fuel cell science - download.e-bookshelf.de€¦ · fuel cell catalysis: a surface science...

FUEL CELL SCIENCE

Wiley Series on Electrocatalysis and Electrochemistry

Andrzej Wieckowski, Series Editor

Fuel Cell Catalysis: A Surface Science Approach, Marc T. M. Koper

Electrochemistry of Functional Supramolecular Systems, Margherita Venturi, Paola Ceroni,

and Alberto Credi

Catalysis in Electrochemistry: From Fundamentals to Strategies for Fuel Cell Development,

Elizabeth Santos and Wolfgang Schmickler

Fuel Cell Science: Theory, Fundamentals, and Biocatalysis, Andrzej Wieckowski and

Jens Nørskov

FUEL CELL SCIENCETHEORY, FUNDAMENTALS,AND BIOCATALYSIS

Edited by

Andrzej WieckowskiJens K. Nørskov

Copyright � 2010 by John Wiley & Sons, Inc. All rights reserved.

Published by John Wiley & Sons, Inc., Hoboken, New Jersey

Published simultaneously in Canada

No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form

or by any means, electronic, mechanical, photocopying, recording, scanning, or otherwise, except as

permitted under Section 107 or 108 of the 1976 United States Copyright Act, without either the prior

written permission of the Publisher, or authorization through payment of the appropriate per-copy fee to

the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, 978-750-8400,

fax 978-750-4470, or on the web at www.copyright.com. Requests to the Publisher for permission

should be addressed to the Permissions Department, John Wiley & Sons, Inc., 111 River Street, Hoboken,

NJ 07030, 201-748-6011, fax 201-748-6008, or online at http://www.wiley.com/go/permission.

Limit of Liability/Disclaimer of Warranty: While the publisher and author have used their best efforts

in preparing this book, they make no representations or warranties with respect to the accuracy or

completeness of the contents of this book and specifically disclaim any implied warranties of

merchantability or fitness for a particular purpose. No warranty may be created or extended by sales

representatives or written sales materials. The advice and strategies contained herein may not be

suitable for your situation. You should consult with a professional where appropriate. Neither the

publisher nor author shall be liable for any loss of profit or any other commercial damages, including

but not limited to special, incidental, consequential, or other damages.

For general information on our other products and services or for technical support, please contact

our Customer Care Department within the United States at 877-762-2974, outside the United States

at 317-572-3993 or fax 317- 572-4002.

Wiley also publishes its books in a variety of electronic formats. Some content that appears in print

may not be available in electronic formats. For more information about Wiley products,

visit our web site at www.wiley.com.

Library of Congress Cataloging in-Publication Data:

Wieckowski, Andrzej

Fuel cell science : theory, fundamentals, and biocatalysis / edited by Andrzej Wieckowski

and Jens Nørskov.

p. cm.

Includes bibliographical references and index.

ISBN 978-0-470-41029-5 (cloth)

Printed in the United States of America

10 9 8 7 6 5 4 3 2 1

www.copyright.com

http://www.wiley.com/go/permission

www.wiley.com

CONTENTS

Foreword vii

Preface xi

Preface to the Wiley Series on Electrocatalysis and Electrochemistry xiii

Contributors xv

1. Hydrogen Reactions on Nanostructured Surfaces 1

Holger Wolfschmidt, Odysseas Paschos, and Ulrich Stimming

2. Comparison of Electrocatalysis and Bioelectrocatalysis

of Hydrogen and Oxygen Redox Reactions 71

Marc T. M. Koper and Hendrik A. Heering

3. Design of Palladium-Based Alloy Electrocatalysts

for Hydrogen Oxidation Reaction in Fuel Cells 111

Sung Jong Yoo and Yung-Eun Sung

4. Mechanism of an Enhanced Oxygen Reduction Reactionat Platinum-Based Electrocatalysts: Identification

and Quantification of Oxygen Species Adsorbed on

Electrodes by X-Ray Photoelectron Spectroscopy 147

Mitsuru Wakisaka, Hiroyuki Uchida, and Masahiro Watanabe

5. Biocathodes for Dioxygen Reduction in Biofuel Cells 169

Renata Bilewicz and Marcin Opallo

6. Platinum Monolayer Electrocatalysts: Improving

Structure and Activity 215

Kotaro Sasaki, Miomir B. Vukmirovic, Jia X. Wang, and Radoslav R. Adzic

7. The Importance of Enzymes: Benchmarks for Electrocatalysts 237

Fraser A. Armstrong

8. Approach to Microbial Fuel Cells and Their Applications 257

Juan Pablo Busalmen, Abraham Esteve-Nunez , and Juan Miguel Feliu

9. Half-Cell Investigations of Cathode Catalysts for PEM Fuel Cells:

From Model Systems to High-Surface-Area Catalysts 283

Matthias Arenz and Nenad M. Markovic

v

10. Nanoscale Phenomena in Catalyst Layers for PEM Fuel Cells:

From Fundamental Physics to Benign Design 317

Karen Chan, Ata Roudgar, Liya Wang, and Michael Eikerling

11. Fuel Cells with Neat Proton-Conducting Salt Electrolytes 371

Dominic Gervasio

12. Vibrational Spectroscopy for the Characterizationof PEM Fuel Cell Membrane Materials 395

Carol Korzeniewski

13. Ab Initio Electrochemical Properties of Electrode Surfaces 415

Ismaila Dabo, Yanli Li, Nic�ephore Bonnet, and Nicola Marzari

14. Electronic Structure and Reactivity of Transition Metal Complexes 433

Heather J. Kulik and Nicola Marzari

15. Quantitative Description of Electron Transfer Reactions 457

Patrick H.-L. Sit, Agostino Migliore, Michael L. Klein, and Nicola Marzari

16. Understanding Electrocatalysts for Low-Temperature Fuel Cells 489

Peter Ferrin, Manos Mavrikakis, Jan Rossmeisl, and Jens K. Nørskov

17. Operando XAS Techniques: Past, Present, and Future 511

Christina Roth and David E. Ramaker

18. Operando X-Ray Absorption Spectroscopy of Polymer

Electrolyte Fuel Cells 545

Eugene S. Smotkin and Carlo U. Segre

19. New Concepts in the Chemistry and Engineering

of Low-Temperature Fuel Cells 565

Fikile R. Brushett, Paul. J. A Kenis, and Andrzej Wieckowski

Index 611

vi CONTENTS

FOREWORD

IS THERE A COMMON “ACTIVITY YARDSTICK” THAT APPLIESTO ALL FUEL CELL ELECTROCATALYSTS?

