nanostructured and photoelectrochemical systems for solar photon conversion, 2008, p.781

Nelson 7.2 Basic principles of photovoltaic conversion in organic materials 457 7.5 Relationship between material and device parameters and photovoltaic performance 473 8 Dye-sensitised

NANOSTRUCTURED AND PHOTOELECTROCHEMICAL

SOLAR PHOTON CONVERSION

SERIES ON PHOTOCONVERSION OF SOLAR ENERGY

Series Editor: Mary D Archer (Cambridge, UK)

Vol 1: Clean Electricity from Photovoltaics

eds Mary D Archer & Robert Hill

Vol 2: Molecular to Global Photosynthesis

eds Mary D Archer & Jim Barber

Vol 3: Nanostructured and Photoelectrochemical Systems

for Solar Photon Conversion

eds Mary D Archer & Arthur J Nozik

Forthcoming

From Solar Photons to Electrons and Molecules

by Mary D Archer

Imperial College Press ICP

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library.

Imperial College Press

57 Shelton Street

Covent Garden

London WC2H 9HE

Distributed by

World Scientific Publishing Co Pte Ltd.

5 Toh Tuck Link, Singapore 596224

USA office: 27 Warren Street, Suite 401-402, Hackensack, NJ 07601

UK office: 57 Shelton Street, Covent Garden, London WC2H 9HE

Printed in Singapore.

For photocopying of material in this volume, please pay a copying fee through the Copyright Clearance Center, Inc., 222 Rosewood Drive, Danvers, MA 01923, USA In this case permission to photocopy is not required from the publisher.

ISBN-13 978-1-86094-255-6

ISBN-10 1-86094-255-5

All rights reserved This book, or parts thereof, may not be reproduced in any form or by any means, electronic or mechanical, including photocopying, recording or any information storage and retrieval system now known or to be invented, without written permission from the Publisher.

Copyright © 2008 by Imperial College Press

Series on Photoconversion of Solar Energy — Vol 3

NANOSTRUCTURED AND PHOTOELECTROCHEMICAL SYSTEMS FOR

SOLAR PHOTON CONVERSION

This volume is dedicated

1A

The vacuum scale of electrode potential and the concept of the

solution Fermi level

24

2.4 Charge transfer at the semiconductor–electrolyte interface 84

3.3 Optical spectroscopy of quantum wells, superlattices and quantum dots 163 3.4 Hot electron and hole cooling dynamics in quantum-confined

semiconductors

167 3.5 High conversion efficiency via multiple exciton generation in quantum dots

176

4 Fundamentals and applications in electron-transfer reactions

M D Archer

4.5 Semiclassical theories of nonadiabatic electron transfer 232 4.6 Electron transfer in donor–bridge–acceptor supermolecules 238

4.9 Optimisation of photoinduced electron transfer in photoconversion 263 App

4A

App

4B

Derivation of high-temperature limit Marcus rate equation for

homogeneous electron transfer using density-of-states approach

266

5 Fundamentals in metal-oxide heterogeneous photocatalysis

N Serpone and A V Emeline

5.2 The complex science underlying metal-oxide photocatalysis 277 5.3 Metal-oxide photochemistry, photophysics and modelling 310

5.6 Evidence for a gas/solid surface reaction being photocatalytic 374

6.3 Preparation of substrates, absorber and transporting layers 403

7 Organic donor–acceptor heterojunction solar cells

J J Benson-Smith and J Nelson

7.2 Basic principles of photovoltaic conversion in organic materials 457

7.5 Relationship between material and device parameters and photovoltaic performance

473

8 Dye-sensitised mesoscopic solar cells

M Grätzel and J R Durrant

8.6 Pilot production of modules, outdoor field tests and commercial DSSC development

527

9 Semiconductor/liquid junction photoelectrochemical solar cells

S Maldonado, A G Fitch and N S Lewis

10 Photoelectrochemical storage cells

S Licht and G Hodes

10.6 High-efficiency multiple-bandgap cells with storage 622

12 Experimental techniques in photoelectrochemistry

L M Peter and H Tributsch

12.3 Photocurrent, photovoltage and microwave reflectance methods 683

12.11 Combination of electrochemistry with vacuum spectroscopy 725

Appendices

xi

Xin Ai received her BS degree from Jilin University, Changchun, China In 2004, she

obtained her PhD in chemistry from Emory University, Atlanta, Georgia, where she worked on the investigation of photoinduced interfacial electron-transfer dynamics on dye molecule and inorganic semiconductor nanocomposite films using femtosecond infrared spectroscopy She then joined National Renewable Energy Laboratory, Golden, Colorado, as a postdoctoral associate Her primary research interest is in the photochemical and photoelectrochemical properties of novel molecular materials, including conjugated polymers, carbon nanotubes and quantum dots, which have been used to fabricate a new generation of solar cells She currently focuses on fundamental understanding of the photoinduced interfacial charge-transfer processes occurring in these materials, using femtosecond transient spectroscopy, terahertz spectroscopy and time-resolved and steady-state photoluminescence spectroscopy The goal of her work is to understand the factors affecting the efficiency of photovoltaic cells and, through this understanding, to provide insight into improving the performance of the working devices

Mary Archer read chemistry at Oxford University and took her PhD from Imperial

College, London, in 1968 From 1968 to 1972, she did post-doctoral work in chemistry with Dr John Albery at Oxford, and she then spent four years at The Royal Institution in London, working with Lord Porter (then Sir George Porter) on photo-electrochemical methods of solar energy conversion She taught chemistry at Cambridge University from 1976 to 1986 From 1991 to 1999, she was a Visiting Professor in the Department of Biochemistry at Imperial College, London, and from

electro-1999 to 2002, she held a Visiting Professorship at ICCEPT (Imperial College Centre for Energy Policy and Technology) She is President of the UK Solar Energy Society and the National Energy Foundation and a Companion of the Energy Institute She was awarded the Melchett Medal of the Energy Institute in 2002 and the Eva Philbin Award of the Institute of Chemistry of Ireland in 2007

Jessica Benson-Smith was awarded the British Marshall Scholarship in 2004 As a

recipient of this fellowship, she became a postgraduate student in the Experimental Solid State Physics Department at Imperial College, London, from which she received her PhD in 2007 She specialises in the spectroscopy of organic bulk heterojunction films for organic solar cell applications

James Durrant is Professor of Photochemistry in the Department of Chemistry at

Imperial College, London After completing his undergraduate studies in physics at the University of Cambridge, he obtained a PhD in biochemistry at Imperial College, London, in 1991, studying the primary reactions of plant photosynthesis After postdoctoral positions and a BBSRC Advanced Fellowship, he joined the Chemistry Department at Imperial College in 1999 His interests are in photochemical approaches to solar energy conversion, electron-transfer dynamics and excitonic solar cells

Alexei Emeline obtained his MSc in Physical Chemistry from Tomsk State University

in 1990 and his PhD in Molecular Physics from St Petersburg State University in

1995 He started his academic research career at St Petersburg State University as a researcher at the V A Fock Institute of Physics in 1995, and later in the same year as

an assistant professor of the Faculty of Physics In 1996 he was awarded a NATO Science Fellowship to take up a post-doctoral fellowship at Concordia University in Canada under the supervision of Professor Nick Serpone From 1998 to 2004, he remained in the same laboratory at Concordia University as an associate researcher In

2005 he was awarded a JSPS Fellowship and spent one year in the group of Professor Akira Fujishima at Kanagawa Academy of Science and Technology in Japan He is currently a senior researcher, working for his DSc in the V A Fock Institute of Physics of St Petersburg State University His research interests focus on fundamental studies of interfacial photophysical and photochemical processes in heterogeneous systems, particularly on the role of photoexcitation conditions on the direction and efficiency of different photoprocesses

Anthony Fitch received his undergraduate degree at the University of Nebraska at

Kearney and is currently pursuing his PhD at Caltech under the advisement of Nathan

