operating under same conditions and equipped with the same M&E assembly. Electrolyte Nafion 117. Catalysts: 85 % Pt-Ru/C and 85 % Pt/C; 2 mg Pt cm-2; methanol 1 M. Reprinted from Ref. [2] with permission from Wiley-VCH.

MEAs in single cells experiments are constructed using a Nafion 117 membrane, Pt-black with a catalyst loading of 4 mg/cm2 for the cathode and a high surface area Pt-Ru alloy (either unsupported or carbon supported) for the anode (2 mg/cm2). A maximum power density of 250 mW/cm2 was achieved for operating conditions of 110 °C, 3 bars O2, 0.5 M methanol and an electrode

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

66

A. S. Aricò, V. Baglio and V. Antonucci

surface area of 3 cm2. Single cell experiments exploring operating conditions at lower temperatures, lower pressures and air being supplied to the cathode electrode utilized similar MEA components as described previously. A maximum power density of 90 mW/cm2 was achieved for operating conditions of 80 ◦C, 1.5 bar air and 0.5 M methanol. These conditions resulted in a maximum power density that was significantly lower than the results obtained for previous experiments using O2 as the oxidant. It was also noted that there was a positive correlation between the air flow rate (25–100 standard cubic centimeter per minute (sccm)) and the cell performance. Siemens AG in Germany, in conjunction with IRF A/S in Denmark and Johnson Matthey Technology Center in the United Kingdom developed a DMFC stack with an electrode area of 550 cm2 under the auspices of the European Union Joule Program [121–123]. The projected cell performance was a potential of 0.5 V at a current density of 100 mA/cm2, with air pressure at 1.5 atm and the desirable stoichiometric flow rate. A 3-cell stack has been demonstrated by operating at a temperature of 110 ◦C and a pressure of 1.5 atm and using 0.75 M methanol, this stack exhibited a performance level of 175 mA/cm2 at 0.5 V per cell; at 200 mA/cm2 the cell potential was 0.48 V. These performances were obtained at a high stoichiometric air flow rate (factor of 10) but in order to reduce auxiliary power requirements, one of the goals at Siemens was to improve the design to lower the air stoichiometric flow to the desired value of about a factor of two. A 0.85 kW air-fed stack composed of 16 cells and operating at 105 ◦C was successively demonstrated with maximum power density of 100 mWcm−2. Los Alamos National Laboratory (LANL) is also actively pursuing the design and development of DMFC cell stacks for electric vehicle applications. According to the latest available information, a five-cell short stack with an active electrode area of 45 cm2 per cell has been demonstrated [68, 89, 124]. The cells were operated at 100 ◦C, an air pressure of 3 atm and a methanol concentration of 0.75 M. The maximum power of this stack was 50 W, which corresponds to a power density of 1 kW/l. At about 80% of the peak power, the efficiency of the cell stack with respect to the consumption of methanol was 37%. Among the recent European community projects dealing with the development of DMFCs for automotive and APU applications, it should be mentioned the DREAMCAR project (ERK6-CT-2000-00315) that was carried out in the framework of the FP5 EC program. Dreamcar was the acronym of Direct Methanol Fuel Cell System for Car Applications; the project was coordinated by THALES ENGINEERING and CONSULTING (France), and included as partners CRF- FIAT (Italy), CNR-ITAE (Italy), SOLVAY (Belgium) and TAU-RAMOT (Israel) [81]. The main objective of the project was to design, manufacture and

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

DMFC Applications

67

test a 5 kW stack at high temperature (up to 140°C). There were three main research topics in the Dreamcar project: higher operating temperature (up to 140°C) to enhance the electrochemical reactions, development of new fluorinated (improvement of the membranes developed in the frame of a previous project NEMECEL JOE3-CT-0063) and hybrid inorganic-organic membranes, development of new carbon supported Pt-alloy catalysts to increase the efficiency of the electrodes and power density [81].

Cell potential / V

0.80

90°C 100°C 110°C 120°C 130°C 140°C

Anode: CNR 85%PtRu/C Cathode: CNR 60%Pt/C Hyflon Membrane

0.60 0.40 0.20

Air Feed 1 M MeOH

0.00 0.5

1

1.5

Current density / A cm

-2

0.3

90°C 100°C 110°C 120°C 130°C 140°C

Air Feed 1 M MeOH

-2

Power density / W cm

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

0

0.2

0.1 Anode: CNR 85%PtRu/C Cathode: CNR 60%Pt/C Hyflon Membrane

0.0 0

0.5

1

Current density /A cm

1.5 -2

Figure 18. DMFC polarization and power density curves for a Hyflon Ion membrane-based MEA. Reprinted from Ref. [64] with permission from Elsevier.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

.

68

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

The Solvay Solexis Hyflon membrane was selected for the final stack. The Hyflon properties are discussed in detail in the section concerning with the present status of DMFC membranes. In order to allow stack operation at high temperature with the Hyflon membrane, the operating pressure was 3-4 bar abs. The performance of the MEAs was first investigated in a single cell configuration based on the same materials of the final stack. The electrochemical behavior is reported at various temperatures in Figure 18 [64]. A performance approaching 300 mW cm-2 with a Pt loading of 2 mg cm-2 at 140 °C and 3 bar air feed was obtained. In the framework of the same project it was also developed a nanoporous proton conducting membrane (NP-PCM) that showed superior performance in the presence of a liquid TFMSA acid electrolyte [66]. Yet, the use of an acidic liquid electrolyte, necessary to make the NP-PCM membrane conductive, was considered incompatible with the materials used in the construction and test of the final stack (severe problems of corrosion and fluid management) [81]. The final stack consisted of 100 cells of 300 cm2 and provided an output electrical power of about 5 kW. This 5 kW DMFC stack developed in the Dreamcar project is shown in Figure 19.

Figure 19. A 5 kW DMFC stack developed in the framework of the DREaMCAR project.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

DMFC Applications

69

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

The specific power output was 110 W/l The average single cell performance in the final stack was about 160 mWcm-2 compared to the 300 mWcm-2 that was almost achieved in the single cell with the same membrane/electrode materials [81]. The main drawbacks concerned with the heat management since a significant heat/energy was dissipated during operation at 130/140°C at 3-4 bar by using an external radiator [81]. Regarding the MEA components, the methanolcross over was recognized as one of the aspects limiting the performance; on the other hand, the reaction kinetics were significantly accelerated by the high operating temperature. Despite the various attempts made to demonstrate the validity of DMFCs for transportation, the technology that is presently considered more appropriate for electro-traction consists in a hybrid system using both pure hydrogen-fed PEMFCs and advanced Li-batteries. This approach is especially investigated by some Japanese car manufacturers including Honda [125]. To make DMFCs competitive with regard to this technology, it is essential to increase the power density, decrease methanol cross-over and reduce the cost. Regarding the electrolyte, it is required an appropriate membrane operating in a range that varies from sub-zero to 130 °C. The same requirements concern with the membranes for PEMFCs. In general, high temperature stack operation will simplify heat and water management.

4.3. DISTRIBUTED AND REMOTE GENERATION OF ELECTRICAL ENERGY Hydrogen fuel must be extracted from fossil fuels or water which are both energy-consuming processes. Once produced, the gas must be compressed or liquefied for distribution, and this process and the distribution itself take yet more energy. By the time the hydrogen has been delivered to the fuel cell for conversion to electricity, then, a significant amount of energy has been lost in these processes. Liquid-feed fuel cells appear a promising alternative to the hydrogen and natural gas consuming fuel cells for specific applications in distributed and remote generation of electrical energy since they allow easy handling and storing of the liquid fuel. Methanol is currently distributed by rail, barge, and truck in several countries for using as an octane enhancer or oxygenate blended with gasoline. As such, methanol infrastructure to the terminal level is complete and the gasoline industry has experience in handling and blending methanol.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

70

A. S. Aricò, V. Baglio and V. Antonucci

Small fuel cells, typically less than 10 kW, are under consideration for many applications that traditional electric utilities have addressed. These applications are a subset of distributed generation. In this context, fuel cells may enable new companies to enter the power-generation business as equipment providers or electricity providers. For example, companies that supply natural gas or propane are forming partnership with fuel-cell companies to bring fuel-cell-based electricity to consumers in remote locations that are off the grid. Among the applications under consideration there are:   

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.



backup for residential, commercial, or industrial users high quality power for commercial or industrial users remote power for residential, commercial, or industrial users in developed or developing counties grid-independent power for residential, commercial, or industrial users

Resource issues are some of the most significant in the stationary power sector, particularly for companies that want to introduce large, grid-connected units. The scale of investments in such units is significant and risks are high if systems do not work properly or integrate properly with the grid. The stationary sector appreciates the following attractive attributes of fuel cells: 





Power quality and reliability. Some manufacturing operations are very sensitive to power fluctuations and disruptions and will pay a premium for high-quality, reliable power. Environmental benefits. Building new conventional power sources, particularly in environmentally sensitive areas or near urban areas, is increasing in difficulty. Remote power. Power transmission and distribution are expensive. Placing a plant near the point of power use avoids these costs.

Smaller plants, such as those for residential units, can properly utilize the DMFC technology. Yet, until now, only small attempts have been made to use the DMFC technology in the field of distributed and remote generation of electrical energy. Any developments in materials or system integration that improve the economics of these technologies will advance fuel cell use in stationary power applications. Deregulation of the electric utility industry in European countries has led to the creation of numerous energy-service companies, which view FCs as an

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

DMFC Applications

71

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

attractive technology for introducing on-site power generation to small and midsize electricity users. Distributed generation is increasingly attractive to utility companies, because it can improve reliability, increase asset use, and avoid transmission and distribution costs. FCs suit distributed generation. Establishing codes and regulations that allow fuel cells to connect safely to the grid is important if they are to achieve widespread commercial use. At the present time, DMFCs are considered in the rated power level of 1-10 kW for remote residential, uninterrupted and standby power. However, PEMFCs are at present more advanced than DMFCs for this application. The competition comes from IC and diesel engine generators but fuel cells have the upper hand because they produce lower levels of environmental and noise pollution. Most of the PEMFCs being developed for residential applications (IFC, Plug Power, H Power) are fuelled with natural gas, with the system providing both electricity and heat [55, 56]. In the last 10 years, the progress in the DMFC technology has raised the hope of its application for electric vehicles; this will in our opinion promote the use of DMFCs also in small stationary applications where the main requirements concern mainly with a simplified system characterized especially by high energy density being the high power density a minor requisite for this application.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Chapter 5

5. FUNDAMENTAL ASPECTS: STATUS OF KNOWLEDGE 5.1. METHANOL OXIDATION PROCESS

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

5.1.1. Electrode-Kinetics and Electrocatalysis Most of fuel cell systems operate with hydrogen as a fuel. Electro-oxidation of hydrogen is a fast reaction on a low loading Pt electrocatalyst (0.05 to 0.1 mg cm-2) in the presence of a proton conducting eelctrolyte. On the other hand, the methanol oxidation reaction rate is at least three to four orders of magnitude slower on a high Pt-Ru loading ( 2 mg cm-2) electrocatalyst. Although thermodynamic characteristics are similar to the hydrogen reaction, especially in terms of reversible oxidation potential, the methanol electro-oxidation reaction is a slow process and it involves the transfer of six electron to the electrode for a complete oxidation to carbon dioxide. Anode reaction:

(1‘)

Various reaction intermediates may be formed during the methanol oxidation [33]. Some of these (CO-like) species are irreversibly adsorbed on the surface of the electrocatalyst and severely poison Pt for the occurrence of the overall reaction, which has the effect of significantly reducing the efficiency for fuel consumption and the power density of the fuel cell. Thus, it is very important to develop new electrocatalysts to inhibit the poisoning and significantly increase the electro-oxidation rate by at least a factor of two-three times. Other species may be

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

74

A. S. Aricò, V. Baglio and V. Antonucci

released with a consequent decrease of fuel efficiency, whereas an efficient catalyst must allow a complete oxidation to CO2. Principal by-products in the methanol oxidation are formaldehyde and formic acid. Methyl formiate and other substances have been found in traces.

(17) (18)

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

(19) DMFCs are characterized by two slow reactions, i.e. methanol electrooxidation and oxygen reduction with the further drawback of the presence of a mixed potential at the cathode determined by the methanol cross-over. On the other hand, as discussed in the previous section, there are specific advantages in using methanol as direct fuel especially for what concerns costs, simplicity of design, large availability, easy handling and distribution. Another concern is that, even though significant progresses have been made in enhancing the electrocatalysis of the four-electron transfer oxygen reduction reaction at low temperatures, the overpotential of this reaction at desired current densities (e.g. 500 mA cm-2) is still about 400 mV in H2/air fuel cells and increases by about 50100 mV in DMFCs because of the effect caused by methanol, which crosses over from the anode to the cathode [1, 7, 8]. In the last decades, a significant amount of investigations has been carried out in the field of electro-oxidation of methanol at low temperature [33] to further elucidate the mechanism that was explored since the 1950s. Most of these fundamental studies have been carried out in half cell configuration and on smooth electrode surfaces, in order to individuate the best electrocatalyst composition [33]. Electrochemical investigations have been generally carried out in combination with spectroscopic techniques in order to elucidate the oxidation mechanism and to investigate the irreversibly adsorbed species on the electrode surface. From a general point of view, almost all electro-oxidation reactions involving low molecular weight organic molecules, such as CO, CH3OH, C2H5OH, HCOOH, HCHO, require the presence of a Pt-based catalyst, at least in the presence of a proton conducting electrolyte [1, 8, 33, 126, 127]. Whereas, cheaper Ni-based anodes can be also used in alkaline electrolytes. In the presence of Pt, all these electro-oxidation reactions give rise to the formation of strongly

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

75

adsorbed CO species in linear or bridge-bonded form. Pt is involved in two key steps occurring during the oxidation route. One is the dehydrogenation step (e.g., in the case of methanol) and the second one is the chemisorption of CO. Studies involving partial substitution of Pt with other transition metals like W, Pd, Ni, Ti, Rh, Mo have not yielded beneficial results [33, 126-131]. The acidic electrolytes have been preferred for several decades because of the carbon dioxide rejecting properties. Accordingly, most of the work has been addressed to the modification of the Pt environment by alloying it with other elements or through the synthesis of multifunctional electrocatalysts [28, 132-145]. The alloying route has given until now the most successful results. Only in a few cases, it has been reported that bimetal catalysts formed by nanosized particles of each element showed superior activity. It was discussed in the section 3.1 that the alloying of Sn and Ru with Pt gives rise to electrocatalysts which strongly promote the oxidation of both methanol and CO. Since the complete oxidation of methanol to CO2 involves the transfer of 6 electrons to the electrode, the overall reaction mechanism involves several steps including dehydrogenation, chemisorption of methanolic residues, rearrangement of adsorbed residues, chemisorption of oxygenated species ( preferentially on the alloying element ) and surface reaction between CO and OH to give rise to CO2. The mechanism of methanol electro-oxidation was elucidated in the last three-four decades by using a variety of experimental procedures [1, 8, 33, 126, 127]. The electrochemical methods used for such investigation include mainly steady-state galvanostatic polarizations, cyclic voltammetry and electrochemical transients (chronoamperometry, chronopotentiometry) [1, 8, 29, 33, 35, 36, 126, 127]. Ac-impedance spectroscopy has been used to a minor extent [20]. In general, electrokinetic parameters, i.e. Tafel slopes, activation energies, reaction orders etc., have been derived by galvanostatic steady-state polarizations and linear potential scans in the presence of acidic liquid electrolytes in a temperature range from 25 to 80 °C [30, 34, 138]. In situ spectroscopic analyses in conjunction with electrochemistry such as FTIR, mass spectrometry, XAS etc. [35-37, 139142] have been demonstrated to be very useful for the investigation of adsorbed species and for determination of intermediate compounds formed during the oxidation process. Ellipsometry has been particularly used to in-situ investigate the potentials at which water discharging was occurring especially on Pt-Ru surfaces [38]. The experimental techniques and the information gained from them have been reviewed in a few excellent papers [1,8, 33, 126, 127]. Beside these methods, in the recent years additional techniques and/or methodologies of investigation of methanol electro-oxidation have been proposed. Among them a, significant interest has arisen around the following procedures:

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

76

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

(i) ultra high vacuum preparation of well defined catalyst surfaces for voltammetric analyses [137]; (ii) in-situ electrochemical microbalance based studies for detection of adsorbed species [143-144]; (iii) in-situ viewing of catalysts surfaces during methanol oxidation by scanning tunneling microscopy [145]; (iv) in-situ electrochemical NMR studies allowing electronic-level description of metallic-shells in nanoparticle catalysts [146]; (v) in-situ carbon monoxide stripping voltammetry [147-150]; (vi) rotating ring disk method with a recast film of catalyst particles/Nafion deposited on an inert glassy carbon electrode to screen electrocatalysts [151-152]; (vii) combinatorial methods for the screening of a large variety of catalyst formulations [44]. Some of these techniques are still dealing with smooth electrode surfaces. But several efforts have been made to carry out such advanced studies by using practical nanosized electrocatalysts. The acquired information served mainly to understand the mechanism at a basic level. Yet, there is a need to develop methodologies which allow to get information on the mechanisms occurring at practical electrodes and during fuel cell operation. In this regard in-situ CO stripping voltammetry is a very interesting methodology for the characterization of the catalyst / solid polymer electrolyte interface under fuel cell conditions allowing the evaluation of active surface area, intrinsic catalytic activity and, in some cases, surface composition of the catalysts [150]. The relationship between the anodic shift of the peak potential of the CO stripping and the catalytic activity towards methanol oxidation provided a clear evidence that CO removal is a rate determining step in the methanol electro-oxidation process [150]. Combinatorial analysis was addressed to the discovery of new catalyst formulations and it has recently attracted a significant interest [44]. These studies have pointed out the need of a systematic screening of a large number of combination of elements as a suitable route to address the problem of finding more active catalysts for methanol oxidation. As a result of these studies Pt-Ru-Os was proposed as one of the most interesting catalyst formulations. The electrokinetic parameters have been mainly determined at Pt-based electrodes. A Tafel slope between 90 and 120 mV dec-1 was determined for the methanol oxidation reaction in the temperature range between 25 and 60 °C [36, 37, 133]; this may vary depending on the potential region under analysis. A fractional positive reaction order with respect to methanol concentration has been

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Fundamental Aspects: Status of Knowledge

77

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

observed in half cell studies in a range of concentrations from 0 to 2.5 M [30]. The Temkin adsorption isotherm has been often used to describe the adsorption of methanolic residues [30, 153, 154] and various kinetic equations describing the variation of current as a function of potential, temperature, coverage of adsorbed species have been proposed [154-155, 156, 157]. Bagotzky and Vassilyev derived functional expressions for methanol electrocatalysis at Pt based electrodes for two potential regions characterized by significantly different rates in water discharging i.e. for slow and fast water discharging process on the surfaces [30]. At a fixed potential, the current density (i / mA cm-2) varies with the coverage on the basis of the following relationships: i = k OH exp (1 r1 Org)

(low potentials)

(20)

i = k exp (1 r1 Org)·exp(2 r2 OH)

(high potentials)

(21)

where k is a kinetic constant, Org and OH are the coverages of methanolic residues and oxygen species, respectively,  and r are electro-kinetic and Temkin parameters, respectively [30]. The latter takes into account the effects of the heterogeneous surface and the interaction between adsorbates. The second relationship holds in the potential region where water discharging and consequent OH chemisorption on the catalyst surface occurs at acceptable rates. For a Pt-Ru surface this region shifts to very low potentials. Although more recent studies have proposed alternative and more complex equations [156, 157], the functional expression proposed by Bagotzky and Vassilyev has been demonstrated to be useful in interpreting kinetic data for a wide variety of catalyst-electrolyte interfaces. The activation energy for methanol electro-oxidation strictly depends on the catalytic system; Pt-Ru shows activation energy of 30-65 kJ mol-1[34, 138, 158]. Yet, this value is significantly high and it clearly indicates the need of high temperature operation. The exchange current density (Io) for methanol oxidation at various Pt-based catalysts is a few orders of magnitude lower than that of H2 oxidation [2, 159]. Mass and specific activities at defined potentials, where the reaction is under kinetic control (e.g. 0.25 V - 0.35 V vs. RHE for high temperature operation), are often utilized to assess catalysts‘ reactivity. This probably depends on the fact that it is not always possible to determine the Tafel slope and/or the Tafel slope changes in the various potential regions [29, 34, 157, 158]. Electro-oxidation of methanol is a structure sensitive reaction. The first studies carried out on Pt surfaces with different crystallographic orientations and for polycristalline Pt evidenced an increased reactivity for methanol oxidation at

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

78

A. S. Aricò, V. Baglio and V. Antonucci

the Pt planes with (110) Miller indexes [160] and for rough surfaces [37, 136]. Yet, Wieckowski et al. [161, 162] showed for Pt-Ru surfaces that the catalytic activity is maximized by the presence of (111) crystallographic planes. These evidences may be explained by the fact that water discharging at Pt sites occurs very slowly at low electrode potentials; thus, for a pure Pt catalyst, Pt sites with high coordination numbers or surfaces with a large number of defects are needed. These sites are capable of chemisorbing oxygen molecules at lower potentials. As opposite, in the Pt-Ru system, water discharging occurs at low potentials on Ru sites, whereas, methanol chemisorption involves three neighboring Pt sites. CO removal which is often the r.d.s. at Pt-Ru surfaces [150] requires the presence of OH species adsorbed on adjacent Ru sites. The latter processes is accelerated by the presence of low Miller index planes. Implications for the development of high surface area catalysts are discussed in a following paragraphs. The electrochemical oxidation of methanol on Pt involves several intermediate steps, i.e. dehydrogenation, CO-like species chemisorption, OH (or H2O) species adsorption, chemical interaction between adsorbed CO and OH compounds, and CO2 evolution. One of these steps is the rate determining step (r.d.s.) depending on the operation temperature and particular catalyst surface (crystallographic orientation, presence of defects, etc.). The state of the art electrocatalysts for the electro-oxidation of methanol in fuel cells are generally based on Pt alloys supported on carbon black [126, 127, 163], even if the use of high surface area unsupported catalysts has recently gained momentum [164]. As stated above, the electrocatalytic activity of Pt is known to be promoted by the presence of a second metal, such as Ru or Sn, acting either as an adatom or a bimetal. The mechanism by which such a synergistic promotion of the methanol oxidation reaction is brought about has been the subject of numerous studies during the last 30-40 years, in which various spectroscopic methods have been employed in conjunction with electrochemistry [1, 8, 33, 126, 127]. A combination of cyclic voltammetry [136], in-situ ellipsometry [38], X-ray absorption spectroscopy [41], on-line mass spectrometry [140] and in-situ FTIR spectroscopic studies [36, 37, 139], revealed that the electro-oxidation of methanol on Pt-based catalysts proceeds through the mechanism (sequence of non-elementary reaction steps) below described. First a sequence of dehydrogenation steps gives rise to adsorbed methanolic residues, according to the following scheme: CH3OH + Pt  Pt-CH2OH + H+ + 1 e-

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

(22)

Fundamental Aspects: Status of Knowledge

79

Pt-CH2OH + Pt  Pt2-CHOH + H+ + 1 e-

(23)

Pt-CHOH + Pt  Pt3-CHO + H+ + 1 e-

(24)

After the initial dehydrogenation steps, the methanolic residue is anchored to three adjacent Pt sites. A surface rearrangement of the methanol oxidation intermediates gives carbon monoxide, linearly or bridge-bonded to Pt, as following: PtCHO  PtCO + H+ + 1e-

(25a)

or Pt PtCHO + Pt 

C=O + H+ + 1ePt

(25b)

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

In the absence of a promoting element, water discharging occurs at high anodic overpotentials on Pt with formation of Pt-OH species at the catalyst surface: Pt + H2O  PtOH + H+ + 1e-

(26)

The final step is the reaction of Pt-OH groups with neighboring methanolic residues to give carbon dioxide: PtOH + PtCO  2Pt + CO2 + H+ + 1e-

(27)

The overall oxidation process of methanol to carbon dioxide proceeds through a six electron donation process; yet, the rate determining step, derived from electrochemical steady-state measurements on Pt, through analysis of the Tafel slope, involves a one-electron step [36]. On a pure Pt surface, the dissociative chemisorption of water on Pt is the rate determining step at potentials below  0.7 V vs. RHE, i.e. in the potential region that is of technical interest [36]. It is generally accepted that an active catalyst for methanol oxidation should give rise to water discharging at low potentials and to "labile" CO chemisorption. Moreover, a good catalyst for methanol oxidation should also catalyze the oxidation of carbon monoxide.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

80

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Even if various theories have been put forward to explain the promoting effect of the additional elements [28, 29, 41, 165], the subject remains controversial. It was determined that transition metal promoters and adatoms improve the electrocatalytic behavior of electrodes, either by minimizing the poisoning species or enhancing the main oxidation process. As commented in section 3.1, three main hypotheses have been made. A first hypothesis suggests that the metal promoters and adatoms either alter the electronic properties of the substrate or act as redox intermediates [31, 41, 166]. A second hypothesis envisages ad-atoms as blocking agents for the poison forming reaction, assumed to occur on a number of sites greater than those required for the main reaction [31]. The influence of a possible steric effect on the enhanced oxidation rate was also considered [31]. A third hypothesis based on the bifunctional theory invokes a mechanism by which the oxidation reaction of either the fuel or the poisoning intermediate is enhanced by the adsorption of oxygen or hydroxyl radicals on promoters or adatoms adjacent to the reacting species [29]. Combining the electronic and bifunctional theories it is derived that the role of the second element is to increase the OH adsorption on the catalyst surface, at lower overpotentials, and to decrease the adsorption strength of the poisoning methanolic residues.

5.1.2. Pt-Alloys and Non-Noble Electrocatalysts: Structural and Morphological Properties As discussed in the section 3.1, the anode catalyst screening studies, indicated essentially the Pt-Ru binary alloy electrocatalyst as the most promising formulation. According to the mechanism shown above it is generally accepted that Pt sites in Pt-Ru alloys are especially involved in both the methanol dehydrogenation step and strong chemisorption of methanol residues. At suitable electrode potentials (0.2 V vs. RHE), water discharging occurs on Ru sites with formation of Ru-OH groups at the catalyst surface [38]: Ru + H2O  Ru-OH + H+ + 1e-

(28)

The final step is the reaction of Ru-OH groups with neighboring methanolic residues adsorbed on Pt to give carbon dioxide: Ru-OH + Pt-CO  Ru + Pt + CO2 + H+ + 1e-

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

(29)

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

81

The chemisorption process of methanol on Ru sites is significantly less favored than on Pt sites, but it is strongly activated by the temperature [135]; methanol dehydrogenation occurs on Pt at potential values below 0.2 V vs. RHE [18]. Besides, the water discharging reaction, producing oxygen species chemisorbed on Ru sites, occurs at potentials as low as 0.2 V vs. NHE. Although at high temperatures (90° to 130 °C) Ru can participate in the methanol chemisorption process at low potentials, the chemisorption energy of oxygen on Ru surfaces is so high (-330 kJmol-1) [167] to inhibit Ru sites to be covered by methanolic residues; hence, they can suitably chemisorb OH groups. The CO removal process is also activated by an increase in temperature and in overpotential [168]. Although, the subject remains still controversial [29, 135, 143, 144, 152], an optimal Ru content in carbon supported Pt-Ru catalysts for the methanol oxidation reaction at high temperature 90-130°C has been found to be 50 at %. The optimum Ru surface composition is referable to the relevant synergism accomplished by a Pt-Ru surface with 50% atomic Ru in maximizing the product of OH (OH coverage) and k (intrinsic rate constant), according to the assumption of the surface reaction between COads and OHads as rds; on the other hand, methanol adsorption and dehydrogenation processes, as well as water discharging, are considered to be in equilibrium at high temperatures. At low temperatures the situation is somewhat different. Adsorption of methanol on Pt requires an ensemble of three neighbor atoms whereas oxidation of methanolic residues requires a chemical interaction with neighbor adsorbed OH groups on Ru sites. Gasteiger et al. [135] have observed that methanol oxidation occurs more readily at room temperature on pure Pt-Ru alloys having low Ru content ( 10%) whereas at intermediate temperatures (60C) the reaction is faster on alloys with increased Ru content ( 33%). An optimal Ru content in unsupported Pt-Ru catalysts for the methanol oxidation reaction at 130 °C has been found to be about 50 at % [169]. These results can be suitably interpreted by considering that dehydrogenation of methanol is the rate determining step at room temperature. Thus, electrocatalysts having a good fraction of ensembles of three neighbor Pt sites appear to give higher electrochemical activities. At higher temperatures methanol dehydrogenation is thermally activated on Ru sites which participate more actively in the methanol chemisorption process. Under such conditions, the rate controlling steps become the water discharging on Ru sites and the chemical interaction between adsorbed methanolic residues and OH groups. Analysis of specific and mass activity of unsupported Pt-Ru catalysts with different composition at high temperature (130 °C) in a DMFC [169] indicates the presence of a maximum for a 1:1 Pt-Ru atomic ratio (Figure 4).