Thinkingwhat should be themessage in the foreword to a book that covers extensively

awide frontier of fuel cell catalysis work, a tempting, albeit somewhat risky, idea kept

coming up in my mind: Is it possible to define a common “activity yardstick” that

applies to a large number of, if not to all, fuel cell electrocatalysts? Is it possible to

make such a generalization when considering the wide variety of catalytic surfaces

studied and practiced in low-temperature fuel cells?

When examining the relevant literature, it appears that themore recent searches for

active metal electrocatalysts and for active molecular electrocatalysts have had

somewhat different priorities. In the case of metal electrocatalysts, the focus has

been on tailoring the electronic properties of metal alloy surfaces to achieve an

optimized bond strength between the metal surface and the relevant adsorbed

intermediates [1]. Such studies have been supported by density functional theory

(DFT) calculations, yielding the energies of the bonds between catalytically active

surfaces and the likely reaction intermediates [2]. In all such studies, the assumption

has been that a complete description of the electrocatalytic process requires consid-

eration of a reactant molecule and a metal surface in contact with water, or aqueous

electrolyte. The electro element of electrocatalysis has been covered all along by

assuming that a change in the interfacial potential difference has an effect on, and only

on, the activation energy of any reaction step involving electron transfer. Accordingly,

the typical rate expression for an electrocatalytic process takes the formof a product of

a preexponential term and a two-component exponential term, with the rate depen-

dence on the electrode potentialE fully covered in the exponential term. For a cathodic

process within the so-called Tafel regime, the rate expression takes the following

general form

JðEÞ ¼ Fk0Acat* Cg

rexpf�DH*act=RTg expf�ðE�E� cell processÞ=bg ð1Þ

where F is the Faraday constant, k0 is a frequency factor, Acat* is the overall catalyst

surface area per unit electrode cross-sectional area, Cr is the concentration of the

reactantmolecule at the electrode surface, g is the reaction orderDHact* is the chemical

component of the activation energy, and b is the so-called Tafel slope. In the case of

molecular electrocatalysts, the more recent achievements in preparation of highly

vii

oxygen reduction reaction (ORR)-active, carbon-supported iron complexes [3],

resulted from efforts to maximize the overall surface density of effective redox

centers, N*. Lefevre et al. [3] showed that an effective center formed when the iron

complex was located on a specific, pretailored carbon surface site. The mechanism of

electrocatalytic processes taking place at such active surface sites is described in terms

of redox mediation, where the electrocatalytic activity at a potential E is expected to

involve a fraction of N*, Nactive (E), defined by

NactiveðEÞ ¼ N*f ðE�E�surface redoxÞ ð2Þ

For example, in the specific case of ORR catalyzed by a molecular iron complex, a

plausible mechanism of ORR at the active complex of iron, X–Fe(II), where X is a

surface anchor site and the iron complex is in its reduced form, has been described [3]

with a first step involving bonding of dioxygen to the active form of the iron complex,

X–Fe(II), assisted by electron transfer from the Fe(II) center:

O2 þX--FeðIIÞ ¼ X--FeðIIIÞ--O--Oþ e ð3aÞ

This step is followed by the completion of the 4e reduction process with regeneration

of the active form of the redox system, written in simplified form as follows:

X--FeðIIIÞ--O--Oþ eþ 3eþ 4Hþ ¼ X--FeðIIÞþ 2H2O ð3bÞ

This mechanism implies that only at a cathode potential sufficiently negative to

generate a significant population of the reduced form of the surface redox couple,

X–Fe(II), can the rate of the process in Equations (3a) and (3b) rise to a measurable

level. In an ideal case where the steady-state population of X–Fe(II) depends on

potential according to a simple Nernst equation, the number of active sites at an

electrode potential E will be given, for a cathodic process, as

NactiveðEÞ ¼ N*½1=ðZþ 1Þ� ð4Þ

whereZ ¼ expfðF=RTÞ ðE�E�surface redoxÞg. Inserting inEquation (1) this dependence

of active-site population on electrode potential, the rate expression will take the form

JðEÞ ¼ Fk0Acat* f ðE�E�

surface redoxÞCgr expf�DHact

* =RTg expf�ðE�E�cell processÞ=bg

ð5Þwhere in the simplest case, f ðE�E�

surface redoxÞ ¼ 1=ðZ þ 1Þ:The significant difference between Equations (5) and (1) is the appearance in (5) of

two sources of rate dependence on electrode potential, associated with two different

values ofE�. One is the dependence of the activation energy at an active surface site onthe overpotential, E�E�

cell process, and the second is a dependence of active-site

population on E�E�surface redox. The former appears in the exponential term of the

rate expression, whereas the latter appears in the preexponential term [4].

viii FOREWORD

The tacit assumption behind the use of the simpler expression [Eq. (1)] for

processes at metal surfaces is that availability of active sites on metal surfaces does

not depend on the electrode potential. This assumptionmisses, however, a key element

of electrocatalysis at metal surfaces [4]. For example, examination of the value of

E�M=M;ox for metal and metal alloy electrocatalysts of high ORR activity, reveals that

“ignition” of theORRprocess is tied consistentlywith the onset of cathodic generation

of some minimum surface population (e.g., 1%) of free metal sites on a surface that is

fully covered under open-circuit conditions by oxygen species that block ORR.

Recognizing that such change in surface composition is required for the onset of the

process, one can describe the ORR process at Pt in terms of surface redox mediation,

involving the Pt/PtOx surface redox system [4]. ORR ignition requires reduction of a

Pt surface oxide, or hydroxide species, for example, according to

2Pt�OHsurface þ 2Hþ þ 2e ¼ 2Ptsurface þ 2H2O ð6aÞ

followednextby reactionofO2atPt andwithmetal sites that becomeavailablebeyond

a threshold potential determined by E�M=MeOH according to

2Ptsurface þO2 þ 2eþ 2Hþ ¼ 2Pt�OHsurface ð6bÞ

Continuous repetition of (6a) þ (6b) sustains a steady-state rate of a 4e ORR process,

taking place at the active (metal) surface sites, with the active site continuously

regenerated beyond a threshold potential determined by E�M=MeOH.