S Lewis

Michael Grätzel is a professor at the École Polytechnique de Lausanne, where he

directs the Laboratory of Photonics and Interfaces He discovered a new type of solar cell based on dye-sensitised mesoscopic oxide particles and pioneered the use of nanocrystalline materials in electroluminescent and electrochromic displays, as well as lithium ion batteries and bioelectronic sensors Author of over 500 publications, two books and inventor of more than 50 patents, his work has received over 40,000 citations so far, ranking him amongst the most highly-cited scientists worldwide He has received several prestigious awards, including the Faraday Medal of the (British) Royal Society of Chemistry, the Dutch Havinga award, the Italgas prize, the European

Millennium award for Innovation, the 2006 World Technology Award in Materials and the Gerischer award In 2006, he was selected by Scientific American as one of

the fifty top researchers in the world He received his doctor’s degree in Natural

Science from the Technical University, Berlin, and holds honorary doctorates from the Universities of Delft, Uppsala and Turin He is a member of the Swiss Chemical Society and the European Academy of Science and an elected honorary member of the Société Vaudoise de Sciences Naturelles

Gary Hodes received his BSc and PhD from Queen’s University of Belfast in 1968

and 1971 respectively, and has been at the Weizmann Institute of Science, Rehovot, Israel since 1972 His research has focused on semiconductor film deposition from solutions (initially electrochemical and later chemical bath deposition) and on various types of solar cells (liquid junction, thin film, polycrystalline and nanoporous) and quantum dots using these films Throughout his career, he has also studied various aspects of semiconductor surface treatments More recently, he is continuing work on various aspects of chemical bath deposition mechanisms and also increasingly concentrating on nanocrystalline, semiconductor-sensitised solar cells

Rolf Könenkamp is the Gertrude-Rempfer Professor of Physics at Portland State

University in Portland, Oregon His present research interests lie in the field of science He has worked extensively on semiconductor devices, such as nanostructured solar cells and nanowire light-emitting diodes and transistors, and he holds several patents in this area He has led the design and construction of a new high-resolution photoelectron microscope since 2002 This will be one of the first aberration-corrected microscopes of this type and it will be used to explore transport and confinement effects on the nanoscale He has worked at NREL, HMI Berlin, Hitachi Tokyo, Princeton University and at the IST in Lisbon, and he is a member of the national R&D team for thin-film photovoltaics in the US

nano-Nathan Lewis is George L Argyros Professor of Chemistry at the California Institute

of Technology, where he has been on the faculty since 1988 He has also served as the Principal Investigator of the Beckman Institute Molecular Materials Resource Center

at Caltech since 1992 From 1981 to 1986, he was on the faculty at Stanford, as assistant professor from 1981 to 1985 and associate professor from 1986 to 1988 He received his PhD in Chemistry from the Massachusetts Institute of Technology He has been an Alfred P Sloan Fellow, a Camille and Henry Dreyfus Teacher–Scholar and a Presidential Young Investigator He received the Fresenius Award in 1990, the ACS Award in Pure Chemistry in 1991, the Orton Memorial Lecture award in 2003

and the Princeton Environmental Award in 2003 He has published over 200 papers and supervised about 50 graduate students and postdoctoral associates His research interests include light-induced electron transfer reactions, both at surfaces and in transition metal complexes, surface chemistry and photochemistry of semi-conductor/liquid interfaces, novel uses of conducting organic polymers and polymer/conductor composites and development of sensor arrays that use pattern recognition algorithms to identify odorants, mimicking the mammalian olfaction process

Tianquan Lian received his BS degree from Xiamen University in 1985, his MS

degree from the Chinese Academy of Sciences in 1988 and his PhD from the University of Pennsylvania in 1993 After postdoctoral training in the University of California at Berkeley, he joined the faculty of chemistry department at Emory University in 1996 He was promoted to associate professor in 2002 and full professor

in 2005 He has been a recipient of the NSF CAREER award and the Sloan fellowship His research interest is focused on the ultrafast dynamics of nanomaterials and interfaces He is particularly interested in fundamental physical chemistry problems related to nanomaterials-based solar energy conversion concepts and devices These problems include the dynamics of electron transfer, energy transfer, vibrational energy relaxation and solvation at interfaces and in nanomaterials

Stuart Licht has over 250 publications in renewable energy chemistry, physical

chemistry and analytical chemistry, and was the recipient of the 2006 Electrochemical Energy Research Award He has developed theory and experiment for the highly efficient solar generation of hydrogen fuel, introduced the contemporary use of caesium to enhance solar cell voltage and established the chemistry of an efficient solar cell that functions day and night He has originated the field of Fe(VI) redox chemistry for charge storage (the ‘Super-Iron Battery’), as well as novel sulphur batteries and a variety of new aluminium electrochemical storage cells En route to new pathways to utilise renewable energy, the Licht group continues to explore a range of fundamental physicochemical processes ranging from quantum mechanics to thermodynamics of water, hydrogen, halide, chalcogenide and transition metal chemistry, and to introduce new analytical methodologies, in dilute, concentrated or molten media, as needed to facilitate the research Licht has chaired a regional section

of the national American Chemical Society and also founded, and chaired, the New England, and the Israel, Sections of the Electrochemical Society

Stephen Maldonado was a Beckman Scholar in 2000–2001 for his work on proton

exchange membrane fuel cell system testing After receiving a BS in Chemistry from the University of Iowa in 2001, he was awarded an NSF Fellowship and a Huntington Fellowship for graduate studies at the University of Texas at Austin His thesis work centred on designing electrocatalytically active graphitic carbon nanotubes In 2006,

he obtained his PhD in Chemistry and joined the research laboratory of Professor Nathan S Lewis as a postdoctoral research scientist at the California Institute of Technology His current research focuses on the electrical and electrochemical properties of metal–silicon contacts using chemically modified silicon

Rüdiger Memming obtained his PhD degree in Physical Chemistry from the

University of Stuttgart, Germany, in 1958, working with Professor Förster, and then did post-doctoral work at the Chemistry Department of the University of Minnesota, Minneapolis, working with Professor R S Livingston for two years In 1960, he started to work in the Philips Research Laboratory in Hamburg, Germany, where he continued until 1987 In addition, he had a research group at the Chemistry Department of the University of Hamburg from 1981 to 1987 After this he started a new government Institute for Solar Energy Research in Hanover, from which he retired in 1994 In 1991, he went to Japan for four months as a JSPS-Fellow

R J Dwayne Miller obtained a BSc Honours degree in 1978 from the University of

Manitoba and his PhD degree in Chemistry from Stanford University in 1983, and then did post-doctoral work as a NATO Science Fellow at the Université de Joseph Fourier, Grenoble He started his academic research career at the University of Rochester in 1984, where he was a faculty member in Chemistry and the Institute of Optics He relocated his research group to the University of Toronto in 1995, where

he is currently the Director of the Institute for Optical Sciences and full professor in the Departments of Chemistry and Physics He is a Fellow of the Royal Society of Canada and the holder of the Canada Research Chair in femtoscience His early research interests focused on the primary events controlling electron transfer at surfaces This work demonstrated how truly fast electron-transfer processes can be at conducting surfaces, and led to his current research, which focuses on femtosecond electron pulse generation to give atomic-level views of transition state processes He welcomed the opportunity to return to his roots to write this review on the photophysical processes at semiconductor surfaces with the hope that this overview will help researchers solve the last hurdles to economically viable solar power

Jenny Nelson is a Professor of Physics at Imperial College, London, where she has

researched novel types of solar cell since 1989 Her current research focuses on photovoltaic energy conversion using molecular materials, characterisation of the charge transport, charge separation and morphological properties of molecular semiconductors, the theory of charge transport in organic semiconductors and modelling of photovoltaic device behaviour She has published over 100 papers on photovoltaic materials and devices and a book on the physics of solar cells

Arthur Nozik graduated from Cornell University in Chemical Engineering in 1959

After a brief spell in the aerospace industry, he entered Yale University to work for a PhD in physical chemistry The birth of his daughter caused him to intermit these studies and join the American Cyanamid Company, but he returned to Yale and finished his PhD in 1967 He then returned to Cyanamid for seven years, introducing Mössbauer spectroscopy to the company In 1974, he joined Allied Chemical Corporation to work on semiconductor photoelectrochemistry as applied to solar photoconversion At Allied, he became the first to demonstrate the ‘zero bias’

photoelectrolysis of water, using an n-TiO2 photoanode and a p-GaP photocathode, and also the photoreduction of dinitrogen on p-GaP He also developed the