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

82

A. S. Aricò, V. Baglio and V. Antonucci

The synergistic promotion exerted by Pt-Ru alloys is supported by X-ray absorption analysis [41]. Accordingly, an increase of Pt d-band vacancies is produced by alloying with Ru; this possibly modifies the adsorption energy of methanolic residues on Pt. Such evidences suggest that the reaction rate is not only dictated by the bifunctional mechanism but it is also influenced by electronic effects occurring on account of the interaction between Pt and Ru [41, 166]. The achievement of significant reaction rates for the Pt-Ru system requires the use of highly dispersed metal phase on electrically conductive carbon black support. In this latter system, the preparation procedure of the electrocatalyst was found to strongly influence its performance for methanol oxidation [8, 152]. Following various studies carried out in this field, the starting point for the development of an active electrocatalyst for CH3OH oxidation is to synthesize a suitable highly dispersed Pt-Ru phase. A few authors have pointed out the need to develop multifunctional catalysts, thus, often the Pt-Ru system is modified with the addition of a third functionality (Pt-Ru-Mo [6], Pt-Ru-W [170], Pt-Ru-Sn [132, 160], Pt-Ru-Os [44]). It was discussed in the section 3.1 that both Pt-Ru and Pt-Sn systems have been reported to be promising catalysts for electro-oxidation of methanol in direct methanol fuel cells 23, 160]. But although there is conclusive evidence on catalytic promotion of methanol electro-oxidation on the Pt-Ru system in relation to Pt, contradictory results have been reported in the literature on the promotional effect of tin for this reaction [28, 137, 171]. While Janssen and Moolhuysen [28] and Watanabe et al. [172] have found a 50-100 times enhancement in the electrocatalytic activity of smooth and electrodeposited Pt electrodes on which Sn was deposited, several other authors [171] have reported considerably smaller and even negative effects. For carbon-supported Pt-Sn electrodes, different methods of are found to exhibit varying effects [173] leading to different Pt-Sn surface properties. The surface characteristics of Pt-Ru and Pt-Sn are of considerable significance for the methanol-oxidation reaction. Since both Pt-Ru and Pt-Sn systems have been reported to promote electrooxidation of methanol, different electronic environments around Pt-sites in Pt-Ru and Pt-Sn should imply different routes for methanol-oxidation reaction. As stated above, the most acceptable mechanism for methanol oxidation on Pt-alloys is based on the bifunctional theory 29]. According to this theory, Pt-sites adsorb methanol through a dehydrogenation step whereas the alloying element, namely Ru or Sn, adsorbs oxygenated species from water. The methanolic residues adsorbed on Pt-sites react with the oxygenated species present on the neighboring Ru or Sn sites in the alloy producing CO2.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

83

In a previous study from our group [174], we observed a 1.1 eV shift in the X-ray near edge spectrum (XANES) of the Pt-Sn/C catalyst with respect to the Pt/C catalyst, from which result one may infer that Sn atoms in the Pt-Sn alloy donate electrons to Pt-atoms and are thereby oxidized. This is in agreement with the electronegativity values for elemental Pt and Sn which differ appreciably. A noticeable charge transfer from Sn to Pt has been also found in an XPS analysis, as reflected by a negative shift in the Pt-4f binding energies of the Pt-Sn samples in relation to Pt/C electrocatalysts [175]. Mukerjee and McBreen [41, 176] have shown in their X-ray absorption spectroscopic studies on Pt-Ru/C and Pt-Sn/C electrocatalysts that alloying Sn with Pt causes partial filling of Pt 5d-bands and a consequent increase in the Pt-Pt bond distances with a contrary effect observed for Pt-Ru catalysts. However, in other investigations, XPS data on Pt-Ru/C samples have not shown any evidence of modification in the Pt-4f spectra as compared with that in Pt/C samples, which showed consistently nearly the similar electronegativity values for elemental Pt (2.2) and Ru (2.1). Beside the work of Mukerjee and McBreen [41], some in situ experiments have also provided evidence for the influence of Ru on the electronic state of Pt [177]. Iwasita et al. [166] observed in their FTIR experiments a shift towards higher frequencies for the CO stretching on Pt-Ru with respect to Pt. This effect has been attributed to a lower chemisorption energy for CO on the Pt-Ru alloy. Frelink et al. [143, 144] measured the stretching frequency of linearly bonded CO as a function of coverage for the Pt and Pt-Ru electrodes. A shift to higher frequencies at various coverages was observed as a result of change in the CO binding strength to the surface induced by Ru through a ligand effect on Pt [143, 144]. But, these authors have also shown by in-situ ellipsometric studies on Pt and Pt-Ru electrodes that the Ru-oxide film is removed by reaction with methanol during its electro-oxidation. This evidence supports the bifunctional mechanism. The latter findings were further corroborated by DEMS studies [143, 144]. In a recent article, Iwasita and co-workers have emphasized the bifunctional mechanism for methanol electro-oxidation on the Pt-Ru system [178]. In our opinion, it may be rationalized from the in-situ experiments described above that the promotional effect of Ru could be due to both the contribution of bifunctional and ligand effects. But, the ligand effect in the Pt-Ru system appears to be significantly less prominent and it has still not found proper support in the results obtained from a few surface techniques including XPS [175]. Further insights on the mechanism can be obtained by analyzing the Ptcarbon monoxide bonding since CO removal from Pt surfaces plays a dominant role in methanol oxidation. The nature of Pt-CO bond in platinum systems has been well documented in the literature [179]. It would, however, be in context to

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

84

A. S. Aricò, V. Baglio and V. Antonucci

elucidate the nature of bonds formed between CO and Pt as well as between OH and Ru (or Sn) in the respective alloy systems. Chemisorption of CO on Pt involves donation of an electron pair from * anti-bonding orbitals of CO to the unfilled 5d-orbitals of Pt. A back donation of electrons from the Pt metal to COorbitals further stabilizes their interaction. Accordingly, it is apparent that the dative electron donation from CO to Pt is a pre-requisite for a strong COchemisorption. If the Pt surface is partially oxidized, the Pt metal can more easily accommodate the electrons donated by CO in its unfilled orbitals. The increase in the Pt 5d-band vacancies produced by Ru atoms, as interpreted by Mukerjee and McBreen 41], would imply an increase in Pt-CO bond strength which in effect would retard the methanol-oxidation reaction. On this basis, the promoting action of Ru arises mainly due to an easier chemisorption of oxygen from water at lower electrode potentials in accordance with the bifunctional theory. Nevertheless, some modifications of the electronic environment around Pt may occur during the interface formation, since the uptake of oxygen functionalities from water could modify the electron accepting characteristics of Ru atoms. On the other hand, a charge transfer from Sn to Pt in the Pt-Sn system, evidenced from XAS and XPS data, causes a substantial increase in electron density around Pt-sites resulting in a weaker chemisorption of CO. This combined with an easier adsorption of OHspecies on the oxidized Sn-sites at lower electrode potentials would enhance the oxidation of irreversibly adsorbed CO to CO2, and would thus promote the methanol-oxidation reaction. The electro-catalytic promotion of CO electrooxidation on Pt-Sn alloy at low electrode potentials has been discussed by Markovic‘ et al. [42]. The Pt3Sn bulk alloy was found significantly more active than the Pt-Ru system for CO oxidation. However, higher electrode potentials produce a negative effect due to strong adsorption of OH-species on Sn-sites and retard the methanol-oxidation reaction on Pt-Sn catalysts. In many cases formation of stable tin-oxide species on the surface during catalyst preparation or electrode activation procedures has been observed [174]. The presence of these compounds on the electrode surface play a significant role during the methanol oxidation reaction since the electrocatalyst is operated in a potential regime in which oxide must not be present but where the supply of active oxygen is important. Formation of labile-bonded oxygen species more easily occurs on an oxide-free surface and in this context the water discharging reaction occurs more easily on Ru than Sn sites. Generally, the achievement of properly reduced Pt-Sn surfaces is brought about with difficulty. In this context, Pt surface containing Sn ad-atoms often perform better than bulk or carbon supported Pt-Sn alloy catalysts [180].

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

85

From all these evidences, one may conclude that the addition of Ru to Pt markedly increases the electrocatalytic activity of Pt through the adsorption of oxygenated species on Ru-sites. By contrast, the promotion of CO oxidation reaction on Pt-Sn catalyst appears to be mainly due to a modification of the electronic environment around Pt-sites. It becomes clear that Pt-Ru catalysts are more effective for methanol oxidation since the reaction desires the electrocatalyst to be operated in a potential regime where labile-bonded oxygen should be present on the surface. In this situation, the supply of active oxygen to the surface is of paramount importance since this apparently would facilitate the oxidation of adsorbed methanolic residues to CO2. The presence of strongly-bonded oxygen species on Sn-sites in the Pt-Sn system limits the oxidation of methanol to CO2. Some Pt-Ru catalysts show a lower Ru content in the Pt-Ru alloy (as detected by XRD) with respect to the nominal composition [10, 164]. In such cases, the remaining Ru atoms are present as RuOx species that may be composed of crystalline tetragonal RuO2 or amorphous Ru-oxide [10, 164, 181]. The oxide species are generally detected by XRD or TEM if they have a crystalline structure (frequently RuO2 tetragonal), whereas X-ray absorption spectroscopy (both EXAFS and XANES) and in some cases XPS may give information on the amorphous Ru-oxide species [181]. Often, TEM micrographs of the carbon supported Pt-Ru sample show some agglomeration of particles having similar dimension but characterized by a different atomic contrast [10]. The interlayer spacing of Pt-Ru alloy particles that may be observed at high magnification is about 2.25-2 Å, corresponding to the (111) planes of the Pt or Pt-Ru fcc lattice. The largest interlayer spacing of the RuO2 structure is 3.17 Å and is comparable to the interlayer spacing d (002) of the graphitic carbon support but they cannot be confused since the carbon lattice is quite irregular like in an amorphous material and is continuous. Amorphous Ru-oxide particles of small size also can not be easily distinguished. In such case XAS and XPS analyses may indicate the presence of amorphous Ru-oxide; it must be noted that XPS gives information only on the outermost layers [10]. It was reported that formation of labile oxygen groups on Ru sites, not directly alloyed with Pt but close enough to Pt sites, can allow chemical interaction between these groups and the adsorbed methanolic residues to produce CO2 [181]. The promoting effect of the RuOx species for methanol oxidation reaction has been extensively investigated by several authors [181-183]; very high performances were obtained in a DMFC with unsupported Pt-RuOx anode electrocatalyst [9]. It was suggested that facile oxygen transfer from Ru to Pt rich regions where adsorption of CO-like residues preferentially occurs, could enhance the catalytic oxidation of methanol [183]. But, oxygenated species, strongly held on Ru surfaces, cannot easily migrate on the surface. In

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

86

A. S. Aricò, V. Baglio and V. Antonucci

general XPS analysis has revealed the presence of Ru(IV) and Ru(VI) species. The oxygen transfer mechanism requires an open oxy-hydroxide structure with a significant number of free Ru-OH bonds [10, 181]. Thus, such mechanism operates better on Ru(IV) that on Ru(VI) species since the formation of labile oxygen bonds occurs preferentially on the first sites. It is thought that both mechanisms involving Pt-Ru alloys and RuOx species could actively participate in the methanol oxidation reaction. Yet, the absence of an atomic mixing of Pt with RuOx species probably imposes an activation barrier for the migration of adsorbed intermediates. In order to assess which mechanism is more effective, in a study in our group we have compared the methanol oxidation behavior on a pure Pt-Ru alloy and a Pt-RuOx electrocatalyst, with comparable particle sizes, at high temperatures (130 °C) [182, 184]. It was observed that both electrocatalysts are active for the methanol oxidation reaction but the electrocatalyst containing a lower amount of oxidized Ru species and characterized by a smaller Ru-O bond strength performed better [182, 184]. The difference in electrochemical results was interpreted on the basis of a need for labile oxygen species on the electrocatalyst surface and not strongly bonded oxygen atoms. In fact, the first oxygen species can be more easily transferred to methanolic residues adsorbed on Pt. The electronic properties of the active phase of the catalyst are also influenced by the methanol-support interaction. Such aspects which are of relevant interest to enhance the anode reaction kinetics have been in-depth investigated for carbon supported Pt-Ru catalysts with nominal 1:1, Pt:Ru, atomic ratio. Various catalysts characterized by different concentrations of metal phase on carbon have been investigated (Table 4) [185]. XRD analysis (Figure 20) of these carbon supported catalysts indicated for all samples the cubic fcc crystallographic structure of the Pt_Ru alloy. A shift towards higher Bragg angles was observed in the XRD patterns of the 85% Pt_Ru/C catalyst with respect to the other carbon-supported catalysts; this indicated a higher degree of alloying in the sample. The particle size, determined by the Debye_Sherrer equation, was smaller in the highly dispersed 30% Pt-Ru/C catalyst. TEM observation of the catalysts (see Figure 21) clearly indicated a significantly higher degree of agglomeration for the 85% Pt_Ru/C catalyst and a lower metal-support interaction in comparison with the other samples. The 30% Pt-Ru/C showed primary Pt-Ru particles well anchored to the carbon black surface. XPS analysis of Pt 4f, Ru 3d and Ru 3p3/2 spectra (Figures 22, 23) showed a shift to higher binding energies with a larger fraction of oxidized species for the catalysts with lower concentrations of metallic phase on carbon, i.e. 30 and 60% Pt_Ru/C as effect of metal support interaction (Table 5).

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Fundamental Aspects: Status of Knowledge

87

Table 4. Physico-chemical characteristics of the carbon supported Pt-Ru catalysts characterised by different concentration of the active phase on the support. Reprinted from Ref. [185] with permission from Elsevier Catalyst

BulkComposition (XRF) (±2%) 30% Pt-Ru/C Pt48Ru52 60% Pt-Ru/C Pt50Ru50 85% Pt-Ru/C Pt51Ru49

Alloy Composition (XRD) (±2%) Pt69Ru31 Pt59Ru41 Pt47Ru53

Surface Composition (XPS) (±5%) Pt52Ru48 Pt51Ru49 Pt55Ru45

Particle size Å(±2Å) 18 22 21

Table 5. Binding energy and relative intensities of different species from curve-fitted XPS spectra in the various Pt-Ru/C catalysts. Reprinted from Ref. [185] with permission from Elsevier Species 85% Pt-Ru/C Pt 4f

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Ru 3p

60% Pt-Ru/C Pt 4f

Ru 3p

30% Pt-Ru/C Pt 4f

Ru 3p

Binding Energy (eV) 4f7/2 71.70 73.10 74.80 3p3/2 461.85 463.65 467.85

4f5/2 75.05 76.45 78.15

4f7/2 71.70 72.90 74.80 3p3/2 461.90 463.40 467.10

4f5/2 75.05 76.25 78.15

4f7/2 71.70 72.90 74.80 3p3/2 461.95 464.39 467.62

4f5/2 75.05 76.25 78.15

Relative peak area (%)

90.09 5.49 4.42 51.57 37.60 10.83

81.18 10.73 8.09 29.20 47.27 23.53

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

81.18 8.09 10.73 21.08 53.96 24.96

88

A. S. Aricò, V. Baglio and V. Antonucci (220)

I

85%Pt-Ru/C

I

Counts /a.u.

60%Pt-Ru/C

30%Pt-Ru/C I

85%Pt-Ru/C

62

66

70

60%Pt-Ru/C

30%Pt-Ru/C 60%Pt/C 15

25

35

45

55

65

75

85

95

Two theta degrees

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Figure 20. XRD patterns of carbon-supported Pt-Ru catalysts. Reprinted from Ref. [185] with permission from Elsevier.

Figure 21. Transmission electron micrographs of carbon-supported Pt-Ru catalysts: (a) 85% Pt-Ru/C catalyst; (b) 60% Pt-Ru/C catalyst; (c) 30% Pt-Ru/C catalyst. Reprinted from Ref. [185] with permission from Elsevier.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Fundamental Aspects: Status of Knowledge

85% Pt-Ru/C

89

Pt°

Pt2+ 60% Pt-Ru/C

Pt°

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Pt2+ 30% Pt-Ru/C

Pt°

Pt2+ Pt4+ 85

80

75

70

65

Binding Energy (eV) Figure 22. X-ray photoelectron spectra of Pt-Ru catalysts (Pt 4f doublet). Reprinted from Ref. [185] with permission from Elsevier.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

90

A. S. Aricò, V. Baglio and V. Antonucci

85% Pt-Ru/C

Ru°

60% Pt-Ru/C

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

RuO2

30% Pt-Ru/C

RuOxHy (RuO3)

475

470

465

460

455

Binding Energy (eV) Figure 23. X-ray photoelectron spectra of Pt-Ru catalysts (Ru 3p3/2 line). Reprinted from Ref. [185] with permission from Elsevier.

A comparison of the in situ stripping behavior of adsorbed methanolic residues for the three Pt-Ru/C catalysts at various temperatures is shown in Figure 24 [185].

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Fundamental Aspects: Status of Knowledge

91

90°C 100°C

80°C

0.1 A cm-2

110°C 120°C

Current density / A cm-2

130°C

85% Pt-Ru/C 90 °C 100 °C

80 °C

0.05 A cm-2

110 °C 120 °C 130 °C

60% Pt-Ru/C 100°C

90°C

80°C

110°C 120°C

0.1 A cm-2

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

130°C

30% Pt-Ru/C 0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Potential / V vs. RHE Figure 24. In situ stripping voltammetry of methanol residues at the various carbonsupported Pt-Ru catalyst / Nafion 117 membrane interfaces, at various temperatures under the DMFC configuration. Anode: 1 M methanol, 1 atm rel. adsorbed for 30 min; cathode: H2 feed 1 atm rel. Reprinted from Ref. [185] with permission from Elsevier.

As the temperature increased above 90 °C, the stripping area of the methanolic residues decreased progressively for all catalysts, whereas the peak shifted towards lower potentials on account of the decrease of the activation energy for CO removal. By comparing the behavior of the various catalysts, it was observed that the 30% Pt_Ru/C sample was characterized by the largest stripping

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

92

A. S. Aricò, V. Baglio and V. Antonucci

area per unit of weight. Such an evidence indicated that methanolic residues in this sample covered a largest amount of catalytic sites. Yet, the stripping peak potential at each temperature was shifted towards negative values for the 85% PtRu/C catalyst. Furthermore, the stripping peaks were significantly sharper for this catalyst indicating higher intrinsic catalytic activity and faster charge transfer kinetics in the sample characterized by smaller carbon support interaction. The stripping voltammetry provided results similar to the X-ray photoelectron spectroscopy. Both techniques deal with the electronic levels on the catalyst surface. Methanol in the voltammetry simply acts as a probing molecule. It is surprising to observe that the stripping voltammetry amplifies the small differences which are observed by XPS. In practice, the cyclic voltammetry (CV) provides information with a resolution that is almost one order of magnitude better than what observed by XPS. In other words, the observed differences in the binding energies among the various catalysts are typically of the order of 0.5 eV i.e. comparable to the resolution of this technique. The stripping peaks differ for the various catalysts by a few tens of mV in the stripping voltammetry. Usually, the resolution is of the order of a few mV in the CV experiments. The CV analysis can also take advantage of the strong effect determined by the temperature change as an additional tool to discriminate among different catalytic properties. By comparing the behavior of the three different catalysts either in terms of single cell and half-cell polarization curves at 90 °C (Figure 25) [185], better performances were achieved for the catalyst showing both lower stripping peak potentials (as expected) but also lower coverage of methanolic residues. Thus, especially at 90 °C, the higher intrinsic catalytic activity (lower activation barrier) appears to be more relevant than the catalyst dispersion. The physicochemical properties of the catalysts have less influence on the anode electrochemical behavior at high temperatures (e.g., 120 °C) since CO poisoning is alleviated under such conditions. Another interesting argument concerning with the amelioration of the PtRu catalytic system is involving the development of decorated catalysts [80]. It is widely accepted that a large decrease of Pt loading in direct methanol fuel cell (DMFC) electrodes would significantly reduce the cost of these electrochemical devices [2]. Fuel cells have to fulfill this pre-requisite to achieve market competitiveness with respect to the present electrical power generators. Ultra-low Pt loading electrodes have been successfully developed in the last decade for H2–air polymer electrolyte fuel cells (PEMFCs) [56]; yet, the Pt loading needs to be increased when the anode is fed with reformate gas. An interesting method for preparing fuel cell electrocatalysts based on the

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Fundamental Aspects: Status of Knowledge

93

spontaneous deposition of Ru on Pt was first proposed by Wieckowski and coworkers [161].

Cell voltage / V

0.8

T = 90°C

85% Pt-Ru 60% Pt-Ru 30% Pt-Ru

0.6 0.4 0.2

a

0 0

0.2

0.4

0.6

0.8

1

Anode potential / V vs. RHE

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Current density / A cm-2 0.8

T = 90°C

0.6 0.4

85% Pt-Ru 60% Pt-Ru 30% Pt-Ru

0.2

b 0 0

0.2

0.4

0.6

0.8

1

Current density / A cm-2

Figure 25. DMFC single cell (a) and anodic half-cell polarization behaviour (b) at 90 °C for various Pt-Ru/C catalysts. Anode: 1 M methanol, 1 atm rel.; cathode: air feed 2.5 atm (a), H2 feed 1 atm rel. (b). Reprinted from Ref. [185] with permission from Elsevier.

Subsequently, Adzic and co-workers [186], using a similar approach, were able to reduce the Pt loading in reformate- fed PEMFCs anodes based on Ru catalysts coated with Pt submonolayers on the surface. Pt-decorated Ru catalysts have been preliminarily investigated as anodes in DMFCs [80]. These catalysts have allowed the Pt loading in the electrodes to be reduced, maintaining acceptable performances. The decoration process has been frequently applied in recent years to fuel cell electro-catalysis including both low temperature polymer electrolyte [80, 186] and intermediate temperature solid oxide fuel cell [187]

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

94

A. S. Aricò, V. Baglio and V. Antonucci

catalysts; it allows to increase the concentration of noble metal nanoparticles at the boundaries of the three-phase reaction zone in the catalytic layer. The approach to deposit small amounts of Pt-nanoparticles on the surface of Ru agglomerates was investigated to prepare practical anode catalysts for DMFCs [80]. Low temperature fuel cell processes, including methanol electro-oxidation, are governed by the surface properties of the electrodes; thus, catalysts containing small amounts of Pt-nanoparticles on the surface of a less expensive metal, which participates to the reaction, may represent a useful approach to reduce the incidence of the catalyst cost in DMFC devices. Figure 26 shows a comparison of the DMFC polarization curves in the potential range of technical interest for the decorated (0.1 mg Pt cm-2), carbon supported Pt–Ru alloy (2 mg Pt cm-2) and bare unsupported Ru catalysts [80, 188].

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Cell potential / V

0.80

0.25 -2

2 mg Pt cm-2 0.2

0.1 mg Pt cm-2

0.60

0.15

0.40

0.1

5 mg Ru cm-2

0.20

0.05

0.00

Power density / W cm

Pt decorated-Ru 60% Pt-Ru/C Ru unsupported

1.00

0 0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1

Current density / A cm-2

100 nm

Decorated catalyst

Figure 26. (a) DMFC single cell polarizations at 130 °C for commercial Pt–Ru/C, Ptdecorated and bare unsupported Ru catalysts. Anode: 1 M methanol, 2 atm rel.; cathode: air feed 2.5 atm., (b) transmission electron micrograph of Pt-decorated Ru catalyst. Reprinted from Ref. [188] with permission from Elsevier.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Fundamental Aspects: Status of Knowledge

95

The carbon supported Pt–Ru alloy based electrode shows lower potential losses in the activation controlled region with respect to the decorated catalyst. However, the Pt loading in the decorated catalyst based anode is 20 times lower. The mass activity is considerably better for the decorated catalyst if the current is normalised by the Pt content. Yet, in the present case, Ru not only acts as a support for the Pt nanoparticles but it also participates in the electrocatalytic oxidation of methanol. The role of ruthenium is to promote the water discharging process, but, it can also participate to the dissociative adsorption of methanol at high temperatures [188]. Figure 27 shows a comparison of the CO and methanol stripping behavior for the Pt decorated Ru catalyst. The stripping profiles are very similar suggesting that CO removal is the rate determining step in the methanol oxidation process

Current density / A cm

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

-2

0.4

CO

0.3

CH3OH

0.2 0.1 0 -0.1 -0.2 0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

Potential / V vs. RHE Figure 27. Comparison of adsorbed CO and methanolic residues stripping profiles for Ptdecorated unsupported Ru catalyst, at 80 °C under a DMFC configuration. Anode: 1 M methanol or 2% CO in argon, 1 atm. rel. adsorbed for 30 min at 0.1 V; cathode: H2 feed 1 atm. rel. Reprinted from Ref. [188] with permission from Elsevier.

The following aspects have been ascertained for the decorated catalysts. A proper location of the Pt particles in the three-phase reaction zone of the catalytic layer appears to enhance the Pt utilization. The promoting effect of the Ptdecorated catalyst can be represented as a bimetallic system composed of single Pt and Ru nanoparticles; (i) it shows lower intrinsic catalytic activity than the Pt–Ru alloy; (ii) the promoting effect of the decorating Pt particles with respect to bare

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

96

A. S. Aricò, V. Baglio and V. Antonucci

Ru relies mainly on an enhancement of methanol dehydrogentation rate; (iii) CH3OH dedydrogenation and water discharging processes in the decorated catalyst most likely occur on different nanoparticle sites; thus, there is no significant competition for the adsorption of electroactive species; (iv) the decorated catalyst approach allows a better Pt utilization and significant decrease of Pt loading. Unfortunately, at the present, Ru is not significantly cheaper than Pt. Thus, this approach may be of interest only if the cost of Ru would decrease significantly. But, this does not seem to be the case; thus, alternative routes should be considered for the decorated catalysts by selecting a support which effectively participates to the mechanism but with a cost significantly lower than the elements used for decoration. It has been stated that the Pt-Ru is the most promising catalyst for methanol oxidation in acidic electrolytes. However, the Pt-Ru formulation suffers the lack of other elements capable to oxidize methanolic residues by an effective redox mechanism. In this light, an interesting approach is the formulation of new electrocatalysts that, although based on Pt-Ru alloy (this has demonstrated irreplaceable characteristics in matching the above requirements), would take advantage of proper surface promoters able to facilitate labile adsorption of methanol residues and their oxidation through redox functionalities. In recent years, electrocatalyst formulations which includes tungsten and molybdenum have been proposed [6, 35, 132, 170]. In an acid electrolyte environment, these elements are generally stable in one or more oxidized forms such as WO2/WO3 and can easily exchange oxidation states by adsorbing hydroxyl ions from water and donate these species to methanol residues adsorbed on Pt. Tseung et al. [170] investigated the Pt-Ru-W system. They observed superior electrocatalytic activity, if tungsten was present in a reduced form, for CO oxidation. Similar evidences were found for the ternary system containing molybdenum [35]. But in this latter system, the electrochemical stability of molybdenum, under prolonged operation conditions, has still not been demonstrated. Tungsten has the ability of coordinating water and a high mobility of hydroxyl species in the WO2/WO3 system in the region near the reversible potential for the oxidation of methanol, i.e. 0.04 V vs. NHE. The formation of hydrogen tungsten bronze, HxWO3, during the dehydrogenation process of methanol was observed for Pt/WO3 electrocatalysts [189]. The presence of a ―spillover‖ effect had a significant influence on the activity of the catalysts. Tungsten oxide-containing electrocatalysts produced an enhancement of the reaction rate by the continuous formation and oxidation of HxWO3 during the dehydrogenation of methanol. It was assumed that, water adsorbed on WO3 at low

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

97

overpotentials and interacted with CO adsorbed on Pt, thus facilitating its oxidation to CO2. Only a few electrocatalyst formulations, alternative to Pt, have been proposed for methanol electro-oxidation in acidic environment. In most cases, the alternatives contain noble metals like gold, which does not solve the problem of the high cost of the materials employed in DMFCs. Only a few formulations without noble metals have been tested: these are mainly based on transition metal alloys like NiZr [130], transition metal oxides and tungsten-based compounds [128, 129, 190]. All these materials showed lower reaction rates than Pt-based electrocatalyst and thus such unsatisfactory preliminary results have not stimulated much work in these directions. Thus, although many studies have been directed to the investigation and optimization of noble metals electrocatalysts, little work has been devoted to explore the possibility of using transition metal oxides as anodes. Transition metal oxides, because of their partially filled orbitals, may be ideal for a labile interaction with carbon monoxide and also be able to chemisorb hydroxyl species on neighbor sites. As transition metal oxide compounds intrinsically possess the thermodynamic characteristics which predict water discharging at low overpotentials, one could expect that their wide range of potential for electroreduction might meet the demand for the essential reaction through a surfaceoxygen vacancy mechanism. As the most favorable condition for water discharging occurs in the region near to the reversible potential for methanol oxidation, transition metals like Mo and W appear ideally suitable for accelerating the reaction rate. The doping of tungsten oxide with rare earth elements or Na-alkali was found to increase the oxidation rate of carbon monoxide (pure or in reformed gas) in a fuel cell [191]. This could open favorable perspectives, depending upon the degree of reduction of the bronzes, and whether water chemisorption can be tuned and the OH mobility adjusted through regulation of density of oxygen vacancies. Beside tungsten oxides, tungsten carbide is also known to possess intrinsic catalytic activity for electro-oxidation reactions such as hydrogen oxidation [129]. WO2 was demonstrated to be very active for CO oxidation in a fuel cell as early as 1969 [128], and, more recently, it has been shown that tungsten oxides and bronzes also have electrocatalytic activity for this reaction as well as methanol oxidation [191]. Furthermore, the concentration of oxygen vacancies on the electrocatalyst surface of tungsten bronzes can be appropriately adjusted by the addition of sodium or rare earth elements, so as to attain the desired range of potential for oxidative removal of the adsorbed CO by the active species generated on the surface of the electrocatalyst [191].