Mediation by a surface redox system is apparently a common feature of a wide

variety of electrocatalysts, whether molecular or metallic, and this insight can lead

to an attempted definition of a “general key to active electrocatalysts.” From

Equation (5), an optimum value of E�surface redox will maximize the product of the

preexponential and exponential-terms at an electrode potential of technical interest,

that is, at a low overpotential-versus-E�cell process. Consequently,E

�surface redox must not

be too far from E�cell process, to electroactivate the mediating surface system and

thereby ignite the faradaic process at a low overpotential. However, too small a

difference between the two standard potentials will mean a small free-energy drive

for the reaction of the reactant molecule with the active form of the surface redox

system [e.g., reaction (6b)], because the standard free-energy change in that reaction

is FðE�cell process�E�

surface redoxÞ. The activation energy of a process like (6b) is

expected to be lower, the higher the ðE�cell process�E�

surface redoxÞ difference and,

conversely, very close proximity of the two standard potentials will likely result

in excessiveDHact* . We are looking, therefore, at a need to optimize the gap between

E�surface redox and E�

cell process, to satisfy the conflicting requirements of a low over-

potantial for electrode surface activation and a sufficient free-energy drive for the

reaction of the reactant molecule with the active surface site.

On the basis of experimental results reported to date, the optimum value of

ðE�cell process�E�

surface redoxÞ for requiring electrocatalytic processes in low-temperature

fuel cells is in the range of 300–400mV. In the case of ORR at unalloyed Pt, for

example, (E�O2=H2O

�E�Pt=PtOx) is near 400mV and can be lowered further by about

FOREWORD ix

100mV by alloying [1], resulting in enhanced ORR activity. The rate enhancement

derived in this case from lowering of ðE�cell process�E�

surface redoxÞ indicates that the

beneficial effect of Pt alloying originates from lowering of the ignition overpotential,

resulting in an increase in the value of the preexponential term in Equation (5) at some

given cathode potential E. A metal surface where (E�O2=H2O

�E�M=MOx) is either

significantly smaller than 300mVor significantly higher than 400mVexhibits ORR

activity that is lower than that of Pt because it is associated with either high DHact* (in

the former case), or an excessive ignition overpotential (in the latter case). An

aggressive goal for the future would be to minimize further the difference between

the twoE� values. A surface redox systemwithE�surface redox removed less than 300mV

from E�cell process, while, at the same time, securing a low DHact

* for reaction of the

reactant molecule with the active surface site, will enable the onset of significant

current at overpotentials lower than those demonstrated to date. The reduction to

practice of suchadesirable surface function is highly challenging, however, becauseof

the low rates typically associated with processes driven by small changes in free

energy.

In summary, at a risk typical for all generalizations, a general rule for active fuel cell

electrocatalysts is proposed here, in the hope that it can provide a common ground for

the evaluation and development of new electrocatalysts. The rule is based on the

recognition that a wide variety of electrocatalytic processes, taking place at either

redox-functionalized or metal surfaces, are surface-redox-mediated, leading, in turn,

to the pursuit of an optimum value for (E�cell process�E�

surface redox) as the guideline for

maximizing the electrocatalytic activity. An optimized gap between these two

standard potentials best addresses the conflicting demands of a minimum over-

potential for surface activation and a high rate of the reaction between the reactant

molecule and the active surface site. Since themaximum rate is expected at an optimal

gap between the E� values, a plot of the rate of the electrocatalytic process versus

(E�cell process�E�

surface redox) will obviously take the famous form of a “volcano”;

however, this typical shape is now projected and explained in terms of a redox

mediation mechanism and the need to optimize the value of (E�cell process�E�

surface redox)

to achieve high rates at low overpotential. Enjoy the book!

S. GOTTESFELD

REFERENCES

1. H. A. Gasteiger and N. M. Markovi�c, Science 324(5923), 48–49 (2009).

2. J. Rossmeisel et al., inFuelCell Catalysis: A Surface ScienceApproach,M. T.M.Koper, ed.,

Wiley, Hoboken, NJ, 2009, pp. 57–93.

3. M. Lefevre et al., Science 324(5923), 71 (2009).

4. S. Gottesfeld, in Fuel Cell Catalysis: A Surface Science Approach, M. T. M. Koper, ed.,

Wiley, Hoboken, NJ, 2009, pp. 1–30.

x FOREWORD

PREFACE

The book covers some essential topics in the science of fuel cell electrocatalysis [1,2].

It shows an increase in importance of theory andmodeling, and the emerging newfield

of electrocatalysis science: bioelectrocatalysis. It shows a spectacular evolution of the

electrocatalysis concepts, froma simple statementof hydrogenevolution/oxidationon

platinum to reactions involving advanced nanoengineering and single-crystal sur-

faces, newmethods to study, and complicated chemical moieties (up to enzymes). It is

basically a materials/theory volume of chemical physics of fuel cell reactions,

including the electron transfer process and structure of the electric double layer, as

seen by a new generation of scientists, not necessarily electrochemists. It also shows

that operando measurements became possible because of the availability of synchro-

tron light. It forecasts thework for the future for the current and incominggenerationof

fuel cell scientists, namely, to use theory and understanding of the process involved

(see Chapter 19 and the Foreword), use the operando (advanced in situ) approach, and

expect theunexpected fromthe emergingnewfieldofbioelectrocatalysis.The future is

bright and exciting; the combination of the intellectual, high technology, and energy

issues makes us strong. We are looking forward.

AWacknowledges the splendid support by theNational ScienceFoundation and the

US Army Research Office toward his research in the preparation of this book.

J. NORSKOV

A. WIECKOWSKI

REFERENCES

1. S.-G. Sun, P.A. Christensen, and A. Wieckowski, eds., In-Situ Spectroscopic Studies of

Adsorption at the Electrode and Electrocatalysis, Elsevier, Amsterdam, 2007.

2. A. Wieckowski, E. Savinova, and C. Vayenas, eds., Catalysis and Electrocatalysis at

Nanoparticle Surfaces, Marcel Dekker, New York, 2003.