‘photochemical diode’, the forerunner of today’s particulate semiconductor suspensions In 1978, he moved to the new Solar Energy Research Institute (now NREL) at Golden, Colorado, where he was Branch Chief of the Photoconversion Branch, 1980–1984, and has been a Senior Research Fellow since 1984 He was Team Leader of the NREL Chemical Sciences Team from 1985 to 2006 and he has been Professor Adjoint at the University of Colorado at Boulder since 1998 At NREL, his research has centred on the behaviour of hot carriers in quantum wells, superlattices and quantum dots In 2005, thirty years of work in solar photon conversion were rewarded when he and his research group demonstrated efficient multiple exciton generation in lead chalcogenide quantum dots He was awarded the 2008 Eni Award for Science and Technology

Laurie Peter gained his PhD in Southampton in 1969, before being awarded a CIBA

Postdoctoral Fellowship to work in the Group of Heinz Gerischer, who was then at the Technische Hochschule in Munich Subsequently, he moved with Gerischer’s group

to the Fritz Haber Institute in Berlin, where he remained as a member of staff until

1975, when he returned to the UK to take up a lectureship in Southampton He remained in Southampton for the next 17 years and was promoted to professor before moving to Bath in 1992 to become Professor of Physical Chemistry and subsequently

Head of Department Laurie Peter was an editor of the Journal of Electroanalytical

Chemistry from 1999 to 2005, and has been awarded the Electrochemistry Prize of the

Royal Society of Chemistry and the Pergamon Medal of the International Society of Electrochemistry He currently leads the UK SUPERGEN Excitonic Solar Cell Consortium (Bath, Cambridge, Edinburgh and Imperial College), which is studying non-classical solar cells

Nick Serpone obtained a BSc Honours Chemistry in 1964 from Sir George Williams

University in Montreal and his PhD degree in Physical-Inorganic Chemistry from Cornell University in 1968, following which he joined the chemistry faculty of Concordia University as Assistant Professor His early research involved NMR studies

of Group IV coordination complexes After sabbatical leaves at the University of Bologna, Italy, in 1975 and at the École Polytechnique Fédérale de Lausanne, Switzerland, as an invited professor in 1983, his research interests focused on the photochemistry of coordination complexes and on fundamental and applied studies in heterogeneous photocatalysis in which, together with others, he has been instrumental

in developing the technology to degrade environmental organic pollutants and to dispose of toxic metals In 1981, he co-founded the Canadian Centre for Picosecond Laser Spectroscopy at Concordia University and was its director until 2002 His other principal research interests have involved studies of ultra-fast photophysical and photochemical events in metal chalcogenide and silver halide semiconductors Following his appointment as a University Research Professor and Professor Emeritus

in 1998, he joined the Chemistry Division of the National Science Foundation in Washington DC as an IBO Program Director from 1998 to 2001 He was a Visiting Professor in Italy’s programme ‘Rientro dei Cervelli’ at the University of Pavia from

2002 to 2005, where he carried out research into the photochemistry of sunscreen active agents

Helmut Tributsch obtained his PhD degree in physical chemistry at the Technical

University, Munich, in 1968, working with Heinz Gerischer, and subsequently continued research with Melvin Calvin at the University of Berkeley For the next ten years he worked in different institutions including Stanford University, the CNRS in Paris and the Fritz-Haber Institute in Berlin Since 1982, he has been Professor of Physical Chemistry at the Free University in Berlin and head of the department at the Hahn-Meitner Institute specialising in research on sustainable energy systems

xix

Thus daily were my sympathies enlarged, And thus the common range of visible things Grew dear to me: already I began

To love the sun, a Boy I lov’d the sun, Not as I since have lov’d him, as a pledge And surety of our earthly life, a light Which while we view we feel we are alive;

But, for this cause, that I had seen him lay His beauty on the morning hills, had seen The western mountain touch his setting orb,

In many a thoughtless hour, when, from excess

Of happiness, my blood appear’d to flow With its own pleasure, and I breath’d with joy

William Wordsworth, The Prelude: Book 2: School-Time, 1805

More solar energy falls on the Earth’s surface every day than the total amount of energy the world’s population would consume in 16 years at present rates of utilisation To harness this potential to provide reliable and economic carbon-free sources of electricity and fuels remains a challenge, even in current times of high energy prices and action to mitigate climate change However, there are encouraging signs The annual global market for photovoltaic (PV) modules was valued at US$12.9bn in 2007 and is predicted to grow by 15% compound per annum Although

crystalline silicon p–n junction cells still dominate this market, a new generation of

photovoltaic and photoelectrochemical devices is emerging to challenge them, many based on the unique properties of matter at the nanoscale

It is this new generation of solar photon conversion devices that are covered in this book They are less highly developed than those described in Volumes 1 and 2 of this series, but their promise is at least as great That promise is two-fold: on the one hand highly efficient devices with sophisticated architectures in which the Shockley–Queisser limit on efficiency is finally overcome, and on the other very low-cost plastic or organic-based devices that are cheap enough to be disposable

The leitmotifs of these devices include bespoke dye sensitisers, space-quantised nanoscale structures that enable hot carrier or multiple exciton generation, molecular and solid-state junction architectures that lead to efficient exciton dissociation and charge separation, and charge collection by percolation through porous or mesoscale phases Another common theme underlying the devices discussed in this book is the

orthogonalisation of the pathways for photon absorption and carrier collection Contrast the classical silicon solar cell, in which the two pathways are parallel with an ETA or bulk heterojunction cell, in which they are orthogonal In the silicon cell, the base layer has to be sufficiently thick to absorb incoming photons, so minority carrier diffusion lengths have to be (and are) as long as 200–500 µm, placing great demands

on materials quality In an ETA or bulk heterojunction cell, the junction architecture allows efficiencies of over 5% to be achieved with exciton or charge carrier diffusion lengths that are as much as one million times shorter, and materials of much lower electronic quality suffice

Photocatalysis is closely related to photoelectrochemistry, and the fundamentals of both disciplines are covered in this volume Their applications to photoelectrolysis and other solar fuel-forming or waste-destroying photochemical and photoelectro-chemical processes will form the main subject matter of the fourth and final volume in this book series

To satisfy the global need for carbon-free energy, the fields of photovoltaics and photoelectrochemistry must continue to develop The key to progress lies in the quality of the fundamental research being conducted in this area It is worrying that global funding streams for research to develop advanced solar photon conversion technologies remain fragile despite the concerted and powerful case for a ‘Manhattan project’ effort to do so made by the international scientific community during a special conference in 2005 on basic research needs for solar energy utilisation promoted by the US Department of Energy’s Office of Science, Basic Energy Sciences Division However, commercialisation of some of these devices is beginning, and a January 2008 report from BCC Research predicts that the market for nanostructured thin films and silicon and dye-sensitised solar cells is set to grow at more than 50% per annum through to 2013 as the technology matures

Our warmest appreciation goes to our fifteen authors, who between them have provided so rich a picture of the scientific frontiers they are exploring We also thank Alexandra Anghel, Carol Burling, Barrie Clark and Stuart Honan for their editorial assistance, David Ginley, John Kelly and Reshef Tenne for providing information, James Bolton for his early input into some of the material in Chapters 1 and 4, the staff of World Scientific Press who expertly drew many of the diagrams, and Lenore Betts, Lizzie Bennett and Katie Lydon of IC Press for guiding us along the winding road to publication

Mary D Archer

March 2008

The major themes of this book are announced by its title: nanostructured and

photo-electrochemical systems for solar energy conversion It deals mainly with the direct, i.e

non-thermal, conversion of solar photonic energy into electrical power by electrochemical or advanced photovoltaic means in extended-junction, mesoporous, nanocomposite or space-quantised structures and devices Other themes are the fundamentals of electron transfer and photoinduced electron transfer in supramolecular assemblies, photocatalytic reactions at semiconductor dispersions, and experimental techniques for the characterisation of semiconductor photoelectrochemical systems Semiconductors have been the electrode materials of choice for solar photon conversion for nearly thirty years, on account of their favourable optoelectronic

photo-properties and chemical versatility Semiconductor bandgap energies E g commonly fall in

the range 1–3 eV, which overlaps well with the spectrum of terrestrial sunlight, as shown

in Fig 9.1, and also with the decomposition potentials of such important reactions as water splitting, as shown in Fig 2.17 Absorption by a semiconductor of photons of energy greater than the bandgap energy leads to the creation of ‘free’ holes and electron (in broadband inorganic semiconductors) and excitons (in organic semiconductors) At the junction of a photovoltaic device, these free carriers or excitons are separated into a flow of electrons in one direction and a flow of holes in the other at a potential difference determined by the light intensity and the junction characteristics, leading to the generation of electric power on illumination