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

98

A. S. Aricò, V. Baglio and V. Antonucci

As discussed above, one of the main requirements for an optimal alloy electrocatalyst, such as Pt-Ru, is its high dispersion. The mass activity (A/g Pt) of the catalyst for methanol electro-oxidation is strictly related to the degree of dispersion, since the reaction rate is generally proportional to its active surface area. For an oxygen reduction electrocatalyst, it was found that there is an optimum particle size for the metallic phase (about 3 nm), which corresponds to a significantly high mass activity [192]. For this reaction it was found that the specific activity (mA cm-2 real surface area) increases with the particle size, but simultaneously decreases with the active surface area. The best compromise was achieved at about 3 nm Pt particle size [192]. In the case of methanol electro-oxidation on carbon-supported Pt electrocatalysts, two different trends were observed. Mc Nicol et al. [8] observed, for their Pt electrocatalysts, a maximum activity at about 80 m2/g surface area. Recently, another group has shown that the specific activity increases as a function of particle size [193]. Thus, a maximum in mass activity vs. particle size should be observed as in the case of oxygen reduction. On the other hand, Watanabe et al. [194] found that the specific activity for methanol oxidation on a carbon supported Pt electrocatalyst does not change for a particle size above 2 nm (Pt fcc structure); thus, the mass activity increases as the dispersion of the metal phase is increased [194]. These latter findings have been in part confirmed for the Pt-Ru system for a particle size above 3 nm [68]. Not much work has been carried out on electrocatalysts with particle sizes below 2 nm or amorphous catalysts. Generally, most of the Pt-Ru fuel cell electrocatalysts have particle size above 2 nm, which are crystalline with a face centered cubic (fcc) structure. For particles with 1-1.5 nm size about 50% of the atoms are on the surface and there is no ordered structure. In a recent study, Pt-Ru catalysts characterized by a particle size smaller than 2 nm were investigated for methanol oxidation. The catalytic performance of the catalysts with mean size for the primary particles of about 11.5 nm was quite poor compared to the conventional catalysts; it was also observed, that in that range of particle size the structure is mainly amorphous [195]. Some kinetic investigations by Wieckowski et al. [161] on Pt surfaces covered with Ru ad-atoms have shown that the surface composed by the (111) Pt crystallographic plane covered by Ru ad-atoms performs better than any other Pt/Ru ad-atoms crystallographic plane. Thus, extending these considerations to high surface area Pt-Ru electrocatalysts, the specific activity should increase with the particle size. In fact, a large-size particle possess a high surface content of (111) planes [192]. Accordingly, a maximum in electrocatalytic activity should be observed as a function of particle size; the data shown by Ren et al. [68] may be

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

99

fitted as well with a ―volcano‖ relationship with maximum at about 2.5-3 nm and shape similar to that generally observed for oxygen reduction. Unfortunately, the scattering due to the variable composition of the Pt-Ru surface, in high surface area electrocatalysts, does not allow to single out a clear behavior. A second aspect concerns the loading of the metal phase on the carbon support. A high Pt wt. percent on the carbon substrate will significantly decrease the anode thickness for the same Pt loading per geometric electrode area (e.g., 1 mg cm-2). Thus, it is possible to enhance mass transport through the electrode and, at the same time, reduce the ohmic drop. However, it has been found that an increase in Pt loading (above 40 % wt.) on the carbon support decreases the dispersion of the electrocatalyst, due to some particle agglomeration (see Figure 21). In general preparation procedures, such as impregnation, colloidal deposition and surface reaction involve the adsorption step of active compounds on a carbon black surface. The synthesis of a highly dispersed electrocatalyst phase in conjunction with a high metal loading on carbon support is one of the goals of the recent activity in the field of DMFCs. In this regard it is of interest to determine which carbon black is most suitable as support. In recent reports [40], the mostly used carbon blacks were: Acetylene Black (BET Area: 50 m2/g), Vulcan XC-72 (BET Area: 250 m2/g) and KETJEN Black (BET Area: 900 m2/g). All these materials have optimal electronic conductivity but, as denoted above, they differ for the BET surface area and thus very probably in morphology. A low surface area carbon black (such as Acetylene Black) will not allow high dispersion of the metal phase, especially for a high metal loading (low carbon content); on the other hand, it does not have micropores in the structure, which could hinder mass transport through the electrocatalytic layer. A high surface area carbon black can easily accommodate a high amount of metal phase, with a high degree of dispersion but, at the same time, the significant amount of micropores on the carbon support will not allow a homogeneous distribution of the electrocatalytic phase through the support, which could lead to mass transport limitations of the reactant as well as its limited access to the inner electrocatalytic sites. Vulcan XC appears to be a suitable compromise with the presence of a small amount of micropores and a reasonable high surface area sufficient to accommodate a high loading of the metal phase. Up to now, this carbon black is the most used carbon support for the preparation of DMFCs catalysts. However, in the recent years, nanostructured mesoporous carbon supports as well as carbon nanotubes have been considered a suitable alternative to conventional carbon blacks.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

100

A. S. Aricò, V. Baglio and V. Antonucci

For what concerns the development of anode catalysts for alkaline direct methanol fuel cells, it is pointed out that the methanol oxidation rate is accelerated at high pH values. Thus, from a kinetic point view, it is advantageous to carry out the methanol oxidation in alkaline electrolytes [60]. Furthermore, being the corrosion constraints less significant in alkaline media, in principle, a wider number of catalyst formulations can be investigated for methanol oxidation as compared to proton conducting electrolytes. These formulations can include some transition metals such as Ni [60], usually discarded by the fact that they are affected by significant corrosion in an acidic medium. Despite these promising aspects, the number of investigations concerning with the development of anode catalysts for alkaline DMFCs is less numerous than in acidic electrolytes. Essentially, two strategies have been used. Due to the enhanced reaction rate at high pHs, alkaline DMFCs can employ non-precious transition metals, e.g. Ni, which are characterised by lower intrinsic activity. Although, the reaction rates are usually faster for PtRu than Ni, the increase of reaction kinetics due to the increased pH compensates in part for this gap in intrinsic activity [60]. The Nibased catalysts can suitably operate in combination with a liquid electrolyte containing a concentrated base such as 5 M KOH or NaOH, characterised by a high pH. Thus, low cost non precious metal anodes can be effectively used in alkaline DMFCs. However, for practical purposes, anion exchange membranes have been recently preferred to the liquid electrolyte [77]. Under such operating conditions and due to carbonation occurring during steady-state operation, the electrolyte in the anode compartment turns progressively into a carbonate/bicarbonate mixture with corresponding lower pH than the KOH solution. There are still some kinetic advantages over Nafion-type membranes, but, not as significant as in the case of KOH. Furthermore, the conductivity decreases. Pt-electrocatalysts have been mainly considered for operation in conjunction with anion exchange membranes. These include the conventional Pt/C catalyst, platinised Ti electrodes [78] and Pt-Ru alloys [60]. Platinised mesh anodes have shown higher catalytic activities than conventional Pt/C electrodes. The main advantage was the direct contact of the reactant to the mesh and membrane [78]. Due to the open area of the mesh, the liquid can easily reach the interface reducing mass transport resistance [78]. The typical activation losses recorded for methanol oxidation at PtRu in acidic systems are less accentuated in alkaline media. The methanol oxidation rate at PtRu electrodes in a carbonate/bicarbonate mixture is about 8 and 2.5 times larger than in sulphuric acid, at 0.35 and 0.45 V RHE, respectively [60]. However, it should be considered that the reaction kinetics in the presence of a Nafion-type ionomer, are faster than

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Fundamental Aspects: Status of Knowledge

101

in sulphuric acid due to the absence of the well known drawback caused by sulphate anion adsorption on the catalyst surface.

5.2. DMFC MEMBRANES AND METHANOL CROSS-OVER ISSUES

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

5.2.1. Proton Conducting Membranes In the section 3.2, we have discussed historical aspects related to electrolyte development for DMFCs. The conclusion was that, until now, no univocal identification of the most appropriate electrolyte system for DMFCs has been made. Thus, the argument is open to various considerations which regard conductivity, cross-over, stability, cost etc. Several excellent reviews have been published on this topic [63, 196, 197]. In the present section, we have restricted our discussion to the characteristics of membranes that are actually considered for commercial DMFC systems. Perfluorosulfonic polymer electrolyte membranes (for example NafionTM) are currently used in H2/air and methanol/air fuel cells because of their excellent conductivity and electrochemical stability [56]. Unfortunately, they suffer from several drawbacks such as methanol cross-over and membrane dehydration. The latter severely hinders the fuel cell operation above 100 °C, which is a prerequisite for the oxidation at high rate of small organic molecules involving formation of strongly adsorbed reaction intermediates such as CO-like species [2]. Alternative membranes based on poly(arylene ether sulfone) [198], sulfonated poly(ether ketone) [199] or block co-polymer ion-channel-forming materials as well as acid-doped polyacrylamid and polybenzoimidazole have been suggested [197-200]. Various relationships between membrane nanostructure and transport characteristics, including conductivity, diffusion, permeation and electro-osmotic drag, have been observed [201]. Interestingly, the presence of less connected hydrophilic channels and larger separation of sulfonic groups in sulfonated poly(ether ketone) reduces water/methanol permeation and electro-osmotic drag with respect to Nafion while maintaining high protonic conductivity [201]. Furthermore, an improvement in thermal and mechanical stability has been shown in nano-separated acid–base polymer blends obtained by combining polymeric Nbases and polymeric sulfonic acids [199]. Considerable efforts have been addressed in the last decade to the development of composite membranes. These include ionomeric membranes modified by dispersing inside their polymeric

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

102

A. S. Aricò, V. Baglio and V. Antonucci

matrix insoluble acids, oxides, zirconium phosphate and so on; other examples are ionomers or inorganic solid acids with high proton conductivity, embedded in porous non proton- conducting polymers [200]. These systems are discussed more in detail in the section dealing with high temperature operation. However, some characteristics are anticipated in this section due to their application in a wide temperature range. In an attempt to reduce the drawbacks of perfluorosulfonic membranes, nanoceramic fillers have been included in the polymer electrolyte network. Stonehart, Watanabe and co-workers [202] have successfully reduced the humidification constraints in PEMFCs by the inclusion of small amounts of SiO2 and Pt/TiO2 (~7 nm) nanoparticles to retain the electrochemically produced water inside the membrane. This approach was used in DMFCs to increase the operating temperature (up to 145 °C) and reduce methanol cross-over by effect of an increase in the tortuosity factor for methanol permeation [70]. Although it has been hypothesized [197, 200] that the inorganic filler induces structural changes in the polymer matrix, the water retention mechanism and the protonic conductivity appears more likely to be favored by the presence of acidic functional groups on the surface of nanoparticle fillers [71]. At present, there are no strong indications that the transport properties are significantly affected by the filler [203]. However, the greater water retention capability of the composite membrane at high temperatures (130 °C–150 °C) and under low humidity [71] should promote the ‗vehicular mechanism‘ of proton conduction as occurs at lower temperatures [204]. As above mentioned, non-perfluorinated ionomer membranes based on thermostable sulfonated aromatic polymers such as polybenzimidazole, polysulfones, polyethersulfones, polyetherketones have been proposed as alternative materials at the traditional perfluorosulfonic membranes [197]. Among them, polysulfones are promising candidates for their low cost, commercial availability and processability and for this, they are under investigation in DMFC technology [74]. In fact, one of the main drawbacks which has limited the market penetration of these systems is the cost of the polymer electrolyte membrane. Commercially available fuel cells stacks are based on Nafion (DuPont) or similar high cost perfluorosulfonic membranes which contribute to a significant extent to the cost of the entire device. Besides, application of Nafion membranes in direct methanol fuel cells is affected by significant cross-over levels which reduce the fuel efficiency of these devices. Thus, DMFCs lose part of their advantages with respect to PEMFCs and internal combustion systems. Some technical solutions have been proposed to address this latter drawback. As above said, most of them involve modification of perfluorosulfonic membranes by addition of ceramic oxides, but also deposition of Pt nanoparticles inside inner polymer channels, coating the surface with barrier

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

103

polymer or deposition of sulfonated polyvinyl alcohol [2, 63, 196, 197, 205, 206] etc. All these efforts have shown that it may be possible to reduce methanol crossover but it can not be completely eliminated while maintaining high conductivity levels. Such evidences may be explained by the fact that an efficient proton transport in the temperature range from ambient to 130 ◦C requires suitable water contents and water transport inside the polymer electrolyte [2, 63, 196, 197, 205, 206]. As methanol is highly soluble in water, it appears that its mass transport through the membrane may be eliminated only at expenses of the electrolyte conductivity. The perfluorinated polymers such as polysulfones, polyetherketones, polyimmides usually combine reduced cross-over and appropriate conductivity levels. However, it is pointed out that the conduction mechanism in these systems is not much different than in Nafion. Thus, methanol cross-over can not be eliminated completely. It appears that the main advantage of these polymers mainly concerns with the cost reduction with respect to Nafion. Alternatively, membranes such as phosphoric acid impregnated-polybenzoimidazole (PBI) which do not need water transport to maintain high proton conductivity may represent a valid approach [69, 205]. However, these electrolytes still present methanol cross-over effects; moreover, suitable DMFC life-time for such membranes has not been yet demonstrated. Another issue is concerning with cold start-up and low temperature operation. In principle, the water uptake properties of sulfonic acid-based membranes may be modulated by selecting a proper concentration and distribution of sulfonic groups inside the polymer. Such an objective is generally pursued in the preparation of grafted polymer membranes [206]. The application of the radiochemical grafting technique to the production of DMFC membranes has been explored in the framework of the NEMECEL European program (see section 4.2). In this procedure the material properties may be properly tailored by varying a few parameters in the synthesis while maintaining the process characteristics and plants for large scale production. The main efforts are addressed to reduce the cost of production through a flexible preparation process and the proper selection of cheap base materials. For this purpose ETFE (ethylene–tetrafluoroethylene, C2H2 and C2F2 groups with 1/1 ratio and nearly perfect alternance) films have been selected as substrate material; the films were radio-chemically grafted with styrene and subsequently sulfonated in order to obtain sulfonic acid anchored groups. The present cost of the base irradiated ETFE material favorably compares with the average industrial cost of the commercial perfluorinated sulfonic membranes. In order to improve the mechanical strength properties of the polymer, increase the thermal resistance and reduce the cross-over of gases or

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

104

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

liquids (such as methanol) through it, while maintaining suitable conductivity, appropriate cross-linking was made during the grafting step by adding a crosslinking agent [206]. As above discussed, Nafion-type membranes are often used as electrolytes in DMFCs; yet, since methanol is rapidly transported across perfluorinated membranes and is chemically oxidized to CO2 and H2O at the cathode, there is a significant decrease in coulombic efficiency for methanol consumption by as much as 20% under practical operation conditions [207]. Thus, it is very important to modify these membranes by, as example, developing composites or finding alternative proton conductors with the capability of inhibiting methanol transport. It is generally accepted that a solidstate proton conductor is preferable for liquid fuel fed DMFCs because it hinders corrosion and rejects carbon dioxide (produced during the methanol oxidation). However, there are some pre-requisites that should be properly considered. The polymer electrolyte should have a high ionic conductivity (5 ·10-2 ohm-1 cm-1) under working conditions and low permeability to methanol (less than 10-6 moles min-1 cm-2). Furthermore, it must be chemically and electrochemically stable under operating conditions. These requirements appear to be potentially met by new classes of solid polymer electrolytes that show promising properties. As mentioned above, the alternative proton conducting membranes include sulfonated poly (ether ether ketone) and poly (ether sulfone) [208], polyvinylidene fluoride [65], styrene grafted and sulfonated membranes [209], zeolites gel films (tin mordenite) and/or membranes doped with heteropolyanions [2]. Sulfonated poly (ether ether ketone) and poly(ether sulfone) electrolytes show promising characteristics in terms of (i) mechanical strength even when these are prepared in thin films (10-50 m) and (ii) acceptable conductivity in their protonic form with low resistance to ion transport and reduced cross-over of methanol. The perspectives of preparing some of these materials as very thin films by casting them on electrodes from solution containing the polymer precursors offer considerable advantages in terms of reduction of ohmic overpotentials in DMFCs, as compared with perfluorinated sulfonic membranes with larger thickness (e. g., Nafion 117). The stability and conductivity properties of sulfonated poly ether ether ketones and poly ether sulfones, as ascertained in DMFCs, are at the present at an insufficient stage with respect to the requirements. The evidence, to date, indicate some instability of the materials in (hot) methanol, which is further increased by sulfonation of the materials [209]. One approach to solve the above mentioned problems is to cast the membranes and then cross-link the polymer. Suitable compounds for carrying out the crosslinking are aliphatic and cyclic diamines, such as diazobicyclooctane and

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

105

aminopyridine (the chain length of the diamines can be varied and optimised so as to minimize the methanol flow through the membrane). An appropriate compromise between sulfonation and cross-linking procedures is generally required to obtain membranes with high ionic conductivity, good mechanical resistivity and low methanol cross-over [209]. Alternatives to these membranes and Nafion are acid-doped polyacrylamid and polybenzoimidazole [69]. The main question about these membranes is the extent of leaching of acids of small molecular weight (H3PO4) entrapped in the polymer, during operation of a fuel cell fed with a hot methanol/water mixture as the anode reactant. In fact, these polymers usually swell at high temperatures in the presence of water and methanol. Probably these problems may be better addressed by using a high molecular weight superacid (such as phosphotungstic acid) that may be physically entrapped in the polymer structure. However, in this case, the uptake of water by the polymer should not be significantly reduced since water is essential for the protonic conduction. Concerning the development of composite membranes [65, 70] that was mentioned above, composite recast Nafion-silica membranes have shown excellent properties in terms of mechanical characteristics, water retention at high temperature, resilience to methanol cross-over and ionic conductivity [70]. These electrolytes can suitable operate at both ambient temperature and high temperature (up to 145 °C) with a significant enhancement in methanol oxidation kinetics [70]. However, the main drawback, at the present time, appears to be the high cost of production, primarily determined by the expensive perfluorinated ionomer necessary for for membrane fabrication. A modification of this procedure concerns with the use of heteropolyacid doped silica entrapped into recast Nafion or zirconium phosphate/ extruded Nafion membranes [210]. The membrane properties are usually evaluated by wide-angle X-ray scattering (WAXS) and small-angle X-ray scattering (SAXS) analyses, membrane conductivity and methanol cross-over measurements, mechanical analyses water/methanol uptake and swelling etc. [67]. Conductivity and cross-over properties are related to the self-diffusion coefficient of water and methanol in the membrane [67]. The transport properties of these species are usually investigated by Nuclear Magnetic Resonance (NMR) spectroscopy [67].

5.2.2. Alkaline Membranes Some aspects regarding the development of alkaline electrolytes for DMFCs were discussed in the section 3.2. It was reported that for several decades, the alkaline electrolyte approach was discarded in favour of the carbon dioxide-

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

106

A. S. Aricò, V. Baglio and V. Antonucci

rejecting acidic electrolyte. A renewed interest has been recently aroused from anion exchange membranes and several excellent reviews have been published on this topic [60, 61]. The aim of this sub-section is to present a critical analysis of the advantages and disadvantages of Alkaline Electrolyte Direct Methanol Fuel Cells (AEDMFCs) with respect to proton conducting electrolytes. As well known, the major drawback of proton-conducting electrolyte-based direct methanol fuel cells concerns with the slow reaction kinetics and fuel crossover. A large amount of precious catalysts is necessary and this causes a considerable impact on the cost of these devices. Furthermore, catalyst corrosion and membrane degradation at low pH values limit the number of materials that can be selected for a long-term stability [63]. On the other hand, the liquid alkaline electrolytes that were initially preferred for DMFCs [58] are affected by practical constraints such as potassium or sodium carbonate formation and precipitation in the catalyst pores, need to frequently regenerate the electrolyte and liquid electrolyte leakage through the electrode. Thus, for several decades, a CO2-rejecting solid-state electrolyte was preferred for low temperature fuel cells. Some of the drawbacks associated to the behavior of liquid alkaline electrolytes in fuel cells can be solved by using new anion exchange membranes [60, 61]. In the anionic polymer electrolytes, OH- ions, responsible of ionic conduction, are formed at the cathode by the water fed to humidify the oxidant stream according to the reaction 5 in section 2. The OH- ions are transported through the membrane to the anode where they react with methanol to form CO2 (reaction 4 in section 2). The CO2 reaction product easily reacts with OH- ions to form carbonate/bicarbonate (CO32-/HCO3-) anions [60]. In the membrane region in contact with the anode, the CO32-/HCO3- ions neutralize the positive charge fixed on the polymeric membrane e.g. quaternary ammonium functionalities affecting conductivity and causing a variation of the local pH in the anode compartment with respect to the cathode [60]. The decrease of pH at the anode will cause a positive shift of the redox potential (in the absolute potential scale) for the oxidation process, thus diminishing the electromotive force. In other words, the pH difference will reduce the voltage thermodynamically. The thermodynamic voltage loss can be 290 mV in the presence of a pH difference of about 4 at 80°C [60]. This loss decreases by increasing the operating temperature. High temperature operation restricts the number of anionic polymers that can be used. Furthermore, the high temperature approach could not be appropriate for portable applications. Beside these aspects, there are several advantages in using anion exchange membranes. As discussed above, the main advantage concerns with the favourable reaction rates in alkaline media with respect to acidic electrolytes for both oxygen electro-reduction and methanol electro-oxidation reactions [60, 61].

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

107

Methanol is oxidised to carbon dioxide in both acid and alkaline electrolytes in the presence of proper catalysts e.g. PtRu [60]. The significant reduction of activation overpotential in alkaline media can compensate the voltage loss due to the thermodynamic effects associated to the pH gradient [60]. However, the enhanced reaction kinetics may allow to use cheaper materials and, possibly, non noble metals. In alkaline media, Ni anodes and Ag cathodes represent a suitable compromise in terms of activity and cost [25, 26, 50]. In general, a significant reduction of the catalyst cost may be envisaged. Further advantages of alkaline versus proton conducting electrolytes are mentioned in the following. The OH- ions migrate from the cathode to the anode; this pathway is opposite to the direction of the electro-osmotic drag in proton exchange membrane direct methanol fuel cells. Thus, the electro-osmotic drag does not contribute to the methanol cross-over in alkaline systems. However, in this regard, it should be pointed out that even at high current densities (in the cell voltage region of technical interest) the contribution of the electro-osmotic drag to the methanol cross-over, in protonic electrolyte-based DMFCs, is quite small with respect to the concentration gradient (diffusion). OH- migration in the membrane is assisted by water as it occurs for protons in the analogous acidic polymer electrolyte. Thus, methanol cross-over can not be completely eliminated by the anion exchange membranes; however, in principle, it can be reduced. In general, the alkaline electrolyte causes lower corrosion problems. This allows to investigate a large number of catalyst formulations especially with regard to the methanol tolerance characteristics. The other aspect is concerning membrane stability. One of the main degradation mechanism of proton exchange membrane during fuel cell operation is promoted by the hydrogen peroxide-type radicals formation during oxygen electro-reduction in acidic environment. This process is less effective at high pH values [60, 61]. Thus, as it occurs for the catalysts, there is a wider range of polymers that can be potentially used as anionic electrolytes in DMFCs. Cheap hydrocarbon-only membranes may be selected [61]; yet, these systems are less appropriate than the fluorinated membranes for what concerns the high temperature stability. Besides the above mentioned thermodynamic effect due to the pH gradient, another significant problem associated to the use of anion exchange membranes, is the low ionic conductivity. This is essentially caused by the lower mobility of anions such as OH- or carbonate-type with respect to the protons. The lack of appropriate anionic membranes with conductivity characteristics similar to Nafion has practically hindered the development of alkaline DMFCs for several decades. It should be mentioned that several researchers such as Ogumi et al. [211] and Yu and Scott [78] have reported

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

108

A. S. Aricò, V. Baglio and V. Antonucci

interesting results with anionic membranes but alkaline solutions were indeed used in those works to enhance the membrane conductivity. Recirculation of the liquid electrolyte through the device beside enhancing conductivity, reduces significantly the pH gradient. Thus, the electrolyte recirculation eliminates the thermodynamic constraints and enables an extension of the three-phase reaction zone from the electrode/membrane interface to the bulk of the electrode favoring the presence of a mixed conductivity in the catalytic layer. Such aspect is quite important in the absence of a suitable ionomer solution. In this regard, it is of considerable interest the synthesis of perfluorinated anion exchange membranes that can be dissolved in non volatile solvents enabling the preparation of mixed conductivity (ionic and electronic) catalytic layers as it occurs for conventional PEMFCs and DMFCs. Perfluorinated anion-exchange membranes are also of interest for their perspectives to extend the operating range to higher temperatures [60]. Of course, these are significantly more expensive than hydrocarbon membranes containing C-H bonds in the backbone. One of the reasons why an increasing number of DMFC developers is recently addressing their interest to alkaline membranes is especially due to the development of new polymers that show appropriate conductivities in a range from ambient temperature to 80 °C [77, 78]. Although, these conductivity values are still smaller than conventional perfluorosulfonic membranes, being the alkaline membranes less affected by methanol permeation, they can be used in a thinner form. This compensates in part the effect of the large cell resistance [60, 61]. Furthermore, for some applications such as portable power sources, these devices operate under passive mode at room temperatures. Suchconditions do not allow to reach elevated current densities; thus, the effect of the ohmic drop is less significant. Among the various types of anion-exchange membranes recently proposed, a significant interest has been stimulated by radiation grafted alkaline membranes [77]. Radiation-grafting of styrene into non fluorinated (LDPE), partially fluorinated (PVDF) and fully fluorinated films (FEP) has been widely considered for PEMFCs and DMFCs [63, 77]. Lower methanol cross-over with respect to Nafion and the perspective of high temperature operation has been demonstrated in DMFCs with grafted protonic membranes [206]. The radiation grafting process is also appealing for anionic membranes. This process is generally carried out on commercial preformed films; several parameters can be modulated such as the process temperature, the level of grating, radiation dose, thickness to obtain particular membrane properties [61]. The methanol cross-over can be further diminished by using appropriate cross-linking procedures. Interesting results were obtained by Yu and Scott using Morgane ADP, a commercial anion exchange membrane produced by Solvay (Belgium) [78]. The

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Fundamental Aspects: Status of Knowledge

109

main drawbacks concerned with conductivity and stability. Another interesting commercial membrane is produced by Tokyzama (Japan) [60, 61]. Recently, high conductivity anion exchange membranes were prepared by radiation grafting of vinylbenzylchloride (VBC) into FEP with subsequent ammination with triethylamine and ion exchange with KOH [77]. These membranes were characterised by a degree of grafting of about 20 % and conductivities of 2 10-2 S cm-1 at 50 °C in aqueous solutions. The conductivities are 20 % about the values obtained for state of art perfluorosulfonic acid membranes. The activation energy for OH- migration in the membrane was twice compared with that observed for protons in fully hydrated Nafion indicating that OH- mobility is strongly temperature dependent [78].

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

5.2.3. Membranes for High Temperature Applications A low cost and high temperature membrane, with suitable ionic conductivity and stability up to 150 °C, would be a potential solution to some of the drawbacks presently affecting reformate-fuelled polymer electrolytes (PEMFCs) as well as direct methanol fuel cells (DMFCs) [2, 4]. Fuel cell operation at elevated temperatures can limit the effects of electrode poisoning by adsorbed CO molecules, increase both methanol oxidation and oxygen reduction kinetics and simplify water and thermal management. Furthermore, high temperature operation can reduce the complexity of the reforming reactor employed for PEMFCs [56]; the temperature range from 130 to 150 °C is ideal for application of these systems in electric vehicles and for distributed power generation. Various protonconducting polymer electrolyte materials have been investigated for high temperature operation. Two categories of membranes can be proposed, depending on whether water is required for proton conduction or is not necessary [67]. Polymer electrolytes involving water molecules in the proton mobility mechanism (e.g., perfluorosulfonic membranes) need humidification to maintain suitable conductivity characteristics. The amount of humidification may vary depending on the operating temperature and membrane properties; it influences the size and complexity of the device. Some other electrolytes do not necessarily involve water molecules in the mechanism of proton conduction (e.g., PBI/H3PO4[205], blends of PBI and polysulfone [197], hybrids of polymers and proton-conducting inorganic compounds such as Zr(HPO4)2[200], etc.); these systems do not strictly need humidification. Yet, there are some drawbacks related to the short-term stability of such systems: phosphoric acid leakage from the membrane during operation, poor extension of the three-phase reaction zone inside the electrodes

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

110

A. S. Aricò, V. Baglio and V. Antonucci

due to the absence of a proper ionomer, and reduced conductivity levels for inorganic proton conductors. These problems have decreased the perspectives of utilization of water-free protonic electrolytes in low temperature fuel cells. Alternatively, composite perfluorosulfonic membranes containing different types of inorganic fillers such as hygroscopic oxides [67, 70], surface modified oxides [71, 72], zeolites [212], inorganic proton conductors [210] and so on have shown an increased conductivity with respect to the bare perfluorosulfonic membranes at high temperature and DMFC operation up to about 150 °C has been demonstrated (Figure 28). The mechanism enhancing proton conduction at such temperatures is presently the subject of debate [67]. There is evidence that such an effect is mainly due to the water retention capability of the filler [67, 70, 71]. In fact some of these compounds, for example, silica, zeolites, and so on are frequently used as desiccant materials in storage systems. In this application, after some time, saturation by the environment humidity occurs. The desiccant materials are ‗‗reactivated‘‘ by desorbing the condensed water at temperatures around 120–150 °C [67]. 0.25 Conductivity/ S cm

Power density / W cm-2

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

0.6

-1

Nafion-SiO2-PWA Nafion-SiO2 Nafion-ZrO2 Nafion- n Al2O3 Nafion- b Al2O3

0.4

SiO2/PWA 0.2

SiO2 0.15

ZrO2 n Al2O3

0.1

b Al2O3

0.05 2

4

6

8

pH filler slurry

0.2

0 0

0.5

1

1.5

2

2.5

3

Current density / A cm-2

Figure 28. DMFC power density curves at 145 °C for MEAs containing different inorganic fillers. Methanol feed 2 M, 2.5 atm; oxygen feed 2.5 atm. Pt loading 2 ± 0.1 mg cm-2. The inset shows the variation of membrane conductivity at 145 °C as a function of the pH of slurry of the filler. Reprinted from Ref. [67] with permission from Elsevier.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

111

This fact indicates that such materials may physically adsorb and retain water on the surface at temperatures close to those ideal for PEMFC/DMFC operation in automotive applications [6,16,17]. In the adsorption process, the first layer involves a chemical interaction between the surface sites of the filler and water. Generally, this causes a water displacement on the surface with formation of a chemical bond between water residues and filler functional groups [71]. Figure 29 shows the correlation between the characteristics of adsorbed water on the filler surface and DMFC power density at high temperature. Additional layers of adsorbed water may form subsequently by physical interaction involving Van der Waals bonds. In this case, no displacement of water should occur. Such bonds become weaker as the distance of the physically adsorbed water from the surface increases. Whereas, chemically adsorbed water can involve up to a monolayer, physical adsorption and water condensation in the pores may build up a shell of water molecules surrounding the primary particles and agglomerates of the inorganic filler [73]. Most of these inorganic materials have intrinsically low proton conduction up to 150 °C. It has been observed that they can be loaded with a proper dispersion in amounts up to 3–5% inside the membrane without affecting significantly the conductivity at or below 90 °C [67, 71]. Whereas, an increase in the operating temperature is possible in the presence of the filler [67, 71]. A proper distribution of the nanoparticle filler in the membrane water channels can maximize the effect of water retention in the conduction path at high temperatures. The presence of hygroscopic inorganic oxides inside the composite membrane, besides extending the operation of perfluorosulfonic membranes (e.g., Nafion) in the high temperature range, reduces the cross-over effects by increasing the ‗‗tortuosity factor‘‘ in the permeation path [70]. Such effects are particularly serious at high temperature in DMFC systems. Presently, these membranes appear to operate better at high pressure since this allows one to maintain a suitable content of liquid water inside the assembly or to facilitate water condensation in the pores. In DMFC devices, cathode operation at high pressure reduces system efficiency because of power consumption by the air compressor; whereas, the power loss for the liquid pump at the anode is less remarkable. Although, significant progress has been achieved in the last few years on the development of composite membrane-based systems [67], the high-pressure requirement, is actually the main feature limiting large application of such composite electrolytes at temperatures above 100°C.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

112

A. S. Aricò, V. Baglio and V. Antonucci 3300

3200

3500

3100

3450

3000

H2O stretching vibration / cm

-1

-1

surface OH groups stretching/ cm

3550

a

3400 2

3

4

5

6

7

8

9

pH filler slurry 1650

b 1640

300

200 1630 100

1620

Power Density / mW cm -2

H2O bending vibration / cm

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

-1

400

0 0

2

4

6

8

10

pH filler slurry

Figure 29. Variation of O–H stretching vibration frequencies for surface OH functionalities and physically adsorbed water versus the pH of slurry of the fillers (a); variation of O–H bending vibration frequencies of physically adsorbed water and DMFC maximum power density versus the pH of slurry of the fillers (b). Reprinted from Ref. [67] with permission from Elsevier.