Note: Color versions of selected figures are available on ftp://ftp.wiley.com/

sci_tech_med/fuel_cell_catalysis.

xi

PREFACE TO THE WILEY SERIES ONELECTROCATALYSIS AND ELECTROCHEMISTRY

TheWiley Series on Electrocatalysis and Electrochemistry covers recent advances in

electrocatalysis and electrochemistry and depicts prospects for their contribution to

the industrial world. The series illustrates the transition of electrochemical sciences

from its beginnings in physical electrochemistry (covering mainly electron transfer

reactions, concepts of electrode potentials, and structure of the electrical double layer)

to a field in which electrochemical reactivity is shown as a unique aspect of

heterogeneous catalysis, is supported by high-level theory, connects to other areas

of science, and focuses on electrode surface structure, reaction environments, and

interfacial spectroscopy.

The scope of this series ranges from electrocatalysis (practice, theory, relevance to

fuel cell science and technology) to electrochemical charge transfer reactions,

biocatalysis and photoelectrochemistry. While individual volumes may appear quite

diverse, the series promises up-to-date and synergistic reports on insights to further the

understanding of properties of electrified solid/liquid systems. Readers of the series

will also find strong reference to theoretical approaches for predicting electrocatalytic

reactivity by high-level theories such as DFT. Beyond the theoretical perspective,

further vehicles for growth are provided by the sound experimental background and

demonstration of the significance of such topics as energy storage, syntheses of

catalytic materials via rational design, nanometer-scale technologies, prospects in

electrosynthesis, new research instrumentation, surface modifications in basic re-

search on charge transfer, and related interfacial reactivity. In this context, one might

notice that new methods that are being developed for one specific field are readily

adapted for application in another.

Electrochemistry has benefited from numerous monographs and review articles

due to its applicability in the practical world. Electrocatalysis has also been the

subject of individual reviews and compilations. TheWiley Series on Electrocatalysis

and Electrochemistry hopes to address the current activity in both of these comple-

mentary fields by containing volumes that individually focus on topics of current and

potential interest and application. At the same time, the chapters intend to demon-

strate the connections of electrochemistry to areas in addition to chemistry and

physics, such as chemical engineering, quantum mechanics, chemical physics,

surface science, biochemistry, and biology, and thereby bring together a vast range

of literature that covers each topic. While the title of each volume informs of the

specific concentration chosen by the volume editors and chapter authors, the integral

outcome of the series aims is to offer a broad-based analysis of the total development

xiii

of the field. The progress of the series will provide a global definition of what

electrocatalysis and electrochemistry are concerned with now and how these fields

will evolve overtime. The purpose is twofold; to provide a modern reference for

graduate instruction and for active researchers in the two disciplines, and to document

that electrocatalysis and electrochemistry are dynamic fields that are ever-expanding

and ever-changing in their scientific profiles.

Creation of each volume required the editor involvement, vision, enthusiasm and

time. The Series Editor thanks all the individual volume editors who graciously

accepted the invitations. Special thanks are for Ms. Anita Lekhwani, the Series

AcquisitionsEditor,whoextended the invitation to theSeriesEditor and is awonderful

help in the assembling process of the Series.

ANDRZEJ WIECKOWSKI

Series Editor

xiv PREFACE TO THE WILEY SERIES ON ELECTROCATALYSIS AND ELECTROCHEMISTRY

CONTRIBUTORS

Radoslav R. Adzic, Brookhaven National Laboratory, Upton, NY 11973

Matthias Arenz, Department of Chemistry, University of Copenhagen,

Copenhagen, Denmark

Fraser A. Armstrong, Department of Chemistry, Oxford University, South Parks

Road, Oxford OX1 3QR, United Kingdom

Renata Bilewicz, Faculty of Chemistry, University of Warsaw, Pasteura 1, 02-093

Warsaw, Poland

Nic�ephore Bonnet, Department of Materials Science and Engineering, Massachu-

setts Institute of Technology, Cambridge, MA 02139

Fikile R. Brushett, Department of Chemical and Biomolecular Engineering,

University of Illinois at Urbana—Champaign, Urbana, IL 61801

Juan Pablo Busalmen, Laboratorio de Bioelectroquımica, INTEMA(CONICET),

UNMdP. Juan B. Justo 4302, B7608FDQ, Mar del Plata, Argentina

Karen Chan, Department of Chemistry, Simon Fraser University, Burnaby, British

Columbia, Canada

Ismaila Dabo, Universit�e Paris-Est, CERMICS, Projet Micmac ENPC-INRIA, 6-8

Avenue Blaise Pascal, 77455 Marne-la-Vall�ee Cedex 2, France

Michael Eikerling, Department of Chemistry, Simon Fraser University, Burnaby,

British Columbia, Canada

AbrahamEsteve-Nunez, DepartamentodeQuımicaAnalıtica e IngenierıaQuımica,

Universidad de Alcal�a, Madrid, Spain

JuanMiguel Feliu, Instituto de Electroquımica, Universidad de Alicante, Apartado

de Correos 99, 03080 Alicante, Spain

Peter Ferrin, Department of Chemical and Biological Engineering, University of

Wisconsin—Madison, Madison, WI 53706

Dominic Gervasio, Department of Chemical and Environmental Engineering,

University of Arizona, Tucson, AZ 85721

Hendrik A. Heering, Leiden Institute of Chemistry, Leiden University, PO Box

9502, 2300 RA Leiden, The Netherlands

xv

Paul J. A. Kenis, Department of Chemical and Biomolecular Engineering, Uni-

versity of Illinois at Urbana—Champaign, Urbana, IL 61801

Michael L. Klein, Institute for Computational Molecular Science, College of

Science and Technology, Temple University, Philadelphia, PA 19122

Marc T.M. Koper, Leiden Institute of Chemistry, Leiden University, PO Box 9502,

2300 RA Leiden, The Netherlands

Carol Korzeniewski, Department of Chemistry and Biochemistry, Texas Tech

University, Lubbock, TX 79409

Heather J.Kulik, Department ofMaterials Science andEngineering,Massachusetts

Institute of Technology, Cambridge, MA 02139

Yanli Li, Universit�e Paris-Est, CERMICS, Projet Micmac ENPC-INRIA, 6-8

Avenue Blaise Pascal, 77455 Marne-la-Vall�ee Cedex 2, France

Nenad M. Markovic, Materials Science Division Argonne National Laboratory,

Argonne, IL 60439

Nicola Marzari, Department of Materials Science and Engineering, Massachusetts

Institute of Technology, Cambridge, MA 02139

Manos Mavrikakis, Department of Chemical and Biological Engineering,

University of Wisconsin—Madison, Madison, WI 53706

Agostino Migliore, Center for Molecular Modeling, Department of Chemistry,

University of Pennsylvania, Philadelphia, PA 19104

JensK.Nørskov, Department ofPhysics,Center forAtomic-ScaleMaterialsDesign,

Technical University of Denmark, DK-2800, Lyngby, Denmark

Marcin Opallo, Institute of Physical Chemistry, Polish Academy of Sciences, ul.