Photoelectrochemical cells for solar photon conversion are usually designed to produce either electric power or solar fuels; this book focuses on the latter Power-producing solar cells are designed to be operated at their maximum-power point to produce electric power at the energy conversion efficiency ηmp

mp mp

o

i V E

where E is the incident solar irradiance, ioS mp is the maximum-power photocurrent

density and Vmp is the maximum-power voltage The ratio between the maximum power

generated and the product of the short-circuit photocurrent density i sc and the open-circuit

voltage Voc is known as the fill factor, ηfill The higher the value of ηfill, the better the quality of the device

mp mp fill

photocurrent must be carried to the junction by electrons through p-type material, and by holes through n-type material Minority carriers are highly susceptible to bulk recomb-

ination, as well as to trapping and interfacial recombination A high level of materials quality and fastidious attention to cell design and fabrication are therefore needed to endow minority carriers in a silicon cell with sufficient lifetime to reach and flow across the junction without loss by hopping or recombination The minimum thickness of a photovoltaic cell is determined by the width of the absorber layer needed to absorb incident light efficiently Since crystalline silicon is an indirect-gap material, it is not intensely absorbing, and so a comparatively thick wafer of it is required to absorb incident sunlight efficiently, even with such refinements as surface texturisation, internal light scattering and back-surface reflection to increase the optical path length of light in the cell Thus the excellent performance of the classical silicon photovoltaic cell is in some ways a triumph of materials and device optimisation over basically unfavourable materials characteristics Few other inorganic semiconductors, and no organic semi-conductors, are capable of being developed to deliver similarly good performance in a

photovoltaic cell of classical, planar-junction architecture Moreover, in a classical p–n

1

Conversion efficiencies are, or should be, quoted for standard test conditions, which are 1000 W m–2 of AM1.5 global insolation and a cell temperature of 25 C

junction cell the same material is required both to absorb light and to permit charge transport along the same dominant parallel pathway, which is perpendicular to the planar junction, as shown in Fig 1.1a

Extended-junction and nanostructured photoconversion devices can escape from these constraints by orthogonalising the pathways for light absorption and charge collection, as illustrated in Fig 1.1b The pathways for charge collection are much shorter, allowing the use of inexpensive low-quality materials, and also of organic semiconductors in which light absorption generates not free charge carriers but short-lived excitons that must reach an interface in order to separate at it and generate photocurrent Additional and important advantages of nanosized semiconductor structures and particles are the increased carrier lifetimes arising from space quantisation, the enhanced redox potentials of photogenerated holes and electrons arising from the increased effective bandgap and the possibility of multiple exciton generation by one absorbed photon in a quantum dot

In this chapter, I give an account of the historical development of semiconductor photoelectrochemistry and nanostructured photovoltaic devices in Section 1.2, and then Sections 1.3–1.6 provide a brief introduction to the major cell types discussed in the remainder of the book: the ETA (extremely thin absorber) cell, organic and hybrid cells, dye-sensitised solar cells (Grätzel cells) and regenerative solar cells

In Chapter 2, Miller and Memming present an advanced treatment of the solid-state physics and photoelectrochemistry of semiconductors In Chapter 3, my co-editor Art Nozik covers the fundamentals and applications of quantum-confined structures and explains how the unique ability of quantum dots to generate multiple pairs of charge carriers with a single high-energy photon could lead to a new generation of photovoltaic cells In Chapter 4, I turn to electron-transfer theory, and how its understanding through

Figure 1.1 (a) Classical planar n-on-p photovoltaic cell junction, showing the dominant parallel direction of

the light and charge separation pathways; (b) Extended, structured junction with interposed absorber layer (shaded in grey), showing the dominant non-parallel direction of the light and charge separation pathways

+ +

n p n absorber p

(a) (b)

the powerful prism of Marcus theory has led to the design and synthesis of molecular dyads, triads and polyads with optimised hole–electron lifetimes and energies, which might in future be linked into energy-funnelling antennae or nanoscopic current-collecting systems to create molecular power-producing cells Nick Serpone and Alexei Emeline provide a comprehensive account of the fundamentals of metal-oxide heterogeneous photocatalysis, with particular emphasis on dispersed titanium dioxide systems, in Chapter 5

In Chapters 6–10, we turn to important cell types: inorganic extended-junction devices are described by Rolf Könenkamp, who has pioneered their development, in Chapter 6, and polymer and polymer-composite cells by Jenny Nelson and Jessica Benson-Smith in Chapter 7 In Chapter 8, dye-sensitised solar cells are discussed by James Durrant and their inventor, Michael Grätzel Another authority in the field of solar photoelectrochemistry, Nate Lewis, and two colleagues, Stephen Maldonado and Anthony Fitch, provide an overview of non-dye-sensitised semiconductor/liquid junction solar cells in Chapter 9 In Chapter 10, Stuart Licht and Gary Hodes describe their own and others’ development of photoelectrochemical storage (PECS) cells, which have the conceptual advantage over the other types of power-producing cell described in this volume of being able to produce continuous rather than intermittent power Figure 1.2 shows how the performance of all these cell types has improved over time Finally, Xin

Ai and Tianquan Lian deal with the measurement of electron-transfer dynamics at the molecule/semiconductor interface in Chapter 11, and Laurie Peter and Helmut Tributsch cover techniques for the characterisation of photoelectrochemical systems in Chapter 12 One type of photoelectrochemical device not covered in this book is the photo-galvanic cell By this term is meant power-producing or storage cells in which the products of an endoergonic photoredox reaction that occurs in solution are harvested at metal (or at any rate photoinactive) electrodes Although Albery and Archer (1977) took

a sanguine view of the maximum power conversion efficiency (5–9%) that might be obtained from such a cell, subsequent experimental studies have shown that the combination of long optical lengths, low diffusivities of short-lived redox products and imperfect electrode selectivity in practice restrict conversion efficiencies to well below 1%, rendering the photogalvanic cell impractical as a power-producing device (Archer and Ferreira, 1980)

Each chapter is comprehensively referenced, and the reader may also find some of the following recent reviews and books helpful: Fujishima and Zhang (2005), Soga (2006),

Durrant et al (2006), Hodes (2007), Kamat (2007) and Licht (2007) The Festschrift issue of the Journal of Physical Chemistry (Vol 100, No 50, 21 December 2006) in

honour of my co-editor’s seventieth birthday also contains many papers of relevance

Figure 1.2 Best efficiencies by year of power-producing photoelectrochemical and nanostructured cells:
regenerative solar cells (RSC): 1 Heller (1981); 2

Gibbons et al (1984): 3 Tufts et al (1987); 4 Licht and Peramunage (1990); 5 Licht et al (1998); 6 Licht et al (1999); • photoelectrochemical cells with storage (PECS): 1 Ang and Sammells (1980); 2 Keita and Nadjao (1984); 3 Licht et al (1987); 4 Licht et al (1999); — liquid dye-sensitised solar cells

(LDSSC): 1 O’Regan and Grätzel (1991); 2 Nazeeruddin et al (2001); 3 Chiba et al (2006); — solid-state dye-sensitised solar cells (SSDSSC): 1 Tennakone

et al (1995); 2 Tennakone et al (1998); 3 Schmidt-Mende et al (2005); 4 Snaith et al (2007); • organic photovoltaic cells (OPV): 1 Tang (1986);

2 Granstrom et al (1998); 3 Shaheen et al (2001); 4 Brabec et al (2002); 5 Li et al (2005); 6 Reyes-Reyes et al (2005); 7 Peet et al (2007); 8 Kim et al.