Significant efforts have been addressed to technical aspects concerning composite membranes development; fewer attempts have been devoted to an indepth analysis of the basic mechanism of operation of such materials. A better understanding of the effects enhancing the proton conductivity at 150 °C in hybrid-membrane systems could allow one to identify new routes to enhance

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

113

conductivity and reduce high-pressure. In this regard, a wide range analysis of the filler and composite membrane properties is mandatory for a basic understanding. It appeared that an appropriate tailoring of the surface chemistry in these nanoparticles is a key step towards enhancing water retention at high temperature [71]. A rational analysis of filler effects on structural, proton transport properties and electrochemical characteristics of composite perfluorosulfonic membranes for direct methanol fuel cells was reported [67, 71 ]. It has been observed that a proper tailoring of the surface acid–base properties of the inorganic filler for application in composite Nafion membranes allows appropriate DMFC operation at high temperatures and with reduced pressures [71]. An increase in both strength and amount of acidic surface functional groups in the fillers would enhance the water retention inside the composite membranes through an electrostatic interaction, in the presence of humidification constraints, in the same way as for the adsorption of hydroxyl ions in solution [67, 71, 73]. Thus, acid–base properties of inorganic fillers play a key role in the water uptake of composite Nafion based-membranes at temperatures close to 150 °C by influencing the proton conductivity of the electrolyte. The presence of acidic OH groups on the particle surface facilitates water coordination which acts as a vehicle molecule for proton migration. Physically adsorbed water, forming a multilayer shell around the filler nanoparticles, is desorbed on the investigated inorganic fillers at around 140 °C, causing an inversion in the conductivity vs. temperature behavior (Figure 30). This determines a minimum in the value of cell resistance around this temperature. DMFC performances of various MandE assemblies based on composite membranes, containing fillers with different acid– base characteristics, increase as the pH of the slurry of the inorganic filler decreases (Figure 29). These results indicate that the ionic conductivity of the composite membranes and their range of operation may be increased by an appropriate tailoring of the surface characteristics of the ceramic oxides inside the membrane. For materials characterized by the same type of surface functional groups, the effect of the filler surface area becomes prevailing in determining the water retention properties of the composite membranes at high temperature. This effect appears to be associated with the larger number of water-adsorbing acidic sites on the filler surface. As expected, the surface properties play a more important role than the crystalline structure of the filler, since the water molecules acting as promoters towards the proton migration are effectively coordinated by the surface groups. The conductivity and performance of composite perfluorosulfonic membranes in DMFCs are strongly related to the surface acidity, which in turn influences the characteristics of the water physically

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

114

A. S. Aricò, V. Baglio and V. Antonucci

adsorbed on the inorganic filler surface. It has been observed that the more acidic the filler surface, the larger the capability to undergo a strong interaction with water through the formation of hydrogen bonds (Figure 29). This latter effect produces a decrease in the O–H stretching and bending frequencies in the physically adsorbed water. Furthermore, an increase in the water uptake in the composite membrane and an enhancement of proton conductivity are observed in the presence of acidic fillers [67, 71 ]. The proton migration inside the membrane appears to be assisted by the water molecules on the surface of the nanofiller particles and could also be promoted by the formation and breaking of hydrogen bonds [67, 73]. Operation at low pressure does not appear to introduce significant technical limitation for the hydration/conductivity characteristics of composite membranes in high temperature DMFCs (150 °C). Reduction of the applied pressure causes a reduction of the fraction of liquid water inside the system and, to some extent, the water retention ability. However, the lower consumption of electrical power needed for the auxiliaries passing from a compressor to an air blower would probably more than compensate for the power output losses due to the small increase in resistance.

0.16

Cell resistance / ohm cm 2

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

0.18

0.14 0.12 0.1 0.08 0.06 Nafion-SiO2/PWA (3%) 2 Nafion-SiO22 (3%) Nafion-ZrO2 2 (3%) Nafion-Al2O3 2 3 n (3%) Nafion-Al2O3 2 3 b (3%) Recast Nafion

0.04 0.02

Error bar for the measurement

0 80

100

120

140

160

Temperature / °C

Figure 30 Variation of cell resistance values as a function of the operating temperature for DMFCs employing different Nafion recast composite membranes. Reprinted from Ref. [67] with permission from Elsevier.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

115

On the other hand, in the absence of methanol-tolerant oxygen reduction catalysts, high oxygen partial pressures are needed for proper cathode operation in the presence of methanol cross-over. The present approach can be used for membranes alternative to Nafion such as sulfonated polyetherketones and polysulfones, as well as in high temperature hydrogen–air PEMFCs where water retention inside the electrolyte plays an even more important role than in DMFCs. Conventional ion-exchange perfluoropolymer membranes such as the wellknown Nafion® membrane are based on long-side-chain polymers (LSC). In the last decades, Solvay Solexis has developed a new short-side-chain (SSC) proton conducting perfluoropolymer membrane, i.e. Hyflon® Ion characterized by excellent chemical stability and equivalent weight (850 g/eq.) lower than conventional Nafion 117 (1100 g/eq.) [64]. Besides the improved conductivity related to the higher degree of sulfonation, the short-side-chain Hyflon® Ion ionomer is characterized, in the protonic form, by a primary transition at around 160°C whereas the conventional Nafion shows this transition at about 110°C. Such a characteristic of the Hyflon® Ion membrane ensures proper operation at high temperatures (100–150°C) provided that sufficient amount of water is supplied to the membrane or retained inside the polymer under these conditions. In principle, the water uptake properties of sulfonic acid-based membranes may be modulated by selecting a proper concentration and distribution of sulfonic groups inside the polymer. In this regard the Hyflon® Ion membrane, due to its equivalent weight lower than conventional polymers is favored for the high temperature operation. Hyflon® Ion membranes have been investigated for applications in direct methanol fuel cells (DMFCs) operating between 90° and 140°C in the Dreamcar project that was described in the transportation section. DMFC assemblies based on these membranes showed low cell resistance and promising performances compared to conventional membranes. The peak power density reached about 290 mW cm–2 at 140°C and 3 bars abs. with 1 M methanol and air feed (Figure 18).

5.2.4. Effects of Cross-Over on DMFC Performance and Efficiency Methanol cross-over through the polymer membrane is known to be one of the most challenging problems affecting the performance of DMFCs [68]. The overall efficiency of a methanol fuel cell is determined by both voltage and faradaic efficiency for consumption of methanol [12]. The faradaic efficiency is mainly influenced by methanol cross-over through the membrane. If we express methanol cross-over in terms of current density practically lost in the parasitic

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

116

A. S. Aricò, V. Baglio and V. Antonucci

reaction occurring at the cathode, Icross-over, the faradaic efficiency is determined by the equation below already reported in the section 2.3:

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

fuel 

I cell

I cross over  I cell

(30)

The methanol cross-over is usually measured indirectly by determining the amount of CO2 produced at the cathode by the oxidation of methanol on the Pt surface [68]. This CO2 can be monitored on line by using an IR-detector or by chromatographic analysis of aliquot samples of cathode outlet stream [70, 207]. The latter method is more precise but time consuming, and it requires a more expensive apparatus. Chromatographic analysis of the cathode outlet reveals the presence of small but still detectable amounts of unreacted methanol [70]. Yet, this latter is often neglected since it is generally less than a few percents of the total cross-over, i.e. comparable to the accuracy level of the IR sensor for CO2. The cross-over of methanol is influenced by both membrane characteristics and temperature, as well as by the operating current density [68, 207]. In general, an increase in temperature causes an increase in the diffusion coefficient of methanol and determines a swelling of the polymer membrane. Both effects contribute to an increase of the methanol cross-over rate. The cross-over includes both methanol permeability due to a concentration gradient and molecular transport caused by electro-osmotic drag in the presence of a proton conducting electrolyte. The latter is directly related to the proton migration through the membrane and it increases with the current density [213]. For DMFCs equipped with an alkaline electrolyte, the electro-osmotic drag is directed towards the anode. Thus, it does not contribute to the cross-over. The methanol permeability, caused by the concentration gradient at the anode-electrolyte interface, depends on the operating current density. In a polarization curve, the onset of diffusional limitations occurs when the rate of reactant supply is lower than the rate of its electrochemical consumption. Thus, if the anode is sufficiently active to oxidize methanol electrochemically to CO2 at a rate comparable to or higher than the rate of methanol supply, the methanol concentration gradient between anode/electrolyte and cathode/electrolyte interfaces could reduce to about zero [11, 68, 207]. This condition can be realized in the presence of electrodes in which there is an extension of the electrocatalyst/electrolyte interface in the electrocatalytic layer (i.e., a composite catalytic layer containing catalyst and micelles of the ionomer) at the limiting current density. Membranes with large thicknesses are effective barriers for reducing methanol cross-over; conversely, an

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

117

increase in thickness causes an increase of ohmic overpotentials. In some cases, it may be more productive to use a thinner membrane with reduced ohmic limitations and select appropriate operating conditions which limit the methanol cross-over. At 90 °C, the methanol cross-over current density through Nafion 117, at open circuit conditions and in the presence of 1M concentration in the anode feed, is larger than 100 mA cm-2 (equivalent current density), but it decreases to less than 40 mA cm-2 (equivalent current density) at about 300 mA cm-2 [68, 207]. Nafion 112 membrane with 50 µm thickness shows a few hundreds of mA of equivalent current density for methanol cross-over close to the open circuit conditions, but this reduces to 60 mA cm-2 at a current density of 500 mA cm-2 (electric output rate of DMFC) in the presence of electrodes allowing efficient methanol consumption and with moderate methanol concentrations (1 M) [11, 68]. This means that a faradaic efficiency close to 90% can be achieved with thin membranes at high current densities with consequent benefits in terms of reduced ohmic drop. Modified composite silica-Nafion membranes with silica particles (about 7 nm size) entrapped in the polymer structure show lower methanol cross-over (40 mA cm-2 equivalent current density at 145 °C and 500 mA cm-2 cell current) [70]. In this case, the silica acts as a physical barrier, especially if its action is further supported by an increase of the crystalline properties of the polymer. In general, a perfluorinated membrane with a larger extent of crystalline region with respect to the amorphous one should show reduced methanol cross-over. This does not necessarily occur at the expense of ionic conductivity as one may expect, since this latter is mainly determined by the water retention properties of the membrane which are enhanced by the addition of hygroscopic inorganic materials [70]. The actual conductivity of the composite membrane at 145 °C and high operating pressure is around 8·10-2 S cm-1. Recently, the usual method of measuring the cross-over has been criticized [214]. Since some fraction of the CO2 produced at the anode during cell operation can permeate through the membrane, the level of methanol cross-over, especially at high current densities, may be significantly affected by this phenomenon. Some studies reporting the variation of equivalent current density vs. cell current do not show a progressive decrease of methanol cross-over as the current density increases, but at high current density an asymptotic behavior towards a constant value of equivalent current density is observed [68]. This would indicate the presence of some background which may be related to the permeation of CO2. Thus, the fuel efficiency of a DMFC operating with low methanol concentrations and equipped with a thin membrane can be, at high current densities, larger than that calculated up to now and it could approach 95% or more. Studies carried out

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

118

A. S. Aricò, V. Baglio and V. Antonucci

at the University of California indicate a fuel utilization of 80-90% under optimum overall conversion efficiency conditions [12].

5.3. OXYGEN ELECTRO-REDUCTION IN DMFCS

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

5.3.1. Electrocatalysis of Oxygen Reduction Although Pt/C electrocatalysts are, at present, the most widely used materials as cathodes in proton conducting electrolyte-based low temperature fuel cells, due to their intrinsic activity and stability in acidic solutions, there is still great interest to develop more active, selective and less expensive electrocatalysts for oxygen reduction. However, there are a few directions that can be investigated to reduce the costs and to improve the electrocatalytic activity of Pt, especially in the presence of methanol cross-over. One is to increase Pt utilization; this can be achieved either by increasing its dispersion on carbon and the interfacial region with the electrolyte. Another successful approach to enhance the electrocatalysis of O2 reduction is by alloying Pt with transition metals. This enhancement in electrocatalytic activity has been differently interpreted, and several studies were made to analyze in depth the surface properties of the proposed alloys combinations [215-217]. Although a comprehensive understanding of the numerous reported evidences has not yet been reached, the observed electrocatalytic effects have been ascribed to several factors (interatomic spacing, preferred orientation, electronic interactions) which play, under fuel cell conditions, a favorable role in enhancing the ORR rate [218-222]. High specific activities of Pt-Cr and Pt-Cr-Co alloy electrocatalysts for oxygen reduction as compared with that on platinum were observed in H2-air polymer electrolyte membrane fuel cells (PEMFCs ) [222]. The formation of a tetragonal ordered structure upon thermal treatment, in the case Pt-Co-Cr, leads to a more active electrocatalyst than that with a Pt fcc structure [218-222]. The state of art Pt-CoCr electrocatalysts have a particle size of 6 nm [218-222]. In addition to the coprecipitation and impregnation methods for preparation of these alloys on a carbon support, colloidal preparation procedures have also been investigated [223]. A carbothermal reduction is often used to form the desired alloy. As said above, there is an increasing interest in developing methanol tolerant catalysts alternative to Pt for the oxygen reduction; however, it should be taken into account that if the methanol which permeates through the membrane is not completely oxidized at the cathode surface to CO2, it would contaminate the water at the outlet of the cathode compartment. This could cause several environmental

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

119

problems in the absence of a proper technical solution. This could be a single chamber DMFC with highly selective anode and cathode catalysts. Alternatively, a proper catalytic burner should be used at the cathode outlet. As discussed in a previous section, various studies carried out in the past, especially on carbon supported Pt electrocatalysts for oxygen reduction in phosphoric acid fuel cells, showed that the electrocatalytic activity (mass activity, mA g-1 Pt, and specific activity, mA cm-2 Pt) depends on the mean particle size. The mass activity for oxygen reduction reaches a maximum at a dimension of about 3 nm, corresponding closely to the particle size at which there is a maximum in the fraction of (111) and (100) surface atoms on Pt particles of cubooctahedral geometry [192]. On the other hand, the specific activity increases gradually with an increase in Pt particle size and closely follows the trend observed between the surface fraction of (111) and (100) Pt atoms and the particle size. These results have indicated that the (111) and (100) surface atoms are more electro-catalytically active than Pt atoms located on high Miller index planes [99]. Platinum atoms at edge and corner sites are considered to be less active than Pt atoms on the crystal faces. Accordingly, both mass and specific activity should decrease significantly as the relative fraction of atoms at edge and corner sites approach unity [192]. This situation occurs with Pt particles smaller than 1 nm diameter. Most of the experimental evidence observed in PEMFCs shows that crystalline electrocatalysts with about 2.5-3 nm particle size have higher activities than amorphous particles and as well as larger crystallite sizes [224]. In the case of DMFCs, an additional aspect should be considered; this is in regard to the methanol cross-over through the membrane. Methanol oxidation and oxygen reduction in the cathode compartment compete for the same sites producing a mixed potential which reduces the cell open circuit potential. In order to better understand these effects, it is worthwhile considering the mechanism for methanol oxidation and oxygen reduction; possible mechanisms are presented below. Oxygen reduction: O2 + Pt  Pt-O2 Pt-O2+H++1e-  Pt-HO2

(31) (32)

Pt-HO2+Pt  Pt-OH+Pt-O Pt-OH+Pt-O+3H++3e- 2Pt+2 H2O

(33) (34)

rds

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

120

A. S. Aricò, V. Baglio and V. Antonucci

Reaction (33) indicates an intermediate step requiring a dual-site reaction, and if it is the rate-determining step, it is more affected by particle size than intermediate step represented by Reaction (32). When the particle size becomes very small, then only the inactive edge and corner atoms will be present and dual sites of the proper orientation would not be available. Thus the activity of the particle should be lower. According to the observed variation of the reaction rate with particle size, one may conclude that the rate-determining step is represented by reaction (33) [192]. Methanol oxidation:

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

CH3OH+3Pt  Pt3-COH + 3H+ + 3ePt3COH  Pt-CO+H++2Pt + 1ePt+H2O  Pt-OH+H++1ePt-OH+PtCO  2Pt+CO2 +H+ + 1e-

(35) (36) (37) (38)

In the case of methanol oxidation at the cathode, three neighboring Pt sites in a proper crystallographic arrangement will favor the methanol chemisorption. Since at high cathodic potentials the water discharging reaction (reaction 37) is largely favored, oxidation of the methanolic residues adsorbed on the surface proceeds very fast producing a parasitic anodic current on this electrode. When the particle size of the electrocatalyst is very small or one has an amorphous Pt electrocatalyst for the oxygen reduction, methanol chemisorption energy could be lower and hence the cathode less poisonable. But, at the same time, due to the fact that only the inactive edge and corner atoms will be present and dual sites of the proper orientation will not be available, the activity of such electrocatalyst for oxygen reduction will be lower. The best compromise is to modulate the structure and the particle size between amorphous and crystalline in order to decrease the poisoning by methanol and enhance oxygen reduction. A second possibility is to use a promoting element for oxygen reduction which simultaneously hinders the methanol chemisorption still maintaining the proper structure and particle size.

5.3.2. Pt-Based Catalysts and Non Noble Metal Electrocatalysts In order to enhance the oxygen reduction kinetics in DMFCs, research activities based on the modification of Pt electrocatalyst by addition of transition metals were carried out in various laboratories [223, 225]. In phosphoric acid and PEMFCs the intrinsic electrocatalytic activity of Pt alloys (Pt-Cr, Pt-Ni, Pt-Cr, Pt-

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

121

Cu, Pt-Fe), with a lattice parameter smaller than that of Pt, was found to be higher than on the base metal [213-222]. This effect has been related to the nearest neighbor distance of Pt-Pt atoms on the surface of the fcc crystals. Since it has already been observed that the rate determining step involves the rupture of the OO bond through a dual site mechanism, a decrease of the Pt-Pt distance favors the dual site O2 adsorption. Most of these evidences were derived from studies carried out for phosphoric acid fuel cells. In such cases leaching of non-noble elements produces a surface roughening with a corresponding increase of the Pt surface area. In PEMFCs where the electrolyte anions are chemically bound to the backbone of perfluorosulfonic acid membranes and the cell temperature is relatively low, corrosion problems are minimized compared to phosphoric acid fuel cells [213, 222]. However, the electrolyte corrosion due to the high cathode potential is an important issue. Also in the case of PEMFCs, there is specific evidence for an enhancement of the reaction rate for O2 reduction on a Pt- alloy electrocatalyst [213, 222, 224]. Similar aspects have been considered to interpret the promoting behavior in PEMFCs. Many investigations in PEMFCs have shown that enhanced electrocatalytic activity for the ORR for some binary Pt based alloy catalysts, such as Pt–M, (where M = Co, Fe, etc.), in comparison with pure Pt [226-231] can be interpreted in terms of the increased Pt d-band vacancy (electronic factor) and by the favorable Pt–Pt interatomic distance (geometric effect). According to what discussed above, a lattice contraction due to alloying would result in a more favorable Pt–Pt distance for the dissociative adsorption of O2. Besides this, an interplay between electronic and geometric factors (Pt d-band vacancy and Ptcoordination numbers) and its relative effect on the OH chemisorption from the electrolyte occurs [232]. As pointed out in a previous sub-section, methanol chemisorption and oxygen reduction reactions require an appropriate geometrical arrangement of Pt atoms. Both processes are favored on a Pt (111) surface, which possesses the reasonable nearest Pt-Pt interatomic distance. Thus, the poisoning effect of methanol crossover should be more significant on the Pt (111) surface. However, it is difficult to quantify the compensation effect due to the increased methanol oxidation rate at the sites where oxygen reduction is favored. Beside these aspects, recent work has also taken into consideration the role played by the promoting element (Co, Cr) for the removal of strongly bonded oxygenated species on Pt through an intraalloy electron transfer [225]. Chemisorption of oxygen molecules occurs more easily on oxide-free Pt surfaces [225]. But, as reported in cyclicvoltammetry studies [136], methanol adsorption and oxidation are favored on a reduced Pt surface rather than on platinum oxide. The addition of Co and Cr to Pt appears to

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

122

A. S. Aricò, V. Baglio and V. Antonucci

simultaneously favor both oxygen reduction and methanol oxidation reactions. As an example, a promoting effect of Cr on Pt for methanol electro-oxidation was already reported [56, 213, 224]. Furthermore, the presence of electropositive elements, alloyed to Pt, favors the chemisorption of OH species on neighboring Pt sites. In the absence of oxygen, a small but noticeable promoting effect for methanol oxidation in a wide range of anodic overpotentials has been observed by Cr, Fe and other elements usually selected as catalytic enhancers for the oxygen reduction reaction in PEFCs [6]. At present, it is difficult to establish if the beneficial effect on oxygen reduction is prevailing with respect to the promoting effect on methanol oxidation at DMFC cathode. Some papers have reported that a limited enhancement of ORR was shown when a Pt-alloy instead of Pt was used in DMFCs [121], whereas in other cases, it was observed that Pt-alloys are less tolerant than Pt to methanol. These aspects are discussed in detail below with particular regard to the Pt-Fe system. As well known, methanol crossover results in a significant loss in efficiency of a DMFC because on the Pt cathode two reactions compete, i.e. O2 reduction (ORR) and CH3OH oxidation. One possibility to solve this problem is to use an oxygen reduction catalyst, inactive towards methanol oxidation or having a high methanol tolerance. Thus, it is necessary a cathode electrocatalyst that catalyzes the oxygen reduction and limits methanol oxidation. Watanabe et al. reported that, after electrochemical testing of a Pt–Fe alloy, the catalyst was covered by a thin Pt skin of less than 1 nm in thickness [229, 230]. Moreover, they suggested that during the adsorption step, a p orbital of O2 interacts with empty d orbitals of Pt and consequent back donation occurs from the partially filled orbital of Pt to p* (anti-bonding) molecular orbital of O2. The increase in d-band vacancies on Pt by alloying produces a strong metal– O2 interaction. This interaction weakens the O–O bonds, resulting in bond cleavage and bond formation between O and H+ of the electrolyte, thus improving the ORR. As for PAFCs and PEMFCs, one concern with Pt alloys in direct methanol fuel cells is dissolution of transition metal. Pourbaix diagrams indicate that most metals such as Co, Cu, Fe, Ni, etc. are soluble at a potential between 0.3 and 1 V versus SHE and at low pH values. Although the cathode potentials in methanol fuel cells are usually lower than for PAFCs and DMFCs, this aspect should be taken into proper consideration. Again, the dissolution of the transition metal would cause an increase in surface area of the residual Pt catalyst. Leaching of Fe ions into the membrane may cause increase of resistance and accelerate the degradation reactions of the polymer. There are different procedures related to catalyst preparation for DMFCs [15, 16]. These will be discussed more in detail in the technology section. Preparation

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Fundamental Aspects: Status of Knowledge

123

methods such as impregnation, colloidal deposition and surface reduction involve the adsorption of active compounds on a carbon black surface. The synthesis of a highly dispersed electrocatalyst phase in conjunction with high metal loading on a carbon support is one of the present goals in DMFCs. Most studies have dealt with the preparation of catalysts characterized by particle size larger than about 46 nm and low concentration of active phase on carbon due to the need of a high temperature treatment (generally a carbothermal reduction) to form bimetallic alloys of transition metals with Pt [218]. Recently, cathode catalysts synthesized by a low-temperature colloidalincipient wetness route characterized by high concentration of metallic phase on carbon black and particle size smaller than 3 nm have been investigated [233, 234]. The new approach allowed to obtain carbon supported bimetallic nanoparticles with a particle size of about 2–2.5 nm and a suitable degree of alloying. XRD patterns of these Pt/C and Pt–M/C catalysts are reported in Figure 31. The physicochemical properties are reported in Table 6. These showed the typical fcc crystallographic structure of Pt. A moderate degree of alloying was found for Pt–Fe catalysts; whereas, the degree of alloying was slightly larger for Pt–Co/C and significantly larger for Pt–Cu/C compared to Pt–Fe by using the same procedure. These evidences were derived by the decrease of the lattice parameter (Table 6). For what concerns the polarization behavior, the Pt–Fe/C (2.4 nm) performed better than the Pt/C, Pt–Cu/C and Pt–Co/C catalysts with similar particle size (2.1–2.8 nm) at 60 ◦C (Figure 32). For the Pt–Fe system, the open circuit voltage (OCV) was also higher than that recorded with the other catalysts; this indicated a higher methanol tolerance of this catalyst, since the same membrane was used in all experiments. Table 6. Physicochemical characteristics of Pt and Pt-M/C catalysts for oxygen reduction in DMFCs. Reprinted from Ref. [233] with permission from Elsevier Catalysts 60% Pt/C

Mean particle size/nm (XRD) 2.8

Mean particle size/nm (TEM) 3.0

Lattice parameter (A0)/nm (XRD) 0.392

Pt/M at. (XRF) -

60% Pt-Fe/C

2.4

2.5

0.390

3.33

60% Pt-Cu/C

2.1

2.5

0.387

3.79

60% Pt-Co/C

2.3

2.5

0.389

3.50

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

A. S. Aricò, V. Baglio and V. Antonucci

Intensity / a.u.

124

Pt/C PtCo/C PtFe/C PtCu/C

15

25

35

45

55

65

75

85

95

2

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Figure 31. XRD patterns of Pt and Pt–based bimetallic catalysts for the oxygen reduction reaction in DMFCs. Reprinted from Ref. [233] with permission from Elsevier.

The better performance of Pt–Fe catalyst for ORR was also confirmed by the cathode polarization curves (Figure 32). Methanolic residues stripping analysis of these catalysts (Figure 33) showed that this enhanced activity possibly derived from better methanol tolerance and higher intrinsic catalytic activity for oxygen reduction. The presence of a significant current density in the hydrogen desorption region (E < 0.4V RHE) for the Pt-Fe even after methanol adsorption, not observed for the other catalysts, indicated suitable methanol tolerance properties. The positive shift of the potential for Pt-Oxide reduction was associated to a better intrinsic catalytic activity. The electrochemical active surface area as derived by the methanolic residues stripping analysis was larger for catalysts with smaller particle size, e.g. PtCu/C. Yet, it appeared that the small increase of electrochemically active surface area in PtCu did not play the same role of the increase of methanol tolerance and intrinsic catalytic activity in PtFe. The electrochemical behavior of Pt–Fe for ORR was also investigated in the presence of methanol in the acidic solution [234] and compared to the Pt catalyst. Figure 34 shows the Tafel slope profiles as obtained from rotating disc experiments for the ORR reaction on Pt and Pt–Fe catalysts with and without methanol in the solution at 1000 rpm. Both catalysts showed a decrease in performance in the presence of methanol. With a methanol concentration of 0.05 M, a shift towards lower potentials was observed for both catalysts. However, the polarization curve on the Pt–Fe catalyst was less negatively shifted in the presence of methanol than that on Pt; this clearly indicated a promoting effect of the bimetallic catalyst in enhancing the ORR and a better tolerance to methanol.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Fundamental Aspects: Status of Knowledge

125

1

80

Cell potential / V

70

60% PtCu/C

60

-2

60% PtFe/C 60% Pt/C 60% PtCo/C

50

0.6

T = 60°C

40

Atmospheric pressure

0.4

30 20

0.2

Power density / mW cm

(a)

0.8

10 0

0 0

0.1

0.2 0.3 0.4 Current density / A cm -2

0.5

0.6

1

Potential / V vs RHE

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

(b)

60% 60% 60% 60%

0.8

Pt/C Pt-Fe/C Pt-Cu/C Pt-Co/C

0.6

0.4 0

0.1

0.2 0.3 Current density / A cm-2

0.4

0.5

Figure 32. (a) Polarization and power density curves for the DMFCs equipped with the various cathode catalysts at 60 °C under atmospheric pressure, and (b) cathodic polarization curves for DMFCs based on the different cathode catalysts recorded at the same operatine conditions. Reprinted from Ref. [233] with permission from Elsevier.

At potentials lower than 0.7 V, the performance of the Pt–Fe catalyst in the presence of methanol appeared to be similar to that of Pt without alcohol. Similar methanol tolerance and enhanced oxygen reduction properties were observed for several Pt alloys with transition metals [235]; these effects were ascribed to a lowered activity for methanol oxidation.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

126

A. S. Aricò, V. Baglio and V. Antonucci 0.15

PtCu/C 0.1

PtFe/C

Current density / A cm

-2

PtFe/C

Pt/C 0.05

0

Pt/C

-0.05 T = 60°C

-0.1

Scan rate: 50 mV sec-1

PtCu/C PtFe/C

-0.15 0

0.2

0.4

0.6

0.8

1

1.2

Potential / V vs RHE

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Figure 33. Adsorbed methanolic residues stripping voltammetry at a scan rate of 50 mVs−1 for the different cathode catalysts at 60 °C. Reprinted from Ref. [233] with permission from Elsevier.

The enhanced catalytic activity for Pt–Fe catalysts has been attributed to the presence of a Pt skin over the alloy together with an electronic effect induced by Fe on Pt, as previously reported by Watanabe and coworkers [229, 230]. The formation of the Pt skin probably occurs because iron on the surface leaches out of the alloy during operation in acidic electrolytes, while Pt atoms are redeposited and rearranged on the surface. For what concerns the electronic effect, it was emphasized that the electronic structures of the Pt skin layers are altered by the underlying alloy substrates, which in turn facilitates the electron transfer to oxygen molecules [229, 230]. On the other hand, Li et al. claimed that the improvement in the performance of Pt–Fe/C for ORR may be partly due to the higher peroxide decomposition activity of Pt in the presence of dissolved Fe favoring the 4e− transfer route [236]. In this respect, EDX measurements after operation showed a decrease in the Fe content in the alloy, revealing a partial ion dissolution during operation [234]. This may cause an increase in roughness and surface area, but the long-term stability of this system has to be confirmed by further experiments. Alternatively to platinum, organic transition metal complexes are known to be good electrocatalysts for the oxygen reduction reaction. Transition metals, such as iron or cobalt organic macrocycles from the families of phenylporphyrins, phthalocyanines and azoannulenes have been tested as O2-reduction electrocatalysts in fuel cells [53, 237-240].

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

RHE vs. RHE V vs. Potential/ V Potential/

Fundamental Aspects: Status of Knowledge

127

1 Pt-Fe/C -76 mV/dec

0.8 Pt/C

0.6

-125 mV/dec 0.1 M HClO4

0.4 -3

-2

-1

0

1

RHE vs.RHE V vs. Potential/ V Potential/

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Log (((J*Jl)/(Jl-J)/ (((J*Jl)/(Jl-J)/mA mAcm cm-2) 1

Pt-Fe/C -93 mV/dec

0.8

Pt/C

0.6 0.4

-138 mV/dec 0.1 M HClO4+ 0.05M CH3OH

-3

-2

-1

0

1

-2

Log (((J*Jl)/(Jl-J)/ (((J*Jl)/(Jl-J)/mA mAcm cm ) Figure 34. Tafel plots for ORR on Pt-Fe/C and Pt/C catalysts in 0.1 M HClO4 with and without methanol in the solution.