Kasprzaka 44/52, 01-224 Warsaw, Poland

Odysseas Paschos, Department of Physics, Technische Universit€at M€unchen,James Franck Strasse 1, D-85748, Garching, Germany

David E. Ramaker, Chemistry Department, George Washington University,

Washington, DC 20052

Jan Rossmeisl, Department of Physics, Center for Atomic-Scale Materials Design,

Technical University of Denmark, DK-2800, Lyngby, Denmark

Christina Roth, Institute for Materials Science, Technische Universit€at, Darmstadt,

Germany

Ata Roudgar, Department of Chemistry, Simon Fraser University, Burnaby, British

Columbia, Canada

Kotaro Sasaki, Brookhaven National Laboratory, Upton, NY 11973

xvi CONTRIBUTORS

Carlo U. Segre, Physics Division, Illinois Institute of Technology, 3101 S. Dearborn

St., Chicago, IL 60616

Patrick H.-L. Sit, Center for Molecular Modeling, Department of Chemistry,

University of Pennsylvania, Philadelphia, PA 19104 and Institute for

Computational Molecular Science, College of Science and Technology,Temple

University, Philadelphia, PA 19122

Eugene S. Smotkin, Department of Chemistry and Chemical Biology, 417 Hurtig

Hall, Northeastern University, Boston, MA 02116

Ulrich Stimming, Department of Physics, Technische Universit€at M€unchen, James

Franck Strasse 1, D-85748 Garching, Germany and ZAE Bayern Division 1,

Walther Meissner Strasse 6, D-85748 Garching, Germany

Yung-Eun Sung, World Class University Program of Chemical Convergence for

Energy and Environment, School of Chemical and Biological Engineering, Seoul

National University, Seoul 151-744, Korea

Hiroyuki Uchida, Clean Energy Research Center, University of Yamanashi,

4 Takeda, Kofu 400-8510, Japan

Miomir B. Vukmirovic, Brookhaven National Laboratory, Upton, NY 11973

Mitsuru Wakisaka, Fuel Cell Nanomaterials Center, University of Yamanashi,


Jia X. Wang, Brookhaven National Laboratory, Upton, NY 11973

Liya Wang, Department of Chemistry, Simon Fraser University, Burnaby, British

Columbia, Canada

Masahiro Watanabe, Fuel Cell Nanomaterials Center, University of Yamanashi,


Andrzej Wieckowski, Department of Chemistry, University of Illinois at Urbana—

Champaign, Urbana, IL 61801

Holger Wolfschmidt, Department of Physics, Technische Universit€at M€unchen,James Franck Strasse 1, D-85748 Garching, Germany

Sung Jong Yoo, Fuel Cell Center, Korea Institute of Science and Technology, Seoul

136-791, Korea

CONTRIBUTORS xvii

CHAPTER 1

Hydrogen Reactionson Nanostructured Surfaces

HOLGER WOLFSCHMIDT and ODYSSEAS PASCHOS

Department of Physics, Technische Universit€at M€unchen, Garching, Germany

ULRICH STIMMING

Department of Physics, Technische Universit€at M€unchen and ZAE Bayern Division 1,Garching, Germany

Hydrogen catalysis is an important scientific field since hydrogen reactions (e.g.,

hydrogen evolution and hydrogen oxidation) play a key role in electrochemical

devices such as fuel cells and electrolyzers. The latter devices have the potential to

provide clean and sustainable energy with high efficiencies. This chapter reviews

hydrogen catalysis in detail. Details on hydrogen reaction studies from theoretical and

experimental perspectives are presented. The former usually complement the results

from experimental studies and are used to strengthen them. Various systems that have

been explored throughout the years are reviewed. These include model surfaces as

well as applied systems. Model catalyst systems comprise Pt and Pd nanoislands

deposited on planar surfaces of inert supports, high-quality single-crystal materials,

or single nanoparticles created with scanning tunneling microscopy tips. Applied

systems consist of metallic nanoparticles deposited on high-surface-area carbon

supports. Theory versus experiment, and model versus applied systems are reviewed

in detail, and useful insights for hydrogen reactions in these systems are demonstrated

1.1 INTRODUCTION

Whereas the nineteenth century was the stage of the steam engine and the twentieth

centurywas the stageof the internal-combustion engine, it is likely that the twenty-first

century will be the stage of the fuel cell. Fuel cells have captured the interest of people

Fuel Cell Science: Theory, Fundamentals, and Biocatalysis,Edited by Andrzej Wieckowski and Jens K. NørskovCopyright � 2010 John Wiley & Sons, Inc.

1

around theworld as one of the next great energy alternative. They are nowon the verge

of being introduced commercially, revolutionizing the present method of power

production. Fuel cells can use hydrogen as fuel and oxygen or air as oxidant, offering

the prospect of supplying the world with clean, sustainable electrical power, heat,

and water.

This chapter focusesonhydrogen reactions such as thehydrogenoxidation reaction

(HOR) and the hydrogen evolution reaction (HER). These reactions are of utmost

importance in developing and improving fuel cell devices. The discussion here is

directed principally toward hydrogen electrocatalysis from an experimental as well as

theoretical perspective. Starting with an overview on the fundamentals of hydrogen

reactions in Section 1.2, studies on single crystals as well-defined and high-quality

surfaces are reviewed. An introduction to theoretical work calculating important

fundamentals for hydrogen catalysis regarding material properties is then discussed.

As predicted by theory, the behavior of nanostructured and bimetallic surfaces differs

from that of bulk material. Similar findings supporting the theoretical predictions are

shown for large nanostructured surfaces as well as single particles. The section

concludes with a short overview of carbon-based catalysts.