(2007);
extremely thin absorber cells (ETA): 1 Ernst et al (2003); 2 Nanu et al (2005) All efficiency values are for standard or near-standard AM1.5G insolation; PECS4, RSC5, RSC6 and OPV8 are multijunction devices; all the others are single-junction devices

1.2 Historical perspective

The first recorded observations of a photoelectrochemical (PEC) phenomenon were made

by the young Henri Becquerel (Becquerel, 1839), who noted the photocurrent and photovoltage produced by sunlight acting on silver chloride-coated platinum electrodes in various electrolytes Having satisfied himself that these effects were not thermal, Becquerel postulated that they resulted from a solid-state photochemical reaction, for which he obtained rough spectral response curves by the use of colour filters Becquerel’s interpretation of the photocurrent as a measure of ‘le nombre de rayons chimiques’ (the number of chemical rays) was vigorously challenged by the powerful Academician Jean Baptiste Biot (Biot, 1839), but over the next twenty years, Becquerel’s view prevailed as

he continued to work on silver halide-coated electrodes, developing a chemical light meter based on AgCl (Becquerel, 1841–1867)

photoelectro-Pre-1940 PEC work on semiconducting films and electrodes was comprehensively

reviewed by Copeland et al (1941) Early work—mainly on the halides, oxides and

sulphides of Ag, Cu and Hg in aqueous solution—was complicated by the photochemical instability of some of the materials used, notably the silver halides Early workers generally interpreted the effects they saw in terms of photochemical reactions, particularly the pH-dependent photolysis of water, light-induced increases in solubility, activation of molecules by light and ‘local cell’ electrochemical activity, but gradually it was recognised that physical rectifying effects analogous to those in solid-state barrier

cells were involved (e.g Goldmann and Brodsky, 1914; Fink and Alpern, 1930; Müller

and Spector, 1932; Roulleau, 1935, 1937)

Time-resolved studies of the photopotential of cyanine-sensitised AgBr revealed a

biphasic response, ascribed by Sheppard et al (1929, 1940) to the differing mobilities of electronic and ionic carriers Work at Eastman Kodak by Leermakers et al (1937)

established the correspondence between the spectral photosensitisation of AgBr emulsions by cyanines and the absorption spectra of the dyes The tendency for open-circuit photopotentials to increase with decreasing temperature was observed by Athanasiu (1925–1935), and the logarithmic dependence of photopotentials on irradiance

by Sichling (1911) Several amperometric ‘electrolytic photoelements’ based on the photocurrent produced by the action of light on semiconductor electrodes, such as the Rayfoto and Arcturus Photolytic Cells (Cu|Cu2O|Pb(NO3)2(aq)|Pb or Cu2O), were commercialised in the 1930s (Fink and Alpern, 1930; Wilson, 1938; Lange, 1938; Fink and Adler, 1940)

During the forties and early fifties, Veselovskii (1941–1952) carried out extensive investigations on AgBr/Ag and oxidised Zn, Fe, Pb, Ag, Au and Pt, confirming by measurements of spectral sensitivity and quantum yield that the observed photo-

electrochemical effects were produced within a fairly thick oxide layer, and not by a surface reaction Hillson and Rideal (1949) attempted to resolve the disputed mechanisms

of hydrogen and oxygen evolution at various oxidised metal electrodes from photocurrent measurements PEC effects were observed in Cu2O (Kalita, 1935; Kasgkarev and Kosonogoya, 1948), oxidised Pb (Ginzburg and Veselovskii, 1952), Se (Pittman, 1953) and chlorophyll-sensitised ZnO, CdO and PbS (Putseiko, 1953)

Much pre-1955 work, carried out on semiconductors of uncontrolled properties and purity, was at best qualitative Only after the crucial role of purity in controlling semi-conductor properties had been recognised, and the nature of holes had been distinguished from that of positive ions, did systematic work on semiconductor electrodes become possible The modern era of inorganic semiconductor electrochemistry began at Bell Telephone Laboratories with Brattain and Garrett’s classical (1955) work on the current–

voltage characteristics of n- and p-type Ge electrodes in aqueous solutions of KOH, KCl

and HCl This established that the current was controlled by the surface concentrations of holes and electrons, which were in turn controlled by the applied bias and could be increased by illumination Dewald’s (1959, 1960a) lucid expositions of the principles of semiconductor electrochemistry laid the foundation for rapid experimental advances in the sixties, when many important concepts were established: the relation between the sign

of the photopotential and the conductivity type of the electrode (Williams, 1960); the concept of the flatband potential (Dewald, 1960b); valence-band and conduction-band electron-transfer kinetics (Gerischer, 1960, 1966); the mechanisms of photocorrosion (Williams, 1960; Turner, 1960) and suppression of semiconductor corrosion by common ions in solution (Barker, 1966); electrode surface states (Boddy and Brattain, 1962;

Lazorenko-Manevich, 1962); work on n-Se and p-Se (Gobrecht et al., 1959); KTaO3(Boddy et al., 1968); oxygen evolution on illuminated anodically biased n-TiO2 (Boddy, 1968) and SnO2 (Möllers and Memming, 1972); and hydrogen evolution on illuminated

cathodically biased p-GaP (Beckmann and Memming, 1969)

Despite these advances, semiconductor photoelectrochemistry remained the domain

of a few specialists until the seminal announcement by Fujishima and Honda (1971,

1972) (prefigured in Fujishima et al., 1969) of the sustained photoelectrolysis of water by the use of an n-TiO2 photoanode, followed by the surge of interest in renewable energy

produced by the 1973 oil price shock Fujishima and Honda’s 1972 Nature paper,

although reporting no hitherto unknown phenomena, reoriented research towards the glittering prize of solar photoelectrochemical water splitting and power production using inexpensive polycrystalline semiconductor electrodes Since then, the pursuit of this and related goals has transformed semiconductor photoelectrochemistry from a specialist domain into a major interdisciplinary subject

The use of nanoscale constructs has given a further major boost to solar photon conversion The scale of nanosized materials such as quantum dots and nanotubes, conventionally taken to lie in the range 1–100 nm, produces very interesting size quant-isation effects in optoelectronic and other properties: bandgaps shift to the blue, carrier lifetimes increase, potent catalytic properties emerge and constructs with very high surface-to-volume ratios can be made Incorporation of nanoscale structures in photovoltaic devices allows these unique properties to be exploited, with conversion efficiencies above the detailed balance limit becoming possible in principle

Nanosized TiO2 powders are of outstanding importance in this context Aqueous suspensions of ~30 nm particulate TiO2 (mostly in the rutile form) are the active agent in many of the photocatalytic systems described by Serpone and Emeline in Chapter 5 Agglomerations of TiO2 nanoparticles into mesoporous films of pore size 2–50 nm which allow the penetration of liquid are the basis of the important dye-sensitised solar cell (DSSC) discussed by Grätzel and Durrant in Chapter 8, as well as most of the hybrid devices described by Nelson and Benson-Smith in Chapter 7, and some of the ETA (Extremely Thin Absorber) cells described by Könenkamp in Chapter 6 The term ‘eta-

solar cell’ was actually introduced by Könenkamp and co-workers (Siebentritt et al.,

1997), who had earlier used the term ‘sensitisation cell’ for the same type of device (Wahi and Könenkamp, 1992) Precursors to ETA cells with liquid electrolytes as hole

conductors were developed by Vogel et al (1990), Ennaoui et al (1992) and Weller

(1993) Similar electrolytic cells with RuS2 (Ashokkumar et al., 1994) and InP (Zaban

et al., 1998) nanoparticle absorbers have also been demonstrated

Size quantisation effects in the photoelectrochemistry of semiconductors were first

demonstrated by Nozik et al (1985) in a strained-layer superlattice electrode with 40

alternating layers of undoped GaAs and GaAs0.5P0.50 (in which charge carriers were spatially confined in one dimension in the GaAs quantum wells) Quantum dots (in which charge carriers are spatially confined in three dimensions) were first used in photo-

electrochemistry by Kietzmann et al (1991) to sensitise TiO2 in Grätzel-type cells, and in

a photovoltaic device by Greenham et al (1996), who observed quantum efficiencies of

up to 12% in conjugated polymer/CdX (X = Se or S) nanocrystal composite solar cells Multiple exciton generation in quantum dots was first observed by Schaller and Klimov (2004) in PbSe QDs, and efforts to incorporate quantum-dot arrays into next-generation solar cells are now underway (Service, 2008)