One major problem with these metal organic macrocyclics is their chemical stability under fuel cell operation at high potentials. In many cases, the metal ions are irreversibly dissolved in the acid electrolyte. However, if the metal-organic macrocyclic is supported on a high surface area carbon and treated at high temperatures (from 500 to 800 C), the residue exhibits promising electrocatalytic activity without any degradation in performance, from which one may infer the good stability of the metal in the electrocatalyst [239]. In some other studies, a few inorganic materials have been proposed as suitable substitutes for platinum in methanol fuel cells due to their selectivity for oxygen reduction, even in the presence of methanol. These materials mainly

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

128

A. S. Aricò, V. Baglio and V. Antonucci

consist of the Chevrel-phase type (Mo4Ru2Se8), transition metal sulfides (MoxRuySz, MoxRhySz) or other transition metal chalcogenides ((Ru1-xMox)SeOz) [52, 241]. Some of these possess semiconducting properties, thus, in theory, they could introduce an additional ohmic drop in the electrode. However, their activity for oxygen reduction is significantly lower than Pt [2]. Carbon supported Ru electrocatalysts are reported to exhibit high selectivity for oxygen reduction in the presence of methanol but their activities are significantly lower [242]. The tolerance of these materials to methanol is due to the absence of adsorption sites for methanol dehydrogenation. In the case of Ru/carbon, at high potentials, the surface is mainly covered by Ru oxides on which methanol chemisorption is hampered [242]. In a previous study [20], we have shown that in the presence of methanol, the Tafel slope for oxygen reduction at 60 °C on a Pt surface is practically unaltered at high potentials (-60 mV/dec), i.e., in the region where the surface is mainly composed of Pt-oxides. At lower potentials, where the metallic surface is exposed, methanol is effectively oxidized in the presence of oxygen; this parasitic reaction determines an increase in Tafel slope from 130 to 170 mV/dec [20] . This effect is the same as that observed in classical cyclic voltammetry of methanol at Pt electrodes [136]; the methanol oxidation peak in the anodic scan is followed by a sharp decrease of current due to the formation of a strong Pt-oxide layer on the surface before oxygen evolution. The presence of Pt oxides on the cathode catalyst surface enhance its methanol tolerance. Probably, this could also explain why Pt-Cr, Pt-Co and Pt-Co-Cr have shown only limited or no enhancement of oxygen reduction kinetics in DMFCs, as compared to that in H2-air SPEFCs. In general, Co and Cr interact with Pt in the alloy by electron donation to 5d Pt orbitals due to their electropositive character. Thus, they remove strongly bonded oxygen species from Pt sites and favor methanol chemisorption on Pt [225]. Regarding the development of cathode catalysts for alkaline DMFCs, it should be pointed out, as for the anode catalyst, that the corrosion problems are minimised by the operation at high pH values. Thus, due to the large variety of catalytic formulations that may be screened it should be more easily to discover a methanol tolerant cathode catalyst. The reaction kinetics for oxygen reduction at the cathode are more favorable in alkaline media. This allows to replace Pt with less noble or non precious catalysts with significant advantages in terms of cost reduction. Among the various cathode formulations, Ag and MnO2 catalysts have shown suitable methanol tolerance and catalytic activity for oxygen reduction [60].

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Chapter 6

6. TECHNOLOGY DEVELOPMENT

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

6.1. CATALYST PREPARATION, ELECTRODE MANUFACTURING AND MEMBRANE-ELECTRODE ASSEMBLIES (MEAS) The MEAs are usually considered as the most important components of a DMFC power source. They contain backing layers, gas diffusion layers, catalytic layers and membrane. However, a significant role is also played by the flow field/current collector, reactant manifold and the stack housing. A stack module is usually formed by a series connection of cells (e.g. through bipolar plates). Several modules can be connected each other in series or in parallel depending on the required electrical characteristics for the power source. Furthermore, several auxiliaries are necessary for thermal and water management, start-up, shut-down and normal operation. These include compressor/blowers, fuel tank and liquid pumps, methanol concentration sensor, gas/liquid separation devices, eventual catalytic burner, DC/DC (step-up) and DC/AC converters. All the above components form a DMFC system and they are the subject of development and integration studies. In this section we will restrict our analysis to a few aspects that we consider of priority interest. The core of the fuel cell device is certainly the electrode-electrolyte assembly. In the previous section, we have discussed both catalyst and membrane properties for DMFCs. In this section, we will discuss mainly electrode properties and manufacturing methods, including catalyst preparation, assembling procedures etc. Under practical operating conditions, the electrode structure plays a role similar to that of catalyst and membrane. Thus, the electrode properties should be tailored as a function of the operating conditions. An appropriate knowledge of the influence of these

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

130

A. S. Aricò, V. Baglio and V. Antonucci

conditions on the electrochemical behavior is fundamental to individuate development strategies for the fuel cell components. An important role is also played by the stack hardware components, such as the flow field, in determining the performance characteristics. A proper integration of the stack in a system containing all the auxiliaries is necessary to determine optimal operating conditions as well as to control the start-up and shut-down procedures. In the following, a few examples of catalyst preparation are presented before discussing electrode properties and manufacturing as well as MEA assembling procedures. The synthesis of a highly dispersed electrocatalyst phase in conjunction with a high metal loading on a carbon support is one of the present goals in DMFCs [2]. One of the main requirements for an optimal electrocatalyst is its high dispersion. The mass activity (A g−1) of the catalyst for an electrochemical reaction is directly related to the degree of dispersion since the reaction rate is generally proportional to the active surface area. However, for both methanol oxidation and oxygen reduction process, it has been found that there is an optimum particle size for Pt electrocatalysts (see section 5), that corresponds to a maximum in mass activity [2]. For the oxygen reduction reaction, it was found that the specific activity (A cm−2 real area) increases with the particle size and thus decreases with the active surface area. Similar evidences have been observed for the methanol oxidation process if small nanosized metal particles (e.g. PtRu with about 1-1.5 nm mean particle size) with significant amorphous character are considered as the lower limit [195]. Another significant aspect concerns the loading of the metal phase on the carbon support. A large Pt concentration on the carbon substrate will reduce the anode thickness for the same Pt loading (mg /cm2). Thus, it is possible to enhance the mass transport through the electrode and at the same time reduce the ohmic drop. Accordingly, it is important to prepare a highly concentrated catalyst phase on a conductive support while maintaining a high level of dispersion. The main routes for the synthesis of Pt-Ru/carbon black electrocatalyst include impregnation, colloidal procedures, self-assembling methods, decoration etc.. The impregnation is characterized by a deposition step of Pt and Ru precursors (e.g., H2PtCl6, RuCl3, Pt(NH3)2(OH)2, Ru3CO12, Pt(NH3)2(NO)2 etc.) followed by a reduction step. This can be a chemical reduction of the electrocatalyst slurry in aqueous solution by using N2H4, NaS2O5, NaS2O3, NaBH4 (liquid-phase reduction) or gas-phase reduction of the impregnated carbon black by a flowing hydrogen stream. For cathode catalysts consisting of Pt alloyed with transition metals such as Co, Fe, Cr, a carbothermal reduction is often used. In this procedure, the oxide precursor supported on a high surface area carbon reacts with the carbon support in an inert stream (N2) at suitable temperatures (600-900

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Technology Development

131

°C). The sintering effect produced by the carbothermal reaction is less significant as compared to the hydrogen reduction. Furthermore, the high temperature process enables the formation of new crystallographic structures with appropriate catalytic activity. It has been shown that the impregnation method can be used for the synthesis of a multifunctional system, e.g. from a bimetallic to a quaternary electrocatalyst (Pt-Ru-Sn-W) [133]. However, it requires the use of high surface area carbon black such as Ketjenblack whose limitations for the operation of a methanol fuel cell have been pointed out in a previous section. Furthermore, this procedure does not allow to obtain high dispersions in the presence of high metal loadings. There are various colloidal deposition routes, developed by Jalan [243], Bönnemann et al. [244], Petrow and Allen [245], etc. The advantages of these preparation routes consist in the attainment of significantly high surface areas in the presence of high metal loading on carbon. The main disadvantages are represented by some complexity of the preparation steps in the overall synthesis, the use of organic compounds/solvents and the higher production costs. An additional method is based on the thermal decomposition of appropriate high molecular weight Pt-precursors such as Pt-carbonyl compounds [246] to obtain unsupported high surface area electrocatalysts. Other methods, such as coprecipitation, self-assembling, sol-gel or physical methods (e.g., sputtering [247]) have stimulated much interest in the past and they are now becoming interesting for the synthesis of catalysts for portable power sources [247]. The preparation of cathode Pt-based electrocatalysts for DMFCs is practically the same as for the anode electrocatalysts. However, as above discussed, there are some differences in terms of thermal treatment of a Pt-Ru or Pt-Sn alloy for methanol oxidation and that of Pt-Cr, Pt-Co, Pt-Co-Cr electrocatalysts for oxygen reduction [213-22]. In the latter case, the activation temperature is usually significantly higher (around 700-900 °C), as compared to the 100-400°C treatments used for the anode alloy. For the cathode electrocatalyst, a complete alloying is generally accompanied, in most cases, by a phase transition from fcc (cubic) to an ordered tetragonal (fct) structure; this treatment causes an increase in mean particle size. Thus, the increase of intrinsic electrocatalytic activity is counteracted by the decrease of the surface area of the electrocatalysts. The choice of the most appropriate preparation procedure relies on the following considerations. It is well known that the preparation procedure of electrocatalysts influences their physico-chemical properties and thus their activity. The performance characteristics of an electrocatalyst depend on its chemical composition (surface and bulk), structure and morphology. Accordingly, the selected methodology of electrocatalyst synthesis should allow one to address

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

132

A. S. Aricò, V. Baglio and V. Antonucci

the process for the attainment of a proper structure (crystalline or amorphous) and with a chemical composition on the surface as close as possible to the nominal or bulk composition. Since the rate of all electrocatalytic reactions is strictly related to the active surface area, besides surface chemistry, the morphology of the electrocatalyst needs to be tailored. Morphology is not only related to the metalphase area but also to the presence of micro- and macro pores in the electrocatalyst support which could facilitate or hinder the mass transport properties. All these characteristics determine the cell performance even if the relative influence of each parameter is still not known in detail. It is thus necessary to select appropriate procedures for an optimization of these characteristics, i.e. composition, structure, particle size, porosity, etc. Generally, a combination of physico-chemical and electrochemical analyses carried out on electrocatalysts with different characteristics indicates the system that better suits the scope of application in a DMFC. The performance of a DMFC is strongly affected by the fabrication procedure of the membrane-electrode assembly (MEA). The conventional technology that was used two decades ago, consisted in the preparation of gas-diffusion electrodes having suitable polytetrafluoroethylene (PTFE) contents in both diffusion and catalyst layers. Nafion ionomer was spread onto the electrocatalyst layer, followed by the preparation of membrane-electrode assembly by a hot-pressing procedure [19]. A disadvantage of this procedure is that the Nafion impregnated into the active layer of the electrode has a limited penetration depth. This drawback reduces the electrochemically active area between electrocatalyst particles and ionomer, thus decreasing the catalyst utilization [243]. The function of PTFE in the catalyst layer is to provide a network for gas transport and to give structural integrity to the layer. However, it has been evidenced that the ionic path in the electrocatalytic layer due to the recast ionomer is hindered on the surface of the electrocatalysts and pores covered and/or blocked by PTFE particles [243]. To avoid these drawbacks, modified electrode preparation methods have been developed in the last decade. These methods are characterized by a procedure involving a direct mixing of the ionomer with the electrocatalyst (Figure 35). Both direct mixing of electrocatalyst with ionomer in appropriate solvents like glycerol and a ―paste process‖ procedure, based on the addition of a colloidal ionomer to the electrocatalytic layer, have been successfully developed in DMFCs in order to obtain PTFE-free electrocatalytic layers [243-246]. Such fabrication methods do not require a laborious procedure. The optimization of the structure of the electrode and/or MEA requires an appropriate investigation of the microstructure of the carbon support, in order to ideally distribute the ionomer on the carbon surface containing Pt or Pt-Ru particles.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Technology Development

133

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Figure 35. SEM-FEG image of a cathode composite DMFC catalytic layer (catalyst and ionomer).

The main problem of direct mixing or ―paste process‖ procedures is that the ionomer does not soak deeply into the smaller pores of the active layer, as it would in the case of a liquid electrolyte. Thus, the reaction area is limited to an interface between Pt particles distributed on the outer surface of carbon agglomerates and ionomer [247]. However, these effects do not explain completely the reduced catalyst utilization. Such a difference is in part determined by agglomeration effects and metal-support interaction. As example, the electrochemical active surface area (ECA) of a highly dispersed carbon supported Pt-Ru catalyst, as determined by CO stripping analysis under fuel cell configuration (catalyst-ionomer composite), and in the presence of high Pt loadings (1-2 mg Pt cm-2, as typically occurs in a DMFC) is about 50% of that determined by XRD or TEM analyses (MSA). The Pt loading could be significantly reduced if the Pt utilization is increased. The enhancement in Pt utilization could be optimally achieved if both the anode and cathode electrocatalysts are tailored to be distributed on the outer surface of the carbon agglomerate; furthermore, an increase in dimension of carbon pores and a decrease in the size of ionomer micelles are needed to increase the contact region with the ionomer. One of the approaches that is recently used especially for PEMFCs concerns with the direct deposition of the catalyst onto the membrane to form a catalyst coated membrane (CCM). The diffusion and backing layers are added subsequently e.g. during cell and stack assembling. In this configuration, there is an intimate contact between the catalytic layer and the membrane; whereas, the

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

134

A. S. Aricò, V. Baglio and V. Antonucci

diffusion layer is just put in contact with the catalytic layer. This method first, developed by Wilson and Gottesfeld [244] and initially reported as ―ink method‖ appears more appropriate to deposit thin catalyst layers with small catalyst loading (0.05-0.4 mg cm-2). The typical catalyst loading in the ink method is often too small with respect to what is required in the present DMFCs. However, there are several modifications that have been developed for this approach and larger noble metal loadings are now possible in the CCM configuration. The fabrication methods include spray-coating, printing, brushing etc. When air is fed to the cathode side, while oxygen reacts to produce water, the nitrogen contained in the feed stream remains entrapped in the pores of the electrode; the entrapped nitrogen is a diffusion barrier for the incoming oxygen, and it results in mass transport overpotential with consequent performance losses even at intermediate current densities. Furthemore, although it is known that oxygen permeability through the ionomer is high at high current densities, transport of this gas to the reaction sites is retarded by flooding of the electrocatalyst layer [2]. Due to this flooding of the active layer, the ionomer swells until it is saturated with water, thus increasing the hydrophilicity of the layer. Such drawbacks have been conveniently reduced in air feed-SPE fuel cells by using thin film electrodes. These are characterised by low electrocatalyst loadings (0.05 - 0.1 mg cm-2) [244]. Due to the lower performance of the oxygen electrode with such low Pt loadings in DMFCs, alternative solutions should be investigated. Some alternatives to the conventional preparation procedure of the electrocatalyst layer have been developed to enhance the oxygen transport properties when air is used as feed stream. Gas channels allowing a fast transport of the reaction gas and an easy removal of the excess N2, can be realized in the cathode layer by means of PTFE-carbon composite ducts [245]. These ducts inside the electrocatalytic layer have been obtained by ultrasonically mixing the electrocatalyst-ionomer mixture with appropriate amounts of a PTFE-carbon composite before spreading the resulting paste onto carbon cloth or carbon paper. The PTFE in this case does not cover Pt sites, being only loaded on the carbon black; at the same time, the PTFE-C mixture contains a significant number of dry pores, thus enhancing the mass transport properties of the gas; its function will be to supply the reacting gas to the electrocatalytic sites and to exhaust the product water as well as the excess nitrogen [245]. As a consequence, this configuration does not affect the electrocatalyst utilization and the continuity of ionomer in the catalyst layer, thus improving the mass-transport properties of the electrode. A second approach is to modify the hydrophobic-hydrophilic properties of the ionomer in the electrocatalyst layer. As example, Nafion ionomer in solution is

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Technology Development

135

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

characterized by a rod-like structure in the form of direct micelles with the perfluorinated matrix forming the inner part and ionic sulfonate groups on the surface. In a completely hydrophilic layer, the gas is not easily provided in the absence of a network of hydrophobic pores within the electrocatalyst layer. Gas transport through pores filled with ionomer and/or water will not be a limiting step if it only occurs over short distances. This is the case of thin film electrodes employed in SPEFCs. A complete hydrophilic layer could become a problem for gas transport in DMFCs since these devices require significant electrocatalyst loadings and, consequently, thick electrodes are generally used. To overcome these problems, the recast Nafion gel inside the electrocatalytic layer is thus modified by annealing at moderate temperatures (150-180°C). The thermal treatment degrades a significant fraction of sulfonic acid groups increasing the degree of hydrophobicity of the layer. This partially hydrophobic layer improves the mass transport properties at high currents and the removal of product water [243]. A third approach is to use pore formers such as (NH4)2CO3 to increase the porosity in the active layer of the cathode, as suggested by Wendt and co-workers [248] and by Appleby et al. [249]. This approach has already shown interesting results for air feed DMFCs. Each of these procedures can have specific advantages depending on the operating conditions e.g. operation temperature, air pressure, level of humidification etc.

6.2. ELECTROCHEMICAL TESTING: INFLUENCE OF OPERATING CONDITIONS The performance of a DMFC is generally reported in terms of maximum power density and power density at a particular cell voltage, e.g. 0.5 V. It strongly depends on the operating conditions (working temperature, type of oxidant, i.e. air or oxygen, back-pressures, reactant flow-rate, mass transport conditions, MandE assembly conditioning) and on the characteristics of the fuel cell components (catalyst loading, fabrication procedures of electrodes and MandE assembly, membrane conductivity, flow-fields). It is very difficult to make a comparison among the various polarization data reported in the literature, mainly on account of the different operation conditions. The best single cell (5-50 cm2 active area) DMFC performances, achieved by various groups in the last decade, are about 300-500 mW cm-2 and 200-300 mW cm-2 as maximum power density, in the presence of oxygen and air feed at the cathode, respectively [9, 11, 13-18, 40, 66, 68, 184]. These performances have been obtained at temperatures close or above

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

136

A. S. Aricò, V. Baglio and V. Antonucci

100 °C, with overall Pt loadings of 2-5 mg cm-2 and, in most cases, under pressurized conditions. At a cell voltage of 0.5 V, the best performances in the presence of air as oxidant are about 200 mW cm-2. If the DMFC is operated at room temperature and under air breathing conditions, the performances become one order of magnitude lower than those achieved under the above conditions. Typical performances achieved at ambient temperature under passive operation mode vary between 10 and 40 mW cm-2 depending on methanol concentration and device characteristics (see section 4.1). The variation of electrochemical performance of the DMFC with methanol concentration generally reflects two phenomena. By increasing the methanol concentration, the coverage of the electrocatalyst sites by methanolic residues increases, but, at the same time the water concentration in the solution decreases. A reaction order of 0.5 with respect to methanol concentration has been observed in half-cell polarization experiments up to 2.5-3 M concentration [30]. During fuel cell operation, a high methanol concentration in the anode feed determines a high concentration gradient across the interface with consequent increase of the crossover through the Nafion membrane. Thus, a delicate balance among the effects of methanol oxidation kinetics and methanol cross-over is required to enhance the performance in the activation and ohmic-controlled regions [10]. On the other hand, the polarization behavior in the mass transfer controlled region is directly related to the methanol concentration. By increasing the methanol concentration, a corresponding increase of limiting current density in the I-V characteristics is generally observed [10]. Generally, the highest power densities, under normal mode operation, have been observed with 1 to 2 M methanol concentrations; 1 M methanol is preferable in the presence of an air-fed DMFC because the effects of methanol cross-over on the cathode polarization and fuel efficiency are less significant with a lower methanol concentration. In the passive mode, the maximum power density often increases up to 5 M methanol concentration. The direct reaction between methanol and oxygen molecules at the cathode due to the cross-over determines an increase of temperature at the catalytic sites that is proportional to the cross-over rate and thus to the concentration. It is often difficult to estimate the real temperature at the reaction sites and thus the effective influence of the methanol concentration on the performance in the passive mode. High cathode back pressures are desirable to achieve suitable oxygen reduction rates in the presence of methanol cross-over. A high oxygen partial pressure is favorable due to the positive reaction order and to impede methanol chemisorption on the cathode surface [11]. High air flow rates counteract the negative effects of flooding and cross-over by increasing the efficiency of oxidation of organic molecules [11]. Yet, a stack operation at high pressure (2-3

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Technology Development

137

atm) and high flow rates requires the use of a suitable compressor which consumes from 10 to 20% of the output power for a 5 kW stack [7]; for operation at lower pressure (1.5 atm) with suitable air stoichiometries, an air blower is sufficient, allowing a significant reduction in energy consumption and cost. Suitable anode back pressures are necessary for high temperature operation. This aspect is crucial to maintain a good hydration of the membrane. The energy consumption of the liquid pump is much lower with respect to the air compressor. However, a pressure gradient between anode and cathode would not be appropriate for the cross-over especially at high temperatures.

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

6.3. STACK HARDWARE AND DESIGN The architecture of a DMFC stack for transportation and stationary applications, including remote and distributed generation of electrical energy, is essentially similar to that of a PEFC stack for the same applications [5, 55, 56]. Whereas, a large variety of approaches and designs has been adopted for portable fuel cell stacks [4, 5]. The requirements for the stack vary depending on the applications. Compact size, fast start-up procedure and high performance are required for transportation applications [7]. Easy handling, miniaturization and rapid fuel refilling are especially important for portable applications [4,5]. There are no particular constraints in terms of volume and weight for stationary uses. However, longer life-time characteristics are expected for such applications as compared to the use in portable systems. In general, high reliability and low fabrication costs are important for a large-scale diffusion. In the case of miniaturized power sources, the fabrication costs may be larger than in other applications but competitive with the present Li-batteries [4]. The conventional PEMFC stack architecture [55, 56] for transportation and stationary applications is based on bipolar plates connecting in series the various cells (MEAs). Besides, there are two end plates enabling current collection, a manifold and appropriate gaskets which allow together with the flow-fields in the bipolar plates to distribute the reactants over the various cells. There are several variants which mainly concern with flow field-type, manifold and gasket designs. The flow fields are often based on flow channels machined into graphite (generally a composite graphite is used) or consist of corrosion resistant alloy bipolar plates; a flow field can also be a corrosion resistant metal foam or a stamped flow pattern in a metal plate. The machined graphite flow field can be a simple design of dots, parallel channels, serpentine or interdigitated design [2]. The flow configuration may be cross-flow, co-flow or counter-flow. All these

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

138

A. S. Aricò, V. Baglio and V. Antonucci

aspects influence significantly mass transport and thermal management, by favoring diffusion or forced convection of the reactants to the catalytic sites and heat removal. The PEMFC stacks often contain cooling cells in between active (MEA equipped) cells. A coolant (e.g. demineralised water) can be used for the heat exchange for operation below 100 °C. Alternatively, an excess of air flow at the cathode can be used for stack cooling. In the case of DMFCs, the approach of the cooling cells is not strictly required because an efficient heat removal can be obtained by increasing the recirculation rate of the liquid mixture of water and methanol at the anode and using an external radiator. This approach is suitable for transportation and it can allow to increase the compactness of the module. It was pointed out in the previous sub-section that the increasing levels of performance that have recently been reached in DMFCs are often associated with a reduction of kinetic and ohmic limitations [9]; however, it is also of significant interest a reduction of the mass-transport limitations. In this regard, significant progress has been obtained by improving the characteristics of the electrode backing layer in terms of composition and thickness. Some investigations have focused on design of the reactant flow fields [11]. The most widely employed flow field in advanced fuel cells is based on the serpentine configuration. In such a configuration, the reactant is constrained to flow in a zig-zag way along parallel channels which are machined in a graphite plate in contact with the electrode backing layer. Such a design has often been adopted for stacked cells. The reactant molecules have access to the electrocatalytic sites through diffusion across the so-called diffusion layer, i.e. the backing layer, made of carbon cloth and carbon black, hydrophobized by appropriate addition of polytetrafluoroethylene (PTFE). A different approach for the flow of reactants and products within the electrode structure, i.e. an interdigitated design, was proposed by Nguyen [250] and Wilson et al. [124] for H2-O2 solid polymer electrolyte fuel cells (SPEFCs). In practice, the reactant gases are forced to enter the electrode pores and exit from them under a gradient pressure by making the inlet and outlet channels deadended. As pointed out by Nguyen [250], the flow through the electrode, in the presence of the interdigitated design, is no more governed by diffusion but becomes convective in nature. This particular design was selected for H2-air SPE fuel cells in order to avoid the water flooding problem at the cathode and to facilitate the removal of inert nitrogen molecules, which accumulate in the pores of the active layer and consequently produces a diffusion barrier [250]. The forced-flow-through characteristics created by the interdigitated flow fields in SPEFCs have been also investigated for DMFCs [11]. In general, it has been shown that enhanced mass transfer characteristics are achieved with the

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Technology Development

139

interdigitated flow field in DMFCs but these beneficial effects are especially observed only at high current densities, corresponding to low values of cell potential. At high cell potentials, i.e. under practical operation conditions (above 0.5 V), a higher efficiency for fuel utilization is obtained with the classical serpentine flow fields due to the smaller effect on methanol cross-over. This problem does not affect the cathode, thus an optimised condition may be represented by a serpentine flow field at the anode and an interdigitated flow field at the cathode. In the fuel cell stacks of significant size, graphite bipolar plates are being replaced with the more economic carbon based composite materials or by metallic foams [251, 252]. With composite materials, the same design of graphite plates may be reproduced, whereas, the metallic foam conceptually operates under conditions similar to serpentine flow fields where the reactant distribution over the electrocatalyst layer is controlled by diffusion. Alternative stack designs have been investigated by Scott and co-workers [253]. These authors have analyzed the possibility of using more open structures at the anode (e.g., dots or open channels) to favor the diffusion of methanol. In other cases, the parallel flow channels pattern has been preferred, due to an optimal combination of simplicity of design and suitable performance. Graphite or carbon composite based bipolar plates exhibit minimal corrosion. In the case of stainless-steel or metallic alloys-based materials, an appropriate evaluation of the chemical and electrochemical stability in the presence of hot methanol/water mixtures is necessary. In some cases surface treatments or special alloys are required to minimize corrosion. The DMFC stacks for portable applications may have different architectures [4, 5] with respect to the classical stack configuration for stationary and transportation applications. This especially occurs if the power output requirements are smaller than 50-100 W and a passive mode operation is required. Several configurations have been proposed for the passive DMFC stacks; the most common are the bi-cell type and monopolar-type [5]. In the bi-cell type, the methanol tank is allocated in between two anodes which belong to two different cells and the cathodes of these two cells are exposed to air [5]. Bi-cell units are grouped in a stack by leaving a gap between two cathodes belonging to two different bi-cells; series connection of the various bi-cells is made externally. In the monopolar configuration all electrodes of the same type, e.g. all anodes, are allocated on one face of the membrane and the cathodes on the other face. Each couple of electrodes forms a cell; the membrane is the same for all the cells. Series connection between two cells is realized by an electric conductor passing through the membrane or by an external circuit [4, 5].

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

140

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

A planar architecture is often used for µDMFC stacks [254]. As example, a catalysed membrane integrated on a silicon or polymeric matrix by using micromachining processes has emerged recently as a possible way to fabricate miniaturized DMFCs [255, 256]. Thanks to the integrated-circuit (IC) fabrication technology [257], micro-channel patterns of µDMFCs bipolar plates, into which reactants are fed, can be featured on Si or polymeric matrix with high resolution and good repeatability. Basically, the µDMFC has a conventional single cell structure, where a membrane electrode assembly (MEA) is sandwiched between two current collectors, made of gold, with fuel/air channels. These designs take advantage of the full wafer-level process capability. Alternatively, micro-channels can be realized on a polymeric substrate such as polycarbonate by a mechanical erosion with a numerical control mechanical device [258]. Micropumps can be used for fuel delivery in µDMFC stacks. For the passive mode operation, several approaches can be used for methanol feed to the anode. These have been recently reviewed by Qian et al [5]. Such approaches are based on natural circulation [259], capillary action [260] or self-pressurization using a controlled three-way valve [261]. It is appropriate to use a non-diluted fuel tank and control methanol and water feed by valves, metering, orifices or pumps. Water should be recovered from the cathode e.g. by favoring back-diffusion through the membrane from the cathode.

6.4. DMFC SYSTEMS The DMFC stack plant is generally designed on the basis of the power output level and the desired application. A few examples of DMFC systems are reported below. Figure 36 shows a typical flow-sheet of a DMFC stack plant useful for stationary and transportation applications [262]. Although a rough estimation of fuel consumption in this system can be derived by integration as a function of time of the stack operating current and the cross-over equivalent current density, an on line sensor of methanol concentration in the inlet fuel stream allows a better dynamic response of the device [262]. A CO2 sensor on the cathode outlet stream allows an evaluation of the fuel lost in the parasitic reaction at the cathode, and thus of the fuel efficiency under different operation conditions. As discussed in a previous section, this method does not take into account the CO2 permeation through the membrane from the anode, that may be significant at high current densities and in the presence of thin membranes [2]. An interesting DMFC system design has been proposed in a recent European project called Morepower [21].

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Technology Development

141

Figure 36. Flow-sheet of the DMFC stack plant equipped with an on-line methanol concentration sensor in the anode feed and a carbon dioxide detector in the cathode exhaust stream. Reprinted from Ref. [2] with permission from Wiley-VCH.

The project regarded the development of a low-cost, low temperature, portable DMFC system, nominal power 250 W, of compact construction and modular design for the potential market area of weather stations, medical devices, signal units, gas sensors and security cameras. The system was designed by the Institut fur Microtechnik of Mainz (Germany) and a modelling was carried out by Specchia et al. at the Politecnico of Turin (Italy) to evaluate heat, mass fluxes and pressure drops, for the integration and optimisation of the DMFC components in a portable Auxiliary Power Unit [263]. The system design and components are reported in Figure 37 [264]. These consist of the DMFC stack, the radiator (E-201) to cool the fuel solution downstream the DMFC anode, the gas–liquid separator (S-201, an atmospheric adiabatic flash unit) to dump up the produced CO2, the catalytic burner (R-401) to burn the residual MeOH vapor before releasing the anode exhausts in the atmosphere, the pump (P-201) to feed the fuel solution to the DMFC anode, the MeOH cartridge (V-201) to feed fresh MeOH into the system,

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

142

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

the water condenser (E-101) to recover and make-up the water lost during operation, the blower (B-101) to feed the fresh air necessary to the cathode reactions. The addition of fresh feed solution from the MeOH cartridge (V-201) to the exhaust solution is controlled via a MeOH sensor (I-201) [264, 265]; the controlled composition feed solution is then pumped into the DMFC, where the overall electrochemical reactions between the fuel and air produce power and heat. All the systems and sub-components necessary for the start-up are also present in the DMFC system. A small fraction of pure MeOH, taken directly from the MeOH cartridge (V-201) via a dedicated pump (P-501), is fed to an evaporator (E-501, electrically heated during the initial phase of the start-up procedure). The obtained MeOH vapour is then burned into a burner (R-501) with fresh air (B103); the produced flue gas is used to heat-up the solution to be fed into the DMFC in the start-up heat exchanger (E-502).