The fundamentals of hydrogen reactions are reviewed in Section 1.2. Starting from

thegeneral reversible hydrogen reaction, the different reaction pathways suggested by

Volmer, Heyrowsky, and Tafel are introduced. Because of the importance of the

hydrogen adsorption mechanism and the important findings with new experimental

techniques, a short overview of results obtained since the late 1990s is given. An

introduction to the correlation between catalytic behavior and the catalyst material

significance of this correlation, completes this section using experimental and

theoretical calculations, with a conclusion regarding the long-range.

Single crystals and well-defined surfaces play a very important role in surface

science. Many scientific contributions are available that study these well-defined

surfaces. Section 1.3 introduces the electrochemical behavior toward hydrogen

reactions on Pt, Au, and Pd surfaces. The quality of single crystals rapidly increased

in the 1990s, resulting in newand different insights. Because of the importance of Pt as

a catalyst, themain part of this section focuses on this element. The dependence of the

crystallographic orientation toward adsorption as well as electrocatalytic activity is

discussed. An introduction to Pd as a catalyst material with the property to absorb

hydrogen and Au as an inert support material is the last topic in that section.

Besides experimental work, numerous theoretical calculations for hydrogen

catalysis have been performed. Computational methods such as density functional

theory (DFT) and Monte Carlo simulations are powerful tools in surface science and

catalysis. Theoretical as well as experimental work has been combined in several

scientific publications and complement each other well. The first principles of

theoretical techniques and theoretical results are shown in Section 1.4. As a main

topic, the adsorption behavior of hydrogen is considered and the d-band model is

introduced. Calculations regarding the hydrogen oxidation reaction and the influence

of different reactions pathways are also shown. Theoretical calculations of metals on

thinfilms and supported onvarious foreignmetals are reviewed and are correlatedwith

experimental findings.

2 HYDROGEN REACTIONS ON NANOSTRUCTURED SURFACES

The chemical behavior ofmetal nanoparticles often differs from that of bulkmetal.

Different effects such as particle size, interparticle distance, and support effects have

to be considered in this nanometer-scale regime. Since Pd and Pt are important

materials in catalysis, much work was done in the last few decades describing the

abovementioned effects. In particular, multilayers, monolayers, and submonolayers

of Pd and Pt onto foreign metal supports have shown unexpected behavior. Pd

on Au(111) regarding several electrochemical properties introduces this section.

Different types of adsorption, absorption, and desorption behavior as well as

electrocatalytic activity toward hydrogen reactions are shown and discussed. The

deposition of Pd onother supports and the influence of hydrogen reactions hindered by

adsorbing foreign adsorbates as well as investigations of Pt overlayers onAu(111) are

also discussed. A summary and detailed discussion in Section 1.5 also includes

theoretical aspects.

As mentioned above, planar surfaces are thoroughly investigated and serve as

widely accepted reference systems with high-quality, reproducible results. For local

investigation of small structures, new approaches and setups have to be designed

and applied. For this purpose, the electrochemical scanning tunneling microscope

(EC-STM) has been modified by several groups in order to create small nanoparticles

and nanoparticle arrays and also to investigate corrosion, deposition, dissolution, and

reactivity. Due to the tunneling effect, high resolution is achievable and thus leads to a

precise techniquewith atomic resolution. TheSTM tip can be used in different ways in

the electrochemical environment to create and investigate local reactivity of nanos-

tructures. Experimental and theoretical results are compared and are shown to

complement each other. Specifically, the activity of a single Pd particle is shown.

Adiscussion of the experimental results of the stability of Pd particles deposited onAu

(111) and their reactivity towardHER follows.A summary completes Section 1.6with

a comparison between results obtained at extended Pd nanostructured Au(111)

surfaces and single Pd particles.

Section 1.7 presents an overview of studies performed on carbon-based systems.

Since carbon has high electrical conductivity, is relatively inexpensive to use, and is

highly available, it has been the favored support material for many years. Of the many

scientific contributions, only a few can be presented here regarding the mechanism of

HER and HOR using metallic nanoparticles with carbon-based supports. The reac-

tivity of these catalysts for hydrogen reactions and CO oxidation is also of major

interest. These catalyst systems include glassy carbon, carbon nanofibers, Vulcan,

and carbon black for support for metallic nanoparticles, and the more highly oriented

and defined pyrolytic graphite (HOPG) are also presented and discussed.

1.2 FUNDAMENTALS OF HYDROGEN REACTIONS

1.2.1 Hydrogen Catalysis

Over the years a number of studies have been performed in order to investigate

the characteristics of hydrogen-related reactions. The general reversible reaction is

as follows:

FUNDAMENTALS OF HYDROGEN REACTIONS 3

Hþ þ e� $ 12H2 ð1:1Þ

Its standard potential is set to 0V. In the case of proton discharge to formmolecular

hydrogen the reaction is called a hydrogen evolution reaction (HER), while the

reverse pathway describes the hydrogen oxidation reaction (HOR). However, for the

reaction to proceed at sufficient rate, it needs to be catalyzed on an electrode surface.

Possible catalyst candidates include various metals such as Pt, Pd, and Ru, as well as

enzymes with active centers. Much research focused on finding parameters that

influence the activity of materials toward hydrogen electrocatalysis. Even though

much progress has been made on this matter, it is still not clearly known how

various properties influence the catalytic activity. More details will be given later

in this section.

Today it is generally accepted that hydrogenevolutiononPtoccursvia twodifferent

pathways consisting of every two reaction steps:

Discharge reaction of a proton to form an adsorbed hydrogen atom, known as the

Volmer reaction [1]:

PtþHþ þ e� !H�Pt ð1:2ÞCombination of two adsorbed hydrogen atoms to form molecular hydrogen,

known as the Tafel reaction [2]:

2ðH�PtÞ!H2 þ 2Pt ð1:3ÞCombination of an adsorbed hydrogen atom with a proton and an electron to form

molecular hydrogen, known as the Heyrovsky reaction [3]:

H�PtþHþ þ e� !H2 þ Pt ð1:4Þ

Two different pathways can occur; the first one is described as a combination of

reactions (1.2) and (1.3), known as the Volmer–Tafel mechanism. With this mecha-

nism, protons from the solution are discharged on the catalyst surface, forming

adsorbed hydrogen atoms. Then, two adjacent adsorbed hydrogen atoms combine to

form molecular hydrogen. The second mechanism, known as the Volmer–Heyrovsky

mechanism, can be described by combining reactions (1.2) and (1.4). A proton from

electrolyte solution is discharged on the catalyst surface to forman adsorbed hydrogen

atom. This step is followed by combination with another proton and electron to form

molecular hydrogen.