The history of organic semiconductors goes back further, although only in recent years have organic solar cells revealed their full potential The first observation of high

conductivity in an organic polymer was made by Bolto et al (1963) in polypyrrole that

had been partially oxidised by iodine doping Semiconducting device (FET) properties

were first noted in organic materials by Heilmeier et al (1964) in copper phthalocyanine,

and later by McGinness et al (1974) in melanin, a naturally derived material which is a

complex copolymer of polyacetylene, polypyrrole and polyaniline In a later paper,

Shirakawa et al (1977) reported high conductivity in partially oxidised polyacetylene,

work for which he and his colleagues Alan MacDiarmid and Alan Heeger won the 2000 Nobel Prize in Chemistry

These discoveries have created a new industry based on organic semiconductors, one spin-off from which has been the development of organic solar cells These function rather differently from ‘classical’ inorganic cells Photon absorption in an organic semi-conductor generates, not ‘free’ charge carriers as in a broad-band inorganic semi-conductor, but excitons which do not spontaneously separate but will dissociate at a heterojunction yielding separate charges, provided the interface presents a sufficient electrochemical potential drop The first ‘modern’ organic solar cell was the planar bilayer device of Tang (1986), but this cell architecture is limited because the exciton must be generated within a few diffusion lengths of the heterojunction to have a reasonable chance of being collected Exciton diffusion lengths are typically <10 nm, and this limits the thickness, and hence the absorbance, of a planar bilayer cell The highly folded architecture of the dispersed or bulk heterojunction (BHJ) cell, which came onto

the scene in the 1990s, overcomes this problem Halls et al (1995) published one of the

first studies showing that a dispersed heterojunction, formed from a phase-separated blend of donor-type and acceptor-type polymers in which the domain size is of the order

of an exciton diffusion length, greatly improved the quantum efficiency of exciton

dissociation and charge generation Yu et al (1995) reported a quantum efficiency of

29% and an energy conversion efficiency of about 2.9% for a BHJ cell containing a blend

of poly(2-methoxy-5-(2′-ethyl-hexyloxy)-1,4-phenylene vinylene (MEH-PPV) with C60

Yoshino et al (1997) introduced a three-layer structure into a polymer donor–acceptor blend cell, which improved performance further Hummelen et al (1995) developed

solubilised fullerenes such as [6,6]-phenyl C61-butyric acid methyl ester (PCBM; Fig 7.1 shows its structure), thus enabling the casting of well-blended bulk heterojunctions of two phases from a single solvent

Further breakthroughs in organic solar cell performance came in 2000–2001 Stephen Forrest and co-workers achieved 2.4% efficiency in a Tang-like copperphthalocyanine/ perylene tetracarboxylic derivative cell by incorporating an exciton-blocking layer between the photoactive organic layers and the metal cathode to prevent exciton

quenching at the cathode (Peumans et al., 2000) The Linz and Groningen groups

improved the efficiency of the conjugated polymer/methanofullerene cell to 2.5% by

using a more intimate blend of the two phases (Shaheen et al., 2001) At the same time,

Michael Grätzel improved the performance of the all-solid-state dye-sensitised solar cell

to 2.56% by blending the hole conductor spiro-OMeTAD with 4–tert–butylpyridine and

Li(CF3SO2)2N to control junction recombination (Krüger et al., 2001)

In the next four sections of this chapter, we shall look at the operating principles of those cells that are the subject of specialist chapters later in the book All are designed to produce electric power Most have highly extended interfaces (as illustrated in Fig 1.1)

to increase light absorption and/or charge separation Given the steady improvements in efficiency shown in Fig 1.2, there is a good prospect of further efficiency gains and wider commercial application of one or more of these newer cell types

1.3 Extremely thin absorber (ETA) cells

ETA (Extremely Thin Absorber) cells are the subject of Chapter 6 by Könenkamp, and here we provide only a brief introduction and overview An ETA cell is an inorganic all-

solid-state photovoltaic cell with a structured electroconducting substrate, usually an

n-type semiconducting oxide such as TiO2 or ZnO, and a solid hole-conducting superstrate,

usually a p-type semiconductor such as CuSCN, CuAlO2, PEDOT or spiro-OMeTAD In

between the two is sandwiched an extremely thin (~5–150 nm) layer of a strongly absorbing inorganic semiconductor such as a-Si:H (Wahi and Könenkamp, 1992), CdTe

or CdS (Siebentritt et al., 1997) or CuInS2 (Möller et al., 1998) The absorber can be in

the form of either a continuous layer or nanoparticles of a semiconductor such as PbS

(Wienke et al., 2003) In some devices, the hole conductor is omitted and the absorber is

contacted directly by a metal contact

Figure 1.3 shows the two most common ETA cell structures, in which the substrate morphology is either mesoporous (Fig 1.3a) or columnar (Fig 1.3b) As in the dye-sensitised cells discussed in Section 1.5, the extended interface interposes many individual absorber layers in the path of incoming light, allowing good total light absorption although each individual absorber layer is very thin The interfacial morphology should be rough on the wavelength scale of sunlight (~500 nm), so that the incoming light beam is strongly scattered, thus increasing the optical path length of light within the cell Light is further trapped by internal reflection in the absorber layers The holes and electrons that are generated by light absorption within the absorber layer have only a very short distance to travel to the interface with the electron and hole conductor, respectively, allowing efficient carrier collection from absorber materials with poor transport properties Hence low-cost materials and inexpensive fabrication processes can

be used to make ETA cells As Könenkamp explains in Chapter 6, the main difficulty in fabricating ETA cells is achieving conformal, void-filling deposition of the absorber on the substrate, and then of the hole conductor (if present) on the absorber

Figure 1.4 shows the band structure of a typical ETA cell Electronically speaking,

these are p-i-n junction cells, the light absorber being the i-layer and the built-in field

persisting throughout the thin transport layers under most bias conditions Incoming light generates hole–electron pairs in the absorber, which separate into free carriers, as is normal in inorganic semiconductors The field in the absorber layer assists carrier transport and hence very short carrier diffusion lengths (as low as ~10 nm) can be tolerated The photogenerated electrons are injected into the conduction band of the electron conductor, and the holes into the valence band of the hole conductor These separated charges then travel across the transport layers to the contacts under majority-carrier transport conditions without the risk of recombination However, the extended junction, despite its optical advantages, may also be responsible for the dominant loss mechanism in most ETA cells, namely interface recombination at defect and boundary sites of the large solid/solid interface areas

At their present state of development, ETA cells typically exhibit open-circuit voltages of 0.6–0.7 V and photocurrents of 5–15 mA cm–2 While the fill factors in the earlier ETA cells were often poor, typically ~20%, more recent cells have improved fill factors, typically ~60% Overall, the solar conversion efficiencies are modest, as shown

in Fig 1.2; to date ηmp values of approximately 5% have been obtained (Nanu et al.,

2005) Insertion of an ultrathin tunnel barrier layer of an insulator such as Al2O3 or MgO

can improve the open-circuit voltage (Wienke et al., 2003) Incorporation of quantum

glass TCO

back contact

ETA (absorber) hole conductor electron conductor

Figure 1.3 Typical ETA cell geometries The interfaces between the absorber and electron and hole transport

layers are structured, usually in porous (Fig 1.3a) or columnar (Fig 1.3b) form The interfaces between the transport layers and the contact layers are planar If the substrate morphology is porous, both transport layers should be transparent to avoid shadowing effects The contact layer on the light entry side must be transparent and the back contact should be reflective, to minimise optical losses outside the absorber layer If the interfacial structuring is not very deep, it is possible to omit the hole transport layer, and deposit the back contact straight onto the ETA layer, which greatly simplifies device fabrication.

dots within the absorber layer may enhance carrier lifetimes, as well as introducing the possibility of a two-state excitation process (Luque and Marti, 1997) or multiple exciton generation (as discussed in Chapter 3)