Figure 37. DMFC process scheme developed in the framework of the MOREPOWER project. Reprinted from Ref. [263] with permission from Elsevier.

This system design may appear more complex than that previously presented in Figure 36; however, the degree of complexity is usually determined by the required characteristics in terms of system control and rapid start-up and shutdown. As example, if a self-start up is preferred with respect to the rapid start-up,

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Technology Development

143

the burner (R-501) with its associated auxiliaries i.e. evaporator (E-501), pump (P-501), blower (B-103) and relative valves in Figure 37 can be removed from the scheme. A self start-up is usually accomplished by allowing the stack to operate at low voltages and at current as high as possible up to reach the appropriate temperature which allows normal operation. The low stack voltage is usually associated to a low electrical efficiency and large heat release. For what concerns cold start-up, a concentrated methanol solution e.g. 10 M remains liquid even at several degrees below zero. A low stack voltage corresponds to a high anode half cell potential. Thus, a quite stable PtRu catalyst is necessary to operate under these conditions especially if the start-up and shut-down cycles are frequent. It has been observed that Ru can dissolve into the membrane and migrate to the cathode [43]. Of course the low temperature condition associated to the start-up mitigates the corrosion problems. For what concerns the CO2 escape, the use of a highly selective hydrophobic membrane can allow to remove the catalytic burner (R-401) with the associated air supply (B-102) in Figure 37; furthermore, the radiator (E201) may be more compact if no loss of water/methanol vapor occurs through the selective membrane even at high temperature. In such a case, there is no need to cool down significantly the unreacted fuel mixture. Accordingly, it would not be necessary to spend much energy in the E-502 pre-heater. The heat released from the stack should be properly used to heat-up the methanol solution fed to the anode. The Morepower DMFC system was designed to operate at a maximum temperature of 60 °C [75]. For transportation applications, a high operating temperature is desired. DMFC systems working at high temperature e.g. 130 °C require special auxiliaries (pumps, tanks, sensors etc.) as well as fittings, valves, tubes capable of operation in this temperature range, unless a large radiator is used to cool down the unreacted anode solution that is recirculated through the preheater [81]. It is more appropraite to use most of the heat released from the stack into the pre-heater instead of dissipating a large part of it through the radiator. The device necessary for recovering water from the cathode condenser to the anode, may be quite compact if part of the water permeates or back-diffuses from a highly hydrophobic cathode to the anode through a membrane containing proper hydrophilic channels. A simple passive methanol fuel cell stack usually does not meed auxiliaries; on the other hand, miniaturised DMFCs may require some auxiliaries such as micropumps etc. Miniaturised systems can also be quite complex. DMFC stacks and systems for portable uses have been recently reviewed [5]. Of course, a simplification of the systems allows to reduce the production costs. In some cases, the proper development of materials for MEAs and auxiliaries may aid to simply the DMFC systems.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Chapter 7

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

7. TECHNO-ECONOMICAL CHALLENGES The most challenging problem for the development of DMFCs has been, and still is, a significant enhancement of electrocatalytic activities for the 6-electron transfer electro-oxidation of methanol. On the other hand, research in this area has enlightened many scientists and engineers to use highly sophisticated electrochemical surface science and material science techniques for unraveling the mysteries of the reaction path, rate determining steps and physicochemical characteristics (electronic and geometric factors, adsorption/desorption energies and electrocatalyst/support interaction) which influence the activities of the various types of electrocatalysts. The sluggishness of the reaction, especially in the presence of protonic electrolytes, is caused by the very strong chemical adsorption of CO-type species on an electrocatalyst subsequent to the dissociative adsorption of methanol (Pt is the best known electrocatalyst for this step). A neighboring chemisorbed labile OH species is vital for the electrooxidation of the strongly adsorbed CO species. To date, a Pt-Ru electrocatalyst (50:50 at. wt %) has shown the best results. There are some promotional effects by the presence of elements such as Sn, Mo, W, Os, as well as some refractory metal oxides (WO3). To date, there has been little success with alternatives to Pt and its alloys in proton conductive electrolytes; those tested include transition metal alloys, oxides and tungsten bronzes (oxide doped with sodium, tungsten carbide). One achievement has been in using carbon-supported electrocatalysts, which has helped to reduce the Pt loading by about a factor of two to four. The reation rates are larger in alkaline environments with respect to protonic electrolytes. This fact and the lower corrosion constraints in alkaline media allow to replace Pt with non-precious metal catalysts e.g. Ni. Alternatively, PtRu can be used in an alkaline electrolyte to take advantage of the lower overpotentials. However, no significant enhancement in terms of power density has been

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

146

A. S. Aricò, V. Baglio and V. Antonucci

achieved because this kinetic advantage is counteracted by the carbonation drawback and reduced ionic conductivity unless concentrated alkaline solutions are used. The performance of the oxygen reduction reaction with a platinum electrocatalyst is affected by the cross-over of methanol from the anode to the cathode through the ion exchange membrane. Firstly, the open circuit potential is reduced by about 200 mV and the second effect is due to the competitive adsorption of dissociated methanol and oxygen species. At present, there is no clear evidence of catalytic enhancement in oxygen reduction for alloys of Pt with Cr, Co and Ni in the presence of methanol cross-over. Non-platinum electrocatalysts, such as heat-treated phthalocyanines and porphyrins, as well as transition metals chalcogenides, have some chance of methanol tolerance but have considerably lower activities than platinum and also raise questions of stability. The near term prospects of replacing platinum as an electrocatalyst is very slim but a great challenge is to reduce the noble metal loading in both electrodes by a factor of about 10 to reduce its cost to about $10/kW. If an anion exchange membrane is used instead of a protonic electrolyte, Ptbased cathode electrocatalysts can be replaced by silver or MnO2 which are much less expensive and methanol tolerant. Although the oxygen reduction in alkaline media is faster than in acidic electrolyte, the performance enhancement achieved with anion exchange membranes is quite limited due to the absence of a suitable ionomer to extend the triple-phase boundary in the electrode bulk and the low anionic conductivity. As discussed for the anode, a potential solution can be to ricirculate KOH solution through the device, yet, this approach is affected by technical problems, e.g. carbonation, need to frequently regenerate the electrolyte, electrolyte leakage through the electrodes, which have limited the development of alkaline devices for several decades. The perfluorosulfonic acid polymer electrolyte in the DMFC is an equally expensive material (about $300/kW, based on the state-of-the-art performance). There has been a lot of research on alternative proton conducting membranes which allow CO2 rejection (sulfonated polyetherketone, polyether sulfone, radiation grafted polystyrene, zeolites, electrolytes doped with heteropolyacids and sulfonated polybenzimidazole), but, it is still a challenge to attain sufficiently high specific conductivity and stability in the DMFC environment. Nafion-based composite membranes with silicon oxide and zirconium hydrogen phosphate have shown beneficial effects on operation up to about 150 ºC with enhanced performance (lower activation and ohmic overpotentials); these can also suitably operate at ambient conditions with reduced cross-over due to an increase of the tortuosity factor.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Techno-Economical Challenges

147

Alternatively, new emphasis has been addressed recently to the anion exchange membranes. Both anodic and cathodic reaction rates are enhanced in alkaline media. Yet, the kinetic advantage is counteracted by a thermodynamic loss due to the presence of a pH gradient between the anode and the cathode. This is caused by the carbonation process occurring at the anode. This drawback can be overcome by recurculating a KOH or carbonate solution through the device, but, several technical problems arise under these conditions i.e. precipitation of carbonate on the electrode pores, need to frequently regenerate the electrolyte etc. Other drawbacks of anion exchange membranes concern with the low anionic conductivity (about five time lower than Nafion at low temperatures), larger activation energy for ion conduction than Nafion, no proper ionomer solution to enable an extension of the three-phase reaction zone in the electrode bulk, reduced stability at high temperature. Some of these problems can be suitably solved by improving the characteristics of the anionic polymer electrolyte. Regarding the methanol cross-over there is no contribution from the elctro-osmotic drag with anion exchange membranes. Yet, as well known, most of the methanol cross-over is due to the concentration gradient between the anode and the cathode and the hydrophilic properties of the present membranes. A critical area to improve overall cell performance is the fabrication of MEAs. Progress on preparation of high performance MEAs has been made by preparing thin electrocatalyst layers (about 10 µm thick) composed of the electrocatalyst and ionomer in the electrode substrate or directly deposited onto the membrane (CCM). Problems caused by barrier layer effects of nitrogen for access of oxygen to the catalytically active sites and electrode flooding need further investigations. Possible solutions to these problems are heat treatments of the recast Nafion gel in the electrocatalytic layer to make it hydrophobic or to use pore formers to increase porosity. Direct methanol single cell development in the last decade has allowed to achieve very interesting performances. Maximum power densities of about 500 mW cm-2 and 300 mW cm-2 under oxygen and air feed operation, respectively, and 200 mW cm-2 at a cell potential of 0.5 V have been reported for cells operating at temperatures close or above 100 °C under pressurized conditions, with Pt loadings of 1–2 mg cm-2. At ambient temperature in the presence of passive mode operation the power density ranges between 10 and 40 mW cm-2. The development of DMFC stacks for both transportation and portable applications has gained momentum in the last two-three years. The rated power output of the DMFC stack varies from a few watts in the case of portable power sources up to a few kW for remote power generator and hybrid battery-fuel cell vehicles. The best results achieved with DMFC stacks for electrotraction are 1

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

148

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

kW/l power density with an overall efficiency of 37 % at 0.5 V/cell. These performances make the DMFCs quite competitive with respect to the reformerH2/air SPE fuel cell, especially if one considers the complexity of the whole system; yet, the Pt loadings are still high in the DMFCs (around 1-2 mg cm-2). Reducing the loading of noble metals or using cheap non-noble metal catalysts is actually one of the breakthroughs which may allow the DMFC to increase its competitiveness on the market of power sources. In the near-term the high energy density of DMFCs and the recent advances in the technology of the realization of miniaturized fuel cells make these systems attractive to replace the current Li-based batteries in cellular phones, lap top computers and other portable systems. This field appears the most promising for the near-term and a successful utilization of such systems; the progress made in manufacturing DMFCs for portable systems may also stimulate new concepts and designs which may aid the further development of these systems for electrotraction.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Chapter 8

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

CONCLUSION It is widely recognized that to reduce greenhouse gases and obey recent environmental laws it is necessary to develop highly efficient and low-cost energy conversion systems. Direct methanol fuel cells possess good potentialities in this regard due to intrinsically low polluting emissions and system simplicity. In general, liquid-fueled fuel cells are a promising alternative to hydrogen fueled devices as electrochemical power sources for application in portable technology and in electric cars. Furthermore, the existing infrastructure for liquid fuel supply and distribution can be used for methanol, thus, reducing the time gap between development and commercialization. Recent results on DMFC stacks in terms of power density output (≈1 kW/l) and overall conversion efficiency (37% at 0.5 V per cell) indicate that these systems are quite competitive with respect to the reformer-H2/air PEMFC units for application in electrotraction as well as in distributed power generation. Yet, significant progress is necessary to further decrease the gap that still exists with respect to conventional power generation systems in terms of power density and costs. The major hurdles concern with the reduction of noble metal loading, methanol cross-over drawbacks and fabrication costs. At present, the most appealing application for DMFCs is in the field of portable power sources where device costs are less critical and power densities are close to those of Li-batteries. Table 7 summarizes the main drawbacks of DMFCs together with some potential solutions. Unfortunately, several proposed approaches cause the occurrence of new drawbacks. A few examples are presented in the following. The increase of operating temperature to enhance the reaction kinetics causes membrane dehydration with most of the conventional membranes. This results in a signinficant increase of the ohmic constraints. On the other hand, the membranes which allow high temperature operation such as phosphoric acid

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

150

A. S. Aricò, V. Baglio and V. Antonucci

doped polybenzoimidazole do not appear appropriate in terms of suitable conductivity under low temperature operation as required for portable power sources. Table 7. Drawbacks and potential solutions of DMFC devices Drawback Low power density

Potential solution Enhance oxidation kinetics

Improve electrode performance

Fuel cross-over

Membranes impermeable to methanol

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Methanol tolerant oxygen reduction catalysts

High cost

Reduce noble metal loading

Membranes alternative to Nafion

Present approach -Multifunctional catalysts -Increase the operating temperature and pH -Highly dispersed catalysts -Thin film electrodes -Optimization of the MEA -Anion exchange membranes -Composite membranes -Polyarylsulfonic membranes -Polyvinyl alcohol treated membranes -Chevrel-phase type (Mo4Ru2Se8), transition metal sulfides (MoxRuySz, MoxRhySz) or other transition metal chalcogenides -Pt-alloys -Non-noble metal catalysts (anode and cathode) in conjunction with alkaline electrolytes - Oxide catalysts -Cathode catalysts based on iron or cobalt organic macrocycles (phenylporphyrins, phthalocyanines) -Cobalt polypyrrolecarbon composite catalysts (CoPPY-C) -Decoration (anode catalyst) -Anion exchange membranes -Grafted membranes -SPEEK, SPSf, etc.

The use of non-noble metal catalysts is presently possible in DMFCs in the presence of alkaline electrolytes. However, low conductivity, carbonation and thermodynamic constraints limit the practical applications of this approach. The

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Conclusion

151

methanol tolerant cathode catalysts such as Chevrel-phase type or transition metal chalcogenides do not allow the oxidation to CO2 of the methanol permeated through the membrane; the cathode outlet will thus contain traces of unreacted methanol that can not be released in the atmosphere. This will require modification in the system and/or cell concept. Reduction of the catalyst layer thickness to reduce mass transport constraints can be achieved by increasing the concentration of the active phase on the support. However, this approach reduces catalyst utilization. The use of highly hydrophobic cathodes to favour oxygen transport and reduce flooding by the water, permeated through the membrane or formed by the reaction, increases the resistance and reduces the interface between catalyst and ionomer (triple phase boundary). These examples show that there are no univocal and radical solutions and a compromise is often necesary to enhance the device characteristics. Furthermore, it also appears that the materials should be tailored for the specific applications. A chemical and dimensional stable electrolyte with high conductivity in a wide temperature range would be more appropriate if the conduction mechanism is not assisted by water. Methanol cross-over is often associated to water permeation; these effects cause cathode poisoning and flooding. From a practical point of view, a carbon dioxide rejecting electrolyte appears more appropriate but new efforts should be addressed to the development of multifunctional catalysts with reduced noble metal loadings. Significant progress in the materials development would be also beneficial to reduce system complexity. The applications of DMFC in portable power sources cover the spectrum of cellular phones, personal organizers, laptop computers, military back power packs, etc. The infusion of semiconductor technology into the development of micro and mini fuel cells by leading organizations such as LANL, JPL, Motorola, has provided an awakening of DMFCs replacing the most advanced type of rechargeable batteries, i.e., lithium ion. For several of these applications, a DMFC working at room temperature and ambient pressure with an efficiency of only about 20% may be sufficient to have a strikingly higher performance than the lithium ion batteries, in respect to operating hours between refueling/recharging because of the high energy density of methanol. Further, the refueling in the case of DMFCs is instantaneous, whereas it requires about 3-5 hrs for lithium ion batteries. There is still a challenge in reducing the weight and volume of the DMFC to a level competitive with lithium ion batteries, as needed for say the cellular phone and laptop applications. What is most attractive in the portable power applications, as compared with the transportation and stationary applications is that the cost per kW or cost per kWh could be higher by a factor of

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

152

A. S. Aricò, V. Baglio and V. Antonucci

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

10 to 100. For this application, there is hardly any competition for lithium ion and DMFCs from any other type of power source. The present analysis indicates that the targets for each application may be achieved through a thoughtful development of materials, device design as well as through an appropriate choice of operating conditions.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

ACKNOWLEDGMENTS We acknowledge the financial support for the DMFC activity from the European Community through the Nemecel (EU Joule), Dreamcar (EU FP5) and Morepower (EU FP6) projects, from Regione Piemonte through the Microcell project and from Pirelli Labs., Solvay-Solexis, De Nora and Nuvera through several contracts. In particular, we would thank the project leaders of these contracts M. Dupont, M. Straumann, H. Hutchinson, G. Bollito, S. Nunes, G. Saracco, S. Specchia, P. Caracino, A. Tavares, A. Ghielmi, R. Ornelas and E. Ramunni. We express our gratitude to our colleagues that have collaborated to the DMFC activity at CNR-ITAE; in particular, A.K. Shukla, H. Kim, S. Srinivasan, C. Yang, R. Dillon, K.M. El-Khatib, Z. Poltarzewski, A.M. Castro Luna, G. Garcia, L.G. Arriaga, I. Nicotera and P. L. Antonucci. We are indebted with our collaborators C. D‘Urso, A. Stassi, A. Di Blasi, S. Siracusano, T. Denaro, F.V. Matera, E. Modica, G. Monforte, P. Cretì for their invaluable contribution.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved. Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

REFERENCES [1] [2] [3]

[4]

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

[5] [6] [7] [8] [9] [10] [11] [12]

[13]

Wasmus, S.; Kuver, A. J. Electroanal. Chemistry 1999, 461, 14-31. Arico, A.S.; Srinivasan, S.; Antonucci, V. Fuel Cells 2001, 1, 2, 133-161. Lamy, H.; Léger, J.-M.; Srinivasan, S. in Direct methanol fuel cells––from a 20th century electrochemists’ dream to a 21st century emerging technology; Bockris, J.O‘M.; Conway, B.E.; Eds.; Modern Aspects of Electrochemistry, Plenum Press, New York, NY, 2000; 34, Chapter 3, 53. Dillon, R.; Srinivasan, S.; Aricò, A.S.; Antonucci, V. Journal of Power Sources 2004, 127, 1-2, 112-126. Qian, W.; Wilkinson, D. P.; Shen, J.; Wang, H.; Zhang, J. Journal of Power Sources 2006, 154, 1, 202-213. Bockris, J. O‘M.; Srinivasan, S. Fuel Cells: Their Electrochemistry; McGraw- Hill Book Company, New York 1969. Shukla, A. K.; Aricò, A. S.; Antonucci, V. Renewable and Sustainable Energy Reviews 2000, 5, 137-155. McNicol, B. D.; Rand, D. A. J.; Williams, K. R.. J. Power Sources 1999, 83, 1-2, 15-31. Ren, X.; Wilson, M.S.; Gottesfeld, S. J. Electrochem. Soc. 1996, 143, 1, L12-L15. Aricò, A.S.; Creti, P.; Kim, H.; Mantegna, R.; Giordano, N.; Antonucci, V. J. Electrochem. Soc. 1996, 143, 3950-3959. Aricò, A.S.; Cretì, P.; Baglio, V.; Modica, E.; Antonucci, V. J. Power Sources 2000 91, 2, 202-209. Moore, R.M.; Gottesfeld, S.; Zelenay, P. in Proton Conducting Membrane Fuel Cells - Second International Symposium; Gottesfeld, S.; Fuller T. F.; Eds.; The Electrochemical Society, Pennington, NJ, 1999; 98-27, 365-379. Narayanan, S.R.; Chun, W.; Valdez, T.I.; Jeffries-Nakamura, B.; Frank, H.; Surampudi, S.; Halpert, G.; Kosek, J.; Cropley, C.; LaConti, A.B.; Smart,

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

156

[14] [15] [16] [17] [18] [19] [20] [21]

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

[22]

[23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36]

A. S. Aricò, V. Baglio and V. Antonucci M.; Wang, Q.; Surya Prakash, G.; Olah, G.A. Program and Abstracts, Fuel Cell Seminar 1996, 525 - 528. Baldauf , M.; Preidel, W.. J. Power Sources 1999, 84, 2, 161-166. Shukla, A.K.; Christensen, P.A.; Hamnett, A.; Hogarth, M.P. J. Power Sources, 1995, 55, 1, 87-91. Scott, K.; Taama, W.M.; Argyropoulos, P.; Sundmacher, K. J. Power Sources, 1999, 83, 1-2, 204-216. Jung, D.H.; Lee, C.H.; Kim, C.S.; Shin, D.R. J. Power Sources 1998, 71, 12, 169-173. Ravikumar, M.K.; Shukla, A.K. J. Electrochem. Soc. 1996, 143, 8, 26012606. Srinivasan, S.; Mosdale, R.; Stevens, P.; Yang, C. Annu. Rev. Energy Environ., 1999, 24, 281-285. Aricò, A. S.; Antonucci, V.; Alderucci, V.; Modica, E.; Giordano, N. J. Appl. Electrochem. 1993, 23, 11, 1107-1116. EU funded project MOREPOWER (compact direct methanol fuel cells for portable applications), project nr. SES6-CT-2003-502652 (2004). Carlstrom, C.; Craft, J.; Fannon, M.; Manning, M.; Marvin, R.; Modi, A.; Reichard, J.; Scartozzi, P.; Dolan, G.; Sievers, B. Hydrogen, Fuel Cells and Infrastructure Technologies, Annual Review, 2006, May 16-19. Apanel, G.; Johnson, E. Fuel Cells Bullettin 2004, 12-17. Kordesch, K.; Simader, G. Fuel Cells and their Applications, Wiley-VCH, Weinheim, 1996. Pavela, T.O. An Acad. Sci. Fennicae 1954, AII, 59, 7-11. Justi, E.W.; Winsel, A.W. (1955). Brit. Patent 821, 688. Cathro, K. J. J. Electrochem. Soc. 1969, 116, 11, 1608-1611. Janssen, M. M. P.; Moolhuysen, J. Electrochim. Acta 1976, 21, 11, 861-868. Watanabe, M.; Motoo, S. J. Electroanal. Chem. 1975, 60, 3, 275-283. Bagotzky, V. S.; Vassiliev, Yu. B. Electrochim. Acta 1967, 12, 9, 13231343. Shibata, M.; Motoo, S. J. Electroanal. Chem. 1985, 194, 2, 261-274. McNicol, B.D.; Short, R.T. J. Electroanal. Chem. 1977, 81, 2, 249-260. Parsons, R.; VanderNoot, T. J. Electroanal. Chem. 1998, 257, 1-2, 9-45. Aramata, A.; Kodera, T.; Masuda, M. J. Appl. Electrochem. 1988, 18, 4, 577-582. Beden, B.; Kardigan, F.; Lamy, C.; Leger, J. M. J. Electroanal. Chem. 1981, 127, 1-3, 75-85. Chandrasekaran, K.; Wass, J. C.; Bockris, J. O' M. J. Electrochem. Soc. 1990, 137, 2, 518-524.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

References

157

[37] Christensen, P. A.; Hamnett, A.; Troughton, G. L. J. Electroanal. Chem. 1993, 362, 1-2, 207-218. [38] Ticianelli, E.; Berry, J. G.; Paffet, M. T.; Gottesfeld, S. J. Electroanal. Chem. 1977, 81, 2, 229-238. [39] Goodenough, J. B.; Hamnett, A.; Kennedy, B. J.; Manoharan, R.; Weeks, S. J. Electroanal. Chem. 1988, 240, 1-2, 133-145. [40] Ravikumar, M.K.; Shukla, A.K. J. Electrochem. Soc. 1996, 143, 8, 26012606. [41] Mc Breen, J.; Mukerjee, S. J. Electrochem. Soc. 1995, 142, 10, 3399-3404. [42] Markovic‘, N.; Widelov, A.; Ross, P.N.; Monteiro, O.R.; Brown, I.G. Catal. Lett. 1997, 43, 1-2, 161-166. [43] Piela, P.; Eickes, C.; Brosha, E.; Garzon, F.; Zelenay, P. J. Electrochem. Soc. 2004, 151, A2053-A2059. [44] Gurau, B.; Viswanathan, R.; Liu, R.X.; Lafrenz, T.J.; Ley, K.L.; Smotkin, E.S.; Reddington, E.; Sapienza, E.A.; Chan, B.C.; Mallouk, T.E.; Sarangapani, S. J. Phys. Chem. B 1998, 102, 49, 9997-10003. [45] Fritts, S.D.; Sen, R.K. Assessment of Methanol Electrooxidation forDirect Methanol-Air Fuel Cells; DOE Report, Contract DE AC06-76 RLO 1830 by Battelle, 1988. [46] Cameron, D.S.; Hards, G.A.; Harrison, B.; Potter, R.J. Platinum Metals Rev. 1987, 31, 4, 173-181. [47] Beden, B.; Kadirgan, F.; Lamy, C.; Leger, J.-M. J. Electroanal. Chem. 1981, 127, 1-3, 75-85. [48] Hamnett, A.; Kennedy, B.J. Electrochimica Acta 1988, 33, 11, 1613-1618. [49] Miyazaki, E. J. Catalysis 1980, 65, 84-92. [50] Goodenough, J.B.; Hamnett, A.; Kennedy, B.J.; Manoharam, R.; Weeks, S.A. Electrochimica Acta 1990, 35, 1, 199-207. [51] Ichinose, O.; Kawaguchi, M.; Furuya, F. J Appl Electrochem. 2004, 34, 1, 55-59. [52] Alonso-Vante, N.; Tributsch, H. Nature 1986, 323, 431-432. [53] Sun, J.R.; Wang, J.T.; Savinell, R.F. J. Appl. Electrochem. 1998, 28, 10, 1087-1093. [54] Imaizunii, S.; Shimanoe, K.; Teraoka, Y.; Yamazoe, N. Electrochemical and Solid-State Letters 2005, 8, 6, A270-A272. [55] Srinivasan, S.; Dillon, R.; Krishnan, L.; Arico, A.S.; Antonucci, V.; Bocarsly, A.B.; Lee, W.J.; Hsueh, K.-L.; Lai, C.-C.; Peng, A. Fuel Cell Science, Engineering and Technology 2003, 529-536. [56] Costamagna, P.; Srinivasan, S. Journal of Power Sources 2001, 102, 1-2, 242-252.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

158 [57] [58] [59] [60] [61] [62]

[63] [64] [65] [66]

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

[67]

[68] [69] [70] [71] [72] [73] [74] [75]

A. S. Aricò, V. Baglio and V. Antonucci Wynn, J.E. Proc. Ann. Power Sources Conf. 1960, 14, 52-57. Murray, J.N.; Grimes, P.G. Fuel Cells, Am. Inst. Chem. Engrs. 1963, 57. Hunger, H.F. Proc. Ann. Power Sources Conf. 1960, 14, 55-59. Wang, Y.; Li, L.; Hu, L.; Zhuang, L.; Lu, J.; Xu, B. Electrochemistry Communications 2003, 5, 8, 662-666. Varcoe, J. R.; Slade, R. C. T. Fuel Cells 2004, 5, 2, 187-200. Pivovar B. S. in Membrane for Energy Conversion, Peinemann, K.-V.; Nunes, S.P.; Eds.; Wiley-VCH Verlag GmbH and Co. KGaA, Weinheim, 2008. Neburchilov, V.; Martin, J.; Wang, H.; Zhang, J. Journal of Power Sources 2007, 169, 2, 221-238. Aricò, A.S.; Baglio, V.; Di Blasi, A.; Antonucci, V.; Cirillo L.; Ghielmi, A.; Arcella, V. Desalination 2006, 199, 271-273. Peled, E.; Duvdevani, T.; Aharon, A.; Melman, A. Electrochem. Solid-State Lett. 2000, 3, 12, 525-528. Peled, E.; Livshits, V.; Rakhman, M.; Aharon, A.; Duvdevani, T.; Philosoph, M.; Feigling, T. Electrochem. Solid-State Lett. 2004, 7, A507A510. Aricò, A.S.; Baglio, V.; Antonucci, V. in Membrane for Energy Conversion, Peinemann, K.-V.; Nunes, S.P.; Eds.; Wiley-VCH Verlag GmbH and Co. KGaA, Weinheim, 2008.; 2, 123-168. Ren, X.; Zelenay, P.; Thomas, S.; Davey, J.; Gottesfeld, S. J. Power Sources 2000, 86, 1, 111-116. Wang, J.; Wasmus, S.; Savinell, R.F. J. Electrochem. Soc. 1995, 142, 42184223. Aricò, A.S.; Creti, P.; Antonucci, P.L.; Antonucci, V. Electrochem. SolidState Lett. 1998, 1, 66-68. Aricò, A.S.; Baglio, V.; Di Blasi, A.; Cretì, P.; Antonucci, P.L.; Antonucci, V. Solid State Ionics 2003, 161, 251-265. Ruffmann, B.; Silva, H.; Schulte, B.; Nunes, S. Solid State Ionics 2003, 162-163, 269-275. Aricò, A.S.; Baglio, V.; Di Blasi, A.; Antonucci, V. Electrochemistry Communications 2003, 5, 862-866. Lufrano, F.; Baglio, V.; Staiti, P.; Arico', A.S.; Antonucci, V. Desalination 2006, 199, 1-3, 283-285. Antonucci, V.; Aricò, A.S.; Baglio, V.; Brunea, J.; Buder, I.; Cabello, N.; Hogarth, M.; Martin, R.; Nunes, S. Desalination 2006, 200, 653-655.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