Hydrogen oxidation reaction on Pt can be described in a similar way using the

reaction pathways in reverse order:

Dissociation of molecular hydrogen into one adsorbed hydrogen atom and

immediate discharge of the other atom into proton and electron, similar to

the Heyrovsky reaction:

H2 þ Pt!H�PtþHþ þ e� ð1:5Þ


Adsorption of molecular hydrogen on the catalyst surface in the form of two

hydrogen atoms, similar to the Tafel reaction:

H2 þ 2Pt! 2ðH�PtÞ ð1:6Þ

Discharge of an adsorbed hydrogen atom to proton and electron, similar to the

Volmer reaction:

H�Pt! PtþHþ þ e� ð1:7Þ

Similar to hydrogen evolution, the hydrogen oxidation reaction can follow two

different pathways. The first mechanism is a combination of reactions (1.5) and (1.7).

A hydrogen molecule is positively charged (H2 !Hþ2 ), and immediately one of its

atoms is discharged into proton and electron,while the other is adsorbed on the surface

of the catalyst. Then the adsorbed hydrogen atom is discharged into proton and

electron. The second one is a combination of reactions (1.6) and (1.7). With this

mechanismmolecular hydrogen is adsorbed on the catalyst surface in the form of two

hydrogen atoms, followed by discharge of the atoms into proton and electron.

1.2.2 Hydrogen Adsorption Mechanism and Experimental Setups

Aswas shown above, in both hydrogen reactions (oxidation and evolution) the step of

forming a hydrogen adsorbate on the catalyst surface exists in both pathways.

Research was performed in order to study the mechanism of hydrogen adsorption

onPt single crystals. Pt is one of themostwidely studied catalysts because of its ability

to catalyze hydrogen reactions with small overpotentials. Initial studies focused on

determining the heat of adsorption of hydrogen on Pt(111) single crystals. An

interesting review was published by Markovic and Ross [4], who showed the values

for the heat of hydrogen adsorption reported in early years to be inconsistent.

Christmann and Ertl [5] reported in 1976 that the value for Pt(111) is approximately

equal to 50–60 kJ/mol. However, McGabe and Schmidt [6] in 1977 and Salmeron

et al. [7] in 1979 reported higher values, between 70 and 90 kJ/mol. Later, it was found

that these corresponded to adsorption of hydrogen on defect sites. Until relatively

recently it was accepted that hydrogen tends to adsorb on highly coordinated sites,

which for the case of Pt(111) would be the threefold hollow sites. These would lead,

though, to a very high coverage of two hydrogen atoms per Pt; therefore, in order to

ensure agreement with experimental values, it was accepted that hydrogen occupies

the threefold next-nearest-neighbor sites (for details, see Section 1.3). Olsen et al. [8],

performing DFT calculations, showed that hydrogen tends to occupy the top sites.

Nevertheless, in all cases the values reported were close to each other.

Depending on the overpotential, the adsorbed hydrogen atom on the catalyst

surface is referred to as under- or overpotential deposited hydrogen (Hupd or Hopd).

Hupd refers to hydrogen atoms adsorbed at potentials positive of the reversible

hydrogen electrode (RHE) potential, while Hopd occurs at potentials negative of the

RHEpotential. The state ofHupd andHopd depends also on the pHof the electrolyte and


is generated from either protons or water molecules [4] and can be described by the

following reactions:

PtþH3Oþ þ e� ! Pt�Hupd þH2O ðpH � 7Þ ð1:8Þ

PtþH2Oþ e�Pt�Hupd þOH� ðpH � 7Þ ð1:9Þ

There are several possible reasonswhy reported experimental values sometimes do

not agree. The first would be the quality of the single crystal. Crystals having defect

sites or impurities adsorbed on their surfaces act differently toward electrochemical

reactions. Also, it has been shown that different single crystal faces of Pt have different

electrocatalytic rates. Markovic and Ross [4] showed that the activity increases in the

order of (111) < (100) < (110) (Fig. 1.1).Barber et al. [11] showed a slightly different result (Fig. 1.2), where the activity for

HER/HOR increases in the order of (100) < (111) < (551) < (110).However, despite these small differences, it is clearly shown that the activity is

strongly affected by the orientation of the Pt single crystal.

The experimental technique also plays a determining role on the results obtained

mainly with the appearance of a limiting current density above certain overpotentials

for the hydrogen reactions. Especially for the case of HORon Pt, the exchange current

density is high in acidic solutions, but simultaneously, because of the low solubility of

hydrogen, the limiting diffusion current is low.Quaino et al. [12] showed that by using

Levich–Koutecky analysis the j(g) dependence for HOR cannot always be obtained

accurately and may be underestimated. Bagotzky and Osetrova [13] were the first

to propose an alternative experimental setup that had the potential to solve many

issues related to the investigation ofHOR. Their setup consisted of Ptmicroelectrodes

2

0

-2

-4

-6-0.2 0.0

i[mA

cm-2

]

0.2 0.4

(a) 0.05 H2SO4, 274 K

0.6 0.8

Pt(110) Pt(100) Pt(111)

FIGURE 1.1 Polarization curves for HER and HOR on Pt(hkl) in 0.1MHClO4 at sweep rate

20mV/s [4].


(�A� and �B�) embedded in fused-glass tubes with two different polished surfaces.

Because of the small thickness of the electrode, therewas an enhanced mass transport

of hydrogen. Therefore, values for limiting diffusion current that were one order of

magnitude higher than those obtained from RDE setups could be reached. However,

the results obtained were affected by the roughness of the microelectrode surface,

which can be clearly seen in Figure 1.3.

0.2

25i / m

A c

m-2

(geo

m)

50

75

0.4ϕr / v

1

4

3

2

0.6 0.8 1.0

FIGURE 1.3 Dependence of hydrogen ionization current on potential on microelectrodes

with different roughness valuesA andB in 0.5MH2SO4: (1)A, (2)B and in 1MKOH: (3)A,

(4) B [13].

-0.20

-0.15

-0.10

-0.05

0.0010-3 10-2 10-1 100

(100)SI (111) (511) (110)

current-density / A cm-2

E v

s. R

HE

/ V

101 3232323232

FIGURE1.2 DerivedTafel plots, less the diffusion effect, for the Pt (100)SI, (511), (111), and

(110) faces as marked on the plot [11].