The technology of ETA cells is not yet mature, and future improvements in their efficiency can reasonably be expected Modelling calculations (Taretto and Rau, 2005) indicate that 15% efficient CdTe ETA cells are possible even at electron diffusion lengths

as low as 10 nm, provided that the built-in voltage is optimised to restrict recombination over the working bias range of the cell If efficiency improvements of this order can be achieved and fabrication methods can be satisfactorily scaled up, ETA cells could offer a low-cost, stable alternative to traditional photovoltaic cells and dye-sensitised solar cells

1.4 Organic solar cells

Organic solar cells are photovoltaic devices containing thin (typically ~100 nm) films of light-absorbing organic semiconductors such as conjugated polymers or small molecules The term ‘organic solar cell’ is something of a misnomer, since the some of the most efficient devices contain (essentially inorganic) fullerene derivatives as electron

Figure 1.4 Band structure in an ETA cell, showing the majority-carrier quasi-Fermi levels E F,n and E F,p in the

electron and hole conductor, respectively, and the photovoltage V The conduction-band edge of the absorber

must be above that of the electron conductor, while the valence-band edge of the absorber must be below that of

the hole conductor The conduction-band and valence-band offsets ∆E c and ∆E v between the electron or hole conductor and the absorber should be sufficient to suppress back electron or hole transfer, but not so great as to

cause a substantial loss of photovoltage; ∆E c and ∆E vvalues of ~0.2–0.3 V are optimal

hole conductor absorber layer

acceptors Nevertheless, the term ‘organic solar cell’ and its equivalent, ‘organic voltaics’ (OPV), are a convenient shorthand which we, like others, shall use for both purely organic and hybrid organic/inorganic devices

photo-Jessica-Benson Smith and Jenny Nelson cover OPV in detail in Chapter 7, and here again we provide only a brief introduction Polymer-based solar cells have been recently

reviewed by Mayer et al (2007) Figure 1.5 shows two typical cell architectures for OPV

(Fig 7.4 shows two more) The first (Fig 1.5a) is a simple bilayer device in which an electron-donating layer and an electron-accepting layer of two different organic semiconductors form a planar junction at which light-generated excitons dissociate into charge carriers and are collected at the electrodes The bilayer copperphthalocyanine/ perylene tetracarboxylic derivative (CuPc/PTC) cell developed by Tang (1986) was the pioneer in this field, with a power conversion efficiency of 0.95% Currently the best-performing bilayer cells have efficiencies of around 5% (see Fig.1.2), and contain a vacuum-deposited molecular layer of a metal phthalocyanine as the light absorber and electron donor and a fullerene derivative as electron acceptor

The other common OPV cell type, shown in Fig 1.5b, is the bulk heterojunction (BHJ) cell, sometimes called the dispersed heterojunction cell These cells contain intimate three-dimensional blends of the electron-donating and electron-accepting

Figure 1.5 Typical organic photovoltaic cell architectures (a) Bilayer cell; (b) bulk heterojunction cell

glass TCO (hole collector)

back contact (electron collector) acceptor phase donor phase

exciton

donor phase

acceptor phase

– –

(a) (b)

+ +

materials, which creates a very large-area interface between the two phases Provided the phase domain size is of the order of twice the exciton diffusion length, the probability of exciton capture and dissociation is much increased The blend may be random, as in Fig 1.5b, or it may be vertically graduated or ordered as in the right-hand diagram of Fig 1.3 Currently the most efficient BHJ cells are made by co-deposition from solution

of a solubilised electron-donating polythiophene polymer together with a solubilised electron-accepting fullerene derivative At the time of writing, the world record efficiency of 5.5% for a single-junction organic solar cell is held by the polymer/

fullerene derivative cell made by Alan Heeger’s group (Peet et al., 2007), in which a few

volume per cent of alkanedithiols is incorporated to improve the bulk heterojunction morphology The related tandem (double-junction) BHJ cell developed by Heeger’s

group (Kim et al., 2007), which has a front cell of a low-bandgap polymer/fullerene

composite connected by a transparent TiO2 layer to a back cell of a lower-bandgap polymer/fullerene composite, has a reported efficiency of 6.5% for light intensity of

200 mW cm–2 Ordered heterojunctions have not so far yielded such good performance: the best ordered nanostructured oxide/polymer hybrid device is only 0.4% efficient The photovoltaic effect in organic solar cells arises, not from formation of free charge

carriers in one or both phases, as in a classical silicon p–n junction cell, but from exciton

dissociation at the junction between the two phases Photon absorption in organic conductors, whether small molecules or polymers, does not generate free holes and

Figure 1.6 Exciton formation and charge separation in a bulk heterojunction, for the case that the donor

phase absorbs the incident light The energies of the HOMO (highest occupied molecular orbital) and LUMO (lowest unoccupied molecular orbital) of the donor (D) and acceptor (A) phases are shown The exciton

binding energy ∆Eb is (∆Ee + ∆Eh)

electrons, but rather hole–electron pairs known as excitons, normally in the singlet state These are bound together by Coulomb attraction and in some cases by local lattice relaxation, with typical binding energies of ~0.5 eV Such excitons are short-lived and have low mobility, with typical diffusion lengths of <10 nm before they recombine However, if such excitons are formed sufficiently close to an interface with appropriate band offsets, they will reach it by diffusion and dissociate into separate charges, as shown

in Fig 1.6 for the case that light is absorbed in the donor phase The exciton, which has

binding energy ∆Eb = ∆Ee + ∆Eh, will dissociate at the heterojunction, injecting an

electron into the acceptor phase, provided that ∆ELUMO > ∆Eb, where ∆ELUMO is the LUMO band offset between the two phases Similarly, an exciton formed in the acceptor

phase will dissociate spontaneously at the heterojunction provided that ∆EHOMO > ∆Eb,

where ∆EHOMO is the HOMO band offset In either case, electrons then diffuse through the acceptor phase to the electron-collecting electrode, and holes through the donor phase

to the hole-collecting electrode The diffusion pathway is straightforward in the bilayer cell, but requires the donor and acceptor phases to form continuous interpenetrating networks in the BHJ

The moving charge usually takes the form of a polaron, which is a charge surrounded

by a locally polarised lattice Polarisation slows the moving charges; fullerene derivatives probably owe their superior performance to the high electron affinity and small lattice distortion (and consequent relatively high mobility) of the fullerene radical anion C60•– Polymer donor phases usually form hole polarons that are delocalised over a segment of the polymer backbone, and hence are also relatively mobile

The electron-collecting electrode should form an ohmic contact to the acceptor LUMO and hence have a relatively low work function, while the hole-collecting electrode should have a relatively high work function Apart from this, the performance

of a bilayer cell is not very sensitive to the nature of the electrodes, since field-assisted migration is not a significant mode of carrier transport The performance of a randomly-blended BHJ cell, on the other hand, is critically dependent on the electrodes being sufficiently different to form a conducting (ohmic) contact to the ‘right’ phase and a blocking contact to the ‘wrong’ phase If this is not the case, the internal hole and electron photocurrents in a random BHJ will be shorted at the electrode(s) An alternative stratagem is to grade the vertical composition of the cell from a 0:100 blend of the two components at one electrode to 100:0 at the other

Organic solar cells show considerable promise They have the advantages over inorganic solar cells of being mechanically flexible, lightweight, and disposable with little environmental impact The constituents can be made soluble so they can be made by low-cost, low-temperature solution-processing methods that should be easily scalable up The rapid advance in the technology of OLEDs (organic light-emitting diodes) is helping

in that it is fertilising research on organic polymer cells, for example through the synthesis of organic semiconductors with improved properties and electrodes with the right electronic properties for selective electron transfer However, OPV have some way

to go before they can become a commercial reality Their efficiency must be improved by extending the absorption edge towards the red or beyond, reducing the energy loss in the charge-separation step by better band-edge alignment, improving charge-carrier mobilities and (in the BHJ) reducing the dead ends and isolated domains that trap charge carriers Device stability also requires attention