References

159

[76] Miyatake, K.; Watanabe, M. in in Membrane for Energy Conversion, Peinemann, K.-V.; Nunes, S.P.; Eds.; Wiley-VCH Verlag GmbH and Co. KGaA, Weinheim, 2008.; 2, 47-60. [77] Slade, R.C.T.; Varcoe, J.R. Solid State Ionics 2005, 176, 5-6, 585-597. [78] Yu, E. H.; Scott, K. Electrochemistry Communications 2004, 6, 4, 361-365. [79] Polyfuel website: http://www.polyfuel.com. [80] Aricò, A.S.; Baglio, V.; Modica, E.; Di Blasi, A.; Antonucci, V. Electrochemistry Communications 2004, 6, 164-169. [81] Final report of the EU funded project DREaMCAR (direct methanol fuel cell system for car applications), project nr. ERK6-CT-2000-00315 (2005). [82] Hockaday, R.G.; DeJohn, M.; Navas, C.; Turner, P.S.; Vaz, H.L.; Vazul, L.L. Proceedings of the Fuel Cell Seminar, Portland, Oregon, USA, 2000; 30 October–2 November, 791-794. [83] Kelley, S.C.; Deluga, G.A.; Smyrl, W.H. Electrochem. Solid-State Lett. 2000, 3, 9, 407-409. [84] Narayanan, S.R.; Valdez, T.I.; Clara, F. Proceedings of the Fuel Cell Seminar, Portland, OR, USA, 2000; 30 October–2 November, 795-798. [85] Jung, D.-H.; Jo, Y.-H.; Jung, J.-H.; Cho, S.-Y.; Kim, C.-S.; Shin, D.-R. Proceedings of the Fuel Cell Seminar, Portland, OR, USA, 2000; 30 October–2 November, 420-423. [86] Bostaph, J.; Koripella, R.; Fisher, A.; Zindel, D.; Hallmark, J. Proceedings of the 199th Meeting on Direct Methanol Fuel Cell, The Electrochemical Society, Washington, DC, USA, 2001; 25–29 March. [87] Hockaday, R.G. (1998). US Patent No. 5,759,712. [88] Witham, C.K.; Chun, W.; Valdez, T.I.; Narayanan, S.R. Electrochem. SolidState Lett. 2000, 3, 11, 497-500. [89] Gottesfeld, S.; Ren, X.; Zelenay, P.; Dinh, H.; Guyon, F.; Davey, J. Proceedings of the Fuel Cell Seminar, Portland, OR, USA, 2000; 30 October–2 November, 799-802. [90] Dohle, H.; Mergel, J.; Scharmaan, H.; Schmitz, H. Proceedings of the 199th Meeting Direct Methanol Fuel Cell Symposium, The Electrochemical Society, Washington, DC, USA, 2001; 25–29 March. [91] Chang, H. The Knowledge Foundation‘s Third Annual International Symposium on Small Fuel Cells and Battery Technologies for Portable Power Applications, Washington, DC, USA, 2001; 22–24 April. [92] Chang, H.; Kim, J.R.; Cho, J.H.; Kim, H.K.; Choi, K.H. Solid State Ionics, 2002, 148, 601-605.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

160

A. S. Aricò, V. Baglio and V. Antonucci

[93] Jung, D.H.; Jo, Y.-H.; Jung, J.-H.; Cho, S.-H.; Kim, C.-S.; Shin, D.-R. Proceedings Fuel Cell Seminar, Portland, OR, USA, 2000; 30 October–2 November, 420-423. [94] Lifton, R.F. The Knowledge Foundation‘s Third Annual International Symposium on Small Fuel Cells and Battery Technologies for Portable Power Applications, Washington, DC, USA, 2001; 22–24 April. [95] Martin, J.J.; Qian, W.; Wang, H.; Neburchilov, V., Zhang, J.; Wilkinson, D.P.;Chang, Z. J. Power Sources 2007, 164, 1, 287-292. [96] Guo, Z.; Faghri, H. J. Power Sources 2006, 160, 1183-1187. [97] Liu, J.G.; Zhao, T.S.; Chen, R.; Wong, C.W. Electrochem. Comm. 2005, 7, 288-293. [98] Chen, R.; Zhao, T.S. J. Power Sources 2007, 167, 2, 455-460. [99] Chen, R.; Zhao, T.S. Electrochem. Comm. 2007, 9, 4, 718-724. [100] Chen, R.; Zhao, T.S. Electrochim. Acta 2007, 52, 13, 4317-4324. [101] Zhang, H.F., Hsing, I.-M. J. Power Sources 2007, 167, 2, 450-454. [102] Abdelkareem, M.A.; Nakagawa, N. J. Power Sources 2007, 165, 2, 685691. [103] Chu, D.; Jiang, R. Electrochim. Acta 2006, 51, 26, 5829-5835. [104] Yang, H. ; Zhang, T.S.; Ye, Q. J. Power Sources 2005, 139, 1-2, 79-90. [105] Di Blasi, A.; Baglio, V.; Denaro, T.; Antonucci, V.; Aricò, A.S. Journal of New Materials for Electrochemical Systems 2008, 11, 165-174. [106] Lu, G.Q.; Wang, C.Y. J. Power Sources 2004, 134, 1, 33-40. [107] Lai, Q.-Z.; Yin, G.-P.; Zhang, J.; Wang, Z.-B.; Cai, K.-D.; Liu, P. J. Power Sources 2008, 175, 1, 458-463. [108] Blum, A.; Duvdevani, T.; Philosoph, M.; Rudoy, N.; Peled, E. J. Power Sources 2003, 117, 1-2, 22-25. [109] Kim, D.; Cho, E.A.; Hong, S.-A.; Oh, I.-H.; Ha, H.-Y. J. Power Sources 2004, 130, 1-2, 172-177. [110] Pan, Y.H. J. Power Sources 2006, 161, 1, 282-289. [111] Lu, G.Q.; Wang, C.Y.; Yen, T.J.; Zhang, X. Electrochim. Acta 2004, 49, 5, 821-828. [112] Motokawa, S.; Mohamedi, M.; Momma, T.; Shoji, S.; Osaka, T. Electrochem. Comm. 2004, 6, 6, 562-565. [113] Sabate, N.; Esquivel, J.P.; Santander, J.; Torres, N.; Gracia, I.; Ivanov, P.; Fonseca, L.; Figueras, E.; Canè, C. J. New Materials for Electrochem. Sys. 2008, 11, 2, 143-146. [114] Lim, S.W.; Kim, S.W.; Kim, J.; Ahn, J.E.; Han, H.S.; Shul, Y.G. J. Power Sources 2006, 161, 1, 27-33.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

References

161

[115] Baglio, V.; Stassi, A.; Matera, F.V.; Di Blasi, A.; Antonucci, V.; Aricò, A.S. J. Power Sources 2008, 180, 2, 797-802. [116] Harris, D. Ballard Power Systems Inc. News Release, 2000, 9 November. [117] Zhang, J.; Colbow, K.M.; Wilkinson, D.P. (2001). US Patent No. 6,187,467. [118] http://www.ird.dk/product.htm/. [119] Buttin, D.; Dupont, M.; Straumann, M.; Gille, R.; Dubois, J.C.; Ornelas, R.; Fleba, G.P.; Ramunni, E.; Antonucci, V.; Arico, A.S.; Creti, P.; Modica, E.; Pham-Thi, M.; Ganne, J.P. J. Appl. Electrochem. 2001, 31, 3, 275-279. [120] Baldauf, M.; Preidel, W. J. Power Sources 1999, 84, 2, 161-166. [121] Baldauf, M.; Frank, M.; Kaltschmidt, R.; Lager, W.; Luft, G.; Poppinger, M.; Preidel, W.; Seeg, H.; Tegeder, V.; Odgaard, M.; Ohlenschloeger, U.; Yde-Andersen, S.; Lindic, M.H.; Drews, T.; Nguyen, C.V.; Tachet, I.; Skou, E.; Engell, J.; Lundsgaard, J.; Hogarth, M.P.; Hards, G.A.; Kelsall, I.; Theobald, B.R.C.; Smith, E.; Thompsett, D.; Gunner, A.; Walsby, N. (1996–1999). Publishable Final Report of The European Commission Joule III Programme. [122] Baldauf, M.; Preidel, W. in Proceedings of the Third International Symposium on Electrocatalysis: Electrocatalysis in direct and indirect methanol PEM fuel cells; Lamy, C.; Wendt, H.; Eds.; Portoroz, Slovenia, 1999; 12–14 September. [123] Baldauf, M.; Preidel, W. J. Appl. Electrochem. 2001, 31, 7, 781-786. [124] Wilson, M.S.; Sringer, T.E.; Davey, J.R.; Gottesfeld, S. in Proton Conducting Membrane Fuel Cells I; Gottesfeld, S.; Halpert, G.; Langrebe, A.; Eds.; Proceedings of The Electrochemical Society, Pennington, NJ, 1995; 95-23, 115. [125] http://world.honda.com/news/. [126] Hamnett, A. Catalysis Today 1997, 38, 4, 445-457. [127] Hogarth, M. P.; Hards, G. A. Platinum Metal Rev. 1996, 40, 150-155. [128] Niedrach, L. W.; Zeliger, H.I. J. Electrochem. Soc. 1969, 116, 1, 152-153. [129] Okamoto, H.; Kawamura, G.; Ishikawa, A.; Kudo, T. J. Electrochem. Soc. 1987, 134, 7, 1653-1658. [130] Hays, C. C.; Manoharan, R.; Goodenough, J. B. J. Power Sources 1993, 45, 3, 291-301. [131] Antonucci, P.L.; Aricò, A. S.; Modica, E.; Antonucci, V. J. Solid State Electrochemistry 1999, 3, 4, 205-209. [132] Gotz, M.; Wendt, H. Electrochim. Acta 1998, 43, 24, 3637-3644. [133] Aricò, A. S.; Poltarzewski, Z.; Kim, H.; Morana, A.; Giordano, N.; Antonucci, V. J. Power Sources 1995, 55, 2, 159-166.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

162

A. S. Aricò, V. Baglio and V. Antonucci

[134] Lasch, K.; Jorissen, L.; Garche, J. in Proceedings of the Third International Symposium on Electrocatalysis: Electrocatalysis in direct and indirect methanol PEM fuel cells; Lamy, C.; Wendt, H.; Eds.; Portoroz, Slovenia, 1999; 12–14 September, 104-106. [135] Gasteiger, H. A.; Markovic, N.; Ross, P. N. Jr.; Cairns, E. J. J. Electrochem. Soc. 1994, 141, 7, 1795-1803. [136] Pletcher, D.; Solis, V. Electrochim. Acta 1982, 27, 6, 775-782. [137] Wang, K.; Gasteiger, H.A.; Markovic, N.M.; Ross, P.N.Jr. Electrochim. Acta, 1996, 41, 16, 2587-2593. [138] Méli, G.; Léger, J.-M.; Lamy, C.; Durand, R. J. Appl. Electrochem. 1993, 23, 3, 197-202. [139] Clavilier, J.; Lamy, C.; Leger, J.-M. J. Electroanal. Chem. 1981, 125, 1, 249-254. [140] Xia, X. H.; Iwasita, T.; Ge, F.; Vielstich, W. Electrochim. Acta 1996, 41, 5, 711-718. [141] Lampitt, R.A.; Carrette, L.P.L.; Hogarth, M.P.; Russell, A.E. J. Electroanal. Chem. 1999, 460, 1-2, 80-87. [142] Vielstich, W.; Christensen, P. A.; Weeks, S. A.; Hamnett, A. J. Electroanal. Chem. 1988, 242, 1-2, 327-333. [143] Frelink, T.; Visscher, W.; Van Veen, J.A.R. Surface Sci. 1995, 335, 1-2, 353-360. [144] Frelink, T.; Visscher, W.; Van Veen, J.A.R. Langmuir 1996, 12, 15, 37023708. [145] Iwasita, T.; Hoster, H.; John-Anacker, A.; Lin, W. F.; Vielstich, W. Langmuir 2000, 16, 2, 522-529. [146] Tong, Y. Y.; Rice, C.; Wieckowski, A.; Oldfield, E. J. Am. Chem. Soc. 2000, 122, 6, 1123-1129. [147] Gasteiger, H.A.; Markovic, N.; Ross, P. N.; Cairns, E.J.Jr. J. Phys. Chem. 1994, 98, 2, 617-625. [148] Friedrich, A. K.; Geyzers, K. P.; Linke, U.; Stimming, U.; Stumper, J. J. Electroanal. Chem. 1996, 402, 1-2, 123-128. [149] Schmidt, T. J.; Noeske, M.; Gasteiger, H. A.; Behm, R. J., Britz,P., Brijoux, W., Bonnemann, H. Langmuir 1997, 13, 10, X-2595. [150] Dinh, H. N.; Ren, X.; Garzon, F. H.; Zelenay, P.; Gottesfeld, S. J. Electroanal. Chem. 2000, 491, 1-2, 222-233. [151] Schmidt, T.J.; Gasteiger, H.A.; Behm, R.J. Electrochemistry Communications 1998, 1, 1-5. [152] Takasu, Y.; Fujiwara, T.; Murakami, Y.; Sasaki, K.; Oguri, M.; Asaki, T.; Sugimoto,W. J. Electrochem. Soc. 2000, 147, 12, 4421-4427.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

References

163

[153] Aricò, A.S.; Modica, E.; Passalacqua, E.; Antonucci, V.; Antonucci, P. L. J. Appl. Electrochem. 1997, 27, 11, 1275-1282. [154] Swathirajan, S.; Mikhail, Y.M. J. Electrochem. Soc. 1991, 138, 9, 26312635. [155] Aramata, A.; Masuda, M.; Kodera, T. J. Electrochem. Soc. 1989, 136, 11, 3288-3296. [156] Léger, J.-M. in Proceedings of the Third International Symposium on Electrocatalysis: Electrocatalysis in direct and indirect methanol PEM fuel cells; Lamy, C.; Wendt, H.; Eds.; Portoroz, Slovenia, 1999; 12–14 September; 9. [157] Inada, R.; Shimazu, K.; Kita, H. J. Electroanal. Chem. 1990, 277, 1-2, 315326. [158] Aramata, A.; Masuda, M. J. Electrochem. Soc. 1991, 138, 7, 1949-1957. [159] Shukla, A.K.; Ravikumar, M.K.; Aricò, A.S.; Antonucci, V.; Giordano, N.; Hamnett, A. J. Appl. Electrochem. 1995, 25, 6, 528-532. [160] Kita, H.; Gao, Y.; Nakato, T.; Hattori, H. J. Electroanal. Chem. 1994, 373, 1-2, 177-183. [161] Chrzanowski, W.; Wieckowski, A. Langmuir 1998, 14, 8, 1967-1970. [162] Kim, H.; Wieckowski, A. in Third International Symp. on New Materials for Fuel Cells and Modern Battery Systems; Savadogo, O.; Ed.; Montrèal, Canada, 1999; 125-126. [163] Liu, L.; Pu, C.; Viswanathan, R.; Fan, Q.; Liu, R.; Smotkin, E.S. Electrochim. Acta 1998, 43, 241, 3657-3663. [164] Aricò, A.S.; Shukla, A.K.; El-Khatib, K.M.; Creti, P.; Antonucci, V. J. Appl. Electrochem. 1999, 29, 6, 671-676. [165] Anderson, A. B.; Grantscharova, E.; Seong, S. J. Electrochem. Soc. 1996, 143, 6, 2075-2082. [166] Iwasita, T.; Nart, F.C.; Vielstich, W. Ber. Bunsenges Phys. Chem. 1990, 94, 9, 1030-1034. [167] Madey, T. E.; Engelhardt, H. A.; Menzel, D. Surf. Sci. 1975, 48, 2, 304-328. [168] Aricò, A. S.; Modica, E.; Cretì, P.; Antonucci, P.L.; Antonucci, V. J. New Mat. Electrochem. Systems 2000, 3, 207-220. [169] Aricò, A. S.; Baglio, V.; Cretì, P.; Antonucci, V. in Fuel Cell Seminar Abstracts 2000, October 30-November 2, Portland, Oregon, 75-78. [170] Chen, K.Y.; Sun, Z.; Tseung, A.C.C. Electrochem. Solid State Lett. 2000, 3, 1, 10-12. [171] Campbell, S. A.; Parsons, R. J. Chem. Soc. Faraday Trans. 1992, 88, 6, 833-841.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

164

A. S. Aricò, V. Baglio and V. Antonucci

[172] Watanabe, M.; Furuuchi, Y.; Motoo, S. J. Electroanal. Chem. 1985, 191, 2, 367-375. [173] Bittins-Cattaneo, B.; Iwasita, T. J. Electroanal. Chem. 1987, 238, 1-2, 151161. [174] Aricò, A.S.; Antonucci, V.; Giordano, N.; Shukla, A.K.; Ravikumar, M.K.; Roy, A.; Barman, S.R.; Sarma, D.D. J. Power Sources 1994, 50, 3, 295-309. [175] Shukla, A.K.; Aricò, A.S.; El-Khatib, K.M.; Kim, H.; Antonucci, P.L.; Antonucci, V. Appl. Surf. Sci. 1999, 137, 1-3, 20-29. [176] Mukerjee, S.; McBreen, J. in Proc. 2nd International Symp. on New Materials for Fuel Cells and Modern Battery Systems; Savadogo, O.; Roberge, P.R.; Eds.; Montréal, Canada, 1977; 548-559. [177] Ianniello, R.; Schmidt, V.M.; Stimming, U.; Stumper, J.; Wallau, A. Electrochim. Acta 1994, 39, 11-12, 1863-1869. [178] Iwasita, T.; Hoster, H.; John-Anacker, A.; Lin, W. F.; Vielstich, W. Langmuir 2000, 16, 2, 522-529. [179] Cotton, F.A.; Wilkinson, G. Advanced Inorganic Chemistry, WileyInterscience, New York, NY, 1980. [180] Frelink, T.; Visscher, W.; Van Veen, J.A.R. Electrochim. Acta 1994, 39, 1112, 1871-1875. [181] Rolison, D.R.; Hagans, P.L.; Swider, K.L.; Long, J.W. Langmuir 1999, 15, 3, 774-779. [182] Aricò, A.S.; Monforte, G.; Modica, E.; Antonucci, P.L.; Antonucci, V.. Electrochemistry Communications 2000, 2, 7, 466-470. [183] Hamnett, A.; Kennedy, B. J.; Wagner, F.E. J. Catal. 1990, 124, 1, 30-40. [184] Aricò, A.S.; Cretì, P.; Modica, E.; Monforte, G.; Baglio, V.; Antonucci, V. Electrochim. Acta 2000, 45, 25-26, 4319-4328. [185] Aricò, A.S.; Baglio, V.; Di Blasi, A.; Modica, E.; Antonucci, P.L.; Antonucci, V. Journal of Electroanalytical Chemistry 2003, 557, 167-176. [186] Brankovic, S.R.; Wang, J.X.; Adzic, R.R. Electrochemical and Solid-State Letters 2001, 4, 12, A217-A220. [187] Hibino, T.; Hashimoto, A.; Yano, M.; Suzuki, M.; Yoshida, S.; Sano, M. Journal of The Electrochemical Society 2002, 149, 2, A133-A136. [188] Arico‘, A.S.; Baglio, V.; Di Blasi, A.; Modica, E.; Monforte, G.; Antonucci, V. Journal of Electroanalytical Chemistry 2005, 576, 161-169. [189] Shen, P.; Chen, K.; Tseung, A.C.C. J. Chem. Soc. Faraday Trans. 1994, 90, 3089-3092. [190] Bashyam, R.; Zelenay, P. Nature 2006, 443, 7107, 63-66. [191] Tseung, A.C.C.; Shen, P.K.; Chen, K.Y. J. Power Sources 1996, 61, 1-2, 223-225.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

References

165

[192] Giordano, N.; Passalacqua, E.; Pino, L.; Aricò, A.S.; Antonucci, V.; Vivaldi, M.; Kinoshita, K. Electrochimica Acta 1991, 36, 13, 1979-1984. [193] Poirier, J.A. The Electrochemical Society Interface 1994, 141, 2, 425, 430. [194] Watanabe, M.; Saegusa, S.; Stonehart, P. J. Electroanal. Chem. 1989, 271, 1-2, 213-220. [195] Garcia, G.; Baglio, V.; Stassi, A.; Pastor, E.; Antonucci, V.; Aricò, A.S. Journal of Solid State Electrochemistry 2007, 11, 1229-1238. [196] Savadogo, O.J. Power Sources 2004, 127, 135-161. [197] Savadogo, O.Journal of New Materials for Electrochemical Systems 1998, 1, 47-66. [198] Nolte, R.; Ledjeff, K.; Bauer, M.; Mulhaupt, R. Journal of Membrane Science 1993, 83, 211-220. [199] Kerres, J.; Ullrich, A.; Meier, F.; Haring, T. Solid State Ionics 1999, 125, 243-249. [200] Alberti, G.; Casciola, M. Annu. Rev. Mater. Res. 2003, 33, 129-154. [201] Kreuer, K.D. J. Membrane Sci. 2001, 185, 29-39. [202] Watanabe, M.; Uchida, H.; Seki, Y.; Emori, M.; Stonehart, P. J. Electrochem. Soc. 1996, 143, 3847-3852. [203] Kreuer, K.D.; Paddison, S.J.; Spohr, E.; Schuster, M. Chemical Reviews 2004, 104, 4637-4678. [204] Kreuer, K.D. Solid State Ionics 1997, 97, 1-15. [205] Li, Q.; Jensen, J.O. in Membrane for Energy Conversion, Peinemann, K.V.; Nunes, S.P.; Eds.; Wiley-VCH Verlag GmbH and Co. KGaA, Weinheim, 2008; 2, 61-96. [206] Aricò, A.S.; Baglio, V.; Cretì, P.; Di Blasi, A.; Antonucci, V.; Brunea, J.; Chapotot, A.; Bozzi, A.; Schoemans, J. Journal of Power Sources 2003, 123, 2, 107-115. [207] Cleghorn, S.; Ren, X.; Thomas, S.; Gottesfeld, S. in Book of Extended Abstracts, ISE-ECS Joint Symposium, Paris, 1997, September, Abstract n. 182, 218-219. [208] Bauer, B.; Jones, D.J.; Roziere, J.; Tchicaya, L.; Alberti, G.; Casciola, M.; Massinelli, I.; Peraio, A.; Besse, S.; Ramunni, E. J. New Materials for Electrochemical Systems 2000, 3, 2, 93-98. [209] Hietala, S.; Koel, M.; Skou, E.; Elomaa, M.; Sundholm, F. J. Mater. Chem. 1998, 8, 5, 1127-1132. [210] Yang, C.; Srinivasan, S.; Aricò, A.S.; Cretì, P.; Baglio, V.; Antonucci, V. Electrochemical and Solid-State Letters 2001, 4, A31-A34. [211] Ogumi, Z.; Matsuoka, K.; Chiba, S.; Matsuoka, M.; Iriyama, Y.; Abe, T.; Inaba, M. Electrochemistry 2002, 70, 12, 980-983.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

166

A. S. Aricò, V. Baglio and V. Antonucci

[212] Jung, D.H.; Cho, S.Y.; Peck, D.H.; Shin, D.R.; Kim, J.S. J. Power Sources 2003, 118, 205-211. [213] Ren, X.; Sringer, T.E.; Gottesfeld, S. in Proton Conducting Membrane Fuel Cells - Second International Symposium, Gottesfeld, S.; Fuller, T. F.; Eds.; The Electrochemical Society, Pennington, New Jersey, 1999; 98-27, 341357. [214] Dohle, H.; Divisek, J.; Mergel, J.; Oetjen, H.F.; Zingler, C.; Stolten, D. in Fuel Cell Seminar Abstracts, 2000; October 30-November 2, Portland, Oregon, 126-129. [215] Paffet, M.T.; Beery, G.J.; Gottesfeld, S. J. Electrochem. Soc. 1988, 135, 6, 1431-1436. [216] Watanabe, M.; Tsurumi, K.; Mizukami, T.; Nakamura, T.; Stonehart, P. J. Electrochem. Soc. 1994, 141, 10, 2659-2668. [217] Freund, A.; Lang, J.; Lehman, T.; Starz, K.A. (1996). Cat. Today, 27, 1-2, 279-283. [218] Jalan, V.; Taylor, E.J. J. Electrochem. Soc. 1983, 130, 2299-2301. [219] Gottesfeld, S.; Paffett, M.T.; Redondo, A. J. Electroanal. Chem. 1986, 205, 163-168. [220] Beard, B.C.; Ross, P.N. J. Electrochem. Soc. 1990, 137, 3368-3374. [221] Kim, K.T.; Hwang, J.T.; Kim, Y.G.; Chung, J.S. J. Electrochem. Soc. 1993, 140, 1, 31-36. [222] Mukerjee, S.; Srinivasan, S.; Soriaga, M.P.; McBreen, J. J. Electrochem. Soc. 1995, 142, 5, 1409-1422. [223] Neergat, M.; Shukla, A.K.; Gandhi, K.S. J. Appl. Electrochem. 2001, 31, 4, 373-378. [224] Takasu, Y.; Iwazaki, T.; Sugimoto, W.; Murakami, Y. Electrochemistry Communications 2000, 2, 9, 671-674. [225] Aricò, A.S.; Shukla, A.K.; Kim, H.; Park, S.; Min, M.; Antonucci, V. Appl. Surf. Sci. 2001, 172, 1-2, 33-40. [226] Paulus, U.A.; Wokaun, A.; Scherer, G.G.; Schmidt, T.J.; Stamenkovic, V.; Radmilovic, V.; Markovic, N.M.; Ross, P.N. J. Phys.Chem. B 2002, 106, 16, 4181-4191. [227] Koffi, R.C. ; Coutanceau, C. ; Garnier, E. ; Leger, J.-M.; Lamy, C. Electrochim. Acta 2005, 50, 20, 4117-4127. [228] Xiong, L.; Manthiram, A. J. Electrochem. Soc. 2005, 152, 4, A697-A703. [229] Uchida, H.; Ozuka, H.; Watanabe, M. Electrochim. Acta 2002, 47, 22-23, 3629-3636. [230] Toda, T.; Igarashi, H.; Watanabe, M. J. Electroanal. Chem. 1999, 460, 1-2, 258-262.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

References

167

[231] Shukla, A.K.; Raman, R.K.; Choudhury, N.A.; Priolkar, K.R.; Sarode, P.R.; Emura, S.; Kumashiro, R. J. Electroanal. Chem. 2004, 563, 2, 181-190. [232] Zinola, C.F., Castro Luna, A.M., Triaca, W.E., Arvia, A.J. J. Applied Electrochem. 1994, 24, 2, 119-125. [233] Baglio, V.; Stassi, A.; Di Blasi, A.; D‘Urso, C.; Antonucci, V.; Aricò, A.S.Electrochimica Acta 2007, 53, 1360–1364. [234] Stassi, A.; D‘Urso, C.; Baglio, V.; Di Blasi, A.; Antonucci, V.; Aricò, A.S.; Castro Luna, A.M.; Bonesi, A.; Triaca, W.E. J. Appl. Electrochem. 2006, 36, 10, 1143-1149. [235] Min, M.; Cho, J.; Cho, K.; Kim, H. Electrochim. Acta 2000, 45, 25-26, 4211-4217. [236] Li, W.; Zhou, W.; Li, H.; Zhou, Z.; Zhou, B.; Sun, G.; Xin, Q. Electrochim. Acta 2004, 49, 7, 1045-1055. [237] Jasinski, R. J. Electrochem. Soc. 1965, 112, 526-530. [238] Franke, R.; Ohms, D.; Wiesener, K. J. Electroanal. Chem. 1989, 260, 1, 6373. [239] Faubert, G.; Lalande, G.; Coté, R.; Guay, D.; Dodelet, J.P.; Weng, L.T.; Bertrand, P.; Dénés, G. Electrochim. Acta 1996, 41, 10, 1689-1701. [240] Elzing, A.; Van der Putten, A.; Visscher, W.; Barendrecht, E. J. Electroanal. Chem. 1987, 233, 1-2, 113-123. [241] Reeve, R.W.; Christensen, P.A.; Hamnett, A.; Haydock, S.A.; Roy, S.C. J. Electrochem. Soc. 1998, 145, 10, 3463-3471. [242] Schmidt, T.J.; Paulus, U.A.; Gasteiger, H.A.; Alonso-Vante, N.; Behm, R.J. J. Electrochem. Soc. 2000, 147, 7, 2620-2624. [243] Aricò, A.S.; Creti, P.; Giordano, N.; Antonucci, V.; Antonucci, P.L.; Chuvilin, A. J. Appl. Electrochem. 1996, 26, 9, 959-967. [244] Wilson, M.S.; Gottesfeld, S. J. Appl. Electrochem. 1992, 22, 1, 1-7. [245] Shukla, A.K.; Stevens, P.; Hamnett, A.; Goodenough, J.B. J. Appl. Electrochem. 1989, 19, 3, 383-386. [246] Uchida, M.; Aoyama, Y.; Eda, N.; Ohta, A. J. Electrochem. Soc. 1995, 142, 2, 463-468. [247] Aricò, A.S.; Creti, P.; Antonucci, P.L.; Cho, J.; Kim, H.; Antonucci, V. Electrochim. Acta 1998, 43, 24, 3719-3729. [248] Fisher, A.; Jindra, J.; Wendt, H. J. Appl. Electrochem. 1998, 28, 3, 277-282. [249] Gamburzev, S.; Boyer, C.; Appleby, A.J. in Proton Conducting Membrane Fuel Cells II -Second International Symposium, Gottesfeld, S.; Fuller, T. F.; Eds.; The Electrochemical Society, Pennington, NJ, 1999; 98-27, 23-29. [250] Nguyen, T.V. J. Electrochem. Soc. 1996, 143, 5, L103-L105.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

168

A. S. Aricò, V. Baglio and V. Antonucci

[251] Busick, D.N.; Wilson, M. in Proton Conducting Membrane Fuel Cells II Second International Symposium, Gottesfeld, S.; Fuller, T. F.; Eds.; The Electrochemical Society, Pennington, NJ, 1999; 98-27, 435-445. [252] Zawodzinski, T.; Wilson, M.S.; Gottesfeld, S. in Proton Conducting Membrane Fuel Cells II -Second International Symposium, Gottesfeld, S.; Fuller, T. F.; Eds.; The Electrochemical Society, Pennington, NJ, 1999; 9827, 446-456. [253] Scott, K.; Taama, W.M.; Argyropoulos, P. J. Applied Electrochem. 1998, 28, 12, 1389-1397. [254] Min, K.B.; Tanaka, S.; Esashi, M. Electrochemistry 2002, 70, 12, 924–927. [255] D‘Arrigo, G.; Spinella, C.; Arena, G.; Lorenti, S. Material Science. Eng. 2003, 23, 13–18. [256] Yao, S.-C.; Tang, X.; Hsieh, C.-C.; Alyousef, Y.; Vladimer, M.; Fedder, G.K.; Amon, C.H. Energy 2006, 31, 5, 636–649. [257] D‘Arrigo, G.; Coffa, S.; Spinella, C. Sensors and Actuators 2006, 99, 112– 118. [258] Motokawa, S.; Mohamedi, M.; Momma, T.; Shoji, S.; Osaka, T. Electrochemistry Communications 2004, 6, 6, 562-565. [259] Ye, Q.; Zhao, T.S. J. Power Sources 2005, 147, 1-2, 196-202. [260] Xie, C.; Bostaph, J.; Pavio, J. J. Power Sources 2004, 136, 2, 55-65. [261] Zhang, J.; Colbow, K.M.; Lee, A.N.L.; Lin, B. (2004) US Patent Pub. No. 20040131898. [262] Narayanan, S.R.; Valdez, T.I.; Chun, W. Electrochem. Solid-State Lett. 2000, 3, 3, 117-120. [263] Icardi, U.A.; Specchia, S.; Fontana, G.J.R.; Saracco, G.; Specchia, V. Journal of Power Sources 2008, 176, 2, 460-467. [264] Specchia, S.; Icardi, U.A.; Specchia, V.; Saracco, G. Int. J. Chem. React. Eng. 2005, 3, A24-A27. [265] Sgroi, M.; Bollito, G.; Innocenti, G.; Saracco, G.; Specchia, S.; Icardi, U.A. J. Fuel Cell Sci. Technol. 2007, 4, 3, 345-349.