Two electrodes with different roughness values, differentiated by the degree of

polishing,were used to study hydrogen oxidation in both acidic and alkaline solutions,

and as can be seen, the results are different for the two electrodes. Quaino et al. [14]

used a similar setup to studyhydrogenoxidation. Theywere able to demonstrate that at

low overpotentials the Tafel–Volmer route dominates the kinetics of HOR. At high

overpotentials the Tafel–Volmer effect diminishes while the Volmer–Heyrovsky

mechanism becomes dominant.

Also, traces of impurities that can be present in unpurified solutions can compete

with the reactions under certain conditions, especially at low current densities [15],

resulting in misleading interpretation of the results.

1.2.3 Correlation between Activity toward Hydrogen Reactionsand Physicochemical Properties of Catalyst Material

In the early years research was focused on finding a relationship between the activity

toward hydrogen evolution and oxidation and a property of the catalyst. Conway and

Bockris [16] reported a correlation between the exchange current density j0 and the

electronic workfunction f.Workfunction is defined as the energy with which electrons

near the Fermi level are bound to the material. According to their study [16], the

relationship between j0 and f arises from the dependence of heat of adsorption on f.

Additionally, they showed that the bond strength between adsorbed hydrogen and

metal calculated from Pauling�s equation was smaller than the one obtained from

experiments. Theyalso, as canbeen seen inFigure1.4, usingvalues from the literature,

demonstrated that for various metals (e.g., Ta, Mo, W, Cu, Ni, Fe, Rh, Pd, Pt) the

logarithm of j0 increases as the heat of adsorption of H decreases, while an opposite

trend is observed for Hg, Cd, Pb, and Tl.

For HER, it was also shown (Fig. 1.5) that the logarithm of j0 increases as the d

character of the material increases.

The latterwas explained by the fact that as the d character increases,more electrons

have paired spins and hence require more energy to extract, them causing DH of

adsorbed hydrogen atoms to decrease.

Parsons [17] studied the relationship between exchange current density and

the ability of the electrode to adsorb atomic hydrogen in terms of the standard free

energy DGH. His theoretical studies showed that log j0 reaches a maximum when

DGH� 0. Even though he mentions a disagreement between experimental and

theoretical results (a similar disagreement is also mentioned by Trasatti [18] for the

heat of adsorption, the observed trend should still be valid. Metals that adsorb

hydrogen weakly (DGH has positive values), such as Hg, Zn, and Sn, have low

exchange current densities. Metals such as Pt that adsorb moderately hydrogen have

high values of j0 andmetals that adsorb hydrogen strongly, such asMo, Ta, andW, also

have low j0 values.

It was shown that there is dependence between the exchange current density for

hydrogen reactions and the workfunction of the catalyst material. However, work-

function values were usually used by electrochemists as obtained from physical

experiments. These values were usually measured using nonelectrochemical inputs


such as adsorption of gas hydrogen on metals and without taking into consideration

the chemical environment surrounding the catalyst. Trasatti [18] published an

interesting review on several aspects in order to obtain more accurate data regarding

the correlation of hydrogen reactions to physicochemical properties of materials.

He argued that the sign of the charge of electrode surface is usually ignored. If the

exchange current density j0 is plotted versus the workfunction (Fig. 1.6), then two

fairly parallel lines can be obtained.

One line consists of data from transition metals and sp metals with positively

charged surfaces,while the other includes data from spmetalswith negatively charged

surfaces. It is also noteworthy that the lines are approximately 0.4 eVapart from each

other. Trasatti explained the division of materials in these two groups in terms of

orientation ofwatermolecules on the catalyst surface. As is shown in Figure 1.7, if the

surface is positively charged, then water will be positioned with an oxygen atom

toward the metal, whereas in the case of a negatively charges catalyst surface, an

opposite orientation is expected.

3.5

−3

−4

−5

−6

−7

−8

−9

−10

−11

−12

−13 xPb

Hg

Nbx

TI

xAIxCd

MoXxCd

CuFex

w

xAu Au

FeNi

Ag

Pt (high c.d.)Rh

Pd

Pt (low c.d.)

4.0 4.5φ IN ELECTRON VOLTS

LOG

10 I 0

(A

mp

Cm

−2)

5.0 5.5

FIGURE 1.4 Linear dependence of log10 of the exchange current (i0) of HER on the

electronic workfunction f. Values of log10 i0 are taken from the literature [16].


Although the plot shows a clear difference between transition and sp metals, it

includes no information regarding themechanismof reaction. This information can be

factored in only if the exchange current density is plotted versus the heat of adsorption

of hydrogen on themetal surface.Asmentioned previously, the rate andmechanismof

HER depends on the bond strength between themetal and the hydrogen atom (M�H).

Parsons reported that it should pass through amaximum, and a similar volcano-shaped

curve was reported by Krishtalik and Delahay [19], as shown in Figure 1.8.

As shown inFigure 1.8, Pt is on the top of thevolcano curvewhere the Pt�Hbond is

neither too strong nor too weak. The general trend observed in the volcano curve is

that for several metals, as the bonding energy of hydrogen to the metal increases,

the activity also increases, reaching a maximum. Then an opposite trend is

observed, where log j0 decreases as the bonding strength of hydrogen to the metal

increases.

A similar study was done by Nørskov et al. [20]. Density functional theory (DFT)

calculations demonstrated a volcano-type behavior of hydrogen chemisorption

energies with respect to exchange current density for hydrogen evolution (Fig. 1.9).

Platinum was again found to be a better catalyst than other metals for HER

primarily because hydrogen evolution reaction on Pt is thermoneutral at the equilib-

rium potential. The findings of this work can be used to predict behavior of other

bimetallic systems for HER as well as HOR. The analysis was reported as a new

method to obtain H adsorption free energies and understand trends for different

systems that are of electrochemical interest.

-8

-7To

Fe

Ni

W

Mo

Pt (high c.d.)

Pt (low c.d.)

PdRh

-6

-5

-4

-3

-236 38 40

% d – CHARACTER42

LOG

10 I 0

(A

mp

Cm

-2)

44 46 48 50

FIGURE 1.5 Log10 i0 for HER as a function of percent d character of the metal [16].


fuel cell science - download.e-bookshelf.de€¦ · fuel cell catalysis: a surface science...

Documents