1.5 Dye-sensitised solar cells (Grätzel cells)

The architecture of the original dye-sensitised solar cell (DSSC) reported by O’Regan and Grätzel (1991) is shown in Fig 1.7a (see also Fig 8.1) This consisted of two conducting glass electrodes, one coated with a compact but highly porous film of TiO2 on which was adsorbed a ruthenium polypyridyl dye, and the other platinised, separated by a solution containing a high concentration of the iodide/triiodide redox couple in an organic solvent These mesoscopic solar cells—to give them the name preferred by their inventors—or Grätzel cells—to give them the name quickly conferred on them by others—had a reported efficiency of 7.1% in simulated sunlight This provoked some raised eyebrows, because the cell architecture broke many of the existing ‘rules’ about how to make an efficient cell—the dye layer was only a monolayer thick, so each layer absorbed only very little light, the interface was hugely extended, leading to abundant opportunities for back reaction, and the nanocrystalline TiO2 semiconductor was full of electron traps However, fifteen years later, the mode of action of the cell is well understood, laboratory DSSCs have reached over 11% efficiency, and commercial cells are becoming available for portable power applications

Grätzel and Durrant discuss the DSSC fully in Chapter 8, and Mori and Yanigada (2006) have provided a recent review In brief, these cells owe their good performance to

a happy combination of factors The TiO2 layer in the DSSC is about 10 µm thick and has

a roughness factor of over 1000, providing a very large surface area for dye adsorption Incident light is efficiently scattered into the TiO2 layer and absorbed by the many successive dye monolayers through which the light passes The photoactive state of the

Ru dye (a long-lived triplet) injects an electron into the TiO2 conduction band in an ultrafast and efficient process, shown in Fig 1.7b and in more detail in Fig 8.4, creating

an open-circuit photopotential of ~0.8V under standard operating conditions The cell is a majority-carrier device, so the injected electrons can diffuse through the TiO2 to the collecting anode with no danger from bulk recombination—there is little or no interfacial

electric field because of the screening effect of ions in the solution in the electrode pores—and the quantum efficiency of charge collection can approach unity in optimal conditions The oxidised dye is quickly restored to its ground state by electron transfer from iodide, and the electric circuit is completed by the reduction of iodine, mainly complexed as triiodide, to iodide at the counter electrode Crucially, the unwanted interfacial charge-recombination reaction

2e–(TiO2) + I2 (solu.) → 2I– (1.3)

is slow compared with electron transport across the TiO2 layer primarily because it is a two-step process involving high-energy I–-like intermediates This reaction is, however, fast at the counter electrode—as it needs to be to avoid significant overpotential loss—because platinum catalyses the reaction by adsorption of the intermediates In some views, the dye layer also contributes to the slowness of the back reaction at the photoelectrode by physically blocking access of triiodide to the TiO2

The various loss mechanisms in the DSSC, and what is being done to reduce their impact, are discussed in Chapter 8 The first is poor light absorption near the band edge

of the dyes commonly used in the cell, which limits device short-circuit current; the

Figure 1.7 Dye-sensitised solar cell (a) cell architecture; (b) electronic energy levels The placement of the

semiconductor band-edge energy E c and the solution Fermi levels S/S+, S*/S+ and I – /I3– on the same scale, the vacuum scale of electronic energy, is explained in Appendix 1A at the end of this chapter

iodide/triiodide electroyte

porous coated TiO 2

dye-–

+

I– / I 3

S / S+S* / S +

dye

TiO 2 particle dye molecules iodide/triiodide electroyte

second is further minimisation of interfacial recombination losses, which, whilst already being relative minor at short circuit, accelerate as the device operation moves towards open circuit and are the primary limitation of device voltage output In practical terms, the ‘classical’ Grätzel cell also suffers from the corrosive nature of iodine and the possibility of leakage of the liquid electrolyte The use of gel electrolytes and ionic liquids has gone some way to solving these problems More radically, the ionic electrolyte has been replaced by electronic hole conductors in a later generation of all-solid-state cells that have now reached efficiencies of over 5% (see Fig 1.2), and the semiconductor (TiO2 or ZnO) has been engineered into nanorods to shorten the pathway

for electron collection (Baxter and Aydil, 2006; Pan et al., 2007)

In commercial terms, the main advantage that DSSCs have over silicon p–n junction

cells is that their cost per watt could be four to five times lower, because of their cost materials and construction techniques Moreover, the efficiency of DSSCs falls off less in low-intensity light or with increasing temperature than that of Si cells, and they

lower-have lower embodied energy (i.e less energy is required to make them) DSSCs can also

be fabricated as translucent panels or laid down as flexible films on non-planar surfaces There is considerable commercial interest in DSSCs They could make striking additions to buildings as power-generating coloured ‘glass’ panels, and a demonstration

of a building-integrated application by the Japanese company Aisin Seiki is described in Chapter 8 In autumn 2006, Konarka Technologies Inc (Lowell, Mass., USA), which has produced some cells for testing by the US military, joined with the Ecole Polytechnique Fédérale de Lausanne, where Michael Grätzel and co-workers developed the original DSSC, to license a 30-MW European facility for the production of flexible, foil-backed DSSCs to G24 Innovations Ltd (www.g24i.com), a new company based in Cardiff, Wales, which has recently started to ship mobile phone charger units Also in autumn

2006, the Australian company Dyesol (www.dyesol.com) acquired the Lausanne-based Greatcell Solar SA and is working towards the industrialization of DSSC The Rhode Island-based Solaris Nanosciences Corporation (www.solarisnano.com) has demon-

strated in-situ replacement of degraded dye in the DSSC, and Hydrogen Solar

(www.hydrogensolar.com), a private company based in Surrey, UK, has developed a

‘Tandem Cell’ in which a DSSC is combined with an Fe2O3-based photoelectrolysis to achieve water splitting

1.6 Regenerative solar cells

Regenerative solar cells (RSCs), also known as semiconductor/liquid chemical cells, electrochemical photovoltaic solar cells, wet photovoltaic cells or liquid-

photoelectro-junction solar cells, are single-compartment electrochemical batteries that produce electrical power on illumination without any net chemical change occurring in the cell

They normally comprise one semiconductor photoelectrode, which may be n- or p-type,

an electrolyte phase containing a redox couple O,R, and a non-light-sensitive counter electrode that is reversible to the same redox couple The source of the cell emf is the photopotential of the illuminated semiconductor electrode Illumination of this electrode drives the redox reaction O + e–
R in the non-spontaneous direction at an electrode potential at which it cannot occur in the dark The reaction is reversed at the counter electrode The cell therefore produces electric power without net chemical change, unlike

a conventional battery in which the reagents are depleted as current is drawn

It is possible to construct a regenerative cell with two photoelectrodes and a suitable redox couple, and hence to increase the output voltage The photoelectrode of an RSC may be dye-sensitised; the Grätzel cell discussed in the preceding section and in Chapter

8 is the prime example of such a cell However, the principles of RSC operation are essentially the same whether the photoelectrode is sensitised or not

In a practical device, the cell geometry must allow for the illumination of the semiconductor/solution interface, and the ohmic resistance of the cell should be minimised to avoid internal power loss This is best achieved in a thin-layer device in which the semiconductor electrode faces the Sun and is illuminated through the electrolyte and the counter electrode The counter electrode should therefore either be translucent or coarsely gridded

A number of RSCs of good efficiency were developed in the late 1970s and the 1980s In Chapter 9, Nate Lewis and his fellow researchers discuss the types of photo-electrode used in RSCs—mainly crystalline or amorphous silicon, III–V semiconductors, cadmium chalcogenides or lamellar or ternary chalcogenides—and the stratagems that have been adopted to stabilise their performance; these include judicious choice of the redox couple and solvent to avoid corrosion, chemical modification or metallisation of the photoelectrode surface and passivation of surface traps and defects Table 9.1, at the end of Chapter 9, gives an extensive compendium of RSC characteristics and perform-ance; Table 1.1 shows selected cells of high efficiency from this table

Figure 1.8 shows cell schematics for two generic non-dye-sensitised regenerative

solar cells (RSCs) based on either an n-type or a p-type semiconductor electrode Figure1.8a shows the cell n-SC | O,R | CE, in which the electrode n-SC is an n-type

semiconductor photoanode, CE is the counter electrode and O,R is the redox couple R is

oxidised to O at the illuminated n-SC, while O is reduced to R at CE Figure1.8b shows the analogous RSC based on the p-type semiconductor photocathode p-SC In both cases, the photoelectrode operates in the potential range between its flatband potential Ufb and the standard potential UO, Ro of the redox couple, as shown in the bottom diagram