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

INDEX

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

A absorption spectroscopy, 110, 119 accuracy, 161 achievement, 86, 115, 118, 193 acid, 5, 7, 8, 23, 26, 34, 35, 36, 38, 40, 41, 43, 44, 45, 46, 47, 48, 49, 88, 98, 104, 133, 139, 140, 142, 143, 145, 148, 151, 152, 157, 160, 165, 167, 174, 183, 194, 197 acidic, 26, 29, 33, 40, 42, 43, 46, 48, 88, 98, 105, 106, 133, 134, 138, 141, 146, 148, 149, 157, 164, 171, 173, 194 acidity, 158 activated carbon, 63 activation, 9, 11, 12, 14, 43, 82, 106, 109, 118, 120, 127, 131, 139, 148, 151, 179, 184, 194, 195 activation energy, 109, 127, 151, 195 active oxygen, 33, 118, 119 active site, 36, 195 adhesion, 45 adiabatic, 189 ADP, 151 adsorption, 9, 11, 17, 30, 31, 32, 33, 42, 47, 67, 108, 110, 112, 114, 115, 118, 120, 132, 133, 137, 139, 153, 154, 157, 167, 168, 169, 171, 175, 193, 194 agent, 112, 143

agricultural, 2 aid, 67, 191, 196 alcohol, xv, 1, 2, 84, 172, 198 alkali, 33, 40, 135 alkaline, 8, 26, 33, 35, 38, 39, 45, 48, 49, 105, 138, 146, 147, 148, 149, 150, 162, 175, 193, 194, 195, 198 alkaline media, 40, 138, 148, 175, 193, 194, 195 alloys, 29, 32, 33, 34, 67, 110, 113, 114, 115, 116, 120, 134, 139, 164, 167, 168, 169, 172, 187, 193, 194, 198 alternative, 1, 7, 28, 29, 34, 38, 41, 44, 45, 49, 50, 73, 100, 109, 133, 134, 138, 141, 144, 160, 165, 182, 193, 194, 197, 198 ambient air, 61, 65, 80 ambient pressure, 10, 17, 56, 67, 69, 199 amelioration, 27, 28, 36, 128 ammonium, 147 amorphous, 119, 136, 163, 165, 167, 178, 180 Anion, 198 annealing, 183 Anode, 48, 55, 57, 59, 65, 66, 76, 80, 90, 103, 105, 126, 129, 131, 132, 134, 139, 148, 187 application, 24, 25, 39, 47, 49, 50, 51, 68, 70, 80, 85, 88, 102, 141, 143, 152, 153, 155, 157, 180, 188, 197, 200

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

170

Index

aqueous solution, 88, 151, 178 argon, 132 argument, 128, 140 asymptotic, 163 atmosphere, 189, 199 atmospheric pressure, 20, 65, 84, 172 atoms, 14, 26, 27, 29, 35, 112, 114, 116, 117, 119, 136, 165, 166, 167, 168, 173 automobiles, 85 automotive applications, 49, 90, 153 availability, 45, 104, 141

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

B barrier, 76, 77, 87, 120, 127, 142, 162, 163, 182, 186, 195 basic research, xiii, xv, 3 batteries, 1, 22, 23, 24, 53, 54, 63, 65, 85, 86, 88, 99, 185, 195, 196, 197, 199 behavior, 12, 13, 19, 28, 29, 46, 71, 98, 112, 120, 125, 127, 132, 137, 147, 158, 163, 167, 170, 171, 178, 184 bending, 156, 158 beneficial effect, 168, 187, 194 benefits, 3, 101, 163 bicarbonate, 40, 139, 147 binding energies, 67, 116, 121, 127 bipolar, 56, 64, 66, 68, 77, 86, 89, 90, 91, 177, 185, 187, 188 blends, 45, 152 blocks, 62 bonding, 117, 169 bonds, 40, 41, 117, 120, 150, 154, 158, 169 breathing, 60, 67, 70, 72, 76, 184 bubbles, 73, 74, 75 burn, 189 by-products, 104

C candidates, 35, 53, 141 capillary, 7, 67, 72, 78, 188

carbide, 42, 135, 193 carbon cloth, 61, 74, 81, 182, 186 carbon dioxide, 7, 26, 67, 103, 105, 111, 112, 113, 144, 146, 148, 189, 199 carbon monoxide, 107, 111, 112, 117, 135 carbon nanotubes, 138 carbon paper, 74, 182 carrier, 88 cast, 145 casting, 144 catalysis, 35, 129 catalytic activity, 29, 32, 49, 107, 109, 127, 132, 135, 171, 173, 175, 179 catalytic effect, 33 catalytic properties, 63, 127 catalytic system, 26, 28, 29, 34, 36, 47, 109, 128 cathode polarization, 9, 11, 170, 184 cell membranes, 39 cellular phone, 53, 54, 61, 63, 64, 68, 196, 199 ceramic, 39, 41, 61, 65, 142, 158 CFD, iii, iv chalcogenides, 34, 35, 175, 194, 198, 199 channels, 24, 42, 44, 74, 77, 78, 140, 154, 182, 185, 186, 187, 188, 191 chemical energy, 20 chemical interaction, 14, 26, 110, 114, 119, 153 chemical oxidation, 19, 29 chemical properties, 45, 179 chemical stability, 41, 160, 174 chemicals, 1 chemisorption, 105, 109, 110, 112, 113, 114, 116, 117, 135, 166, 167, 168, 175, 184 chromatography, 67 circulation, 188 classes, 35, 144 classical, 38, 45, 175, 187 cleavage, 169 clusters, 44 coal, 2 cobalt, 173, 198

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Index codes, 101 collaboration, 61, 65, 87 combustion, xv, 1, 24, 85, 142 commercialization, 85, 197 communication, 49, 65 communication systems, 65 community, 96 compensation, 168 competition, 17, 102, 133, 200 competitiveness, 128, 196 complexity, 37, 75, 152, 179, 190, 196, 199 components, 3, 38, 45, 49, 50, 70, 95, 99, 177, 183, 189, 190 composites, 144 composition, 78, 105, 107, 114, 119, 137, 179, 186, 190 compounds, 7, 35, 36, 106, 110, 118, 134, 135, 137, 145, 152, 169, 179 condensation, 154, 155 conditioning, 183 conducting polymers, 141 conduction, 42, 44, 46, 141, 142, 145, 147, 152, 154, 195, 199 conductor, 49, 144, 187 configuration, 61, 63, 65, 66, 70, 76, 97, 105, 126, 132, 181, 182, 185, 186, 187 Congress, ix constraints, 12, 38, 40, 42, 48, 50, 138, 141, 147, 149, 157, 185, 193, 197, 198 construction, 66, 68, 84, 98, 189 consumers, 100 consumption, 14, 17, 19, 21, 43, 65, 75, 85, 96, 104, 144, 155, 159, 161, 162, 185, 188 continuity, 182 contracts, 201 control, 6, 9, 11, 43, 76, 109, 178, 188, 190 convection, 6, 64, 66, 84, 186 conversion, 73, 99, 163, 197 cooling, 73, 186 correlation, 30, 32, 33, 154 corrosion, 26, 46, 89, 98, 138, 144, 146, 148, 167, 175, 185, 187, 191, 193 cost-effective, 62

171

costs, 45, 47, 87, 101, 104, 164, 179, 185, 191, 197 couples, 81 covering, 79 cracking, 63 critical analysis, 146 crossing over, 77 cross-linking, 143, 145, 151 crystalline, 119, 136, 158, 163, 165, 167, 180 crystals, 167 cycles, 191 cyclic voltammetry, 10, 33, 106, 110, 127, 175

D decomposition, 173, 179 defects, 109, 110 degradation, 39, 44, 46, 146, 149, 169, 174 degradation mechanism, 149 dehydration, 7, 140, 197 dehydrogenation, 13, 32, 105, 110, 111, 113, 114, 116, 134, 175 delivery, 2, 61, 188 deposition, 62, 64, 77, 128, 137, 142, 169, 178, 179, 181 desorption, 171, 193 detection, 107 diamines, 145 diesel, 85, 87, 102 diesel fuel, 87 diffusion, 5, 27, 38, 61, 67, 71, 72, 76, 81, 84, 92, 140, 146, 148, 161, 177, 180, 181, 182, 185, 186, 187, 188 disclosure, 37 dispersion, 128, 136, 137, 154, 164, 178 displacement, 154 distributed generation, xv, 2, 100, 101, 185 distribution, xv, 1, 22, 64, 67, 68, 80, 86, 87, 92, 99, 101, 104, 138, 143, 154, 160, 187, 197 doped, 7, 43, 44, 144, 146, 193, 194, 198 doping, 135

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

172

Index

dream, 203 drying, 73, 88 durability, 86

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

E earth, 135 economics, 101 electric cars, 197 electric circuit, 81 electric power, 92 electric utilities, 100 electrical conductivity, 79 electrical power, 6, 48, 98, 128, 159 electrical resistance, 70 electricity, 99, 100, 101, 102 electrocatalysis, 3, 35, 36, 104, 108, 164 electrochemical reaction, 38, 96, 178, 190 electrochemistry, 106, 110 electromotive force, xv, 1, 9, 147 electron, 5, 77, 103, 104, 112, 117, 123, 131, 168, 173, 175, 193 electron beam, 77 electron density, 118 electronegativity, 30, 32, 116 electronic structure, 35, 173 electrons, 106, 116, 117 electroplating, 77 emission, 1, 85, 87 energy, xv, 1, 2, 3, 6, 8, 20, 21, 22, 23, 27, 31, 36, 47, 53, 62, 66, 69, 73, 85, 87, 99, 101, 102, 109, 113, 115, 116, 121, 167, 185, 191, 196, 197, 199 energy consumption, 185 energy density, xv, 1, 2, 3, 21, 22, 23, 47, 53, 62, 66, 87, 102, 196, 199 energy efficiency, 8, 73 energy supply, 53 engines, 1 enthusiasm, 85 environment, 26, 29, 32, 33, 46, 105, 118, 133, 134, 149, 153, 194

equilibrium, 114 ERD, 62, 63 erosion, 188 etching, 77, 78 ethanol, iv, 2, 23, 36, 41, 62, 75, 76, 153, 188 ethylene, 44, 143 European Commission, 209 European Community, 19, 201 European Union, 92, 95, 201, 204, 207 evaporation, 21 evolution, 110, 175 EXAFS, 27, 119 exposure, 80 Exxon, 48

F fabrication, xiii, xvi, 3, 62, 69, 70, 77, 78, 88, 146, 180, 182, 183, 185, 188, 195, 197 fermentation, 2 fiber, 88 filler surface, 43, 154, 158 fillers, 84, 141, 156, 157, 158 film, 61, 63, 80, 107, 117, 143, 144, 150, 198 financial support, 92, 201 flexibility, 70 flooding, 12, 14, 17, 34, 72, 73, 182, 184, 186, 195, 199 flow, 13, 15, 17, 19, 62, 65, 67, 68, 69, 72, 73, 76, 79, 86, 89, 95, 145, 177, 183, 184, 185, 186, 187, 188 flow field, 13, 65, 68, 72, 89, 177, 185, 186, 187 flow rate, 15, 19, 62, 69, 73, 79, 95, 184 fluctuations, 101 flue gas, 190 fluid, 98 fluoride, 144 fluorinated, 44, 96, 149, 150 fluorine, 41 foams, 187 focusing, 40

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Index Ford, 86 formaldehyde, 8, 104 fossil fuels, 2, 99 free energy, 8 FTIR, 106, 110, 116 fuel efficiency, 21, 67, 104, 142, 163, 184, 188

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

G gas, 2, 12, 16, 61, 67, 71, 74, 76, 81, 84, 86, 99, 100, 102, 128, 135, 143, 177, 178, 180, 182, 186, 189, 190 gas chromatograph, 67 gas diffusion, 61, 71, 76, 177 gas sensors, 84, 189 gasoline, 47, 86, 87, 100 gel, 144, 183, 195 General Electric, 40 General Motors, 86 generation, xv, 1, 22, 40, 45, 47, 53, 100, 101, 152, 185, 197 generators, 102, 128 Gibbs free energy, 8 glass, 41 glass transition temperature, 41 glycerol, 180 glycol, 23 goals, 85, 95, 137, 169, 178 gold, 63, 80, 134, 188 grafting, 84, 143, 150, 151 graphite, 36, 68, 72, 89, 94, 185, 186, 187 gravity, 78 greenhouse gas, 85, 197 groups, 27, 41, 73, 88, 111, 113, 114, 119, 140, 143, 154, 157, 160, 183

173

heat removal, 186 heme, 35 heterogeneous, 108 high power density, 102 high pressure, 87, 155, 184 high resolution, 188 homogeneity, 92 Honda, 85, 86, 99 hot pressing, 69, 89 House, iv housing, 75, 177 humidity, 62, 141, 153 hybrid, 67, 69, 78, 85, 86, 96, 99, 152, 157, 195 hydration, 63, 158, 185 hydride, 2, 22, 86 hydro, 12, 17, 24, 68, 74, 140, 182, 191, 195 hydrocarbon, xv, 1, 2, 46, 149 hydrocarbon fuels, xv, 1 hydrogen, xv, 1, 2, 3, 7, 9, 11, 12, 32, 40, 44, 48, 62, 69, 86, 87, 99, 103, 134, 135, 149, 158, 160, 171, 178, 194, 197 hydrogen bonds, 158 hydrogen peroxide, 40, 149 hydrophilic, 12, 18, 24, 68, 74, 140, 182, 191, 195 hydrophilicity, 74, 182 hydrophobic, 5, 34, 63, 64, 74, 76, 182, 191, 195, 199 hydrophobic properties, 5 hydrophobicity, 74, 183 hydroxide, 120 hydroxyl, 29, 112, 133, 134, 135, 157 hypothesis, 29, 112

I H handling, xv, 2, 100, 104, 185 heat, 18, 19, 20, 71, 99, 102, 186, 189, 190, 191, 194, 195 heat release, 19, 20, 191

ICE, 24 identification, 140 impedance spectroscopy, 106 impregnation, 88, 137, 165, 169, 178, 179 in situ, 31, 67, 73, 116, 125

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

174

Index

inactive, 166, 167, 169 incidence, 130 inclusion, 141 industrial, 33, 100, 143 industry, 62, 100, 101 inert, 16, 107, 178, 186 infrared spectroscopy, 27 infrastructure, xv, 1, 2, 22, 86, 87, 100, 197 injury, ix Innovation, 57, 69, 88 inorganic, 43, 68, 96, 141, 152, 153, 154, 157, 163, 174 inorganic filler, 43, 141, 152, 153, 154, 157 instability, 145 institutions, 87 integration, 101, 177, 188, 189 integrity, 180 interaction, 30, 31, 32, 108, 114, 115, 117, 120, 127, 134, 157, 164, 169, 181, 193 interface, 16, 43, 72, 107, 118, 139, 149, 162, 181, 184, 199 internal combustion, 1, 24, 85, 142 intrinsic, 86, 107, 114, 127, 132, 135, 138, 164, 167, 171, 179 inversion, 158 ion transport, 144 ionic, 5, 30, 33, 38, 45, 84, 144, 145, 147, 149, 151, 158, 163, 180, 183, 194 ionic conduction, 147 ions, 46, 62, 133, 147, 148, 157, 169, 174 iron, 173, 198

J Japanese, 99 Jatropha, iv Jet Propulsion Laboratory, 64

K ketones, 145 kinetic equations, 27, 108

kinetics, 3, 7, 12, 26, 28, 39, 42, 50, 89, 99, 120, 127, 138, 146, 151, 167, 175, 184, 197, 198 KOH, 26, 33, 39, 40, 46, 48, 139, 151, 194, 195

L Langmuir, 210, 211, 212 laptop, 23, 53, 54, 199 large-scale, xv, 1, 185 lattice, 119, 167, 168, 170 law, 15, 19, 21, 197 leaches, 173 leaching, 43, 145, 167 leakage, 147, 152, 194 legislation, 85 lifetime, 85 ligand, 117 Li-ion batteries, 22 limitations, 66, 138, 158, 162, 179, 186 linear, 31, 32, 105, 106 liquid fuels, xv, 1, 22 liquid water, 17, 155, 159 liquids, 143 lithium, 23, 53, 65, 199 lithium ion batteries, 199 location, 132 losses, 73, 82, 86, 131, 139, 159, 182 low molecular weight, 105 low power, 54, 61, 67 low temperatures, xv, 2, 38, 42, 104, 114, 169, 195

M macrocyclics, 174 magnetic, ix management, 24, 42, 66, 74, 75, 98, 99, 151, 177, 186 manifold, 64, 177, 185 manufacturing, xiii, xvi, 3, 101, 177, 178, 196

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Index market, 23, 39, 53, 54, 84, 92, 128, 142, 189, 196 market penetration, 142 mass spectrometry, 27, 106, 110 mass transfer, 70, 71, 184, 186 mass-transport, 182, 186 matrix, 41, 141, 183, 188 measurement, 15 media, 26, 36, 40, 139, 148, 175, 193, 194, 195 MEMS, 77 metal content, 50 metal ions, 174 metal nanoparticles, 130 metal oxides, 35, 134, 193 metals, 30, 31, 32, 34, 35, 46, 105, 134, 135, 138, 148, 164, 167, 169, 172, 173, 178, 194, 196 methanol poisoning, 17, 35 MFC, 65, 70, 71, 74, 79, 128 micelles, 162, 181, 183 microscopy, 107 microstructure, 180 migration, 28, 39, 120, 148, 151, 157, 162 military, 199 miniaturization, 22, 185 mixing, 14, 61, 120, 180, 181, 182 mobile phone, 23, 53, 67, 84 mobility, 134, 135, 149, 151, 152 modules, 177 molecular oxygen, 35 molecular weight, 22, 105, 145, 179 molecules, 5, 8, 16, 19, 32, 36, 42, 43, 75, 105, 109, 140, 151, 154, 158, 168, 173, 184, 186 molybdenum, 133 momentum, 110, 195 monolayer, 14, 154 mordenite, 144 morphology, 137, 179

175

N nanofiller, 158 nanoparticles, 42, 130, 131, 132, 141, 157, 158, 169 nanotubes, 138 natural, 2, 6, 64, 84, 100, 102, 188 natural gas, 2, 100, 102 network, 86, 87, 141, 180, 183 niche market, 92 niobium, 70 nitrogen, 182, 186, 195 NMR, 107, 146 noble metals, 34, 35, 46, 134, 148, 196 noise, 102 non-noble, 35, 38, 46, 167, 196, 198 non-uniform, 64 normal, 15, 17, 68, 177, 184, 191 novelty, 77 nuclear, 62, 63

O occlusion, 48 octane, 100 on-line, 110, 189 optimization, 74, 134, 180 organ, 35, 36 organic, xv, 2, 8, 32, 105, 140, 173, 174, 179, 184, 198 organic compounds, 179 organometallic, 35, 36 orientation, 74, 79, 110, 164, 166, 167 osmotic, 15, 16, 38, 46, 140, 148, 162, 195 oxidation rate, 27, 30, 104, 112, 135, 138, 168 oxidative, 40, 135 oxide, 35, 39, 43, 117, 118, 119, 130, 134, 135, 141, 152, 154, 158, 168, 175, 178, 193, 194

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

176

Index

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

P packaging, 66 PAFC, 34 palladium, 63 parameter, 42, 167, 170, 180 particles, 50, 63, 64, 105, 107, 119, 121, 132, 136, 154, 158, 163, 165, 178, 180, 181 partnership, 100 passenger, 87 passive, 6, 21, 24, 47, 50, 69, 70, 71, 72, 73, 74, 75, 78, 79, 80, 150, 184, 187, 188, 191, 195 PBI, 47, 142, 152 PCBs, 80 PCM, 58, 75, 98 PCP, 72 PEEK, 44, 45 PEMFC, 3, 7, 12, 17, 44, 86, 87, 88, 153, 185, 197 periodic table, 26 permeability, 84, 144, 162, 182 permeation, 14, 15, 16, 38, 44, 45, 68, 140, 150, 154, 163, 188, 199 perovskite, 35 peroxide, 44, 173 peroxide radical, 44 pH values, 138, 147, 149, 169, 175 phone, 65, 67, 199 phosphate, 43, 141, 146, 194 phosphoric acid fuel cell, 34, 165, 167 photoelectron spectroscopy, 127 photographs, 74 photolithography, 77, 78, 79 phthalocyanines, 36, 173, 194, 198 physical interaction, 154 physicochemical, 30, 128, 169, 193 physicochemical properties, 30, 128, 169, 179 pitch, 66 planar, 36, 55, 61, 65, 70, 188 plants, 101, 143 plastic, 62, 63, 66, 75, 80

platinum, 27, 48, 63, 64, 70, 78, 87, 117, 164, 168, 173, 174, 194 play, 118, 157, 164, 171 poison, 29, 30, 31, 32, 104, 112 poisoning, 17, 25, 29, 31, 32, 35, 39, 50, 104, 112, 128, 151, 167, 168, 199 polarization, 9, 10, 11, 12, 14, 15, 19, 37, 41, 67, 71, 72, 73, 77, 82, 83, 92, 97, 127, 129, 130, 162, 170, 171, 172, 183, 184 pollutants, 85 pollution, 102 polycarbonate, 70, 188 polyether, 194 polyimide, 45 polymer blends, 140 polymer electrolytes, 49, 144, 147, 151 polymer matrix, 141 polymer membranes, 143 polymer structure, 145, 163 polystyrene, 40, 194 polytetrafluoroethylene, 180, 186 polyvinyl alcohol, 142 poor, 7, 28, 39, 40, 42, 50, 70, 136, 152 pores, 40, 48, 62, 63, 73, 147, 154, 155, 180, 181, 182, 183, 186, 195 porosity, 180, 183, 195 porous, 41, 61, 63, 67, 71, 72, 141 porphyrins, 36, 194 ports, 63 positive correlation, 95 potassium, 48, 147 powder, 63, 64, 69 power plants, 1, 85 precipitation, 40, 48, 147, 179, 195 premium, 101 pressure, 7, 9, 10, 12, 17, 20, 21, 43, 56, 65, 67, 69, 79, 84, 92, 95, 96, 97, 155, 157, 158, 163, 172, 183, 184, 186, 189, 199 printing, 182 production, xv, 1, 62, 143, 146, 179, 191 production costs, 179, 191 program, 40, 96, 143 propane, 100

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Index property, ix proton exchange membrane, xv, 2, 47, 84, 148, 149 protons, 16, 148, 149, 151 prototype, 49, 61, 67, 87, 88 PTFE, 34, 180, 182, 186 pumps, 6, 21, 62, 69, 177, 188, 191

Q quantum, 67 quaternary ammonium, 147

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

R radiation, 44, 150, 151, 194 radio, 143 radius, 30, 33, 34 rail, 100 range, 7, 9, 12, 15, 24, 39, 43, 44, 47, 68, 79, 84, 85, 88, 99, 106, 108, 130, 135, 136, 141, 149, 150, 152, 154, 157, 158, 168, 191, 199 rare earth elements, 135 reactant, 6, 50, 62, 92, 94, 138, 139, 145, 162, 177, 183, 185, 186, 187, 188 reaction mechanism, 106 reaction order, 106, 108, 184 reaction rate, 14, 29, 38, 50, 103, 115, 134, 135, 136, 138, 148, 166, 167, 178, 195 reaction zone, 130, 132, 149, 152, 195 reactive ion, 77, 78 reactivity, xv, 2, 109 recovery, 24 recycling, 75 redox, 19, 29, 112, 133, 147 refractory, 193 regenerate, 46, 48, 147, 194, 195 regulation, 102, 135 rejection, 194 relationship, 107, 108, 109, 137, 140 relevance, 3

177

reliability, 51, 101, 185 repeatability, 188 reservoir, 79, 80 residential, 100, 101, 102 residues, 14, 27, 106, 108, 111, 112, 113, 114, 115, 116, 119, 125, 126, 127, 132, 133, 154, 167, 170, 173, 184 resilience, 145 resistance, 15, 35, 42, 48, 67, 70, 71, 139, 143, 144, 150, 158, 159, 160, 169, 199 resistivity, 145 resolution, 127, 188 resources, 87 retention, 41, 43, 141, 145, 152, 154, 157, 158, 159, 160, 163 RIE, 78 risks, 101 roads, iv room temperature, 21, 67, 71, 75, 76, 77, 114, 150, 184, 199 roughness, 173 ruthenium, 64, 78, 131

S safety, 2 sample, 119, 121, 127 saturation, 153 scanning tunneling microscopy, 107 scattering, 27, 137, 146 search, 25, 26 second generation, 62 security, 189 seed, 79 selecting, 133, 143, 160 selectivity, 2, 174 self-assembling, 178, 179 SEM, 6, 181 semiconductor, 62, 199 sensors, 6, 62, 84, 189, 191 separation, 140, 177

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

178

Index

series, 6, 61, 62, 64, 66, 68, 70, 77, 81, 177, 185, 187 services, ix shape, 11, 137 Shell, 48 short-term, 152 side effects, 21 silica, 42, 43, 145, 152, 163 silicon, 42, 76, 77, 78, 188, 194 silver, 25, 33, 40, 48, 194 sintering, 61, 179 sites, 5, 6, 14, 30, 36, 82, 84, 89, 109, 111, 112, 113, 114, 116, 118, 119, 127, 133, 135, 138, 153, 158, 165, 166, 167, 168, 175, 182, 184, 186, 195 skin, 169, 173 sodium, 135, 147, 193 sol-gel, 43, 179 solid oxide fuel cells, 35, 39 solid-state, 46, 49, 144, 147 solvents, 88, 150, 179, 180 species, 9, 14, 17, 27, 29, 31, 33, 42, 103, 105, 106, 107, 108, 110, 111, 112, 113, 116, 118, 119, 121, 133, 134, 135, 140, 146, 168, 175, 193, 194 spectroscopic methods, 110 spectroscopy, 27, 106, 146 spectrum, 71, 116, 199 speed, 88 sputtering, 62, 179 stability, 36, 39, 40, 41, 44, 47, 72, 133, 140, 145, 147, 149, 151, 160, 164, 173, 174, 187, 194, 195 stainless steel, 66, 70, 74, 92 steel, 66, 70, 74, 79, 92, 187 steric, 27, 29, 30, 32, 112 stock, 69 stoichiometry, 17, 21, 55, 65 storage, xv, 1, 2, 68, 87, 153 strategies, 43, 138, 178 strength, 32, 112, 117, 120, 143, 144, 157 stretching, 116, 156, 158 strong interaction, 32, 158

structural changes, 141 styrene, 143, 144, 150 substances, 104 substitutes, 35, 105, 174 substrates, 61, 79, 89, 173 sulfuric acid, 26, 36, 41, 48, 88 sulphate, 139 superacids, 46 supply, 6, 17, 53, 68, 72, 78, 100, 118, 119, 162, 182, 191, 197 surface area, 28, 64, 92, 95, 107, 110, 136, 137, 167, 169, 171, 173, 174, 178, 179, 180, 181 surface chemistry, 157, 180 surface properties, 115, 130, 158, 164 surface treatment, 187 swelling, 39, 44, 146, 161 synergistic, 110, 115 synthesis, 105, 137, 143, 150, 169, 178, 179 systems, xv, 1, 2, 7, 23, 28, 30, 37, 44, 47, 50, 53, 62, 68, 85, 101, 103, 115, 117, 139, 140, 141, 142, 148, 149, 152, 153, 154, 155, 157, 185, 188, 190, 191, 196, 197

T tanks, 191 targets, 86, 200 technology, xiii, xvi, 1, 2, 3, 47, 61, 65, 77, 84, 85, 86, 87, 99, 101, 102, 142, 169, 180, 188, 196, 197, 199, 203 Teflon, 5, 41, 64 TEM, 119, 121, 170, 181 TEOS, 43 thermal decomposition, 179 thermal resistance, 143 thermal treatment, 164, 179, 183 thermodynamic, 1, 3, 8, 9, 33, 103, 135, 147, 149, 195, 198 thin film, 65, 144, 182, 183

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,

Copyright © 2010. Nova Science Publishers, Incorporated. All rights reserved.

Index time, 15, 19, 24, 25, 30, 38, 47, 48, 61, 65, 75, 84, 85, 99, 102, 137, 138, 142, 146, 153, 161, 167, 178, 182, 184, 185, 188, 195, 197 time consuming, 161 tin, 115, 118, 144 titanium, 43 tolerance, 3, 149, 169, 170, 171, 172, 175, 194 Toyota, 85, 86 toys, 76 traction, 1, 3, 99 trans, 53, 69 transfer, 30, 71, 103, 104, 106, 116, 118, 120, 127, 168, 173, 193 transition, 30, 31, 32, 87, 105, 112, 134, 135, 138, 160, 164, 167, 169, 172, 173, 175, 178, 179, 193, 194, 198, 199 transition elements, 32 transition metal, 31, 32, 105, 112, 134, 135, 138, 164, 167, 169, 172, 173, 175, 178, 193, 194, 198, 199 transition temperature, 41 transmission, 101, 131 transparent, 73 transport, 5, 7, 11, 35, 38, 44, 48, 50, 71, 72, 76, 77, 78, 82, 84, 87, 137, 139, 140, 142, 144, 146, 157, 162, 178, 180, 182, 183, 186, 199 transportation, xv, 2, 22, 24, 47, 49, 85, 87, 88, 89, 99, 160, 185, 187, 188, 191, 195, 199 tungsten, 35, 133, 134, 135, 193 tungsten carbide, 135, 193 tunneling, 107

179

V vacancies, 115, 117, 135, 169 vacuum, 107 validity, 99 values, 14, 30, 45, 48, 75, 79, 113, 116, 127, 138, 147, 149, 150, 151, 159, 169, 175, 187 Van der Waals bonds, 154 vapor, 18, 189, 191 variation, 12, 14, 15, 108, 147, 153, 156, 163, 166, 184 vehicles, 85, 86, 102, 152, 195 ventilators, 66 vibration, 156 visualization, 74 Volkswagen, 86 voltammetric, 107

W water recycling, 75 WAXS, 146 wettability, 73, 89 windows, 78 workers, 117, 129, 141, 183, 187 working conditions, 6, 144

X XANES, 116, 119 XPS, 27, 116, 118, 119, 121, 127 X-ray absorption, 110, 115, 116, 119 X-ray photoelectron spectroscopy (XPS), 27 XRD, 119, 120, 121, 122, 169, 170, 181

U uniform, 64, 68, 74, 79 urban areas, 101

Z zeolites, 144, 152, 194 zirconium, 43, 141, 146, 194

Direct Methanol Fuel Cells, Nova Science Publishers, Incorporated, 2010. ProQuest Ebook Central,